university of groningen health economics of screening for ......paranimfs pham thu hiendidik...

University of Groningen

Health economics of screening for hypertension in VietnamNguyen, Thi Phuong Lan

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2016

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Nguyen, T. P. L. (2016). Health economics of screening for hypertension in Vietnam. University ofGroningen.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 10-11-2020

https://www.rug.nl/research/portal/en/publications/health-economics-of-screening-for-hypertension-in-vietnam(a17b9852-6eed-4b9d-b7ff-4e71932b04cd).html

https://www.rug.nl/research/portal/en/publications/health-economics-of-screening-for-hypertension-in-vietnam(a17b9852-6eed-4b9d-b7ff-4e71932b04cd).html

Health Economics of Screening for Hypertensionin Vietnam

Nguyen Thi Phuong Lan

The research for this thesis received financial support from two projects:“Strengthening Teaching and Research Capacity of Preventive Medicine in Vietnam” supported by the NUFFIC (Netherlands) and “Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Vietnam”, supported by the Royal Netherlands Embassy in Hanoi.

______________________

Nguyen Thi Phuong Lan

Health Economics of Screening for Hypertension in Vietnam

Thesis University of GroningenISBN: 978-90-367-9002-4

Printed by Proefschriftmaken.nlPhotos of Thien Cung cave at Ha Long bay and Tulip flowers taken by Maarten J. Postma and Tran Van Bay.Cover design by Nguyen Dinh Vu

Printing of this thesis was financial supported by the Graduate School for Health Research SHARE, the Groningen Graduate School of Science (GGSS) and the University of Groningen (RUG).

© 2016 Nguyen Thi Phuong Lan. No part of this book may be reproduced or converted into any form or by any means without written permission of the author. The copyright of previously published chapters of this thesis remains in the publisher or journal.

Health Economics of Screeningfor Hypertension in Vietnam

PhD thesis

to obtain the degree of PhDat the University of Groningen

on the authority of the Rector Magnificus Prof. E.Sterken

and in accordance withthe decision by the College of Deans.

This thesis will be defended in public on

Friday 23 September 2016 at 09.00hours

by

Thi Phuong Lan Nguyen

born on 14 March 1975in Ninh Binh, Vietnam

Paranimfs Pham Thu Hien

Didik Setiawan

SupervisorProf. M.J. Postma

Co-supervisorsDr. C.C.M Schuilinga-Veninga Dr. Nguyen Thi Bach YenDr. E.P. Wright

Assessment CommitteeProf. K. TaxisProf. H.V. Hogerzeil Prof. L. Annemans

To my colleagues, family and friends

To my colleagues, family and friends

Table of contents

Chapter 1 Introduction 1

Chapter 2 Models to predict the burden of 19 cardiovascular disease risk in a rural mountainous region of Vietnam

Chapter 3 Adherence to hypertension 51 medication: quantitative and qualitative investigations in a rural Northern Vietnamese community

Chapter 4 Direct costs of hypertensive 81 patients admitted to hospital in Vietnam – A bottom-up microcosting analysis

Chapter 5 Utilities of patients with 105 hypertension in Northern Vietnam

Chapter 6 Cost-effectiveness analysis of 123 screening for and managing identifiedhypertensionfor cardiovascular disease prevention in Vietnam

Chapter 7 General discussion 161

Appendix 179

Summary 196

List of publications 203

Acknowledgements 206

Research Institute SHARE 209

Curriculum vitae 213

Chapter 1

Introduction

2

Introduction

Non-communicable diseases have become a main contributor to the global burden of disease, accounting for 58% of total disability adjusted life years (DALYs) in 2013. Notably, cardiovascular and cerebrovascular diseases accounted for 13.5% of total DALYs [1]worldwide. The trend is similar in developing countries, where these conditions contributed 12.3 % to total DALYs [1].

Using DALYs to estimate the burden of disease in Vietnam, non-communicable diseases accounted for 66%, injuries for 16% and infections and other diseases for 18% of the total in 2013 [1].Regarding the economic consequences, an estimated 20 million US$, equal to 0.033% of annual national GDP, was lost due to non-communicable diseases in 2005, which was expected to double by 2015 if no interventions were put in place [2].

The priorities for investment in the health sector are based on changes in the patterns of disease, illustrated in Fig 1 [1]. In Vietnam, the health sector invested in communicable diseases, in previous years, and succeeded in controlling them to a certain extent. In the coming years, solutions to prevent and manage non-communicable disease need investments to be strengthened [2-4].

Chapter 1

Among the non-communicable diseases, cardiovascular disease (CVD) dominates. For example, looking at the burden of disease due to non-communicable diseases, the biggest condition contributing to loss of DALYs is CVD, at around 23% [1]. Therefore, this thesis will focus on the prevention of CVD as an example of an important non-communicable disease burden for Vietnam with potentials for improved prevention and treatment. This thesis concentrates on the integration of results from empirical research on risk factors in the Vietnamese population into relevant models to predict final clinical outcomes and on the cost-effectiveness of different approaches to prevention. The ultimate aim is to provide recommendations on CVD prevention in comparable developing countries based on economicconsiderations.

Health system and health services in Vietnam

Vietnam’s health system is organised from the national to the provincial to the district then to the community (grassroots) level. Primary health care services, which should be the first place to access health care, are provided at grassroots level. In practice, people can access health services at higher levels directly without a primary health-care physician’s referral and can buy medicine without prescriptions in the free market.

In recent years, health services for non-communicable diseases were expanded at community health station and district levels [5]. This is seen as a major improvement. However, challenging issues in the characteristics of the current health care system need to be considered.Firstly, multiple entrances for health services lead to difficulties in following up patients and in estimating the total burden of diseases. Secondly, the absence of health information systems and lack of integration among different health services, especially for patients

Chap

ter 1

3

Introduction

Non-communicable diseases have become a main contributor to the global burden of disease, accounting for 58% of total disability adjusted life years (DALYs) in 2013. Notably, cardiovascular and cerebrovascular diseases accounted for 13.5% of total DALYs [1]worldwide. The trend is similar in developing countries, where these conditions contributed 12.3 % to total DALYs [1].

Using DALYs to estimate the burden of disease in Vietnam, non-communicable diseases accounted for 66%, injuries for 16% and infections and other diseases for 18% of the total in 2013 [1].Regarding the economic consequences, an estimated 20 million US$, equal to 0.033% of annual national GDP, was lost due to non-communicable diseases in 2005, which was expected to double by 2015 if no interventions were put in place [2].

The priorities for investment in the health sector are based on changes in the patterns of disease, illustrated in Fig 1 [1]. In Vietnam, the health sector invested in communicable diseases, in previous years, and succeeded in controlling them to a certain extent. In the coming years, solutions to prevent and manage non-communicable disease need investments to be strengthened [2-4].

Chapter 1

Among the non-communicable diseases, cardiovascular disease (CVD) dominates. For example, looking at the burden of disease due to non-communicable diseases, the biggest condition contributing to loss of DALYs is CVD, at around 23% [1]. Therefore, this thesis will focus on the prevention of CVD as an example of an important non-communicable disease burden for Vietnam with potentials for improved prevention and treatment. This thesis concentrates on the integration of results from empirical research on risk factors in the Vietnamese population into relevant models to predict final clinical outcomes and on the cost-effectiveness of different approaches to prevention. The ultimate aim is to provide recommendations on CVD prevention in comparable developing countries based on economicconsiderations.

Health system and health services in Vietnam

Vietnam’s health system is organised from the national to the provincial to the district then to the community (grassroots) level. Primary health care services, which should be the first place to access health care, are provided at grassroots level. In practice, people can access health services at higher levels directly without a primary health-care physician’s referral and can buy medicine without prescriptions in the free market.

In recent years, health services for non-communicable diseases were expanded at community health station and district levels [5]. This is seen as a major improvement. However, challenging issues in the characteristics of the current health care system need to be considered.Firstly, multiple entrances for health services lead to difficulties in following up patients and in estimating the total burden of diseases. Secondly, the absence of health information systems and lack of integration among different health services, especially for patients

4

Introduction

with chronic diseases, lead to inconvenient or ineffective treatment and unnecessary lifelong health care. Thirdly, lack of capacity and of budgets for continuous and lifelong management of chronic diseases at community health stations lead to under-detection, under-diagnosis and under-treatment of disease [5, 6]. This thesis aims to investigate solutions for better CVD prevention and treatment in Vietnam.

In the next sections of this chapter, the methods and solutions used in the studies described in this thesis will be discussed. In the final section, the main contents of each chapter are introduced.

Approaches to CVD – feasibility & potential application in Vietnam

A number of lessons and recommendations on early detection and management of CVD are available from different countries around the world, some of which may have potential for application in Vietnam. Strategic priorities on prevention and management of CVD are often guided by the World Health Organization (WHO). The WHO recommends three strategies: (1) reducing CVD risk factors effectively; (2) developing solutions to manage CVD with consideration for cost effectiveness and equity; and (3) monitoring CVD and its risk factors [7]. These strategies focus on both primary and secondary prevention.

Primary CVD prevention strategies have been targeted to people with CVD risk factors such as behavioural or metabolic factors. Population-based interventions often focus on behavioural risk factors such as smoking, excessive alcohol intake, unhealthy diet, physical inactivity or having stress. Individual-based interventions commonly comprise screening and treatment focussing on metabolic CVD risk factors such as hypertension, diabetes, obesity and dyslipidaemia.

Chapter 1

Especially hypertension is known to be an important factor contributing to the burden of CVD [8]. The prevalence of hypertension has been increasing in recent years [8, 9]. Unfortunately, the percentage of undiagnosed people with hypertension is estimated to be quite high, both globally and in Vietnam: 53% globally [10] and approximately 52% in Vietnam [9]. This high number of undiagnosed hypertensive persons who are not yet on appropriate treatment is one key obstacle to reducing the burden of this disease. Screening for hypertension is therefore an important first step to increase both awareness and treatment of hypertension [11], which would finally reduce the burden of CVD [12].

Several studies on screening for hypertension have been conducted. In Canada, it is recommended to check blood pressure of all adults during regular visits to physicians [13] but there is no evidenceconcerning those who seldom access health care. The optimal frequency of screening is still unknown and no evidence is available about the best screening interval [13]. In the USA, screening is recommended for all adults every one or two years [14]. There, it was found that annual screening was more specific (with an overall 11% improvement) in comparison with routine screening at every visit [15]. In the UK, a population-based screening programme for hypertension was not recommended [16]. In The Netherlands, the recommendation is to screen for hypertension based on age, history of CVD and other CVD risk factors [17]. In Japan, a hospital-based study in a population not taking antihypertensive medicine at baseline and older than 20 years suggested that the recommended screening interval for hypertension, either every two or three years, depends on the systolic blood pressure level [18].

Up to now, there is insufficient evidence regarding the impact on mortality and morbidity of population-wide screening for

Chap

ter 1

5

Introduction

with chronic diseases, lead to inconvenient or ineffective treatment and unnecessary lifelong health care. Thirdly, lack of capacity and of budgets for continuous and lifelong management of chronic diseases at community health stations lead to under-detection, under-diagnosis and under-treatment of disease [5, 6]. This thesis aims to investigate solutions for better CVD prevention and treatment in Vietnam.

In the next sections of this chapter, the methods and solutions used in the studies described in this thesis will be discussed. In the final section, the main contents of each chapter are introduced.

Approaches to CVD – feasibility & potential application in Vietnam

A number of lessons and recommendations on early detection and management of CVD are available from different countries around the world, some of which may have potential for application in Vietnam. Strategic priorities on prevention and management of CVD are often guided by the World Health Organization (WHO). The WHO recommends three strategies: (1) reducing CVD risk factors effectively; (2) developing solutions to manage CVD with consideration for cost effectiveness and equity; and (3) monitoring CVD and its risk factors [7]. These strategies focus on both primary and secondary prevention.

Primary CVD prevention strategies have been targeted to people with CVD risk factors such as behavioural or metabolic factors. Population-based interventions often focus on behavioural risk factors such as smoking, excessive alcohol intake, unhealthy diet, physical inactivity or having stress. Individual-based interventions commonly comprise screening and treatment focussing on metabolic CVD risk factors such as hypertension, diabetes, obesity and dyslipidaemia.

Chapter 1

Especially hypertension is known to be an important factor contributing to the burden of CVD [8]. The prevalence of hypertension has been increasing in recent years [8, 9]. Unfortunately, the percentage of undiagnosed people with hypertension is estimated to be quite high, both globally and in Vietnam: 53% globally [10] and approximately 52% in Vietnam [9]. This high number of undiagnosed hypertensive persons who are not yet on appropriate treatment is one key obstacle to reducing the burden of this disease. Screening for hypertension is therefore an important first step to increase both awareness and treatment of hypertension [11], which would finally reduce the burden of CVD [12].

Several studies on screening for hypertension have been conducted. In Canada, it is recommended to check blood pressure of all adults during regular visits to physicians [13] but there is no evidenceconcerning those who seldom access health care. The optimal frequency of screening is still unknown and no evidence is available about the best screening interval [13]. In the USA, screening is recommended for all adults every one or two years [14]. There, it was found that annual screening was more specific (with an overall 11% improvement) in comparison with routine screening at every visit [15]. In the UK, a population-based screening programme for hypertension was not recommended [16]. In The Netherlands, the recommendation is to screen for hypertension based on age, history of CVD and other CVD risk factors [17]. In Japan, a hospital-based study in a population not taking antihypertensive medicine at baseline and older than 20 years suggested that the recommended screening interval for hypertension, either every two or three years, depends on the systolic blood pressure level [18].

Up to now, there is insufficient evidence regarding the impact on mortality and morbidity of population-wide screening for

6

Introduction

hypertension in developing countries [19]. Furthermore, the optimal strategy for screening under given conditions is still unclear. For example, on which part of the population should the screening focus: population-wide or only those people with high risk? Which screening interval should be applied?

After detection of hypertension, appropriate treatment plays an important role in controlling blood pressure and reducing the risk of CVD [20, 21]. This comprises both prevalence of treatment and adherence. Although evidence exists that long-term adherence to medicine among hypertensive patients reduces CVD events [22, 23],adherence to medicine can be very low. For example, one meta-analysis covering 376,162 patients using medicine to prevent CVD found the percentage of adherent patients to be only 57% [24].Information about adherence, factors affecting adherence, and the effect on outcomes when adherence improves is not available for the Vietnamese population.

This thesis aims to investigate solutions for better CVD prevention and treatment in Vietnam, including adherence issues. The research questions and sub-research questions are listed below.

Research Questions

The main research question reads “What are cost-effective ways of screening for and managing hypertension to prevent cardiovascular diseases in Vietnam?”

Sub-questions were:

1. Which model is most appropriate to measure CVD risk in Vietnam?

Chapter 1

2. How good is adherence to hypertension treatment and which factors affect adherence in the study population?

3. How will adherence to hypertension treatment influence outcomes?

4. What are the future risks of CVD in case of non-treatment or non-adherence?

5. What quality of life is experienced by hypertensive patients in the Vietnamese population?

6. What is the burden of disease if hypertension is not treated or controlled?

7. Which is the most cost-effective population screening strategy to identify patients with hypertension at risk for future complications, in the context of modelling various screening and treatment coverage scenarios?

8. What recommendations can be given for national programs to identify patients with hypertension and to minimize risks of further disease development in those patients?

Methods available from previous studies and applied in this thesis

Predicting outcomes of CVD interventions

To identify the efficacy of an intervention, the preferred approach is to run clinical trials, especially double-blind, randomized, controlled trials (RCT). However, there are obstacles to conducting RCTs, such as, for example, the large scale and long duration, resulting in high costs. Additionally, there are ethical issues concerning the inclusion of patients in control groups or exposing them to untested treatments. Finally, RCTs often measure intermediate outcomes rather than clinical endpoints of morbidity and mortality. To help overcome these limitations, several models have been developed to predict CVD events in populations. For example, the Framingham risk score, the

Chap

ter 1

7

Introduction

hypertension in developing countries [19]. Furthermore, the optimal strategy for screening under given conditions is still unclear. For example, on which part of the population should the screening focus: population-wide or only those people with high risk? Which screening interval should be applied?

After detection of hypertension, appropriate treatment plays an important role in controlling blood pressure and reducing the risk of CVD [20, 21]. This comprises both prevalence of treatment and adherence. Although evidence exists that long-term adherence to medicine among hypertensive patients reduces CVD events [22, 23],adherence to medicine can be very low. For example, one meta-analysis covering 376,162 patients using medicine to prevent CVD found the percentage of adherent patients to be only 57% [24].Information about adherence, factors affecting adherence, and the effect on outcomes when adherence improves is not available for the Vietnamese population.

This thesis aims to investigate solutions for better CVD prevention and treatment in Vietnam, including adherence issues. The research questions and sub-research questions are listed below.

Research Questions

The main research question reads “What are cost-effective ways of screening for and managing hypertension to prevent cardiovascular diseases in Vietnam?”

Sub-questions were:

1. Which model is most appropriate to measure CVD risk in Vietnam?

Chapter 1


3. How will adherence to hypertension treatment influence outcomes?


5. What quality of life is experienced by hypertensive patients in the Vietnamese population?

6. What is the burden of disease if hypertension is not treated or controlled?

7. Which is the most cost-effective population screening strategy to identify patients with hypertension at risk for future complications, in the context of modelling various screening and treatment coverage scenarios?

8. What recommendations can be given for national programs to identify patients with hypertension and to minimize risks of further disease development in those patients?

Methods available from previous studies and applied in this thesis

Predicting outcomes of CVD interventions

To identify the efficacy of an intervention, the preferred approach is to run clinical trials, especially double-blind, randomized, controlled trials (RCT). However, there are obstacles to conducting RCTs, such as, for example, the large scale and long duration, resulting in high costs. Additionally, there are ethical issues concerning the inclusion of patients in control groups or exposing them to untested treatments. Finally, RCTs often measure intermediate outcomes rather than clinical endpoints of morbidity and mortality. To help overcome these limitations, several models have been developed to predict CVD events in populations. For example, the Framingham risk score, the

8

Introduction

“assessing cardiovascular risk to Scottish Intercollegiate Guidelines Network to assign preventative treatment” (ASSIGN) score, systematic coronary risk evaluation (SCORE) score, Prospective Cardiovascular Münster (PROCAM) score, QRESEARCH cardiovascular risk algorithms, and the World Health Organization/International Society of Hypertension (WHO/ISH) models are well-known and frequently used in research. In Asian countries, the Asian and Chinese Multiple-provincial Cohort Study (CMCS) models have been developed in recent years. The main differences among these models are the parameters used, which may provide different outcomes [25, 26]. It is crucial to choose the most appropriate model when one is needed for application to a population different from the original population for which it was developed. A comparison of different models to adequately predict the outcome of CVD interventions in Vietnam will be presented and discussed in Chapter 2.

Adherence measurement

Prediction of the outcomes of different treatment assumes an adequateadherence to the treatment. As noted above, good adherence seems to be crucial for optimal effectiveness of hypertensive treatment to prevent CVD. Understanding the factors that influence adherence can help to improve the effectiveness of treatment. Measurement of adherence and its effects on clinical outcomes can help to quantify the real effect of blood pressure treatment interventions. Therefore, the following issues are addressed in the next section of this thesis: (1) how to measure adherence; (2) how to identify factors affecting adherence; (3) which recommendations could be made to improve adherence to hypertensive treatment; and (4) what is known about the association between adherence and clinical outcomes.

Chapter 1

A range of methods, questionnaires, tools and scales have been introduced to measure adherence among patients with chronic diseases[27]. Besides quantifying the level of adherence, also several reasons for non-adherence have been revealed. The different tools can be divided into five groups, according to aspects of the adherence they focus on: (1) medication-taking behaviours; (2) medication-taking behaviours and barriers to adherence; (3) barriers to adherence; (4) beliefs associated with adherence; and (5) barriers and beliefs associated with adherence. Each method has its own advantages and disadvantages in estimating adherence depending on the simplicity of collecting data, the reliability of the measurements, over- or underestimation of the outcome, and the costs related to the use of the method. At present, none of those methods is considered the best way to measure adherence. They often can be considered complimentary rather than competitive and mutually exclusive.

Various factors are known to affect adherence. They can be classified into seven groups: demographic factors, social factors, cognitive factors, interactions between health care providers and patients, health care system characteristics, the medication involved, and the general health profiles of the patients [28].

Low adherence is more likely to result in uncontrolled high blood pressure (HBP) compared to high compliance with medication [29]and is related to hospitalisation [30] or associated with CVD events [23]. A link between a specific level of adherence and occurrence of CVD has been demonstrated [23].

Economic evaluation

Economic evaluation can be displayed in two forms: either partial or full economic evaluations are considered. Partial evaluation may include outcome description, efficacy/effectiveness evaluation, cost description, and cost analysis. Full economic evaluation includes cost-

Chap

ter 1

9

Introduction

“assessing cardiovascular risk to Scottish Intercollegiate Guidelines Network to assign preventative treatment” (ASSIGN) score, systematic coronary risk evaluation (SCORE) score, Prospective Cardiovascular Münster (PROCAM) score, QRESEARCH cardiovascular risk algorithms, and the World Health Organization/International Society of Hypertension (WHO/ISH) models are well-known and frequently used in research. In Asian countries, the Asian and Chinese Multiple-provincial Cohort Study (CMCS) models have been developed in recent years. The main differences among these models are the parameters used, which may provide different outcomes [25, 26]. It is crucial to choose the most appropriate model when one is needed for application to a population different from the original population for which it was developed. A comparison of different models to adequately predict the outcome of CVD interventions in Vietnam will be presented and discussed in Chapter 2.

Adherence measurement

Prediction of the outcomes of different treatment assumes an adequateadherence to the treatment. As noted above, good adherence seems to be crucial for optimal effectiveness of hypertensive treatment to prevent CVD. Understanding the factors that influence adherence can help to improve the effectiveness of treatment. Measurement of adherence and its effects on clinical outcomes can help to quantify the real effect of blood pressure treatment interventions. Therefore, the following issues are addressed in the next section of this thesis: (1) how to measure adherence; (2) how to identify factors affecting adherence; (3) which recommendations could be made to improve adherence to hypertensive treatment; and (4) what is known about the association between adherence and clinical outcomes.

Chapter 1

A range of methods, questionnaires, tools and scales have been introduced to measure adherence among patients with chronic diseases[27]. Besides quantifying the level of adherence, also several reasons for non-adherence have been revealed. The different tools can be divided into five groups, according to aspects of the adherence they focus on: (1) medication-taking behaviours; (2) medication-taking behaviours and barriers to adherence; (3) barriers to adherence; (4) beliefs associated with adherence; and (5) barriers and beliefs associated with adherence. Each method has its own advantages and disadvantages in estimating adherence depending on the simplicity of collecting data, the reliability of the measurements, over- or underestimation of the outcome, and the costs related to the use of the method. At present, none of those methods is considered the best way to measure adherence. They often can be considered complimentary rather than competitive and mutually exclusive.

Various factors are known to affect adherence. They can be classified into seven groups: demographic factors, social factors, cognitive factors, interactions between health care providers and patients, health care system characteristics, the medication involved, and the general health profiles of the patients [28].

Low adherence is more likely to result in uncontrolled high blood pressure (HBP) compared to high compliance with medication [29]and is related to hospitalisation [30] or associated with CVD events [23]. A link between a specific level of adherence and occurrence of CVD has been demonstrated [23].

Economic evaluation

Economic evaluation can be displayed in two forms: either partial or full economic evaluations are considered. Partial evaluation may include outcome description, efficacy/effectiveness evaluation, cost description, and cost analysis. Full economic evaluation includes cost-

10

Introduction

effectiveness analysis, cost-utility analysis, and cost-benefit analysis. Cost-effectiveness analysis and cost utility analysis are applied for an assessment of whether a program is worth doing, using references, for example, the threshold for cost-effectiveness ratio, a decision on a limited budget, or how best to allocate an existing budget. Cost-utility analysis is a specific type of cost-effectiveness analysis which uses QALYs as outcome, while cost-effectiveness analysis considers clinical effectiveness as endpoint, which may vary, for example, cost per case detected, cost per mmHg reduction in hypertensive patients, cost per episode-free day, or cost per life-year gained.

Economic evaluation works in several perspectives, which depends on the data available and the preference of policy makers; it may be considered from the perspective of the health care system, the patients or society. In our context, modelling of cost-effectiveness was considered most feasible and appropriate to be performed, talking the health care provider perspective, because data limitations hampered talking a broader perspective.

Health utilities

The concept of quality-adjusted life years (QALYs), which accounts for both quantity and quality of life, has been used widely in cost-effectiveness analysis. In recent years, several tools and algorithms have been developed to measure health-related quality of life expressed in utilities. The most common tools are the Quality of Well-Being Scale, the Health Utilities Index, the EuroQoL EQ-5D and the Short-Form 36. Several studies in different populations have shown that different tools provide different values [31-33]. So far, there is no recommendation on the best instrument to measure health utilities in a given situation.

Chapter 1

Main contents of thesis and each chapter

In this thesis, several of the methods described above will be applied to answer the research questions, generating information for clinical recommendations and policy documents.

In Chapter 2, an empirical research was conducted to measure the prevalence of risk factors for CVD in Vietnam. Data from this survey was applied to and compared using three prediction models with the aim of identifying the most appropriate model to predict CVD in a rural area in Northern Vietnam. In Chapter 3, we used a mixed-methods study with qualitative and quantitative data collection to determine levels of adherence to treatment and factors influencing adherence among hypertensive patients. It is notable that methods for quantifying levels of adherence used in many countries are based on prescriptions, while we found the issue to be more complex in the Vietnamese context, where patients may buy medicines without a prescription.

In Chapter 4, we conduct a partial economic evaluation to estimate the economic burden of CVD for the health care system. In this study, we focused on hypertensive patients who require hospitalization.

Chapter 5 describes the use of SF-36 to provide evidence on health utility and its predictors, to quantify the quality of life. SF-36 was chosen because it has been used widely in other countries among hypertensive patients. Health utilities were derived from the SF-36 by the algorithm produced by Brazier et al [34].

In Chapter 6, a decision tree was combined with a Markov model to compare the cost-effectiveness of different strategies for screening and treatment for hypertension. The model was designed based on findings from Chapters 2, 3, 4 and 5 along with other evidence available from studies on the Vietnamese population and meta-analysis studies.

Chap

ter 1

11

Introduction

effectiveness analysis, cost-utility analysis, and cost-benefit analysis. Cost-effectiveness analysis and cost utility analysis are applied for an assessment of whether a program is worth doing, using references, for example, the threshold for cost-effectiveness ratio, a decision on a limited budget, or how best to allocate an existing budget. Cost-utility analysis is a specific type of cost-effectiveness analysis which uses QALYs as outcome, while cost-effectiveness analysis considers clinical effectiveness as endpoint, which may vary, for example, cost per case detected, cost per mmHg reduction in hypertensive patients, cost per episode-free day, or cost per life-year gained.

Economic evaluation works in several perspectives, which depends on the data available and the preference of policy makers; it may be considered from the perspective of the health care system, the patients or society. In our context, modelling of cost-effectiveness was considered most feasible and appropriate to be performed, talking the health care provider perspective, because data limitations hampered talking a broader perspective.

Health utilities

The concept of quality-adjusted life years (QALYs), which accounts for both quantity and quality of life, has been used widely in cost-effectiveness analysis. In recent years, several tools and algorithms have been developed to measure health-related quality of life expressed in utilities. The most common tools are the Quality of Well-Being Scale, the Health Utilities Index, the EuroQoL EQ-5D and the Short-Form 36. Several studies in different populations have shown that different tools provide different values [31-33]. So far, there is no recommendation on the best instrument to measure health utilities in a given situation.

Chapter 1

Main contents of thesis and each chapter

In this thesis, several of the methods described above will be applied to answer the research questions, generating information for clinical recommendations and policy documents.

In Chapter 2, an empirical research was conducted to measure the prevalence of risk factors for CVD in Vietnam. Data from this survey was applied to and compared using three prediction models with the aim of identifying the most appropriate model to predict CVD in a rural area in Northern Vietnam. In Chapter 3, we used a mixed-methods study with qualitative and quantitative data collection to determine levels of adherence to treatment and factors influencing adherence among hypertensive patients. It is notable that methods for quantifying levels of adherence used in many countries are based on prescriptions, while we found the issue to be more complex in the Vietnamese context, where patients may buy medicines without a prescription.

In Chapter 4, we conduct a partial economic evaluation to estimate the economic burden of CVD for the health care system. In this study, we focused on hypertensive patients who require hospitalization.

Chapter 5 describes the use of SF-36 to provide evidence on health utility and its predictors, to quantify the quality of life. SF-36 was chosen because it has been used widely in other countries among hypertensive patients. Health utilities were derived from the SF-36 by the algorithm produced by Brazier et al [34].

In Chapter 6, a decision tree was combined with a Markov model to compare the cost-effectiveness of different strategies for screening and treatment for hypertension. The model was designed based on findings from Chapters 2, 3, 4 and 5 along with other evidence available from studies on the Vietnamese population and meta-analysis studies.

12

Introduction

Chapter 7 summarizes and discusses the main findings and provides suggestions for further research.

Chapter 1

References

1. Institute-for-Health-Metrics-and-Evaluation. (2013) Global Burden Disease Data Visualizations. Available from: http://vizhub.healthdata.org/gbd-compare/.

2. Van Minh H, Lan Huong D, Bao Giang K and Byass P. (2009) Economic aspects of chronic diseases in Vietnam. Glob Health Action 2.

3. Ministry-of-Health (2010) Joint Annual Health Review 2011 Strengthening management capacity and reforming health financing to implement the five-year health sector plan 2011 –2015. Ha Noi.

4. Ministry-of-Health. (2015) Join Annual Health Review 2014. Strengthening prevention and control of non-communicable disease.

5. Ministry-of-Health (2015) Join Annual Health Review 2015.Strengthening primary health care at the grassroots towards universal health coverage.

6. Ministry-of-Health. (2013) Join Annual Health Review 2013. Towards universal health coverage. .

7. World-Health-Organization (2007) Prevention of Cardiovascular Disease - Guidelines for assessment and management of cardiovascular risk.

8. Catherine H (2011) Vietnam Noncommunicable Disease Prevention and Control Programme 2002-2010.

9. Son PT, Quang NN, Viet NL, Khai PG, Wall S, et al. (2012) Prevalence, awareness, treatment and control of hypertension in Vietnam-results from a national survey. J Hum Hypertens 26: 268-280.

Chap

ter 1

13

Introduction

Chapter 7 summarizes and discusses the main findings and provides suggestions for further research.

Chapter 1

References



3. Ministry-of-Health (2010) Joint Annual Health Review 2011 Strengthening management capacity and reforming health financing to implement the five-year health sector plan 2011 –2015. Ha Noi.


5. Ministry-of-Health (2015) Join Annual Health Review 2015.Strengthening primary health care at the grassroots towards universal health coverage.

6. Ministry-of-Health. (2013) Join Annual Health Review 2013. Towards universal health coverage. .


8. Catherine H (2011) Vietnam Noncommunicable Disease Prevention and Control Programme 2002-2010.


14

Introduction

10. Chow CK, Teo KK, Rangarajan S, Islam S, Gupta R, et al. (2013) Prevalence, awareness, treatment, and control of hypertension in rural and urban communities in high-, middle-, and low-income countries. JAMA 310: 959-968.

11. Kotwani P, Balzer L, Kwarisiima D, Clark TD, Kabami J, et al. (2014) Evaluating linkage to care for hypertension after community-based screening in rural Uganda. Trop Med IntHealth 19: 459-468.

12. Maurer J and Ramos A. (2015) One-year routine opportunistic screening for hypertension in formal medical settings and potential improvements in hypertension awareness among older persons in developing countries: evidence from the Study on Global Ageing and Adult Health (SAGE). Am J Epidemiol 181: 180-184.

13. Lindsay P, Connor Gorber S, Joffres M, Birtwhistle R, McKay D, et al. (2013) Recommendations on screening for high blood pressure in Canadian adults. Can Fam Physician 59: 927-933, e393-400.

14. (2007) Screening for high blood pressure: U.S. Preventive Services Task Force reaffirmation recommendation statement. 2007/12/07:[783-786].

15. Garrison GM and Oberhelman S. (2013) Screening for hypertension annually compared with current practice. Ann Fam Med 11: 116-121.

16. The-UK-National-Screening-Committee. (2013) The UK NSC recommendation on Hypertension screening in adults. Available from: http://legacy.screening.nhs.uk/hypertension-adult.

Chapter 1

17. van Buuren S, Boshuizen HC and Reijneveld SA. (2006) Toward targeted hypertension screening guidelines. Med Decis Making 26: 145-153.

18. Takahashi O, Glasziou PP, Perera R, Shimbo T and Fukui T. (2012) Blood pressure re-screening for healthy adults: what is the best measure and interval? J Hum Hypertens 26: 540-546.

19. Durao S, Ajumobi O, Kredo T, Naude C, Levitt NS, et al. (2015) Evidence insufficient to confirm the value of population screening for diabetes and hypertension in low- and-middle-income settings. S Afr Med J 105: 98-102.

20. Law MR, Morris JK and Wald NJ. (2009) Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies. BMJ 338: b1665.

21. Yano Y, Briasoulis A, Bakris GL, Hoshide S, Wang JG, et al. (2014) Effects of antihypertensive treatment in Asian populations: a meta-analysis of prospective randomized controlled studies (CARdiovascular protectioN group in Asia: CARNA). J Am Soc Hypertens 8: 103-116.

22. Dragomir A, Cote R, Roy L, Blais L, Lalonde L, et al. (2010) Impact of adherence to antihypertensive agents on clinical outcomes and hospitalization costs. Med Care 48: 418-425.

23. Mazzaglia G, Ambrosioni E, Alacqua M, Filippi A, Sessa E, et al. (2009) Adherence to antihypertensive medications and cardiovascular morbidity among newly diagnosed hypertensive patients. Circulation 120: 1598-1605.

24. Naderi SH, Bestwick JP and Wald DS. (2012) Adherence to drugs that prevent cardiovascular disease: meta-analysis on 376,162 patients. Am J Med 125: 882-887 e881.

Chap

ter 1

15

Introduction

10. Chow CK, Teo KK, Rangarajan S, Islam S, Gupta R, et al. (2013) Prevalence, awareness, treatment, and control of hypertension in rural and urban communities in high-, middle-, and low-income countries. JAMA 310: 959-968.

11. Kotwani P, Balzer L, Kwarisiima D, Clark TD, Kabami J, et al. (2014) Evaluating linkage to care for hypertension after community-based screening in rural Uganda. Trop Med IntHealth 19: 459-468.

12. Maurer J and Ramos A. (2015) One-year routine opportunistic screening for hypertension in formal medical settings and potential improvements in hypertension awareness among older persons in developing countries: evidence from the Study on Global Ageing and Adult Health (SAGE). Am J Epidemiol 181: 180-184.

13. Lindsay P, Connor Gorber S, Joffres M, Birtwhistle R, McKay D, et al. (2013) Recommendations on screening for high blood pressure in Canadian adults. Can Fam Physician 59: 927-933, e393-400.

14. (2007) Screening for high blood pressure: U.S. Preventive Services Task Force reaffirmation recommendation statement. 2007/12/07:[783-786].

15. Garrison GM and Oberhelman S. (2013) Screening for hypertension annually compared with current practice. Ann Fam Med 11: 116-121.

16. The-UK-National-Screening-Committee. (2013) The UK NSC recommendation on Hypertension screening in adults. Available from: http://legacy.screening.nhs.uk/hypertension-adult.

Chapter 1

17. van Buuren S, Boshuizen HC and Reijneveld SA. (2006) Toward targeted hypertension screening guidelines. Med Decis Making 26: 145-153.

18. Takahashi O, Glasziou PP, Perera R, Shimbo T and Fukui T. (2012) Blood pressure re-screening for healthy adults: what is the best measure and interval? J Hum Hypertens 26: 540-546.

19. Durao S, Ajumobi O, Kredo T, Naude C, Levitt NS, et al. (2015) Evidence insufficient to confirm the value of population screening for diabetes and hypertension in low- and-middle-income settings. S Afr Med J 105: 98-102.



22. Dragomir A, Cote R, Roy L, Blais L, Lalonde L, et al. (2010) Impact of adherence to antihypertensive agents on clinical outcomes and hospitalization costs. Med Care 48: 418-425.


24. Naderi SH, Bestwick JP and Wald DS. (2012) Adherence to drugs that prevent cardiovascular disease: meta-analysis on 376,162 patients. Am J Med 125: 882-887 e881.

16

Introduction

25. Siontis GC, Tzoulaki I, Siontis KC and Ioannidis JP. (2012) Comparisons of established risk prediction models for cardiovascular disease: systematic review. BMJ 344: e3318.

26. Selvarajah S, Kaur G, Haniff J, Cheong KC, Hiong TG, et al. (2014) Comparison of the Framingham Risk Score, SCORE and WHO/ISH cardiovascular risk prediction models in an Asian population. Int J Cardiol 176: 211-218.

27. Nguyen TM, La Caze A and Cottrell N. (2014) What are validated self-report adherence scales really measuring?: a systematic review. Br J Clin Pharmacol 77: 427-445.

28. Thi-Phuong-Lan N, Catharina CM S-V, Thi Bach Yen N, Thu-Hang V, E. Pamela W, et al. (2016) Adherence to antihypertensive medication: quantitative and qualitative investigations in a rural Northern Vietnamese community

29. Krousel-Wood MA, Muntner P, Islam T, Morisky DE and Webber LS. (2009) Barriers to and determinants of medication adherence in hypertension management: perspective of the cohort study of medication adherence among older adults. Med Clin North Am 93: 753-769.

30. Addison CC, Jenkins BW, Sarpong D, Wilson G, Champion C, et al. (2011) Relationship between medication use and cardiovascular disease health outcomes in the Jackson Heart Study. Int J Environ Res Public Health 8: 2505-2515.

31. McDonough CM, Grove MR, Tosteson TD, Lurie JD, Hilibrand AS, et al. (2005) Comparison of EQ-5D, HUI, and SF-36-derived societal health state values among spine patient outcomes research trial (SPORT) participants. Qual Life Res 14: 1321-1332.

Chapter 1

32. Brazier J, Roberts J, Tsuchiya A and Busschbach J. (2004) A comparison of the EQ-5D and SF-6D across seven patient groups. Health Econ 13: 873-884.

33. Glasziou P, Alexander J, Beller E and Clarke P. (2007) Which health-related quality of life score? A comparison of alternative utility measures in patients with Type 2 diabetes in the ADVANCE trial. Health Qual Life Outcomes 5: 21.

34. Brazier J, Roberts J and Deverill M. (2002) The estimation of a preference-based measure of health from the SF-36. J Health Econ 21: 271-292.

Chap

ter 1

17

Introduction

25. Siontis GC, Tzoulaki I, Siontis KC and Ioannidis JP. (2012) Comparisons of established risk prediction models for cardiovascular disease: systematic review. BMJ 344: e3318.

26. Selvarajah S, Kaur G, Haniff J, Cheong KC, Hiong TG, et al. (2014) Comparison of the Framingham Risk Score, SCORE and WHO/ISH cardiovascular risk prediction models in an Asian population. Int J Cardiol 176: 211-218.



29. Krousel-Wood MA, Muntner P, Islam T, Morisky DE and Webber LS. (2009) Barriers to and determinants of medication adherence in hypertension management: perspective of the cohort study of medication adherence among older adults. Med Clin North Am 93: 753-769.

30. Addison CC, Jenkins BW, Sarpong D, Wilson G, Champion C, et al. (2011) Relationship between medication use and cardiovascular disease health outcomes in the Jackson Heart Study. Int J Environ Res Public Health 8: 2505-2515.

31. McDonough CM, Grove MR, Tosteson TD, Lurie JD, Hilibrand AS, et al. (2005) Comparison of EQ-5D, HUI, and SF-36-derived societal health state values among spine patient outcomes research trial (SPORT) participants. Qual Life Res 14: 1321-1332.

Chapter 1

32. Brazier J, Roberts J, Tsuchiya A and Busschbach J. (2004) A comparison of the EQ-5D and SF-6D across seven patient groups. Health Econ 13: 873-884.

33. Glasziou P, Alexander J, Beller E and Clarke P. (2007) Which health-related quality of life score? A comparison of alternative utility measures in patients with Type 2 diabetes in the ADVANCE trial. Health Qual Life Outcomes 5: 21.

34. Brazier J, Roberts J and Deverill M. (2002) The estimation of a preference-based measure of health from the SF-36. J Health Econ 21: 271-292.

Chapter 2

Models to predict the burden of cardiovascular disease risk in a rural

mountainous region of Vietnam

Thi-Phuong-Lan Nguyen, C.C.M.Schuiling-Veninga,

Thi-Bach-Yen-Nguyen, Thu-Hang Vu, E.P.Wright, M.J. Postma

Value in Health Regional Issues:

Volume 3, May 2014, Pages 87–93

20

Models to predict the burden of cardiovascular disease risk in Vietnam

Abstract

Objectives

To compare and identify the most appropriate model to predict cardiovascular disease in a rural area in Northern Vietnam, using data on hypertension from the communities.

Methods

A cross sectional survey was conducted including all residents in selected communities, aged 35 to 64 years, during April to August 2012 in Thai Nguyen province. Data on age, sex, smoking status, blood pressure, and blood tests (glucose, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein) were collected to identify prevalence of high blood pressure and to use as input variables for the models. We compared three models: Asian, Chinese Multiple-provincial Cohort Study (CMCS) and Framingham, to estimate cardiovascular risk in the coming years in this context and compare these models and outcomes.

Results

The prevalence of high blood pressure in these communities was lower than reported nationally (12.3%). CVD risk differed greatly depending on the model applied: approximately 21% of subjects according to the CMCS and Asian models, but 37% using the Framingham model had more than 10% risk for CVD. In the group without current CVD, these numbers decreased to 9% using the CMCS and Asian models, but 28% according to the Framingham model. There were no significant differences between the Asian and

Chapter 2

CMCS models but differences were highly significant when comparing Asian vs. Framingham or CMCS vs. Framingham model.

Conclusion

The Asian and CMCS models provided similar results in predicting CVD risk in the Vietnamese population in Thai Nguyen. The Framingham model provided vastly different results and might be less suitable for these populations.

Chap

ter 2

21


Abstract

Objectives

To compare and identify the most appropriate model to predict cardiovascular disease in a rural area in Northern Vietnam, using data on hypertension from the communities.

Methods

A cross sectional survey was conducted including all residents in selected communities, aged 35 to 64 years, during April to August 2012 in Thai Nguyen province. Data on age, sex, smoking status, blood pressure, and blood tests (glucose, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein) were collected to identify prevalence of high blood pressure and to use as input variables for the models. We compared three models: Asian, Chinese Multiple-provincial Cohort Study (CMCS) and Framingham, to estimate cardiovascular risk in the coming years in this context and compare these models and outcomes.

Results

The prevalence of high blood pressure in these communities was lower than reported nationally (12.3%). CVD risk differed greatly depending on the model applied: approximately 21% of subjects according to the CMCS and Asian models, but 37% using the Framingham model had more than 10% risk for CVD. In the group without current CVD, these numbers decreased to 9% using the CMCS and Asian models, but 28% according to the Framingham model. There were no significant differences between the Asian and

Chapter 2

CMCS models but differences were highly significant when comparing Asian vs. Framingham or CMCS vs. Framingham model.

Conclusion

The Asian and CMCS models provided similar results in predicting CVD risk in the Vietnamese population in Thai Nguyen. The Framingham model provided vastly different results and might be less suitable for these populations.

22


Introduction

The burden of cardiovascular diseases (CVD) is increasing worldwide; it is a leading cause of death both globally and in Asian countries [1-3]. Of all CVD, ischemic heart disease and cerebrovascular disease/stroke were the most common causes of death in 1990 and 2010. The mortality rate of both diseases has increased over the past two decades: approximately 35% for ischemic heart disease and 26% for stroke. In 2010, both diseases also belonged to the top five causes of years of life lost [1]. Hypertensive heart disease increased from 18.3 in 1990 to 14.2 in 2010 in the global ranking of causes of death. In Vietnam, CVD was among the top 10 leading causes of death in 2006, 2007 and 2009 [4, 5]. Around 32% of deaths from non-communicable diseases in rural areas are caused by cardiovascular diseases [6, 7].

To make best use of limited resources, planning of health care interventions is important. Predictive modeling of the risk of CVD can render valuable information for planning these health interventions.

The first step in the prediction of the CVD risk is the selection of the best-fitting model for the Vietnamese population. Over the last decade, several models have been developed and validated. The first well-known model was developed in the US using data from the Framingham study. However, this model might be less reliable in other populations and has overestimated or underestimated the CVD risks in specific settings [8, 9]. Therefore, other models were developed, such as the SCORE in Europe[10], QRISK in the United Kingdom [11], PROCAM in Germany [12], ASSIGN in Scotland [13], and CUORE in Italy [14]. In Asia, models to predict CVD risk have been developed in Thailand, China (Chinese Multiple-provincial Cohort Study (CMCS)), Japan, Malaysia, and Singapore. For all Asian populations, a tool was developed based on six cohorts in this region

Chapter 2

(Asian Model) [8, 15-21]. As all these models vary in different aspects like the time horizon used, characteristics of study population included, input variables, and outcome, they may all produce different results. In Vietnam, with its own environment of biological, behavioral and social characteristics, there is not yet a specific model to predict CVD. The largest survey on risk factors for CVD done in Vietnam applied the Framingham model only [22]. Also, the Framingham model is explicitly introduced on the Vietnam Heart Association website. The aim of this study is to calculate the CVD risk using two prediction models developed in Asia and assess to what extent these results coincides with CVD risks calculated using the long-established Framingham model. In this way, we hope to identify the most suitable model to predict future patterns of disease for health planners in Vietnam.

The second step for modeling of the CVD risk is information on risk factors and demographic characteristics of the population. It is well-known that improved CVD management could be approached through changing life style. Additionally, there are concerns on the capacity of health services to adequately detect risk factors such as hypertension. It is estimated that half of the expected 25% hypertensive individuals in the population are not aware of their hypertension [23]. Better CVD management is public health concern and hypertension detection and treatment are crucial elements in that management. There have been large surveys on hypertension in Vietnam [22, 23]. This report describes a smaller but potentially more intensive study. A specific sub-aim of our study was to collect complete and reliable data to assess the prevalence of hypertension and distribution of risk factors in four rural communities, in a mountainous area in the North of Vietnam, subsequently to be used as input variables for CVD models.

Chap

ter 2

23


Introduction

The burden of cardiovascular diseases (CVD) is increasing worldwide; it is a leading cause of death both globally and in Asian countries [1-3]. Of all CVD, ischemic heart disease and cerebrovascular disease/stroke were the most common causes of death in 1990 and 2010. The mortality rate of both diseases has increased over the past two decades: approximately 35% for ischemic heart disease and 26% for stroke. In 2010, both diseases also belonged to the top five causes of years of life lost [1]. Hypertensive heart disease increased from 18.3 in 1990 to 14.2 in 2010 in the global ranking of causes of death. In Vietnam, CVD was among the top 10 leading causes of death in 2006, 2007 and 2009 [4, 5]. Around 32% of deaths from non-communicable diseases in rural areas are caused by cardiovascular diseases [6, 7].

To make best use of limited resources, planning of health care interventions is important. Predictive modeling of the risk of CVD can render valuable information for planning these health interventions.

The first step in the prediction of the CVD risk is the selection of the best-fitting model for the Vietnamese population. Over the last decade, several models have been developed and validated. The first well-known model was developed in the US using data from the Framingham study. However, this model might be less reliable in other populations and has overestimated or underestimated the CVD risks in specific settings [8, 9]. Therefore, other models were developed, such as the SCORE in Europe[10], QRISK in the United Kingdom [11], PROCAM in Germany [12], ASSIGN in Scotland [13], and CUORE in Italy [14]. In Asia, models to predict CVD risk have been developed in Thailand, China (Chinese Multiple-provincial Cohort Study (CMCS)), Japan, Malaysia, and Singapore. For all Asian populations, a tool was developed based on six cohorts in this region

Chapter 2

(Asian Model) [8, 15-21]. As all these models vary in different aspects like the time horizon used, characteristics of study population included, input variables, and outcome, they may all produce different results. In Vietnam, with its own environment of biological, behavioral and social characteristics, there is not yet a specific model to predict CVD. The largest survey on risk factors for CVD done in Vietnam applied the Framingham model only [22]. Also, the Framingham model is explicitly introduced on the Vietnam Heart Association website. The aim of this study is to calculate the CVD risk using two prediction models developed in Asia and assess to what extent these results coincides with CVD risks calculated using the long-established Framingham model. In this way, we hope to identify the most suitable model to predict future patterns of disease for health planners in Vietnam.

The second step for modeling of the CVD risk is information on risk factors and demographic characteristics of the population. It is well-known that improved CVD management could be approached through changing life style. Additionally, there are concerns on the capacity of health services to adequately detect risk factors such as hypertension. It is estimated that half of the expected 25% hypertensive individuals in the population are not aware of their hypertension [23]. Better CVD management is public health concern and hypertension detection and treatment are crucial elements in that management. There have been large surveys on hypertension in Vietnam [22, 23]. This report describes a smaller but potentially more intensive study. A specific sub-aim of our study was to collect complete and reliable data to assess the prevalence of hypertension and distribution of risk factors in four rural communities, in a mountainous area in the North of Vietnam, subsequently to be used as input variables for CVD models.

24


Methods

Models

As described above, several CVD prediction models are available around the world. In this study, we selected two models developed in Asia, because their target populations may share a similar background of disease and risk factors for CVD to Vietnam. This would avert the potential over- or underestimation that may appear when the Framingham model or others developed for Western countries are used. The Asian model was validated based on six cohorts in Asia and may be most representative for Asian countries. The Chinese Multiple-provincial Cohort Study (CMCS) model was developed based on cohorts in 11 provinces in China. We are interested in this model as well because Vietnamese share a background of mortality patterns with the Chinese: CVD was 40% in Vietnam and 38% in China among deaths due to non-communicable disease, which account for 80% of all deaths in China and 75% of all deaths in Vietnam. Age-standardized death rate per 100,000 due to CVD and diabetes was similar in China and Vietnam (male: 311.5 and 381.5 in China and Vietnam; female: 259.6 and 298.2 in China and Vietnam, respectively). High BP in 2010 was 38.2% in China and 33% in Vietnam. Patterns of obesity or overweight may, however, be different [24, 25].

We also applied the Framingham model, which is widely used and also - as mentioned - is presented on the Vietnam Heart Association website. This tool was developed from cohort of Americans who were free of coronary heart disease and diabetes, to predict the 10-year risk of myocardial infarction and coronary heart death.

Risk factors included in the models differ to some extent. Information on sex, age and smoking status is, however, included in all three

Chapter 2

models. Furthermore, the CMCS model also includes information on diabetes, stage of hypertension, total cholesterol, and high-density lipoprotein cholesterol. In the Asian model, systolic blood pressure and total cholesterol are used as input variables. In the Framingham model, the following variables are used: total cholesterol, high-density lipoprotein cholesterol and systolic blood pressure.

We calculated the CVD risk of each individual using equations of CMCS and Asian model which consider above input variables, and their own mean survival rate, the regression coefficients, and mean values of the risk factors in CMCS or Asian models. In the Framingham model, we calculated the point score of each input variable and translated point scores into risk levels, guided by the Framingham Heart Study. The CMCS, Asian and Framingham models have been described in detail [8, 26, 27].

The CVD risk was assessed using all three models for two scenarios. The base case excluded all cases with current CVD, whereas in the sensitivity analysis all cases, both with and without current CVD, were included.

Study population and study design

A cross sectional study was conducted from April to August 2012 in Thai Nguyen, a mountainous province in the North of Vietnam. Four districts, all around 40 kilometers from Thai Nguyen city, were purposively selected; we then chose one commune in the middle of each district. Within each commune, all villages having an active health care worker were listed. From this list, we selected randomly enough villages to cover 45 to 50% of the total population in that commune.

All residents from 35 to 64 years old in each village were invited to participate in the study. Exclusion criteria, based on self-reported

Chap

ter 2

25


Methods

Models

As described above, several CVD prediction models are available around the world. In this study, we selected two models developed in Asia, because their target populations may share a similar background of disease and risk factors for CVD to Vietnam. This would avert the potential over- or underestimation that may appear when the Framingham model or others developed for Western countries are used. The Asian model was validated based on six cohorts in Asia and may be most representative for Asian countries. The Chinese Multiple-provincial Cohort Study (CMCS) model was developed based on cohorts in 11 provinces in China. We are interested in this model as well because Vietnamese share a background of mortality patterns with the Chinese: CVD was 40% in Vietnam and 38% in China among deaths due to non-communicable disease, which account for 80% of all deaths in China and 75% of all deaths in Vietnam. Age-standardized death rate per 100,000 due to CVD and diabetes was similar in China and Vietnam (male: 311.5 and 381.5 in China and Vietnam; female: 259.6 and 298.2 in China and Vietnam, respectively). High BP in 2010 was 38.2% in China and 33% in Vietnam. Patterns of obesity or overweight may, however, be different [24, 25].

We also applied the Framingham model, which is widely used and also - as mentioned - is presented on the Vietnam Heart Association website. This tool was developed from cohort of Americans who were free of coronary heart disease and diabetes, to predict the 10-year risk of myocardial infarction and coronary heart death.

Risk factors included in the models differ to some extent. Information on sex, age and smoking status is, however, included in all three

Chapter 2

models. Furthermore, the CMCS model also includes information on diabetes, stage of hypertension, total cholesterol, and high-density lipoprotein cholesterol. In the Asian model, systolic blood pressure and total cholesterol are used as input variables. In the Framingham model, the following variables are used: total cholesterol, high-density lipoprotein cholesterol and systolic blood pressure.

We calculated the CVD risk of each individual using equations of CMCS and Asian model which consider above input variables, and their own mean survival rate, the regression coefficients, and mean values of the risk factors in CMCS or Asian models. In the Framingham model, we calculated the point score of each input variable and translated point scores into risk levels, guided by the Framingham Heart Study. The CMCS, Asian and Framingham models have been described in detail [8, 26, 27].

The CVD risk was assessed using all three models for two scenarios. The base case excluded all cases with current CVD, whereas in the sensitivity analysis all cases, both with and without current CVD, were included.

Study population and study design

A cross sectional study was conducted from April to August 2012 in Thai Nguyen, a mountainous province in the North of Vietnam. Four districts, all around 40 kilometers from Thai Nguyen city, were purposively selected; we then chose one commune in the middle of each district. Within each commune, all villages having an active health care worker were listed. From this list, we selected randomly enough villages to cover 45 to 50% of the total population in that commune.

All residents from 35 to 64 years old in each village were invited to participate in the study. Exclusion criteria, based on self-reported

26


medical history or alcohol consumption, were current pregnancy, mental health disease, heavy alcohol consumption (often drunk), cancer, AIDS, current treatment of hypertension or diabetes, or a medical history of myocardial infarction, stroke, heart failure or kidney failure.

Medical students and village health workers, who were trained on blood pressure (BP) measurement and on using the study questionnaires, collected data at a village center or in the household. Data were collected on demographic characteristics (sex, age, ethnicity, education, occupation, and residence), risk factors for CVD (smoking, alcohol consumption) and BP was measured. All subjects who had systolic blood pressure (SBP) ≥140 mm Hg and/or diastolic blood pressure (DBP) ≥90 mm Hg were invited to visit the community health center for a second BP measurement, 5 to 7 days after the first measurement. All subjects who still had SBP ≥140 mm Hg and/or DBP ≥90 mm Hg on the second day were medically examined including an ECG and blood samples were collected. These people were advised to manage high BP at their community health center or were referred to a higher level for further investigation. All subjects who had SBP < 140 mm Hg and/or DBP < 90 mm Hg were advised to check their BP regularly.

Measurement and classification

BP measurement: Participants were advised to avoid drinking alcohol the day before the visit, and to stop cigarette smoking, drinking coffee/tea and taking exercise for at least 30 minutes before their BP measurement. Automatic sphygmomanometers (OMRON, model HEM-7200-C1, Healthcare, Japan) with an appropriate sized cuff were used. BP was measured twice for everyone, with participants in a sitting position after at least 10 minutes of rest. A third measurement was performed if the difference between the first two was over 10

Chapter 2

mmHg for SBP or DBP. Average SBP and/or DBP of the first and second or second and third measurements were used for analysis [28, 29]. Hypertension was defined as value of SBP ≥ 140 mm Hg and/or DBP ≥90 mm Hg on the second visit. It has been recommended that in a setting of scarce resources, the value of blood pressure at the second visit would be suitable for a reliable measurement of the real prevalence of hypertension in the community [30]. Hypertension was classified according to the Seventh Joint National Committee on Hypertension Classification[31].

Occupations were assigned to one of seven categories: farmer, worker, businessperson, wage earner, housewife, government officer, and other.

In the second visit, patients were examined by medical doctors from Thai Nguyen University of Medicine and Pharmacy. For this study, we defined coronary heart disease (including angina, left ventricular hypertrophy, myocardial infarction, or heart failure) by medical history, physical examination and result of electrocardiogram. Blood tests (glucose, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein) were performed by a team from Thai Nguyen General Hospital. Diabetes was defined as fasting blood glucose level ≥ 7 mmol/l [32]. Subjects who smoked cigarettes or pipe during the last month were defined as current smokers.

Statistical analysis

Chi square, student’s t-test or Mann-Whitney test were used to compare differences in biological factors among different visits, different groups, and the distribution of CVD in different models. Chi square testing was used to compare model differences. Based on the Guideline for Assessment and Management of Cardiovascular Risk, we used 10% as cutoff point for CVD risks over a 10-year period. In particular, it is stated that with a 10-year risk of CVD greater than 10%, subjects need to be monitored regularly [33]. All statistical analysis were performed using STATA 10.

Chap

ter 2

27


medical history or alcohol consumption, were current pregnancy, mental health disease, heavy alcohol consumption (often drunk), cancer, AIDS, current treatment of hypertension or diabetes, or a medical history of myocardial infarction, stroke, heart failure or kidney failure.

Medical students and village health workers, who were trained on blood pressure (BP) measurement and on using the study questionnaires, collected data at a village center or in the household. Data were collected on demographic characteristics (sex, age, ethnicity, education, occupation, and residence), risk factors for CVD (smoking, alcohol consumption) and BP was measured. All subjects who had systolic blood pressure (SBP) ≥140 mm Hg and/or diastolic blood pressure (DBP) ≥90 mm Hg were invited to visit the community health center for a second BP measurement, 5 to 7 days after the first measurement. All subjects who still had SBP ≥140 mm Hg and/or DBP ≥90 mm Hg on the second day were medically examined including an ECG and blood samples were collected. These people were advised to manage high BP at their community health center or were referred to a higher level for further investigation. All subjects who had SBP < 140 mm Hg and/or DBP < 90 mm Hg were advised to check their BP regularly.

Measurement and classification

BP measurement: Participants were advised to avoid drinking alcohol the day before the visit, and to stop cigarette smoking, drinking coffee/tea and taking exercise for at least 30 minutes before their BP measurement. Automatic sphygmomanometers (OMRON, model HEM-7200-C1, Healthcare, Japan) with an appropriate sized cuff were used. BP was measured twice for everyone, with participants in a sitting position after at least 10 minutes of rest. A third measurement was performed if the difference between the first two was over 10

Chapter 2

mmHg for SBP or DBP. Average SBP and/or DBP of the first and second or second and third measurements were used for analysis [28, 29]. Hypertension was defined as value of SBP ≥ 140 mm Hg and/or DBP ≥90 mm Hg on the second visit. It has been recommended that in a setting of scarce resources, the value of blood pressure at the second visit would be suitable for a reliable measurement of the real prevalence of hypertension in the community [30]. Hypertension was classified according to the Seventh Joint National Committee on Hypertension Classification[31].

Occupations were assigned to one of seven categories: farmer, worker, businessperson, wage earner, housewife, government officer, and other.

In the second visit, patients were examined by medical doctors from Thai Nguyen University of Medicine and Pharmacy. For this study, we defined coronary heart disease (including angina, left ventricular hypertrophy, myocardial infarction, or heart failure) by medical history, physical examination and result of electrocardiogram. Blood tests (glucose, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein) were performed by a team from Thai Nguyen General Hospital. Diabetes was defined as fasting blood glucose level ≥ 7 mmol/l [32]. Subjects who smoked cigarettes or pipe during the last month were defined as current smokers.


Chi square, student’s t-test or Mann-Whitney test were used to compare differences in biological factors among different visits, different groups, and the distribution of CVD in different models. Chi square testing was used to compare model differences. Based on the Guideline for Assessment and Management of Cardiovascular Risk, we used 10% as cutoff point for CVD risks over a 10-year period. In particular, it is stated that with a 10-year risk of CVD greater than 10%, subjects need to be monitored regularly [33]. All statistical analysis were performed using STATA 10.

28


Ethical issues

The research proposal was approved by the Institutional Review Board in biomedical research in the Institute of Social and Medical Studies in Vietnam. The study was explained to all those invited to join, and all subjects signed a consent form before participating in the study. All participants had the right to withdraw from the study at any time. Patients who were found to have high BP were advised to manage it at their community health station or were referred to a higher level for specialist examination if needed. Subjects without high BP were advised to monitor their BP regularly.

Results

Study population characteristics

As shown in Fig 1, the 3,779 subjects who participated in the study comprised 86.5 % of all those who were invited. They included 43.3% males and 84% was ethnically Kinh (the majority ethnic group in Vietnam). The most common occupation (72.5%) was farmer and the mean age was 47.4 ± 8 years.

Chapter 2

Fig 1: Flow chart of selecting the final sample

Notes: We selected all subjects with hypertension in the second measurement. We required blood samples from all cases to have values for all variables to be included in the analysis with three models.

In this study, hypertension prevalence is a key factor and an important input variable. We therefore first present the data on hypertension prevalence, followed by the CVD model results.

Present in communities during screening for

hypertension: 4386 subjects selected

589 subjects declined to participate the screening

Attended the screening: 3779 subjects

(3779*100/4386 = 86.5%)

3003 subjects without hypertension in the first

measurement

Hypertension during the first measurement: 776 subjects

(776*100/3779=20.5%)10 subjects not followed

up and 300 subjects without hypertension in the second measurement

Hypertension during the second measurement: 466 subjects

(466*100/766=60.8%)

24 subjects did not provide blood samples

Blood samples: 442 subjects (entering the models)

Chap

ter 2

29


Ethical issues

The research proposal was approved by the Institutional Review Board in biomedical research in the Institute of Social and Medical Studies in Vietnam. The study was explained to all those invited to join, and all subjects signed a consent form before participating in the study. All participants had the right to withdraw from the study at any time. Patients who were found to have high BP were advised to manage it at their community health station or were referred to a higher level for specialist examination if needed. Subjects without high BP were advised to monitor their BP regularly.

Results

Study population characteristics

As shown in Fig 1, the 3,779 subjects who participated in the study comprised 86.5 % of all those who were invited. They included 43.3% males and 84% was ethnically Kinh (the majority ethnic group in Vietnam). The most common occupation (72.5%) was farmer and the mean age was 47.4 ± 8 years.

Chapter 2

Fig 1: Flow chart of selecting the final sample

Notes: We selected all subjects with hypertension in the second measurement. We required blood samples from all cases to have values for all variables to be included in the analysis with three models.

In this study, hypertension prevalence is a key factor and an important input variable. We therefore first present the data on hypertension prevalence, followed by the CVD model results.

Present in communities during screening for

hypertension: 4386 subjects selected

589 subjects declined to participate the screening


(3779*100/4386 = 86.5%)


measurement

Hypertension during the first measurement: 776 subjects

(776*100/3779=20.5%)10 subjects not followed

up and 300 subjects without hypertension in the second measurement

Hypertension during the second measurement: 466 subjects

(466*100/766=60.8%)


Blood samples: 442 subjects (entering the models)

30


Prevalence of hypertension

The distribution of BP among 3779 subjects in the first visit is presented in Table 1. Among this study population, 776 subjects (20.5%) had SBP ≥140 mm Hg and/or DBP ≥ 90 mm Hg during the first visit. The prevalence of elevate BP was greater among men than women (28.3% vs. 14.6%, p<0.001) and increased with age (10.5%, 23.9%, 33.4% in 35-44, 45-54 and 55-64 age groups, respectively, p<0.001). During the second visit, at which 766 subjects appeared, 466 subjects (60.8%) still had a high BP. Using the BP values of this second visit, the overall prevalence of hypertension in the total population (3779 subjects) was 12.3%, including 16% of men and 9.6% of women (p<0.001).Table 1: Prevalence of blood pressure classification in study population,

stratified by sex and age group (%) at 1st visit (n=3779)

Normal Prehypertension Stage 1 Stage 2 p value

Men

35-44 30.6 52.9 13.2 3.3

0.00045-54 22.0 45.2 21.8 11.0

55-64 19.1 38.0 26.0 16.9

Women

35-44 64.5 29.6 4.3 1.5

0.00045-54 46.7 36.0 12.6 4.7

55-64 36.5 37.3 18.1 8.1

Total

Men 24.9 46.8 19.2 9.10.000

Women 51.7 33.7 10.5 4.2

Both sexes 40.1 39.4 14.2 6.3

Note: Blood pressure level was clasified according to JNC 7: Normal: SBP/DBP<120/80 mm Hg; Prehypertension: SBP/DBP is 120-139/80-89 mm Hg; Stage 1: SBP/DBP is 140-159/90-99 mm Hg; Stage 2: SBP/DBP ≥160/100 mm Hg; men vs women: p=0.000;

Chapter 2

Both the median and mean value of the SBP and DBP were lower at the second visit compared to measurements during the first visit (mean: 151.3 mm Hg and 147.7 mm Hg; median: 148 mm Hg and 145 mm Hg in the first and second visits respectively, p<0.001). However, the range of BP in the second visit was greater than in the first visit. Biological factors among hypertensive subjects (at second measurement) are presented in Table 2. As blood tests of 24 subjects were missing, data on only 442 subjects are presented. Age, fasting glucose, and SBP were comparable between men and women (p>0.05). Total cholesterol and LDL-cholesterol were lower in men (p<0.05 and p<0.01 respectively) whereas HDL-cholesterol was higher in men (p<0.05).

Chap

ter 2

31


Prevalence of hypertension

The distribution of BP among 3779 subjects in the first visit is presented in Table 1. Among this study population, 776 subjects (20.5%) had SBP ≥140 mm Hg and/or DBP ≥ 90 mm Hg during the first visit. The prevalence of elevate BP was greater among men than women (28.3% vs. 14.6%, p<0.001) and increased with age (10.5%, 23.9%, 33.4% in 35-44, 45-54 and 55-64 age groups, respectively, p<0.001). During the second visit, at which 766 subjects appeared, 466 subjects (60.8%) still had a high BP. Using the BP values of this second visit, the overall prevalence of hypertension in the total population (3779 subjects) was 12.3%, including 16% of men and 9.6% of women (p<0.001).Table 1: Prevalence of blood pressure classification in study population,

stratified by sex and age group (%) at 1st visit (n=3779)

Normal Prehypertension Stage 1 Stage 2 p value

Men

35-44 30.6 52.9 13.2 3.3

0.00045-54 22.0 45.2 21.8 11.0

55-64 19.1 38.0 26.0 16.9

Women

35-44 64.5 29.6 4.3 1.5

0.00045-54 46.7 36.0 12.6 4.7

55-64 36.5 37.3 18.1 8.1

Total

Men 24.9 46.8 19.2 9.10.000

Women 51.7 33.7 10.5 4.2

Both sexes 40.1 39.4 14.2 6.3

Note: Blood pressure level was clasified according to JNC 7: Normal: SBP/DBP<120/80 mm Hg; Prehypertension: SBP/DBP is 120-139/80-89 mm Hg; Stage 1: SBP/DBP is 140-159/90-99 mm Hg; Stage 2: SBP/DBP ≥160/100 mm Hg; men vs women: p=0.000;

Chapter 2

Both the median and mean value of the SBP and DBP were lower at the second visit compared to measurements during the first visit (mean: 151.3 mm Hg and 147.7 mm Hg; median: 148 mm Hg and 145 mm Hg in the first and second visits respectively, p<0.001). However, the range of BP in the second visit was greater than in the first visit. Biological factors among hypertensive subjects (at second measurement) are presented in Table 2. As blood tests of 24 subjects were missing, data on only 442 subjects are presented. Age, fasting glucose, and SBP were comparable between men and women (p>0.05). Total cholesterol and LDL-cholesterol were lower in men (p<0.05 and p<0.01 respectively) whereas HDL-cholesterol was higher in men (p<0.05).

32


Table 2: Characteristics of hypertensive patients n=442 subjects)

CharacteristicsUsing in model

Men (n=244)

Women (n=198)

Both sexes

p value

Age (mean ±SD)CMCS, Asian, Framingham

52.1(± 7.8)

53(± 7.3)

52(± 7.6) 0.249

SmokingCMCS, Asian, Framingham

38.93% 0% 21.49% 0.000

Diabetes (cases) CMCS 3 2 5 0.828

SBP (mm Hg)

(mean ± SD)Asian

159

(± 15.7)

158

(± 14.7)

158

(±15)0.505

Stage of hypertension (cases) CMCS, Framingham

SBP 140 – 159

mm Hg or DBP

90 – 99 mmHg

132 116 248

0.344

SBP ≥160 mm Hg or

DBP ≥ 100mmHg112 82 194

TC (mmol/l)

(mean ± SD)Asian

5.03

(±1.51)

5.22

(±1.17)

5.11

(±1.37)0.015

TC (cases) CMCS, Framingham

<160 (mg/dL) 59 36 95 0.127

160 – 199 (mg/dL) 93 64 157 0.206

200 – 239 (mg/dL) 58 60 118 0.123

240 – 279 (mg/dL) 23 28 51 0.123

≥ 280 (mg/dL) 11 10 21 0.790

HDL-C (cases) CMCS, Framingham

< 35 (mg/dL) 21 22 43 0.377

35 – 44 (mg/dL) 63 58 121 0.415

45 – 49 (mg/dL) 45 35 80 0.835

50 – 59 (mg/dL) 57 53 110 0.410

≥ 60 (mg/dL) 58 30 88 0.024

Chapter 2

Comparison of future patterns of CVD using CMCS, Asian and Framingham model

Risks of coronary heart disease were estimated using three models: the CMCS, Asian and Framingham models. In the base case, we excluded cases with a diagnosis of CVD at the time of examination in the second visit in all models (Table 3). The prevalence of ‘more than 10% CVD risk’ when the Framingham model was applied was 28.24%, higher than for the CMCS (11.7 %) and Asian models ((9.1%). For women, 24.6% were at ‘more than 10% CVD risk’ according to the CMCS model, compared to only 1.6% and 0.5% in the Asian and Framingham models. In the age group 55 to 64 years, the prevalence of ‘more than 10% at CVD risk’ was highest (46%) using the Framingham model and lowest (15.3%) with the Asian model. The differences in the results between CMCS model vs. Framingham model and Asian model vs. Framingham model were highly significant (p<0.001). There were, however, no significant differences between results from the CMCS and the Asian model (p>0.05).

Chap

ter 2

33


Table 2: Characteristics of hypertensive patients n=442 subjects)

CharacteristicsUsing in model

Men (n=244)

Women (n=198)

Both sexes

p value

Age (mean ±SD)CMCS, Asian, Framingham

52.1(± 7.8)

53(± 7.3)

52(± 7.6) 0.249

SmokingCMCS, Asian, Framingham

38.93% 0% 21.49% 0.000

Diabetes (cases) CMCS 3 2 5 0.828

SBP (mm Hg)

(mean ± SD)Asian

159

(± 15.7)

158

(± 14.7)

158

(±15)0.505

Stage of hypertension (cases) CMCS, Framingham

SBP 140 – 159

mm Hg or DBP

90 – 99 mmHg

132 116 248

0.344

SBP ≥160 mm Hg or

DBP ≥ 100mmHg112 82 194

TC (mmol/l)

(mean ± SD)Asian

5.03

(±1.51)

5.22

(±1.17)

5.11

(±1.37)0.015

TC (cases) CMCS, Framingham

<160 (mg/dL) 59 36 95 0.127

160 – 199 (mg/dL) 93 64 157 0.206

200 – 239 (mg/dL) 58 60 118 0.123

240 – 279 (mg/dL) 23 28 51 0.123

≥ 280 (mg/dL) 11 10 21 0.790

HDL-C (cases) CMCS, Framingham

< 35 (mg/dL) 21 22 43 0.377

35 – 44 (mg/dL) 63 58 121 0.415

45 – 49 (mg/dL) 45 35 80 0.835

50 – 59 (mg/dL) 57 53 110 0.410

≥ 60 (mg/dL) 58 30 88 0.024

Chapter 2

Comparison of future patterns of CVD using CMCS, Asian and Framingham model

Risks of coronary heart disease were estimated using three models: the CMCS, Asian and Framingham models. In the base case, we excluded cases with a diagnosis of CVD at the time of examination in the second visit in all models (Table 3). The prevalence of ‘more than 10% CVD risk’ when the Framingham model was applied was 28.24%, higher than for the CMCS (11.7 %) and Asian models ((9.1%). For women, 24.6% were at ‘more than 10% CVD risk’ according to the CMCS model, compared to only 1.6% and 0.5% in the Asian and Framingham models. In the age group 55 to 64 years, the prevalence of ‘more than 10% at CVD risk’ was highest (46%) using the Framingham model and lowest (15.3%) with the Asian model. The differences in the results between CMCS model vs. Framingham model and Asian model vs. Framingham model were highly significant (p<0.001). There were, however, no significant differences between results from the CMCS and the Asian model (p>0.05).

34

T

able

3: P

reva

lenc

e of

risk

of c

ardi

ovas

cula

r eve

nts a

ccor

ding

to d

iffer

ent m

odel

s in

rura

l mou

ntai

nous

Vie

tnam

, stra

tifie

d by

sex

and

age

grou

p (e

xclu

ding

cas

es h

avin

g C

VD

)(%

)

CM

CS

Asi

anFr

amin

gham

<10%

>=10

%p

valu

e<1

0%>=

10%

p va

lue

<10%

>=10

%p

valu

e

Age

gro

up0.

000

0.00

00.

000

35-4

410

010

096

.67

3.33

45-5

498

.77

1.23

93.8

76.

1380

.37

19.6

3

55-6

473

.62

26.3

884

.66

15.3

453

.99

46.0

1

Sexe

s0.

000

0.00

00.

000

Men

(n=2

03)

100

84.2

415

.76

46.8

53.2

Wom

en (n

=183

)75

.41

24.5

998

.36

1.64

99.4

50.

55

Bot

h se

xes (

n=38

6)88

.34

11.6

690

.93

9.07

71.7

628

.24

p va

lue

0.00

0

Not

es:

CM

CS

mod

el v

s A

sian

mod

el:

p= 0

.238

; C

MC

S m

odel

vs

Fram

ingh

am m

odel

: p=

0.00

0; A

sian

mod

el v

s Fr

amin

gham

mod

el:

p=0.

000

Chapter 2

In the sensitivity analysis, we included cases with a diagnosis of CVD at the examination in the second visit in each of the three models, and placed those cases in groups with more than 10% CVD risk (Table 4). The prevalence of the group with ‘more than 10% CVD risk’ was highest in the Framingham model (22.2%, 20.6% and 37.3% in CMCS, Asian and Framingham respectively; p<0.01).The prevalence of the group ‘more than 10% CVD risk’ in women was similar in the Asian and Framingham models (8% and 9% in Framingham and Asian respectively) but it was three times higher when the CMCS model was used (30.3%). In contrast, the prevalence of ‘more than 10% CVD risk’ in men was lowest using the CMCS model (16.8%)and highest using the Framingham model (61.07%). CVD risk increased with age at all models (p<0.001). In the group from 55 to 64 years old, the prevalence of ‘more than 10% CVD risk’ was 25.4%, 35.1% and 52.4% in the Asian, CMCS and Framingham model results, respectively.

Chap

ter 2

35

T

able

3: P

reva

lenc

e of

risk

of c

ardi

ovas

cula

r eve

nts a

ccor

ding

to d

iffer

ent m

odel

s in

rura

l mou

ntai

nous

Vie

tnam

, stra

tifie

d by

sex

and

age

grou

p (e

xclu

ding

cas

es h

avin

g C

VD

)(%

)

CM

CS

Asi

anFr

amin

gham

<10%

>=10

%p

valu

e<1

0%>=

10%

p va

lue

<10%

>=10

%p

valu

e

Age

gro

up0.

000

0.00

00.

000

35-4

410

010

096

.67

3.33

45-5

498

.77

1.23

93.8

76.

1380

.37

19.6

3

55-6

473

.62

26.3

884

.66

15.3

453

.99

46.0

1

Sexe

s0.

000

0.00

00.

000

Men

(n=2

03)

100

84.2

415

.76

46.8

53.2

Wom

en (n

=183

)75

.41

24.5

998

.36

1.64

99.4

50.

55

Bot

h se

xes (

n=38

6)88

.34

11.6

690

.93

9.07

71.7

628

.24

p va

lue

0.00

0

Not

es:

CM

CS

mod

el v

s A

sian

mod

el:

p= 0

.238

; C

MC

S m

odel

vs

Fram

ingh

am m

odel

: p=

0.00

0; A

sian

mod

el v

s Fr

amin

gham

mod

el:

p=0.

000

Chapter 2

In the sensitivity analysis, we included cases with a diagnosis of CVD at the examination in the second visit in each of the three models, and placed those cases in groups with more than 10% CVD risk (Table 4). The prevalence of the group with ‘more than 10% CVD risk’ was highest in the Framingham model (22.2%, 20.6% and 37.3% in CMCS, Asian and Framingham respectively; p<0.01).The prevalence of the group ‘more than 10% CVD risk’ in women was similar in the Asian and Framingham models (8% and 9% in Framingham and Asian respectively) but it was three times higher when the CMCS model was used (30.3%). In contrast, the prevalence of ‘more than 10% CVD risk’ in men was lowest using the CMCS model (16.8%)and highest using the Framingham model (61.07%). CVD risk increased with age at all models (p<0.001). In the group from 55 to 64 years old, the prevalence of ‘more than 10% CVD risk’ was 25.4%, 35.1% and 52.4% in the Asian, CMCS and Framingham model results, respectively.

36

Tab

le 4

: Pre

vale

nce

of ri

sk o

f car

diov

ascu

lar e

vent

s acc

ordi

ng to

diff

eren

t mod

els i

n ru

ral m

ount

aino

us V

ietn

am,

stra

tifie

d by

sex

and

age

grou

p (in

clud

ing

case

s hav

ing

CV

D) (

%)

CM

CS

mod

elA

sian

mod

elFr

amin

gham

mod

el

<10%

>=10

%p

valu

e<1

0%>=

10%

p va

lue

<10%

>=10

%p

valu

e

Age

gro

up0.

000

0.04

60.

000

35-4

488

.24

11.7

688

.24

11.7

685

.29

14.7

1

45-5

485

.19

14.8

180

.95

19.0

569

.31

30.6

9

55-6

464

.86

35.1

474

.59

25.4

147

.57

52.4

3

Sexe

s0.

001

0.00

00.

000

Men

(n=2

44)

83.2

16.8

70.0

829

.92

38.9

361

.07

Wom

en (n

=198

)69

.730

.390

.91

9.09

91.9

28.

08

Bot

h se

xes (

n=44

2)77

.85

22.1

579

.41

20.5

962

.67

37.3

3

p va

lue

0.00

0

Not

es: C

MC

S m

odel

vs

Asi

an m

odel

: p=

0.41

5; C

MC

S m

odel

vs

Fram

ingh

am m

odel

: p=0

.000

; Asi

an m

odel

vs

Fram

ingh

am m

odel

: p=0

.000

;m

en v

s wom

en p

=0.0

00

Chapter 2

Discussion

Hypertension is key risk factor for CVD, which is a leading cause of death in Asia and in Vietnam. Prediction of CVD risk by using the most appropriate model for Vietnamese population would support management of CVD, in curative and preventive health care. In this study, we predicted CVD risk among hypertensive patients by using and comparing three established prediction models.

BP measurements were taken during a first visit to the community and for those with high BP, again on a second visit a few days later [30].The prevalence of high BP decreased from 20.5% in the first visit to 12.3% in the second visit, which is low compared to results from previously reported studies in Vietnam (25.1% and 26.4%) [22, 23].The difference could be explained by differences in study methods and geographical setting. A previous study used sampling in a population ≥ 25 years old, in both urban and rural areas; blood tests were conducted in 20% of the urban study population, and BP was measured during one visit. Our study was conducted in a rural mountainous area, measuring all of the village population from 34 to 65 years old, and the BP was measured on two separate occasions in all subjects with high BP in the first visit. The prevalence of high BP in our study was within the range reported in other low and middle income countries in Asia (15%, 29% and 44.6% in Cambodia, Malaysia and Mongolia respectively) [34]. Again, these studies only measured the BP at a single visit, which may lead to overestimates of high BP prevalence. This finding suggests that researchers must be careful in identifying burden of disease in a specific setting and group; accurate data are needed for planning and allocation of scarce resources.

CVD risk differed largely depending on the model applied: approximately 9% of subjects according to the CMCS and Asian

Chap

ter 2

37

Tab

le 4

: Pre

vale

nce

of ri

sk o

f car

diov

ascu

lar e

vent

s acc

ordi

ng to

diff

eren

t mod

els i

n ru

ral m

ount

aino

us V

ietn

am,

stra

tifie

d by

sex

and

age

grou

p (in

clud

ing

case

s hav

ing

CV

D) (

%)

CM

CS

mod

elA

sian

mod

elFr

amin

gham

mod

el

<10%

>=10

%p

valu

e<1

0%>=

10%

p va

lue

<10%

>=10

%p

valu

e

Age

gro

up0.

000

0.04

60.

000

35-4

488

.24

11.7

688

.24

11.7

685

.29

14.7

1

45-5

485

.19

14.8

180

.95

19.0

569

.31

30.6

9

55-6

464

.86

35.1

474

.59

25.4

147

.57

52.4

3

Sexe

s0.

001

0.00

00.

000

Men

(n=2

44)

83.2

16.8

70.0

829

.92

38.9

361

.07

Wom

en (n

=198

)69

.730

.390

.91

9.09

91.9

28.

08

Bot

h se

xes (

n=44

2)77

.85

22.1

579

.41

20.5

962

.67

37.3

3

p va

lue

0.00

0

Not

es: C

MC

S m

odel

vs

Asi

an m

odel

: p=

0.41

5; C

MC

S m

odel

vs

Fram

ingh

am m

odel

: p=0

.000

; Asi

an m

odel

vs

Fram

ingh

am m

odel

: p=0

.000

;m

en v

s wom

en p

=0.0

00

Chapter 2

Discussion

Hypertension is key risk factor for CVD, which is a leading cause of death in Asia and in Vietnam. Prediction of CVD risk by using the most appropriate model for Vietnamese population would support management of CVD, in curative and preventive health care. In this study, we predicted CVD risk among hypertensive patients by using and comparing three established prediction models.

BP measurements were taken during a first visit to the community and for those with high BP, again on a second visit a few days later [30].The prevalence of high BP decreased from 20.5% in the first visit to 12.3% in the second visit, which is low compared to results from previously reported studies in Vietnam (25.1% and 26.4%) [22, 23].The difference could be explained by differences in study methods and geographical setting. A previous study used sampling in a population ≥ 25 years old, in both urban and rural areas; blood tests were conducted in 20% of the urban study population, and BP was measured during one visit. Our study was conducted in a rural mountainous area, measuring all of the village population from 34 to 65 years old, and the BP was measured on two separate occasions in all subjects with high BP in the first visit. The prevalence of high BP in our study was within the range reported in other low and middle income countries in Asia (15%, 29% and 44.6% in Cambodia, Malaysia and Mongolia respectively) [34]. Again, these studies only measured the BP at a single visit, which may lead to overestimates of high BP prevalence. This finding suggests that researchers must be careful in identifying burden of disease in a specific setting and group; accurate data are needed for planning and allocation of scarce resources.

CVD risk differed largely depending on the model applied: approximately 9% of subjects according to the CMCS and Asian

38


models, but using the Framingham model, 28% had a more than 10% risk for CVD. The prevalence of subjects having a ‘more than 10% CVD risk’ in our study (9 – 28%) was lower than reported in previous studies in Vietnam (40.9%), possibly because they used the Framingham model only. It was high compared to the numbers reported from other Asian countries in the age group 40 to 64 years ( 3%, 5.6%, 3.9% and 11.4% in Cambodia, Malaysia, China and Mongolia respectively) [22, 34, 35]. However, this comparison is of only limited validity because different methods were used to estimate CVD risk in these studies. Studies in other Asia countries used WHO/ISH risk prediction charts and were aware of potential underestimation of CVD risk [34]. The earlier study in Vietnam may have overestimated CVD risk also because it only measured BP at one visit and only did blood tests on the urban sample. These limitations may present a challenge for both curative and preventive care in CVD management based on such data.

In our study, inclusion or exclusion of current CVD was found to affect the predicted future pattern of CVD risk differently in each model. The results from the CMCS and Asian models were comparable but results from the Framingham model differed significantly. None of these models have yet been validated in Vietnam. Previous studies have suggested that applying the Framingham model in Europe and Asia resulted in overestimates of the burden of disease, where there were lower rates of CVD [8, 13, 36-39]. The differences when using the three models can be explained by the different components in each model, such as characteristics of study population, input variables, outcome variables, time horizon, and values in each equation. The Asian model considers the mean age, SBP, total cholesterol and smoke/non-smoke; outcome is defined as cardiovascular death, non-fatal myocardial infarction, and non-fatal

Chapter 2

cerebrovascular events in the following eight years. The CMCS equation uses a subgroup of risk factors, no mean values as in the Asian equation, and outcome is defined as coronary death and myocardial infarction in the coming 10 years.

The implication of our study is that Vietnam should of course ideally have its own cohort to build up its own model to predict CVD risk. However, due to the long time needed for that, and due to resource constraints, it may be legitimate to use a model from countries with similar contextual factors, such as the CMCS or the Asian models, to predict patterns of CVD risk in the coming years as a basis for planning and management of services in Vietnam. Applying the Framingham model seems to provide deviating results from both other models and might be less appropriate for this specific setting.

There are limitations to our study. First, we could not confirm high BP using ambulatory blood pressure monitoring. However, measurements taken at two visits are considered acceptable for low resource settings [30]. Secondly, we identified cases with current CVD based only on medical history, clinical exam and ECG so we may have lost cases with CVD without clinical symptoms and signs in ECG. Finally, the CMCS, Asian and Framingham original models can be applied to general populations. In this study, we only tested them in hypertensive patients, which reflects a limitation of our study. Notably, this may reduce the generalizability of our findings.

Conclusion

In conclusion, the prevalence of high BP in a rural mountainous area of Vietnam was found to be lower than previously reported for Vietnam, but still sufficient to warrant attention from health services. Our results demonstrate that the Asian and CMCS models provided

Chap

ter 2

39


models, but using the Framingham model, 28% had a more than 10% risk for CVD. The prevalence of subjects having a ‘more than 10% CVD risk’ in our study (9 – 28%) was lower than reported in previous studies in Vietnam (40.9%), possibly because they used the Framingham model only. It was high compared to the numbers reported from other Asian countries in the age group 40 to 64 years ( 3%, 5.6%, 3.9% and 11.4% in Cambodia, Malaysia, China and Mongolia respectively) [22, 34, 35]. However, this comparison is of only limited validity because different methods were used to estimate CVD risk in these studies. Studies in other Asia countries used WHO/ISH risk prediction charts and were aware of potential underestimation of CVD risk [34]. The earlier study in Vietnam may have overestimated CVD risk also because it only measured BP at one visit and only did blood tests on the urban sample. These limitations may present a challenge for both curative and preventive care in CVD management based on such data.

In our study, inclusion or exclusion of current CVD was found to affect the predicted future pattern of CVD risk differently in each model. The results from the CMCS and Asian models were comparable but results from the Framingham model differed significantly. None of these models have yet been validated in Vietnam. Previous studies have suggested that applying the Framingham model in Europe and Asia resulted in overestimates of the burden of disease, where there were lower rates of CVD [8, 13, 36-39]. The differences when using the three models can be explained by the different components in each model, such as characteristics of study population, input variables, outcome variables, time horizon, and values in each equation. The Asian model considers the mean age, SBP, total cholesterol and smoke/non-smoke; outcome is defined as cardiovascular death, non-fatal myocardial infarction, and non-fatal

Chapter 2

cerebrovascular events in the following eight years. The CMCS equation uses a subgroup of risk factors, no mean values as in the Asian equation, and outcome is defined as coronary death and myocardial infarction in the coming 10 years.

The implication of our study is that Vietnam should of course ideally have its own cohort to build up its own model to predict CVD risk. However, due to the long time needed for that, and due to resource constraints, it may be legitimate to use a model from countries with similar contextual factors, such as the CMCS or the Asian models, to predict patterns of CVD risk in the coming years as a basis for planning and management of services in Vietnam. Applying the Framingham model seems to provide deviating results from both other models and might be less appropriate for this specific setting.

There are limitations to our study. First, we could not confirm high BP using ambulatory blood pressure monitoring. However, measurements taken at two visits are considered acceptable for low resource settings [30]. Secondly, we identified cases with current CVD based only on medical history, clinical exam and ECG so we may have lost cases with CVD without clinical symptoms and signs in ECG. Finally, the CMCS, Asian and Framingham original models can be applied to general populations. In this study, we only tested them in hypertensive patients, which reflects a limitation of our study. Notably, this may reduce the generalizability of our findings.

Conclusion

In conclusion, the prevalence of high BP in a rural mountainous area of Vietnam was found to be lower than previously reported for Vietnam, but still sufficient to warrant attention from health services. Our results demonstrate that the Asian and CMCS models provided

40


similar outcomes in predicting CVD risk in Vietnamese population. Prediction of risk using the Asian and CMCS models is likely to be more appropriate than the Framingham for the Vietnamese context, as they are from the same region. Framingham provides vastly different results and might be less suitable for this populations. The results of this study are useful to health planners, and will provide a basic for a further study on cost-effectiveness modeling of CVD management.

Acknowledgements

This research was funded by a project called “Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Viet Nam”, financed by the Royal Netherlands Embassy in Hanoi. The authors thank Dr. Nguyen Tien Dung, Dr. Duong Danh Liem and Dr. Vu Tien Thang for their support for patient examinations at community health center and Dr. Hac Van Vinh for his communication with health services. Thanks also go to Trung Thanh, Cu Van, Tuc Tranh and La Hien for cooperation in data collection. We thank to Prof. Anushka Patel and Dr. Federica Barzi for providing specific information on the Asian model.

Chapter 2

References

1. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2095-128. Epub 2012/12/19. doi: S0140-6736(12)61728-0 [pii] 10.1016/S0140-6736(12)61728-0 [doi]. PubMed PMID: 23245604.

2. Hata J, Kiyohara Y. Epidemiology of stroke and coronary artery disease in Asia. Circ J. 2013;77(8):1923-32. Epub 2013/07/12. doi: DN/JST.JSTAGE/circj/CJ-13-0786 [pii]. PubMed PMID: 23842096.

3. WHO. The 10 leading causes of death in the world, 2000 and 2011 2013 [cited 2013 August 5]. Available from: http://who.int/mediacentre/factsheets/fs310/en/.

4. Nguyen Phuong Hoa CR, Damian G Hoy, Nguyen Duc Hinh, Nguyen Thi Kim Chuc & Duc Anh Ngo. Mortality measures from sample-based surveillance: evidence of the epidemiological transition in Viet Nam 2012 [cited 2013 August, 5]. Available from: http://www.who.int/bulletin/volumes/90/10/11-100750/en/.

5. Ngo AD, Rao C, Hoa NP, Adair T, Chuc NT. Mortality patterns in Vietnam, 2006: Findings from a national verbal autopsy survey. BMC Res Notes. 2010;3:78. Epub 2010/03/20. doi: 1756-0500-3-78 [pii] 10.1186/1756-0500-3-78 [doi]. PubMed PMID: 20236551; PubMed Central PMCID: PMC2851717.

Chap

ter 2

41


similar outcomes in predicting CVD risk in Vietnamese population. Prediction of risk using the Asian and CMCS models is likely to be more appropriate than the Framingham for the Vietnamese context, as they are from the same region. Framingham provides vastly different results and might be less suitable for this populations. The results of this study are useful to health planners, and will provide a basic for a further study on cost-effectiveness modeling of CVD management.

Acknowledgements

This research was funded by a project called “Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Viet Nam”, financed by the Royal Netherlands Embassy in Hanoi. The authors thank Dr. Nguyen Tien Dung, Dr. Duong Danh Liem and Dr. Vu Tien Thang for their support for patient examinations at community health center and Dr. Hac Van Vinh for his communication with health services. Thanks also go to Trung Thanh, Cu Van, Tuc Tranh and La Hien for cooperation in data collection. We thank to Prof. Anushka Patel and Dr. Federica Barzi for providing specific information on the Asian model.

Chapter 2

References

1. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2095-128. Epub 2012/12/19. doi: S0140-6736(12)61728-0 [pii] 10.1016/S0140-6736(12)61728-0 [doi]. PubMed PMID: 23245604.

2. Hata J, Kiyohara Y. Epidemiology of stroke and coronary artery disease in Asia. Circ J. 2013;77(8):1923-32. Epub 2013/07/12. doi: DN/JST.JSTAGE/circj/CJ-13-0786 [pii]. PubMed PMID: 23842096.

3. WHO. The 10 leading causes of death in the world, 2000 and 2011 2013 [cited 2013 August 5]. Available from: http://who.int/mediacentre/factsheets/fs310/en/.

4. Nguyen Phuong Hoa CR, Damian G Hoy, Nguyen Duc Hinh, Nguyen Thi Kim Chuc & Duc Anh Ngo. Mortality measures from sample-based surveillance: evidence of the epidemiological transition in Viet Nam 2012 [cited 2013 August, 5]. Available from: http://www.who.int/bulletin/volumes/90/10/11-100750/en/.


42


6. Hoang VM, Dao LH, Wall S, Nguyen TK, Byass P. Cardiovascular disease mortality and its association with socioeconomic status: findings from a population-based cohort study in rural Vietnam, 1999-2003. Prev Chronic Dis. 2006;3(3):A89. Epub 2006/06/17. doi: A89 [pii]. PubMed PMID: 16776890; PubMed Central PMCID: PMC1637797.

7. Huong DL, Van Minh H, Janlert U, Van do D, Byass P. Socio-economic status inequality and major causes of death in adults: a 5-year follow-up study in rural Vietnam. Public Health. 2006;120(6):497-504. Epub 2006/05/19. doi: S0033-3506(06)00068-0 [pii]10.1016/j.puhe.2006.03.003 [doi]. PubMed PMID: 16707143.

8. Liu J, Hong Y, D'Agostino RB, Sr., Wu Z, Wang W, Sun J, et al. Predictive value for the Chinese population of the Framingham CHD risk assessment tool compared with the Chinese Multi-Provincial Cohort Study. JAMA. 2004;291(21):2591-9. Epub 2004/06/03. doi: 10.1001/jama.291.21.2591 [doi] 291/21/2591 [pii]. PubMed PMID: 15173150.

9. Kanjilal S, Rao VS, Mukherjee M, Natesha BK, Renuka KS, Sibi K, et al. Application of cardiovascular disease risk prediction models and the relevance of novel biomarkers to risk stratification in Asian Indians. Vasc Health Risk Manag. 2008;4(1):199-211. Epub 2008/07/17. PubMed PMID: 18629375; PubMed Central PMCID: PMC2464770.

Chapter 2

10. Conroy RM, Pyorala K, Fitzgerald AP, Sans S, Menotti A, De Backer G, et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J. 2003;24(11):987-1003. Epub 2003/06/06. doi: S0195668X03001143 [pii]. PubMed PMID: 12788299.

11. Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, May M, Brindle P. Derivation and validation of QRISK, a new cardiovascular disease risk score for the United Kingdom: prospective open cohort study. BMJ. 2007;335(7611):136. Epub 2007/07/07. doi: bmj.39261.471806.55 [pii] 10.1136/bmj.39261.471806.55 [doi]. PubMed PMID: 17615182; PubMed Central PMCID: PMC1925200.

12. Assmann G, Cullen P, Schulte H. Simple scoring scheme for calculating the risk of acute coronary events based on the 10-year follow-up of the prospective cardiovascular Munster (PROCAM) study. Circulation. 2002;105(3):310-5. Epub 2002/01/24. PubMed PMID: 11804985.

13. Woodward M, Brindle P, Tunstall-Pedoe H. Adding social deprivation and family history to cardiovascular risk assessment: the ASSIGN score from the Scottish Heart Health Extended Cohort (SHHEC). Heart. 2007;93(2):172-6. Epub 2006/11/09. doi: hrt.2006.108167 [pii]10.1136/hrt.2006.108167 [doi]. PubMed PMID: 17090561; PubMed Central PMCID: PMC1861393.

14. Ferrario M, Chiodini P, Chambless LE, Cesana G, Vanuzzo D, Panico S, et al. Prediction of coronary events in a low incidence population. Assessing accuracy of the CUORE Cohort Study prediction equation. Int J Epidemiol. 2005;34(2):413-21. Epub 2005/01/22. doi: dyh405 [pii] 10.1093/ije/dyh405 [doi]. PubMed PMID: 15659467.

Chap

ter 2

43


6. Hoang VM, Dao LH, Wall S, Nguyen TK, Byass P. Cardiovascular disease mortality and its association with socioeconomic status: findings from a population-based cohort study in rural Vietnam, 1999-2003. Prev Chronic Dis. 2006;3(3):A89. Epub 2006/06/17. doi: A89 [pii]. PubMed PMID: 16776890; PubMed Central PMCID: PMC1637797.

7. Huong DL, Van Minh H, Janlert U, Van do D, Byass P. Socio-economic status inequality and major causes of death in adults: a 5-year follow-up study in rural Vietnam. Public Health. 2006;120(6):497-504. Epub 2006/05/19. doi: S0033-3506(06)00068-0 [pii]10.1016/j.puhe.2006.03.003 [doi]. PubMed PMID: 16707143.

8. Liu J, Hong Y, D'Agostino RB, Sr., Wu Z, Wang W, Sun J, et al. Predictive value for the Chinese population of the Framingham CHD risk assessment tool compared with the Chinese Multi-Provincial Cohort Study. JAMA. 2004;291(21):2591-9. Epub 2004/06/03. doi: 10.1001/jama.291.21.2591 [doi] 291/21/2591 [pii]. PubMed PMID: 15173150.

9. Kanjilal S, Rao VS, Mukherjee M, Natesha BK, Renuka KS, Sibi K, et al. Application of cardiovascular disease risk prediction models and the relevance of novel biomarkers to risk stratification in Asian Indians. Vasc Health Risk Manag. 2008;4(1):199-211. Epub 2008/07/17. PubMed PMID: 18629375; PubMed Central PMCID: PMC2464770.

Chapter 2

10. Conroy RM, Pyorala K, Fitzgerald AP, Sans S, Menotti A, De Backer G, et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J. 2003;24(11):987-1003. Epub 2003/06/06. doi: S0195668X03001143 [pii]. PubMed PMID: 12788299.

11. Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, May M, Brindle P. Derivation and validation of QRISK, a new cardiovascular disease risk score for the United Kingdom: prospective open cohort study. BMJ. 2007;335(7611):136. Epub 2007/07/07. doi: bmj.39261.471806.55 [pii] 10.1136/bmj.39261.471806.55 [doi]. PubMed PMID: 17615182; PubMed Central PMCID: PMC1925200.

12. Assmann G, Cullen P, Schulte H. Simple scoring scheme for calculating the risk of acute coronary events based on the 10-year follow-up of the prospective cardiovascular Munster (PROCAM) study. Circulation. 2002;105(3):310-5. Epub 2002/01/24. PubMed PMID: 11804985.

13. Woodward M, Brindle P, Tunstall-Pedoe H. Adding social deprivation and family history to cardiovascular risk assessment: the ASSIGN score from the Scottish Heart Health Extended Cohort (SHHEC). Heart. 2007;93(2):172-6. Epub 2006/11/09. doi: hrt.2006.108167 [pii]10.1136/hrt.2006.108167 [doi]. PubMed PMID: 17090561; PubMed Central PMCID: PMC1861393.

14. Ferrario M, Chiodini P, Chambless LE, Cesana G, Vanuzzo D, Panico S, et al. Prediction of coronary events in a low incidence population. Assessing accuracy of the CUORE Cohort Study prediction equation. Int J Epidemiol. 2005;34(2):413-21. Epub 2005/01/22. doi: dyh405 [pii] 10.1093/ije/dyh405 [doi]. PubMed PMID: 15659467.

44


15. Sritara P, Cheepudomwit S, Chapman N, Woodward M, Kositchaiwat C, Tunlayadechanont S, et al. Twelve-year changes in vascular risk factors and their associations with mortality in a cohort of 3499 Thais: the Electricity Generating Authority of Thailand Study. Int J Epidemiol. 2003;32(3):461-8. Epub 2003/06/05. PubMed PMID: 12777437.

16. Wu Y, Liu X, Li X, Li Y, Zhao L, Chen Z, et al. Estimation of 10-year risk of fatal and nonfatal ischemic cardiovascular diseases in Chinese adults. Circulation. 2006;114(21):2217-25.Epub 2006/11/08. doi: CIRCULATIONAHA.105.607499 [pii] 10.1161/CIRCULATIONAHA.105.607499 [doi]. PubMed PMID: 17088464.

17. Saito I, Sato S, Nakamura M, Kokubo Y, Mannami T, Adachi H, et al. A low level of C-reactive protein in Japanese adults and its association with cardiovascular risk factors: the Japan NCVC-Collaborative Inflammation Cohort (JNIC) study. Atherosclerosis. 2007;194(1):238-44. Epub 2006/09/12. doi: S0021-9150(06)00456-4 [pii] 10.1016/j.atherosclerosis.2006.07.032 [doi]. PubMed PMID: 16963054.

18. Arima H, Yonemoto K, Doi Y, Ninomiya T, Hata J, Tanizaki Y, et al. Development and validation of a cardiovascular risk prediction model for Japanese: the Hisayama study. Hypertens Res. 2009;32(12):1119-22. Epub 2009/09/19. doi: hr2009161 [pii] 10.1038/hr.2009.161 [doi]. PubMed PMID: 19763133.

19. Purwanto, Eswaran C, Logeswaran R, Abdul Rahman AR. Prediction models for early risk detection of cardiovascular event. J Med Syst. 2012;36(2):521-31. Epub 2012/06/08. PubMed PMID: 22675726.

Chapter 2

20. Baik I, Cho NH, Kim SH, Shin C. Dietary information improves cardiovascular disease risk prediction models. Eur J Clin Nutr. 2013;67(1):25-30. Epub 2012/11/15. doi: ejcn2012175 [pii] 10.1038/ejcn.2012.175 [doi]. PubMed PMID: 23149979.

21. Lee J, Heng D, Ma S, Chew SK, Hughes K, Tai ES. The metabolic syndrome and mortality: the Singapore Cardiovascular Cohort Study. Clin Endocrinol (Oxf).2008;69(2):225-30. Epub 2008/01/23. doi: CEN3174 [pii] 10.1111/j.1365-2265.2008.03174.x [doi]. PubMed PMID: 18208579.

22. Nguyen QN, Pham ST, Do LD, Nguyen VL, Wall S, Weinehall L, et al. Cardiovascular disease risk factor patterns and their implications for intervention strategies in Vietnam. Int J Hypertens. 2012;2012:560397. Epub 2012/04/14. doi: 10.1155/2012/560397 [doi]. PubMed PMID: 22500217; PubMed Central PMCID: PMC3303616.

23. Son PT, Quang NN, Viet NL, Khai PG, Wall S, Weinehall L, et al. Prevalence, awareness, treatment and control of hypertension in Vietnam-results from a national survey. J Hum Hypertens. 2012;26(4):268-80. Epub 2011/03/04. doi: jhh201118 [pii]10.1038/jhh.2011.18 [doi]. PubMed PMID: 21368775.

24. Organization WH. Country profile 2010 [cited 2013 August, 5]. Available from: http://www.who.int/countries/vnm/en/.

25. World_Health_Organization. Country profile 2012 [cited 2013 August, 5]. Available from: http://www.who.int/countries/chn/en/.

Chap

ter 2

45


15. Sritara P, Cheepudomwit S, Chapman N, Woodward M, Kositchaiwat C, Tunlayadechanont S, et al. Twelve-year changes in vascular risk factors and their associations with mortality in a cohort of 3499 Thais: the Electricity Generating Authority of Thailand Study. Int J Epidemiol. 2003;32(3):461-8. Epub 2003/06/05. PubMed PMID: 12777437.

16. Wu Y, Liu X, Li X, Li Y, Zhao L, Chen Z, et al. Estimation of 10-year risk of fatal and nonfatal ischemic cardiovascular diseases in Chinese adults. Circulation. 2006;114(21):2217-25.Epub 2006/11/08. doi: CIRCULATIONAHA.105.607499 [pii] 10.1161/CIRCULATIONAHA.105.607499 [doi]. PubMed PMID: 17088464.

17. Saito I, Sato S, Nakamura M, Kokubo Y, Mannami T, Adachi H, et al. A low level of C-reactive protein in Japanese adults and its association with cardiovascular risk factors: the Japan NCVC-Collaborative Inflammation Cohort (JNIC) study. Atherosclerosis. 2007;194(1):238-44. Epub 2006/09/12. doi: S0021-9150(06)00456-4 [pii] 10.1016/j.atherosclerosis.2006.07.032 [doi]. PubMed PMID: 16963054.

18. Arima H, Yonemoto K, Doi Y, Ninomiya T, Hata J, Tanizaki Y, et al. Development and validation of a cardiovascular risk prediction model for Japanese: the Hisayama study. Hypertens Res. 2009;32(12):1119-22. Epub 2009/09/19. doi: hr2009161 [pii] 10.1038/hr.2009.161 [doi]. PubMed PMID: 19763133.

19. Purwanto, Eswaran C, Logeswaran R, Abdul Rahman AR. Prediction models for early risk detection of cardiovascular event. J Med Syst. 2012;36(2):521-31. Epub 2012/06/08. PubMed PMID: 22675726.

Chapter 2

20. Baik I, Cho NH, Kim SH, Shin C. Dietary information improves cardiovascular disease risk prediction models. Eur J Clin Nutr. 2013;67(1):25-30. Epub 2012/11/15. doi: ejcn2012175 [pii] 10.1038/ejcn.2012.175 [doi]. PubMed PMID: 23149979.

21. Lee J, Heng D, Ma S, Chew SK, Hughes K, Tai ES. The metabolic syndrome and mortality: the Singapore Cardiovascular Cohort Study. Clin Endocrinol (Oxf).2008;69(2):225-30. Epub 2008/01/23. doi: CEN3174 [pii] 10.1111/j.1365-2265.2008.03174.x [doi]. PubMed PMID: 18208579.


23. Son PT, Quang NN, Viet NL, Khai PG, Wall S, Weinehall L, et al. Prevalence, awareness, treatment and control of hypertension in Vietnam-results from a national survey. J Hum Hypertens. 2012;26(4):268-80. Epub 2011/03/04. doi: jhh201118 [pii]10.1038/jhh.2011.18 [doi]. PubMed PMID: 21368775.

24. Organization WH. Country profile 2010 [cited 2013 August, 5]. Available from: http://www.who.int/countries/vnm/en/.

25. World_Health_Organization. Country profile 2012 [cited 2013 August, 5]. Available from: http://www.who.int/countries/chn/en/.

46


26. Framingham_heart_study. Hard Coronary Heart Disease (10-year risk) [cited 2013 August, 5]. Available from: http://www.framinghamheartstudy.org/risk-functions/coronary-heart-disease/hard-10-year-risk.php.

27. Barzi F, Patel A, Gu D, Sritara P, Lam TH, Rodgers A, et al. Cardiovascular risk prediction tools for populations in Asia. J Epidemiol Community Health. 2007;61(2):115-21. Epub 2007/01/20. doi: 61/2/115 [pii] 10.1136/jech.2005.044842 [doi]. PubMed PMID: 17234869; PubMed Central PMCID: PMC2465638.

28. Williams B, Poulter NR, Brown MJ, Davis M, McInnes GT, Potter JF, et al. Guidelines for management of hypertension: report of the fourth working party of the British Hypertension Society, 2004-BHS IV. J Hum Hypertens. 2004;18(3):139-85. Epub 2004/02/20. doi: 10.1038/sj.jhh.1001683 [doi] 1001683 [pii]. PubMed PMID: 14973512.

29. Parati G, Mendis S, Abegunde D, Asmar R, Mieke S, Murray A, et al. Recommendations for blood pressure measuring devices for office/clinic use in low resource settings. Blood Press Monit. 2005;10(1):3-10. Epub 2005/02/03. doi: 00126097-200502000-00002 [pii]. PubMed PMID: 15687867.

30. Bovet P, Gervasoni JP, Ross AG, Mkamba M, Mtasiwa DM, Lengeler C, et al. Assessing the prevalence of hypertension in populations: are we doing it right? J Hypertens. 2003;21(3):509-17. Epub 2003/03/18. doi: 10.1097/01.hjh.0000052473.40108.07 [doi]. PubMed PMID: 12640244.

Chapter 2

31. Cuddy ML. Treatment of hypertension: guidelines from JNC 7 (the seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure 1). J Pract Nurs. 2005;55(4):17-21; quiz 2-3.Epub 2006/03/04. PubMed PMID: 16512265.

32. Genuth S, Alberti KG, Bennett P, Buse J, Defronzo R, Kahn R, et al. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care. 2003;26(11):3160-7. Epub 2003/10/28. PubMed PMID: 14578255.

33. World Health Organization. Prevention of Cardiovascular Disease 2007 [cited 2013 August, 5]. Available from: http://www.who.int/cardiovascular_diseases/guidelines/PocketGL.ENGLISH.AFR-D-E.rev1.pdf.

34. Dugee O, Oum S, Buckley BS, Bonita R. Assessment of total cardiovascular risk using WHO/ISH risk prediction charts in three low and middle income countries in Asia. BMC Public Health. 2013;13(1):539. Epub 2013/06/06. doi: 1471-2458-13-539 [pii]10.1186/1471-2458-13-539 [doi]. PubMed PMID: 23734670; PubMed Central PMCID: PMC3679976.

35. Mendis S, Lindholm LH, Anderson SG, Alwan A, Koju R, Onwubere BJ, et al. Total cardiovascular risk approach to improve efficiency of cardiovascular prevention in resource constrain settings. J Clin Epidemiol. 2011;64(12):1451-62. Epub 2011/05/03. doi: S0895-4356(11)00050-3 [pii] 10.1016/j.jclinepi.2011.02.001 [doi]. PubMed PMID: 21530172.

Chap

ter 2

47


26. Framingham_heart_study. Hard Coronary Heart Disease (10-year risk) [cited 2013 August, 5]. Available from: http://www.framinghamheartstudy.org/risk-functions/coronary-heart-disease/hard-10-year-risk.php.

27. Barzi F, Patel A, Gu D, Sritara P, Lam TH, Rodgers A, et al. Cardiovascular risk prediction tools for populations in Asia. J Epidemiol Community Health. 2007;61(2):115-21. Epub 2007/01/20. doi: 61/2/115 [pii] 10.1136/jech.2005.044842 [doi]. PubMed PMID: 17234869; PubMed Central PMCID: PMC2465638.

28. Williams B, Poulter NR, Brown MJ, Davis M, McInnes GT, Potter JF, et al. Guidelines for management of hypertension: report of the fourth working party of the British Hypertension Society, 2004-BHS IV. J Hum Hypertens. 2004;18(3):139-85. Epub 2004/02/20. doi: 10.1038/sj.jhh.1001683 [doi] 1001683 [pii]. PubMed PMID: 14973512.

29. Parati G, Mendis S, Abegunde D, Asmar R, Mieke S, Murray A, et al. Recommendations for blood pressure measuring devices for office/clinic use in low resource settings. Blood Press Monit. 2005;10(1):3-10. Epub 2005/02/03. doi: 00126097-200502000-00002 [pii]. PubMed PMID: 15687867.

30. Bovet P, Gervasoni JP, Ross AG, Mkamba M, Mtasiwa DM, Lengeler C, et al. Assessing the prevalence of hypertension in populations: are we doing it right? J Hypertens. 2003;21(3):509-17. Epub 2003/03/18. doi: 10.1097/01.hjh.0000052473.40108.07 [doi]. PubMed PMID: 12640244.

Chapter 2

31. Cuddy ML. Treatment of hypertension: guidelines from JNC 7 (the seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure 1). J Pract Nurs. 2005;55(4):17-21; quiz 2-3.Epub 2006/03/04. PubMed PMID: 16512265.

32. Genuth S, Alberti KG, Bennett P, Buse J, Defronzo R, Kahn R, et al. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care. 2003;26(11):3160-7. Epub 2003/10/28. PubMed PMID: 14578255.

33. World Health Organization. Prevention of Cardiovascular Disease 2007 [cited 2013 August, 5]. Available from: http://www.who.int/cardiovascular_diseases/guidelines/PocketGL.ENGLISH.AFR-D-E.rev1.pdf.

34. Dugee O, Oum S, Buckley BS, Bonita R. Assessment of total cardiovascular risk using WHO/ISH risk prediction charts in three low and middle income countries in Asia. BMC Public Health. 2013;13(1):539. Epub 2013/06/06. doi: 1471-2458-13-539 [pii]10.1186/1471-2458-13-539 [doi]. PubMed PMID: 23734670; PubMed Central PMCID: PMC3679976.

35. Mendis S, Lindholm LH, Anderson SG, Alwan A, Koju R, Onwubere BJ, et al. Total cardiovascular risk approach to improve efficiency of cardiovascular prevention in resource constrain settings. J Clin Epidemiol. 2011;64(12):1451-62. Epub 2011/05/03. doi: S0895-4356(11)00050-3 [pii] 10.1016/j.jclinepi.2011.02.001 [doi]. PubMed PMID: 21530172.

48


36. Menotti A, Puddu PE, Lanti M. Comparison of the Framingham risk function-based coronary chart with risk function from an Italian population study. Eur Heart J. 2000;21(5):365-70. Epub 2000/02/10. doi: 10.1053/euhj.1999.1864 [doi] S0195668X9991864X [pii]. PubMed PMID: 10666350.

37. Brindle P, Emberson J, Lampe F, Walker M, Whincup P, Fahey T, et al. Predictive accuracy of the Framingham coronary risk score in British men: prospective cohort study. BMJ. 2003;327(7426):1267. Epub 2003/12/04. doi: 10.1136/bmj.327.7426.1267 [doi] 327/7426/1267 [pii]. PubMed PMID: 14644971; PubMed Central PMCID: PMC286248.

38. Hense HW, Schulte H, Lowel H, Assmann G, Keil U. Framingham risk function overestimates risk of coronary heart disease in men and women from Germany--results from the MONICA Augsburg and the PROCAM cohorts. Eur Heart J. 2003;24(10):937-45. Epub 2003/04/26. doi: S0195668X03000812 [pii]. PubMed PMID: 12714025.

39. Eichler K, Puhan MA, Steurer J, Bachmann LM. Prediction of first coronary events with the Framingham score: a systematic review. Am Heart J. 2007;153(5):722-31, 31 e1-8. Epub 2007/04/25. doi: S0002-8703(07)00162-7 [pii] 10.1016/j.ahj.2007.02.027 [doi]. PubMed PMID: 17452145.

Chap

ter 2

49


36. Menotti A, Puddu PE, Lanti M. Comparison of the Framingham risk function-based coronary chart with risk function from an Italian population study. Eur Heart J. 2000;21(5):365-70. Epub 2000/02/10. doi: 10.1053/euhj.1999.1864 [doi] S0195668X9991864X [pii]. PubMed PMID: 10666350.

37. Brindle P, Emberson J, Lampe F, Walker M, Whincup P, Fahey T, et al. Predictive accuracy of the Framingham coronary risk score in British men: prospective cohort study. BMJ. 2003;327(7426):1267. Epub 2003/12/04. doi: 10.1136/bmj.327.7426.1267 [doi] 327/7426/1267 [pii]. PubMed PMID: 14644971; PubMed Central PMCID: PMC286248.

38. Hense HW, Schulte H, Lowel H, Assmann G, Keil U. Framingham risk function overestimates risk of coronary heart disease in men and women from Germany--results from the MONICA Augsburg and the PROCAM cohorts. Eur Heart J. 2003;24(10):937-45. Epub 2003/04/26. doi: S0195668X03000812 [pii]. PubMed PMID: 12714025.

39. Eichler K, Puhan MA, Steurer J, Bachmann LM. Prediction of first coronary events with the Framingham score: a systematic review. Am Heart J. 2007;153(5):722-31, 31 e1-8. Epub 2007/04/25. doi: S0002-8703(07)00162-7 [pii] 10.1016/j.ahj.2007.02.027 [doi]. PubMed PMID: 17452145.

Chapter 3

Adherence to hypertension medication: quantitative and qualitative

investigations in a rural Northern Vietnamese community

Thi-Phuong-Lan Nguyen, C.C.M. Schuiling-Veninga,Thi-Bach-Yen Nguyen, Thu-Hang Vu, E. P. Wright, and M.J. Postma

ƠSubmitted for publication

52

Adherence to hypertension medication: quantitative and qualitative research

Abstract

Objectives

The purposes of this study were to assess the adherence to medication of hypertensive patients visiting community health stations in a rural area in Vietnam, to examine the relationship between levels of adherence and cardiovascular risk among hypertensive patients and to further understand factors influencing adherence.

Methods

This study is part of a prospective one-year study conducted on hypertension management in a population aged 35 to 64 years. Data on age, sex, blood pressure and blood test results were collected at baseline. Cardiovascular risk was based on the Cardiovascular Risk Prediction Model for populations in Asia. To calculate medication adherence, the number of days the drug was taken was divided by the number of days since the first day of the prescription. A threshold of 80% was applied to differentiate between adherence and non-adherence. In-depth interviews were conducted among 18 subjects, including subjects classified as adherent and as non-adherent.

Results

Among 315 patients analyzed, 49.8% of the patients were adherent. Qualitative investigation revealed discrepancies in classification of adherence and non-adherence based on quantitative analysis and interviews. No significant difference in medication adherence between two cardiovascular disease risk groups (<10% vs. >10% risk) was found, also not after controlling for age, sex, and ethnicity (adjusted odds ratio at 1.068; 95 % CI: 0.614 to 1.857). The odds of medication

Chapter 3

adherence in females was 1.531 times higher than in males but the difference was not statistically significant (95% CI: 0.957 to 2.448). Each one-year increase in age resulted in patients being 1.036 times more likely to be compliant (95% CI: 1.002 to 1.072). Awareness of complications related to hypertension was given as the main reason for adherence to therapy.

Conclusion

Medication adherence rate was relatively low among hypertensive subjects. The data suggest that rather than risk profile, the factor of age should be considered for guiding the choice on who to target for improving medication adherence.

Chap

ter 3

53


Abstract

Objectives

The purposes of this study were to assess the adherence to medication of hypertensive patients visiting community health stations in a rural area in Vietnam, to examine the relationship between levels of adherence and cardiovascular risk among hypertensive patients and to further understand factors influencing adherence.

Methods

This study is part of a prospective one-year study conducted on hypertension management in a population aged 35 to 64 years. Data on age, sex, blood pressure and blood test results were collected at baseline. Cardiovascular risk was based on the Cardiovascular Risk Prediction Model for populations in Asia. To calculate medication adherence, the number of days the drug was taken was divided by the number of days since the first day of the prescription. A threshold of 80% was applied to differentiate between adherence and non-adherence. In-depth interviews were conducted among 18 subjects, including subjects classified as adherent and as non-adherent.

Results

Among 315 patients analyzed, 49.8% of the patients were adherent. Qualitative investigation revealed discrepancies in classification of adherence and non-adherence based on quantitative analysis and interviews. No significant difference in medication adherence between two cardiovascular disease risk groups (<10% vs. >10% risk) was found, also not after controlling for age, sex, and ethnicity (adjusted odds ratio at 1.068; 95 % CI: 0.614 to 1.857). The odds of medication

Chapter 3

adherence in females was 1.531 times higher than in males but the difference was not statistically significant (95% CI: 0.957 to 2.448). Each one-year increase in age resulted in patients being 1.036 times more likely to be compliant (95% CI: 1.002 to 1.072). Awareness of complications related to hypertension was given as the main reason for adherence to therapy.

Conclusion

Medication adherence rate was relatively low among hypertensive subjects. The data suggest that rather than risk profile, the factor of age should be considered for guiding the choice on who to target for improving medication adherence.

54


Introduction

Many studies have suggested that a high level of adherence to antihypertensive drug treatment is related to better blood pressure (BP) control and a reduced risk of cardiovascular disease (CVD) [1-15]. Notably, however, there has been a lack of well-planned studies to identify any effects of adherence on clinical outcome, in particular in developing countries [16]. The scarce studies available so far suggest that adherence to antihypertensive medication is often relatively low. A meta-analysis on data of 376,162 patients showed an adherence to medication for preventing cardiovascular disease of only 57% [17]. Similarly, a study conducted in Italy showed that approximately 60% of the patients had a good-to-excellent adherence to antihypertensive medication [2], whereas in Poland only 26% of the cardiovascular patients used their drugs as prescribed [18]. Many Asian studies mimic this same trend of low medication adherence; the percentage of patients showing good adherence was 53% in Malaysia [8], 65% among Chinese populations [19], 55 % in Korea [20] and 66% in Vietnam [21].

Knowledge of factors that affect adherence could play an important role in the development of interventions to improve it. From both qualitative and quantitative studies described in the literature, many factors potentially affecting adherence to medication are known, including demographic, social and cognitive factors, interactions between health care providers and patients, health care system characteristics, the medication involved, and the general health profile of the patient [14, 17, 21-34]. For instance, one meta-analysis showed that patients with a history of CVD were more often adherent to medications such as aspirin, BP-lowering drugs and statins, compared to patients without previous CVD [17]. Another Polish study conducted at primary care level revealed that medication adherence

Chapter 3

was higher in patients at lower level of CVD risk, and that there was a weak correlation between CVD risk level and non-adherence to medication [18].

As there is a lack of evidence on the association between CVD risk and the adherence to antihypertensive medication in general and in developing countries such as Vietnam in particular, we conducted this study to (i) assess the level of adherence of hypertensive patients visiting community health stations (CHS) in a rural area in Vietnam; (ii) examine the relationship between level of adherence and cardiovascular risk among hypertensive patients; and (iii) get a better understanding of adherence and factors influencing adherence among these patients. Information on adherence is crucial for estimating the effectiveness of antihypertensive drugs, as opposed to efficacy from the clinical trials. Together with data on the costs of screening and treatment, these findings will provide input parameters for future modeling in a full-fledged cost-effectiveness study of community programs for the control of hypertension in Vietnam.

Methods

Study design and setting

This study was conducted in rural mountainous communes in the North of Vietnam among subjects aged from 35 to 64 years. The selection of study locations and the baseline surveys have been described in detail elsewhere [35]. Both quantitative and qualitative methods were applied in order to get both evidence on the level of medication adherence and factors potentially explaining adherence or non-adherence. Notably, the quantitative study used a prospective design with a 1-year time frame and the qualitative study involved an in-depth interview 18 months after the start of the baseline survey.

Chap

ter 3

55


Introduction

Many studies have suggested that a high level of adherence to antihypertensive drug treatment is related to better blood pressure (BP) control and a reduced risk of cardiovascular disease (CVD) [1-15]. Notably, however, there has been a lack of well-planned studies to identify any effects of adherence on clinical outcome, in particular in developing countries [16]. The scarce studies available so far suggest that adherence to antihypertensive medication is often relatively low. A meta-analysis on data of 376,162 patients showed an adherence to medication for preventing cardiovascular disease of only 57% [17]. Similarly, a study conducted in Italy showed that approximately 60% of the patients had a good-to-excellent adherence to antihypertensive medication [2], whereas in Poland only 26% of the cardiovascular patients used their drugs as prescribed [18]. Many Asian studies mimic this same trend of low medication adherence; the percentage of patients showing good adherence was 53% in Malaysia [8], 65% among Chinese populations [19], 55 % in Korea [20] and 66% in Vietnam [21].

Knowledge of factors that affect adherence could play an important role in the development of interventions to improve it. From both qualitative and quantitative studies described in the literature, many factors potentially affecting adherence to medication are known, including demographic, social and cognitive factors, interactions between health care providers and patients, health care system characteristics, the medication involved, and the general health profile of the patient [14, 17, 21-34]. For instance, one meta-analysis showed that patients with a history of CVD were more often adherent to medications such as aspirin, BP-lowering drugs and statins, compared to patients without previous CVD [17]. Another Polish study conducted at primary care level revealed that medication adherence

Chapter 3

was higher in patients at lower level of CVD risk, and that there was a weak correlation between CVD risk level and non-adherence to medication [18].

As there is a lack of evidence on the association between CVD risk and the adherence to antihypertensive medication in general and in developing countries such as Vietnam in particular, we conducted this study to (i) assess the level of adherence of hypertensive patients visiting community health stations (CHS) in a rural area in Vietnam; (ii) examine the relationship between level of adherence and cardiovascular risk among hypertensive patients; and (iii) get a better understanding of adherence and factors influencing adherence among these patients. Information on adherence is crucial for estimating the effectiveness of antihypertensive drugs, as opposed to efficacy from the clinical trials. Together with data on the costs of screening and treatment, these findings will provide input parameters for future modeling in a full-fledged cost-effectiveness study of community programs for the control of hypertension in Vietnam.

Methods


This study was conducted in rural mountainous communes in the North of Vietnam among subjects aged from 35 to 64 years. The selection of study locations and the baseline surveys have been described in detail elsewhere [35]. Both quantitative and qualitative methods were applied in order to get both evidence on the level of medication adherence and factors potentially explaining adherence or non-adherence. Notably, the quantitative study used a prospective design with a 1-year time frame and the qualitative study involved an in-depth interview 18 months after the start of the baseline survey.

56


The quantitative study included the data on medication use and adherence and patient characteristics potentially associated with adherence. The in-depth interviews allowed us to get further understanding of factors influencing the level of adherence. Factors that have been suggested in the literature to influence adherence are listed in Table 1 [14, 17, 21-34, 36].

Chapter 3

Table 1. Factors potentially affecting medication adherence reportedin the literature

Group FactorsDemographic factors

AgeSexRaceEducation level

Social factorsSocial economic statusSocial support

Cognitive factorsUnderstanding of cause and effect of hypertensionAwareness of hypertension riskAwareness of BP target and medication indication, forgetfulness and self-efficacy and sensing timing to take medication

Health care system characteristics

Communication between providers and patientsFrequency of visits to health-care providersAvailability of spare time to see doctorQuality of communication when in the office


Health insuranceHealth care system typeProviders’ typology

Medication involvedInclusive drug classMultiple/single dosageComplexity of regimenPotential and actual side-effectsShortages of drugsTotal number of pills per day

General health profile of the patient

History of cardiovascular diseaseComorbidityDepressionExact BP levelPossible symptoms of hypertensionQuality of life

Chap

ter 3

57


The quantitative study included the data on medication use and adherence and patient characteristics potentially associated with adherence. The in-depth interviews allowed us to get further understanding of factors influencing the level of adherence. Factors that have been suggested in the literature to influence adherence are listed in Table 1 [14, 17, 21-34, 36].

Chapter 3

Table 1. Factors potentially affecting medication adherence reportedin the literature

Group FactorsDemographic factors

AgeSexRaceEducation level

Social factorsSocial economic statusSocial support

Cognitive factorsUnderstanding of cause and effect of hypertensionAwareness of hypertension riskAwareness of BP target and medication indication, forgetfulness and self-efficacy and sensing timing to take medication


Communication between providers and patientsFrequency of visits to health-care providersAvailability of spare time to see doctorQuality of communication when in the office


Health insuranceHealth care system typeProviders’ typology

Medication involvedInclusive drug classMultiple/single dosageComplexity of regimenPotential and actual side-effectsShortages of drugsTotal number of pills per day

General health profile of the patient

History of cardiovascular diseaseComorbidityDepressionExact BP levelPossible symptoms of hypertensionQuality of life

58


Quantitative study

Patient selection

Patient selection at the time of the baseline survey has been described in detail elsewhere [35]. We randomly selected enough villages to cover 45 to 50% of the population in each commune, and one commune per district among four districts in Thai Nguyen province. In each location, we measured the blood pressure of all residents from 35 to 64 years old to identify subjects with hypertension [35]. After the baseline survey, subjects were either advised to manage their BP at a CHS or referred to second-line healthcare services. The latter choice was made if the local physician could not clearly identify acardiovascular issue or suspected a potentially serious health condition better treated in hospital. We followed up those subjects who met the following inclusion and exclusion criteria. The main inclusion criteria were that subjects must have a medication prescription for at least onemonth and had at least 90 days of follow up since the first prescription. Exclusion criteria were: (i) history of myocardial infarction or other serious heart disease(s), or any heart diseases which need to be treated in second-line facilities; (ii) referral to second-line if, despite strictly following the prescribed regimen, BP was inadequately controlled or organ damage was suspected; (iii) referral to second-line because patients requested it, generally thinking that their hypertension would be better managed there; (iv) patients had moved to another place to live; (v) patients no longer needed to take anti-hypertensive drugs; and (vi) patients missed getting a prescription for two months or more between two doses, because their BP had decreased to below 140 mm Hg during that time so their physicians decided they could stop medication.

Chapter 3

Follow-up

Medical doctors at participating CHSs attended a 3-day training program in hypertension management conducted by a cardiologist from Thai Nguyen University of Medicine and Pharmacy. Subjects managed at the CHS were advised to visit the CHS every month to check their BP and receive anti-hypertensive drugs if needed. Subjectswere also advised to visit the CHS more frequently if monitoring of BP or symptoms was needed, especially in cases where the types of medicine were changed. Subjects were advised to quit smoking, reduce salt intake, stop drinking alcohol, and of course, to take the drugs as prescribed. We followed subjects for one year after the date of registering at the CHS and recorded their BP and drug prescriptions at every visit.

Variables and measurements

In a previous paper, we suggested that the Asian or Chinese risk models could be used to predict CVD risk in South-East Asia [35]. In this study, we applied the Asian model because it was developed using data from six cohorts in Asia and therefore likely more representative for non-Chinese Asian countries, such as Vietnam. Input data for this model were collected at baseline, including systolic BP, cholesterol level, smoking status, age and sex of subjects. Subsequently, we divided the subjects into two groups: those having less than 10% and those having more than 10% risk of CVD in the coming eight yearsaccording to the Asian risk model. Details on measurements of all variables were described in a previous report [35].

To assess the level of adherence to antihypertensive medication, the number of pill-days covered was calculated; i.e., the number of days the drug was taken divided by the number of days since the first day of prescribing. A threshold of 80% was applied to differentiate between adherence and non-adherence [37, 38].

Chap

ter 3

59


Quantitative study

Patient selection

Patient selection at the time of the baseline survey has been described in detail elsewhere [35]. We randomly selected enough villages to cover 45 to 50% of the population in each commune, and one commune per district among four districts in Thai Nguyen province. In each location, we measured the blood pressure of all residents from 35 to 64 years old to identify subjects with hypertension [35]. After the baseline survey, subjects were either advised to manage their BP at a CHS or referred to second-line healthcare services. The latter choice was made if the local physician could not clearly identify acardiovascular issue or suspected a potentially serious health condition better treated in hospital. We followed up those subjects who met the following inclusion and exclusion criteria. The main inclusion criteria were that subjects must have a medication prescription for at least onemonth and had at least 90 days of follow up since the first prescription. Exclusion criteria were: (i) history of myocardial infarction or other serious heart disease(s), or any heart diseases which need to be treated in second-line facilities; (ii) referral to second-line if, despite strictly following the prescribed regimen, BP was inadequately controlled or organ damage was suspected; (iii) referral to second-line because patients requested it, generally thinking that their hypertension would be better managed there; (iv) patients had moved to another place to live; (v) patients no longer needed to take anti-hypertensive drugs; and (vi) patients missed getting a prescription for two months or more between two doses, because their BP had decreased to below 140 mm Hg during that time so their physicians decided they could stop medication.

Chapter 3

Follow-up

Medical doctors at participating CHSs attended a 3-day training program in hypertension management conducted by a cardiologist from Thai Nguyen University of Medicine and Pharmacy. Subjects managed at the CHS were advised to visit the CHS every month to check their BP and receive anti-hypertensive drugs if needed. Subjectswere also advised to visit the CHS more frequently if monitoring of BP or symptoms was needed, especially in cases where the types of medicine were changed. Subjects were advised to quit smoking, reduce salt intake, stop drinking alcohol, and of course, to take the drugs as prescribed. We followed subjects for one year after the date of registering at the CHS and recorded their BP and drug prescriptions at every visit.

Variables and measurements

In a previous paper, we suggested that the Asian or Chinese risk models could be used to predict CVD risk in South-East Asia [35]. In this study, we applied the Asian model because it was developed using data from six cohorts in Asia and therefore likely more representative for non-Chinese Asian countries, such as Vietnam. Input data for this model were collected at baseline, including systolic BP, cholesterol level, smoking status, age and sex of subjects. Subsequently, we divided the subjects into two groups: those having less than 10% and those having more than 10% risk of CVD in the coming eight yearsaccording to the Asian risk model. Details on measurements of all variables were described in a previous report [35].

To assess the level of adherence to antihypertensive medication, the number of pill-days covered was calculated; i.e., the number of days the drug was taken divided by the number of days since the first day of prescribing. A threshold of 80% was applied to differentiate between adherence and non-adherence [37, 38].

60


We used Chi square and Wilcoxon tests to examine whether adherent and non-adherent patients differed with regard to their CVD risk, sex and age. We then used logistic regression to investigate whether adherence could be predicted by CVD risk, sex, age and ethnicity. Continuous variables were presented as means ± standard deviation (SD) and non-continuous variables were presented as percentages.

Qualitative study

For the qualitative study, we used purposive sampling to select both adherent and non-adherent subjects from two different communes. Information on adherence came from the quantitative study. The number of subjects interviewed was determined by the results of the interviews and decided by the two main researchers, based on all notes after each interview. New subjects were interviewed until the point of saturation, when no extra information was collected with additional interviews.

Three interviewers were trained by the primary researcher. The face-to-face interviews lasted 30 minutes and were conducted using a semi-structured questionnaire, in a private room at the CHS. Each interview was recorded and later transcribed. The main topics of the interview included: (i) drugs used; (ii) reasons for (non-)adherence; (iii) side effects; and (iv) information received from physicians at baseline. The in-depth interview guideline is presented in full in Appendix 1.

For data analysis, two researchers read all 18 transcripts to understand the whole situation and subsequently used the extraction method to identify and group meaningful statements according to the main topics. Adherence in the qualitative part was considered as missing no more than one day per week on a regular basis, according to self-reporting during the interview. Non-adherent subjects were those who reported that they had stopped taking their medication, or took the

Chapter 3

medicines less than 20 days per month, or forgot their medicines on average two or more days per week.

Human subjects and ethical issue clearance

The research proposal was approved by the Institutional Review Board in biomedical research in the Institute of Social and Medical Studies in Hanoi, Vietnam. Written informed consent was collected at the time of the baseline survey. Study subjects received a small compensation for their time at every visit (25,000 VND, about 1.20 USD).

Results

Quantitative results

As shown in Fig 1, 338 subjects entered into the one-year prospective study. Table 2 presents the characteristics of the subjects: 54% males and 82% ethnically Kinh (the majority ethnic group in Vietnam). The mean age was 53.7 years and the CVD risk - measured according to the Asian risk model - was less than 10% in the coming eight years for 76% of the subjects.

Chap

ter 3

61


We used Chi square and Wilcoxon tests to examine whether adherent and non-adherent patients differed with regard to their CVD risk, sex and age. We then used logistic regression to investigate whether adherence could be predicted by CVD risk, sex, age and ethnicity. Continuous variables were presented as means ± standard deviation (SD) and non-continuous variables were presented as percentages.

Qualitative study

For the qualitative study, we used purposive sampling to select both adherent and non-adherent subjects from two different communes. Information on adherence came from the quantitative study. The number of subjects interviewed was determined by the results of the interviews and decided by the two main researchers, based on all notes after each interview. New subjects were interviewed until the point of saturation, when no extra information was collected with additional interviews.

Three interviewers were trained by the primary researcher. The face-to-face interviews lasted 30 minutes and were conducted using a semi-structured questionnaire, in a private room at the CHS. Each interview was recorded and later transcribed. The main topics of the interview included: (i) drugs used; (ii) reasons for (non-)adherence; (iii) side effects; and (iv) information received from physicians at baseline. The in-depth interview guideline is presented in full in Appendix 1.

For data analysis, two researchers read all 18 transcripts to understand the whole situation and subsequently used the extraction method to identify and group meaningful statements according to the main topics. Adherence in the qualitative part was considered as missing no more than one day per week on a regular basis, according to self-reporting during the interview. Non-adherent subjects were those who reported that they had stopped taking their medication, or took the

Chapter 3

medicines less than 20 days per month, or forgot their medicines on average two or more days per week.

Human subjects and ethical issue clearance

The research proposal was approved by the Institutional Review Board in biomedical research in the Institute of Social and Medical Studies in Hanoi, Vietnam. Written informed consent was collected at the time of the baseline survey. Study subjects received a small compensation for their time at every visit (25,000 VND, about 1.20 USD).

Results

Quantitative results

As shown in Fig 1, 338 subjects entered into the one-year prospective study. Table 2 presents the characteristics of the subjects: 54% males and 82% ethnically Kinh (the majority ethnic group in Vietnam). The mean age was 53.7 years and the CVD risk - measured according to the Asian risk model - was less than 10% in the coming eight years for 76% of the subjects.

62


Fig 1: Flow chart of selecting patients after baseline survey


Hypertension during the first measurement: 776

subjects

Hypertension during the second measurement:

466 subjects

Blood samples: 442 subjects


measurement

10 subjects not followed up and 300 subjects

without hypertension in the second measurement


No need for medication prescription: 55 subjects

Refer to higher level: 7 subjects

Moved to other location: 3 subjects

Died during follow up:3 subjects

Drop out after baseline survey: 36 subjects

Prescription less than 30 days: 3 subjects; Following up less than 90 days since the first prescription: 5 subjects; Patients missed gettinga prescription for two months or more between two doses, because their BP had decreased to below 140 mm Hg during that time so their doctors decided they could stop medication: 15 subjects. In total: 23 subjects

Entering the study: 315 subjects

Chapter 3

Table 2. Characteristics of patients (n=315)

Characteristic of patients n (%)

Sex

Female 144 (45.7)

Male 171 (54.3)

Ethnic

Kinh 259 (82.2)

Others 56 (17.8)

% CVD risk

≥10% 75 (23.8)

<10% 240 (76.2)

Mean age (years +/- SD) 53.7 +/- 6.95

In the CHS setting, 49.8% of the patients were adherent to their antihypertensive medication. As presented in Table 3, age did differ significantly between adherent and non-adherent subjects. Each one year increase in age resulted in subjects being 1.04 times more likely to be adherent (95% CI from: 1.002 to 1.072; p = 0.04). We found no association between CVD risk (<10% vs. ≥ 10% risk) and being adherent (odds ratio at 1.07; 95 % CI from 0.61 to 1.86; p = 0.81). Also, females were 1.53 times more likely to be adherent than males but the difference was not significant (95% CI from 0.96 to 2.45; p = 0.07).

Chap

ter 3

63


Fig 1: Flow chart of selecting patients after baseline survey


Hypertension during the first measurement: 776

subjects

Hypertension during the second measurement:

466 subjects

Blood samples: 442 subjects


measurement

10 subjects not followed up and 300 subjects

without hypertension in the second measurement


No need for medication prescription: 55 subjects

Refer to higher level: 7 subjects

Moved to other location: 3 subjects

Died during follow up:3 subjects

Drop out after baseline survey: 36 subjects

Prescription less than 30 days: 3 subjects; Following up less than 90 days since the first prescription: 5 subjects; Patients missed gettinga prescription for two months or more between two doses, because their BP had decreased to below 140 mm Hg during that time so their doctors decided they could stop medication: 15 subjects. In total: 23 subjects

Entering the study: 315 subjects

Chapter 3

Table 2. Characteristics of patients (n=315)

Characteristic of patients n (%)

Sex

Female 144 (45.7)

Male 171 (54.3)

Ethnic

Kinh 259 (82.2)

Others 56 (17.8)

% CVD risk

≥10% 75 (23.8)

<10% 240 (76.2)

Mean age (years +/- SD) 53.7 +/- 6.95

In the CHS setting, 49.8% of the patients were adherent to their antihypertensive medication. As presented in Table 3, age did differ significantly between adherent and non-adherent subjects. Each one year increase in age resulted in subjects being 1.04 times more likely to be adherent (95% CI from: 1.002 to 1.072; p = 0.04). We found no association between CVD risk (<10% vs. ≥ 10% risk) and being adherent (odds ratio at 1.07; 95 % CI from 0.61 to 1.86; p = 0.81). Also, females were 1.53 times more likely to be adherent than males but the difference was not significant (95% CI from 0.96 to 2.45; p = 0.07).

64

Tab

le 3

. A

ssoc

iatio

ns (C

OR

, AO

R*,

CI,

p) o

f (no

n)ad

here

nce

and

patie

nt's

char

acte

ristic

Para

met

ers

Adh

eren

cen

(%)

Non

-adh

eren

cen

(%)

Adh

eren

ce v

s. no

n-ad

here

nce

Cru

de a

naly

sis

Adj

uste

d an

alys

isC

OR

(CI 9

5%)

PA

OR

(CI 9

5%)

pC

VD

ris

k

CV

D ri

sk <

10%

(ref

)11

9(4

9.6)

121

(50.

4)1

1

CV

D ri

sk ≥

10%

38(5

0.7)

37(4

9.3)

1.04

4(0

.622

-1.7

54)

0.89

1.06

8(0

.614

–1.

857)

0.81

5

Sex

Mal

e (r

ef.)

77(4

5.0)

94(5

5.0)

11

Fem

ale

80(5

5.6)

64(4

4.4)

1.52

6(0

.977

-2.3

83)

0.07

11.

531

(0.9

57–

2.44

8)0.

076

Eth

nic

Kin

h (r

ef)

136

(52.

5)12

3(4

7.5)

11

Oth

er21

(37.

5)35

(62.

5)0.

584

(0.3

2-1.

067)

0.09

0.59

9(0

.328

–1.

095)

0.09

6

Age

(yea

rs)

54.6

52.8

0.02

1.03

6(1

.002

-1.0

72)

0.03

6

Not

e: B

inar

y lo

gist

ic re

gres

sion

, adj

ustin

g fo

r the

oth

er fa

ctor

s sho

wn

in th

e ta

ble

(CV

D ri

sk, s

ex, e

thni

c, a

ge).

CO

R: c

rude

odd

ratio

; AO

R:

Adj

uste

d od

d ra

tio; C

I: co

nfid

ent i

nter

val.

Chapter 3

Qualitative results

To deepen our understanding of the background of adherence versus non-adherence, we interviewed 18 subjects (12 males), of whom 11 had been identified as adherent based on the criteria and data in the quantitative study.

Interestingly, a range of reasons for adherence and non-adherence were mentioned. Being aware of the complications of high BP or experience within the family with complications was mentioned as a factor enhancing good adherence. Most of the occasions when patients did not take the medicine were related to forgetting, when they were busy or changed their daily activity pattern; only two of 12 actively decided not to take the pills because they felt better. For example, one man said: “I reduced the number of pills... and I did not see different results”. Side effects appeared as another influential factor; six subjects had side-effects during the treatment, such as cough, headache, nausea or fatigue, which might have led to interruptions in taking medicines. Subjects explicitly recounted: “After taking the first pills, I felt serious headaches…. so I stopped taking the medicines,….no one advised me,....I waited until next month to see doctor then she changed medicine for me” or “I got headache....then I changed the medicine”.

Notably, the adherence classification from the quantitative results, based on follow-up data, was sometimes inconsistent with the information reported during interviews. Two of the 11 quantitatively adherent subjects reported periods of non-adherence, while four of the seven quantitatively non-adherent subjects could be classified as adherent based on the interviews. Five of the 11 adherent subjects stated in the interview that they had never forgotten to take the medicines, only maybe sometimes changed from morning intake to afternoon intake. Of the 12 subjects who mentioned having forgotten

Chap

ter 3

65

Tab

le 3

. A

ssoc

iatio

ns (C

OR

, AO

R*,

CI,

p) o

f (no

n)ad

here

nce

and

patie

nt's

char

acte

ristic

Para

met

ers

Adh

eren

cen

(%)

Non

-adh

eren

cen

(%)

Adh

eren

ce v

s. no

n-ad

here

nce

Cru

de a

naly

sis

Adj

uste

d an

alys

isC

OR

(CI 9

5%)

PA

OR

(CI 9

5%)

pC

VD

ris

k

CV

D ri

sk <

10%

(ref

)11

9(4

9.6)

121

(50.

4)1

1

CV

D ri

sk ≥

10%

38(5

0.7)

37(4

9.3)

1.04

4(0

.622

-1.7

54)

0.89

1.06

8(0

.614

–1.

857)

0.81

5

Sex

Mal

e (r

ef.)

77(4

5.0)

94(5

5.0)

11

Fem

ale

80(5

5.6)

64(4

4.4)

1.52

6(0

.977

-2.3

83)

0.07

11.

531

(0.9

57–

2.44

8)0.

076

Eth

nic

Kin

h (r

ef)

136

(52.

5)12

3(4

7.5)

11

Oth

er21

(37.

5)35

(62.

5)0.

584

(0.3

2-1.

067)

0.09

0.59

9(0

.328

–1.

095)

0.09

6

Age

(yea

rs)

54.6

52.8

0.02

1.03

6(1

.002

-1.0

72)

0.03

6

Not

e: B

inar

y lo

gist

ic re

gres

sion

, adj

ustin

g fo

r the

oth

er fa

ctor

s sho

wn

in th

e ta

ble

(CV

D ri

sk, s

ex, e

thni

c, a

ge).

CO

R: c

rude

odd

ratio

; AO

R:

Adj

uste

d od

d ra

tio; C

I: co

nfid

ent i

nter

val.

Chapter 3

Qualitative results

To deepen our understanding of the background of adherence versus non-adherence, we interviewed 18 subjects (12 males), of whom 11 had been identified as adherent based on the criteria and data in the quantitative study.

Interestingly, a range of reasons for adherence and non-adherence were mentioned. Being aware of the complications of high BP or experience within the family with complications was mentioned as a factor enhancing good adherence. Most of the occasions when patients did not take the medicine were related to forgetting, when they were busy or changed their daily activity pattern; only two of 12 actively decided not to take the pills because they felt better. For example, one man said: “I reduced the number of pills... and I did not see different results”. Side effects appeared as another influential factor; six subjects had side-effects during the treatment, such as cough, headache, nausea or fatigue, which might have led to interruptions in taking medicines. Subjects explicitly recounted: “After taking the first pills, I felt serious headaches…. so I stopped taking the medicines,….no one advised me,....I waited until next month to see doctor then she changed medicine for me” or “I got headache....then I changed the medicine”.

Notably, the adherence classification from the quantitative results, based on follow-up data, was sometimes inconsistent with the information reported during interviews. Two of the 11 quantitatively adherent subjects reported periods of non-adherence, while four of the seven quantitatively non-adherent subjects could be classified as adherent based on the interviews. Five of the 11 adherent subjects stated in the interview that they had never forgotten to take the medicines, only maybe sometimes changed from morning intake to afternoon intake. Of the 12 subjects who mentioned having forgotten

66


to take the medicines, two took the medicine still within a day but at a different time, while the others did not take the drug at all that day. It was interesting that two subjects classified as non-adherent by the follow-up quantitative study, reported in interviews that during some months that they didn’t get medicines from the CHS, they bought the medicines themselves (this is possible in the open market for pharmaceuticals in Vietnam), so they were actually adherent after all.

Discussion

The results of our quantitative study showed that only 50% of the hypertensive patients managed at CHS in Vietnam were adherent to their medication, when we applied the 80% cut-off point for pill-days covered. There was no significant difference in adherence between patients with a high or low risk for CVD. However, adherence seemed to be influenced by age, as older patients used their medication more often in accordance with the doctor’s advice.

Adherence to hypertensive medication found in our study is similar to other studies and to results from a meta-analysis of adherence in selected databases, ranging from 26 to 78% [28, 33, 39, 40]. In our case, village health workers sometimes reminded subjects to visit the CHS for their check-up, which may have contributed to a relatively high level of adherence. A previous quantitative study in Vietnam also considered adherence of subjects during a 17 months study period. However the adherence was measured by numbers of check-ups and the threshold applied was appearance for at least one check-up per one or two months [21], which could have led to somewhat different interpretations and results.

In our study, women appeared to have a higher medication adherence than men, although not statistically significant. In a study in rural

Chapter 3

Bangladesh, women had higher medication adherence than men [41],while the opposite was found in Korea and Taiwan [20, 28].Differences might be explained by differing ways to measure adherence [28] and by the different settings (hospital vs. rural communities). Increasing age was also associated with better medication adherence in our study, which is similar to previous studies on that issue [20, 28, 40-42].

The level of CVD risk has been reported as a factor influencing adherence with antihypertensive and lipid-lowering therapy [43]. We could not confirm this phenomenon in our quantitative study. The disagreement between these studies might be elucidated by differing ways to clarify CVD risk. For example, the previous study classified CVD into three levels: (i) angina or coronary angiography, (ii) coronary artery bypass graft, percutaneous transluminal coronary angioplasty or history of coronary heart disease and (iii) acute or prior myocardial infarction [43], whereas in this study we used the Asian model for risk estimation, which allows us to measure the 8-year-risk. Patients were divided into only two groups (<10% vs. ≥10% risk) [44]. In our study, physicians did not explicitly inform their patients about their level of CVD risk at baseline. Not knowing about their own risk level could explain why we did not find an association between level of CVD risk and adherence.

The major strength of this study is the combination of qualitative insights and quantitative analyses. Therefore, estimation of medication adherence could be based both on medical records and on data from in-depth interviews within a selected sample.

A small sample was interviewed in depth to obtain more detailed information about their adherence and the factors influencing whether they took their drugs as prescribed. The findings from the in-depth interviews indicate similar issues as were detected in previous studies

Chap

ter 3

67


to take the medicines, two took the medicine still within a day but at a different time, while the others did not take the drug at all that day. It was interesting that two subjects classified as non-adherent by the follow-up quantitative study, reported in interviews that during some months that they didn’t get medicines from the CHS, they bought the medicines themselves (this is possible in the open market for pharmaceuticals in Vietnam), so they were actually adherent after all.

Discussion

The results of our quantitative study showed that only 50% of the hypertensive patients managed at CHS in Vietnam were adherent to their medication, when we applied the 80% cut-off point for pill-days covered. There was no significant difference in adherence between patients with a high or low risk for CVD. However, adherence seemed to be influenced by age, as older patients used their medication more often in accordance with the doctor’s advice.

Adherence to hypertensive medication found in our study is similar to other studies and to results from a meta-analysis of adherence in selected databases, ranging from 26 to 78% [28, 33, 39, 40]. In our case, village health workers sometimes reminded subjects to visit the CHS for their check-up, which may have contributed to a relatively high level of adherence. A previous quantitative study in Vietnam also considered adherence of subjects during a 17 months study period. However the adherence was measured by numbers of check-ups and the threshold applied was appearance for at least one check-up per one or two months [21], which could have led to somewhat different interpretations and results.

In our study, women appeared to have a higher medication adherence than men, although not statistically significant. In a study in rural

Chapter 3

Bangladesh, women had higher medication adherence than men [41],while the opposite was found in Korea and Taiwan [20, 28].Differences might be explained by differing ways to measure adherence [28] and by the different settings (hospital vs. rural communities). Increasing age was also associated with better medication adherence in our study, which is similar to previous studies on that issue [20, 28, 40-42].

The level of CVD risk has been reported as a factor influencing adherence with antihypertensive and lipid-lowering therapy [43]. We could not confirm this phenomenon in our quantitative study. The disagreement between these studies might be elucidated by differing ways to clarify CVD risk. For example, the previous study classified CVD into three levels: (i) angina or coronary angiography, (ii) coronary artery bypass graft, percutaneous transluminal coronary angioplasty or history of coronary heart disease and (iii) acute or prior myocardial infarction [43], whereas in this study we used the Asian model for risk estimation, which allows us to measure the 8-year-risk. Patients were divided into only two groups (<10% vs. ≥10% risk) [44]. In our study, physicians did not explicitly inform their patients about their level of CVD risk at baseline. Not knowing about their own risk level could explain why we did not find an association between level of CVD risk and adherence.

The major strength of this study is the combination of qualitative insights and quantitative analyses. Therefore, estimation of medication adherence could be based both on medical records and on data from in-depth interviews within a selected sample.

A small sample was interviewed in depth to obtain more detailed information about their adherence and the factors influencing whether they took their drugs as prescribed. The findings from the in-depth interviews indicate similar issues as were detected in previous studies

68


[45-47], suggesting that (non-) adherence is related to awareness of risks for complications, to the presence of and wish to deal with side effects, and to absence of symptoms of high BP (“feeling healthy”). All adherent subjects mentioned that they were motivated to adhere because they were aware of potential complications. This illustrates how the qualitative study adds details to results from the quantitative part.

Adherence of about one third of the 18 interviewed patients showed discrepancies compared to the qualitative data results. Two patients, initially classified as quantitatively adherent, were found to be non-adherent, while four patients initially thought to be non-adherent turned out to be adherent. Furthermore, the qualitative results suggested that one main reason for subjects to adhere to therapy was their awareness of the seriousness of hypertension complications. These results show the importance of understanding how patients decide on and exhibit adherence to medication, and demonstrate that it is insufficient to assume that hypertension patients do follow the prescription, as has sometimes been suggested [48].

The discrepancy between qualitative and quantitative data in classifying patients as adherent or not suggests that researchers must be careful in interpreting results from quantitative studies. Similar to previous studies, exact measurements and analyses of drug use may result in both over- and under-estimation of adherence. For example, we used information on numbers of drugs prescribed, but actual use of the medication can differ largely. The discrepancies in adherence could also be explained by the fact that in our context, patients may buy medicines for hypertension without prescription in the pharmacy. This will potentially be similar in other developing countries, which may lead to underestimation of adherence when only health facility

Chapter 3

data are used to estimate it. Also, self-reporting may have its limitations; for example, recall bias may be an issue.

Conclusions

Medication adherence was relatively low among hypertensive subjects in Vietnam but similar to that in many other countries. CVD risk at baseline survey did not significantly differentiate adherent from non-adherent subjects. Yet, significant differences in adherence were found for age. This may suggest that rather than risk profile, age should be considered for guiding the choice on who to target for improving medication adherence. Our qualitative study enabled further detailing of factors influencing adherence and indicated that the quantitative results should be interpreted with caution.

Acknowledgement

The research was funded by a NUFFIC (Netherlands) project for PhD research. Thanks go to the staff of the community health stations in Trung Thanh, Cu Van, La Hien and Tuc Tranh for their co-operation in the study.

Chap

ter 3

69


[45-47], suggesting that (non-) adherence is related to awareness of risks for complications, to the presence of and wish to deal with side effects, and to absence of symptoms of high BP (“feeling healthy”). All adherent subjects mentioned that they were motivated to adhere because they were aware of potential complications. This illustrates how the qualitative study adds details to results from the quantitative part.

Adherence of about one third of the 18 interviewed patients showed discrepancies compared to the qualitative data results. Two patients, initially classified as quantitatively adherent, were found to be non-adherent, while four patients initially thought to be non-adherent turned out to be adherent. Furthermore, the qualitative results suggested that one main reason for subjects to adhere to therapy was their awareness of the seriousness of hypertension complications. These results show the importance of understanding how patients decide on and exhibit adherence to medication, and demonstrate that it is insufficient to assume that hypertension patients do follow the prescription, as has sometimes been suggested [48].

The discrepancy between qualitative and quantitative data in classifying patients as adherent or not suggests that researchers must be careful in interpreting results from quantitative studies. Similar to previous studies, exact measurements and analyses of drug use may result in both over- and under-estimation of adherence. For example, we used information on numbers of drugs prescribed, but actual use of the medication can differ largely. The discrepancies in adherence could also be explained by the fact that in our context, patients may buy medicines for hypertension without prescription in the pharmacy. This will potentially be similar in other developing countries, which may lead to underestimation of adherence when only health facility

Chapter 3

data are used to estimate it. Also, self-reporting may have its limitations; for example, recall bias may be an issue.

Conclusions

Medication adherence was relatively low among hypertensive subjects in Vietnam but similar to that in many other countries. CVD risk at baseline survey did not significantly differentiate adherent from non-adherent subjects. Yet, significant differences in adherence were found for age. This may suggest that rather than risk profile, age should be considered for guiding the choice on who to target for improving medication adherence. Our qualitative study enabled further detailing of factors influencing adherence and indicated that the quantitative results should be interpreted with caution.

Acknowledgement

The research was funded by a NUFFIC (Netherlands) project for PhD research. Thanks go to the staff of the community health stations in Trung Thanh, Cu Van, La Hien and Tuc Tranh for their co-operation in the study.

70


References

1. Dragomir A, Cote R, Roy L, Blais L, Lalonde L, Berard A, et al. Impact of adherence to antihypertensive agents on clinical outcomes and hospitalization costs. Med Care. 2010;48(5):418-25. Epub 2010/04/16. doi: 10.1097/MLR.0b013e3181d567bd. PubMed PMID: 20393367.

2. Esposti LD, Saragoni S, Benemei S, Batacchi P, Geppetti P, Di Bari M, et al. Adherence to antihypertensive medications and health outcomes among newly treated hypertensive patients. Clinicoecon Outcomes Res. 2011;3:47-54. Epub 2011/09/22. doi: 10.2147/ceor.s15619. PubMed PMID: 21935332; PubMed Central PMCID: PMCPmc3169972.

3. Breekveldt-Postma NS, Penning-van Beest FJ, Siiskonen SJ, Falvey H, Vincze G, Klungel OH, et al. The effect of discontinuation of antihypertensives on the risk of acute myocardial infarction and stroke. Curr Med Res Opin. 2008;24(1):121-7. Epub 2007/11/23. doi: 10.1185/030079908x253843. PubMed PMID: 18031596.

4. Shin S, Song H, Oh SK, Choi KE, Kim H, Jang S. Effect of antihypertensive medication adherence on hospitalization for cardiovascular disease and mortality in hypertensive patients. Hypertens Res. 2013;36 (11):1000-5. Epub 2013/08/24. doi: 10.1038/hr.2013.85. PubMed PMID: 23966057.

Chapter 3

5. Corrao G, Parodi A, Nicotra F, Zambon A, Merlino L, Cesana G, et al. Better compliance to antihypertensive medications reduces cardiovascular risk. J Hypertens. 2011;29 (3):610-8.Epub 2010/12/16. doi: 10.1097/HJH.0b013e328342ca97 [doi]. PubMed PMID: 21157368.

6. Campbell NR, Brant R, Johansen H, Walker RL, Wielgosz A, Onysko J, et al. Increases in antihypertensive prescriptions and reductions in cardiovascular events in Canada. Hypertension. 2009;53(2):128-34. Epub 2008/12/31. doi: 10.1161/hypertensionaha.108.119784. PubMed PMID: 19114646.

7. Bramley TJ, Gerbino PP, Nightengale BS, Frech-Tamas F. Relationship of blood pressure control to adherence with antihypertensive monotherapy in 13 managed care organizations. J Manag Care Pharm. 2006;12(3):239-45. Epub 2006/04/21. PubMed PMID: 16623608.

8. Ramli A, Ahmad NS, Paraidathathu T. Medication adherence among hypertensive patients of primary health clinics in Malaysia. Patient Prefer Adherence. 2012;6:613-22. Epub 2012/09/13. doi: 10.2147/ppa.s34704. PubMed PMID: 22969292; PubMed Central PMCID: PMCPmc3437910.

9. Wu PH, Yang CY, Yao ZL, Lin WZ, Wu LW, Chang CC. Relationship of blood pressure control and hospitalization risk to medication adherence among patients with hypertension in Taiwan. Am J Hypertens. 2010;23(2):155-60. Epub 2009/11/21. doi: 10.1038/ajh.2009.210. PubMed PMID: 19927135.

Chap

ter 3

71


References

1. Dragomir A, Cote R, Roy L, Blais L, Lalonde L, Berard A, et al. Impact of adherence to antihypertensive agents on clinical outcomes and hospitalization costs. Med Care. 2010;48(5):418-25. Epub 2010/04/16. doi: 10.1097/MLR.0b013e3181d567bd. PubMed PMID: 20393367.

2. Esposti LD, Saragoni S, Benemei S, Batacchi P, Geppetti P, Di Bari M, et al. Adherence to antihypertensive medications and health outcomes among newly treated hypertensive patients. Clinicoecon Outcomes Res. 2011;3:47-54. Epub 2011/09/22. doi: 10.2147/ceor.s15619. PubMed PMID: 21935332; PubMed Central PMCID: PMCPmc3169972.

3. Breekveldt-Postma NS, Penning-van Beest FJ, Siiskonen SJ, Falvey H, Vincze G, Klungel OH, et al. The effect of discontinuation of antihypertensives on the risk of acute myocardial infarction and stroke. Curr Med Res Opin. 2008;24(1):121-7. Epub 2007/11/23. doi: 10.1185/030079908x253843. PubMed PMID: 18031596.

4. Shin S, Song H, Oh SK, Choi KE, Kim H, Jang S. Effect of antihypertensive medication adherence on hospitalization for cardiovascular disease and mortality in hypertensive patients. Hypertens Res. 2013;36 (11):1000-5. Epub 2013/08/24. doi: 10.1038/hr.2013.85. PubMed PMID: 23966057.

Chapter 3

5. Corrao G, Parodi A, Nicotra F, Zambon A, Merlino L, Cesana G, et al. Better compliance to antihypertensive medications reduces cardiovascular risk. J Hypertens. 2011;29 (3):610-8.Epub 2010/12/16. doi: 10.1097/HJH.0b013e328342ca97 [doi]. PubMed PMID: 21157368.

6. Campbell NR, Brant R, Johansen H, Walker RL, Wielgosz A, Onysko J, et al. Increases in antihypertensive prescriptions and reductions in cardiovascular events in Canada. Hypertension. 2009;53(2):128-34. Epub 2008/12/31. doi: 10.1161/hypertensionaha.108.119784. PubMed PMID: 19114646.

7. Bramley TJ, Gerbino PP, Nightengale BS, Frech-Tamas F. Relationship of blood pressure control to adherence with antihypertensive monotherapy in 13 managed care organizations. J Manag Care Pharm. 2006;12(3):239-45. Epub 2006/04/21. PubMed PMID: 16623608.

8. Ramli A, Ahmad NS, Paraidathathu T. Medication adherence among hypertensive patients of primary health clinics in Malaysia. Patient Prefer Adherence. 2012;6:613-22. Epub 2012/09/13. doi: 10.2147/ppa.s34704. PubMed PMID: 22969292; PubMed Central PMCID: PMCPmc3437910.

9. Wu PH, Yang CY, Yao ZL, Lin WZ, Wu LW, Chang CC. Relationship of blood pressure control and hospitalization risk to medication adherence among patients with hypertension in Taiwan. Am J Hypertens. 2010;23(2):155-60. Epub 2009/11/21. doi: 10.1038/ajh.2009.210. PubMed PMID: 19927135.

72


10. Burnier M, Brede Y, Lowy A. Impact of prolonged antihypertensive duration of action on predicted clinical outcomes in imperfectly adherent patients: comparison of aliskiren, irbesartan and ramipril. Int J Clin Pract. 2011;65(2):127-33. Epub 2011/01/07. doi: 10.1111/j.1742-1241.2010.02616.x. PubMed PMID: 21208354.

11. Grosso G, Raciti T, Marventano S, Romeo I, Mistretta A. [Adherence to antihypertensive and lipid-lowering medications: a problem of public health, not yet resolved]. Ann Ig. 2011;23(2):173-84. Epub 2011/07/21. PubMed PMID: 21770233.

12. Lowy A, Munk VC, Ong SH, Burnier M, Vrijens B, Tousset EP, et al. Effects on blood pressure and cardiovascular risk of variations in patients' adherence to prescribed antihypertensive drugs: role of duration of drug action. Int J Clin Pract. 2011;65(1):41-53. Epub 2010/11/26. doi: 10.1111/j.1742-1241.2010.02569.x. PubMed PMID: 21091596.

13. Matsumura K, Arima H, Tominaga M, Ohtsubo T, Sasaguri T, Fujii K, et al. Impact of antihypertensive medication adherence on blood pressure control in hypertension: the COMFORT study. Qjm. 2013;106(10):909-14. Epub 2013/05/23. doi: 10.1093/qjmed/hct121. PubMed PMID: 23696676.

14. Krousel-Wood M, Thomas S, Muntner P, Morisky D. Medication adherence: a key factor in achieving blood pressure control and good clinical outcomes in hypertensive patients. Curr Opin Cardiol. 2004;19(4):357-62. Epub 2004/06/26. PubMed PMID: 15218396.

Chapter 3

15. Grassi G, Seravalle G, Mancia G. Cardiovascular consequences of poor compliance to antihypertensive therapy. Blood Press. 2011;20(4):196-203. Epub 2011/02/09. doi: 10.3109/08037051.2011.557902 [doi]. PubMed PMID: 21299436.

16. Gwadry-Sridhar FH, Manias E, Lal L, Salas M, Hughes DA, Ratzki-Leewing A, et al. Impact of interventions on medication adherence and blood pressure control in patients with essential hypertension: a systematic review by the ISPOR medication adherence and persistence special interest group.Value Health. 2013;16(5):863-71. Epub 2013/08/21. doi: 10.1016/j.jval.2013.03.1631. PubMed PMID: 23947982.

17. Naderi SH, Bestwick JP, Wald DS. Adherence to drugs that prevent cardiovascular disease: meta-analysis on 376,162 patients. Am J Med. 2012;125(9):882-7.e1. Epub 2012/07/04. doi: 10.1016/j.amjmed.2011.12.013. PubMed PMID: 22748400.

18. Wilinski J, Dabrowski M. Medication adherence in hypertensive patients of different cardiovascular risk treated in primary health care. Przegl Lek. 2013;70(6):377-80. Epub 2013/09/24. PubMed PMID: 24052973.

19. Lee GK, Wang HH, Liu KQ, Cheung Y, Morisky DE, Wong MC. Determinants of medication adherence to antihypertensive medications among a Chinese population using Morisky Medication Adherence Scale. PLoS One. 2013;8(4):e62775. Epub 2013/05/03. doi: 10.1371/journal.pone.0062775. PubMed PMID: 23638143; PubMed Central PMCID: PMCPmc3636185.

Chap

ter 3

73


10. Burnier M, Brede Y, Lowy A. Impact of prolonged antihypertensive duration of action on predicted clinical outcomes in imperfectly adherent patients: comparison of aliskiren, irbesartan and ramipril. Int J Clin Pract. 2011;65(2):127-33. Epub 2011/01/07. doi: 10.1111/j.1742-1241.2010.02616.x. PubMed PMID: 21208354.

11. Grosso G, Raciti T, Marventano S, Romeo I, Mistretta A. [Adherence to antihypertensive and lipid-lowering medications: a problem of public health, not yet resolved]. Ann Ig. 2011;23(2):173-84. Epub 2011/07/21. PubMed PMID: 21770233.

12. Lowy A, Munk VC, Ong SH, Burnier M, Vrijens B, Tousset EP, et al. Effects on blood pressure and cardiovascular risk of variations in patients' adherence to prescribed antihypertensive drugs: role of duration of drug action. Int J Clin Pract. 2011;65(1):41-53. Epub 2010/11/26. doi: 10.1111/j.1742-1241.2010.02569.x. PubMed PMID: 21091596.

13. Matsumura K, Arima H, Tominaga M, Ohtsubo T, Sasaguri T, Fujii K, et al. Impact of antihypertensive medication adherence on blood pressure control in hypertension: the COMFORT study. Qjm. 2013;106(10):909-14. Epub 2013/05/23. doi: 10.1093/qjmed/hct121. PubMed PMID: 23696676.

14. Krousel-Wood M, Thomas S, Muntner P, Morisky D. Medication adherence: a key factor in achieving blood pressure control and good clinical outcomes in hypertensive patients. Curr Opin Cardiol. 2004;19(4):357-62. Epub 2004/06/26. PubMed PMID: 15218396.

Chapter 3

15. Grassi G, Seravalle G, Mancia G. Cardiovascular consequences of poor compliance to antihypertensive therapy. Blood Press. 2011;20(4):196-203. Epub 2011/02/09. doi: 10.3109/08037051.2011.557902 [doi]. PubMed PMID: 21299436.

16. Gwadry-Sridhar FH, Manias E, Lal L, Salas M, Hughes DA, Ratzki-Leewing A, et al. Impact of interventions on medication adherence and blood pressure control in patients with essential hypertension: a systematic review by the ISPOR medication adherence and persistence special interest group.Value Health. 2013;16(5):863-71. Epub 2013/08/21. doi: 10.1016/j.jval.2013.03.1631. PubMed PMID: 23947982.

17. Naderi SH, Bestwick JP, Wald DS. Adherence to drugs that prevent cardiovascular disease: meta-analysis on 376,162 patients. Am J Med. 2012;125(9):882-7.e1. Epub 2012/07/04. doi: 10.1016/j.amjmed.2011.12.013. PubMed PMID: 22748400.

18. Wilinski J, Dabrowski M. Medication adherence in hypertensive patients of different cardiovascular risk treated in primary health care. Przegl Lek. 2013;70(6):377-80. Epub 2013/09/24. PubMed PMID: 24052973.

19. Lee GK, Wang HH, Liu KQ, Cheung Y, Morisky DE, Wong MC. Determinants of medication adherence to antihypertensive medications among a Chinese population using Morisky Medication Adherence Scale. PLoS One. 2013;8(4):e62775. Epub 2013/05/03. doi: 10.1371/journal.pone.0062775. PubMed PMID: 23638143; PubMed Central PMCID: PMCPmc3636185.

74


20. Park JH, Shin Y, Lee SY, Lee SI. Antihypertensive drug medication adherence and its affecting factors in South Korea. Int J Cardiol. 2008;128(3):392-8. Epub 2007/07/24. doi: S0167-5273(07)01112-6[pii] 10.1016/j.ijcard.2007.04.114 [doi]. PubMed PMID: 17643514.

21. Nguyen QN, Pham ST, Nguyen VL, Wall S, Weinehall L, Bonita R, et al. Implementing a hypertension management programme in a rural area: local approaches and experiences from Ba-Vi district, Vietnam. BMC Public Health. 2011;11:325. Epub 2011/05/19. doi: 1471-2458-11-325 [pii] 10.1186/1471-2458-11-325 [doi]. PubMed PMID: 21586119; PubMed Central PMCID: PMC3112133.

22. Bailey JE, Lee MD, Somes GW, Graham RL. Risk factors for antihypertensive medication refill failure by patients under Medicaid managed care. Clin Ther. 1996;18(6):1252-62. Epub 1996/11/01. PubMed PMID: 9001841.

23. Wong MC, Jiang JY, Griffiths SM. Factors associated with compliance to thiazide diuretics among 8551 Chinese patients. J Clin Pharm Ther. 2011;36(2):179-86. Epub 2011/03/04. doi: 10.1111/j.1365-2710.2010.01174.x. PubMed PMID: 21366647.

24. Marshall IJ, Wolfe CD, McKevitt C. Lay perspectives on hypertension and drug adherence: systematic review of qualitative research. Bmj. 2012;345:e3953. Epub 2012/07/11. doi: 10.1136/bmj.e3953. PubMed PMID: 22777025; PubMed Central PMCID: PMCPmc3392078.

Chapter 3

25. Iskedjian M, Einarson TR, MacKeigan LD, Shear N, Addis A, Mittmann N, et al. Relationship between daily dose frequency and adherence to antihypertensive pharmacotherapy: evidence from a meta-analysis. Clin Ther. 2002;24(2):302-16. Epub 2002/03/26. PubMed PMID: 11911560.

26. Lewis LM, Schoenthaler AM, Ogedegbe G. Patient factors, but not provider and health care system factors, predict medication adherence in hypertensive black men. J Clin Hypertens (Greenwich). 2012;14(4):250-5. Epub 2012/03/31. doi: 10.1111/j.1751-7176.2012.00591.x. PubMed PMID: 22458747.

27. Lewis LM. Factors associated with medication adherence in hypertensive blacks: a review of the literature. J Cardiovasc Nurs. 2012;27(3):208-19. Epub 2011/07/16. doi: 10.1097/JCN.0b013e318215bb8f. PubMed PMID: 21760525.

28. Chen SL, Lee WL, Liang T, Liao IC. Factors associated with gender differences in medication adherence: a longitudinal study. J Adv Nurs. 2014;70(9):2031-40. Epub 2014/02/11. doi: 10.1111/jan.12361 [doi]. PubMed PMID: 24506542.

29. Hashmi SK, Afridi MB, Abbas K, Sajwani RA, Saleheen D, Frossard PM, et al. Factors associated with adherence to anti-hypertensive treatment in Pakistan. PLoS One. 2007;2(3):e280. Epub 2007/03/16. doi: 10.1371/journal.pone.0000280. PubMed PMID: 17356691; PubMed Central PMCID: PMCPmc1805684.

Chap

ter 3

75


20. Park JH, Shin Y, Lee SY, Lee SI. Antihypertensive drug medication adherence and its affecting factors in South Korea. Int J Cardiol. 2008;128(3):392-8. Epub 2007/07/24. doi: S0167-5273(07)01112-6[pii] 10.1016/j.ijcard.2007.04.114 [doi]. PubMed PMID: 17643514.

21. Nguyen QN, Pham ST, Nguyen VL, Wall S, Weinehall L, Bonita R, et al. Implementing a hypertension management programme in a rural area: local approaches and experiences from Ba-Vi district, Vietnam. BMC Public Health. 2011;11:325. Epub 2011/05/19. doi: 1471-2458-11-325 [pii] 10.1186/1471-2458-11-325 [doi]. PubMed PMID: 21586119; PubMed Central PMCID: PMC3112133.

22. Bailey JE, Lee MD, Somes GW, Graham RL. Risk factors for antihypertensive medication refill failure by patients under Medicaid managed care. Clin Ther. 1996;18(6):1252-62. Epub 1996/11/01. PubMed PMID: 9001841.

23. Wong MC, Jiang JY, Griffiths SM. Factors associated with compliance to thiazide diuretics among 8551 Chinese patients. J Clin Pharm Ther. 2011;36(2):179-86. Epub 2011/03/04. doi: 10.1111/j.1365-2710.2010.01174.x. PubMed PMID: 21366647.

24. Marshall IJ, Wolfe CD, McKevitt C. Lay perspectives on hypertension and drug adherence: systematic review of qualitative research. Bmj. 2012;345:e3953. Epub 2012/07/11. doi: 10.1136/bmj.e3953. PubMed PMID: 22777025; PubMed Central PMCID: PMCPmc3392078.

Chapter 3

25. Iskedjian M, Einarson TR, MacKeigan LD, Shear N, Addis A, Mittmann N, et al. Relationship between daily dose frequency and adherence to antihypertensive pharmacotherapy: evidence from a meta-analysis. Clin Ther. 2002;24(2):302-16. Epub 2002/03/26. PubMed PMID: 11911560.

26. Lewis LM, Schoenthaler AM, Ogedegbe G. Patient factors, but not provider and health care system factors, predict medication adherence in hypertensive black men. J Clin Hypertens (Greenwich). 2012;14(4):250-5. Epub 2012/03/31. doi: 10.1111/j.1751-7176.2012.00591.x. PubMed PMID: 22458747.

27. Lewis LM. Factors associated with medication adherence in hypertensive blacks: a review of the literature. J Cardiovasc Nurs. 2012;27(3):208-19. Epub 2011/07/16. doi: 10.1097/JCN.0b013e318215bb8f. PubMed PMID: 21760525.

28. Chen SL, Lee WL, Liang T, Liao IC. Factors associated with gender differences in medication adherence: a longitudinal study. J Adv Nurs. 2014;70(9):2031-40. Epub 2014/02/11. doi: 10.1111/jan.12361 [doi]. PubMed PMID: 24506542.

29. Hashmi SK, Afridi MB, Abbas K, Sajwani RA, Saleheen D, Frossard PM, et al. Factors associated with adherence to anti-hypertensive treatment in Pakistan. PLoS One. 2007;2(3):e280. Epub 2007/03/16. doi: 10.1371/journal.pone.0000280. PubMed PMID: 17356691; PubMed Central PMCID: PMCPmc1805684.

76


30. Tsiantou V, Pantzou P, Pavi E, Koulierakis G, Kyriopoulos J. Factors affecting adherence to antihypertensive medication in Greece: results from a qualitative study. Patient Prefer Adherence. 2010;4:335-43. Epub 2010/09/23. PubMed PMID: 20859460; PubMed Central PMCID: PMCPmc2943225.

31. Rimando M. Factors influencing medication compliance among hypertensive older African American adults. Ethn Dis. 2013;23(4):469-73. Epub 2014/01/08. PubMed PMID: 24392610.

32. Schoenthaler A, Chaplin WF, Allegrante JP, Fernandez S,Diaz-Gloster M, Tobin JN, et al. Provider communication effects medication adherence in hypertensive African Americans. Patient Educ Couns. 2009;75(2):185-91. Epub 2008/11/18. doi: 10.1016/j.pec.2008.09.018. PubMed PMID: 19013740; PubMed Central PMCID: PMCPmc2698021.

33. Lagi A, Rossi A, Passaleva MT, Cartei A, Cencetti S. Compliance with therapy in hypertensive patients. Intern Emerg Med. 2006;1(3):204-8. Epub 2006/11/24. PubMed PMID: 17120466.

34. Simonyi G. [Patient adherence in antihypertensive treatment]. Orv Hetil. 2013;154(40):1587-91. Epub 2013/10/01. doi: 10.1556/oh.2013.29692. PubMed PMID: 24077162.

35. Thi Phuong Lan Nguyen CCMS-V, Thi Bach Yen Nguyen, Vu Thi Thu Hang, E. Pamela Wright, M.J. Postma. Models to Predict the Burden of Cardiovascular Disease Risk in a Rural Mountainous Region of Vietnam. Value in Health Regional Issues. 2014; 3(null):87-93.

Chapter 3

36. Alsabbagh MH, Lemstra M, Eurich D, Lix LM, Wilson TW, Watson E, et al. Socioeconomic status and nonadherence to antihypertensive drugs: a systematic review and meta-analysis. Value Health. 2014;17(2):288-96. Epub 2014/03/19. doi: 10.1016/j.jval.2013.11.011. PubMed PMID: 24636389.

37. Peterson AM, Nau DP, Cramer JA, Benner J, Gwadry-Sridhar F, Nichol M. A checklist for medication compliance and persistence studies using retrospective databases. Value Health. 2007;10(1):3-12. Epub 2007/01/31. doi: 10.1111/j.1524-4733.2006.00139.x. PubMed PMID: 17261111.

38. Hughes D, Cowell W, Koncz T, Cramer J. Methods for integrating medication compliance and persistence in pharmacoeconomic evaluations. Value Health. 2007;10(6):498-509. Epub 2007/11/01. doi: VHE205 [pii] 10.1111/j.1524-4733.2007.00205.x [doi]. PubMed PMID: 17970932.

39. Fitz-Simon N, Bennett K, Feely J. A review of studies of adherence with antihypertensive drugs using prescription databases. Ther Clin Risk Manag. 2005;1(2):93-106. Epub 2008/03/25. PubMed PMID: 18360549; PubMed Central PMCID: PMC1661615.

40. Wong MC, Tam WW, Cheung CS, Tong EL, Sek AC, Cheung NT, et al. Medication adherence to first-line antihypertensive drug class in a large Chinese population. Int J Cardiol. 2013;167(4):1438-42. Epub 2012/05/09. doi: S0167-5273(12)00481-0 [pii] 10.1016/j.ijcard.2012.04.060 [doi]. PubMed PMID: 22560948.

Chap

ter 3

77


30. Tsiantou V, Pantzou P, Pavi E, Koulierakis G, Kyriopoulos J. Factors affecting adherence to antihypertensive medication in Greece: results from a qualitative study. Patient Prefer Adherence. 2010;4:335-43. Epub 2010/09/23. PubMed PMID: 20859460; PubMed Central PMCID: PMCPmc2943225.

31. Rimando M. Factors influencing medication compliance among hypertensive older African American adults. Ethn Dis. 2013;23(4):469-73. Epub 2014/01/08. PubMed PMID: 24392610.

32. Schoenthaler A, Chaplin WF, Allegrante JP, Fernandez S,Diaz-Gloster M, Tobin JN, et al. Provider communication effects medication adherence in hypertensive African Americans. Patient Educ Couns. 2009;75(2):185-91. Epub 2008/11/18. doi: 10.1016/j.pec.2008.09.018. PubMed PMID: 19013740; PubMed Central PMCID: PMCPmc2698021.

33. Lagi A, Rossi A, Passaleva MT, Cartei A, Cencetti S. Compliance with therapy in hypertensive patients. Intern Emerg Med. 2006;1(3):204-8. Epub 2006/11/24. PubMed PMID: 17120466.

34. Simonyi G. [Patient adherence in antihypertensive treatment]. Orv Hetil. 2013;154(40):1587-91. Epub 2013/10/01. doi: 10.1556/oh.2013.29692. PubMed PMID: 24077162.

35. Thi Phuong Lan Nguyen CCMS-V, Thi Bach Yen Nguyen, Vu Thi Thu Hang, E. Pamela Wright, M.J. Postma. Models to Predict the Burden of Cardiovascular Disease Risk in a Rural Mountainous Region of Vietnam. Value in Health Regional Issues. 2014; 3(null):87-93.

Chapter 3

36. Alsabbagh MH, Lemstra M, Eurich D, Lix LM, Wilson TW, Watson E, et al. Socioeconomic status and nonadherence to antihypertensive drugs: a systematic review and meta-analysis. Value Health. 2014;17(2):288-96. Epub 2014/03/19. doi: 10.1016/j.jval.2013.11.011. PubMed PMID: 24636389.

37. Peterson AM, Nau DP, Cramer JA, Benner J, Gwadry-Sridhar F, Nichol M. A checklist for medication compliance and persistence studies using retrospective databases. Value Health. 2007;10(1):3-12. Epub 2007/01/31. doi: 10.1111/j.1524-4733.2006.00139.x. PubMed PMID: 17261111.

38. Hughes D, Cowell W, Koncz T, Cramer J. Methods for integrating medication compliance and persistence in pharmacoeconomic evaluations. Value Health. 2007;10(6):498-509. Epub 2007/11/01. doi: VHE205 [pii] 10.1111/j.1524-4733.2007.00205.x [doi]. PubMed PMID: 17970932.

39. Fitz-Simon N, Bennett K, Feely J. A review of studies of adherence with antihypertensive drugs using prescription databases. Ther Clin Risk Manag. 2005;1(2):93-106. Epub 2008/03/25. PubMed PMID: 18360549; PubMed Central PMCID: PMC1661615.

40. Wong MC, Tam WW, Cheung CS, Tong EL, Sek AC, Cheung NT, et al. Medication adherence to first-line antihypertensive drug class in a large Chinese population. Int J Cardiol. 2013;167(4):1438-42. Epub 2012/05/09. doi: S0167-5273(12)00481-0 [pii] 10.1016/j.ijcard.2012.04.060 [doi]. PubMed PMID: 22560948.

78


41. Khanam MA, Lindeboom W, Koehlmoos TL, Alam DS, Niessen L, Milton AH. Hypertension: adherence to treatment in rural Bangladesh - findings from a population-based study. Glob Health Action. 2014;7:25028. Epub 2014/11/02. doi: 25028 [pii]. PubMed PMID: 25361723; PubMed Central PMCID: PMC4212079.

42. Ross S, Walker A, MacLeod MJ. Patient compliance in hypertension: role of illness perceptions and treatment beliefs. J Hum Hypertens. 2004;18(9):607-13. Epub 2004/03/19. doi: 10.1038/sj.jhh.1001721 [doi] 1001721 [pii]. PubMed PMID: 15029218.

43. Chapman RH1 BJ, Petrilla AA, Tierce JC, Collins SR, Battleman DS, Schwartz JS. Predictors of adherence with antihypertensive and lipid-lowering therapy. Arch Intern Med 2005;165(10):1147-52.

44. World-Health-Organization. Prevention of cardiovascular disease. [cited 2013 August 5,]. Available from: http://www.who.int/cardiovascular_diseases/guidelines/PocketGL.ENGLISH.AFR-D-E.rev1.pdf.

45. Fongwa MN, Evangelista LS, Hays RD, Martins DS, Elashoff D, Cowan MJ, et al. Adherence treatment factors in hypertensive African American women. Vasc Health Risk Manag. 2008;4(1):157-66. Epub 2008/07/17. PubMed PMID: 18629350; PubMed Central PMCID: PMC2464745.

Chapter 3

46. Gascon JJ, Sanchez-Ortuno M, Llor B, Skidmore D, Saturno PJ. Why hypertensive patients do not comply with the treatment: results from a qualitative study. Fam Pract. 2004;21(2):125-30. Epub 2004/03/17. PubMed PMID: 15020377.

47. Osamor PE, Owumi BE. Factors associated with treatment compliance in hypertension in southwest Nigeria. J Health Popul Nutr. 2011;29(6):619-28. Epub 2012/01/31. PubMed PMID: 22283036; PubMed Central PMCID: PMC3259725.

48. Rosenbaum L. Beyond belief--how people feel about taking medications for heart disease. N Engl J Med. 2015;372(2):183-7. Epub 2015/01/08. doi: 10.1056/NEJMms1409015 [doi]. PubMed PMID: 25564902.

Chap

ter 3

79


41. Khanam MA, Lindeboom W, Koehlmoos TL, Alam DS, Niessen L, Milton AH. Hypertension: adherence to treatment in rural Bangladesh - findings from a population-based study. Glob Health Action. 2014;7:25028. Epub 2014/11/02. doi: 25028 [pii]. PubMed PMID: 25361723; PubMed Central PMCID: PMC4212079.

42. Ross S, Walker A, MacLeod MJ. Patient compliance in hypertension: role of illness perceptions and treatment beliefs. J Hum Hypertens. 2004;18(9):607-13. Epub 2004/03/19. doi: 10.1038/sj.jhh.1001721 [doi] 1001721 [pii]. PubMed PMID: 15029218.

43. Chapman RH1 BJ, Petrilla AA, Tierce JC, Collins SR, Battleman DS, Schwartz JS. Predictors of adherence with antihypertensive and lipid-lowering therapy. Arch Intern Med 2005;165(10):1147-52.

44. World-Health-Organization. Prevention of cardiovascular disease. [cited 2013 August 5,]. Available from: http://www.who.int/cardiovascular_diseases/guidelines/PocketGL.ENGLISH.AFR-D-E.rev1.pdf.

45. Fongwa MN, Evangelista LS, Hays RD, Martins DS, Elashoff D, Cowan MJ, et al. Adherence treatment factors in hypertensive African American women. Vasc Health Risk Manag. 2008;4(1):157-66. Epub 2008/07/17. PubMed PMID: 18629350; PubMed Central PMCID: PMC2464745.

Chapter 3

46. Gascon JJ, Sanchez-Ortuno M, Llor B, Skidmore D, Saturno PJ. Why hypertensive patients do not comply with the treatment: results from a qualitative study. Fam Pract. 2004;21(2):125-30. Epub 2004/03/17. PubMed PMID: 15020377.

47. Osamor PE, Owumi BE. Factors associated with treatment compliance in hypertension in southwest Nigeria. J Health Popul Nutr. 2011;29(6):619-28. Epub 2012/01/31. PubMed PMID: 22283036; PubMed Central PMCID: PMC3259725.

48. Rosenbaum L. Beyond belief--how people feel about taking medications for heart disease. N Engl J Med. 2015;372(2):183-7. Epub 2015/01/08. doi: 10.1056/NEJMms1409015 [doi]. PubMed PMID: 25564902.

Chapter 4

Direct costs of hypertensive patients admitted to hospital in Vietnam –

A bottom-up micro-costing analysis

Thi-Phuong-Lan Nguyen, Thi-Bach-Yen Nguyen, Thanh-Trung Nguyen, Van-Vinh Hac, Hoa H Le, C.C.M. Schuiling-Veninga, M.J.Postma

BMC Health Serv Res. 2014; 14: 514.

DOI: 10.1186/s12913-014-0514-4.

82

Direct costs of hypertensive inpatients

Abstract

Objectives

There is an economic burden associated with hypertension both worldwide and in Vietnam. In Vietnam, patients with uncontrolledhigh blood pressure are hospitalized for further diagnosis and initiation of treatment. Because there is no evidence on costs of inpatient carefor hypertensive patients available yet to inform policy makers, health insurance and hospitals, this study aims to quantify direct costs ofinpatient care for these patients in Vietnam.

Methods

A retrospective study was conducted in a hospital in Vietnam. Directcosts were analyzed from the health-care provider’s perspective.Hospital-based costing was performed using both bottom-up andmicro-costing methods. Patients with sole essential or primaryhypertension (ICD-code I10) and those comorbid with sphingolipid metabolism or other lipid storage disorders (ICD-code E75) wereselected. Costs were quantified based on financial and other records ofthe hospital. Total cost per patient resulted from an aggregation oflaboratory test costs, drug costs, inpatient-days’ costs and otherremaining costs, including appropriate allocation of overheads. Both mean and medians, as well as interquartile ranges (IQRs) werecalculated. In addition to a base-case analysis, specific scenarios wereanalyzed.

Results

230 patients were included in the study (147 cases with I10 code onlyand 83 cases with I10 combined with E75). Median length

Chapter 4

of hospital stay was 6 days. Median total direct costs per patient wereUS$65 (IQR: 37 -95). Total costs per patient were higher in thecombined hypertensive and lipid population than in the solehypertensive population at US$78 and US$53, respectively. In allscenarios, hospital inpatient days’ costs were identified as the majorcost driver in the total costs.

Conclusion

Costs of hospitalization of hypertensive patients is relatively high compared to annual medication treatment at a community health station for hypertension as well as to the total health expenditure per capita in Vietnam. Given that untreated/undetected hypertension likely leads to more expensive treatments of complications, these findings may justify investments by the Vietnamese health-care sector to control high blood pressure in order to save downstream health care budgets.

Chap

ter 4

83


Abstract

Objectives

There is an economic burden associated with hypertension both worldwide and in Vietnam. In Vietnam, patients with uncontrolledhigh blood pressure are hospitalized for further diagnosis and initiation of treatment. Because there is no evidence on costs of inpatient carefor hypertensive patients available yet to inform policy makers, health insurance and hospitals, this study aims to quantify direct costs ofinpatient care for these patients in Vietnam.

Methods

A retrospective study was conducted in a hospital in Vietnam. Directcosts were analyzed from the health-care provider’s perspective.Hospital-based costing was performed using both bottom-up andmicro-costing methods. Patients with sole essential or primaryhypertension (ICD-code I10) and those comorbid with sphingolipid metabolism or other lipid storage disorders (ICD-code E75) wereselected. Costs were quantified based on financial and other records ofthe hospital. Total cost per patient resulted from an aggregation oflaboratory test costs, drug costs, inpatient-days’ costs and otherremaining costs, including appropriate allocation of overheads. Both mean and medians, as well as interquartile ranges (IQRs) werecalculated. In addition to a base-case analysis, specific scenarios wereanalyzed.

Results

230 patients were included in the study (147 cases with I10 code onlyand 83 cases with I10 combined with E75). Median length

Chapter 4

of hospital stay was 6 days. Median total direct costs per patient wereUS$65 (IQR: 37 -95). Total costs per patient were higher in thecombined hypertensive and lipid population than in the solehypertensive population at US$78 and US$53, respectively. In allscenarios, hospital inpatient days’ costs were identified as the majorcost driver in the total costs.

Conclusion

Costs of hospitalization of hypertensive patients is relatively high compared to annual medication treatment at a community health station for hypertension as well as to the total health expenditure per capita in Vietnam. Given that untreated/undetected hypertension likely leads to more expensive treatments of complications, these findings may justify investments by the Vietnamese health-care sector to control high blood pressure in order to save downstream health care budgets.

84


Introduction

Hypertension represents a health and economic burden worldwide. In2000, approximately one quarter of the adult populations hadhypertension, equaling to approximately 972 million adults [1].Furthermore, it has been projected that almost 30% of the world’sadult population will be hypertensive by 2025 [1]. High blood pressure and associated diseases may be responsible for up to 7 million deaths annually worldwide [2]. In the Eastern European and CentralAsian regions, high blood pressure is estimated to directly orindirectly account for 25% of all health expenditures [3]. In SouthwestChina, a cost-of-illness analysis from the societal perspective in 2010estimated the cost of hypertension to be US$9,393 per patient [4]. Inthe Philippines, a health insurance company reported reimbursementfor hypertension- related diagnoses during 3.5 years to be US$56 million for 360,016 patients [5]. This equals to 34% of their financialbudget for hospital spending [5].

Among the group of developing countries, the prevalence ofhypertension in Vietnam can be rated as intermediate with estimatesranging from 20% to 30% in adults [6]. In recent years, hypertension has been one of the main contributors to the overall burden of diseasein Vietnam [7, 8]. A national survey, conducted between 2002 and 2008, estimated prevalence rates for overall, male, and femalepopulations to be 25.1%, 28.3% and 23.1%, respectively [8]. Usingthe Framingham general cardiovascular risk score, it was estimatedthat the prevalence of people with overall 10-year cardiovasculardisease (CVD) risks ≥ 10%, 20% and 30% are 27%, 10% and 3.9%,respectively [9, 10]. For patients who are 60-and-above years old, hypertension as the underlying cause of death ranks third for males(6.2% of all deaths) and second for females (6.6% of all deaths). Forall ages, hypertension ranks sixth for males (3.8% of all deaths) and

Chapter 4

second for females (5.1% of all deaths) [7]. Treatment of hypertensionis crucial to avert these high risks, associated costs and related deaths.

Compared to surrounding countries, hypertension may impact evenhigher costs to society in Vietnam because diagnosis and initiation oftreatments for hypertension often take place in the hospital. This maybe explained by several factors, including the absence of some specificservices in primary care for all regions of the country, the perception that hospital care offers a higher quality of care and the preference forspecific services to be in a hospital setting for administrative and convenience reasons. At present, the inpatient and outpatient economic burdens of hypertension in Vietnam have not been estimatedprecisely. In addition, studies on the topic are scarce. One study at thecommunity-health station level on the cost of drug treatment for thewhole population over a 10-years period estimated costs of 9,808billion VND associated with grade 1 hypertension and 11,192 billion VND associated with grade 2 and 3 hypertension [11]. The sum ofthese figures represent approximately 14% of the total health expenditure in 2010 [12, 13]. Notably, among hypertensive patients, only 29.6% was treated and 10.7% achieved target blood pressurecontrol [8]. Patients, who are not treated and have uncontrolled highblood pressure, are at higher risks of complications requiringhospitalization, which add to the economic burdens for the health caresystem.

The health-care provider payment system in Vietnam is currentlybeing reformed. The fee- for-service system will be replaced by case-mix payments and eventually diagnostic-related groups [14]. Thus, to support reimbursement decisions based on diagnostic-related groups, it would be useful for the Vietnamese health insurance system to haveaccurate estimates of costs associated with hypertension.

Chap

ter 4

85


Introduction

Hypertension represents a health and economic burden worldwide. In2000, approximately one quarter of the adult populations hadhypertension, equaling to approximately 972 million adults [1].Furthermore, it has been projected that almost 30% of the world’sadult population will be hypertensive by 2025 [1]. High blood pressure and associated diseases may be responsible for up to 7 million deaths annually worldwide [2]. In the Eastern European and CentralAsian regions, high blood pressure is estimated to directly orindirectly account for 25% of all health expenditures [3]. In SouthwestChina, a cost-of-illness analysis from the societal perspective in 2010estimated the cost of hypertension to be US$9,393 per patient [4]. Inthe Philippines, a health insurance company reported reimbursementfor hypertension- related diagnoses during 3.5 years to be US$56 million for 360,016 patients [5]. This equals to 34% of their financialbudget for hospital spending [5].

Among the group of developing countries, the prevalence ofhypertension in Vietnam can be rated as intermediate with estimatesranging from 20% to 30% in adults [6]. In recent years, hypertension has been one of the main contributors to the overall burden of diseasein Vietnam [7, 8]. A national survey, conducted between 2002 and 2008, estimated prevalence rates for overall, male, and femalepopulations to be 25.1%, 28.3% and 23.1%, respectively [8]. Usingthe Framingham general cardiovascular risk score, it was estimatedthat the prevalence of people with overall 10-year cardiovasculardisease (CVD) risks ≥ 10%, 20% and 30% are 27%, 10% and 3.9%,respectively [9, 10]. For patients who are 60-and-above years old, hypertension as the underlying cause of death ranks third for males(6.2% of all deaths) and second for females (6.6% of all deaths). Forall ages, hypertension ranks sixth for males (3.8% of all deaths) and

Chapter 4

second for females (5.1% of all deaths) [7]. Treatment of hypertensionis crucial to avert these high risks, associated costs and related deaths.

Compared to surrounding countries, hypertension may impact evenhigher costs to society in Vietnam because diagnosis and initiation oftreatments for hypertension often take place in the hospital. This maybe explained by several factors, including the absence of some specificservices in primary care for all regions of the country, the perception that hospital care offers a higher quality of care and the preference forspecific services to be in a hospital setting for administrative and convenience reasons. At present, the inpatient and outpatient economic burdens of hypertension in Vietnam have not been estimatedprecisely. In addition, studies on the topic are scarce. One study at thecommunity-health station level on the cost of drug treatment for thewhole population over a 10-years period estimated costs of 9,808billion VND associated with grade 1 hypertension and 11,192 billion VND associated with grade 2 and 3 hypertension [11]. The sum ofthese figures represent approximately 14% of the total health expenditure in 2010 [12, 13]. Notably, among hypertensive patients, only 29.6% was treated and 10.7% achieved target blood pressurecontrol [8]. Patients, who are not treated and have uncontrolled highblood pressure, are at higher risks of complications requiringhospitalization, which add to the economic burdens for the health caresystem.

The health-care provider payment system in Vietnam is currentlybeing reformed. The fee- for-service system will be replaced by case-mix payments and eventually diagnostic-related groups [14]. Thus, to support reimbursement decisions based on diagnostic-related groups, it would be useful for the Vietnamese health insurance system to haveaccurate estimates of costs associated with hypertension.

86


In addition, hospitals in Vietnam are gaining greater autonomy.Having knowledge on their expenditures for inpatient care serviceswill help these hospitals to improve their financial management aswell as to adequately issue fees for both insured and non-insuredpatients. Currently, neither information on the costs of inpatient carefor the treatment of hypertension and its consequences nor solid estimates on the comprehensive costs of hypertension management is available in Vietnam.

To address these gaps of information, we conducted this study to quantify the direct costs of inpatient care for hypertensive patients. Furthermore, these results may 1) contribute to a better understanding of the economic burden of hypertension in Vietnam, 2) help with reimbursement decision making for insured patients, 3) support setting potential fees and charges to be issued for non-insured patients and 4) inform on potential impacts of preventive policies.

Methods


We conducted a retrospective study in the Thai Nguyen hospital. Data was collected from the financial records during October 1st to September 30th 2011. The study was conducted at a regional hospital with 800 beds in the city of Thai Nguyen, which is located in amountainous area that is approximately 100 km North of Hanoi. Itserves patients from Thai Nguyen and neighboring provinces.

All costing of resource utilization was adjusted to 2011 l e v e l s and presented in US$ using the exchange rate of US$1 to VND20.830[15].

Chapter 4

Study participants

Using the International Classification of Diseases 10th version (ICD-10), we identified and retrieved information on all patients with codesI10 alone (essential or primary hypertension) or comorbid patients with I10 combined with E75 (sphingolipid metabolism and other lipid storage disorders). The latter group was included because hypertension often coincides with these disorders. The combination of diagnosesmight reflect a relatively large share of the overall hypertensive patient group with deviating costs. ICD-10 codes were taken from hospitaldatabases and individual patients’ records.

Information on age, sex and ICD code was available for all of the selected patients with ICD code I10 or I10 combined with E75.

However, data might be missing as administrators may forget to enterclassification data or misclassify; i.e. patients and/or patient-relateddata might be missing. In addition, by doing this retrospectively, wehad no certainty on whether each financial record was included or not. Nurses should enter all consumed items such as drugs, tests, medicalmaterials into the database system before she/he can get them fromstore or lab services. Other cost items such as patient-days andexaminations were automatically recorded for every patient. This ensures that all these items were recorded and financial records arelikely complete. Whenever financial records seemed grosslyincomplete or absent at all, patients were excluded. This was howevernot the case. It was therefore plausibly assumed that missing data waslimited and random. Thus, no specific bias was expected.

Costing perspective and cost components

Direct costs were analyzed from the provider perspective. Data oncosts included all charges to patients for drugs, materials (both medical and non-medical) and laboratory testing, all crucial elements

Chap

ter 4

87


In addition, hospitals in Vietnam are gaining greater autonomy.Having knowledge on their expenditures for inpatient care serviceswill help these hospitals to improve their financial management aswell as to adequately issue fees for both insured and non-insuredpatients. Currently, neither information on the costs of inpatient carefor the treatment of hypertension and its consequences nor solid estimates on the comprehensive costs of hypertension management is available in Vietnam.

To address these gaps of information, we conducted this study to quantify the direct costs of inpatient care for hypertensive patients. Furthermore, these results may 1) contribute to a better understanding of the economic burden of hypertension in Vietnam, 2) help with reimbursement decision making for insured patients, 3) support setting potential fees and charges to be issued for non-insured patients and 4) inform on potential impacts of preventive policies.

Methods


We conducted a retrospective study in the Thai Nguyen hospital. Data was collected from the financial records during October 1st to September 30th 2011. The study was conducted at a regional hospital with 800 beds in the city of Thai Nguyen, which is located in amountainous area that is approximately 100 km North of Hanoi. Itserves patients from Thai Nguyen and neighboring provinces.

All costing of resource utilization was adjusted to 2011 l e v e l s and presented in US$ using the exchange rate of US$1 to VND20.830[15].

Chapter 4

Study participants

Using the International Classification of Diseases 10th version (ICD-10), we identified and retrieved information on all patients with codesI10 alone (essential or primary hypertension) or comorbid patients with I10 combined with E75 (sphingolipid metabolism and other lipid storage disorders). The latter group was included because hypertension often coincides with these disorders. The combination of diagnosesmight reflect a relatively large share of the overall hypertensive patient group with deviating costs. ICD-10 codes were taken from hospitaldatabases and individual patients’ records.

Information on age, sex and ICD code was available for all of the selected patients with ICD code I10 or I10 combined with E75.

However, data might be missing as administrators may forget to enterclassification data or misclassify; i.e. patients and/or patient-relateddata might be missing. In addition, by doing this retrospectively, wehad no certainty on whether each financial record was included or not. Nurses should enter all consumed items such as drugs, tests, medicalmaterials into the database system before she/he can get them fromstore or lab services. Other cost items such as patient-days andexaminations were automatically recorded for every patient. This ensures that all these items were recorded and financial records arelikely complete. Whenever financial records seemed grosslyincomplete or absent at all, patients were excluded. This was howevernot the case. It was therefore plausibly assumed that missing data waslimited and random. Thus, no specific bias was expected.

Costing perspective and cost components

Direct costs were analyzed from the provider perspective. Data oncosts included all charges to patients for drugs, materials (both medical and non-medical) and laboratory testing, all crucial elements

88


in the financial records and reflecting adequate charging. In particular,hospital-based costing was performed using both bottom-up and micro-costing methods and aggregation was subsequently conductedfor all costs related to medical services used by a group of hospitalizedpatients [16, 17]. For this purpose, costs of inpatient care were brokendown into two parts:

Inpatient – days costs and other costs (laboratory testing, drugs, medical materials charged to patients directly, admission, andexaminations by specialists). Costs of inpatient care can be expressedas:

Costs of inpatient care = inpatient-days costs +laboratory test costs +drug costs + costs of medical materials charged to patients directly +admission cost + costs of examinations by specialists

Method of cost measurements

Calculation of inpatient – days’ costs. The step-down allocation method- partially adjusted for interaction between overheaddepartments- was applied for allocating overhead costs [17]. In thebase-case, the discount rate for medical equipment and building was3% [17]. The formula applied here may be expressed as:

Total costs of each department = total labor costs + total costs ofmaterials/infrastructure (both medical and non-medical, not charged to patients directly) + total costs of capital (both medical and non-medical equipment and buildings).

Labor costs were calculated based on the actual payment for labor bythe hospital every month during a year; funding was provided bygovernment or specific funds for services organized by the hospital.Material/infrastructure costs were calculated by items used at eachdepartment multiplied with the price market which the hospital paid based on again monthly updated records of the hospital. Capital costs

Chapter 4

were calculated based on assets’ records of the hospital, which is updated every year.

Notably, non-medical costs of materials/infrastructure primarilycomprise of power, telephone, uniforms, stationary and cloths. In theabsence of any detailed information, an assumption was made thatequal costs for an inpatient day would apply for every disease within each department.

The inpatient–day costs per department resulted from the total costs divided by the total number of inpatient days for that department:

Total costs of a given department

Inpatient–day costs =

Total number of inpatient-days at that department

Other costs and total patient costs.

The cost-to-charge ratio method was applied to calculate laboratorytesting costs [17]. While the input data to calculate cost of each test -such as number of chemicals used, time investment of staff membersto run each test, etcetera - were limited, total input of each department, total number of tests per type and prices were available. Therefore,cost-to-charge ratio was considered to be the best method forcalculating laboratory costs. At each laboratory department, wetracked the number of each test and multiplied these by the chargesthat apply in the hospital (base on the recommendation of the Ministryof Health). We subsequently summed up all those items to identifytotal finances of that department. Cost-to-charge ratio in eachdepartment was calculated dividing total input costs of eachdepartment in terms of building, labor, material, equipment, etceteraby those total finances generated through charges. Information on numbers of test was only available for the last half of the year because

Chap

ter 4

89


in the financial records and reflecting adequate charging. In particular,hospital-based costing was performed using both bottom-up and micro-costing methods and aggregation was subsequently conductedfor all costs related to medical services used by a group of hospitalizedpatients [16, 17]. For this purpose, costs of inpatient care were brokendown into two parts:

Inpatient – days costs and other costs (laboratory testing, drugs, medical materials charged to patients directly, admission, andexaminations by specialists). Costs of inpatient care can be expressedas:

Costs of inpatient care = inpatient-days costs +laboratory test costs +drug costs + costs of medical materials charged to patients directly +admission cost + costs of examinations by specialists

Method of cost measurements

Calculation of inpatient – days’ costs. The step-down allocation method- partially adjusted for interaction between overheaddepartments- was applied for allocating overhead costs [17]. In thebase-case, the discount rate for medical equipment and building was3% [17]. The formula applied here may be expressed as:

Total costs of each department = total labor costs + total costs ofmaterials/infrastructure (both medical and non-medical, not charged to patients directly) + total costs of capital (both medical and non-medical equipment and buildings).

Labor costs were calculated based on the actual payment for labor bythe hospital every month during a year; funding was provided bygovernment or specific funds for services organized by the hospital.Material/infrastructure costs were calculated by items used at eachdepartment multiplied with the price market which the hospital paid based on again monthly updated records of the hospital. Capital costs

Chapter 4

were calculated based on assets’ records of the hospital, which is updated every year.

Notably, non-medical costs of materials/infrastructure primarilycomprise of power, telephone, uniforms, stationary and cloths. In theabsence of any detailed information, an assumption was made thatequal costs for an inpatient day would apply for every disease within each department.

The inpatient–day costs per department resulted from the total costs divided by the total number of inpatient days for that department:

Total costs of a given department

Inpatient–day costs =

Total number of inpatient-days at that department

Other costs and total patient costs.

The cost-to-charge ratio method was applied to calculate laboratorytesting costs [17]. While the input data to calculate cost of each test -such as number of chemicals used, time investment of staff membersto run each test, etcetera - were limited, total input of each department, total number of tests per type and prices were available. Therefore,cost-to-charge ratio was considered to be the best method forcalculating laboratory costs. At each laboratory department, wetracked the number of each test and multiplied these by the chargesthat apply in the hospital (base on the recommendation of the Ministryof Health). We subsequently summed up all those items to identifytotal finances of that department. Cost-to-charge ratio in eachdepartment was calculated dividing total input costs of eachdepartment in terms of building, labor, material, equipment, etceteraby those total finances generated through charges. Information on numbers of test was only available for the last half of the year because

90


of a mistake in the software to extract total number of each test. To annualize, the total numbers for each test was multiplied by two.

In this study, standard national prices for drugs and materials were not available. As Thai Nguyen hospital is a non-profit hospital, weassumed that the use of drug and material charges from this hospitalwould well approach actual costs. Cost of drugs and materials werecalculated based on the current market prices, which the hospital paid. Therefore, we could validly impute these charges into our cost analysis for each drug and material used [17]. Both drugs and medical materials were charged to patients directly and accordingly inserted in the analysis. These costs were consistently calculated based on numbers of items used multiplied by the specific prices of each item.

Admission cost, which is required for every intake examination foradmission to the hospital, was calculated similarly as was done forinpatient – day’ costs. Admission costs per patient is equal to the totalcosts of outpatient department divided by the total number patient visits at outpatient.

Examination costs at specialist departments were calculated forpatients who had specific examinations. Examinations were countedand monetized using charges as determined by the recommendations of the Ministry of Health.

Total costs per patient resulted from the aggregation of inpatient-days’costs multiplied by the length of stay, laboratory test costs, drugcosts, medical material costs, admission costs, and costs ofexaminations at specialist departments.

Sources of data

Labor costs (wages and allowances) were quantified based on financialrecords of the hospital. Material (not charged to patients directly) andcapital (both medical and non-medical equipment and building costs)

Chapter 4

were quantified based on the financial records, administration ofmaterials used and capital inventories of the hospital.

Age, sex, department, length of stay, total numbers of eachlaboratory test, numbers of each drug and medical materials chargedto patients directly were retrieved from patient-based databases ofthe hospital. Prices for specific services such as medication, materials/ disposables and laboratory tests were available from the specificdatabases of the hospital. Furniture and land costs were not included in this study as no information was available.

The number of tests completed for each patient and its charges wereobtained from the individual patient sheets. We selected all tests, including test for the diagnosis and/or treatment of hypertension andother condition. Total number of all tests in 6 months was multipliedby 2 and then divided by the total number of patients in 2011, equating11 in the whole hospital. For the hypertensive in-patient group, theaverage number of tests per patient was 16.8, 16.3 and 17.1 in both groups, I10 and I10+E75, respectively. Of course for costing, eachindividual test was priced separately.

Statistical and sensitivity analysis

Means, medians and Inter Quartile Range (IQRs) of costs weremeasured as outcome in this study. Univariate sensitivity analyseswere performed to explore the robustness of the analyses [18].Notably, hypertensive patients may be admitted to hospital with comorbidities.

In this context, it is very important to rule out the costs resulting from diagnosing and treating these other diseases. Therefore, we conductedone sensitivity analysis with a scenario that excluded all costs resulting from diagnosing and treating comorbid diseases. We limitedto specific costs for diagnosing and treating I10 and E75 diseases in

Chap

ter 4

91


of a mistake in the software to extract total number of each test. To annualize, the total numbers for each test was multiplied by two.

In this study, standard national prices for drugs and materials were not available. As Thai Nguyen hospital is a non-profit hospital, weassumed that the use of drug and material charges from this hospitalwould well approach actual costs. Cost of drugs and materials werecalculated based on the current market prices, which the hospital paid. Therefore, we could validly impute these charges into our cost analysis for each drug and material used [17]. Both drugs and medical materials were charged to patients directly and accordingly inserted in the analysis. These costs were consistently calculated based on numbers of items used multiplied by the specific prices of each item.

Admission cost, which is required for every intake examination foradmission to the hospital, was calculated similarly as was done forinpatient – day’ costs. Admission costs per patient is equal to the totalcosts of outpatient department divided by the total number patient visits at outpatient.

Examination costs at specialist departments were calculated forpatients who had specific examinations. Examinations were countedand monetized using charges as determined by the recommendations of the Ministry of Health.

Total costs per patient resulted from the aggregation of inpatient-days’costs multiplied by the length of stay, laboratory test costs, drugcosts, medical material costs, admission costs, and costs ofexaminations at specialist departments.

Sources of data

Labor costs (wages and allowances) were quantified based on financialrecords of the hospital. Material (not charged to patients directly) andcapital (both medical and non-medical equipment and building costs)

Chapter 4

were quantified based on the financial records, administration ofmaterials used and capital inventories of the hospital.

Age, sex, department, length of stay, total numbers of eachlaboratory test, numbers of each drug and medical materials chargedto patients directly were retrieved from patient-based databases ofthe hospital. Prices for specific services such as medication, materials/ disposables and laboratory tests were available from the specificdatabases of the hospital. Furniture and land costs were not included in this study as no information was available.

The number of tests completed for each patient and its charges wereobtained from the individual patient sheets. We selected all tests, including test for the diagnosis and/or treatment of hypertension andother condition. Total number of all tests in 6 months was multipliedby 2 and then divided by the total number of patients in 2011, equating11 in the whole hospital. For the hypertensive in-patient group, theaverage number of tests per patient was 16.8, 16.3 and 17.1 in both groups, I10 and I10+E75, respectively. Of course for costing, eachindividual test was priced separately.

Statistical and sensitivity analysis

Means, medians and Inter Quartile Range (IQRs) of costs weremeasured as outcome in this study. Univariate sensitivity analyseswere performed to explore the robustness of the analyses [18].Notably, hypertensive patients may be admitted to hospital with comorbidities.

In this context, it is very important to rule out the costs resulting from diagnosing and treating these other diseases. Therefore, we conductedone sensitivity analysis with a scenario that excluded all costs resulting from diagnosing and treating comorbid diseases. We limitedto specific costs for diagnosing and treating I10 and E75 diseases in

92


this scenario. In addition, the discount rate for capital was varied from 1% to 5%. As data on furniture were not available, these costs werenot included in the base case. Additionally, a scenario where anestimated 5% furniture cost was added to the total capital cost wasexplored [19]. Finally and in the absence of standardized nationalprices, we analyzed sensitivity t o laboratory tests by using laboratorytest charges instead of laboratory test costs as used in the base case.Multivariate sensitivity analyses were also performed to explore thecontribution of two or three parameters at once to the uncertainty in the total costs.

Results

The analyses were based on 230 patients who met the inclusion criteria, including 147 cases with essential (primary) hypertension (I10) and 83 cases with hypertension combined with sphingolipid metabolism and other lipid storage disorders (I10+E75). Mean agewas 64.3 (SD+/-14.7) and 53.5% was female. Characteristics of patients are indicated in table 1. Median length of stay was 6 days with an IQR of 3-8 in the whole study population, and 5 days (IQR: 2-7)and 7 days (IQR: 5-9) in the in the I10 and I10+E75 groups,respectively. Hypertensive patients were admitted in 3 departments ofthe hospital; i.e., Cardiovascular Internal Medicine, Geriatric InternalMedicine and Neurology. For these departments, base- case inpatient–day costs were US$4.99, US$5.05 and US$5.33, respectively.Notably, the most expensive per day costs were associated withpatients admitted to Neurology department.

Chapter 4

Table 1: Characteristic of patientsCharacteristic n = 230 Percentage (%)

Gender

Female 123 53.5Male 107 46.5Patient group

I10 147 64%I10 + E75 83 36%Age (mean ± SD) 64.3 (± 14.7)

Costs of treatment for these hypertensive inpatients are presented in Table 2 as the result of the base case, which considers a 3% discount, uses laboratory test charges and includes no furniture costs. Mediantotal direct costs per patient were US$65 (IQR: 37-95). Total costs perpatient were higher in the combined hypertensive and lipid population (US$78) than in the sole hypertensive patients (US$53). However, themedian costs per day were slightly higher in the sole hypertensivepatients compared to the combined group with estimates of US$11.4 and US$10.9, respectively. In the base-case, inpatient-day costs, at41% of the total cost, represented the highest cost component. Costs ofdrugs followed as second largest cost component with 34%, aspresented in Table 2.

Results of the sensitivity analyses, including 12 scenarios reflectingrelevant alternative options, are presented in Fig 1. In all scenarios,inpatient-day costs appeared to be the most important cost driver in thetotal costs per patient. In the scenario where costs of comorbiddiseases were excluded, the median total direct costs per patient wereUS$64.6 (IQR: 37 -95). Using laboratory testing charges instead ofcosts showed changes in the specific scenarios with increases between15% and 17% (or US$75 to US$76 for total cost) compared to thebase-case. Other scenarios, such as changing discount rates to 1% and5%, adding 5% furniture to the capital costs produced minor impact on the total costs (from minus 0.8 % to 1.1 % changes in total costs).

Chap

ter 4

93


this scenario. In addition, the discount rate for capital was varied from 1% to 5%. As data on furniture were not available, these costs werenot included in the base case. Additionally, a scenario where anestimated 5% furniture cost was added to the total capital cost wasexplored [19]. Finally and in the absence of standardized nationalprices, we analyzed sensitivity t o laboratory tests by using laboratorytest charges instead of laboratory test costs as used in the base case.Multivariate sensitivity analyses were also performed to explore thecontribution of two or three parameters at once to the uncertainty in the total costs.

Results

The analyses were based on 230 patients who met the inclusion criteria, including 147 cases with essential (primary) hypertension (I10) and 83 cases with hypertension combined with sphingolipid metabolism and other lipid storage disorders (I10+E75). Mean agewas 64.3 (SD+/-14.7) and 53.5% was female. Characteristics of patients are indicated in table 1. Median length of stay was 6 days with an IQR of 3-8 in the whole study population, and 5 days (IQR: 2-7)and 7 days (IQR: 5-9) in the in the I10 and I10+E75 groups,respectively. Hypertensive patients were admitted in 3 departments ofthe hospital; i.e., Cardiovascular Internal Medicine, Geriatric InternalMedicine and Neurology. For these departments, base- case inpatient–day costs were US$4.99, US$5.05 and US$5.33, respectively.Notably, the most expensive per day costs were associated withpatients admitted to Neurology department.

Chapter 4

Table 1: Characteristic of patientsCharacteristic n = 230 Percentage (%)

Gender

Female 123 53.5Male 107 46.5Patient group

I10 147 64%I10 + E75 83 36%Age (mean ± SD) 64.3 (± 14.7)

Costs of treatment for these hypertensive inpatients are presented in Table 2 as the result of the base case, which considers a 3% discount, uses laboratory test charges and includes no furniture costs. Mediantotal direct costs per patient were US$65 (IQR: 37-95). Total costs perpatient were higher in the combined hypertensive and lipid population (US$78) than in the sole hypertensive patients (US$53). However, themedian costs per day were slightly higher in the sole hypertensivepatients compared to the combined group with estimates of US$11.4 and US$10.9, respectively. In the base-case, inpatient-day costs, at41% of the total cost, represented the highest cost component. Costs ofdrugs followed as second largest cost component with 34%, aspresented in Table 2.

Results of the sensitivity analyses, including 12 scenarios reflectingrelevant alternative options, are presented in Fig 1. In all scenarios,inpatient-day costs appeared to be the most important cost driver in thetotal costs per patient. In the scenario where costs of comorbiddiseases were excluded, the median total direct costs per patient wereUS$64.6 (IQR: 37 -95). Using laboratory testing charges instead ofcosts showed changes in the specific scenarios with increases between15% and 17% (or US$75 to US$76 for total cost) compared to thebase-case. Other scenarios, such as changing discount rates to 1% and5%, adding 5% furniture to the capital costs produced minor impact on the total costs (from minus 0.8 % to 1.1 % changes in total costs).

94

Tab

le2:

Dire

ctco

sts(

US$

)per

hype

rtens

ivei

npat

ient

bydi

seas

ecl

assi

ficat

ion

and

cost

cate

gory

Cos

tca

tego

ry

Med

ian

(and

IQR

)ofc

osts

Mea

n(±

SD

)ofc

osts

%of

tota

lco

sts

of a

ll pa

tient

s

% o

f to

tal

costs

of I1

0 gr

oup

% o

f to

tal

costs

of I1

0 co

mbi

ne

E75

All

patie

nts

(n=2

30)

I10

grou

p(n

=147

)

I10

com

bine

d E7

5gr

oup

(n=8

3)

All

patie

nts

(n=2

30)

I10

grou

p(n

=147

)

I10

com

bine

dE7

5gr

oup

(n=8

3)

Inpa

tient

Day

s

29.9

4

(14.

97-3

9.92

)

24.9

5

(9.9

8 -3

4.93

)

34.9

3

(24.

95 -

44.9

1)

29.0

5

(±16

.45)

24.8

6

(±15

.94)

36.4

8

(±14

.69)

41.3

740

.01

43.1

4

Dru

gs11

.67

(4.6

3-3

9.76

)

8.26

(3.6

9-7.

02)

16.1

3

(8.4

7 -5

1.47

)

23.8

5

(±25

.4)

21.2

4

(±25

.26)

28.4

5

(±25

.12)

33.9

634

.19

33.6

5

Test

s13

.00

(6.8

9-1

9.24

)

11.5

6

(5.6

2 -1

5.76

)

13.6

9

(7.8

1 -2

5.76

)

15.8

4

(±13

.88)

14.4

9

(±13

.46)

18.2

2

(±14

.37)

22.5

523

.33

21.5

4

Oth

ers

1.23

(0.8

4-1

.68)

1.17

(0.8

2 -1

.56)

1.41

(0.8

5 -1

.83)

1.49

(±2.

26)

1.53

(±2.

79)

1.41

(±0.

6)2.

122.

471.

67

Tota

l cos

ts65

(37-

95)

53

(30

-82)

78

(57

-107

)

70.2

3

(±42

.89)

62.1

3

(±43

.76)

84.5

6

(±37

.46)

100

100

100

Not

es Scen

ario

1:

1%di

scou

ntin

gfo

rcap

italc

osts

Scen

ario

2:

5%di

scou

ntin

gfo

rcap

italc

osts

Scen

ario

3:

Add

ing

a5%

furn

iture

cost

to to

talc

apita

lcos

tsSc

enar

io 4

: U

sing

labo

rato

ryte

st c

harg

es in

stea

d of

cost

pric

esSc

enar

io 5

: 5%

disc

ount

ing

forc

apita

lcos

ts a

nd a

ddin

g a

5%fu

rnitu

reco

st to

tota

lcap

italc

ost

Scen

ario

6:

5%di

scou

ntin

g fo

rcap

italc

osts

and

usi

ngla

bora

tory

test

cha

rges

Scen

ario

7:

1%di

scou

ntin

g fo

rcap

italc

osts

and

add

ing

a5%

furn

iture

cost

to to

talc

apita

lcos

tsSc

enar

io 8

: 1%

disc

ount

ing

forc

apita

lcos

ts a

nd u

sing

labo

rato

ryte

st c

harg

esSc

enar

io 9

: A

ddin

ga

5%fu

rnitu

reco

st to

tota

lcap

italc

osts

and

usi

ngla

bora

tory

test

cha

rges

Scen

ario

10:

1%

disc

ount

ing

forc

apita

lcos

t, ad

ding

a5%

furn

iture

cost

to to

talc

apita

lcos

ts, u

sing

labo

rato

ryte

st c

harg

esSc

enar

io 1

1: 5

%di

scou

ntin

g fo

rcap

italc

ost,

addi

nga

5%fu

rnitu

reco

st to

tota

lcap

italc

osts

and

usin

g la

bora

tory

test

cha

rges

Scen

ario

12:

Sep

arat

ing

out c

osts

resu

lting

from

dia

gnos

ing

and

treat

ing

com

orbi

d di

seas

es

Chap

ter 4

95

Tab

le2:

Dire

ctco

sts(

US$

)per

hype

rtens

ivei

npat

ient

bydi

seas

ecl

assi

ficat

ion

and

cost

cate

gory

Cos

tca

tego

ry

Med

ian

(and

IQR

)ofc

osts

Mea

n(±

SD

)ofc

osts

%of

tota

lco

sts

of a

ll pa

tient

s

% o

f to

tal

costs

of I1

0 gr

oup

% o

f to

tal

costs

of I1

0 co

mbi

ne

E75

All

patie

nts

(n=2

30)

I10

grou

p(n

=147

)

I10

com

bine

d E7

5gr

oup

(n=8

3)

All

patie

nts

(n=2

30)

I10

grou

p(n

=147

)

I10

com

bine

dE7

5gr

oup

(n=8

3)

Inpa

tient

Day

s

29.9

4

(14.

97-3

9.92

)

24.9

5

(9.9

8 -3

4.93

)

34.9

3

(24.

95 -

44.9

1)

29.0

5

(±16

.45)

24.8

6

(±15

.94)

36.4

8

(±14

.69)

41.3

740

.01

43.1

4

Dru

gs11

.67

(4.6

3-3

9.76

)

8.26

(3.6

9-7.

02)

16.1

3

(8.4

7 -5

1.47

)

23.8

5

(±25

.4)

21.2

4

(±25

.26)

28.4

5

(±25

.12)

33.9

634

.19

33.6

5

Test

s13

.00

(6.8

9-1

9.24

)

11.5

6

(5.6

2 -1

5.76

)

13.6

9

(7.8

1 -2

5.76

)

15.8

4

(±13

.88)

14.4

9

(±13

.46)

18.2

2

(±14

.37)

22.5

523

.33

21.5

4

Oth

ers

1.23

(0.8

4-1

.68)

1.17

(0.8

2 -1

.56)

1.41

(0.8

5 -1

.83)

1.49

(±2.

26)

1.53

(±2.

79)

1.41

(±0.

6)2.

122.

471.

67

Tota

l cos

ts65

(37-

95)

53

(30

-82)

78

(57

-107

)

70.2

3

(±42

.89)

62.1

3

(±43

.76)

84.5

6

(±37

.46)

100

100

100

Not

es Scen

ario

1:

1%di

scou

ntin

gfo

rcap

italc

osts

Scen

ario

2:

5%di

scou

ntin

gfo

rcap

italc

osts

Scen

ario

3:

Add

ing

a5%

furn

iture

cost

to to

talc

apita

lcos

tsSc

enar

io 4

: U

sing

labo

rato

ryte

st c

harg

es in

stea

d of

cost

pric

esSc

enar

io 5

: 5%

disc

ount

ing

forc

apita

lcos

ts a

nd a

ddin

g a

5%fu

rnitu

reco

st to

tota

lcap

italc

ost

Scen

ario

6:

5%di

scou

ntin

g fo

rcap

italc

osts

and

usi

ngla

bora

tory

test

cha

rges

Scen

ario

7:

1%di

scou

ntin

g fo

rcap

italc

osts

and

add

ing

a5%

furn

iture

cost

to to

talc

apita

lcos

tsSc

enar

io 8

: 1%

disc

ount

ing

forc

apita

lcos

ts a

nd u

sing

labo

rato

ryte

st c

harg

esSc

enar

io 9

: A

ddin

ga

5%fu

rnitu

reco

st to

tota

lcap

italc

osts

and

usi

ngla

bora

tory

test

cha

rges

Scen

ario

10:

1%

disc

ount

ing

forc

apita

lcos

t, ad

ding

a5%

furn

iture

cost

to to

talc

apita

lcos

ts, u

sing

labo

rato

ryte

st c

harg

esSc

enar

io 1

1: 5

%di

scou

ntin

g fo

rcap

italc

ost,

addi

nga

5%fu

rnitu

reco

st to

tota

lcap

italc

osts

and

usin

g la

bora

tory

test

cha

rges

Scen

ario

12:

Sep

arat

ing

out c

osts

resu

lting

from

dia

gnos

ing

and

treat

ing

com

orbi

d di

seas

es

96


Discussion

Hypertension is a major and increasing public health problem in Vietnam, including 5.7 million patients unaware of their status, 2.1 million aware but untreated, 2.0 million treated but uncontrolled and 1.2million treated and controlled [8]. Notably, the proportion of patients treated and controlled is modest, illustrating a potential for improvement. Vietnam spends huge amounts of money on thesegroups, especially when they need hospitalization for untreated and uncontrolled hypertension. The present study is the first to estimate the direct medical hospital costs of hypertensive inpatients treatment (including ICD-codes I10 or I10combined with E75) in Vietnam. In this study, we both present both mean and median costs in the relevant table.

Our analysis estimates that the median total cost of inpatient treatment and care is US$65 per hypertensive patient per hospital stay. To put this in perspective, the total health expenditures in Vietnam per capita in 2010 was only US$83 [12]. As costs of inpatient treatment ofhypertension are high, less costly options such as programs for earlier detection, treatment and subsequent control may avert these high hospitalization costs. For example, an earlier study demonstrated that cost of drug treatment at a community health station was US$9.4/patient/year for grade 1 hypertension and US$27 /patient/year for grade 2 and 3 hypertension [11]. When comparing those figures to US$53 for hospitalization of uncontrolled sole hypertension in this study, the potential value of early drug treatment over hospitalization of uncontrolled diseases if evident. Hospitalization for uncontrolled hypertensive patients was two times and over five times the costs of drug treatment for grade 2 or 3 and grade 1 hypertension, respectively. The high cost for inpatient care for hypertension in Vietnam is consistent with a previous study in the Philippines, in which the

Chapter 4

lowest median hospitalization costs (US$57) were reported for essential or secondary hypertension among all hypertension-related hospitalization costs [5].

As expected, the costs of the sole hypertensive patients (US$53) were lower than the costs for the combined hypertensive and lipid patients (US$78). It is quite plausible that comorbid conditions can drive up costs, particularly due to increases in length of stay, compared to a single disease condition. In this study, median lengths of stay were 5 and 7 days for I10 and I10 combined with E75, respectively. These inpatients days were associated with high costs, representing 41% of the total cost. This finding differs from the study conducted in the Philippines, which reported medication costs (34% of the total) as the highest cost component [5]. However, this comparison between Vietnam and the Philippines may only have limited validity due to differences in perspectives and methods to estimate costs, differencesin drug prices, and differences in overall investments in health care and health-care systems between both countries.

In this study, the number of inpatient days for sole-hypertensive patients was 5 days. These patients may have undetected and uncontrolled hypertension and now requires hospitalization to be diagnosed for primary or secondary hypertension. In addition, there is no ambulatory blood-pressure monitoring service available in the community. Patients must be admitted to a hospital for diagnosis. As an alternative to expensive hospital care, community level health services may serve as a cheaper option, but these centers need to be strengthened to enable blood-pressure monitoring for the broader population.

The present findings must be interpreted in the context of potential limitations. Firstly, the study was conducted in one hospital, which may not be representative hospitals across the country. However, it is

Chap

ter 4

97


Discussion

Hypertension is a major and increasing public health problem in Vietnam, including 5.7 million patients unaware of their status, 2.1 million aware but untreated, 2.0 million treated but uncontrolled and 1.2million treated and controlled [8]. Notably, the proportion of patients treated and controlled is modest, illustrating a potential for improvement. Vietnam spends huge amounts of money on thesegroups, especially when they need hospitalization for untreated and uncontrolled hypertension. The present study is the first to estimate the direct medical hospital costs of hypertensive inpatients treatment (including ICD-codes I10 or I10combined with E75) in Vietnam. In this study, we both present both mean and median costs in the relevant table.

Our analysis estimates that the median total cost of inpatient treatment and care is US$65 per hypertensive patient per hospital stay. To put this in perspective, the total health expenditures in Vietnam per capita in 2010 was only US$83 [12]. As costs of inpatient treatment ofhypertension are high, less costly options such as programs for earlier detection, treatment and subsequent control may avert these high hospitalization costs. For example, an earlier study demonstrated that cost of drug treatment at a community health station was US$9.4/patient/year for grade 1 hypertension and US$27 /patient/year for grade 2 and 3 hypertension [11]. When comparing those figures to US$53 for hospitalization of uncontrolled sole hypertension in this study, the potential value of early drug treatment over hospitalization of uncontrolled diseases if evident. Hospitalization for uncontrolled hypertensive patients was two times and over five times the costs of drug treatment for grade 2 or 3 and grade 1 hypertension, respectively. The high cost for inpatient care for hypertension in Vietnam is consistent with a previous study in the Philippines, in which the

Chapter 4

lowest median hospitalization costs (US$57) were reported for essential or secondary hypertension among all hypertension-related hospitalization costs [5].

As expected, the costs of the sole hypertensive patients (US$53) were lower than the costs for the combined hypertensive and lipid patients (US$78). It is quite plausible that comorbid conditions can drive up costs, particularly due to increases in length of stay, compared to a single disease condition. In this study, median lengths of stay were 5 and 7 days for I10 and I10 combined with E75, respectively. These inpatients days were associated with high costs, representing 41% of the total cost. This finding differs from the study conducted in the Philippines, which reported medication costs (34% of the total) as the highest cost component [5]. However, this comparison between Vietnam and the Philippines may only have limited validity due to differences in perspectives and methods to estimate costs, differencesin drug prices, and differences in overall investments in health care and health-care systems between both countries.

In this study, the number of inpatient days for sole-hypertensive patients was 5 days. These patients may have undetected and uncontrolled hypertension and now requires hospitalization to be diagnosed for primary or secondary hypertension. In addition, there is no ambulatory blood-pressure monitoring service available in the community. Patients must be admitted to a hospital for diagnosis. As an alternative to expensive hospital care, community level health services may serve as a cheaper option, but these centers need to be strengthened to enable blood-pressure monitoring for the broader population.

The present findings must be interpreted in the context of potential limitations. Firstly, the study was conducted in one hospital, which may not be representative hospitals across the country. However, it is

98


a governmental and not-for-profit hospital that is potentially similar to other hospitals regarding governmental investments in hospital beds [17]. Thus, inpatient-day costs may be extrapolated to other hospitals, which have similar investments with the explicit notice that costs of each patient may be different as different care needs exist among patients. Secondly, we could not identify the grade of hypertension nor the exact type of sphingolipid metabolism or lipid storage disorder so the association between costs and seriousness of disease could not be made at such detailed level, or even at the level of exact blood pressure values. However, others have suggested that the level of systolic or diastolic blood pressure is not a valid predictor for cost outcomes [20]. Thirdly, we may have excluded some patients who were sole hypertensive or had combined lipid disorder but were not identified with I10 (combined) codes. Fourthly and conversely, there may be patients who were miscoded for I10 or E75, and thus incorrectly included in our analysis. We cannot estimate the size of this problem. However, it is reasonable to assume that it is small and likely random and would therefore not introduce specific biases in our study. Furthermore, we calculated tests’ costs based on the assumption the that the total numbers for each test in quarters 3 and 4 would be equal to the total numbers in quarters 1 and 2, with lack of data for the whole year. Notably, some seasonality in testing may have some influence on our results. Finally, while available data did not allow for measurement of cost per day at each department for each disease as well as detailed admission costs and costs of examination at specialist department, we did perform specific analyses assuming price weights, reflecting the average of all patients in the department. However, this implies that the estimate may be artificially under- or overestimatedcosts.

Chapter 4

Conclusions

In comparison to annual medication treatment at a community health station for hypertension and total health expenditure per capita in Vietnam, the costs of hospitalization of hypertensive patients is high. The main driver of the costs is related to inpatient days rather than to treatment, laboratory or other cost categories. Our findings have important implications for health policies. Costs of treatment forhypertension and combined disorders of sphingolipid metabolism and other lipid storage disorders in this study could become the reference case for reimbursement when health insurance companies apply reimbursement by fee for diagnostic-related groups. The findings in this study, particularly the high cost of hospitalization for untreated and uncontrolled hypertension, justify increasing current expenditures by the Vietnamese health-care sector on effective interventions to control high blood pressure, which may produce savings to the health care budget by preventing expensive complications.

Acknowledgement

The research was funded by a project called “Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Vietnam”.

Chap

ter 4

99


a governmental and not-for-profit hospital that is potentially similar to other hospitals regarding governmental investments in hospital beds [17]. Thus, inpatient-day costs may be extrapolated to other hospitals, which have similar investments with the explicit notice that costs of each patient may be different as different care needs exist among patients. Secondly, we could not identify the grade of hypertension nor the exact type of sphingolipid metabolism or lipid storage disorder so the association between costs and seriousness of disease could not be made at such detailed level, or even at the level of exact blood pressure values. However, others have suggested that the level of systolic or diastolic blood pressure is not a valid predictor for cost outcomes [20]. Thirdly, we may have excluded some patients who were sole hypertensive or had combined lipid disorder but were not identified with I10 (combined) codes. Fourthly and conversely, there may be patients who were miscoded for I10 or E75, and thus incorrectly included in our analysis. We cannot estimate the size of this problem. However, it is reasonable to assume that it is small and likely random and would therefore not introduce specific biases in our study. Furthermore, we calculated tests’ costs based on the assumption the that the total numbers for each test in quarters 3 and 4 would be equal to the total numbers in quarters 1 and 2, with lack of data for the whole year. Notably, some seasonality in testing may have some influence on our results. Finally, while available data did not allow for measurement of cost per day at each department for each disease as well as detailed admission costs and costs of examination at specialist department, we did perform specific analyses assuming price weights, reflecting the average of all patients in the department. However, this implies that the estimate may be artificially under- or overestimatedcosts.

Chapter 4

Conclusions

In comparison to annual medication treatment at a community health station for hypertension and total health expenditure per capita in Vietnam, the costs of hospitalization of hypertensive patients is high. The main driver of the costs is related to inpatient days rather than to treatment, laboratory or other cost categories. Our findings have important implications for health policies. Costs of treatment forhypertension and combined disorders of sphingolipid metabolism and other lipid storage disorders in this study could become the reference case for reimbursement when health insurance companies apply reimbursement by fee for diagnostic-related groups. The findings in this study, particularly the high cost of hospitalization for untreated and uncontrolled hypertension, justify increasing current expenditures by the Vietnamese health-care sector on effective interventions to control high blood pressure, which may produce savings to the health care budget by preventing expensive complications.

Acknowledgement

The research was funded by a project called “Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Vietnam”.

100


References

1. Kearney PM, Whelton M, Reynolds K, Muntner P, Whelton PK, He J. Global burden of hypertension: analysis of worldwide data. Lancet. 2005;365(9455):217-23. Epub 2005/01/18. doi: S0140-6736(05)17741-1 [pii] 10.1016/S0140-6736(05)17741-1 [doi]. PubMed PMID: 15652604.

2. Perkovic V, Huxley R, Wu Y, Prabhakaran D, MacMahon S. The burden of blood pressure-related disease: a neglected priority for global health. Hypertension. 2007;50(6):991-7. Epub 2007/10/24. doi: HYPERTENSIONAHA.107.095497 [pii] 10.1161/HYPERTENSIONAHA.107.095497 [doi]. PubMed PMID: 17954719.

3. Gaziano TA, Bitton A, Anand S, Weinstein MC. The global cost of nonoptimal blood pressure. J Hypertens.2009;27(7):1472-7. Epub 2009/05/29. doi: 10.1097/HJH.0b013e32832a9ba3 [doi]. PubMed PMID: 19474763.

4. Le C, Zhankun S, Jun D, Keying Z. The economic burden of hypertension in rural south-west China. Trop Med Int Health. 2012. Epub 2012/09/15. doi: 10.1111/j.1365-3156.2012.03087.x [doi]. PubMed PMID: 22973901.

5. Wagner AK, Valera M, Graves AJ, Lavina S, Ross-Degnan D. Costs of hospital care for hypertension in an insured population without an outpatient medicines benefit: an observational study in the Philippines. BMC Health Serv Res. 2008;8:161. Epub 2008/07/31. doi: 1472-6963-8-161 [pii] 10.1186/1472-6963-8-161 [doi]. PubMed PMID: 18664285; PubMed Central PMCID: PMC2518143.

Chapter 4

6. Ibrahim MM, Damasceno A. Hypertension in developing countries. Lancet. 2012;380(9841):611-9. Epub 2012/08/14. doi: S0140-6736(12)60861-7 [pii] 10.1016/S0140-6736(12)60861-7 [doi]. PubMed PMID: 22883510.


8. Son PT, Quang NN, Viet NL, Khai PG, Wall S, Weinehall L, et al. Prevalence, awareness, treatment and control of hypertension in Vietnam-results from a national survey. J Hum Hypertens. 2012;26(4):268-80. Epub 2011/03/04. doi: jhh201118 [pii] 10.1038/jhh.2011.18 [doi]. PubMed PMID: 21368775.

9. Framingham_Heart_Study. Hard Coronary Heart Disease (10-year risk) 2001 [cited 2013 August 15]. Available from: https://www.framinghamheartstudy.org/risk-functions/coronary-heart-disease/hard-10-year-risk.php.


11. Health_strategy_and_Policy_Institute. Cost effectiveness of hypertensive interventions in Vietnam. Research report. 2011.

Chap

ter 4

101


References

1. Kearney PM, Whelton M, Reynolds K, Muntner P, Whelton PK, He J. Global burden of hypertension: analysis of worldwide data. Lancet. 2005;365(9455):217-23. Epub 2005/01/18. doi: S0140-6736(05)17741-1 [pii] 10.1016/S0140-6736(05)17741-1 [doi]. PubMed PMID: 15652604.

2. Perkovic V, Huxley R, Wu Y, Prabhakaran D, MacMahon S. The burden of blood pressure-related disease: a neglected priority for global health. Hypertension. 2007;50(6):991-7. Epub 2007/10/24. doi: HYPERTENSIONAHA.107.095497 [pii] 10.1161/HYPERTENSIONAHA.107.095497 [doi]. PubMed PMID: 17954719.

3. Gaziano TA, Bitton A, Anand S, Weinstein MC. The global cost of nonoptimal blood pressure. J Hypertens.2009;27(7):1472-7. Epub 2009/05/29. doi: 10.1097/HJH.0b013e32832a9ba3 [doi]. PubMed PMID: 19474763.

4. Le C, Zhankun S, Jun D, Keying Z. The economic burden of hypertension in rural south-west China. Trop Med Int Health. 2012. Epub 2012/09/15. doi: 10.1111/j.1365-3156.2012.03087.x [doi]. PubMed PMID: 22973901.

5. Wagner AK, Valera M, Graves AJ, Lavina S, Ross-Degnan D. Costs of hospital care for hypertension in an insured population without an outpatient medicines benefit: an observational study in the Philippines. BMC Health Serv Res. 2008;8:161. Epub 2008/07/31. doi: 1472-6963-8-161 [pii] 10.1186/1472-6963-8-161 [doi]. PubMed PMID: 18664285; PubMed Central PMCID: PMC2518143.

Chapter 4

6. Ibrahim MM, Damasceno A. Hypertension in developing countries. Lancet. 2012;380(9841):611-9. Epub 2012/08/14. doi: S0140-6736(12)60861-7 [pii] 10.1016/S0140-6736(12)60861-7 [doi]. PubMed PMID: 22883510.


8. Son PT, Quang NN, Viet NL, Khai PG, Wall S, Weinehall L, et al. Prevalence, awareness, treatment and control of hypertension in Vietnam-results from a national survey. J Hum Hypertens. 2012;26(4):268-80. Epub 2011/03/04. doi: jhh201118 [pii] 10.1038/jhh.2011.18 [doi]. PubMed PMID: 21368775.

9. Framingham_Heart_Study. Hard Coronary Heart Disease (10-year risk) 2001 [cited 2013 August 15]. Available from: https://www.framinghamheartstudy.org/risk-functions/coronary-heart-disease/hard-10-year-risk.php.


11. Health_strategy_and_Policy_Institute. Cost effectiveness of hypertensive interventions in Vietnam. Research report. 2011.

102


12. World_bank. Health expenditure, total (% of GDP) 2012 [cited 2012 October 28]. Available from: http://data.worldbank.org/indicator/SH.XPD.TOTL.ZS.

13. World_bank. Population (total) 2010 [cited 2012 October 28]. Available from: http://data.worldbank.org/indicator/SP.POP.TOTL.

14. Ministry_of_Health, Health_Parnership_group. Joint Annual Health Review 2012: Improving quality of Medical services. 2012.

15. Vietcombank. Exchange rate. 2011 [cited 2011 October 1]. Available from: http://www.vietcombank.com.vn/exchangerates/Default.aspx.

16. Tan SS, Rutten FF, van Ineveld BM, Redekop WK, Hakkaart-van Roijen L. Comparing methodologies for the cost estimation of hospital services. Eur J Health Econ. 2009;10(1):39-45. Epub 2008/03/15. doi: 10.1007/s10198-008-0101-x [doi]. PubMed PMID: 18340472.

17. Drummond MF SM, Torrance GW, O'Brien BJ, Stoddart GL. Methods for the Economic Evaluation of Health care programs: Oxford: Oxford University Press; 2005.

18. Andronis L, Barton P, Bryan S. Sensitivity analysis in economic evaluation: an audit of NICE current practice and a review of its use and value in decision-making. Health Technol Assess. 2009;13(29):iii, ix-xi, 1-61. Epub 2009/06/09. doi: 10.3310/hta13290 [doi]. PubMed PMID: 19500484.

19. Nguyen TBY. Economic burden of Shigellosis treatment in Nha Trang, Vietnam. Ha Noi: Ha Noi Univesity of Medicine; 2009.

20. Linjer E, Jornmark J, Hedner T, Jonsson B. Predictors for high costs of hospital care in elderly hypertensive patients. Blood Press. 2006;15(4):245-50. Epub 2006/11/03. PubMed PMID: 17078183.

Chap

ter 4

103


12. World_bank. Health expenditure, total (% of GDP) 2012 [cited 2012 October 28]. Available from: http://data.worldbank.org/indicator/SH.XPD.TOTL.ZS.

13. World_bank. Population (total) 2010 [cited 2012 October 28]. Available from: http://data.worldbank.org/indicator/SP.POP.TOTL.

14. Ministry_of_Health, Health_Parnership_group. Joint Annual Health Review 2012: Improving quality of Medical services. 2012.

15. Vietcombank. Exchange rate. 2011 [cited 2011 October 1]. Available from: http://www.vietcombank.com.vn/exchangerates/Default.aspx.

16. Tan SS, Rutten FF, van Ineveld BM, Redekop WK, Hakkaart-van Roijen L. Comparing methodologies for the cost estimation of hospital services. Eur J Health Econ. 2009;10(1):39-45. Epub 2008/03/15. doi: 10.1007/s10198-008-0101-x [doi]. PubMed PMID: 18340472.

17. Drummond MF SM, Torrance GW, O'Brien BJ, Stoddart GL. Methods for the Economic Evaluation of Health care programs: Oxford: Oxford University Press; 2005.

18. Andronis L, Barton P, Bryan S. Sensitivity analysis in economic evaluation: an audit of NICE current practice and a review of its use and value in decision-making. Health Technol Assess. 2009;13(29):iii, ix-xi, 1-61. Epub 2009/06/09. doi: 10.3310/hta13290 [doi]. PubMed PMID: 19500484.

19. Nguyen TBY. Economic burden of Shigellosis treatment in Nha Trang, Vietnam. Ha Noi: Ha Noi Univesity of Medicine; 2009.

20. Linjer E, Jornmark J, Hedner T, Jonsson B. Predictors for high costs of hospital care in elderly hypertensive patients. Blood Press. 2006;15(4):245-50. Epub 2006/11/03. PubMed PMID: 17078183.

Chapter 5

Utilities of patients with hypertension in Northern Vietnam

Thi-Phuong-Lan Nguyen, Paul F M Krabbe, Thi-Bach-Yen Nguyen, C.C.M. Schuiling-Veninga, E.P. Wright, M.J.Postma

PloS One. 2015 Oct 27;10(10):e0139560. doi: 10.1371/journal.pone.0139560. eCollection 2015

106

Utilities of patients with hypertension

AbstractObjectives

The study aimed to inform cost-effectiveness analysis of hypertension management in Vietnam by providing utilities and predictors of utilities in patients with hypertension.

Methods

Hypertensive patients up to 80 years old visiting the hospital were invited to participate in a survey using Quality Metric’s Short-form 36v2TM

translated into Vietnamese. Health-state utilities were estimated by applying a previously published algorithm.

Results

The mean utility of the 691 patients interviewed was 0.73. Controlling for age, sex, blood pressure (BP) stage, and history of stroke, the utilities in older patients were lower than those in younger ones, and statistically significantly different between the extremes of youngest and oldest groups (p=0.03). Utility in males was higher than in females (p=0.002). Patients with a history of stroke appeared to exhibit lower utilities than patients without such history, but the difference was not statistically significant (p=0.73). Patients with more than three comorbidities did have lower utilities than patients without comorbidity (p=0.01).

Conclusion

Health-state utilities found among hypertensive patients in Vietnamwere similar to those found in other international studies. Statistically significant predictors of lower utility were female gender, age over 70 and having more than three comorbidities, but not BP stage or history of stroke. Our results can be used and may contribute to cost-effectiveness analysis of management of hypertension.

Chapter 5

Introduction

Hypertension and its complications represent key public health problems around the world especially among older adults. More than 40% of hypertensive person are older than 25 years [1]. Every year, 9.4 million deaths occur due to complications of hypertension [2].Globally, diagnosis and treatment are relatively accessible and affordable, but even when patients are on treatment, there are health risks in the long term.

Different outcome measures can be applied to assess the effects of diagnosis and treatment, including health-state utilities, which are core inputs for cost–effectiveness analysis of disease management. Therefore, detailed measurements of utilities of hypertensive patients will support health economic evaluations of planned hypertension management programs. Currently, only limited information on health-state utilities of hypertensive patients is available for Asian countries, including Vietnam.

In many countries, the Short-form 36 version 2TM (SF-36v2) questionnaire, as developed by Quality Metrics, has been applied to collect data on health-related quality of life (HRQoL) [3]. This instrument has also been applied for hypertensive patients in various countries [4-11], enhancing plausibility and acceptability of cross-country comparisons of HRQoL [9,12,13]. In Asia this instrument has already been validated in China, Singapore, Korea and the Philippines [14-17] which enhances plausibility that it is suitable for use in Vietnam.

In this study, we used SF-36v2TM data from patients visiting a hospital out-patient clinic in an urban setting in Northern Vietnam to measure their HRQoL expressed in utilities. Furthermore, we aim to evaluate differences in health-state utilities related to characteristics of these patients to identify potential risk factors.

Chap

ter 5

107


AbstractObjectives

The study aimed to inform cost-effectiveness analysis of hypertension management in Vietnam by providing utilities and predictors of utilities in patients with hypertension.

Methods

Hypertensive patients up to 80 years old visiting the hospital were invited to participate in a survey using Quality Metric’s Short-form 36v2TM

translated into Vietnamese. Health-state utilities were estimated by applying a previously published algorithm.

Results

The mean utility of the 691 patients interviewed was 0.73. Controlling for age, sex, blood pressure (BP) stage, and history of stroke, the utilities in older patients were lower than those in younger ones, and statistically significantly different between the extremes of youngest and oldest groups (p=0.03). Utility in males was higher than in females (p=0.002). Patients with a history of stroke appeared to exhibit lower utilities than patients without such history, but the difference was not statistically significant (p=0.73). Patients with more than three comorbidities did have lower utilities than patients without comorbidity (p=0.01).

Conclusion

Health-state utilities found among hypertensive patients in Vietnamwere similar to those found in other international studies. Statistically significant predictors of lower utility were female gender, age over 70 and having more than three comorbidities, but not BP stage or history of stroke. Our results can be used and may contribute to cost-effectiveness analysis of management of hypertension.

Chapter 5

Introduction

Hypertension and its complications represent key public health problems around the world especially among older adults. More than 40% of hypertensive person are older than 25 years [1]. Every year, 9.4 million deaths occur due to complications of hypertension [2].Globally, diagnosis and treatment are relatively accessible and affordable, but even when patients are on treatment, there are health risks in the long term.

Different outcome measures can be applied to assess the effects of diagnosis and treatment, including health-state utilities, which are core inputs for cost–effectiveness analysis of disease management. Therefore, detailed measurements of utilities of hypertensive patients will support health economic evaluations of planned hypertension management programs. Currently, only limited information on health-state utilities of hypertensive patients is available for Asian countries, including Vietnam.

In many countries, the Short-form 36 version 2TM (SF-36v2) questionnaire, as developed by Quality Metrics, has been applied to collect data on health-related quality of life (HRQoL) [3]. This instrument has also been applied for hypertensive patients in various countries [4-11], enhancing plausibility and acceptability of cross-country comparisons of HRQoL [9,12,13]. In Asia this instrument has already been validated in China, Singapore, Korea and the Philippines [14-17] which enhances plausibility that it is suitable for use in Vietnam.

In this study, we used SF-36v2TM data from patients visiting a hospital out-patient clinic in an urban setting in Northern Vietnam to measure their HRQoL expressed in utilities. Furthermore, we aim to evaluate differences in health-state utilities related to characteristics of these patients to identify potential risk factors.

108


Methods

Setting

We conducted the survey from April to May 2013 among patients attending the hypertension unit of the outpatient clinic in Thai Nguyen General Hospital; further details are described elsewhere [18]. The patients included new and existing cases coming to the outpatient clinic for the management of their hypertension.

Participants

All confirmed hypertensive patients who were able to read and write, were no more than 80 years old and were willing to participate in this study were recruited. We invited patients to take the survey while they were waiting for examination; interviewers collected the data in the outpatient clinic. As one of our purposes was to investigate the relation between utilities and different patient characteristics, only records with full information were included in the analysis. Among 722 cases agreed to provide data, 712 records had full information on patients’ characteristic. Finally, 691 cases were complete and could be used to calculate health utility.

Procedures and measurements

We applied the Quality Metric Short Form, SF-36v2TM, translated into Vietnamese, to collect data on health status. Health-state utilities were estimated by applying the algorithm derived by Brazier et al. [19].This model has previously been shown to explain a reasonable share of variance (approximately 60%) and is considered sensitive in the measurement of impacts on health, especially in those cases when only relatively small differences in health are expected, as in the case of hypertension management. For this model, eleven specific items of SF-36 were selected to estimate health-state utilities: in the physical domain, items number 3, 4 and 12 were included; in the role domain,

Chapter 5

items number 15 and 18; item number 32 in the social domain; items number 21 and 22 in the pain domain; items number 24 and 28 in the mental domain; and item number 27 in the vitality domain. Utilities were subsequently estimated based on responses elicited by standard gambling techniques[19].

During interviews we collected data on patient characteristics such as age, gender, occupation and ethnicity. Blood pressure (BP) was measured once at the time of the medical examination. Other health status parameters were collected from hard copies of medical records, including comorbidities such as diabetes, gout, heart valve disease, peripheral artery disease, chronic arthritis, lipid metabolism disorder, coronary artery disease, other chronic diseases, and history of stroke as a complication of hypertension.


Descriptive listings, including percentages, frequencies, means and SDs for the various groups distinguished, were computed. We also applied independent t-tests, one-way ANOVA test to test the significance of differences in mean utilities between groups. Finally, multiple regression was applied to estimate influences of the individual variables on health-state utilities.

Ethical considerations

This study was approved by the ethical committee of the Thai Nguyen General Hospital. When patients were invited to join the study, the purpose was explained to them and they were told they could choose to join or not, and could stop the interview at any time. Those patients who did agree to provide the information gave verbal consent to the interviewer.

Chap

ter 5

109


Methods

Setting

We conducted the survey from April to May 2013 among patients attending the hypertension unit of the outpatient clinic in Thai Nguyen General Hospital; further details are described elsewhere [18]. The patients included new and existing cases coming to the outpatient clinic for the management of their hypertension.

Participants

All confirmed hypertensive patients who were able to read and write, were no more than 80 years old and were willing to participate in this study were recruited. We invited patients to take the survey while they were waiting for examination; interviewers collected the data in the outpatient clinic. As one of our purposes was to investigate the relation between utilities and different patient characteristics, only records with full information were included in the analysis. Among 722 cases agreed to provide data, 712 records had full information on patients’ characteristic. Finally, 691 cases were complete and could be used to calculate health utility.

Procedures and measurements

We applied the Quality Metric Short Form, SF-36v2TM, translated into Vietnamese, to collect data on health status. Health-state utilities were estimated by applying the algorithm derived by Brazier et al. [19].This model has previously been shown to explain a reasonable share of variance (approximately 60%) and is considered sensitive in the measurement of impacts on health, especially in those cases when only relatively small differences in health are expected, as in the case of hypertension management. For this model, eleven specific items of SF-36 were selected to estimate health-state utilities: in the physical domain, items number 3, 4 and 12 were included; in the role domain,

Chapter 5

items number 15 and 18; item number 32 in the social domain; items number 21 and 22 in the pain domain; items number 24 and 28 in the mental domain; and item number 27 in the vitality domain. Utilities were subsequently estimated based on responses elicited by standard gambling techniques[19].

During interviews we collected data on patient characteristics such as age, gender, occupation and ethnicity. Blood pressure (BP) was measured once at the time of the medical examination. Other health status parameters were collected from hard copies of medical records, including comorbidities such as diabetes, gout, heart valve disease, peripheral artery disease, chronic arthritis, lipid metabolism disorder, coronary artery disease, other chronic diseases, and history of stroke as a complication of hypertension.


Descriptive listings, including percentages, frequencies, means and SDs for the various groups distinguished, were computed. We also applied independent t-tests, one-way ANOVA test to test the significance of differences in mean utilities between groups. Finally, multiple regression was applied to estimate influences of the individual variables on health-state utilities.

Ethical considerations

This study was approved by the ethical committee of the Thai Nguyen General Hospital. When patients were invited to join the study, the purpose was explained to them and they were told they could choose to join or not, and could stop the interview at any time. Those patients who did agree to provide the information gave verbal consent to the interviewer.

110


Results

Main characteristics of patients

The main characteristics of the 712 patients who met the inclusion criteria and joined the study are presented in Table 1. Patients’ ages ranged from 41 to 80 years; half belonged to the group 60-69 years and 56% was female. Only 10% of patients still worked for wages. Among the patients, 66% did not meet their target BP at the time of the survey, while 70% had comorbidities and 11% had a history of stroke.

Table 1. Characteristics of patients included in the study (n=712)

Characteristics Number (%)Age (years)41 – 49 12 (1.7)50 – 59 142 (19.9)60 – 69 367 (51.5)70 – 80 191 (26.8)SexMale 314 (44.1)Working statusBeing retired or doing housework 637 (89.5)Working and earning money 75 (10.5)EthnicityKinh 651 (91.4)Others 61 (8.6)Stage of BP at the time of surveyTarget-BP 242 (34.0)Stage one 360 (50.6)Stage two 110 (15.4)ComorbidityComorbidity(ies) 500 (70.2)StrokeHistory of stroke 78 (11)Health-state utilityHealth-state utility (mean +/- SD (n = 691) 0.73 (+/- 0.14)

Note: Health utility was calculated in 691 patients, as syntax of transferring SF-36 data to utility did not work in 21 cases (3%) with missing value of any one of the items in the SF-36 form.

Chapter 5

Health-state utilities

Mean utility was estimated at 0.73 (+/- 0.14) in this hypertensive population. Mean health-state utilities were significantly different with respect to gender and age groups (all p<0.01). Table 2 shows that relatively higher BP appeared to be associated with lower health-state utility at 0.73, 0.73 and 0.71 in the groups with target BP, stage one and stage two, respectively. However, these differences were not statistically significant (p=0.42). Regarding comorbidities or history of stroke, we did not find significant differences between the mean health-state utilities of these groups (Table 2).

Table 2. Distribution of health-state utilities by patients’ characteristic (n=691)

Characteristics MeanStandard deviation

P value

Age (years)41 – 49 0.787 0.16

0.00450 – 59 0.749 0.1460 – 69 0.730 0.1470 – 80 0.700 0.13Sex

0.005Male 0.743 0.14Female 0.713 0.14Stage of BP at the time of survey

0.422Target-BP 0.734 0.14Stage one 0.726 0.13Stage two 0.712 0.15Comorbidity

0.717No comorbidity 0.729 0.14Comorbidity(ies) 0.725 0.14History of stroke

0.729No stroke 0.727 0.14Stroke 0.721 0.15

Chap

ter 5

111


Results

Main characteristics of patients

The main characteristics of the 712 patients who met the inclusion criteria and joined the study are presented in Table 1. Patients’ ages ranged from 41 to 80 years; half belonged to the group 60-69 years and 56% was female. Only 10% of patients still worked for wages. Among the patients, 66% did not meet their target BP at the time of the survey, while 70% had comorbidities and 11% had a history of stroke.

Table 1. Characteristics of patients included in the study (n=712)

Characteristics Number (%)Age (years)41 – 49 12 (1.7)50 – 59 142 (19.9)60 – 69 367 (51.5)70 – 80 191 (26.8)SexMale 314 (44.1)Working statusBeing retired or doing housework 637 (89.5)Working and earning money 75 (10.5)EthnicityKinh 651 (91.4)Others 61 (8.6)Stage of BP at the time of surveyTarget-BP 242 (34.0)Stage one 360 (50.6)Stage two 110 (15.4)ComorbidityComorbidity(ies) 500 (70.2)StrokeHistory of stroke 78 (11)Health-state utilityHealth-state utility (mean +/- SD (n = 691) 0.73 (+/- 0.14)

Note: Health utility was calculated in 691 patients, as syntax of transferring SF-36 data to utility did not work in 21 cases (3%) with missing value of any one of the items in the SF-36 form.

Chapter 5

Health-state utilities

Mean utility was estimated at 0.73 (+/- 0.14) in this hypertensive population. Mean health-state utilities were significantly different with respect to gender and age groups (all p<0.01). Table 2 shows that relatively higher BP appeared to be associated with lower health-state utility at 0.73, 0.73 and 0.71 in the groups with target BP, stage one and stage two, respectively. However, these differences were not statistically significant (p=0.42). Regarding comorbidities or history of stroke, we did not find significant differences between the mean health-state utilities of these groups (Table 2).

Table 2. Distribution of health-state utilities by patients’ characteristic (n=691)

Characteristics MeanStandard deviation

P value

Age (years)41 – 49 0.787 0.16

0.00450 – 59 0.749 0.1460 – 69 0.730 0.1470 – 80 0.700 0.13Sex

0.005Male 0.743 0.14Female 0.713 0.14Stage of BP at the time of survey

0.422Target-BP 0.734 0.14Stage one 0.726 0.13Stage two 0.712 0.15Comorbidity

0.717No comorbidity 0.729 0.14Comorbidity(ies) 0.725 0.14History of stroke

0.729No stroke 0.727 0.14Stroke 0.721 0.15

112


Controlling for other factors with adjusted ORs, we found that utilities in patients from 70 to 80 years old were lower than those of the younger groups; the difference between the oldest and youngest groups was statistically significant (Table 3; p=0.03). Health-state utility in males was higher than in females even after adjustment for other factors (p=0.002). Patients with a history of stroke exhibited an apparently lower health-state utility than patients without stroke, but the difference was not statistically significant (p=0.79). Patients with more than three comorbidities had lower utilities than patients with no comorbidity, after adjustment.

Table 3. Predictors of health-state utilities in hypertensive patients (n=691)

Coefficients95% CI

PLower bound Upper bound

Age (years)41-49 (ref)50-59 -0.04 -0.12 0.04 0.3460-69 -0.06 -0.13 0.02 0.1670-80 -0.09 -0.17 -0.01 0.03Sex

-0.03 -0.05 -0.01 0.002Male (ref)FemaleBP at the time of surveyTarget-BP (ref)BP Stage one -0.01 -0.03 0.01 0.43BP stage two -0.02 -0.05 0.02 0.32Stroke

-0.004 -0.04 0.03 0.79No history of stroke (ref)History of strokeComorbidityNo comorbidity (ref)One comorbidity 0.01 -0.02 0.03 0.53Two comorbidities -0.004 -0.03 0.03 0.81Three comorbidities -0.04 -0.08 0.02 0.20> Three comorbidities -0.11 -0.18 -0.03 0.01

Chapter 5

Discussion

This study aimed to fill gaps in the evidence on utilities for populations with high BP in low- and middle-income countries, in particular Vietnam. Accurate utility scores are crucial for a planned analysis of cost-effectiveness of disease management.

The mean utilities of our population of hypertensive patients were somewhat lower than the utilities of the general population in Vietnam. For example, the health-state utility of an older population in a rural community in Vietnam, measured by EQ-5D, was 0.88 while the utility in our population was 0.73 [20]. However, this comparison should be interpreted with caution as the studies used different instruments to measure utility [9,12,13]. Still, in comparison with other populations in Asian countries, health-state utility was lower in this hypertensive population. For example, a general population in Singapore had mean health-state utility of 0.80 (+/-0.12), using the same method for health-state utility measurement [21].

Differences in utilities were associated with age, gender, and a high number of comorbidities. However, no such statistically significant differences were found among BP-stages or with presence of history of stroke. Health-state utility among these hypertensive patients differed from an Australian study in which the utility of 0.63 was lower than the utility of 0.73 in our population, using the same instrument and method [22]. However, the population in that study consisted of prisoners and persons with an impaired mental health status, which may explain the lower health-state utilities. Results in other studies vary in health-state utility. For example, health-state utility in a hypertensive population (Beaver Dams study) was 0.72 or 0.83, measured with the Quality of Well-Being (QWB) and time trade-off (TTO) methods, respectively [13]. In a Swedish population with high BP, health-state utility was 0.73 when using a rating scale

Chap

ter 5

113


Controlling for other factors with adjusted ORs, we found that utilities in patients from 70 to 80 years old were lower than those of the younger groups; the difference between the oldest and youngest groups was statistically significant (Table 3; p=0.03). Health-state utility in males was higher than in females even after adjustment for other factors (p=0.002). Patients with a history of stroke exhibited an apparently lower health-state utility than patients without stroke, but the difference was not statistically significant (p=0.79). Patients with more than three comorbidities had lower utilities than patients with no comorbidity, after adjustment.

Table 3. Predictors of health-state utilities in hypertensive patients (n=691)

Coefficients95% CI

PLower bound Upper bound

Age (years)41-49 (ref)50-59 -0.04 -0.12 0.04 0.3460-69 -0.06 -0.13 0.02 0.1670-80 -0.09 -0.17 -0.01 0.03Sex

-0.03 -0.05 -0.01 0.002Male (ref)FemaleBP at the time of surveyTarget-BP (ref)BP Stage one -0.01 -0.03 0.01 0.43BP stage two -0.02 -0.05 0.02 0.32Stroke

-0.004 -0.04 0.03 0.79No history of stroke (ref)History of strokeComorbidityNo comorbidity (ref)One comorbidity 0.01 -0.02 0.03 0.53Two comorbidities -0.004 -0.03 0.03 0.81Three comorbidities -0.04 -0.08 0.02 0.20> Three comorbidities -0.11 -0.18 -0.03 0.01

Chapter 5

Discussion

This study aimed to fill gaps in the evidence on utilities for populations with high BP in low- and middle-income countries, in particular Vietnam. Accurate utility scores are crucial for a planned analysis of cost-effectiveness of disease management.

The mean utilities of our population of hypertensive patients were somewhat lower than the utilities of the general population in Vietnam. For example, the health-state utility of an older population in a rural community in Vietnam, measured by EQ-5D, was 0.88 while the utility in our population was 0.73 [20]. However, this comparison should be interpreted with caution as the studies used different instruments to measure utility [9,12,13]. Still, in comparison with other populations in Asian countries, health-state utility was lower in this hypertensive population. For example, a general population in Singapore had mean health-state utility of 0.80 (+/-0.12), using the same method for health-state utility measurement [21].

Differences in utilities were associated with age, gender, and a high number of comorbidities. However, no such statistically significant differences were found among BP-stages or with presence of history of stroke. Health-state utility among these hypertensive patients differed from an Australian study in which the utility of 0.63 was lower than the utility of 0.73 in our population, using the same instrument and method [22]. However, the population in that study consisted of prisoners and persons with an impaired mental health status, which may explain the lower health-state utilities. Results in other studies vary in health-state utility. For example, health-state utility in a hypertensive population (Beaver Dams study) was 0.72 or 0.83, measured with the Quality of Well-Being (QWB) and time trade-off (TTO) methods, respectively [13]. In a Swedish population with high BP, health-state utility was 0.73 when using a rating scale

114


and 0.81 when using TTO [6]. In a Nigerian hypertensive population, it was 0.35(+/-0.42), using the HUI3 method [23]. Again, these comparisons should be interpreted cautiously because different instruments and methods were applied [6]. Finally, in contrast with the aforementioned Nigerian study, in our study, health-state utilities did not significantly differ for different BP-levels. The methods of analysis between these two studies differed, with BP specified as a continuous variable in the Nigerian study but as a categorical one in our study [23]. In addition, in our study BP was measured only once at the time of examination; we do not know adherence to medication in this population or whether target-BP was achieved.

Opposite to our finding, it might be expected that patients with a history of stroke would have lower utilities than those without such a history [23]. Our finding that history of stroke does not significantly predicts utilities in our study population is however in line with a similarly study in a Swedish population [6]. The reason may be that severity of health impacts and complications after stroke were relatively mild and all patients generally recovered.

In general, older patients had lower utilities than younger ones. When controlling for sex and health status, the only significant difference we found was between the oldest group and youngest group. Obviously and in general, utility decreases with age and in other studies it has indeed been quantified that each year of increasing age was reported to reduce health-state utility [22,23]. In line with other studies, males had 0.03 higher mean health-state utility than females [22,23]. We found an association between comorbidities and utilities, but this association was only manifest for patients having more than three comorbidities. In the Australian study, researchers reported that each increase in the number of comorbidities was associated with reduced utility [22]. It should be noted that our results appeared without

Chapter 5

controlling for other variables that may affect HRQoL, such as body mass index, educational level, smoking, marital status, income/socioeconomic factors, serious events in the past, antihypertensive drugs, household position and awareness of hypertension [5,9,11,20,22,24-27].

Regarding the limitations of the study, it should be mentioned that reliability and validity of applying the equation to obtain utilities from SF-36 in Vietnam should be investigated further, as it was developed for a Western population. We should be aware of potential similar problems as reported for other instruments [28,29]. However, it has already been used in other Asian countries [21,30] and currently does seem the optimal approach, most likely being acceptable for our measurements.

Conclusions

Health-state utilities in a Vietnamese hypertensive population were similar to those found in other international studies. Lower levels of health-state utilities were found among those patients who were older, female or had more than three co-morbidities. The data from this study provide a reference on health-state utility of hypertensive patients in Vietnam as an input for future cost-effectiveness analyses of interventions in such populations.

Acknowledgements

The research was funded by the NUFFIC Vietnam – Netherlands project on Preventive Medicine and a project called “Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Vietnam”. Our thanks go to the Outpatient

Chap

ter 5

115


and 0.81 when using TTO [6]. In a Nigerian hypertensive population, it was 0.35(+/-0.42), using the HUI3 method [23]. Again, these comparisons should be interpreted cautiously because different instruments and methods were applied [6]. Finally, in contrast with the aforementioned Nigerian study, in our study, health-state utilities did not significantly differ for different BP-levels. The methods of analysis between these two studies differed, with BP specified as a continuous variable in the Nigerian study but as a categorical one in our study [23]. In addition, in our study BP was measured only once at the time of examination; we do not know adherence to medication in this population or whether target-BP was achieved.

Opposite to our finding, it might be expected that patients with a history of stroke would have lower utilities than those without such a history [23]. Our finding that history of stroke does not significantly predicts utilities in our study population is however in line with a similarly study in a Swedish population [6]. The reason may be that severity of health impacts and complications after stroke were relatively mild and all patients generally recovered.

In general, older patients had lower utilities than younger ones. When controlling for sex and health status, the only significant difference we found was between the oldest group and youngest group. Obviously and in general, utility decreases with age and in other studies it has indeed been quantified that each year of increasing age was reported to reduce health-state utility [22,23]. In line with other studies, males had 0.03 higher mean health-state utility than females [22,23]. We found an association between comorbidities and utilities, but this association was only manifest for patients having more than three comorbidities. In the Australian study, researchers reported that each increase in the number of comorbidities was associated with reduced utility [22]. It should be noted that our results appeared without

Chapter 5

controlling for other variables that may affect HRQoL, such as body mass index, educational level, smoking, marital status, income/socioeconomic factors, serious events in the past, antihypertensive drugs, household position and awareness of hypertension [5,9,11,20,22,24-27].

Regarding the limitations of the study, it should be mentioned that reliability and validity of applying the equation to obtain utilities from SF-36 in Vietnam should be investigated further, as it was developed for a Western population. We should be aware of potential similar problems as reported for other instruments [28,29]. However, it has already been used in other Asian countries [21,30] and currently does seem the optimal approach, most likely being acceptable for our measurements.

Conclusions

Health-state utilities in a Vietnamese hypertensive population were similar to those found in other international studies. Lower levels of health-state utilities were found among those patients who were older, female or had more than three co-morbidities. The data from this study provide a reference on health-state utility of hypertensive patients in Vietnam as an input for future cost-effectiveness analyses of interventions in such populations.

Acknowledgements

The research was funded by the NUFFIC Vietnam – Netherlands project on Preventive Medicine and a project called “Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Vietnam”. Our thanks go to the Outpatient

116


Department at Thai Nguyen General Hospital for their co-operation during the study. We also thank Professor John E Brazier for providing the algorithm to transfer SF-36v2TM to health-state utilities score.

Chapter 5

References

1. WHO. A global brief on Hypertension - Silent killer, global public health crisis. 2013; [cited 30 December 2014]. Available: http://apps.who.int/iris/bitstream/10665/79059/1/WHO_DCO_WHD_2013.2_eng.pdf.

2. Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, et al. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012; 380: 2224-2260.

3. John E. Ware J. SF-36® Health Survey Update. 2013; [cited 30 December 2014]. Available: http://www.sf-36.org/tools/SF36.shtml.

4. Aydemir O, Ozdemir C, Koroglu E The impact of co-morbid conditions on the SF-36: a primary-care-based study among hypertensives. Arch Med Res. 2005; 36: 136-141.

5. Banegas JR, Guallar-Castillon P, Rodriguez-Artalejo F, Graciani A, Lopez-Garcia E, et al. Association between awareness, treatment, and control of hypertension, and quality of life among older adults in Spain. Am J Hypertens. 2006; 19: 686-693.

6. Bardage C, Isacson D, Ring L, Bingefors K A Swedish population-based study on the relationship between the SF-36and health utilities to measure health in hypertension. Blood Press. 2003; 12: 203-210.

7. Bardage C, Isacson DG Hypertension and health-related quality of life. an epidemiological study in Sweden. J Clin Epidemiol. 2001; 54: 172-181.

Chap

ter 5

117


Department at Thai Nguyen General Hospital for their co-operation during the study. We also thank Professor John E Brazier for providing the algorithm to transfer SF-36v2TM to health-state utilities score.

Chapter 5

References

1. WHO. A global brief on Hypertension - Silent killer, global public health crisis. 2013; [cited 30 December 2014]. Available: http://apps.who.int/iris/bitstream/10665/79059/1/WHO_DCO_WHD_2013.2_eng.pdf.

2. Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, et al. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012; 380: 2224-2260.

3. John E. Ware J. SF-36® Health Survey Update. 2013; [cited 30 December 2014]. Available: http://www.sf-36.org/tools/SF36.shtml.

4. Aydemir O, Ozdemir C, Koroglu E The impact of co-morbid conditions on the SF-36: a primary-care-based study among hypertensives. Arch Med Res. 2005; 36: 136-141.

5. Banegas JR, Guallar-Castillon P, Rodriguez-Artalejo F, Graciani A, Lopez-Garcia E, et al. Association between awareness, treatment, and control of hypertension, and quality of life among older adults in Spain. Am J Hypertens. 2006; 19: 686-693.

6. Bardage C, Isacson D, Ring L, Bingefors K A Swedish population-based study on the relationship between the SF-36and health utilities to measure health in hypertension. Blood Press. 2003; 12: 203-210.

7. Bardage C, Isacson DG Hypertension and health-related quality of life. an epidemiological study in Sweden. J Clin Epidemiol. 2001; 54: 172-181.

118


8. Gusmao JL, Mion D, Jr., Pierin AM Health-related quality of life and blood pressure control in hypertensive patients with and without complications. Clinics (Sao Paulo). 2009; 64: 619-628.

9. Lawrence WF, Fryback DG, Martin PA, Klein R, Klein BE Health status and hypertension: a population-based study. J Clin Epidemiol. 1996; 49: 1239-1245.

10. Mena-Martin FJ, Martin-Escudero JC, Simal-Blanco F, Carretero-Ares JL, Arzua-Mouronte D, et al. Health-related quality of life of subjects with known and unknown hypertension: results from the population-based Hortega study. J Hypertens. 2003; 21: 1283-1289.

11. Li W, Liu L, Puente JG, Li Y, Jiang X, et al. Hypertension and health-related quality of life: an epidemiological study in patients attending hospital clinics in China. J Hypertens. 2005; 23: 1667-1676.

12. Kaplan RM, Groessl EJ, Sengupta N, Sieber WJ, Ganiats TG Comparison of measured utility scores and imputed scores from the SF-36 in patients with rheumatoid arthritis. Med Care. 2005; 43: 79-87.

13. Fryback DG, Dasbach EJ, Klein R, Klein BE, Dorn N, et al. The Beaver Dam Health Outcomes Study: initial catalog of health-state quality factors. Med Decis Making. 1993; 13: 89-102.

14. Zhou KN, Zhang M, Wu Q, Ji ZH, Zhang XM, et al. Reliability, validity and sensitivity of the Chinese (simple) short form 36 health survey version 2 (SF-36v2) in patients with chronic hepatitis B. J Viral Hepat. 2013; 20: e47-55.

Chapter 5

15. Thumboo J, Wu Y, Tai ES, Gandek B, Lee J, et al. Reliability and validity of the English (Singapore) and Chinese (Singapore) versions of the Short-Form 36 version 2 in a multi-ethnic urban Asian population in Singapore. Qual Life Res. 2013; 22: 2501-2508.

16. Kim SH, Jo MW, Lee SI Psychometric properties of the Korean short form-36 health survey version 2 for assessing the general population. Asian Nurs Res (Korean Soc Nurs Sci). 2013; 7: 61-66.

17. Castillo-Carandang NT, Sison OT, Grefal ML, Sy RG, Alix OC, et al. A community-based validation study of the short-form 36 version 2 Philippines (Tagalog) in two cities in the Philippines. PLoS One. 2013; 8: e83794.

18. Nguyen TP, Nguyen TB, Nguyen TT, Vinh Hac V, Le HH, et al. Direct costs of hypertensive patients admitted to hospital in Vietnam- a bottom-up micro-costing analysis. BMC Health Serv Res. 2014; 14: 514.

19. Brazier J, Roberts J, Deverill M The estimation of a preference-based measure of health from the SF-36. J Health Econ. 2002; 21: 271-292.

20. Hoi le V, Chuc NT, Lindholm L Health-related quality of life, and its determinants, among older people in rural Vietnam. BMC Public Health. 2010; 10: 549.

21. Wee HL, Cheung YB, Fong KY, Luo N, Machin D, et al. Are English- and Chinese-language versions of the SF-6D equivalent? A comparison from a population-based study. Clin Ther. 2004; 26: 1137-1148.

22. Chong CA, Li S, Nguyen GC, Sutton A, Levy MH, et al. Health-state utilities in a prisoner population: a cross-sectional survey. Health Qual Life Outcomes. 2009; 7: 78.

Chap

ter 5

119


8. Gusmao JL, Mion D, Jr., Pierin AM Health-related quality of life and blood pressure control in hypertensive patients with and without complications. Clinics (Sao Paulo). 2009; 64: 619-628.

9. Lawrence WF, Fryback DG, Martin PA, Klein R, Klein BE Health status and hypertension: a population-based study. J Clin Epidemiol. 1996; 49: 1239-1245.

10. Mena-Martin FJ, Martin-Escudero JC, Simal-Blanco F, Carretero-Ares JL, Arzua-Mouronte D, et al. Health-related quality of life of subjects with known and unknown hypertension: results from the population-based Hortega study. J Hypertens. 2003; 21: 1283-1289.

11. Li W, Liu L, Puente JG, Li Y, Jiang X, et al. Hypertension and health-related quality of life: an epidemiological study in patients attending hospital clinics in China. J Hypertens. 2005; 23: 1667-1676.

12. Kaplan RM, Groessl EJ, Sengupta N, Sieber WJ, Ganiats TG Comparison of measured utility scores and imputed scores from the SF-36 in patients with rheumatoid arthritis. Med Care. 2005; 43: 79-87.

13. Fryback DG, Dasbach EJ, Klein R, Klein BE, Dorn N, et al. The Beaver Dam Health Outcomes Study: initial catalog of health-state quality factors. Med Decis Making. 1993; 13: 89-102.

14. Zhou KN, Zhang M, Wu Q, Ji ZH, Zhang XM, et al. Reliability, validity and sensitivity of the Chinese (simple) short form 36 health survey version 2 (SF-36v2) in patients with chronic hepatitis B. J Viral Hepat. 2013; 20: e47-55.

Chapter 5

15. Thumboo J, Wu Y, Tai ES, Gandek B, Lee J, et al. Reliability and validity of the English (Singapore) and Chinese (Singapore) versions of the Short-Form 36 version 2 in a multi-ethnic urban Asian population in Singapore. Qual Life Res. 2013; 22: 2501-2508.

16. Kim SH, Jo MW, Lee SI Psychometric properties of the Korean short form-36 health survey version 2 for assessing the general population. Asian Nurs Res (Korean Soc Nurs Sci). 2013; 7: 61-66.

17. Castillo-Carandang NT, Sison OT, Grefal ML, Sy RG, Alix OC, et al. A community-based validation study of the short-form 36 version 2 Philippines (Tagalog) in two cities in the Philippines. PLoS One. 2013; 8: e83794.

18. Nguyen TP, Nguyen TB, Nguyen TT, Vinh Hac V, Le HH, et al. Direct costs of hypertensive patients admitted to hospital in Vietnam- a bottom-up micro-costing analysis. BMC Health Serv Res. 2014; 14: 514.

19. Brazier J, Roberts J, Deverill M The estimation of a preference-based measure of health from the SF-36. J Health Econ. 2002; 21: 271-292.

20. Hoi le V, Chuc NT, Lindholm L Health-related quality of life, and its determinants, among older people in rural Vietnam. BMC Public Health. 2010; 10: 549.

21. Wee HL, Cheung YB, Fong KY, Luo N, Machin D, et al. Are English- and Chinese-language versions of the SF-6D equivalent? A comparison from a population-based study. Clin Ther. 2004; 26: 1137-1148.

22. Chong CA, Li S, Nguyen GC, Sutton A, Levy MH, et al. Health-state utilities in a prisoner population: a cross-sectional survey. Health Qual Life Outcomes. 2009; 7: 78.

120


23. Ekwunife OI, Aguwa CN, Adibe MO, Barikpaoar E, Onwuka CJ Health state utilities of a population of Nigerian hypertensive patients. BMC Res Notes. 2011; 4: 528.

24. Sazlina SG, Zaiton A, Nor Afiah MZ, Hayati KS Predictors of health related quality of life in older people with non-communicable diseases attending three primary care clinics in Malaysia. J Nutr Health Aging. 2012; 16: 498-502.

25. Konerman M, Weeks KR, Shands JR, Tilburt JC, Dy S, et al. Short Form (SF-36) Health Survey measures are associated with decreased adherence among urban African Americans with severe, poorly controlled hypertension. J Clin Hypertens (Greenwich). 2011; 13: 385-390.

26. Qian Y, Zhang J, Lin Y, Dong M, Xu M, et al. A tailored target intervention on influence factors of quality of life in Chinese patients with hypertension. Clin Exp Hypertens. 2009; 31: 71-82.

27. Carvalho MV, Siqueira LB, Sousa AL, Jardim PC The influence of hypertension on quality of life. Arq Bras Cardiol. 2013; 100: 164-174.

28. Johnson JA, Luo N, Shaw JW, Kind P, Coons SJ Valuations of EQ-5D health states: are the United States and United Kingdom different? Med Care. 2005; 43: 221-228.

29. Havranek EP, Steiner JF Valuation of health states in the US versus the UK: two measures divided by a common language? Med Care. 2005; 43: 201-202.

30. Leger D, Morin CM, Uchiyama M, Hakimi Z, Cure S, et al. Chronic insomnia, quality-of-life, and utility scores: comparison with good sleepers in a cross-sectional international survey. Sleep Med. 2012; 13: 43-51.

Chap

ter 5

121


23. Ekwunife OI, Aguwa CN, Adibe MO, Barikpaoar E, Onwuka CJ Health state utilities of a population of Nigerian hypertensive patients. BMC Res Notes. 2011; 4: 528.

24. Sazlina SG, Zaiton A, Nor Afiah MZ, Hayati KS Predictors of health related quality of life in older people with non-communicable diseases attending three primary care clinics in Malaysia. J Nutr Health Aging. 2012; 16: 498-502.

25. Konerman M, Weeks KR, Shands JR, Tilburt JC, Dy S, et al. Short Form (SF-36) Health Survey measures are associated with decreased adherence among urban African Americans with severe, poorly controlled hypertension. J Clin Hypertens (Greenwich). 2011; 13: 385-390.

26. Qian Y, Zhang J, Lin Y, Dong M, Xu M, et al. A tailored target intervention on influence factors of quality of life in Chinese patients with hypertension. Clin Exp Hypertens. 2009; 31: 71-82.

27. Carvalho MV, Siqueira LB, Sousa AL, Jardim PC The influence of hypertension on quality of life. Arq Bras Cardiol. 2013; 100: 164-174.

28. Johnson JA, Luo N, Shaw JW, Kind P, Coons SJ Valuations of EQ-5D health states: are the United States and United Kingdom different? Med Care. 2005; 43: 221-228.

29. Havranek EP, Steiner JF Valuation of health states in the US versus the UK: two measures divided by a common language? Med Care. 2005; 43: 201-202.

30. Leger D, Morin CM, Uchiyama M, Hakimi Z, Cure S, et al. Chronic insomnia, quality-of-life, and utility scores: comparison with good sleepers in a cross-sectional international survey. Sleep Med. 2012; 13: 43-51.

Chapter 6

Cost-effectiveness analysis of screening for and managing identified

hypertension for cardiovascular disease prevention in Vietnam

Thi-Phuong-Lan Nguyen, E.P. Wright, Thanh-Trung Nguyen, C.C.M. Schuiling-Veninga, M.J. Bijlsma, Thi-Bach-Yen Nguyen,

M.J. Postma

Plos One. May 18, 2016. http://dx.doi.org/10.1371/journal.pone.0155699

124

Cost-effectiveness analysis of screening for and managing identified hypertension

Abstract

Objectives

To inform development of guidelines for hypertension management in Vietnam, we evaluated the cost-effectiveness of different strategies on screening for hypertension in preventing cardiovascular disease (CVD).

Methods

A decision tree was combined with a Markov model to measure incremental cost-effectiveness of different approaches to hypertension screening. Values used as input parameters for the model were taken from different sources. Various screening intervals (one-off, annually, biannually) and starting ages to screen (35, 45 or 55 years) and coverage of treatment were analysed. We ran both a ten-year and a lifetime horizon. Input parameters for the models were extracted from local and regional data. Probabilistic sensitivity analysis was used to evaluate parameter uncertainty. A threshold of three times GDP per capita was applied.

Results

Cost per quality adjusted life year (QALY) gained varied in different screening scenarios. In a ten-year horizon, the cost-effectiveness of screening for hypertension ranged from cost saving to Int$ 758,695 per QALY gained. For screening of men starting at 55 years, all screening scenarios gave a high probability of being cost-effective. For screening of females starting at 55 years, the probability of favourable cost-effectiveness was 90% with one-off screening. In a lifetime horizon, cost per QALY gained was lower than the threshold

Chapter 6

of Int$ 15,883 in all screening scenarios among males. Similar results were found in females when starting screening at 55 years. Starting screening in females at 45 years had a high probability of being cost-effective if screening biannually was combined with increasing coverage of treatment by 20% or even if sole biannual screening was considered.

Conclusion

From a health economic perspective, integrating screening for hypertension into routine medical examination and related coverage by health insurance could be recommended. Screening for hypertension has a high probability of being cost-effective in preventing CVD. An adequate screening strategy can best be selected based on age, sex and screening interval.

Chap

ter 6

125


Abstract

Objectives

To inform development of guidelines for hypertension management in Vietnam, we evaluated the cost-effectiveness of different strategies on screening for hypertension in preventing cardiovascular disease (CVD).

Methods

A decision tree was combined with a Markov model to measure incremental cost-effectiveness of different approaches to hypertension screening. Values used as input parameters for the model were taken from different sources. Various screening intervals (one-off, annually, biannually) and starting ages to screen (35, 45 or 55 years) and coverage of treatment were analysed. We ran both a ten-year and a lifetime horizon. Input parameters for the models were extracted from local and regional data. Probabilistic sensitivity analysis was used to evaluate parameter uncertainty. A threshold of three times GDP per capita was applied.

Results

Cost per quality adjusted life year (QALY) gained varied in different screening scenarios. In a ten-year horizon, the cost-effectiveness of screening for hypertension ranged from cost saving to Int$ 758,695 per QALY gained. For screening of men starting at 55 years, all screening scenarios gave a high probability of being cost-effective. For screening of females starting at 55 years, the probability of favourable cost-effectiveness was 90% with one-off screening. In a lifetime horizon, cost per QALY gained was lower than the threshold

Chapter 6

of Int$ 15,883 in all screening scenarios among males. Similar results were found in females when starting screening at 55 years. Starting screening in females at 45 years had a high probability of being cost-effective if screening biannually was combined with increasing coverage of treatment by 20% or even if sole biannual screening was considered.

Conclusion

From a health economic perspective, integrating screening for hypertension into routine medical examination and related coverage by health insurance could be recommended. Screening for hypertension has a high probability of being cost-effective in preventing CVD. An adequate screening strategy can best be selected based on age, sex and screening interval.

126


Introduction

Similar to the trend in global burden of disease, hypertension is a leading cause of cardiovascular disease (CVD) in Vietnam [1, 2].According to Vietnam data in the Global Burden of Disease (GBD) report in 2013, cerebrovascular disease accounted for 9.7% of total disability adjusted life years (DALYs), of which hypertension contributed 57%. Ischemic heart disease (IHD) accounted for 2.4% of total DALYs, 51% of which is attributed to hypertension [2].Treatment of hypertension is known to be effective in reducing the burden of CVD [3, 4]. A high prevalence of undiagnosed hypertension is an obstacle to treatment and prevention of complications. Globally, the prevalence of undiagnosed hypertension is 53% [5]; Vietnam is similar at 52% [6].

To increase knowledge and awareness of hypertension, early detection by measuring blood pressure (BP) is recommended [1]. Several studies on screening for hypertension and treatment in populations at risk, because of history of CVD, high cholesterol, diabetes or age, have demonstrated the cost-effectiveness of this approach in preventing cardiovascular or kidney disease [7-11]. Although there are other risk factors for CVD such as diabetes and high cholesterol, high blood pressure (HBP) may be considered the logical first focus, with age and sex as obvious potentially key guiding factors to the hypertension risk. Notably, HBP as an essential major CVD risk factor is incorporated into all risk predictors, potentially making BP screening the most essential factor for risk assessment in general.

WHO recommends that all adults check their BP regularly [1], either annually or every two years, depending on exact previous levels of BP [12]. However, in a limited-resource setting like Vietnam, it is not possible to screen everyone and it would be potentially too expensive

Chapter 6

to start screening for hypertension from 18 years of age. After many years of a focus on infectious diseases and safe motherhood, the Ministry of Health is currently developing plans to deal with non-communicable diseases, which have been increasing in the past decades. The Ministry of Health in Vietnam would like to integrate screening for hypertension into routine medical examination at all levels, especially at community health stations (CHS), and has proposed that these services will be covered by health insurance [13].Therefore, policymakers and planners now urgently need evidence to guide the choices in their guidelines and recommendations. For example, policy makers may plan to postpone screening for hypertension to an older age when risk is higher, for example from the age of 40 [13]. Up to now, however, there is little evidence on the cost-effectiveness of screening for hypertension in developing countries. Previous studies have not provided sufficient evidence on the best screening strategies regarding screening intervals, or specific ages and sexes to be targeted. Information on the most effective strategies for low resource settings is scarce.

To meet that demand, we conducted a study to quantify quality-adjusted life years (QALYs) and incremental cost-effectiveness of the following approaches: (1) no screening; (2) one-off screening; (3) screening every two years;(4) annual screening; and (5) screening in combination with increased coverage of treatment in both sexes and different ages.

Methods

Model

The model combines a decision tree with a Markov model to estimate the cost-effectiveness of screening compared with non-screening for

Chap

ter 6

127


Introduction

Similar to the trend in global burden of disease, hypertension is a leading cause of cardiovascular disease (CVD) in Vietnam [1, 2].According to Vietnam data in the Global Burden of Disease (GBD) report in 2013, cerebrovascular disease accounted for 9.7% of total disability adjusted life years (DALYs), of which hypertension contributed 57%. Ischemic heart disease (IHD) accounted for 2.4% of total DALYs, 51% of which is attributed to hypertension [2].Treatment of hypertension is known to be effective in reducing the burden of CVD [3, 4]. A high prevalence of undiagnosed hypertension is an obstacle to treatment and prevention of complications. Globally, the prevalence of undiagnosed hypertension is 53% [5]; Vietnam is similar at 52% [6].

To increase knowledge and awareness of hypertension, early detection by measuring blood pressure (BP) is recommended [1]. Several studies on screening for hypertension and treatment in populations at risk, because of history of CVD, high cholesterol, diabetes or age, have demonstrated the cost-effectiveness of this approach in preventing cardiovascular or kidney disease [7-11]. Although there are other risk factors for CVD such as diabetes and high cholesterol, high blood pressure (HBP) may be considered the logical first focus, with age and sex as obvious potentially key guiding factors to the hypertension risk. Notably, HBP as an essential major CVD risk factor is incorporated into all risk predictors, potentially making BP screening the most essential factor for risk assessment in general.

WHO recommends that all adults check their BP regularly [1], either annually or every two years, depending on exact previous levels of BP [12]. However, in a limited-resource setting like Vietnam, it is not possible to screen everyone and it would be potentially too expensive

Chapter 6

to start screening for hypertension from 18 years of age. After many years of a focus on infectious diseases and safe motherhood, the Ministry of Health is currently developing plans to deal with non-communicable diseases, which have been increasing in the past decades. The Ministry of Health in Vietnam would like to integrate screening for hypertension into routine medical examination at all levels, especially at community health stations (CHS), and has proposed that these services will be covered by health insurance [13].Therefore, policymakers and planners now urgently need evidence to guide the choices in their guidelines and recommendations. For example, policy makers may plan to postpone screening for hypertension to an older age when risk is higher, for example from the age of 40 [13]. Up to now, however, there is little evidence on the cost-effectiveness of screening for hypertension in developing countries. Previous studies have not provided sufficient evidence on the best screening strategies regarding screening intervals, or specific ages and sexes to be targeted. Information on the most effective strategies for low resource settings is scarce.

To meet that demand, we conducted a study to quantify quality-adjusted life years (QALYs) and incremental cost-effectiveness of the following approaches: (1) no screening; (2) one-off screening; (3) screening every two years;(4) annual screening; and (5) screening in combination with increased coverage of treatment in both sexes and different ages.

Methods

Model

The model combines a decision tree with a Markov model to estimate the cost-effectiveness of screening compared with non-screening for

128


hypertension. Various intervals for screening (one-off, annually, biannually, biannually until 55 or 60 years old and then annually until death) and varying ages to start screening (35 or 45 or 55 years old), during a one-year cycle were applied and are presented in Table 1. The description of screening strategies is presented in the appendix 2.

Table 1. Intervention scenarios and time horizon.

Scenario Description

Ten-years horizon

No No screening, treatment

One-off One-off screening in the first year, treatment

E1 Annual screening and treatment

E2 Biannual screening, treatment

E1&T.20% Annual screening and increase coverage of treatment by 20%

E2&T.20% Biannual screening and increase coverage of treatment by 20%

Lifetime horizon


E1 Annual screening, treatment


E2 until

55+ E1

Biannual screening until 55 years, then annual screening

until death and treatment

E2 until

60+ E1





Notes: 48%, 62% treatment among hypertension were applied in this study in male

and female, respectively

In the model, for either screening or non-screening scenarios, people are divided into three groups: treated hypertension, untreated hypertension and healthy (non-hypertensive). In the Markov model, patients start in the initial hypertension state. Patients can remain in

Chapter 6

this state or move to either acute CVD or CVD/non-CVD death. At the end of each cycle in the acute CVD state, patients can move to stable CVD or CVD/non-CVD death state, or may experience recurrent CVD events and then stay in the same state. From the stable CVD state, patients can experience CVD/non-CVD death or stay in the same health state, or they may have a recurrence of CVD and move to acute CVD. Fig 1 illustrates the complete model.

We defined a composite CVD-outcome that included myocardial infarction (ICD-10 code I21) and cerebrovascular diseases (ICD-10codes I60 to I66). Data on the incidence of myocardial infarction (MI) and of cerebrovascular disease in Vietnam were not available to weight the overall CVD. To overcome this limitation, data from a meta-analysis among hypertensive patients in an Asian population were used to design a composite CVD-outcome comprising 78% stroke (cerebrovascular diseases) and 22% MI [3]. Notably, this distribution was used to weight relative risk, costs and utilities in the overall composite CVD.

Fig 1A: Structure of decision tree on screening and treatment for hypertension

Population

No

Screening

HBP & treatment

HBP & non-treatment

Healthy

HBP & treatment

Healthy

HBP & non-treatment

Chap

ter 6

129


hypertension. Various intervals for screening (one-off, annually, biannually, biannually until 55 or 60 years old and then annually until death) and varying ages to start screening (35 or 45 or 55 years old), during a one-year cycle were applied and are presented in Table 1. The description of screening strategies is presented in the appendix 2.

Table 1. Intervention scenarios and time horizon.

Scenario Description

Ten-years horizon


One-off One-off screening in the first year, treatment

E1 Annual screening and treatment




Lifetime horizon


E1 Annual screening, treatment


E2 until

55+ E1



E2 until

60+ E1





Notes: 48%, 62% treatment among hypertension were applied in this study in male

and female, respectively

In the model, for either screening or non-screening scenarios, people are divided into three groups: treated hypertension, untreated hypertension and healthy (non-hypertensive). In the Markov model, patients start in the initial hypertension state. Patients can remain in

Chapter 6

this state or move to either acute CVD or CVD/non-CVD death. At the end of each cycle in the acute CVD state, patients can move to stable CVD or CVD/non-CVD death state, or may experience recurrent CVD events and then stay in the same state. From the stable CVD state, patients can experience CVD/non-CVD death or stay in the same health state, or they may have a recurrence of CVD and move to acute CVD. Fig 1 illustrates the complete model.

We defined a composite CVD-outcome that included myocardial infarction (ICD-10 code I21) and cerebrovascular diseases (ICD-10codes I60 to I66). Data on the incidence of myocardial infarction (MI) and of cerebrovascular disease in Vietnam were not available to weight the overall CVD. To overcome this limitation, data from a meta-analysis among hypertensive patients in an Asian population were used to design a composite CVD-outcome comprising 78% stroke (cerebrovascular diseases) and 22% MI [3]. Notably, this distribution was used to weight relative risk, costs and utilities in the overall composite CVD.

Fig 1A: Structure of decision tree on screening and treatment for hypertension

Population

No

Screening

HBP & treatment

HBP & non-treatment

Healthy

HBP & treatment

Healthy

HBP & non-treatment

130


Fig 1B: Different health states in the Markov model

Fig 1: Decision tree and Markov model for estimating cost-effectiveness of screening for hypertension

Notes: HBP: high blood pressure; CVD: cardiovascular disease. Patients start in the initial hypertension state. Patients can remain in this state or move to either acute CVD or CVD/non-CVD death. From the acute CVD state, patients can move to stable CVD or CVD/non-CVD death state, or may experience recurrent CVD events. From the stable CVD state, patients may stay in the same health state, or they may have a recurrence of CVD or can move to CVD/non-CVD death.

Screening and hypertension management

We previously conducted community screening for hypertension among currently untreated, undiagnosed adults aged 35-64 years, measuring BP during two visits [14]. Individual cases with BP in the hypertensive range were confirmed by a doctor at the CHS. The details of this field work were described previously [14]. For the non-screening scenario/current practice, patients are assumed to visit the CHS for diagnosis and management when they have symptoms of hypertension.

Diagnosed patients in both the screening and non-screening scenarios were assumed to be receiving treatment for hypertension at the CHS.

Acute CVD

Stable CVDHBP

CVD and non-CVD

Chapter 6

Antihypertensive drugs would be prescribed according to the Ministry of Health guidelines [15]. We assumed that patients with CVD were referred to a hospital for treatment.

Prevalence and incidence of hypertension

In the screening scenario, total prevalence, stratified according to different age groups and sex, was quantified by prevalence of hypertension detected by screening plus prevalence diagnosed by symptoms. Prevalence of current treatment, after diagnosis through symptoms, was recalculated from a national survey on a sex-specific basis [6]. Among hypertensive patients, we estimated the prevalence of patients going or not going into treatment in accordance with a previous study [6]. The number of patients in the group going into treatment was aggregated with those currently under treatment, and the number of hypertensive patients detected by screening multiplied by the percentage on treatment among diagnosed patients. Coverage of treatment reflects the percentage of diagnosed patients who received treatment, additionally assuming that patients classified as on treatment strictly comply with therapy. Although in reality not all patients may be strictly adherent, we did not have reliable data to use as inputs in our analysis. Also, we could not find appropriate evidence on, for example, the association of adherence level and cardiovascular events, particularly in stroke and myocardial infarction events, in developing or Asian countries.

We also updated the data on annual incidence of hypertension in the model, extracting it from a previous study in a Vietnamese population [16]. The annual incidence was converted into an annual transition probability [17]. Although we expect that the probability is age and sex dependent, due to limited data, we had to assume that the probability remains constant with age and sex.

Chap

ter 6

131


Fig 1B: Different health states in the Markov model

Fig 1: Decision tree and Markov model for estimating cost-effectiveness of screening for hypertension

Notes: HBP: high blood pressure; CVD: cardiovascular disease. Patients start in the initial hypertension state. Patients can remain in this state or move to either acute CVD or CVD/non-CVD death. From the acute CVD state, patients can move to stable CVD or CVD/non-CVD death state, or may experience recurrent CVD events. From the stable CVD state, patients may stay in the same health state, or they may have a recurrence of CVD or can move to CVD/non-CVD death.

Screening and hypertension management

We previously conducted community screening for hypertension among currently untreated, undiagnosed adults aged 35-64 years, measuring BP during two visits [14]. Individual cases with BP in the hypertensive range were confirmed by a doctor at the CHS. The details of this field work were described previously [14]. For the non-screening scenario/current practice, patients are assumed to visit the CHS for diagnosis and management when they have symptoms of hypertension.

Diagnosed patients in both the screening and non-screening scenarios were assumed to be receiving treatment for hypertension at the CHS.

Acute CVD

Stable CVDHBP

CVD and non-CVD

Chapter 6

Antihypertensive drugs would be prescribed according to the Ministry of Health guidelines [15]. We assumed that patients with CVD were referred to a hospital for treatment.

Prevalence and incidence of hypertension

In the screening scenario, total prevalence, stratified according to different age groups and sex, was quantified by prevalence of hypertension detected by screening plus prevalence diagnosed by symptoms. Prevalence of current treatment, after diagnosis through symptoms, was recalculated from a national survey on a sex-specific basis [6]. Among hypertensive patients, we estimated the prevalence of patients going or not going into treatment in accordance with a previous study [6]. The number of patients in the group going into treatment was aggregated with those currently under treatment, and the number of hypertensive patients detected by screening multiplied by the percentage on treatment among diagnosed patients. Coverage of treatment reflects the percentage of diagnosed patients who received treatment, additionally assuming that patients classified as on treatment strictly comply with therapy. Although in reality not all patients may be strictly adherent, we did not have reliable data to use as inputs in our analysis. Also, we could not find appropriate evidence on, for example, the association of adherence level and cardiovascular events, particularly in stroke and myocardial infarction events, in developing or Asian countries.

We also updated the data on annual incidence of hypertension in the model, extracting it from a previous study in a Vietnamese population [16]. The annual incidence was converted into an annual transition probability [17]. Although we expect that the probability is age and sex dependent, due to limited data, we had to assume that the probability remains constant with age and sex.

132


Transition probabilities and relative risks

Probabilities for transitions between health states were extracted from previous studies. We started with a known population, for which data on age, sex, BP, cholesterol level and smoking status were available. The Asia Cardiovascular Risk Prediction model was applied to estimate the cumulative eight-year CVD risk (fatal CVD, MI, cerebrovascular diseases) for hypertensive patients [14, 18],subsequently converted into annual probabilities [17]. Notably, as mentioned above, age, sex, BP, smoking, and cholesterol level were used to predict CVD risk [18]. These data were extracted from our fieldwork because we could not access data from the national survey [14]. Probabilities were classified into two categories: fatal CVD and non-fatal CVD. The weighting scale for this measurement was re-calculated from individual studies of a meta-analysis in an Asian population for non-treatment groups (reference group) [3]. In the end, only age and sex were applied in the transition probabilities. To get equivalence of the coefficients for age and sex as in a model including cholesterol, BP level and smoking status, we fitted the transition probability CVD model with the grand-mean-centred predictors for BP, cholesterol level and smoking status [19, 20]. In addition, when there was no evidence for history of CVD among hypertensive patients, we assumed that they had no history of CVD at the time of data entry. If they did have CVD history, they would still have the same probability of transition from hypertension to acute CVD or CVD death as patients with no history of CVD [21].

The transition probability of acute to fatal CVD was calculated from individual studies of a meta-analysis among an Asian population for the non-treatment group [3]. We assumed this transition probability was stable for different ages and sexes.

Chapter 6

The probability of recurrent CVD events of patients at the end of the acute or stable state may be expected to be higher than for subjects without a history of CVD events. However, there was no appropriate equation nor data to estimate the probability for those patients, so we assumed that the annual transition probability of patients with a history of CVD was the same as for those without CVD, when they were of the same age and sex.

Data from the Vietnam Life Table 2013 was used to quantify the transition probability from any health state to crude death (death due to all causes) [22] and to separate transition probability of non-(composite) CVD from crude death. The weighting scale for this estimate was re-calculated from a study on CVD mortality in Vietnam, assuming that the rate of death in CVD and non-CVD was stable by age [23] (possibly overestimated because CVD death may be from other heart disease).

Relative risks (RRs) of acute CVD and CVD death during treatment were also estimated based on individual studies of a meta-analysis among an Asian population [3, 24]. In the base case, we did not consider age-and sex-dependence for these RRs. Furthermore, results of the meta-analysis revealed no significantly different reductions in acute MI and stroke events, comparing different classes of antihypertensive drugs under similar BP control.

Costs

Direct costs were quantified from the health service perspective, integrating four components: screening for hypertension, annual hypertension treatment, acute CVD treatment, and stable CVD treatment. All costs were calculated in international dollars (Int $) for the year 2013. To convert VND to Int $, we divided the amount in VND by the Purchasing Power Parity (PPP) exchange rate (in this

Chap

ter 6

133


Transition probabilities and relative risks

Probabilities for transitions between health states were extracted from previous studies. We started with a known population, for which data on age, sex, BP, cholesterol level and smoking status were available. The Asia Cardiovascular Risk Prediction model was applied to estimate the cumulative eight-year CVD risk (fatal CVD, MI, cerebrovascular diseases) for hypertensive patients [14, 18],subsequently converted into annual probabilities [17]. Notably, as mentioned above, age, sex, BP, smoking, and cholesterol level were used to predict CVD risk [18]. These data were extracted from our fieldwork because we could not access data from the national survey [14]. Probabilities were classified into two categories: fatal CVD and non-fatal CVD. The weighting scale for this measurement was re-calculated from individual studies of a meta-analysis in an Asian population for non-treatment groups (reference group) [3]. In the end, only age and sex were applied in the transition probabilities. To get equivalence of the coefficients for age and sex as in a model including cholesterol, BP level and smoking status, we fitted the transition probability CVD model with the grand-mean-centred predictors for BP, cholesterol level and smoking status [19, 20]. In addition, when there was no evidence for history of CVD among hypertensive patients, we assumed that they had no history of CVD at the time of data entry. If they did have CVD history, they would still have the same probability of transition from hypertension to acute CVD or CVD death as patients with no history of CVD [21].

The transition probability of acute to fatal CVD was calculated from individual studies of a meta-analysis among an Asian population for the non-treatment group [3]. We assumed this transition probability was stable for different ages and sexes.

Chapter 6

The probability of recurrent CVD events of patients at the end of the acute or stable state may be expected to be higher than for subjects without a history of CVD events. However, there was no appropriate equation nor data to estimate the probability for those patients, so we assumed that the annual transition probability of patients with a history of CVD was the same as for those without CVD, when they were of the same age and sex.

Data from the Vietnam Life Table 2013 was used to quantify the transition probability from any health state to crude death (death due to all causes) [22] and to separate transition probability of non-(composite) CVD from crude death. The weighting scale for this estimate was re-calculated from a study on CVD mortality in Vietnam, assuming that the rate of death in CVD and non-CVD was stable by age [23] (possibly overestimated because CVD death may be from other heart disease).

Relative risks (RRs) of acute CVD and CVD death during treatment were also estimated based on individual studies of a meta-analysis among an Asian population [3, 24]. In the base case, we did not consider age-and sex-dependence for these RRs. Furthermore, results of the meta-analysis revealed no significantly different reductions in acute MI and stroke events, comparing different classes of antihypertensive drugs under similar BP control.

Costs

Direct costs were quantified from the health service perspective, integrating four components: screening for hypertension, annual hypertension treatment, acute CVD treatment, and stable CVD treatment. All costs were calculated in international dollars (Int $) for the year 2013. To convert VND to Int $, we divided the amount in VND by the Purchasing Power Parity (PPP) exchange rate (in this

134


case: 7,546.6) [25]. The discount rate for costs was 3% in the base case.

Screening costs were estimated from various sources. During fieldwork, village health workers and medical students measured people’s BP in communities, making it difficult to quantify the costs of these measurements in real-life practice. We made an assumption that each patient visited the CHS three times to be screened for hypertension, at 10 minutes per visit. The total cost of screening per person equalled the number of visits multiplied by the cost per primary care visit, taken from a previous study [26, 27].

Total cost of annual hypertension treatment per person was the sum of drug costs and CHS visits per month multiplied by twelve. Drug cost per month was obtained by multiplying number of pills on prescriptions for 338 patients treated at CHS, by prices extracted from an international drug price indicator guide 2013 [28], adding 30% for cost of transportation and distribution [26, 29]. Monthly visits to CHS were allocated 20 minutes on average; cost for visits were taken from a citation in a previous study [26, 27].

Costs of acute and stable CVD treatment were extracted from the database of Thai Nguyen Hospital, as described previously [30]. All patients with the relevant ICD codes, either code I21 or codes from I60 to I66, were included in the study. Treatment cost of the first and following years after the first acute MI event was calculated based on expert opinion. We applied a proportion of 68.5% patients with a first acute MI who had percutaneous coronary intervention or coronary artery bypass surgery [31] in the first-year treatment. We also assumed that there was no specific treatment or rehabilitation for stroke patients after acute events, which is common in Vietnam, where rehabilitation and long-term care is done by family members.

Chapter 6

Health utilities

Quality of life weights (utilities) for healthy, hypertensive, acute CVD and stable CVD cases were applied in the model. From published studies, we weighted utilities of all health states to the same scale of SF-6D. For example, utility of the general population in Vietnam was 0.88 by EQ-5D measurement [32]; subsequently applying the weighting scale between EQ-5D and SF-6D [33] identified utilities in SF-6D at 0.93. Health utility of hypertension was measured by SF-6D in Vietnam, assuming that utility is equal between treated and non-treated patients [34]. Heath utility of hypertension was also used to weight for MI which was not yet available in Vietnam. The health utility of stable MI was extracted from a Korean study and weighted for Vietnam, using a ratio of utility for hypertension patients between Korean and Vietnamese [35]. Health utility of stable stroke was extracted from a previous study in Vietnam [34]. Utility for the acute state was weighted from the stable state according to the ratio between acute and stable MI and stroke cited previously [34, 36, 37]. We did not discount utility in the base case.

Base case and Sensitivity analysis

Cost-effectiveness was calculated in the base case and sensitivity analysis. Screening scenarios were combined with assumed increased levels of treatment among diagnosed hypertension (at 20% increase compared to baseline). Probabilistic sensitivity analysis in 5,000 repetitions of a Monte Carlo simulation examined the uncertainty of input parameters; Table 2 shows input values and distributions. Gamma distributions were applied for costs, beta distributions for health utilities, lognormal distributions for RRs and beta distributions for transition probabilities. We employed Cholesky decomposition to provide correlated draws, generated from a parameter’s multivariate normal distribution for transition probability of hypertension to death and acute CVD [17].

Chap

ter 6

135


case: 7,546.6) [25]. The discount rate for costs was 3% in the base case.

Screening costs were estimated from various sources. During fieldwork, village health workers and medical students measured people’s BP in communities, making it difficult to quantify the costs of these measurements in real-life practice. We made an assumption that each patient visited the CHS three times to be screened for hypertension, at 10 minutes per visit. The total cost of screening per person equalled the number of visits multiplied by the cost per primary care visit, taken from a previous study [26, 27].

Total cost of annual hypertension treatment per person was the sum of drug costs and CHS visits per month multiplied by twelve. Drug cost per month was obtained by multiplying number of pills on prescriptions for 338 patients treated at CHS, by prices extracted from an international drug price indicator guide 2013 [28], adding 30% for cost of transportation and distribution [26, 29]. Monthly visits to CHS were allocated 20 minutes on average; cost for visits were taken from a citation in a previous study [26, 27].

Costs of acute and stable CVD treatment were extracted from the database of Thai Nguyen Hospital, as described previously [30]. All patients with the relevant ICD codes, either code I21 or codes from I60 to I66, were included in the study. Treatment cost of the first and following years after the first acute MI event was calculated based on expert opinion. We applied a proportion of 68.5% patients with a first acute MI who had percutaneous coronary intervention or coronary artery bypass surgery [31] in the first-year treatment. We also assumed that there was no specific treatment or rehabilitation for stroke patients after acute events, which is common in Vietnam, where rehabilitation and long-term care is done by family members.

Chapter 6

Health utilities




136

Tab

le 2

: Bas

e-ca

se m

odel

inpu

ts a

nd d

istri

butio

n

Var

iabl

esD

ata

Dis

trib

utio

nSo

urce

sPr

eval

ence

of H

BP

5% to

41%

(age

and

sex

depe

nden

t)Fi

xed

Re-

calc

ulat

ion

[6, 1

4]Pr

eval

ence

of H

BP

dete

cted

by

scre

enin

g 2.

8% to

29.

7% (a

ge a

nd se

x de

pend

ent)

Fixe

dR

e-ca

lcul

atio

n [6

, 14]

Prev

alen

ce o

f HB

P no

t det

ecte

d by

scre

enin

g 2.

2 %

to 1

5.3

% (a

ge a

nd se

x de

pend

ent)

Fixe

dR

e-ca

lcul

atio

n [6

]R

ate

of g

oing

to tr

eat a

mon

g aw

are

hype

rtens

ives

62%

, 48%

in

fem

ale

and

mal

e, re

spec

tivel

yFi

xed

Re-

calc

ulat

ion

[6]

One

-yea

r tra

nsiti

on p

roba

bilit

y fro

m h

ealth

y to

hyp

erte

nsio

n

0.00

65 o

r 0.0

164

in fe

mal

e an

d m

ale,

resp

ectiv

ely

Bet

aR

e-ca

lcul

atio

n [1

6, 1

7]

One

-yea

r tra

nsiti

on p

roba

bilit

y fro

m H

BP

to n

on-fa

tal C

VD

Age

and

sex

depe

nden

t

Cho

lesk

yR

e-ca

lcul

atio

n fro

m A

SIA

CV

D

pred

ictio

n m

odel

[3, 1

4, 1

7]C

onsta

nt: (

-9.5

4)C

oeffi

cien

ts fo

r age

: 0.0

7C

oeffi

cien

ts fo

r sex

: 0.5

5

One

-yea

r tra

nsiti

on p

roba

bilit

y fro

m H

BP

to fa

tal C

VD

Age

and

sex

depe

nden

t

Cho

lesk

yR

e-ca

lcul

atio

n [3

, 14,

17]

Con

stant

: (-1

0.67

)C

oeffi

cien

ts fo

r age

: 0.0

7C

oeffi

cien

ts fo

r sex

: 0.5

5O

ne-y

ear t

rans

ition

pro

babi

lity

from

acu

te C

VD

to d

eath

in

the

non-

treat

men

t gro

up

0.00

8B

eta

[3, 1

7]

Cos

t of s

cree

ning

for h

yper

tens

ion

6.

05G

amm

aC

alcu

latio

n ba

sed

on c

ost o

f pr

imar

y ca

re re

porte

d in

pre

viou

s st

udy

[26]

Cos

t of H

BP

treat

men

t in

the

com

mun

ity

70.8

2G

amm

aC

alcu

latio

n ba

sed

on

pres

crip

tions

at C

HSs

and

cost

per o

utpa

tient

vis

its [

27]

Cos

t of a

cute

CV

D a

nd tr

eatm

ent f

irst y

ear

3,72

3.24

Gam

ma

Cal

cula

tion

base

d on

dat

abas

e of

Th

ai N

guye

n ge

nera

l hos

pita

l, V

ietn

am a

nd e

xper

t’s o

pini

on

Cos

t of s

tabl

e C

VD

trea

tmen

t in

follo

wed

yea

r79

.39

Gam

ma

Cal

cula

tion

base

d in

an

expe

rt’s

opin

ions

Util

ity in

hea

lthy

stat

e0.

93B

eta

Re-

calc

ulat

ion

[32,

33]

Util

ity in

HB

P-st

ate

0.74

or 0

.71

in m

ale

and

fem

ale,

resp

ectiv

ely

Bet

a[3

4]U

tility

in a

cute

CV

D-s

tate

0.67

Bet

aR

e-ca

lcul

atio

n [3

, 34-

37]

Util

ity in

stab

le C

VD

-sta

te0.

72 o

r 0.7

1 in

mal

e an

d fe

mal

e, re

spec

tivel

yB

eta

Re-

calc

ulat

ion

[3, 3

4, 3

5]R

elat

ive

risk

HB

P to

acu

te C

VD

0.

72Lo

gnor

mal

Re-

calc

ulat

ion

[3, 2

4]R

elat

ive

risk

CV

D-d

eath

0.

82Lo

gnor

mal

Re-

calc

ulat

ion

[3, 2

4]

Chapter 6

Univariate sensitivity analysis was undertaken to examine the robustness of our model assumptions and data sources. This analysis included ± 25% of the transition probability from BP to fatal and non-fatal CVD and costs of screening, BP treatment and CVD treatment. Also, reducing health utilities of CVD by 10% and 20% and 1% or 3% discounting of QALYs was investigated.

For the purpose of international comparison, health utilities of MI and stroke were also based on disabilities extracted from the GBD 2010 study [38] and then weighted for health utilities of acute and stable CVD to quantify the effectiveness of the intervention. As disability weights of hypertension and healthy were not available at international level, they are weighted based on the values of the base case and these respective values from the GBD study. Notably, utility in acute CVD-state, stable CVD-state, healthy state and HBP-state were 0.685, 0.717, 0.957 and 0.835, respectively [38, 39].

The HBP prevalence of the national survey was higher than in our survey, therefore HBP prevalence according to the national survey was analyzed in a separate scenario. A key point for applying HBP prevalence from the national survey in a scenario analysis and not in the base case was that BP in the national survey was measured during one visit while in our study it was done on two different occasions, with a decrease in BP prevalence from 20.5% at the first visit to 12.3% in the second visit. We feel our survey is more likely to be accurate than the national survey. Furthermore, we suppose that the prevalence of hypertension in urban settings may be higher than in rural areas, as found in the national survey conducted in both rural and urban areas.

A meta-analysis report showed that the RR reduction of CVD is associated with age [4]. Lacking data in our setting, we conducted a scenario analysis which took the RR reduction of CVD age-

Chapter 6

Health utilities




Chap

ter 6

137

Tab

le 2

: Bas

e-ca

se m

odel

inpu

ts a

nd d

istri

butio

n

Var

iabl

esD

ata

Dis

trib

utio

nSo

urce

sPr

eval

ence

of H

BP

5% to

41%

(age

and

sex

depe

nden

t)Fi

xed

Re-

calc

ulat

ion

[6, 1

4]Pr

eval

ence

of H

BP

dete

cted

by

scre

enin

g 2.

8% to

29.

7% (a

ge a

nd se

x de

pend

ent)

Fixe

dR

e-ca

lcul

atio

n [6

, 14]

Prev

alen

ce o

f HB

P no

t det

ecte

d by

scre

enin

g 2.

2 %

to 1

5.3

% (a

ge a

nd se

x de

pend

ent)

Fixe

dR

e-ca

lcul

atio

n [6

]R

ate

of g

oing

to tr

eat a

mon

g aw

are

hype

rtens

ives

62%

, 48%

in

fem

ale

and

mal

e, re

spec

tivel

yFi

xed

Re-

calc

ulat

ion

[6]

One

-yea

r tra

nsiti

on p

roba

bilit

y fro

m h

ealth

y to

hyp

erte

nsio

n

0.00

65 o

r 0.0

164

in fe

mal

e an

d m

ale,

resp

ectiv

ely

Bet

aR

e-ca

lcul

atio

n [1

6, 1

7]

One

-yea

r tra

nsiti

on p

roba

bilit

y fro

m H

BP

to n

on-fa

tal C

VD

Age

and

sex

depe

nden

t

Cho

lesk

yR

e-ca

lcul

atio

n fro

m A

SIA

CV

D

pred

ictio

n m

odel

[3, 1

4, 1

7]C

onsta

nt: (

-9.5

4)C

oeffi

cien

ts fo

r age

: 0.0

7C

oeffi

cien

ts fo

r sex

: 0.5

5

One

-yea

r tra

nsiti

on p

roba

bilit

y fro

m H

BP

to fa

tal C

VD

Age

and

sex

depe

nden

t

Cho

lesk

yR

e-ca

lcul

atio

n [3

, 14,

17]

Con

stant

: (-1

0.67

)C

oeffi

cien

ts fo

r age

: 0.0

7C

oeffi

cien

ts fo

r sex

: 0.5

5O

ne-y

ear t

rans

ition

pro

babi

lity

from

acu

te C

VD

to d

eath

in

the

non-

treat

men

t gro

up

0.00

8B

eta

[3, 1

7]

Cos

t of s

cree

ning

for h

yper

tens

ion

6.

05G

amm

aC

alcu

latio

n ba

sed

on c

ost o

f pr

imar

y ca

re re

porte

d in

pre

viou

s st

udy

[26]

Cos

t of H

BP

treat

men

t in

the

com

mun

ity

70.8

2G

amm

aC

alcu

latio

n ba

sed

on

pres

crip

tions

at C

HSs

and

cost

per o

utpa

tient

vis

its [

27]

Cos

t of a

cute

CV

D a

nd tr

eatm

ent f

irst y

ear

3,72

3.24

Gam

ma

Cal

cula

tion

base

d on

dat

abas

e of

Th

ai N

guye

n ge

nera

l hos

pita

l, V

ietn

am a

nd e

xper

t’s o

pini

on

Cos

t of s

tabl

e C

VD

trea

tmen

t in

follo

wed

yea

r79

.39

Gam

ma

Cal

cula

tion

base

d in

an

expe

rt’s

opin

ions

Util

ity in

hea

lthy

stat

e0.

93B

eta

Re-

calc

ulat

ion

[32,

33]

Util

ity in

HB

P-st

ate

0.74

or 0

.71

in m

ale

and

fem

ale,

resp

ectiv

ely

Bet

a[3

4]U

tility

in a

cute

CV

D-s

tate

0.67

Bet

aR

e-ca

lcul

atio

n [3

, 34-

37]

Util

ity in

stab

le C

VD

-sta

te0.

72 o

r 0.7

1 in

mal

e an

d fe

mal

e, re

spec

tivel

yB

eta

Re-

calc

ulat

ion

[3, 3

4, 3

5]R

elat

ive

risk

HB

P to

acu

te C

VD

0.

72Lo

gnor

mal

Re-

calc

ulat

ion

[3, 2

4]R

elat

ive

risk

CV

D-d

eath

0.

82Lo

gnor

mal

Re-

calc

ulat

ion

[3, 2

4]

Chapter 6

Univariate sensitivity analysis was undertaken to examine the robustness of our model assumptions and data sources. This analysis included ± 25% of the transition probability from BP to fatal and non-fatal CVD and costs of screening, BP treatment and CVD treatment. Also, reducing health utilities of CVD by 10% and 20% and 1% or 3% discounting of QALYs was investigated.

For the purpose of international comparison, health utilities of MI and stroke were also based on disabilities extracted from the GBD 2010 study [38] and then weighted for health utilities of acute and stable CVD to quantify the effectiveness of the intervention. As disability weights of hypertension and healthy were not available at international level, they are weighted based on the values of the base case and these respective values from the GBD study. Notably, utility in acute CVD-state, stable CVD-state, healthy state and HBP-state were 0.685, 0.717, 0.957 and 0.835, respectively [38, 39].

The HBP prevalence of the national survey was higher than in our survey, therefore HBP prevalence according to the national survey was analyzed in a separate scenario. A key point for applying HBP prevalence from the national survey in a scenario analysis and not in the base case was that BP in the national survey was measured during one visit while in our study it was done on two different occasions, with a decrease in BP prevalence from 20.5% at the first visit to 12.3% in the second visit. We feel our survey is more likely to be accurate than the national survey. Furthermore, we suppose that the prevalence of hypertension in urban settings may be higher than in rural areas, as found in the national survey conducted in both rural and urban areas.

A meta-analysis report showed that the RR reduction of CVD is associated with age [4]. Lacking data in our setting, we conducted a scenario analysis which took the RR reduction of CVD age-

Chapter 6

Health utilities




138


dependently into account. We estimated the RR based on values in the base case and the results of a meta-analysis among 147 randomized controlled trials [4].

We applied a CVD prediction model that estimated CVD risk in eight cumulative years. Therefore, a conservative ten-year time horizon was applied to examine the cost-effectiveness of the intervention in the base case. In addition, we ran a lifetime-horizon model to capture long-term effects of the screening and treatment.

Incremental cost-effectiveness ratios for each scenario were investigated from the health care provider perspective. The threshold for willingness to pay was three times the gross domestic product (GDP) [40], taking the PPP exchange rate into account [41]. One GDP per capita which is based on PPP in Vietnam in 2013 was 5,294current Int$ [42].

Ethical statement

The study on costs of stroke and MI was approved by the Ethical Committee of the Thai Nguyen General Hospital; the extraction of cost data was done with permission of the Planning Department of this hospital. The research proposal and consent procedure of the survey involving human participants was approved by the Institutional Review Board in Biomedical Research of the Institute of Social and Medical Studies in Vietnam. Participants were informed about the objectives and methods of the study and signed a consent form when they agreed to participate.

Chapter 6

Results

Deterministic analysis

Results on the numbers of QALYs gained, incremental costs and cost-effectiveness ratios of the ten-year time horizon model are presented in Table 3. In comparison with no screening, all screening strategies resulted in QALYs gained. The number of QALYs gained from screening every two years was very similar to the number gained by an annual screening, in all corresponding age and sex groups. For example, QALYs gained in screening males from age 35 onwards was 0.25 and 0.26 in biannual and annual screening, respectively. Cost per QALYs gained varied by groups and screening strategies, from cost saving to Int$ 758,695 per QALY gained. Screening was not cost-effective for screening strategies starting at 35 years for both sexes in all strategies. Among females, one-off screening was cost-effective starting at 45 years at Int$ 12,070 per QALY gained. For screening starting at 55 years, one-off or biannual screening cost per QALY gained amounted to Int$ 871 and Int$ 11,189, respectively. Screening males from 55 years onwards, was cost saving, Int$ 2,076 and Int$ 7,638 for one-off, biannual and annual screening, respectively.

Chapter 6

Health utilities




Chap

ter 6

139


dependently into account. We estimated the RR based on values in the base case and the results of a meta-analysis among 147 randomized controlled trials [4].

We applied a CVD prediction model that estimated CVD risk in eight cumulative years. Therefore, a conservative ten-year time horizon was applied to examine the cost-effectiveness of the intervention in the base case. In addition, we ran a lifetime-horizon model to capture long-term effects of the screening and treatment.

Incremental cost-effectiveness ratios for each scenario were investigated from the health care provider perspective. The threshold for willingness to pay was three times the gross domestic product (GDP) [40], taking the PPP exchange rate into account [41]. One GDP per capita which is based on PPP in Vietnam in 2013 was 5,294current Int$ [42].

Ethical statement

The study on costs of stroke and MI was approved by the Ethical Committee of the Thai Nguyen General Hospital; the extraction of cost data was done with permission of the Planning Department of this hospital. The research proposal and consent procedure of the survey involving human participants was approved by the Institutional Review Board in Biomedical Research of the Institute of Social and Medical Studies in Vietnam. Participants were informed about the objectives and methods of the study and signed a consent form when they agreed to participate.

Chapter 6

Results

Deterministic analysis

Results on the numbers of QALYs gained, incremental costs and cost-effectiveness ratios of the ten-year time horizon model are presented in Table 3. In comparison with no screening, all screening strategies resulted in QALYs gained. The number of QALYs gained from screening every two years was very similar to the number gained by an annual screening, in all corresponding age and sex groups. For example, QALYs gained in screening males from age 35 onwards was 0.25 and 0.26 in biannual and annual screening, respectively. Cost per QALYs gained varied by groups and screening strategies, from cost saving to Int$ 758,695 per QALY gained. Screening was not cost-effective for screening strategies starting at 35 years for both sexes in all strategies. Among females, one-off screening was cost-effective starting at 45 years at Int$ 12,070 per QALY gained. For screening starting at 55 years, one-off or biannual screening cost per QALY gained amounted to Int$ 871 and Int$ 11,189, respectively. Screening males from 55 years onwards, was cost saving, Int$ 2,076 and Int$ 7,638 for one-off, biannual and annual screening, respectively.

Chapter 6

Health utilities




140


Table 3: Cost – effectiveness of screening for hypertension in alternative screening strategies for hypertension by age and sex in the 10 years model (per 1,000 people).

Cost (Int$) QALY Incremental

costIncremental

QALY

ICER (cost (Int$) per

QALY gained)

Start screening at age of 35 years, female

No 47,676 8,942

One-off 53,249 8,942 5,573 0.044 127,715

E1 94,728 8,942 47,052 0.062 758,695

E2 70,713 8,942 23,037 0.060 386,851

E1&T.20% 94,640 8,942 46,964 0.082 572,679

E2&T.20% 70,629 8,942 22,952 0.079 291,476


No 143,627 8,421

One-off 147,633 8,421 4,006 0.332 12,070

E1 181,734 8,422 38,107 0.361 105,525

E2 161,962 8,422 18,336 0.357 51,335

E1&T.20% 181,249 8,422 37,622 0.478 78,786

E2&T.20% 161,483 8,422 17,856 0.472 37,806


No 231,456 7,278

One-off 232,311 7,279 855 0.982 871

E1 257,492 7,279 26,036 1.022 25,471

E2 242,831 7,279 11,375 1.017 11,189

E1&T.20% 256,092 7,279 24,635 1.352 18,226

E2&T.20% 241,439 7,279 9,982 1.344 7,425

Start screening at age of 35 years, male

No 99,013 8,436

One-off 104,150 8,436 5,138 0.175 29,433

E1 140,394 8,436 41,381 0.262 158,147

E2 117,314 8,436 18,301 0.250 73,227

E1&T.20% 139,978 8,437 40,965 0.370 110,602

E2&T.20% 116,916 8,437 17,903 0.354 50,607

Chapter 6


costIncremental

QALY


QALY gained)


No 174,638 7,681

One-off 177,658 7,682 3,020 0.722 4,183

E1 206,873 7,682 32,235 0.858 37,580

E2 188,094 7,682 13,456 0.839 16,035

E1&T.20% 205,542 7,682 30,904 1.214 25,453

E2&T.20% 186,793 7,682 12,155 1.188 10,233


No 274,570 5,954

One-off 272,510 5,956 Dominant 1.920 Dominant

E1 290,298 5,956 15,728 2.059 7,638

E2 278,804 5,956 4,234 2.039 2,076

E1&T.20% 286,994 5,957 12,424 2.915 4,262

E2&T.20% 275,533 5,957 963 2.887 334

Note: No: No screening, One-off: screening once at the first year. E1: Annual screening, E2: Biannual

screening, E1&T.20%: Annual screening combined with increasing coverage of treatment by 20%,

E2&T.20%: Biannual screening combined with increasing coverage of treatment by 20%.

When we ran models for different starting ages (35 years, 45 years and 55 years) and sexes in comparison with no screening for the lifetime horizon, more QALYs were gained in all strategies. Especially combining screening with increasing treatment by 20% produced relatively high gains in QALYs, compared to only screening, in all ages and sexes. For example, this combination resulted in 4.88 vs 3.69 QALYs gained in the strategies of screening females from 35 years onwards. For the same age and screening strategy, QALYs gained in males were always higher than in females. For example, we saw 5.12 QALYs added for females versus 6.27 for males with annual screening starting at 45 years. The exact numbers of QALYs gained for each strategy is presented in Fig 2.

Chapter 6

Health utilities




Chap

ter 6

141


Table 3: Cost – effectiveness of screening for hypertension in alternative screening strategies for hypertension by age and sex in the 10 years model (per 1,000 people).


costIncremental

QALY


QALY gained)


No 47,676 8,942

One-off 53,249 8,942 5,573 0.044 127,715

E1 94,728 8,942 47,052 0.062 758,695

E2 70,713 8,942 23,037 0.060 386,851

E1&T.20% 94,640 8,942 46,964 0.082 572,679

E2&T.20% 70,629 8,942 22,952 0.079 291,476


No 143,627 8,421

One-off 147,633 8,421 4,006 0.332 12,070

E1 181,734 8,422 38,107 0.361 105,525

E2 161,962 8,422 18,336 0.357 51,335

E1&T.20% 181,249 8,422 37,622 0.478 78,786

E2&T.20% 161,483 8,422 17,856 0.472 37,806


No 231,456 7,278

One-off 232,311 7,279 855 0.982 871

E1 257,492 7,279 26,036 1.022 25,471

E2 242,831 7,279 11,375 1.017 11,189

E1&T.20% 256,092 7,279 24,635 1.352 18,226

E2&T.20% 241,439 7,279 9,982 1.344 7,425


No 99,013 8,436

One-off 104,150 8,436 5,138 0.175 29,433

E1 140,394 8,436 41,381 0.262 158,147

E2 117,314 8,436 18,301 0.250 73,227

E1&T.20% 139,978 8,437 40,965 0.370 110,602

E2&T.20% 116,916 8,437 17,903 0.354 50,607

Chapter 6


costIncremental

QALY


QALY gained)


No 174,638 7,681

One-off 177,658 7,682 3,020 0.722 4,183

E1 206,873 7,682 32,235 0.858 37,580

E2 188,094 7,682 13,456 0.839 16,035

E1&T.20% 205,542 7,682 30,904 1.214 25,453

E2&T.20% 186,793 7,682 12,155 1.188 10,233


No 274,570 5,954

One-off 272,510 5,956 Dominant 1.920 Dominant

E1 290,298 5,956 15,728 2.059 7,638

E2 278,804 5,956 4,234 2.039 2,076

E1&T.20% 286,994 5,957 12,424 2.915 4,262

E2&T.20% 275,533 5,957 963 2.887 334

Note: No: No screening, One-off: screening once at the first year. E1: Annual screening, E2: Biannual

screening, E1&T.20%: Annual screening combined with increasing coverage of treatment by 20%,

E2&T.20%: Biannual screening combined with increasing coverage of treatment by 20%.

When we ran models for different starting ages (35 years, 45 years and 55 years) and sexes in comparison with no screening for the lifetime horizon, more QALYs were gained in all strategies. Especially combining screening with increasing treatment by 20% produced relatively high gains in QALYs, compared to only screening, in all ages and sexes. For example, this combination resulted in 4.88 vs 3.69 QALYs gained in the strategies of screening females from 35 years onwards. For the same age and screening strategy, QALYs gained in males were always higher than in females. For example, we saw 5.12 QALYs added for females versus 6.27 for males with annual screening starting at 45 years. The exact numbers of QALYs gained for each strategy is presented in Fig 2.

Chapter 6

Health utilities




142


Note: E1: Annual screening, E2: Biannual screening, E2 until 60+ E1: Biannual screening until 60 years

old then annual screening until died, E2 until 55+ E1: Biannual screening until 55 years old then annual

screening until died, E1&T.20%: Annual screening combined with increasing coverage of treatment by

20%, E2&T.20%: Biannual screening combined with increasing coverage of treatment by 20%.

Fig 2: QALYs gained by different screening strategies, lifetime model (per 1000 people)

Cost per QALY gained was less than the threshold of Int$ 15,883 per QALY in all scenarios, except for females from 35 years onwards. For example, cost per QALY in females from 35 years onwards were Int$ 27,944, Int$ 20,850, Int$ 17,280 and Int$ 16,115 in strategies of annual, annual combined with increased treatment by 20% and biannual up to 55 or 60 years and then annual screening, respectively. Cost per QALY gained in males was always lower than in females for the same age group and screening strategy. For example, in the annual strategy starting at 45 years, for males the cost was Int$ 6,834 per QALY gained, compared to Int$ 13,331 per QALY for females. Costs per QALY gained are shown in Fig 3.

Chapter 6

Note: E1: Annual screening, E2: Biannual screening, E2 until 60+ E1: Biannual screening until 60 years old then annual screening until died, E2 until 55+ E1: Biannual screening until 55 years old then annual screening until died, E1&T.20%: Annual screening combined with increasing coverage of treatment by 20%, E2&T.20%: Biannual screening combined with increasing coverage of treatment by 20%.

Fig 3: Cost per QALY by different strategies and age group, lifetime model

Probabilistic analysis

Full probabilistic sensitivity analysis of cost-effectiveness for a ten-year horizon is presented in Fig 4. In the scenario of screening starting at 45 years, we found a 65% probability of being cost-effective in one-off screening in females and 95% in males. For the screening strategy in females starting at 55 years, the probability of favourable cost-effectiveness was 90% with one-off screening. For screening males starting at 55 years, all strategies gave a high probability of being cost-effective.

Chapter 6

Health utilities




Chap

ter 6

143


Note: E1: Annual screening, E2: Biannual screening, E2 until 60+ E1: Biannual screening until 60 years

old then annual screening until died, E2 until 55+ E1: Biannual screening until 55 years old then annual

screening until died, E1&T.20%: Annual screening combined with increasing coverage of treatment by

20%, E2&T.20%: Biannual screening combined with increasing coverage of treatment by 20%.

Fig 2: QALYs gained by different screening strategies, lifetime model (per 1000 people)

Cost per QALY gained was less than the threshold of Int$ 15,883 per QALY in all scenarios, except for females from 35 years onwards. For example, cost per QALY in females from 35 years onwards were Int$ 27,944, Int$ 20,850, Int$ 17,280 and Int$ 16,115 in strategies of annual, annual combined with increased treatment by 20% and biannual up to 55 or 60 years and then annual screening, respectively. Cost per QALY gained in males was always lower than in females for the same age group and screening strategy. For example, in the annual strategy starting at 45 years, for males the cost was Int$ 6,834 per QALY gained, compared to Int$ 13,331 per QALY for females. Costs per QALY gained are shown in Fig 3.

Chapter 6

Note: E1: Annual screening, E2: Biannual screening, E2 until 60+ E1: Biannual screening until 60 years old then annual screening until died, E2 until 55+ E1: Biannual screening until 55 years old then annual screening until died, E1&T.20%: Annual screening combined with increasing coverage of treatment by 20%, E2&T.20%: Biannual screening combined with increasing coverage of treatment by 20%.

Fig 3: Cost per QALY by different strategies and age group, lifetime model

Probabilistic analysis

Full probabilistic sensitivity analysis of cost-effectiveness for a ten-year horizon is presented in Fig 4. In the scenario of screening starting at 45 years, we found a 65% probability of being cost-effective in one-off screening in females and 95% in males. For the screening strategy in females starting at 55 years, the probability of favourable cost-effectiveness was 90% with one-off screening. For screening males starting at 55 years, all strategies gave a high probability of being cost-effective.

Chapter 6

Health utilities




144


Note: one-off: screening once at the first year, E1: Annual screening, E2: Biannual screening, E1&20%: Annual screening combined with increasing coverage of treatment by 20%, E2&20%: Biannual screening combined with increasing coverage of treatment by 20%.

Fig 4: Cost-effectiveness acceptability curves of different screening strategies, 10 years horizon model.

The probability of favourable cost-effectiveness for each scenario using the lifetime horizon is shown in Fig 5. The strategy of screening repeated every two years combined with increasing treatment by 20% always gave the highest probability of being cost-effective, followed by biannual screening. Annual screening gave the lowest probability of being cost-effective in all scenarios.

Chapter 6

*Note: E1: Annual screening, E2: Biannual screening, E2 till 60+ E1: Biannual

screening until 60 years old then annual screening until died, E2 till 55+ E1:

Biannual screening until 55 years old then annual screening until died, E1&T.20%:

Annual screening combined with increasing coverage of treatment by 20%,

E2&T.20%: Biannual screening combined with increasing coverage of treatment by

20%.

Figure 5 Cost-effectiveness acceptability curves of different screening strategies, --

lifetime horizon model.

Chapter 6

Health utilities




Chap

ter 6

145


Note: one-off: screening once at the first year, E1: Annual screening, E2: Biannual screening, E1&20%: Annual screening combined with increasing coverage of treatment by 20%, E2&20%: Biannual screening combined with increasing coverage of treatment by 20%.

Fig 4: Cost-effectiveness acceptability curves of different screening strategies, 10 years horizon model.

The probability of favourable cost-effectiveness for each scenario using the lifetime horizon is shown in Fig 5. The strategy of screening repeated every two years combined with increasing treatment by 20% always gave the highest probability of being cost-effective, followed by biannual screening. Annual screening gave the lowest probability of being cost-effective in all scenarios.

Chapter 6

*Note: E1: Annual screening, E2: Biannual screening, E2 till 60+ E1: Biannual

screening until 60 years old then annual screening until died, E2 till 55+ E1:

Biannual screening until 55 years old then annual screening until died, E1&T.20%:

Annual screening combined with increasing coverage of treatment by 20%,

E2&T.20%: Biannual screening combined with increasing coverage of treatment by

20%.

Figure 5 Cost-effectiveness acceptability curves of different screening strategies, --

lifetime horizon model.

Chapter 6

Health utilities




146


Univariate sensitivity analysis

To evaluate how the cost per QALY gained is affected by the variation in input variables, we conducted univariate sensitivity analysis in both time-horizon models. Results of cost per QALY gained in sensitivity analysis are presented in Appendix 3 for the ten-year time horizon and in Appendix 4 for lifetime. In the ten-year model, the results seem relatively insensitive to changing values of input parameters, except for a few cases. One example for the latter is decreased utilities of CVD by 20% for annual screening in females 55 years, where the result was lower than the threshold value, while in the base case it was higher than the threshold in the scenarios of annual screening alone or combined with 20% increased treatment. Also, in one-off screening, applying HBP prevalence of the national survey in males from 35 years onwards, the result became lower than the threshold, while it was higher than the threshold in the base case. In the scenario of applying utilities based on GBD, cost per QALY was much lower than in the base case but had similar trends in comparison with the threshold.

In the lifetime model, the cost per QALY gained was quite insensitive to changing values of input parameters in all different screening options, age and sex groups, with the exception of applying a 3% discount rate on utility and applying HBP prevalence by the national survey. Cost per QALY was much lower after changing the HBP prevalence and higher in a 3% utility discount scenario in comparison with the base case. Taking the threshold into account, results contrasted with the base case for screening females starting at 35 years in a sensitivity analysis of 25% decreased screening costs, 10% or 20% decreased CVD utility, or applying HBP prevalence from the national survey. Notably, the cost per QALY gained in these analyses changed from over to below the threshold in strategies of every two

Chapter 6

years to 55 or 60 years and then annual screening, or annual screening combined with 20% increased treatment. Screening was no longer cost-effective for annual screening of females starting at 45 years if increasing screening costs by 25%, reducing transition probability from hypertension to CVD by 25%, or discounting utilities.

Discussion

This is one of the few studies on health economics analysis of screening for hypertension. It was conducted in the context of the current plans of the Ministry of Health in Vietnam to implement screening for hypertension as part of routine medical examinations to be covered by health insurance. However, in the context of limited resources, it may be useful to consider starting screening at a different age than that recommended by WHO. There is an urgent need for evidence to inform the new policies and planning, which should be based on local data to identify locally appropriate strategies, such as target age groups. Screening for hypertension starting at 35-, 45- or 55 years in males and at 55 years in females until death was cost-effective. The best screening strategies were biannual screening alone or combined with 20% increased treatment coverage. Also highly probable to be cost-effective was bi-annual screening of females starting at 45 years until death, either alone or combined with 20% increased treatment.

We found a negligible probability of favourable cost-effectiveness for screening females starting at 35 years using different strategies. These results may be explained by the lower HBP prevalence in this group compared to others. Overall, cost per QALY gained was lower than the threshold and showed a relatively high probability of being cost-effective in various screening strategies.

Chapter 6

Health utilities




Chap

ter 6

147


Univariate sensitivity analysis

To evaluate how the cost per QALY gained is affected by the variation in input variables, we conducted univariate sensitivity analysis in both time-horizon models. Results of cost per QALY gained in sensitivity analysis are presented in Appendix 3 for the ten-year time horizon and in Appendix 4 for lifetime. In the ten-year model, the results seem relatively insensitive to changing values of input parameters, except for a few cases. One example for the latter is decreased utilities of CVD by 20% for annual screening in females 55 years, where the result was lower than the threshold value, while in the base case it was higher than the threshold in the scenarios of annual screening alone or combined with 20% increased treatment. Also, in one-off screening, applying HBP prevalence of the national survey in males from 35 years onwards, the result became lower than the threshold, while it was higher than the threshold in the base case. In the scenario of applying utilities based on GBD, cost per QALY was much lower than in the base case but had similar trends in comparison with the threshold.

In the lifetime model, the cost per QALY gained was quite insensitive to changing values of input parameters in all different screening options, age and sex groups, with the exception of applying a 3% discount rate on utility and applying HBP prevalence by the national survey. Cost per QALY was much lower after changing the HBP prevalence and higher in a 3% utility discount scenario in comparison with the base case. Taking the threshold into account, results contrasted with the base case for screening females starting at 35 years in a sensitivity analysis of 25% decreased screening costs, 10% or 20% decreased CVD utility, or applying HBP prevalence from the national survey. Notably, the cost per QALY gained in these analyses changed from over to below the threshold in strategies of every two

Chapter 6

years to 55 or 60 years and then annual screening, or annual screening combined with 20% increased treatment. Screening was no longer cost-effective for annual screening of females starting at 45 years if increasing screening costs by 25%, reducing transition probability from hypertension to CVD by 25%, or discounting utilities.

Discussion

This is one of the few studies on health economics analysis of screening for hypertension. It was conducted in the context of the current plans of the Ministry of Health in Vietnam to implement screening for hypertension as part of routine medical examinations to be covered by health insurance. However, in the context of limited resources, it may be useful to consider starting screening at a different age than that recommended by WHO. There is an urgent need for evidence to inform the new policies and planning, which should be based on local data to identify locally appropriate strategies, such as target age groups. Screening for hypertension starting at 35-, 45- or 55 years in males and at 55 years in females until death was cost-effective. The best screening strategies were biannual screening alone or combined with 20% increased treatment coverage. Also highly probable to be cost-effective was bi-annual screening of females starting at 45 years until death, either alone or combined with 20% increased treatment.

We found a negligible probability of favourable cost-effectiveness for screening females starting at 35 years using different strategies. These results may be explained by the lower HBP prevalence in this group compared to others. Overall, cost per QALY gained was lower than the threshold and showed a relatively high probability of being cost-effective in various screening strategies.

Chapter 6

Health utilities




148


Regarding the initial age at screening, cost per QALY gained was lower in the older group in all strategies. This finding is consistent with a study in the Netherlands [9]. Results on cost-effectiveness of screening for hypertension and anti-hypertensive treatment in preventing CVD and kidney disease were also in line with our results [8].

Even though we found that various screening strategies could be cost-effective, biannual screening combined with increased treatment by 20% always had the highest probability of being cost-effective, among different strategies for the same age and sex. This finding agrees with previous studies demonstrating higher QALY gains with treatment, compared to no treatment [43]. Ergo, combining screening with increased treatment among those diagnosed with hypertension arises as a strong recommendation from our analysis.

In the 10 years model, cost per QALY gained was much higher than in the lifetime model, comparing the same age, sex, and screening strategy. This result was similar to that from a previous study comparing one-year, three-year or five-year horizons on the costs and effects of consistently similar hypertension treatment: as the time horizon increases, cost per QALY decreases [44]. We found a high probability of being cost-effective only in the scenario of screening females starting at 55 years and males at 45 years, when screening is done one-off or biannually combined with 20% increased treatment. For the screening strategy in males starting at 55 years, all strategies were cost-effective. This finding contributes sufficient evidence to recommend continuous screening and long-term treatment for HBP in these specific groups.

We applied coverage of treatment among patients who are aware of their diagnosis, using current treatment levels extracted from the national survey. This prevalence was estimated at a cut-off point

Chapter 6

taking adherence into account, at the time of the survey. We also ran the models with the assumption that all treated patients strictly comply with therapy. Previous studies in Vietnam found that among patients managed at CHS, patients’ adherence with anti-hypertension medicine (at cut-off point 80%) was 57% and 46% among females and males, respectively [45]. A similar study in China showed that the incremental cost-effectiveness ratio was sensitive to adherence levels. However, it was still lower than the threshold of three times the GDP, applying adherence levels of 75% or even down to 40% [46].

We built the models to estimate cost-effectiveness of screening and treatment for hypertension in a context of limited data. We were conservative in selecting values for input parameters and model to predict CVD of hypertensive patients. Yet, the first limitation to be noted does concern the input parameters which may not always exactly represent the Vietnamese population. The limited availability of data on health utilities should also be noted. Health utilities of acute strokes, acute and stable MI were extracted from data available from other countries and these were used for the Vietnamese context.

Several types of data/sources of information were not available in Vietnam. For example, there was no model to predict CVD risk or relative risk of treatment versus no treatment for hypertension or CVD. However, we had previously shown that an existing Asian model may be appropriate to predict CVD in Vietnam, as used in this study [14]. We also applied the results from a meta-analysis among Asian studies, which seems reasonable to weight the different CVD components and estimate the RR of treatment for HBP [3]. It should be pointed out that the RR reduction of CVD reported in the base case is not based on age-specific BP at pre-treatment. To examine uncertainty introduced here, one specific scenario was analyzed. The result of the scenario where the RR reduction of CVD depended on

Chapter 6

Health utilities




Chap

ter 6

149


Regarding the initial age at screening, cost per QALY gained was lower in the older group in all strategies. This finding is consistent with a study in the Netherlands [9]. Results on cost-effectiveness of screening for hypertension and anti-hypertensive treatment in preventing CVD and kidney disease were also in line with our results [8].

Even though we found that various screening strategies could be cost-effective, biannual screening combined with increased treatment by 20% always had the highest probability of being cost-effective, among different strategies for the same age and sex. This finding agrees with previous studies demonstrating higher QALY gains with treatment, compared to no treatment [43]. Ergo, combining screening with increased treatment among those diagnosed with hypertension arises as a strong recommendation from our analysis.

In the 10 years model, cost per QALY gained was much higher than in the lifetime model, comparing the same age, sex, and screening strategy. This result was similar to that from a previous study comparing one-year, three-year or five-year horizons on the costs and effects of consistently similar hypertension treatment: as the time horizon increases, cost per QALY decreases [44]. We found a high probability of being cost-effective only in the scenario of screening females starting at 55 years and males at 45 years, when screening is done one-off or biannually combined with 20% increased treatment. For the screening strategy in males starting at 55 years, all strategies were cost-effective. This finding contributes sufficient evidence to recommend continuous screening and long-term treatment for HBP in these specific groups.

We applied coverage of treatment among patients who are aware of their diagnosis, using current treatment levels extracted from the national survey. This prevalence was estimated at a cut-off point

Chapter 6

taking adherence into account, at the time of the survey. We also ran the models with the assumption that all treated patients strictly comply with therapy. Previous studies in Vietnam found that among patients managed at CHS, patients’ adherence with anti-hypertension medicine (at cut-off point 80%) was 57% and 46% among females and males, respectively [45]. A similar study in China showed that the incremental cost-effectiveness ratio was sensitive to adherence levels. However, it was still lower than the threshold of three times the GDP, applying adherence levels of 75% or even down to 40% [46].

We built the models to estimate cost-effectiveness of screening and treatment for hypertension in a context of limited data. We were conservative in selecting values for input parameters and model to predict CVD of hypertensive patients. Yet, the first limitation to be noted does concern the input parameters which may not always exactly represent the Vietnamese population. The limited availability of data on health utilities should also be noted. Health utilities of acute strokes, acute and stable MI were extracted from data available from other countries and these were used for the Vietnamese context.

Several types of data/sources of information were not available in Vietnam. For example, there was no model to predict CVD risk or relative risk of treatment versus no treatment for hypertension or CVD. However, we had previously shown that an existing Asian model may be appropriate to predict CVD in Vietnam, as used in this study [14]. We also applied the results from a meta-analysis among Asian studies, which seems reasonable to weight the different CVD components and estimate the RR of treatment for HBP [3]. It should be pointed out that the RR reduction of CVD reported in the base case is not based on age-specific BP at pre-treatment. To examine uncertainty introduced here, one specific scenario was analyzed. The result of the scenario where the RR reduction of CVD depended on

Chapter 6

Health utilities




150


age showed that there were only small changes in the cost per QALY compared with the base case and the same trend remained when compared with the threshold.

Data on individual profiles (age, sex, BP, cholesterol level, smoking status) for hypertensive patients were not available. To overcome this limitation, we conducted a survey including almost 4,000 subjects, which provided the prevalence of hypertension and profile data of hypertensive patients as inputs to the model [14]. However, that study did not include urban regions. We examined the potential effects of this parameter by applying HBP prevalence from the national survey, which included both rural and urban areas, in a separate scenario.

We considered parameter estimates for the non-treatment group as adequately reflected by the less intensive treatment or placebo/reference group in a meta-analysis among Asian populations [3]. The relative risk of treatment and non-treatment groups that we extracted from that study may be considered appropriate for our purposes, despite potential adherence issues.

With regard to adherence, in a report from China the estimated adherence was 40% in both sexes (50% continuation of prescribed medications, with 10% of doses missed among patients continuing treatment). In a previous study in Vietnam [45] we estimated adherence among patients whose HBP was managed at CHS; patients managed at other health facilities were not included. When new data becomes available in future, we can update the model to explore how the cost-effectiveness ratio changes if the level of adherence changes. In general, limitations in available data meant we could not validate the model as extensively as recommended [47].

Chapter 6

ConclusionFrom a health economics perspective, integrating screening for hypertension into routine medical examination and related coverage by health insurance could be recommended. Our current model suggests that screening for hypertension provides a high probability of being cost-effective in preventing CVD. A screening strategy should be selected based on age, sex and screening interval. Screening of males starting at 35-, 45-, or 55 years and females at 55 years until death displayed high probabilities of being cost-effective in all strategies. Screening in females starting at 45 years displayed a high probability of being cost-effective for biannual screening alone or combined with 20% increased treatment. Screening females starting at 35 years gave a low probability of being cost-effective in all strategies, except for biannual screening with increased treatment which still had a 70% probability of being cost-effective. Consistently, lower costs per QALY were found for males than for females. The lifetime model provided lower cost per QALY gained than the 10-year model. Our results may inform and help managers and policy-makers in developing guidelines for hypertension management in Vietnam.

Supporting Information in the appendixS2 Table. Results of univariate sensitivity analysis in 10 years horizon model.S3 Table. Results of univariate sensitivity analysis in lifetime horizon model.

AcknowledgmentsWe thank the staffs of the Thai Nguyen General Hospital for providing data on costs of CVD. We also give special thanks to Doctor Dang Duc Minh, a CVD expert, for his great advice on CVD treatment in Vietnam.

Chapter 6

Health utilities




Chap

ter 6

151


age showed that there were only small changes in the cost per QALY compared with the base case and the same trend remained when compared with the threshold.

Data on individual profiles (age, sex, BP, cholesterol level, smoking status) for hypertensive patients were not available. To overcome this limitation, we conducted a survey including almost 4,000 subjects, which provided the prevalence of hypertension and profile data of hypertensive patients as inputs to the model [14]. However, that study did not include urban regions. We examined the potential effects of this parameter by applying HBP prevalence from the national survey, which included both rural and urban areas, in a separate scenario.

We considered parameter estimates for the non-treatment group as adequately reflected by the less intensive treatment or placebo/reference group in a meta-analysis among Asian populations [3]. The relative risk of treatment and non-treatment groups that we extracted from that study may be considered appropriate for our purposes, despite potential adherence issues.

With regard to adherence, in a report from China the estimated adherence was 40% in both sexes (50% continuation of prescribed medications, with 10% of doses missed among patients continuing treatment). In a previous study in Vietnam [45] we estimated adherence among patients whose HBP was managed at CHS; patients managed at other health facilities were not included. When new data becomes available in future, we can update the model to explore how the cost-effectiveness ratio changes if the level of adherence changes. In general, limitations in available data meant we could not validate the model as extensively as recommended [47].

Chapter 6

ConclusionFrom a health economics perspective, integrating screening for hypertension into routine medical examination and related coverage by health insurance could be recommended. Our current model suggests that screening for hypertension provides a high probability of being cost-effective in preventing CVD. A screening strategy should be selected based on age, sex and screening interval. Screening of males starting at 35-, 45-, or 55 years and females at 55 years until death displayed high probabilities of being cost-effective in all strategies. Screening in females starting at 45 years displayed a high probability of being cost-effective for biannual screening alone or combined with 20% increased treatment. Screening females starting at 35 years gave a low probability of being cost-effective in all strategies, except for biannual screening with increased treatment which still had a 70% probability of being cost-effective. Consistently, lower costs per QALY were found for males than for females. The lifetime model provided lower cost per QALY gained than the 10-year model. Our results may inform and help managers and policy-makers in developing guidelines for hypertension management in Vietnam.

Supporting Information in the appendixS2 Table. Results of univariate sensitivity analysis in 10 years horizon model.S3 Table. Results of univariate sensitivity analysis in lifetime horizon model.

AcknowledgmentsWe thank the staffs of the Thai Nguyen General Hospital for providing data on costs of CVD. We also give special thanks to Doctor Dang Duc Minh, a CVD expert, for his great advice on CVD treatment in Vietnam.

Chapter 6

Health utilities




152


References1. World Health Organization. (2013) A global brief on

hypertension. Available from: http://apps.who.int/iris/bitstream/10665/79059/1/WHO_DCO_WHD_2013.2_eng.pdf?ua=1.

2. Institute for Health Metrics and Evaluation. (2016) Global Burden of Disease. Available from: http://vizhub.healthdata.org/gbd-compare/.



5. Chow CK, Teo KK, Rangarajan S, Islam S, Gupta R, et al. (2013) Prevalence, awareness, treatment, and control of hypertension in rural and urban communities in high-, middle-,and low-income countries. JAMA 310: 959-968.


Chapter 6

7. Yamagishi K, Sato S, Kitamura A, Kiyama M, Okada T, et al. (2012) Cost-effectiveness and budget impact analyses of a long-term hypertension detection and control program for stroke prevention. J Hypertens 30: 1874-1879.

8. Howard K, White S, Salkeld G, McDonald S, Craig JC, et al. (2010) Cost-effectiveness of screening and optimal management for diabetes, hypertension, and chronic kidney disease: a modeled analysis. Value Health 13: 196-208.

9. Van Buuren S, Boshuizen HC and Reijneveld SA. (2006) Toward targeted hypertension screening guidelines. Med Decis Making 26: 145-153.

10. Wang YC, Cheung AM, Bibbins-Domingo K, Prosser LA, Cook NR, et al. (2011) Effectiveness and cost-effectiveness of blood pressure screening in adolescents in the United States. J Pediatr 158: 257-264 e251-257.

11. Deng BH, Liu HW, Pan PC, Mau LW and Chiu HC. (2007) Cost-effectiveness of elderly health examination program: the example of hypertension screening. Kaohsiung J Med Sci 23: 17-24.

12. Cuddy ML. (2005) Treatment of hypertension: guidelines from JNC 7 (the seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure 1). J Pract Nurs 55: 17-21; quiz 22-13.

13. Ministry-of-Health. (2015) Joint annual Health review 2014 -Strengthening prevention and control of non-communicable disease.

Chapter 6

Health utilities




Chap

ter 6

153


References1. World Health Organization. (2013) A global brief on

hypertension. Available from: http://apps.who.int/iris/bitstream/10665/79059/1/WHO_DCO_WHD_2013.2_eng.pdf?ua=1.

2. Institute for Health Metrics and Evaluation. (2016) Global Burden of Disease. Available from: http://vizhub.healthdata.org/gbd-compare/.





Chapter 6

7. Yamagishi K, Sato S, Kitamura A, Kiyama M, Okada T, et al. (2012) Cost-effectiveness and budget impact analyses of a long-term hypertension detection and control program for stroke prevention. J Hypertens 30: 1874-1879.

8. Howard K, White S, Salkeld G, McDonald S, Craig JC, et al. (2010) Cost-effectiveness of screening and optimal management for diabetes, hypertension, and chronic kidney disease: a modeled analysis. Value Health 13: 196-208.

9. Van Buuren S, Boshuizen HC and Reijneveld SA. (2006) Toward targeted hypertension screening guidelines. Med Decis Making 26: 145-153.

10. Wang YC, Cheung AM, Bibbins-Domingo K, Prosser LA, Cook NR, et al. (2011) Effectiveness and cost-effectiveness of blood pressure screening in adolescents in the United States. J Pediatr 158: 257-264 e251-257.

11. Deng BH, Liu HW, Pan PC, Mau LW and Chiu HC. (2007) Cost-effectiveness of elderly health examination program: the example of hypertension screening. Kaohsiung J Med Sci 23: 17-24.

12. Cuddy ML. (2005) Treatment of hypertension: guidelines from JNC 7 (the seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure 1). J Pract Nurs 55: 17-21; quiz 22-13.

13. Ministry-of-Health. (2015) Joint annual Health review 2014 -Strengthening prevention and control of non-communicable disease.

Chapter 6

Health utilities




154


14. Thi-Phuong-Lan Nguyen, CCMS-V, T-B-YN, Vu Thi-Thu-Hang, E. Pamela Wright, M.J. Postma. (2014) Models to Predict the Burden of Cardiovascular Disease Risk in a Rural Mountainous Region of Vietnam. Value in Health 3: 87-93.

15. Ministry of Health of Vietnam. (2010) Diagnosis and Treatment Guideline for hypertension (decision number 3192/MoH, issued on 31/08/2010). Available from: http://thuvienphapluat.vn/archive/Quyet-dinh/Quyet-dinh-3192-QD-BYT-Huong-dan-chan-doan-dieu-tri-tang-huyet-ap-vb112471t17.aspx.

16. Nguyen QN, Pham ST, Nguyen VL, Weinehall L, Bonita R, et al. (2012) Time trends in blood pressure, body mass index and smoking in the Vietnamese population: a meta-analysis from multiple cross-sectional surveys. PLoS One 7: e42825.

17. Briggs. A, Sculpher. M and Claxton K.(2006) Decision Modelling for Health Economic Evaluation. Oxford, England: Oxford University Press.

18. Barzi F, Patel A, Gu D, Sritara P, Lam TH, et al. (2007) Cardiovascular risk prediction tools for populations in Asia. J Epidemiol Community Health 61: 115-121.

19. Hox. JJ.(2010) Multilevel analysis. Techniques and applications. . New York: Routledge.

20. A. F.(2005) Discovering Statistics Using SPSS.: London: Sage Publications.

Chapter 6

21. Liew D, Park HJ and Ko SK. (2009) Results of a Markov model analysis to assess the cost-effectiveness of a single tablet of fixed-dose amlodipine and atorvastatin for the primary prevention of cardiovascular disease in Korea. Clin Ther 31: 2189-2203; discussion 2150-2181.

22. World Health Organization. (2013) Global Health Observatory Data Repository. Available from: http://apps.who.int/gho/data/?theme=main&vid=61830.

23. Hoang VM, Dao LH, Wall S, Nguyen TK and Byass P. (2006) Cardiovascular disease mortality and its association with socioeconomic status: findings from a population-based cohort study in rural Vietnam, 1999-2003. Prev Chronic Dis 3: A89.

24. Grant RL. (2014) Converting an odds ratio to a range of plausible relative risks for better communication of research findings. BMJ 348: f7450.

25. The-World-Bank. (2016) World Development Indicators: Exchange rates and prices. Available from: http://wdi.worldbank.org/table/4.16#.

26. Ha DA and Chisholm D. (2011) Cost-effectiveness analysis of interventions to prevent cardiovascular disease in Vietnam. Health Policy Plan 26: 210-222.

27. Flessa S and Dung NT. (2004) Costing of services of Vietnamese hospitals: identifying costs in one central, two provincial and two district hospitals using a standard methodology. Int J Health Plann Manage 19: 63-77.

Chapter 6

Health utilities




Chap

ter 6

155


14. Thi-Phuong-Lan Nguyen, CCMS-V, T-B-YN, Vu Thi-Thu-Hang, E. Pamela Wright, M.J. Postma. (2014) Models to Predict the Burden of Cardiovascular Disease Risk in a Rural Mountainous Region of Vietnam. Value in Health 3: 87-93.

15. Ministry of Health of Vietnam. (2010) Diagnosis and Treatment Guideline for hypertension (decision number 3192/MoH, issued on 31/08/2010). Available from: http://thuvienphapluat.vn/archive/Quyet-dinh/Quyet-dinh-3192-QD-BYT-Huong-dan-chan-doan-dieu-tri-tang-huyet-ap-vb112471t17.aspx.

16. Nguyen QN, Pham ST, Nguyen VL, Weinehall L, Bonita R, et al. (2012) Time trends in blood pressure, body mass index and smoking in the Vietnamese population: a meta-analysis from multiple cross-sectional surveys. PLoS One 7: e42825.

17. Briggs. A, Sculpher. M and Claxton K.(2006) Decision Modelling for Health Economic Evaluation. Oxford, England: Oxford University Press.


19. Hox. JJ.(2010) Multilevel analysis. Techniques and applications. . New York: Routledge.

20. A. F.(2005) Discovering Statistics Using SPSS.: London: Sage Publications.

Chapter 6

21. Liew D, Park HJ and Ko SK. (2009) Results of a Markov model analysis to assess the cost-effectiveness of a single tablet of fixed-dose amlodipine and atorvastatin for the primary prevention of cardiovascular disease in Korea. Clin Ther 31: 2189-2203; discussion 2150-2181.

22. World Health Organization. (2013) Global Health Observatory Data Repository. Available from: http://apps.who.int/gho/data/?theme=main&vid=61830.

23. Hoang VM, Dao LH, Wall S, Nguyen TK and Byass P. (2006) Cardiovascular disease mortality and its association with socioeconomic status: findings from a population-based cohort study in rural Vietnam, 1999-2003. Prev Chronic Dis 3: A89.

24. Grant RL. (2014) Converting an odds ratio to a range of plausible relative risks for better communication of research findings. BMJ 348: f7450.

25. The-World-Bank. (2016) World Development Indicators: Exchange rates and prices. Available from: http://wdi.worldbank.org/table/4.16#.

26. Ha DA and Chisholm D. (2011) Cost-effectiveness analysis of interventions to prevent cardiovascular disease in Vietnam. Health Policy Plan 26: 210-222.

27. Flessa S and Dung NT. (2004) Costing of services of Vietnamese hospitals: identifying costs in one central, two provincial and two district hospitals using a standard methodology. Int J Health Plann Manage 19: 63-77.

Chapter 6

Health utilities




156


28. MSH. (2014) International Drug Price Indicator Guide.: Medford, Mass:MSH; . 2013 Edition. :[Available from: http://apps.who.int/medicinedocs/documents/s21497en/s21497en.pdf.

29. Hutton G and Baltussen R. (2005) Cost valuation in resource-poor settings. Health Policy Plan 20: 252-259.

30. Nguyen TP, Nguyen TB, Nguyen TT, Vinh Hac V, Le HH, et al. (2014) Direct costs of hypertensive patients admitted to hospital in Vietnam- a bottom-up micro-costing analysis. BMC Health Serv Res 14: 514.

31. Nguyen HL, Nguyen QN, Ha DA, Phan DT, Nguyen NH, et al. (2014) Prevalence of comorbidities and their impact on hospital management and short-term outcomes in Vietnamese patients hospitalized with a first acute myocardial infarction. PLoS One 9: e108998.

32. Hoi le V, Chuc NT and Lindholm L. (2010) Health-related quality of life, and its determinants, among older people in rural Vietnam. BMC Public Health 10: 549.

33. Wee HL, Machin D, Loke WC, Li SC, Cheung YB, et al. (2007) Assessing differences in utility scores: a comparison of four widely used preference-based instruments. Value Health 10: 256-265.

34. Nguyen TP, Krabbe PF, Nguyen TB, Schuiling-Veninga CC, Wright EP, et al. (2015) Utilities of Patients with Hypertension in Northern Vietnam. PLoS One 10: e0139560.

Chapter 6

35. Kang EJ and Ko SK. (2009) A catalogue of EQ-5D utility weights for chronic diseases among noninstitutionalized community residents in Korea. Value Health 12 Suppl 3: S114-117.

36. McManus RJ, Mant J, Bray EP, Holder R, Jones MI, et al. (2010) Telemonitoring and self-management in the control of hypertension (TASMINH2): a randomised controlled trial. Lancet 376: 163-172.

37. Luengo-Fernandez R, Gray AM, Bull L, Welch S, Cuthbertson F, et al. (2013) Quality of life after TIA and stroke: ten-year results of the Oxford Vascular Study. Neurology 81: 1588-1595.

38. Salomon JA, Vos T, Hogan DR, Gagnon M, Naghavi M, et al. (2012) Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010. Lancet 380: 2129-2143.

39. Smith DW, Davies EW, Wissinger E, Huelin R, Matza LS, et al. (2013) A systematic literature review of cardiovascular event utilities. Expert Rev Pharmacoecon Outcomes Res 13: 767-790.

40. World-Health-Organization. (Cost effectiveness and strategic planning (WHO-CHOICE). Available from: http://www.who.int/choice/costs/CER_levels/en/.

41. World-Health-Organization. (Choosing Interventions that are Cost Effective (WHO-CHOICE). Available from: http://www.who.int/choice/costs/ppp/en/.

Chapter 6

Health utilities




Chap

ter 6

157


28. MSH. (2014) International Drug Price Indicator Guide.: Medford, Mass:MSH; . 2013 Edition. :[Available from: http://apps.who.int/medicinedocs/documents/s21497en/s21497en.pdf.

29. Hutton G and Baltussen R. (2005) Cost valuation in resource-poor settings. Health Policy Plan 20: 252-259.


31. Nguyen HL, Nguyen QN, Ha DA, Phan DT, Nguyen NH, et al. (2014) Prevalence of comorbidities and their impact on hospital management and short-term outcomes in Vietnamese patients hospitalized with a first acute myocardial infarction. PLoS One 9: e108998.


33. Wee HL, Machin D, Loke WC, Li SC, Cheung YB, et al. (2007) Assessing differences in utility scores: a comparison of four widely used preference-based instruments. Value Health 10: 256-265.


Chapter 6

35. Kang EJ and Ko SK. (2009) A catalogue of EQ-5D utility weights for chronic diseases among noninstitutionalized community residents in Korea. Value Health 12 Suppl 3: S114-117.

36. McManus RJ, Mant J, Bray EP, Holder R, Jones MI, et al. (2010) Telemonitoring and self-management in the control of hypertension (TASMINH2): a randomised controlled trial. Lancet 376: 163-172.

37. Luengo-Fernandez R, Gray AM, Bull L, Welch S, Cuthbertson F, et al. (2013) Quality of life after TIA and stroke: ten-year results of the Oxford Vascular Study. Neurology 81: 1588-1595.

38. Salomon JA, Vos T, Hogan DR, Gagnon M, Naghavi M, et al. (2012) Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010. Lancet 380: 2129-2143.

39. Smith DW, Davies EW, Wissinger E, Huelin R, Matza LS, et al. (2013) A systematic literature review of cardiovascular event utilities. Expert Rev Pharmacoecon Outcomes Res 13: 767-790.

40. World-Health-Organization. (Cost effectiveness and strategic planning (WHO-CHOICE). Available from: http://www.who.int/choice/costs/CER_levels/en/.

41. World-Health-Organization. (Choosing Interventions that are Cost Effective (WHO-CHOICE). Available from: http://www.who.int/choice/costs/ppp/en/.

Chapter 6

Health utilities




158


42. World-Bank. (GDP per capita, PPP (current international $). Available from: http://data.worldbank.org/indicator/NY.GDP.PCAP.PP.CD.

43. Montgomery AA, Fahey T, Ben-Shlomo Y and Harding J. (2003) The influence of absolute cardiovascular risk, patient utilities, and costs on the decision to treat hypertension: a Markov decision analysis. J Hypertens 21: 1753-1759.

44. Ly D, Fu AZ and Hebert C. (2009) Cost effectiveness analysis of a hypertension management program in patients with type 2 diabetes. J Clin Hypertens (Greenwich) 11: 116-124.

45. Nguyen T-P-L, CCMS-V, MJP. (2014) Cardiovascular risk, gender and medication adherence in a rural area of Vietnam. Available from: http://www.ispor.org/research_pdfs/48/pdffiles/PCV114.pdf.

46. Gu D, He J, Coxson PG, Rasmussen PW, Huang C, et al. (2015) The Cost-Effectiveness of Low-Cost Essential Antihypertensive Medicines for Hypertension Control in China: A Modelling Study. PLoS Med 12: e1001860.

47. Eddy DM, Hollingworth W, Caro JJ, Tsevat J, McDonald KM, et al. (2012) Model transparency and validation: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force--7. Value Health 15: 843-850.

Chap

ter 6

159


42. World-Bank. (GDP per capita, PPP (current international $). Available from: http://data.worldbank.org/indicator/NY.GDP.PCAP.PP.CD.

43. Montgomery AA, Fahey T, Ben-Shlomo Y and Harding J. (2003) The influence of absolute cardiovascular risk, patient utilities, and costs on the decision to treat hypertension: a Markov decision analysis. J Hypertens 21: 1753-1759.

44. Ly D, Fu AZ and Hebert C. (2009) Cost effectiveness analysis of a hypertension management program in patients with type 2 diabetes. J Clin Hypertens (Greenwich) 11: 116-124.

45. Nguyen T-P-L, CCMS-V, MJP. (2014) Cardiovascular risk, gender and medication adherence in a rural area of Vietnam. Available from: http://www.ispor.org/research_pdfs/48/pdffiles/PCV114.pdf.

46. Gu D, He J, Coxson PG, Rasmussen PW, Huang C, et al. (2015) The Cost-Effectiveness of Low-Cost Essential Antihypertensive Medicines for Hypertension Control in China: A Modelling Study. PLoS Med 12: e1001860.

47. Eddy DM, Hollingworth W, Caro JJ, Tsevat J, McDonald KM, et al. (2012) Model transparency and validation: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force--7. Value Health 15: 843-850.

Chapter 7

General discussion

162

Chapter 7

Background

Vietnam is experiencing a rapid socio-economic development and as a consequence, there is also a shift in disease patterns [1]. At present, the health sector in Vietnam is urged to deal with the increasing problems of non-communicable disease, while it lacks experience and evidence to identify appropriate and effective solutions. Cardiovascular diseases (CVD) are currently a main health problem, accounting for 14% of the burden of disease in the world, and 12% in Vietnam [1, 2]. To deal with this problem, interventions focusing on both primary and secondary prevention are needed using a range of strategies [3]. The example of CVD as a non-communicable disease[4] is taken to conduct several studies to provide evidence that could inform decision makers in Vietnam and in other developing countries with similar conditions, to assist them in making choices and allocating resources in this area.

Research questions

As outlined in the introduction section, the main research question read “What are cost-effective ways of screening for and managing hypertension to prevent development of CVD in Vietnam?”.

Sub-questions concerned:

1. Which model is most appropriate to measure CVD risks in Vietnam?


3. How will adherence to treatment influence outcomes?4. What is the burden of disease if hypertension is not treated or

controlled?

Chap

ter 7

163

Chapter 7

Background

Vietnam is experiencing a rapid socio-economic development and as a consequence, there is also a shift in disease patterns [1]. At present, the health sector in Vietnam is urged to deal with the increasing problems of non-communicable disease, while it lacks experience and evidence to identify appropriate and effective solutions. Cardiovascular diseases (CVD) are currently a main health problem, accounting for 14% of the burden of disease in the world, and 12% in Vietnam [1, 2]. To deal with this problem, interventions focusing on both primary and secondary prevention are needed using a range of strategies [3]. The example of CVD as a non-communicable disease[4] is taken to conduct several studies to provide evidence that could inform decision makers in Vietnam and in other developing countries with similar conditions, to assist them in making choices and allocating resources in this area.

Research questions

As outlined in the introduction section, the main research question read “What are cost-effective ways of screening for and managing hypertension to prevent development of CVD in Vietnam?”.

Sub-questions concerned:

1. Which model is most appropriate to measure CVD risks in Vietnam?


3. How will adherence to treatment influence outcomes?4. What is the burden of disease if hypertension is not treated or

controlled?

164

General discussion


6. What quality of life is experienced by hypertensive patients in Vietnamese population?

7. Which is the most cost-effective population screening strategy to identify patients with hypertension at risk for future complications, modelling various screening and treatment coverage scenarios?

8. What recommendations can be given to the national programs to identify patients with hypertension and to minimize risks of further disease development in those patients?

Below, these research (sub)questions are revisited to analyze objectives/findings achieved and still remaining further work to be addressed in follow-up studies.

Main findings

We have confirmed the importance of hypertension as a public health problem in Vietnam. In one rural region, the prevalence of undiagnosed hypertension was 12.3%. These hypertensive patients were not aware of their health problem before the screening was done (Chapter 2) [5]; ergo, the relevance of our study was confirmed. To estimate the effects of this large number of undiagnosed hypertensive patients on future health consequences, a cost analysis was applied which suggested that if high blood pressure was left untreated or uncontrolled, the burden to the hospital and health care system in the future would be considerable (Chapter 3; sub-question 4). For example, for hypertension, the length of hospital stay was six days, with a total direct cost only for hypertension at Int$ 147 per hospitalization, while the cost of annual hypertension treatment was

Chapter 7

Int$ 69 [6, 7]. Furthermore, if complications arose, such as myocardial infarction (MI), the cost per acute hospitalization rose to Int$ 12,371, in the context of a total per capita expenditure on health of Int$ 308 in 2013 [6, 8, 9]. The burden of hypertension for the patients was also shown to be relevant, measured by health utility (Chapter 4; sub-question 4 and 6), at 0.73 among hypertensive patients, while health utility among the general population in Vietnam was 0.93 if converted in the same instrument [10,11]. Furthermore, the screening and management strategy most likely to be (cost-)effective was identified by modeling. Notably, this analysis indicated that screening combined with increasing the coverage of treatment was most cost-effective in both ten years and lifetime scenarios (Chapter 6; sub-question 7). The key factors to guide screening were age and sex; screening the older aged population was more cost-effective, and screening males was more cost-effective than screening females of the same age [7].Obviously, such models require good input data and include assumptions, which are prone to limitations because many key data are not available in a reliable form. For example, our investigation into adherence to treatment revealed that it was more complex than it at first appeared; providing prescriptions or medicines was not a guarantee that patients would use them, but also, patients could and did access medicines on the free market even if they are not using medicines prescribed by the health care stations (sub-question 2 and 3)[12]. Together, all of the above pieces of information can support the development of appropriate strategies and guidelines for the health system in Vietnam to deal effectively with hypertension and its consequences (sub-question 8). Our approach has demonstrated the usefulness of combining health economics and modelling to guide development of policies to manage non-communicable diseases. These diseases are expected to present the main health problems in the

Chap

ter 7

165

General discussion


6. What quality of life is experienced by hypertensive patients in Vietnamese population?

7. Which is the most cost-effective population screening strategy to identify patients with hypertension at risk for future complications, modelling various screening and treatment coverage scenarios?

8. What recommendations can be given to the national programs to identify patients with hypertension and to minimize risks of further disease development in those patients?

Below, these research (sub)questions are revisited to analyze objectives/findings achieved and still remaining further work to be addressed in follow-up studies.

Main findings

We have confirmed the importance of hypertension as a public health problem in Vietnam. In one rural region, the prevalence of undiagnosed hypertension was 12.3%. These hypertensive patients were not aware of their health problem before the screening was done (Chapter 2) [5]; ergo, the relevance of our study was confirmed. To estimate the effects of this large number of undiagnosed hypertensive patients on future health consequences, a cost analysis was applied which suggested that if high blood pressure was left untreated or uncontrolled, the burden to the hospital and health care system in the future would be considerable (Chapter 3; sub-question 4). For example, for hypertension, the length of hospital stay was six days, with a total direct cost only for hypertension at Int$ 147 per hospitalization, while the cost of annual hypertension treatment was

Chapter 7

Int$ 69 [6, 7]. Furthermore, if complications arose, such as myocardial infarction (MI), the cost per acute hospitalization rose to Int$ 12,371, in the context of a total per capita expenditure on health of Int$ 308 in 2013 [6, 8, 9]. The burden of hypertension for the patients was also shown to be relevant, measured by health utility (Chapter 4; sub-question 4 and 6), at 0.73 among hypertensive patients, while health utility among the general population in Vietnam was 0.93 if converted in the same instrument [10,11]. Furthermore, the screening and management strategy most likely to be (cost-)effective was identified by modeling. Notably, this analysis indicated that screening combined with increasing the coverage of treatment was most cost-effective in both ten years and lifetime scenarios (Chapter 6; sub-question 7). The key factors to guide screening were age and sex; screening the older aged population was more cost-effective, and screening males was more cost-effective than screening females of the same age [7].Obviously, such models require good input data and include assumptions, which are prone to limitations because many key data are not available in a reliable form. For example, our investigation into adherence to treatment revealed that it was more complex than it at first appeared; providing prescriptions or medicines was not a guarantee that patients would use them, but also, patients could and did access medicines on the free market even if they are not using medicines prescribed by the health care stations (sub-question 2 and 3)[12]. Together, all of the above pieces of information can support the development of appropriate strategies and guidelines for the health system in Vietnam to deal effectively with hypertension and its consequences (sub-question 8). Our approach has demonstrated the usefulness of combining health economics and modelling to guide development of policies to manage non-communicable diseases. These diseases are expected to present the main health problems in the

166

General discussion

future in Vietnam, including CVD, along with diabetes, cancer and chronic obstructive pulmonary disease (COPD).

Implications

CVD risk model

To predict the CVD risks and consequences, models are used based on data collected either in dedicated studies or taken from routine national databases. In the first study reported here (Chapter 2), we identified the most suitable CVD prediction model for Vietnam. Among various available models, we explored three that seemedrepresentative for Asian populations or potentially useful for application in Vietnam [13-15]. One challenge was that the model structures and CVD risk outcome varies between the models, because of the different components in each model. For example, input variables, outcome variables, time horizon and values in each equation differed [5, 16]. Notably, patients with a history of CVD have not yet been included in these models, which should be noted in all further analysis or clinical settings [17]. Although the existing evidence did not permit us to identify definitively the best prediction model for the Vietnam population, based on best available evidence we could recommend that the Asian and Chinese Multiple-provincial Cohort Study as the most appropriate for Vietnam (sub-questions 1,5 and 8).

Adherence to hypertensive treatment

In most modelling exercises, we had to make assumptions about the percentage of patients receiving medical treatment and we assumedthat these patients are actually following the treatment regime as prescribed. Nevertheless, it is widely known that not all patients will follow the doctor’s instructions exactly [18]. This is especially true for chronic conditions when patients are asked to continue using the medication for years to avoid the development of complications and

Chapter 7

more serious diseases. Similar to a previous review [18], our study on adherence to antihypertensive medication (Chapter 3; sub-questions 2 and 3) revealed that only half of the patients was adherent to the prescribed treatment [12]. Various methods have been applied to measure adherence to medicine, each with advantages and disadvantages [19]. One finding from our study, possibly also relevant for other research in developing countries, is that quantifying the level of adherence based on the uptake of prescriptions may underestimate actual intake of medication when patients can buy medicines without a prescription. Furthermore, factors were identified for doctors to focus upon in their efforts to improve adherence of patients, such as age and awareness of complications of hypertension (sub-question 8). Having a history of CVD, and thus a higher risk of CVD in the future, was elsewhere reported to be a predictor to better adherence [20] but we could not confirm this in our study [12]. Adherence improvement is associated with better outcomes, but few studies on solutions to improve adherence to medicine could demonstrate success [21], so there is still much room for improvement in enhancing adherence to medicine in real life practice.

Burden of disease

An economic evaluation to estimate the burden of disease (Chapter 4) indicated that undiagnosed or incorrect treatment of hypertension involves major resources and costs [6, 22], even when patients are admitted to a hospital only for hypertension. This came out of the analysis even though we did not take indirect costs into account, such as production loss due to illness. The total costs would then be much higher than only the direct costs, in particular in the Vietnamese setting where most nursing care is provided by family members. Based on the burden of disease, the results of this study can also

Chap

ter 7

167

General discussion

future in Vietnam, including CVD, along with diabetes, cancer and chronic obstructive pulmonary disease (COPD).

Implications

CVD risk model

To predict the CVD risks and consequences, models are used based on data collected either in dedicated studies or taken from routine national databases. In the first study reported here (Chapter 2), we identified the most suitable CVD prediction model for Vietnam. Among various available models, we explored three that seemedrepresentative for Asian populations or potentially useful for application in Vietnam [13-15]. One challenge was that the model structures and CVD risk outcome varies between the models, because of the different components in each model. For example, input variables, outcome variables, time horizon and values in each equation differed [5, 16]. Notably, patients with a history of CVD have not yet been included in these models, which should be noted in all further analysis or clinical settings [17]. Although the existing evidence did not permit us to identify definitively the best prediction model for the Vietnam population, based on best available evidence we could recommend that the Asian and Chinese Multiple-provincial Cohort Study as the most appropriate for Vietnam (sub-questions 1,5 and 8).

Adherence to hypertensive treatment

In most modelling exercises, we had to make assumptions about the percentage of patients receiving medical treatment and we assumedthat these patients are actually following the treatment regime as prescribed. Nevertheless, it is widely known that not all patients will follow the doctor’s instructions exactly [18]. This is especially true for chronic conditions when patients are asked to continue using the medication for years to avoid the development of complications and

Chapter 7

more serious diseases. Similar to a previous review [18], our study on adherence to antihypertensive medication (Chapter 3; sub-questions 2 and 3) revealed that only half of the patients was adherent to the prescribed treatment [12]. Various methods have been applied to measure adherence to medicine, each with advantages and disadvantages [19]. One finding from our study, possibly also relevant for other research in developing countries, is that quantifying the level of adherence based on the uptake of prescriptions may underestimate actual intake of medication when patients can buy medicines without a prescription. Furthermore, factors were identified for doctors to focus upon in their efforts to improve adherence of patients, such as age and awareness of complications of hypertension (sub-question 8). Having a history of CVD, and thus a higher risk of CVD in the future, was elsewhere reported to be a predictor to better adherence [20] but we could not confirm this in our study [12]. Adherence improvement is associated with better outcomes, but few studies on solutions to improve adherence to medicine could demonstrate success [21], so there is still much room for improvement in enhancing adherence to medicine in real life practice.

Burden of disease

An economic evaluation to estimate the burden of disease (Chapter 4) indicated that undiagnosed or incorrect treatment of hypertension involves major resources and costs [6, 22], even when patients are admitted to a hospital only for hypertension. This came out of the analysis even though we did not take indirect costs into account, such as production loss due to illness. The total costs would then be much higher than only the direct costs, in particular in the Vietnamese setting where most nursing care is provided by family members. Based on the burden of disease, the results of this study can also

168

General discussion

suggest a specific amount to be applied for health insurance reimbursement for those patients (sub-questions 4 and 8).

Health utilities

Another way of measuring the burden of disease from the perspective of patient-reported outcomes is to measure health utilities. In Chapter 5 we not only presented evidence on health utility among hypertensive patients in Vietnam but also used that result to weight the health utility of other health states, such as the ‘healthy’, ‘stroke’, or ‘myocardial infarction’ states, citing utility values from other studies when needed for a further analysis [7]. Health utilities in this hypertensive population were indeed lower than in the general population [11]. We also found lower health utilities among patients who were older, female or had more than three comorbidities [10]. The value of health utilities among hypertensive patients could potentially contribute to analysis of global health issues; it has, for example, not yet been shown in the Global Burden of Disease database [1].

Cost effectiveness of screening

In the context of the current lack of evidence to guide hypertension management in Vietnam, there was a need to identify how to integrate screening for hypertension into routine care and coverage by health insurance. We therefore conducted a study on cost-effectiveness of different strategies to screen for hypertension that applied local data. The result suggested that an optimal screening strategy would be based on age, sex and screening interval (such as annual or biannual) and should be combined with enhanced coverage of treatment. Astrategy of regular screening of males starting at 35-, 45-, or 55 years and of females at 55 years exhibited high probabilities of being cost-effective in all variants (main question and sub-questions 7 and 8).Even though we believe that our model provides the best evidence so

Chapter 7

far to guide screening in Vietnam, we do note considerations in the interpretation of the modelling study, because certain important aspects have yet not been integrated in the model. First, differences in reductions in relative risk for CVD among patients with or without ahistory of CVD are not integrated in the model. However, a recent meta-analysis showed that the effect of treatment of high blood pressure in reducing the risk of CVD (including IHD and stroke) was similar in patients with or without a history of CVD. Notably, the risk of CVD is higher in people with a history of CVD but the reduction in risk of CVD is also larger [23]. That means, for the modelling exercise, we could justify applying the same risk reduction rate for both groups of patients. Second, adherence to medicine is a huge issue as it is directly associated with clinical outcomes [24], but adherence has not yet been adequately integrated in the model. Therefore, the model should be updated on this aspect when further adherence datain relation to clinical outcomes become available.

Limitations

Most of the research questions posed at the start of this thesis have been answered; two still remain with specific aspects to be addressed. The first question is on how adherence to treatment will influenceclinical outcomes and the second concerns the future risks of CVD in various scenarios of non-treatment or non-adherence. Studies on the effects of medication adherence among hypertensive patients on outcomes such as MI or stroke have been conducted in Italy and The Netherlands [25, 26], but we still lack such evidence for developing countries. We conducted a prospective study in which patients were followed for one year, however we were not able to record complications such as MI or stroke for various reasons. Firstly, up to now, Vietnam does not have a health information system recording

Chap

ter 7

169

General discussion

suggest a specific amount to be applied for health insurance reimbursement for those patients (sub-questions 4 and 8).

Health utilities

Another way of measuring the burden of disease from the perspective of patient-reported outcomes is to measure health utilities. In Chapter 5 we not only presented evidence on health utility among hypertensive patients in Vietnam but also used that result to weight the health utility of other health states, such as the ‘healthy’, ‘stroke’, or ‘myocardial infarction’ states, citing utility values from other studies when needed for a further analysis [7]. Health utilities in this hypertensive population were indeed lower than in the general population [11]. We also found lower health utilities among patients who were older, female or had more than three comorbidities [10]. The value of health utilities among hypertensive patients could potentially contribute to analysis of global health issues; it has, for example, not yet been shown in the Global Burden of Disease database [1].

Cost effectiveness of screening

In the context of the current lack of evidence to guide hypertension management in Vietnam, there was a need to identify how to integrate screening for hypertension into routine care and coverage by health insurance. We therefore conducted a study on cost-effectiveness of different strategies to screen for hypertension that applied local data. The result suggested that an optimal screening strategy would be based on age, sex and screening interval (such as annual or biannual) and should be combined with enhanced coverage of treatment. Astrategy of regular screening of males starting at 35-, 45-, or 55 years and of females at 55 years exhibited high probabilities of being cost-effective in all variants (main question and sub-questions 7 and 8).Even though we believe that our model provides the best evidence so

Chapter 7

far to guide screening in Vietnam, we do note considerations in the interpretation of the modelling study, because certain important aspects have yet not been integrated in the model. First, differences in reductions in relative risk for CVD among patients with or without ahistory of CVD are not integrated in the model. However, a recent meta-analysis showed that the effect of treatment of high blood pressure in reducing the risk of CVD (including IHD and stroke) was similar in patients with or without a history of CVD. Notably, the risk of CVD is higher in people with a history of CVD but the reduction in risk of CVD is also larger [23]. That means, for the modelling exercise, we could justify applying the same risk reduction rate for both groups of patients. Second, adherence to medicine is a huge issue as it is directly associated with clinical outcomes [24], but adherence has not yet been adequately integrated in the model. Therefore, the model should be updated on this aspect when further adherence datain relation to clinical outcomes become available.

Limitations

Most of the research questions posed at the start of this thesis have been answered; two still remain with specific aspects to be addressed. The first question is on how adherence to treatment will influenceclinical outcomes and the second concerns the future risks of CVD in various scenarios of non-treatment or non-adherence. Studies on the effects of medication adherence among hypertensive patients on outcomes such as MI or stroke have been conducted in Italy and The Netherlands [25, 26], but we still lack such evidence for developing countries. We conducted a prospective study in which patients were followed for one year, however we were not able to record complications such as MI or stroke for various reasons. Firstly, up to now, Vietnam does not have a health information system recording

170

General discussion

clinical data for these types of diseases. Secondly, when patients are to be hospitalized, they have the right to choose which health facility they will use, and there is no formal feedback system to ensure that the relevant data come back to the community health center. Thirdly, it may be difficult for patients to remember exactly what their diagnosis was, so when trying to obtain data by interviewing them later, we expect a strong effect of recall bias.

The number of subgroups integrated into the modelling exercises was lower than specified as optimal.

The first reason was that half of the patients were non-adherent to medication and further detailing per patient subgroup appeared difficult [12]. In Chapter 4, the finding is reported that costs for hospitalization as a result of hypertension were high compared to annual treatment [6]. However, we could not obtain data on the actual prevalence of hospitalization due to uncontrolled high blood pressureamong those either on current treatment or non-treated or non-adherent. That is why we could not take those groups separately into account in the model.

Secondly, no data are available on the association between adherence to medical treatment and the occurrence of CVD complications. Therefore, we had to assume that patients on treatment are adherent and we could not deal with the potential issue of lower adherence and the resulting higher CVD risk.

Finally, we applied data from a CVD risk prediction model to explore trends in Vietnam. The current model has not yet been validated for application in the Vietnamese population [27]. This has to be done in future when the required data become available.

Chapter 7

Future prospects

Apart from addressing the data limitations described above, further research could focus on influence of adherence to treatment on outcomes and the future risks of CVD in case of non-treatment or non-adherence. These studies should be conducted to provide the data needed for predictions and planning of hypertension management.Association of adherence and its outcome will take into account for a scenario of the cost-effectiveness analysis.

In Chapter 4, we suggested several factors that may influence adherence to medicines. A review on interventions to improve adherence to treatment has been published [28] but we still need to identify appropriate solutions to increase coverage of treatment and adherence to medicines in the Vietnamese setting. Furthermore, interventions to improve adherence and an analysis of the cost-effectiveness of improving adherence are needed for developingcountries, within the context of a health-economics adequately incorporating adherence as one important element.

Recommendations & Conclusions

Previous reports on health economic evaluation recommended several interventions to reduce the burden of cardiovascular diseases, such as treatment of hypertension, treatment of hypertension combined with treatment of CVD events, and population-based mass media strategies aimed at reducing levels of cholesterol and dietary salt intake [2]. Our study adds additional evidence to inform managers and policy-makers who are developing guidelines for management of hypertension in Vietnam (sub-question 8). Screening for hypertension is cost-effective in cardiovascular prevention and therefore screening for hypertension

Chap

ter 7

171

General discussion

clinical data for these types of diseases. Secondly, when patients are to be hospitalized, they have the right to choose which health facility they will use, and there is no formal feedback system to ensure that the relevant data come back to the community health center. Thirdly, it may be difficult for patients to remember exactly what their diagnosis was, so when trying to obtain data by interviewing them later, we expect a strong effect of recall bias.

The number of subgroups integrated into the modelling exercises was lower than specified as optimal.

The first reason was that half of the patients were non-adherent to medication and further detailing per patient subgroup appeared difficult [12]. In Chapter 4, the finding is reported that costs for hospitalization as a result of hypertension were high compared to annual treatment [6]. However, we could not obtain data on the actual prevalence of hospitalization due to uncontrolled high blood pressureamong those either on current treatment or non-treated or non-adherent. That is why we could not take those groups separately into account in the model.

Secondly, no data are available on the association between adherence to medical treatment and the occurrence of CVD complications. Therefore, we had to assume that patients on treatment are adherent and we could not deal with the potential issue of lower adherence and the resulting higher CVD risk.

Finally, we applied data from a CVD risk prediction model to explore trends in Vietnam. The current model has not yet been validated for application in the Vietnamese population [27]. This has to be done in future when the required data become available.

Chapter 7

Future prospects

Apart from addressing the data limitations described above, further research could focus on influence of adherence to treatment on outcomes and the future risks of CVD in case of non-treatment or non-adherence. These studies should be conducted to provide the data needed for predictions and planning of hypertension management.Association of adherence and its outcome will take into account for a scenario of the cost-effectiveness analysis.

In Chapter 4, we suggested several factors that may influence adherence to medicines. A review on interventions to improve adherence to treatment has been published [28] but we still need to identify appropriate solutions to increase coverage of treatment and adherence to medicines in the Vietnamese setting. Furthermore, interventions to improve adherence and an analysis of the cost-effectiveness of improving adherence are needed for developingcountries, within the context of a health-economics adequately incorporating adherence as one important element.

Recommendations & Conclusions

Previous reports on health economic evaluation recommended several interventions to reduce the burden of cardiovascular diseases, such as treatment of hypertension, treatment of hypertension combined with treatment of CVD events, and population-based mass media strategies aimed at reducing levels of cholesterol and dietary salt intake [2]. Our study adds additional evidence to inform managers and policy-makers who are developing guidelines for management of hypertension in Vietnam (sub-question 8). Screening for hypertension is cost-effective in cardiovascular prevention and therefore screening for hypertension

172

General discussion

could be integrated into routine health practice with a focus on those over 35 years for males and over 55 years for females.

Even though efforts to encourage patients to adhere to medicines as prescribed is challenging from both the clinical and the economic perspectives, we found that increasing the coverage of treatment and improving adherence to treatment are very important interventions toreduce the burden of disease. In addition, from fieldwork and a previous study [1], we recognized that adherence to screening also needs to be emphasized. Therefore, with regard to adherence, experience from our study suggests that programs to encourage people to undergo screening for hypertension as well as adherence to screening need attention, especially in remote or mountainous areas.

Multiple models to predict CVD risk have been introduced in recent years. Doctors have several choices but that also makes it difficult to select the most suitable model to apply in Vietnam. Our study suggested that models developed on The Asian and Chinese Multiple-provincial Cohort Study are potentially appropriate to apply in the current context in which real data is lacking to develop a local model.

For continued monitoring and planning of hypertension management, the Ministry of Health should create a database system that can share information on disease history, clinical data and treatment for chronic disease, as these aspects require continuous attention. The data could also serve to update the model. An electronic health information system to follow chronic patients would create major benefits for clinical practice, monitoring risk factors, research and ultimately for policy makers to manage non-communicable disease.

The results presented in this thesis provide additional evidence that hypertension importantly contributes to the current burden of disease in Vietnam, from both clinical and economic perspectives. Prediction

Chapter 7

of risk using the Asian and China models may be used as tools in estimating CVD risk in Vietnam and seem valid, while further analyses using real-life data for confirmation are needed. Health utilities of hypertensive patients could serve as a reference in a similar setting, although not yet reported in the Global Burden of Disease database. The value of the costs of hospitalization of hypertensive patients could serve as a reference for reimbursement by insurance companies. Medication adherence is an issue that needs to be considered in hypertension management.

All the evidence gathered, together with the integration into results of the modelling study, leads me to strongly recommend that screening and managing hypertension should be integrated into routine care, using age as guiding factor.

Chap

ter 7

173

General discussion

could be integrated into routine health practice with a focus on those over 35 years for males and over 55 years for females.

Even though efforts to encourage patients to adhere to medicines as prescribed is challenging from both the clinical and the economic perspectives, we found that increasing the coverage of treatment and improving adherence to treatment are very important interventions toreduce the burden of disease. In addition, from fieldwork and a previous study [1], we recognized that adherence to screening also needs to be emphasized. Therefore, with regard to adherence, experience from our study suggests that programs to encourage people to undergo screening for hypertension as well as adherence to screening need attention, especially in remote or mountainous areas.

Multiple models to predict CVD risk have been introduced in recent years. Doctors have several choices but that also makes it difficult to select the most suitable model to apply in Vietnam. Our study suggested that models developed on The Asian and Chinese Multiple-provincial Cohort Study are potentially appropriate to apply in the current context in which real data is lacking to develop a local model.

For continued monitoring and planning of hypertension management, the Ministry of Health should create a database system that can share information on disease history, clinical data and treatment for chronic disease, as these aspects require continuous attention. The data could also serve to update the model. An electronic health information system to follow chronic patients would create major benefits for clinical practice, monitoring risk factors, research and ultimately for policy makers to manage non-communicable disease.

The results presented in this thesis provide additional evidence that hypertension importantly contributes to the current burden of disease in Vietnam, from both clinical and economic perspectives. Prediction

Chapter 7

of risk using the Asian and China models may be used as tools in estimating CVD risk in Vietnam and seem valid, while further analyses using real-life data for confirmation are needed. Health utilities of hypertensive patients could serve as a reference in a similar setting, although not yet reported in the Global Burden of Disease database. The value of the costs of hospitalization of hypertensive patients could serve as a reference for reimbursement by insurance companies. Medication adherence is an issue that needs to be considered in hypertension management.

All the evidence gathered, together with the integration into results of the modelling study, leads me to strongly recommend that screening and managing hypertension should be integrated into routine care, using age as guiding factor.

174

General discussion

References





5. Thi Phuong Lan N, C.C.M S-V, Thi Bach Yen N, Thi Thu Hang V, E. Pamela W, et al. (2014) Models to Predict the Burden of Cardiovascular Disease Risk in a Rural Mountainous Region of Vietnam. Value in Health Regional Issue: 87-93.


7. Thi-Phuong-Lan N, E. Pamela W, Thanh-Trung N, C.C.M. S-V, MJ. B, et al. (2016) Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam.

8. World-Health-Organization. (2013) Country profile.Available from: http://www.who.int/countries/vnm/en/.

Chapter 7

9. TPL N, TT N and MJP P. (2015) Economic burden of acute myocardial infarction in Vietnam. Available from: http://www.ispor.org/Event/ProgramList/2015Milan?type=Poster&sess=I&acode=PCV.




13. Liu J, Hong Y, D'Agostino RB, Sr., Wu Z, Wang W, et al. (2004) Predictive value for the Chinese population of the Framingham CHD risk assessment tool compared with the Chinese Multi-Provincial Cohort Study. JAMA 291: 2591-2599.

14. Framingham-Heart-Study. (2013) Hard coronary heart disease (10-year risk). Available from: http://www.framinghamheartstudy.org/risk-functions/coronary-heart-disease/hard-10-year-risk.php.


Chap

ter 7

175

General discussion

References





5. Thi Phuong Lan N, C.C.M S-V, Thi Bach Yen N, Thi Thu Hang V, E. Pamela W, et al. (2014) Models to Predict the Burden of Cardiovascular Disease Risk in a Rural Mountainous Region of Vietnam. Value in Health Regional Issue: 87-93.


7. Thi-Phuong-Lan N, E. Pamela W, Thanh-Trung N, C.C.M. S-V, MJ. B, et al. (2016) Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam.

8. World-Health-Organization. (2013) Country profile.Available from: http://www.who.int/countries/vnm/en/.

Chapter 7

9. TPL N, TT N and MJP P. (2015) Economic burden of acute myocardial infarction in Vietnam. Available from: http://www.ispor.org/Event/ProgramList/2015Milan?type=Poster&sess=I&acode=PCV.




13. Liu J, Hong Y, D'Agostino RB, Sr., Wu Z, Wang W, et al. (2004) Predictive value for the Chinese population of the Framingham CHD risk assessment tool compared with the Chinese Multi-Provincial Cohort Study. JAMA 291: 2591-2599.

14. Framingham-Heart-Study. (2013) Hard coronary heart disease (10-year risk). Available from: http://www.framinghamheartstudy.org/risk-functions/coronary-heart-disease/hard-10-year-risk.php.


176

General discussion

16. Stevanovic J, Postma MJ and Pechlivanoglou P. (2012) A systematic review on the application of cardiovascular risk prediction models in pharmacoeconomics, with a focus on primary prevention. Eur J Prev Cardiol 19: 42-53.




20. Chapman RH, Benner JS, Petrilla AA, Tierce JC, Collins SR, et al. (2005) Predictors of adherence with antihypertensive and lipid-lowering therapy. Arch Intern Med 165: 1147-1152.

21. Matthes J and Albus C. (2014) Improving adherence with medication: a selective literature review based on the example of hypertension treatment. Dtsch Arztebl Int 111: 41-47.

22. Sokol MC, McGuigan KA, Verbrugge RR and Epstein RS. (2005) Impact of medication adherence on hospitalization risk and healthcare cost. Med Care 43: 521-530.

Chapter 7

23. Ettehad D, Emdin CA, Kiran A, Anderson SG, Callender T, et al. (2016) Blood pressure lowering for prevention of cardiovascular disease and death: a systematic review and meta-analysis. Lancet 387: 957-967.

24. Krousel-Wood M, Thomas S, Muntner P and Morisky D. (2004) Medication adherence: a key factor in achieving blood pressure control and good clinical outcomes in hypertensive patients. Curr Opin Cardiol 19: 357-362.


26. Breekveldt-Postma NS, Penning-van Beest FJ, Siiskonen SJ, Falvey H, Vincze G, et al. (2008) The effect of discontinuation of antihypertensives on the risk of acute myocardial infarction and stroke. Curr Med Res Opin 24: 121-127.

27. Vemer P, Corro Ramos I, van Voorn GA, Al MJ and Feenstra TL. (2016) AdViSHE: A Validation-Assessment Tool of Health-Economic Models for Decision Makers and Model Users. Pharmacoeconomics 34: 349-361.

28. van Dulmen S, Sluijs E, van Dijk L, de Ridder D, Heerdink R, et al. (2007) Patient adherence to medical treatment: a review of reviews. BMC Health Serv Res 7: 55.

Chap

ter 7

177

General discussion

16. Stevanovic J, Postma MJ and Pechlivanoglou P. (2012) A systematic review on the application of cardiovascular risk prediction models in pharmacoeconomics, with a focus on primary prevention. Eur J Prev Cardiol 19: 42-53.




20. Chapman RH, Benner JS, Petrilla AA, Tierce JC, Collins SR, et al. (2005) Predictors of adherence with antihypertensive and lipid-lowering therapy. Arch Intern Med 165: 1147-1152.

21. Matthes J and Albus C. (2014) Improving adherence with medication: a selective literature review based on the example of hypertension treatment. Dtsch Arztebl Int 111: 41-47.

22. Sokol MC, McGuigan KA, Verbrugge RR and Epstein RS. (2005) Impact of medication adherence on hospitalization risk and healthcare cost. Med Care 43: 521-530.

Chapter 7

23. Ettehad D, Emdin CA, Kiran A, Anderson SG, Callender T, et al. (2016) Blood pressure lowering for prevention of cardiovascular disease and death: a systematic review and meta-analysis. Lancet 387: 957-967.

24. Krousel-Wood M, Thomas S, Muntner P and Morisky D. (2004) Medication adherence: a key factor in achieving blood pressure control and good clinical outcomes in hypertensive patients. Curr Opin Cardiol 19: 357-362.


26. Breekveldt-Postma NS, Penning-van Beest FJ, Siiskonen SJ, Falvey H, Vincze G, et al. (2008) The effect of discontinuation of antihypertensives on the risk of acute myocardial infarction and stroke. Curr Med Res Opin 24: 121-127.

27. Vemer P, Corro Ramos I, van Voorn GA, Al MJ and Feenstra TL. (2016) AdViSHE: A Validation-Assessment Tool of Health-Economic Models for Decision Makers and Model Users. Pharmacoeconomics 34: 349-361.

28. van Dulmen S, Sluijs E, van Dijk L, de Ridder D, Heerdink R, et al. (2007) Patient adherence to medical treatment: a review of reviews. BMC Health Serv Res 7: 55.

General discussion

Appendix

180

Appendix

Appendix 1:

In-depth interview guideline: Adherence of hypertensive patients at CHSs (chapter 3)

Introduction

Please let me introduce myself. I am a researcher at TNUMP. As far as I know you are hypertensive patient. We would like to have your ideas on hypertension and using hypertensive drugs. It will take around 30 minutes of your time. You have the right to withdraw at any time during our talk.

Administration

Name, sex, age, address, marital status, living address, job?

Hypertension and health service use

When did you start to recognize that you are a hypertensive patient? How serious was it at that time?

Why do you choose this kind of heath service? Why did you do not choose the others? (Distance, convenience, good health service, good drugs….)? Is it easy or difficult for you to access health care services? What makes it easy/difficult?

How often do you visit health service for hypertension? Why do you go with this frequency?

If a patient does not visit the health services frequently for hypertension management, ask, Why don’t you visit the health service more often to manage hypertension?

Where do you get hypertension drugs? If a patient goes to private pharmacy stores to buy drugs, ask why he/she didn’t get

181

Appendix

Appendix 1:

In-depth interview guideline: Adherence of hypertensive patients at CHSs (chapter 3)

Introduction

Please let me introduce myself. I am a researcher at TNUMP. As far as I know you are hypertensive patient. We would like to have your ideas on hypertension and using hypertensive drugs. It will take around 30 minutes of your time. You have the right to withdraw at any time during our talk.

Administration

Name, sex, age, address, marital status, living address, job?

Hypertension and health service use

When did you start to recognize that you are a hypertensive patient? How serious was it at that time?

Why do you choose this kind of heath service? Why did you do not choose the others? (Distance, convenience, good health service, good drugs….)? Is it easy or difficult for you to access health care services? What makes it easy/difficult?

How often do you visit health service for hypertension? Why do you go with this frequency?

If a patient does not visit the health services frequently for hypertension management, ask, Why don’t you visit the health service more often to manage hypertension?

Where do you get hypertension drugs? If a patient goes to private pharmacy stores to buy drugs, ask why he/she didn’t get

182

Appendix

drugs from the health service? Who gives you advice on drug use or therapy?

How much do you have to pay per month to buy drugs? Is this expensive? Can you afford to pay? Do you consider it a lot of money for hypertension treatment?

Do you get doctor’s advice on hypertensive drugs use? Are you satisfied with the advice? What kind of information do you still needin terms of therapy/drug use and hypertension, including adherence?

Hypertensive drugs use

How many pills per day do you generally take? How many times per day do you generally take pills? Count with them and investigate whether they forget some pills (list answers). Explain how you use drugs (time of day, before/after food, etc…)?

Do you follow doctor’s prescription or do you use drugs by yourself?

Do you forget to take drugs? Do you stop to take drugs on some days? Do you sometimes forget to take drugs in the morning or afternoon? How often does this happen? (ask this with all questions).

Did the drug therapy change, such as a change in the number of pills per day or name of drugs? Did you change by yourself or did the doctor change the therapy? Why did you have to change drugs or therapy?

What helps you to remember to take drugs every day? Is there anyone who helps you to remember to take the drugs frequently?

Does the therapy adequately control your blood pressure level? What kinds of drug do you like/don’t like and why?

Appendix

Do you get any side effects (problems) from taking drugs? If yes, what are they? How do you react when side effects occur (stop using drugs or go to see doctor or change drug by yourself…)?

Do you know any herbal medicines to treat hypertension? Do you use them? Why do you use them? Is this helpful? Do you prefer herbal medicine or Western medicine?

Do you encounter any difficulties in taking drugs? What helps you using drug easily? What skills do you need to take drugs as often as according to the doctor’s prescription?

Do you know of problem that may occur if you do not take drugs or do not take drugs frequently? What are they?

When you stop or forget to take drugs, is there any problem resulting for you?

Sometimes you forgot to take the drug or you did not take drug actively, why was that? What factors cause you to be adherent or non-adherent with drug therapy?

What factors can help you to improve adherence with the therapy?

What do you expect from health services to support you to improve adherence with the therapy?

Do you have any questions?

Thank you very much for your time to answer our questions!

183

Appendix

Do you get any side effects (problems) from taking drugs? If yes, what are they? How do you react when side effects occur (stop using drugs or go to see doctor or change drug by yourself…)?

Do you know any herbal medicines to treat hypertension? Do you use them? Why do you use them? Is this helpful? Do you prefer herbal medicine or Western medicine?

Do you encounter any difficulties in taking drugs? What helps you using drug easily? What skills do you need to take drugs as often as according to the doctor’s prescription?

Do you know of problem that may occur if you do not take drugs or do not take drugs frequently? What are they?

When you stop or forget to take drugs, is there any problem resulting for you?

Sometimes you forgot to take the drug or you did not take drug actively, why was that? What factors cause you to be adherent or non-adherent with drug therapy?

What factors can help you to improve adherence with the therapy?

What do you expect from health services to support you to improve adherence with the therapy?

Do you have any questions?

Thank you very much for your time to answer our questions!

184

Appendix

Appendix 2: Description of screening strategies (Chapter 6)

Ten-years horizon Screening scenario Description

Male, 35 years old One-off screening Cohort of 35 years old. Screening once in the first year. Follow up until they are 45 years old or died.

Annual screening Cohort of 35 years old. Start screening at age 35 years old and repeat every year until 45 years old or died.

Biannual screening Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 45 years old or died.

Annual screening and increase coverage of treatment by 20%

Cohort of 35 years old. Start screening at age 35 years old and repeat every year until 45 years old or die. And increase coverage of treatment by 20%.

Biannual screening and increase coverage of treatment by 20%

Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 45 years old or died. And increase coverage of treatment by 20%.





Cohort of 45 years old. Start screening at age 45 years old and repeat every year until 55 years old or died. And increase coverage of treatment by 20%.

Appendix


Cohort of 45 years old. Start screening at age45 years old and repeat every two years until 55 years old or died. And increase coverage of treatment by 20%.

Male, 55 years old One-off screening Cohort of 55 years old. Screening once in the first year. Follow up until they are 65 years or died.




Cohort of 55 years old. Start screening at age 55 years old and repeat every year until 65 years old or died. And increase coverage of treatment by 20%


Cohort of 55 years old. Start screening at age55 years old and repeat every two years until 65 years old or died. And increase coverage of treatment by 20%

Female, 35 years old

One-off screening Cohort of 35 years old. Screening once in the first year. Follow up until they are 45 years or died.





185

Appendix

Appendix 2: Description of screening strategies (Chapter 6)

Ten-years horizon Screening scenario Description





Cohort of 35 years old. Start screening at age 35 years old and repeat every year until 45 years old or die. And increase coverage of treatment by 20%.


Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 45 years old or died. And increase coverage of treatment by 20%.






Appendix



Male, 55 years old One-off screening Cohort of 55 years old. Screening once in the first year. Follow up until they are 65 years or died.






Cohort of 55 years old. Start screening at age55 years old and repeat every two years until 65 years old or died. And increase coverage of treatment by 20%







186

Appendix


Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 45 years old or died. And increase coverage of treatment by 20%















Appendix



Lifetime horizon Screening scenario DescriptionFemale, 35 years old

Annual screening Cohort of 35 years old. Start screening at age 35 years old and repeat every year until 82 years old or died (47 years follow up).

Biannual screening Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 82 years old or died (47 years follow up).


Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 55 years old or died. Then repeat screening every year, from 56 to 82 years old or die (47 years follow up).


Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 60 years old or died. Then, repeat screening every year, from 61 to 82 years old or died (47 years follow up).


Cohort of 35 years old. Start screening at age 35 years old and repeat every year until 82 years old or died (47 years follow up). And increase coverage of treatment by 20%.


Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 82 years old or died (47 years follow up). And increase coverage of treatment by 20%.

187

Appendix

















Appendix



Lifetime horizon Screening scenario DescriptionFemale, 35 years old




Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 55 years old or died. Then repeat screening every year, from 56 to 82 years old or die (47 years follow up).







188

Appendix













Annual screening Cohort of 55 years old. Screening in every year, starting from 55 to 84 years old or died (29 years follow up).

Appendix

Biannual screening Cohort of 55 years old. Screening in every two years, starting from 55 to 84 years old or died (29 years follow up).


Cohort of 55 years old. Screening in every year, starting from 55 to 84 years old or died (29 years follow up). And increase coverage of treatment by 20% .


Cohort of 55 years old. Screening in every two years, starting from 55 to 84 years old or died (29 years follow up). And increase coverage of treatment by 20%.

Male, 35 years old Annual screening Cohort of 35 years old. Start screening at age 35 years old and repeat every year until 74 years old or died (39 years follow up).



Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 55 years old or died. Then repeat screening every years, from 56 to 74 years old or died (39 years follow up).



189

Appendix













Annual screening Cohort of 55 years old. Screening in every year, starting from 55 to 84 years old or died (29 years follow up).

Appendix

Biannual screening Cohort of 55 years old. Screening in every two years, starting from 55 to 84 years old or died (29 years follow up).


Cohort of 55 years old. Screening in every year, starting from 55 to 84 years old or died (29 years follow up). And increase coverage of treatment by 20% .


Cohort of 55 years old. Screening in every two years, starting from 55 to 84 years old or died (29 years follow up). And increase coverage of treatment by 20%.




Cohort of 35 years old. Start screening at age 35 years old and repeat every two years until 55 years old or died. Then repeat screening every years, from 56 to 74 years old or died (39 years follow up).



190

Appendix






Biannual screening Cohort of 45 years old. Start screening at age 45 years old and repeat every two years until 76 years old or died (31 years follow up)


Cohort of 45 years old. Start screening at age 45 years old and repeat every two years until 55 years old. Then, repeat screening every year, from 56 to 76 years old or died (31 years follow up)





Appendix



Male, 55 years old Annual screening Cohort of 55 years old. Screening in every year, starting from 55 to 77 years old or died (22 years follow up)

Biannual screening Cohort of 55 years old. Screening in every two years, starting from 55 to 77 years old or died (22 years follow up)


Cohort of 55 years old.Start screening at age 55 years old and repeat every year until 77 years old or died (22 years follow up). And increase coverage of treatment by 20%.



191

Appendix






Biannual screening Cohort of 45 years old. Start screening at age 45 years old and repeat every two years until 76 years old or died (31 years follow up)







Appendix



Male, 55 years old Annual screening Cohort of 55 years old. Screening in every year, starting from 55 to 77 years old or died (22 years follow up)

Biannual screening Cohort of 55 years old. Screening in every two years, starting from 55 to 77 years old or died (22 years follow up)


Cohort of 55 years old.Start screening at age 55 years old and repeat every year until 77 years old or died (22 years follow up). And increase coverage of treatment by 20%.



192

App

endi

x 3:

Res

ults

of u

niva

riat

e se

nsiti

vity

ana

lysi

s in

the

10-y

ear

time-

hori

zon

mod

el(C

hapt

er 6

)

Base

cas

e(+

) 25%

sc

reen

ing

cost

(-) 2

5%

scre

enin

g co

st

(+) 2

5%

hype

rtens

ion

treat

men

t co

st

(-) 2

5%

hype

rtens

ion

treat

men

t co

st

(+) 2

5%

CV

D

cost

(-) 2

5%

CV

D

cost

(-) 1

0%

CV

D

utili

ty

(-) 2

0%

CV

D

utili

ty

(+) 2

5%

trans

ition

prob

abilit

y fro

m

hype

rtens

ionto

CVD

(-) 25

%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion

to C

VD

1%

utili

ty

disc

ount

3%

utili

ty

disc

ount

RR

redu

ctio

n of

CV

D

depe

nd

on a

ge

App

lyin

g ut

ilitie

s fro

m

Glo

bal

Bur

den

Dise

ase

hype

rtens

ion

prev

alen

ce

by n

atio

nal

surv

ey

Star

t scr

eeni

ng a

t age

of 3

5 ye

ars,

fem

ales

One

-off

127,

715

160,

645

94,7

8412

7,73

412

7,69

612

6,69

412

8,73

590

,409

69,9

7010

1,58

917

1,25

413

7,06

015

7,25

011

0,55

869

,737

43,3

62

E175

8,69

594

9,47

056

7,92

075

8,71

375

8,67

675

7,57

575

9,81

553

8,26

541

7,08

660

7,25

61,

011,

093

816,

576

942,

381

658,

539

415,

909

306,

054

E238

6,85

148

8,75

328

4,94

938

2,78

339

0,92

038

5,73

038

7,97

227

4,46

321

2,67

730

9,21

951

6,23

641

6,38

448

0,57

633

5,61

321

2,07

815

0,70

7

E1 &

T.20

%57

2,67

971

6,95

042

8,40

857

2,69

857

2,66

157

1,56

057

3,79

940

6,29

431

4,82

545

8,15

476

3,55

461

6,37

071

1,33

049

6,99

431

3,93

723

0,41

8

E2 &

T.20

%29

1,47

636

8,53

921

4,41

428

8,40

429

4,54

929

0,35

629

2,59

720

6,79

716

0,24

323

2,76

738

9,32

431

3,72

836

2,09

425

2,78

415

9,79

211

2,93

9

Star

t scr

eeni

ng a

t age

of 4

5 yea

rs, f

emal

esO

ne-o

ff12

,070

16,1

038,

038

12,0

8912

,051

11,0

3613

,104

8,54

36,

611

8,88

617

,369

12,9

4814

,843

9,31

36,

590

5,01

8

E110

5,52

513

2,94

978

,102

105,

544

105,

506

104,

464

106,

586

74,7

3557

,854

83,9

3314

1,50

611

3,29

013

0,08

185

,054

57,6

7853

,532

E251

,335

65,7

6836

,902

50,7

9651

,874

50,2

7552

,395

36,3

5628

,144

40,4

0769

,542

55,1

1263

,278

41,1

0128

,058

25,1

50

E1 &

T.20

%78

,786

99,5

2558

,048

78,8

0578

,767

77,7

2579

,847

55,7

9843

,194

62,4

5510

5,99

984

,584

97,1

1963

,368

43,0

6339

,476

E2 &

T.20

%37

,806

48,7

2126

,891

37,4

03`3

8,20

936

,746

38,8

6626

,775

20,7

2729

,539

51,5

7840

,587

46,6

0230

,129

20,6

6418

,013

Star

t scr

eeni

ng a

t age

of 5

5 yea

rs, f

emal

esO

ne-o

ff87

12,

139

Dom

inan

t89

085

2D

omina

nt1,

940

616

477

Dom

inan

t2,

530

933

1,06

746

647

6D

omin

ant

E125

,471

32,9

0118

,040

25,4

9025

,452

24,3

8926

,552

18,0

3213

,956

19,6

9535

,082

27,2

9931

,240

21,4

0713

,917

18,4

53

E211

,189

15,1

907,

187

11,0

6511

,312

10,1

0712

,270

7,92

16,

131

8,17

816

,191

11,9

9213

,723

9,25

46,

113

7,63

9

E1 &

T.20

%18

,226

23,8

4512

,607

18,2

4518

,207

17,1

4419

,307

12,9

039,

987

13,8

5225

,500

19,5

3422

,354

15,2

429,

959

12,9

21

E2 &

T.20

%7,

425

10,4

524,

399

7,33

77,

514

6,34

48,

507

5,25

74,

069

5,14

311

,214

7,95

89,

107

6,05

24,

057

4,74

3

Base

cas

e(+

) 25%

sc

reen

ing

cost

(-) 2

5%

scre

enin

g co

st

(+) 2

5%

hype

rtens

ion

treat

men

t co

st

(-) 2

5%

hype

rtens

ion

treat

men

t co

st

(+) 2

5%

CV

D

cost

(-) 2

5%

CV

D

cost

(-) 1

0%

CV

D

utili

ty

(-) 2

0%

CV

D

utili

ty

(+) 2

5%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion

to C

VD

(-) 25

%

trans

ition

prob

abilit

yfro

m

hype

rtens

ion

to C

VD

1%

utili

ty

disc

ount

3%

utili

ty

disc

ount

RR

redu

ctio

n of

CV

D

depe

nd

on a

ge

App

lyin

g ut

ilitie

s fro

m

Glo

bal

Bur

den

Dise

ase

hype

rtens

ion

prev

alen

ce

by n

atio

nal

surv

ey

Star

t scr

eeni

ng a

t age

of 3

5 yea

rs, m

ales

One

-off

29,4

3337

,655

21,2

1129

,449

29,4

1728

,554

30,3

1221

,752

24,8

8422

,942

40,2

4631

,570

36,1

8325

,427

18,9

4810

,796

E115

8,14

719

8,64

011

7,65

415

8,16

315

8,13

115

7,17

515

9,11

911

7,10

314

8,33

112

6,16

721

1,44

317

0,18

919

6,35

613

7,83

710

2,16

274

,341

E273

,227

95,4

3051

,023

70,3

0476

,149

72,2

5274

,201

54,2

2575

,592

58,0

1898

,570

78,8

0990

,941

63,6

5847

,307

32,6

02

E1 &

T.20

%11

0,60

213

9,20

981

,995

110,

618

110,

586

109,

630

111,

574

81,8

9811

1,96

388

,007

148,

257

119,

024

137,

324

96,3

0671

,448

51,4

46

E2 &

T.20

%50

,607

66,2

9234

,921

48,5

4752

,667

49,6

3351

,581

37,4

7456

,955

39,8

6068

,513

54,4

6462

,849

43,9

0032

,694

21,9

59

Star

t scr

eeni

ng a

t age

of 4

5 yea

rs, m

ales

One

-off

4,18

36,

119

2,24

74,

199

4,16

73,

277

5,08

93,

091

2,35

12,

661

6,70

94,

482

5,12

62,

991

2,69

31,

618

E137

,580

47,9

0927

,252

37,5

9637

,564

36,6

3138

,530

27,7

9820

,575

29,5

2350

,999

40,3

3146

,276

30,1

9624

,238

22,8

57

E216

,035

21,6

9410

,377

15,3

3416

,736

15,0

8616

,984

11,8

6110

,003

12,1

7522

,458

17,2

0919

,745

12,6

0510

,342

9,07

9

E1 &

T.20

%25

,453

32,7

5018

,157

25,4

6925

,437

24,5

0426

,403

18,8

2815

,362

19,7

5634

,939

27,3

1731

,343

20,2

9716

,417

15,0

68

E2 &

T.20

%10

,233

14,2

306,

235

9,74

210

,723

9,28

411

,182

7,56

97,

367

7,50

014

,775

10,9

8212

,600

7,87

06,

600

5,33

5

Star

t scr

eeni

ng a

t age

of 5

5 yea

rs, m

ales

One

-off

Dom

inant

Dom

inan

tD

omin

ant

Dom

inan

tD

omin

ant

Dom

inant

Dom

inant

Dom

inant

170

Dom

inan

tD

omin

ant

Dom

inant

Dom

inant

Dom

inan

tD

omin

ant

Dom

inan

t

E17,

638

10,5

134,

763

7,65

47,

622

6,65

78,

619

5,64

74,

964

5,42

611

,302

8,17

09,

311

6,25

44,

925

5,92

8

E22,

076

3,72

442

81,

928

2,22

41,

095

3,05

71,

535

2,17

991

83,

983

2,22

12,

531

1,51

01,

339

1,25

3

E1 &

T.20

%4,

262

6,29

32,

232

4,27

94,

246

3,28

15,

244

3,15

23,

552

2,68

96,

862

4,55

95,

196

3,37

42,

748

3,05

7

E2 &

T.20

%33

41,

498

Dom

inan

t23

443

3D

omina

nt1,

315

247

1,44

6D

omin

ant

1,69

135

740

723

215

Dom

inan

t

*Not

e: O

ne-o

ff: s

cree

ning

onc

e in

the

first

yea

r, E1

: Ann

ual s

cree

ning

, E2:

Bia

nnua

l scr

eeni

ng, E

1&T.

20%

: Ann

ual s

cree

ning

com

bine

d w

ithin

crea

sing

cov

erag

e of

tre

atm

ent

by 2

0%,

E2&

T.20

%:

Bia

nnua

l sc

reen

ing

com

bine

d w

ith i

ncre

asin

g co

vera

ge o

f tre

atm

ent

by20

%,

CV

D:

card

iova

scus

lar d

isea

se.

193

App

endi

x 3:

Res

ults

of u

niva

riat

e se

nsiti

vity

ana

lysi

s in

the

10-y

ear

time-

hori

zon

mod

el(C

hapt

er 6

)

Base

cas

e(+

) 25%

sc

reen

ing

cost

(-) 2

5%

scre

enin

g co

st

(+) 2

5%

hype

rtens

ion

treat

men

t co

st

(-) 2

5%

hype

rtens

ion

treat

men

t co

st

(+) 2

5%

CV

D

cost

(-) 2

5%

CV

D

cost

(-) 1

0%

CV

D

utili

ty

(-) 2

0%

CV

D

utili

ty

(+) 2

5%

trans

ition

prob

abilit

y fro

m

hype

rtens

ionto

CVD

(-) 25

%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion

to C

VD

1%

utili

ty

disc

ount

3%

utili

ty

disc

ount

RR

redu

ctio

n of

CV

D

depe

nd

on a

ge

App

lyin

g ut

ilitie

s fro

m

Glo

bal

Bur

den

Dise

ase

hype

rtens

ion

prev

alen

ce

by n

atio

nal

surv

ey

Star

t scr

eeni

ng a

t age

of 3

5 ye

ars,

fem

ales

One

-off

127,

715

160,

645

94,7

8412

7,73

412

7,69

612

6,69

412

8,73

590

,409

69,9

7010

1,58

917

1,25

413

7,06

015

7,25

011

0,55

869

,737

43,3

62

E175

8,69

594

9,47

056

7,92

075

8,71

375

8,67

675

7,57

575

9,81

553

8,26

541

7,08

660

7,25

61,

011,

093

816,

576

942,

381

658,

539

415,

909

306,

054

E238

6,85

148

8,75

328

4,94

938

2,78

339

0,92

038

5,73

038

7,97

227

4,46

321

2,67

730

9,21

951

6,23

641

6,38

448

0,57

633

5,61

321

2,07

815

0,70

7

E1 &

T.20

%57

2,67

971

6,95

042

8,40

857

2,69

857

2,66

157

1,56

057

3,79

940

6,29

431

4,82

545

8,15

476

3,55

461

6,37

071

1,33

049

6,99

431

3,93

723

0,41

8

E2 &

T.20

%29

1,47

636

8,53

921

4,41

428

8,40

429

4,54

929

0,35

629

2,59

720

6,79

716

0,24

323

2,76

738

9,32

431

3,72

836

2,09

425

2,78

415

9,79

211

2,93

9

Star

t scr

eeni

ng a

t age

of 4

5 yea

rs, f

emal

esO

ne-o

ff12

,070

16,1

038,

038

12,0

8912

,051

11,0

3613

,104

8,54

36,

611

8,88

617

,369

12,9

4814

,843

9,31

36,

590

5,01

8

E110

5,52

513

2,94

978

,102

105,

544

105,

506

104,

464

106,

586

74,7

3557

,854

83,9

3314

1,50

611

3,29

013

0,08

185

,054

57,6

7853

,532

E251

,335

65,7

6836

,902

50,7

9651

,874

50,2

7552

,395

36,3

5628

,144

40,4

0769

,542

55,1

1263

,278

41,1

0128

,058

25,1

50

E1 &

T.20

%78

,786

99,5

2558

,048

78,8

0578

,767

77,7

2579

,847

55,7

9843

,194

62,4

5510

5,99

984

,584

97,1

1963

,368

43,0

6339

,476

E2 &

T.20

%37

,806

48,7

2126

,891

37,4

03`3

8,20

936

,746

38,8

6626

,775

20,7

2729

,539

51,5

7840

,587

46,6

0230

,129

20,6

6418

,013

Star

t scr

eeni

ng a

t age

of 5

5 yea

rs, f

emal

esO

ne-o

ff87

12,

139

Dom

inan

t89

085

2D

omina

nt1,

940

616

477

Dom

inan

t2,

530

933

1,06

746

647

6D

omin

ant

E125

,471

32,9

0118

,040

25,4

9025

,452

24,3

8926

,552

18,0

3213

,956

19,6

9535

,082

27,2

9931

,240

21,4

0713

,917

18,4

53

E211

,189

15,1

907,

187

11,0

6511

,312

10,1

0712

,270

7,92

16,

131

8,17

816

,191

11,9

9213

,723

9,25

46,

113

7,63

9

E1 &

T.20

%18

,226

23,8

4512

,607

18,2

4518

,207

17,1

4419

,307

12,9

039,

987

13,8

5225

,500

19,5

3422

,354

15,2

429,

959

12,9

21

E2 &

T.20

%7,

425

10,4

524,

399

7,33

77,

514

6,34

48,

507

5,25

74,

069

5,14

311

,214

7,95

89,

107

6,05

24,

057

4,74

3

Base

cas

e(+

) 25%

sc

reen

ing

cost

(-) 2

5%

scre

enin

g co

st

(+) 2

5%

hype

rtens

ion

treat

men

t co

st

(-) 2

5%

hype

rtens

ion

treat

men

t co

st

(+) 2

5%

CV

D

cost

(-) 2

5%

CV

D

cost

(-) 1

0%

CV

D

utili

ty

(-) 2

0%

CV

D

utili

ty

(+) 2

5%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion

to C

VD

(-) 25

%

trans

ition

prob

abilit

yfro

m

hype

rtens

ion

to C

VD

1%

utili

ty

disc

ount

3%

utili

ty

disc

ount

RR

redu

ctio

n of

CV

D

depe

nd

on a

ge

App

lyin

g ut

ilitie

s fro

m

Glo

bal

Bur

den

Dise

ase

hype

rtens

ion

prev

alen

ce

by n

atio

nal

surv

ey

Star

t scr

eeni

ng a

t age

of 3

5 yea

rs, m

ales

One

-off

29,4

3337

,655

21,2

1129

,449

29,4

1728

,554

30,3

1221

,752

24,8

8422

,942

40,2

4631

,570

36,1

8325

,427

18,9

4810

,796

E115

8,14

719

8,64

011

7,65

415

8,16

315

8,13

115

7,17

515

9,11

911

7,10

314

8,33

112

6,16

721

1,44

317

0,18

919

6,35

613

7,83

710

2,16

274

,341

E273

,227

95,4

3051

,023

70,3

0476

,149

72,2

5274

,201

54,2

2575

,592

58,0

1898

,570

78,8

0990

,941

63,6

5847

,307

32,6

02

E1 &

T.20

%11

0,60

213

9,20

981

,995

110,

618

110,

586

109,

630

111,

574

81,8

9811

1,96

388

,007

148,

257

119,

024

137,

324

96,3

0671

,448

51,4

46

E2 &

T.20

%50

,607

66,2

9234

,921

48,5

4752

,667

49,6

3351

,581

37,4

7456

,955

39,8

6068

,513

54,4

6462

,849

43,9

0032

,694

21,9

59

Star

t scr

eeni

ng a

t age

of 4

5 yea

rs, m

ales

One

-off

4,18

36,

119

2,24

74,

199

4,16

73,

277

5,08

93,

091

2,35

12,

661

6,70

94,

482

5,12

62,

991

2,69

31,

618

E137

,580

47,9

0927

,252

37,5

9637

,564

36,6

3138

,530

27,7

9820

,575

29,5

2350

,999

40,3

3146

,276

30,1

9624

,238

22,8

57

E216

,035

21,6

9410

,377

15,3

3416

,736

15,0

8616

,984

11,8

6110

,003

12,1

7522

,458

17,2

0919

,745

12,6

0510

,342

9,07

9

E1 &

T.20

%25

,453

32,7

5018

,157

25,4

6925

,437

24,5

0426

,403

18,8

2815

,362

19,7

5634

,939

27,3

1731

,343

20,2

9716

,417

15,0

68

E2 &

T.20

%10

,233

14,2

306,

235

9,74

210

,723

9,28

411

,182

7,56

97,

367

7,50

014

,775

10,9

8212

,600

7,87

06,

600

5,33

5

Star

t scr

eeni

ng a

t age

of 5

5 yea

rs, m

ales

One

-off

Dom

inant

Dom

inan

tD

omin

ant

Dom

inan

tD

omin

ant

Dom

inant

Dom

inant

Dom

inant

170

Dom

inan

tD

omin

ant

Dom

inant

Dom

inant

Dom

inan

tD

omin

ant

Dom

inan

t

E17,

638

10,5

134,

763

7,65

47,

622

6,65

78,

619

5,64

74,

964

5,42

611

,302

8,17

09,

311

6,25

44,

925

5,92

8

E22,

076

3,72

442

81,

928

2,22

41,

095

3,05

71,

535

2,17

991

83,

983

2,22

12,

531

1,51

01,

339

1,25

3

E1 &

T.20

%4,

262

6,29

32,

232

4,27

94,

246

3,28

15,

244

3,15

23,

552

2,68

96,

862

4,55

95,

196

3,37

42,

748

3,05

7

E2 &

T.20

%33

41,

498

Dom

inan

t23

443

3D

omina

nt1,

315

247

1,44

6D

omin

ant

1,69

135

740

723

215

Dom

inan

t

*Not

e: O

ne-o

ff: s

cree

ning

onc

e in

the

first

yea

r, E1

: Ann

ual s

cree

ning

, E2:

Bia

nnua

l scr

eeni

ng, E

1&T.

20%

: Ann

ual s

cree

ning

com

bine

d w

ithin

crea

sing

cov

erag

e of

tre

atm

ent

by 2

0%,

E2&

T.20

%:

Bia

nnua

l sc

reen

ing

com

bine

d w

ith i

ncre

asin

g co

vera

ge o

f tre

atm

ent

by20

%,

CV

D:

card

iova

scus

lar d

isea

se.

194

App

endi

x 4:

Res

ults

of u

niva

riat

e se

nsiti

vity

ana

lysi

s in

the

lifet

ime

time-

hori

zon

mod

el(C

hapt

er 6

)

Base

case

(+) 2

5%

scree

ning

cost

(-) 25

%

scree

ning

cost

(+) 2

5%

hype

rtens

ion t

reatm

ent

cost

(-) 25

%

hype

rtens

iont

reatm

ent

cost

(+) 2

5%

CVD

treatm

ent

cost

(-) 25

%

CVD

treatm

ent

cost

(-) 10

%

CVD

utilit

y

(-) 20

%

CVD

utilit

y

(+) 2

5%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion to

CV

D

(-) 25

%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion to

CV

D

1% ut

ility

disco

unt

3%

utilit

y dis

coun

t

RR re

ducti

on

of C

VD

de

pend

on

age

Appl

ying

ut

ilities

from

G

lobal

Burd

en

Dise

ase

Hyp

erten

sion

prev

alenc

e by

natio

nal s

urve

y

Star

t scr

eeni

ng at

age o

f 35 y

ears

, fem

ales

E127

,944

35,22

020

,669

27,95

427

,935

27,64

628

,243

19,51

014

,987

22,89

936

,366

38,72

871

,742

28,58

714

,938

14,63

9

E213

,470

17,27

79,6

6413

,330

13,61

013

,171

13,76

99,4

057,2

2510

,924

17,71

218

,678

34,63

913

,849

7,201

6,654

E2 un

til 55

+ E1

17,28

022

,000

12,56

017

,179

17,38

116

,981

17,57

912

,065

9,268

14,07

522

,623

23,96

744

,478

17,74

99,2

388,7

59E2

until

60+

E116

,115

20,55

611

,675

16,00

316

,228

15,81

616

,414

11,25

28,6

4413

,111

21,12

222

,353

41,48

316

,571

8,616

8,116

E1&

T.20

%20

,850

26,35

215

,348

20,86

020

,841

20,55

221

,149

14,55

811

,183

17,02

827

,226

28,89

653

,529

21,35

311

,146

10,79

2

E2&

T.20

%9,9

0412

,783

7,026

9,801

10,00

89,6

0510

,203

6,915

5,312

7,971

13,11

913

,733

25,46

910

,207

5,295

4,754

Star

t scr

eeni

ng at

age o

f 45 y

ears

, fem

ales

E113

,331

17,04

99,6

1213

,343

13,31

912

,933

13,72

89,3

057,1

4710

,748

17,62

416

,916

26,48

013

,855

7,125

6,847

E25,9

607,9

104,0

115,8

986,0

225,5

636,3

584,1

613,1

964,6

398,1

447,5

6411

,842

6,271

3,186

2,682

E2 un

til 55

+ E1

7,253

9,514

4,993

7,204

7,303

6,855

7,651

5,063

3,889

5,710

9,807

9,206

14,41

97,6

113,8

773,4

12E2

until

60+

E16,7

158,8

464,5

846,6

606,7

706,3

177,1

134,6

873,6

005,2

649,1

158,5

2313

,347

7,056

3,589

3,108

E1&

T.20

%9,7

0512

,517

6,893

9,717

9,693

9,308

10,10

36,7

755,2

037,7

4212

,962

12,31

619

,278

10,12

35,1

874,8

06

E2&

T.20

%4,1

325,6

062,6

574,0

884,1

753,7

344,5

292,8

842,2

153,1

225,7

935,2

438,2

094,3

872,2

081,6

55

Star

t scr

eeni

ng at

age o

f 55 y

ears

, fem

ales

E16,5

068,6

704,3

426,5

216,4

925,9

547,0

584,5

483,4

964,9

789,0

237,7

0410

,606

7,991

3,486

4,518

E22,3

293,4

901,1

682,3

032,3

551,7

772,8

811,6

281,2

511,5

103,6

592,7

583,7

973,1

031,2

481,3

14E2

until

60+

E12,4

273,6

121,2

432,4

022,4

531,8

752,9

801,6

971,3

041,5

913,7

862,8

753,9

573,2

201,3

011,3

90

E1&

T.20

%4,3

966,0

322,7

594,4

104,3

813,8

434,9

483,0

732,3

623,2

256,3

145,2

057,1

665,5

212,3

552,8

94

E2&

T.20

%1,2

372,1

1535

91,2

211,2

5368

51,7

8986

566

560

32,2

571,4

652,0

161,8

2566

347

1

Base

case

(+) 2

5%

scree

ning

cost

(-) 25

%

scree

ning

cost

(+) 2

5%

hype

rtens

ion t

reatm

ent

cost

(-) 25

%

hype

rtens

ion t

reatm

ent

cost

(+) 2

5%

CVD

treatm

ent

cost

(-) 25

%

CVD

treatm

ent

cost

(-) 10

%

CVD

utilit

y

(-) 20

%

CVD

utilit

y

(+) 2

5%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion to

CV

D

(-) 25

%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion to

CV

D

1% ut

ility

disco

unt

3%

utilit

y dis

coun

t

RR re

ducti

on

of C

VD

de

pend

on

age

Appl

ying

ut

ilities

from

G

lobal

Burd

en

Dise

ase

Hyp

erten

sion

prev

alenc

e by

natio

nal s

urve

y

Star

t scr

eeni

ng at

age o

f 35 y

ears

, mal

es

E110

,834

13,86

57,8

0210

,843

10,82

410

,501

11,16

77,9

266,2

498,7

0314

,376

14,05

022

,974

9,468

6,873

6,237

E24,3

135,9

262,7

014,1

124,5

143,9

804,6

463,1

562,4

883,3

135,9

645,5

979,1

653,7

362,7

362,1

37

E2 un

til 55

+ E1

5,276

7,098

3,453

5,106

5,446

4,942

5,609

3,860

3,043

4,108

7,207

6,849

11,22

14,5

843,3

472,7

42

E2 un

til 60

+ E1

4,866

6,599

3,132

4,682

5,049

4,532

5,199

3,560

2,807

3,769

6,677

6,316

10,34

74,2

253,0

872,4

84

E1&

T.20

%7,2

749,4

165,1

327,2

837,2

646,9

417,6

075,3

224,1

965,7

5912

,737

9,434

15,42

56,3

374,6

144,0

32

E2&

T.20

%2,6

673,8

061,5

282,5

282,8

062,3

343,0

001,9

511,5

381,9

513,8

423,4

615,6

672,2

871,6

921,1

36

Star

t scr

eeni

ng at

age o

f 45 y

ears

, mal

es

E16,8

348,9

824,6

866,8

466,8

236,3

837,2

855,0

023,9

455,3

229,3

328,2

5211

,784

6,088

4,340

4,204

E22,3

033,4

641,1

422,1

692,4

371,8

522,7

551,6

861,3

291,5

653,5

062,7

813,9

742,0

181,4

631,0

67

E2 un

til 55

+ E1

2,631

3,865

1,398

2,508

2,755

2,180

3,083

1,926

1,519

1,836

3,929

3,178

4,542

2,315

1,671

1,294

E2 un

til 60

+ E1

2,434

3,624

1,244

2,304

2,563

1,982

2,885

1,781

1,405

1,673

3,674

2,939

4,200

2,137

1,546

1,157

E1&

T.20

%4,3

125,8

302,7

944,3

244,3

003,8

614,7

633,1

562,4

893,2

296,0

915,2

067,4

353,8

212,7

382,4

59

E2&

T.20

%1,1

111,9

3129

01,0

191,2

0265

91,5

6281

364

157

51,9

741,3

411,9

1694

670

524

2

Star

t scr

eeni

ng at

age o

f 55 y

ears

, mal

es

E12,6

643,9

541,3

742,6

782,6

502,0

263,3

021,9

541,5

421,7

194,1

973,0

203,8

352,6

671,6

981,9

25

E212

184

9D

omina

nt62

181

Dom

inant

759

8970

Dom

inant

931

137

174

173

77D

omina

nt

E2 un

til 60

+ E1

1,617

2,676

558

1,600

1,634

979

2,255

1,186

936

850

2,853

1,834

2,329

1,643

1,031

1,036

E1&

T.20

%1,1

492,0

6123

81,1

631,1

3651

21,7

8784

366

6D

omina

nt2,2

531,3

031,6

551,1

8173

362

9

E2&

T.20

%D

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntDo

mina

ntDo

mina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

nt

*Not

e: E

1: A

nnua

l scr

eeni

ng, E

2: B

iann

ual s

cree

ning

, E2

until

60+

E1:

Bia

nnua

l scr

eeni

ng u

ntil

60 y

ears

old

then

ann

ual s

cree

ning

unt

il di

ed,

E2 u

ntil

55+

E1: B

iann

ual s

cree

ning

unt

il 55

yea

rs o

ldth

en a

nnua

l scr

eeni

ng u

ntil

died

, E1&

T.20

%: A

nnua

l scr

eeni

ng c

ombi

ned

with

incr

easi

ng

cove

rage

of t

reat

men

t by

20%

, E2&

T.20

%: B

iann

ual s

cree

ning

com

bine

d w

ith in

crea

sing

cov

erag

e of

trea

tmen

t by

20%

, CV

D: C

ardi

ovas

cula

r di

seas

e.

195

App

endi

x 4:

Res

ults

of u

niva

riat

e se

nsiti

vity

ana

lysi

s in

the

lifet

ime

time-

hori

zon

mod

el(C

hapt

er 6

)

Base

case

(+) 2

5%

scree

ning

cost

(-) 25

%

scree

ning

cost

(+) 2

5%

hype

rtens

ion t

reatm

ent

cost

(-) 25

%

hype

rtens

iont

reatm

ent

cost

(+) 2

5%

CVD

treatm

ent

cost

(-) 25

%

CVD

treatm

ent

cost

(-) 10

%

CVD

utilit

y

(-) 20

%

CVD

utilit

y

(+) 2

5%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion to

CV

D

(-) 25

%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion to

CV

D

1% ut

ility

disco

unt

3%

utilit

y dis

coun

t

RR re

ducti

on

of C

VD

de

pend

on

age

Appl

ying

ut

ilities

from

G

lobal

Burd

en

Dise

ase

Hyp

erten

sion

prev

alenc

e by

natio

nal s

urve

y

Star

t scr

eeni

ng at

age o

f 35 y

ears

, fem

ales

E127

,944

35,22

020

,669

27,95

427

,935

27,64

628

,243

19,51

014

,987

22,89

936

,366

38,72

871

,742

28,58

714

,938

14,63

9

E213

,470

17,27

79,6

6413

,330

13,61

013

,171

13,76

99,4

057,2

2510

,924

17,71

218

,678

34,63

913

,849

7,201

6,654

E2 un

til 55

+ E1

17,28

022

,000

12,56

017

,179

17,38

116

,981

17,57

912

,065

9,268

14,07

522

,623

23,96

744

,478

17,74

99,2

388,7

59E2

until

60+

E116

,115

20,55

611

,675

16,00

316

,228

15,81

616

,414

11,25

28,6

4413

,111

21,12

222

,353

41,48

316

,571

8,616

8,116

E1&

T.20

%20

,850

26,35

215

,348

20,86

020

,841

20,55

221

,149

14,55

811

,183

17,02

827

,226

28,89

653

,529

21,35

311

,146

10,79

2

E2&

T.20

%9,9

0412

,783

7,026

9,801

10,00

89,6

0510

,203

6,915

5,312

7,971

13,11

913

,733

25,46

910

,207

5,295

4,754

Star

t scr

eeni

ng at

age o

f 45 y

ears

, fem

ales

E113

,331

17,04

99,6

1213

,343

13,31

912

,933

13,72

89,3

057,1

4710

,748

17,62

416

,916

26,48

013

,855

7,125

6,847

E25,9

607,9

104,0

115,8

986,0

225,5

636,3

584,1

613,1

964,6

398,1

447,5

6411

,842

6,271

3,186

2,682

E2 un

til 55

+ E1

7,253

9,514

4,993

7,204

7,303

6,855

7,651

5,063

3,889

5,710

9,807

9,206

14,41

97,6

113,8

773,4

12E2

until

60+

E16,7

158,8

464,5

846,6

606,7

706,3

177,1

134,6

873,6

005,2

649,1

158,5

2313

,347

7,056

3,589

3,108

E1&

T.20

%9,7

0512

,517

6,893

9,717

9,693

9,308

10,10

36,7

755,2

037,7

4212

,962

12,31

619

,278

10,12

35,1

874,8

06

E2&

T.20

%4,1

325,6

062,6

574,0

884,1

753,7

344,5

292,8

842,2

153,1

225,7

935,2

438,2

094,3

872,2

081,6

55

Star

t scr

eeni

ng at

age o

f 55 y

ears

, fem

ales

E16,5

068,6

704,3

426,5

216,4

925,9

547,0

584,5

483,4

964,9

789,0

237,7

0410

,606

7,991

3,486

4,518

E22,3

293,4

901,1

682,3

032,3

551,7

772,8

811,6

281,2

511,5

103,6

592,7

583,7

973,1

031,2

481,3

14E2

until

60+

E12,4

273,6

121,2

432,4

022,4

531,8

752,9

801,6

971,3

041,5

913,7

862,8

753,9

573,2

201,3

011,3

90

E1&

T.20

%4,3

966,0

322,7

594,4

104,3

813,8

434,9

483,0

732,3

623,2

256,3

145,2

057,1

665,5

212,3

552,8

94

E2&

T.20

%1,2

372,1

1535

91,2

211,2

5368

51,7

8986

566

560

32,2

571,4

652,0

161,8

2566

347

1

Base

case

(+) 2

5%

scree

ning

cost

(-) 25

%

scree

ning

cost

(+) 2

5%

hype

rtens

ion t

reatm

ent

cost

(-) 25

%

hype

rtens

ion t

reatm

ent

cost

(+) 2

5%

CVD

treatm

ent

cost

(-) 25

%

CVD

treatm

ent

cost

(-) 10

%

CVD

utilit

y

(-) 20

%

CVD

utilit

y

(+) 2

5%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion to

CV

D

(-) 25

%

trans

ition

prob

abilit

y fro

m

hype

rtens

ion to

CV

D

1% ut

ility

disco

unt

3%

utilit

y dis

coun

t

RR re

ducti

on

of C

VD

de

pend

on

age

Appl

ying

ut

ilities

from

G

lobal

Burd

en

Dise

ase

Hyp

erten

sion

prev

alenc

e by

natio

nal s

urve

y

Star

t scr

eeni

ng at

age o

f 35 y

ears

, mal

es

E110

,834

13,86

57,8

0210

,843

10,82

410

,501

11,16

77,9

266,2

498,7

0314

,376

14,05

022

,974

9,468

6,873

6,237

E24,3

135,9

262,7

014,1

124,5

143,9

804,6

463,1

562,4

883,3

135,9

645,5

979,1

653,7

362,7

362,1

37

E2 un

til 55

+ E1

5,276

7,098

3,453

5,106

5,446

4,942

5,609

3,860

3,043

4,108

7,207

6,849

11,22

14,5

843,3

472,7

42

E2 un

til 60

+ E1

4,866

6,599

3,132

4,682

5,049

4,532

5,199

3,560

2,807

3,769

6,677

6,316

10,34

74,2

253,0

872,4

84

E1&

T.20

%7,2

749,4

165,1

327,2

837,2

646,9

417,6

075,3

224,1

965,7

5912

,737

9,434

15,42

56,3

374,6

144,0

32

E2&

T.20

%2,6

673,8

061,5

282,5

282,8

062,3

343,0

001,9

511,5

381,9

513,8

423,4

615,6

672,2

871,6

921,1

36

Star

t scr

eeni

ng at

age o

f 45 y

ears

, mal

es

E16,8

348,9

824,6

866,8

466,8

236,3

837,2

855,0

023,9

455,3

229,3

328,2

5211

,784

6,088

4,340

4,204

E22,3

033,4

641,1

422,1

692,4

371,8

522,7

551,6

861,3

291,5

653,5

062,7

813,9

742,0

181,4

631,0

67

E2 un

til 55

+ E1

2,631

3,865

1,398

2,508

2,755

2,180

3,083

1,926

1,519

1,836

3,929

3,178

4,542

2,315

1,671

1,294

E2 un

til 60

+ E1

2,434

3,624

1,244

2,304

2,563

1,982

2,885

1,781

1,405

1,673

3,674

2,939

4,200

2,137

1,546

1,157

E1&

T.20

%4,3

125,8

302,7

944,3

244,3

003,8

614,7

633,1

562,4

893,2

296,0

915,2

067,4

353,8

212,7

382,4

59

E2&

T.20

%1,1

111,9

3129

01,0

191,2

0265

91,5

6281

364

157

51,9

741,3

411,9

1694

670

524

2

Star

t scr

eeni

ng at

age o

f 55 y

ears

, mal

es

E12,6

643,9

541,3

742,6

782,6

502,0

263,3

021,9

541,5

421,7

194,1

973,0

203,8

352,6

671,6

981,9

25

E212

184

9D

omina

nt62

181

Dom

inant

759

8970

Dom

inant

931

137

174

173

77D

omina

nt

E2 un

til 60

+ E1

1,617

2,676

558

1,600

1,634

979

2,255

1,186

936

850

2,853

1,834

2,329

1,643

1,031

1,036

E1&

T.20

%1,1

492,0

6123

81,1

631,1

3651

21,7

8784

366

6D

omina

nt2,2

531,3

031,6

551,1

8173

362

9

E2&

T.20

%D

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntDo

mina

ntDo

mina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

ntD

omina

nt

*Not

e: E

1: A

nnua

l scr

eeni

ng, E

2: B

iann

ual s

cree

ning

, E2

until

60+

E1:

Bia

nnua

l scr

eeni

ng u

ntil

60 y

ears

old

then

ann

ual s

cree

ning

unt

il di

ed,

E2 u

ntil

55+

E1: B

iann

ual s

cree

ning

unt

il 55

yea

rs o

ldth

en a

nnua

l scr

eeni

ng u

ntil

died

, E1&

T.20

%: A

nnua

l scr

eeni

ng c

ombi

ned

with

incr

easi

ng

cove

rage

of t

reat

men

t by

20%

, E2&

T.20

%: B

iann

ual s

cree

ning

com

bine

d w

ith in

crea

sing

cov

erag

e of

trea

tmen

t by

20%

, CV

D: C

ardi

o vas

cula

r di

seas

e.

196

Summary

SummaryNon-communicable diseases, such as cardiovascular diseases are a public health problem all over the world, also in developing counties. Strategies in both primary and secondary prevention are applied to deal with this problem. The research in this thesis was conducted to derive evidence which could inform policy makers and health planners to develop better plans for hypertension management, with a special focus on cost-effectiveness of screening policies to optimize control of the burden of hypertension in Vietnam.

This research applied several methods to identify the burden of a chronic disease and solutions to manage it. Regarding methodology, a few highlights should be noted as lessons learnt from this work. First, among various models available to predict CVD risks in hypertensive patients, the Asian and Chinese Multiple-provincial Cohort Study seems the most suitable for the Vietnamese population (Chapter 2: Models to predict the burden of cardiovascular disease risk in a rural mountainous region of Vietnam). This result revealed the necessity to choose models that are appropriate, especially when they were developed for a different population (or a population with a different background) and that real-life/local data may be required for confirmation. Second, to examine adherence to medicine among chronic patients, I applied quantitative methods but also did a qualitative study to obtain a deeper understanding of the issues surrounding quantification of levels of adherence and factors influencing it (Chapter 3: Adherence to hypertension medication: quantitative and qualitative investigations in a rural Northern Vietnamese community). One issue that arose was that whether patients receive medicine does not always reflect how much medicine the patients actually take; they may not take all they receive, or they may take more by buying on the free market. This finding complicates

Summary

interpretation of routine data on treatment. Third, there are severalways to quantify the burden of disease, such as prevalence of disease, costs of disease, costs of its complications, or measures of health utility status (Chapters 4 and 5). The evidence on these issues can be important to guide resource allocation by investigating different strategies for disease management using models considering all of these issues. Finally, the first cost-effectiveness models for Vietnam in these areas were built, requiring a number of input parameters. While local evidence on these parameters are not yet available, we can apply data synthesized and gathered from populations with a similar background. Chapter 6: Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam, integrates all the above aspects into an overall cost-effectiveness model for hypertension screening in the Vietnamese context.

The studies reported here confirm that hypertension is a great burden in Vietnam, from economic, clinical and social perspectives. Although hypertension management can reduce the burden of cardiovascular disease, early detection of hypertension and adherence to medicinesstill need to be greatly improved in the Vietnamese population. It would be cost-effective if screening for hypertension is started at 35, 45, or 55 years among males and at 55 years among females. Aside from cardiovascular risk, age also seems a promising factor to guide efforts to improve adherence to treatment and corresponding cost-effectiveness. Awareness of complications related to hypertension was given as the main reason for adherence to therapy, so that awareness should be increased.

In conclusion, studies reported in this thesis confirm that hypertension strongly contributes to the overall burden of cardiovascular disease in Vietnam. Adherence to medicines is an important issue to be improved; awareness of complications and age are guiding factors for

197

Summary

SummaryNon-communicable diseases, such as cardiovascular diseases are a public health problem all over the world, also in developing counties. Strategies in both primary and secondary prevention are applied to deal with this problem. The research in this thesis was conducted to derive evidence which could inform policy makers and health planners to develop better plans for hypertension management, with a special focus on cost-effectiveness of screening policies to optimize control of the burden of hypertension in Vietnam.

This research applied several methods to identify the burden of a chronic disease and solutions to manage it. Regarding methodology, a few highlights should be noted as lessons learnt from this work. First, among various models available to predict CVD risks in hypertensive patients, the Asian and Chinese Multiple-provincial Cohort Study seems the most suitable for the Vietnamese population (Chapter 2: Models to predict the burden of cardiovascular disease risk in a rural mountainous region of Vietnam). This result revealed the necessity to choose models that are appropriate, especially when they were developed for a different population (or a population with a different background) and that real-life/local data may be required for confirmation. Second, to examine adherence to medicine among chronic patients, I applied quantitative methods but also did a qualitative study to obtain a deeper understanding of the issues surrounding quantification of levels of adherence and factors influencing it (Chapter 3: Adherence to hypertension medication: quantitative and qualitative investigations in a rural Northern Vietnamese community). One issue that arose was that whether patients receive medicine does not always reflect how much medicine the patients actually take; they may not take all they receive, or they may take more by buying on the free market. This finding complicates

Summary

interpretation of routine data on treatment. Third, there are severalways to quantify the burden of disease, such as prevalence of disease, costs of disease, costs of its complications, or measures of health utility status (Chapters 4 and 5). The evidence on these issues can be important to guide resource allocation by investigating different strategies for disease management using models considering all of these issues. Finally, the first cost-effectiveness models for Vietnam in these areas were built, requiring a number of input parameters. While local evidence on these parameters are not yet available, we can apply data synthesized and gathered from populations with a similar background. Chapter 6: Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam, integrates all the above aspects into an overall cost-effectiveness model for hypertension screening in the Vietnamese context.

The studies reported here confirm that hypertension is a great burden in Vietnam, from economic, clinical and social perspectives. Although hypertension management can reduce the burden of cardiovascular disease, early detection of hypertension and adherence to medicinesstill need to be greatly improved in the Vietnamese population. It would be cost-effective if screening for hypertension is started at 35, 45, or 55 years among males and at 55 years among females. Aside from cardiovascular risk, age also seems a promising factor to guide efforts to improve adherence to treatment and corresponding cost-effectiveness. Awareness of complications related to hypertension was given as the main reason for adherence to therapy, so that awareness should be increased.

In conclusion, studies reported in this thesis confirm that hypertension strongly contributes to the overall burden of cardiovascular disease in Vietnam. Adherence to medicines is an important issue to be improved; awareness of complications and age are guiding factors for

198

Summary

interventions to enhance medical adherence. Based on potentiallyfavourable cost-effectiveness, we recommend that screening and managing hypertension should be integrated into routine care in Vietnam. These results and conclusions are relevant for other countries with similar conditions and states of development especially with regard to their health systems.

Summary

Nederlandse SamenvattingChronische ziekten presenteren een “public health” probleem in zowel Westerse- als ontwikkelingslanden. Er is daarbij een combinatie nodig van strategieën in zowel de eerste- als tweedelijns zorg. Het onderzoek gerapporteerd in dit proefschrift beoogt “building blocks” aan te leveren zodat goede planning voor de management van hypertensie mogelijk wordt, in het bijzonder inzake de kosten-effectiviteit van screening in Vietnam.

Diverse methoden werden gebruik en sommige kunnen aangevinkt worden als “highlights”, met name in de context van Vietnam. Ten eerste, van de diverse beschikbare risicomodellen om cardiovasculaire ziekte te voorspellen, zijn de “Asian” en “Chinese Multiple-provincial Cohort Study” het meest geschikt voor toepassing op de Vietnamese bevolking (Chapter 2: Models to predict the burden of cardiovascular disease risk in a rural mountainous region of Vietnam). Dit hoofdstuk onderschrijft het belang van het gebruik van modellen die adequaat zijn voor de specifieke omgeving en bevolking, zeker indien ontwikkeld voor andere landen, en toont de noodzaak deze modellen met lokale data te valideren. Ten tweede, werden voor het bestuderen van adherentie aan medicatie bij chronische patiënten met name ook kwalitatieve methoden gehanteerd naast kwantitatieve, met name om een nog beter begrip te krijgen van de beweegredenen van patiënten niet-adherent te zijn (Chapter 3: Adherence to hypertension medication: quantitative and qualitative investigations in a rural Northern Vietnamese community). Duidelijk werd o.a. dat de medicatie die patiënten ontvangen lang niet altijd weergeeft wat ze ook daadwerkelijk gebruiken; ze laten het mogelijk deels ongebruikt of kopen extra op de vrije markt. Ten derde, bleken er diverse methoden manieren te zijn om o.b.v. Vietnamese data de ziektelast,prevalentie, kosten van complicaties en utilities te schatten (Chapters

199

Summary

interventions to enhance medical adherence. Based on potentiallyfavourable cost-effectiveness, we recommend that screening and managing hypertension should be integrated into routine care in Vietnam. These results and conclusions are relevant for other countries with similar conditions and states of development especially with regard to their health systems.

Summary

Nederlandse SamenvattingChronische ziekten presenteren een “public health” probleem in zowel Westerse- als ontwikkelingslanden. Er is daarbij een combinatie nodig van strategieën in zowel de eerste- als tweedelijns zorg. Het onderzoek gerapporteerd in dit proefschrift beoogt “building blocks” aan te leveren zodat goede planning voor de management van hypertensie mogelijk wordt, in het bijzonder inzake de kosten-effectiviteit van screening in Vietnam.

Diverse methoden werden gebruik en sommige kunnen aangevinkt worden als “highlights”, met name in de context van Vietnam. Ten eerste, van de diverse beschikbare risicomodellen om cardiovasculaire ziekte te voorspellen, zijn de “Asian” en “Chinese Multiple-provincial Cohort Study” het meest geschikt voor toepassing op de Vietnamese bevolking (Chapter 2: Models to predict the burden of cardiovascular disease risk in a rural mountainous region of Vietnam). Dit hoofdstuk onderschrijft het belang van het gebruik van modellen die adequaat zijn voor de specifieke omgeving en bevolking, zeker indien ontwikkeld voor andere landen, en toont de noodzaak deze modellen met lokale data te valideren. Ten tweede, werden voor het bestuderen van adherentie aan medicatie bij chronische patiënten met name ook kwalitatieve methoden gehanteerd naast kwantitatieve, met name om een nog beter begrip te krijgen van de beweegredenen van patiënten niet-adherent te zijn (Chapter 3: Adherence to hypertension medication: quantitative and qualitative investigations in a rural Northern Vietnamese community). Duidelijk werd o.a. dat de medicatie die patiënten ontvangen lang niet altijd weergeeft wat ze ook daadwerkelijk gebruiken; ze laten het mogelijk deels ongebruikt of kopen extra op de vrije markt. Ten derde, bleken er diverse methoden manieren te zijn om o.b.v. Vietnamese data de ziektelast,prevalentie, kosten van complicaties en utilities te schatten (Chapters

200

Summary

4 and 5). Tenslotte, werden de eerste kosten-effectiviteitsmodellen voor Vietnam op dit gebied gebouwd, met een noodzaak voor data als verzameld en verdere data verzameling. Met de kennis vergaard via deze methoden hoop ik bij te kunnen dragen aan betere allocatie van middelen door rationele investeringen o.b.v. kosten-effectiviteitsanalyse. Dergelijke analyses kunnen gegevens gebruiken afkomstig uit vergelijkbare context en vice versa, kunnen andere vergelijkbare landen baat hebben bij de Vietnamese analyses. Het laatste hoofdstuk (Chapter 6: Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam) integreert alle bovenstaande aspecten in de eerste kosten-effectiviteitsanalyse van hypertensie screening voor de Vietnamese context.

Onze studies bevestigen dat hypertensie een forse last geeft in Vietnam, zowel van het economische, alswel het klinische en sociale perspectief. Hoewel behandeling van hypertensie de ziektelast aanzienlijk kan verlagen, kunnen vroege diagnose en adherentie aan medicatie verdere belangrijke bijdragen leveren aan het oplosen van het probleem. Het bleek kosten-effectief te zijn te screenen van 35 jaar bij mannen en vanaf 55 jaar bij vrouwen. Ergo, naast een cardiovasculaire risicoscore blijkt ook leeftijd een belangrijke factor om op geleide daarvan inspanningen te verrichten behandeling te verbeteren en adherentie te verhogen. Ook kennis van complicaties bleek van groot belang, met name werd de opvatting dat er geen ernstige complicaties zijn vaak als belangrijke reden gegeven niet adherent te zijn.

Concluderend is te stellen dat de studies in dit proefschrift de relevantie van hypertensie bevestigen als oorzaak voor de “overall burden” van cardiovasculaire ziekte in Vietnam. Adherentie is een

Summary

belangrijk aspect in het geheel en toegenomen bewustwording van complicaties en een leeftijd-specifieke aanpak kunnen leidend zijn in verbetering van de situatie. Op basis van gunstige kosten-effectiviteit kan n.a.lv. de studies aanbevolen worden screening op en management van hypertensie in de routine zorg in Vietnam op te nemen. Onze bevindingen en conclusies zijn allicht van belang voor Vietnam, maar mogelijk ook voor andere landen in vergelijkbare stadia van ontwikkeling.

201

Summary

4 and 5). Tenslotte, werden de eerste kosten-effectiviteitsmodellen voor Vietnam op dit gebied gebouwd, met een noodzaak voor data als verzameld en verdere data verzameling. Met de kennis vergaard via deze methoden hoop ik bij te kunnen dragen aan betere allocatie van middelen door rationele investeringen o.b.v. kosten-effectiviteitsanalyse. Dergelijke analyses kunnen gegevens gebruiken afkomstig uit vergelijkbare context en vice versa, kunnen andere vergelijkbare landen baat hebben bij de Vietnamese analyses. Het laatste hoofdstuk (Chapter 6: Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam) integreert alle bovenstaande aspecten in de eerste kosten-effectiviteitsanalyse van hypertensie screening voor de Vietnamese context.

Onze studies bevestigen dat hypertensie een forse last geeft in Vietnam, zowel van het economische, alswel het klinische en sociale perspectief. Hoewel behandeling van hypertensie de ziektelast aanzienlijk kan verlagen, kunnen vroege diagnose en adherentie aan medicatie verdere belangrijke bijdragen leveren aan het oplosen van het probleem. Het bleek kosten-effectief te zijn te screenen van 35 jaar bij mannen en vanaf 55 jaar bij vrouwen. Ergo, naast een cardiovasculaire risicoscore blijkt ook leeftijd een belangrijke factor om op geleide daarvan inspanningen te verrichten behandeling te verbeteren en adherentie te verhogen. Ook kennis van complicaties bleek van groot belang, met name werd de opvatting dat er geen ernstige complicaties zijn vaak als belangrijke reden gegeven niet adherent te zijn.

Concluderend is te stellen dat de studies in dit proefschrift de relevantie van hypertensie bevestigen als oorzaak voor de “overall burden” van cardiovasculaire ziekte in Vietnam. Adherentie is een

Summary

belangrijk aspect in het geheel en toegenomen bewustwording van complicaties en een leeftijd-specifieke aanpak kunnen leidend zijn in verbetering van de situatie. Op basis van gunstige kosten-effectiviteit kan n.a.lv. de studies aanbevolen worden screening op en management van hypertensie in de routine zorg in Vietnam op te nemen. Onze bevindingen en conclusies zijn allicht van belang voor Vietnam, maar mogelijk ook voor andere landen in vergelijkbare stadia van ontwikkeling.

202

List of publications

List of publicationsPeer reviewed international publications

Thi-Phuong-Lan Nguyen, E. P. Wright, Thanh-Trung Nguyen, C.C.M. Schuiling-Veninga, M.J. Bijlsma , Thi-Bach-Yen Nguyen, M.J. Postma. Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam. PLoS One, May 18, 2016.

Thi-Phuong-Lan Nguyen, C.C.M. Schuiling-Veninga,Thi-Bach-Yen Nguyen, Thu-Hang Vu, E. P. Wright, and M.J. Postma. Adherence to hypertension medication: quantitative and qualitative investigations in a rural Northern Vietnamese community. (submitted for publication – under review)

Thi-Phuong-Lan Nguyen, P.F. Krabbe, Thi-Bach-Yen Nguyen,C.C.M. Schuiling-Veninga, E.P. Wright, M.J.Postma. Utilities of Patients with Hypertension in Northern Vietnam. PLoS One, October 27, 2015.

Thi-Phuong-Lan Nguyen, Thi-Bach-Yen Nguyen, Thanh-Trung Nguyen, Van-Vinh Hac, H.H. Le, M.J. Postma. Direct costs of hypertensive patients admitted to hospital in Vietnam - a bottom-up micro-costing analysis. BMC Health Serv Res, October 28, 2014.

Thi-Phuong-Lan Nguyen, C.C.M. Schuiling-Veninga, Thi-Thu-Hang Vu, E. P. Wright, M.J. Postma. Models to Predict the Burden of Cardiovascular Disease Risk in a Rural Mountainous Region of Vietnam. Value in Health, May 2014.

Swee May Cripe, Thi-Thanh-Tu Phung, Thi-Phuong-Lan Nguyen,M.A. Williams. Risk Factors Associated with Stillbirth in Thai Nguyen Province, Vietnam. Tropical Pediatrics, October 2007.

203


List of publicationsPeer reviewed international publications

Thi-Phuong-Lan Nguyen, E. P. Wright, Thanh-Trung Nguyen, C.C.M. Schuiling-Veninga, M.J. Bijlsma , Thi-Bach-Yen Nguyen, M.J. Postma. Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam. PLoS One, May 18, 2016.

Thi-Phuong-Lan Nguyen, C.C.M. Schuiling-Veninga,Thi-Bach-Yen Nguyen, Thu-Hang Vu, E. P. Wright, and M.J. Postma. Adherence to hypertension medication: quantitative and qualitative investigations in a rural Northern Vietnamese community. (submitted for publication – under review)

Thi-Phuong-Lan Nguyen, P.F. Krabbe, Thi-Bach-Yen Nguyen,C.C.M. Schuiling-Veninga, E.P. Wright, M.J.Postma. Utilities of Patients with Hypertension in Northern Vietnam. PLoS One, October 27, 2015.

Thi-Phuong-Lan Nguyen, Thi-Bach-Yen Nguyen, Thanh-Trung Nguyen, Van-Vinh Hac, H.H. Le, M.J. Postma. Direct costs of hypertensive patients admitted to hospital in Vietnam - a bottom-up micro-costing analysis. BMC Health Serv Res, October 28, 2014.

Thi-Phuong-Lan Nguyen, C.C.M. Schuiling-Veninga, Thi-Thu-Hang Vu, E. P. Wright, M.J. Postma. Models to Predict the Burden of Cardiovascular Disease Risk in a Rural Mountainous Region of Vietnam. Value in Health, May 2014.

Swee May Cripe, Thi-Thanh-Tu Phung, Thi-Phuong-Lan Nguyen,M.A. Williams. Risk Factors Associated with Stillbirth in Thai Nguyen Province, Vietnam. Tropical Pediatrics, October 2007.

204


Vietnamese publication

Thi-Phuong-Lan Nguyen, Thi-Tuyet Dam. Comparison of SF-36and EQ-5D in hypertensive patients. Journal of Practical Medicine, May 2015.

International Conference presentations

Thi-Phuong-Lan Nguyen, E.P.Wright, Thanh-Trung Nguyen, C.C.M. Schuiling-Veninga, M.J. Bijlsma, Thi-Bach-Yen Nguyen, M.J. Postma. Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam. Poster presentation at ISPOR 7th Asia-Pacific Conference. 2016.

Thi-Phuong-Lan Nguyen, Thanh-Trung Nguyen, M.J. Postma. Economic burden of acute myocardial infarction in Vietnam. Poster presentation at ISPOR 18th Annual European Congress. 2015.

Thi-Phuong-Lan Nguyen, C.C.M. Schuiling-Veninga, M.J. Postma. Health utilities of hypertensive patients in Vietnam. Poster presentation at ISPOR 20th Annual International Congress. 2015

Thi-Phuong-Lan Nguyen, C.C.M. Schuiling-Veninga, M.J. Postma. Cardiovascular risk, gender and medication adherence in a rural area of Vietnam. Poster presentation at ISPOR 17th Annual European Congress. 2014.

Research reports

Thi-Bach-Yen Nguyen, Hoang-Lan Nguyen, Thi-Phuong-Lan Nguyen, Huu-Thang Nguyen. Cost- benefit analysis of providing family planning services for vulnerable populations in Lam Dong


and Quang Binh Provinces. (2015). For Marie Stopes International Vietnam.

Vinh-Van Hac, Thi-Phuong-Lan Nguyen, Thi-Bach-Yen, et all. Cost Analysis of Medical Doctor Training among eight medical faculties in Vietnam. (2013). Sponsored by Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Viet Nam.

Van-Anh Hoang, Thi-Phuong-Lan Nguyen, Thi-Bach-Yen Nguyen. Costs of kidney failure patients admitted to Thai Nguyen hospital, Vietnam. (2013). Sponsored by Centers of Excellence Health Economic: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Viet Nam.

B.P. Teerawichitchainan, Van-Vinh Hac, Thi-Phuong- Lan Nguyen. Changing transition to adulthood in Vietnam's remote Northern uplands, a focus on ethnic minority youth and their family. (2008) New York: Population Council.

205


Vietnamese publication

Thi-Phuong-Lan Nguyen, Thi-Tuyet Dam. Comparison of SF-36and EQ-5D in hypertensive patients. Journal of Practical Medicine, May 2015.

International Conference presentations

Thi-Phuong-Lan Nguyen, E.P.Wright, Thanh-Trung Nguyen, C.C.M. Schuiling-Veninga, M.J. Bijlsma, Thi-Bach-Yen Nguyen, M.J. Postma. Cost-effectiveness analysis of screening for and managing identified hypertension for cardiovascular disease prevention in Vietnam. Poster presentation at ISPOR 7th Asia-Pacific Conference. 2016.

Thi-Phuong-Lan Nguyen, Thanh-Trung Nguyen, M.J. Postma. Economic burden of acute myocardial infarction in Vietnam. Poster presentation at ISPOR 18th Annual European Congress. 2015.

Thi-Phuong-Lan Nguyen, C.C.M. Schuiling-Veninga, M.J. Postma. Health utilities of hypertensive patients in Vietnam. Poster presentation at ISPOR 20th Annual International Congress. 2015

Thi-Phuong-Lan Nguyen, C.C.M. Schuiling-Veninga, M.J. Postma. Cardiovascular risk, gender and medication adherence in a rural area of Vietnam. Poster presentation at ISPOR 17th Annual European Congress. 2014.

Research reports

Thi-Bach-Yen Nguyen, Hoang-Lan Nguyen, Thi-Phuong-Lan Nguyen, Huu-Thang Nguyen. Cost- benefit analysis of providing family planning services for vulnerable populations in Lam Dong


and Quang Binh Provinces. (2015). For Marie Stopes International Vietnam.

Vinh-Van Hac, Thi-Phuong-Lan Nguyen, Thi-Bach-Yen, et all. Cost Analysis of Medical Doctor Training among eight medical faculties in Vietnam. (2013). Sponsored by Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Viet Nam.

Van-Anh Hoang, Thi-Phuong-Lan Nguyen, Thi-Bach-Yen Nguyen. Costs of kidney failure patients admitted to Thai Nguyen hospital, Vietnam. (2013). Sponsored by Centers of Excellence Health Economic: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Viet Nam.

B.P. Teerawichitchainan, Van-Vinh Hac, Thi-Phuong- Lan Nguyen. Changing transition to adulthood in Vietnam's remote Northern uplands, a focus on ethnic minority youth and their family. (2008) New York: Population Council.

206

Acknowledgements

AcknowledgementsThis work for this thesis has been done with great support from my supervisors, colleagues and friends in an international environment and from my lovely family.

I would like to offer the nicest words to my family. Mum, you gave special supports to your oldest child. You not only encouraged me to go abroad but also I found that no one can take care of my child as you did. Because of those contributions, I had more motivations to leave home to study with less worries about how the little girl was growing up. Thanks to my lovely child for your smiles, no complaints to me, being so friendly with your grandmother and making her smile while I was away. I am especially very thankful to my father for his being the model in my life. His encouragement "always believe in yourself and you will have success" goes along with me throughout my PhD research as well as my career. My dears: Viet, Lam, Linh, thanks a lot for your encouragement and support during my trips to the Netherlands. Far from you all for those periods of time made me understand how important you are in my life!

I would like to give special thanks to my promotor, Prof. Maarten J. Postma. You have been the best boss and teacher in my life! Although I was not sure what should I do for PhD research after our first meeting in Hanoi, I had a positive feeling and believed at that moment that I could go further under your supervision. The stress of my PhD study has been gone since that moment. Thanks for giving me a good chance to work in our creative group, freedom to go on with my studies and for always answering me whenever I have questions. “Never make a choice among various options, try all” was a very interesting lesson!

It will be big mistake if I give less thanks to my co-supervisors, Yen, Pamela and Nynke. All of your support during my study is highly

Acknowledgements

appreciated. This work could not have been done without your valuable comments on doing fieldwork, writing manuscripts, editing papers, dealing with daily life and having fun. I enjoyed to work and talk with you all! Many thanks to Nynke for joining Maarten on meetings in Hanoi/Thai Nguyen with the whole supervisory group from Vietnam (Yen and Pamela).

To a friend, great thanks to Hong Anh, I still remember my feeling/motivation when I first made a call to you. Thanks a lot for introducing me to M.J.P. and our group and then always considering my PhD progress. Your contribution and success in scientific works inspired me to try my best during my study here in the peaceful Groningen city and also later in my career.

I would like to thank to Professors John Stewart, Peter Coyte, Bart Van Bergen and Peter Mason. It was really a great opportunity to work together in the CoE project and I got a lot of lessons learnt which could be applied in my PhD study.

To my friends and fellow students in Vietnam, Dung, Minh, Tien Thang, Hang and her husband, Hoang, Son, Trang, Hanh and many others, all of you gave me great advice and support in data collection whether I was in Vietnam or in Groningen. We suffered from the sun of a terrible summer during the fieldwork in 2012. I know that all of us were extremely tired, very patient with the patients and overcame difficult roads in mountain areas, But I learned great lessons from you, the patients and our work, and also had a lot of fun. Many thanks go to Doctor Sy and his wife, Xuan and health staffs at the field sites, accompanying me during the fieldwork. And thanks to Dr. Trung for his agreement for me to collect data at the Thai Nguyen hospital.

To my colleagues and friends in Groningen: thanks to you all for sharing scientific information and ideas, having fun and inspiring my

207

Acknowledgements

AcknowledgementsThis work for this thesis has been done with great support from my supervisors, colleagues and friends in an international environment and from my lovely family.

I would like to offer the nicest words to my family. Mum, you gave special supports to your oldest child. You not only encouraged me to go abroad but also I found that no one can take care of my child as you did. Because of those contributions, I had more motivations to leave home to study with less worries about how the little girl was growing up. Thanks to my lovely child for your smiles, no complaints to me, being so friendly with your grandmother and making her smile while I was away. I am especially very thankful to my father for his being the model in my life. His encouragement "always believe in yourself and you will have success" goes along with me throughout my PhD research as well as my career. My dears: Viet, Lam, Linh, thanks a lot for your encouragement and support during my trips to the Netherlands. Far from you all for those periods of time made me understand how important you are in my life!

I would like to give special thanks to my promotor, Prof. Maarten J. Postma. You have been the best boss and teacher in my life! Although I was not sure what should I do for PhD research after our first meeting in Hanoi, I had a positive feeling and believed at that moment that I could go further under your supervision. The stress of my PhD study has been gone since that moment. Thanks for giving me a good chance to work in our creative group, freedom to go on with my studies and for always answering me whenever I have questions. “Never make a choice among various options, try all” was a very interesting lesson!

It will be big mistake if I give less thanks to my co-supervisors, Yen, Pamela and Nynke. All of your support during my study is highly

Acknowledgements

appreciated. This work could not have been done without your valuable comments on doing fieldwork, writing manuscripts, editing papers, dealing with daily life and having fun. I enjoyed to work and talk with you all! Many thanks to Nynke for joining Maarten on meetings in Hanoi/Thai Nguyen with the whole supervisory group from Vietnam (Yen and Pamela).

To a friend, great thanks to Hong Anh, I still remember my feeling/motivation when I first made a call to you. Thanks a lot for introducing me to M.J.P. and our group and then always considering my PhD progress. Your contribution and success in scientific works inspired me to try my best during my study here in the peaceful Groningen city and also later in my career.

I would like to thank to Professors John Stewart, Peter Coyte, Bart Van Bergen and Peter Mason. It was really a great opportunity to work together in the CoE project and I got a lot of lessons learnt which could be applied in my PhD study.

To my friends and fellow students in Vietnam, Dung, Minh, Tien Thang, Hang and her husband, Hoang, Son, Trang, Hanh and many others, all of you gave me great advice and support in data collection whether I was in Vietnam or in Groningen. We suffered from the sun of a terrible summer during the fieldwork in 2012. I know that all of us were extremely tired, very patient with the patients and overcame difficult roads in mountain areas, But I learned great lessons from you, the patients and our work, and also had a lot of fun. Many thanks go to Doctor Sy and his wife, Xuan and health staffs at the field sites, accompanying me during the fieldwork. And thanks to Dr. Trung for his agreement for me to collect data at the Thai Nguyen hospital.

To my colleagues and friends in Groningen: thanks to you all for sharing scientific information and ideas, having fun and inspiring my

208

Acknowledgements

life, Bob, Eelko, Jannie, Bert, Jelena, Job, Pepijn, Tia, Thea (special pizza maker), Christian, Thang, Hoa, Thao, Pieter, Marcy, Maarten B., Ury, Ira, Neily, Jurjen (BBQ chef) and the other new guys/girls. Hao and Hieu, thanks for introducing life in Groningen and sharing the experience with me. Have a lot of success and enjoy your life!

Thank you a lot for organizing wonderful trips and financial support: Wiebe Zijlstra, Mike Robertson and Thuy Hanh from Hanoi Medical University.

Huge thanks go to Thai Nguyen University of Medicine and Pharmacy, the Board of Deans and Doctor Hac Van Vinh and my colleagues at the Faculty of Public Health for all of your advice and support.

Finally, I would give my special thanks to those who worked with me until the last minutes to complete this book. Thanks to Pamela for your editing. Thanks to the photographers who gave me the beautiful photos presented on the front and back covers, Maarten for photo of Thien Cung cave in Ha Long bay and Bay for photo of Tulip flowers. The two images are symbols of the two countries, my Vietnam and the Netherlands, where I have been happy and could have success in my life.

Research Institute SHARE

Research Institute SHARE This thesis is published within the Research Institute SHARE(Science in Healthy Ageing and healthcaRE) of the University Medical Center Groningen / University of Groningen.

Further information regarding the institute and its research can be obtained from our internetsite: http://www.share.umcg.nl/

More recent theses can be found in the list below.

((co-) supervisors are between brackets)

2016

Küpers LKThe first 1000 days and beyond(prof H Snieder, prof RP Stolk, dr E Corpeleijn)

Kuiper JSThe importance of social relationships in the process of cognitive ageing(prof RC Oude Voshaar, prof RP Stolk, dr N Smidt)

Göhner, CPlacental particles in pregnancy and preeclampsia(prof SA Scherjon,prof E Schleuβner, dr MM Faas, dr T Plösch)

Vries AJ dePatellar tendinopathy; causes, consequences and the use of orthoses(prof RL Diercks, dr I van den Akker-Scheek, dr J Zwerver, dr H van der Worp)

209

Acknowledgements

life, Bob, Eelko, Jannie, Bert, Jelena, Job, Pepijn, Tia, Thea (special pizza maker), Christian, Thang, Hoa, Thao, Pieter, Marcy, Maarten B., Ury, Ira, Neily, Jurjen (BBQ chef) and the other new guys/girls. Hao and Hieu, thanks for introducing life in Groningen and sharing the experience with me. Have a lot of success and enjoy your life!

Thank you a lot for organizing wonderful trips and financial support: Wiebe Zijlstra, Mike Robertson and Thuy Hanh from Hanoi Medical University.

Huge thanks go to Thai Nguyen University of Medicine and Pharmacy, the Board of Deans and Doctor Hac Van Vinh and my colleagues at the Faculty of Public Health for all of your advice and support.

Finally, I would give my special thanks to those who worked with me until the last minutes to complete this book. Thanks to Pamela for your editing. Thanks to the photographers who gave me the beautiful photos presented on the front and back covers, Maarten for photo of Thien Cung cave in Ha Long bay and Bay for photo of Tulip flowers. The two images are symbols of the two countries, my Vietnam and the Netherlands, where I have been happy and could have success in my life.


Research Institute SHARE This thesis is published within the Research Institute SHARE(Science in Healthy Ageing and healthcaRE) of the University Medical Center Groningen / University of Groningen.

Further information regarding the institute and its research can be obtained from our internetsite: http://www.share.umcg.nl/

More recent theses can be found in the list below.

((co-) supervisors are between brackets)

2016

Küpers LKThe first 1000 days and beyond(prof H Snieder, prof RP Stolk, dr E Corpeleijn)

Kuiper JSThe importance of social relationships in the process of cognitive ageing(prof RC Oude Voshaar, prof RP Stolk, dr N Smidt)

Göhner, CPlacental particles in pregnancy and preeclampsia(prof SA Scherjon,prof E Schleuβner, dr MM Faas, dr T Plösch)

Vries AJ dePatellar tendinopathy; causes, consequences and the use of orthoses(prof RL Diercks, dr I van den Akker-Scheek, dr J Zwerver, dr H van der Worp)

210


Holland B van

Promotion of sustainable employment; occupational health in the meat processing industry(prof MF Reneman, prof S Brouwer, dr R Soer, dr MR de Boer)13.04.2016 EXPAND

Otter TA

Monitoring endurance athletes; a multidisciplinary approach(prof KAPM Lemmink, prof RL Diercks, dr MS Brink)

Bielderman JH

Active ageing and quality of life; community-dwelling older adults in deprived neighbourhoods(prof CP van der Schans, dr MHG de Greef, dr GH Schout)

Bijlsma MJ

Age-period-cohort methodology; confounding by birth in cardiovascular pharmacoepidemiology(prof E Hak, prof S Vansteelandt, dr F Janssen)

Dingemans EAA

Working after retirement; determinants and conzequences of bridge employment(prof CJIM Henkens, dr ir H van Solinge)

Jonge L de

Data quality and methodology in studies on maternal medication use in relation to congenital anomalies(prof IM van Langen, prof LTW de Jong-van den Berg, dr MK Bakker)


Vries FM de

Statin treatment in type 2 diabetes patients(prof E Hak, prof P Denig, prof MJ Postma)

Jager M

Unraveling the role of client-professional communcation in adolescent psychosocial care(prof SA Reijneveld, prof EJ Knorth, dr AF de Winter, dr J Metselaar)

Mulder B

Medication use during pregnancy and atopic diseases in childhood(prof E Hak, prof SS Jick, dr CCM Schuling-Veninga, dr TW de Vries)

Romkema S

Intermanual transfer in prosthetic training(prof CK van der Sluis, dr RM Bongers)

Diest M van

Developing an exergame for unsupervised home-based balance training in older adults(prof GJ Verkerke, prof K Postema, dr CJC Lamoth, dr J Stegenga)

Waterschoot FPC

Nice to have or need to have? Unraveling dosage of pain rehabilitation(prof MF Reneman, prof JHB Geertzen, prof PU Dijkstra)

211


Holland B van

Promotion of sustainable employment; occupational health in the meat processing industry(prof MF Reneman, prof S Brouwer, dr R Soer, dr MR de Boer)13.04.2016 EXPAND

Otter TA

Monitoring endurance athletes; a multidisciplinary approach(prof KAPM Lemmink, prof RL Diercks, dr MS Brink)

Bielderman JH

Active ageing and quality of life; community-dwelling older adults in deprived neighbourhoods(prof CP van der Schans, dr MHG de Greef, dr GH Schout)

Bijlsma MJ

Age-period-cohort methodology; confounding by birth in cardiovascular pharmacoepidemiology(prof E Hak, prof S Vansteelandt, dr F Janssen)

Dingemans EAA

Working after retirement; determinants and conzequences of bridge employment(prof CJIM Henkens, dr ir H van Solinge)

Jonge L de

Data quality and methodology in studies on maternal medication use in relation to congenital anomalies(prof IM van Langen, prof LTW de Jong-van den Berg, dr MK Bakker)


Vries FM de

Statin treatment in type 2 diabetes patients(prof E Hak, prof P Denig, prof MJ Postma)

Jager M

Unraveling the role of client-professional communcation in adolescent psychosocial care(prof SA Reijneveld, prof EJ Knorth, dr AF de Winter, dr J Metselaar)

Mulder B

Medication use during pregnancy and atopic diseases in childhood(prof E Hak, prof SS Jick, dr CCM Schuling-Veninga, dr TW de Vries)

Romkema S

Intermanual transfer in prosthetic training(prof CK van der Sluis, dr RM Bongers)

Diest M van

Developing an exergame for unsupervised home-based balance training in older adults(prof GJ Verkerke, prof K Postema, dr CJC Lamoth, dr J Stegenga)

Waterschoot FPC

Nice to have or need to have? Unraveling dosage of pain rehabilitation(prof MF Reneman, prof JHB Geertzen, prof PU Dijkstra)

212


Zijlema WL

(Un)healthy in the city; adverse health effects of traffic-related noise and air pollution(Prof JGM Rosmalen, prof RP Stolk)

Zetstra-van der Woude AP

Data collection on risk factors in pregnancy(prof LTW de Jong-van den Berg, dr H Wang)

Mohammadi S

The intersecting system of patients with chronic pain and their family caregivers; cognitions, behaviors, and well-being(prof M Hagedoorn, prof R Sanderman, dr M Deghani)

Verbeek T

Pregnancy and psychopathology(prof MY Berger, prof CLH Bockting, dr H Burger, dr MG van Pampus

For more 2015 and earlier theses visit our website.

Curriculum vitae

Curriculum vitae Nguyen Thi Phuong Lan (1975) was born in Ninh Binh, Vietnam. She graduated from Thai Nguyen University of Medicine in 1998 with the degree of medical doctor, then worked as a lecturer and researcher at that university in the Faculty of Medicine. In 2001, she went to Thailand and gained a Master certification in Health System Development at Chulalongkorn University in May 2002. Upon her return to Vietnam she continued her work in the Thai Nguyen University of Medicine and Pharmacy.

During her working time in Thai Nguyen University of Medicine and Pharmacy, she also worked as co-operator in several international projects, such as the Vietnam-Sweden co-operation program to develop a Community Based Medical Education program, two projects with the Netherlands: “Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Viet Nam”, and “Strengthening Teaching and Research Capacity of Preventive Medicine in Vietnam” and Population Council Vietnam’s project on strengthening health services in Thai Nguyen province. Most of these works were related to medical education and capacity development of teaching staff and health services.

In September 2012, she started her PhD study at the Unit of Pharmaco-Therapy, Epidemiology & Economics (PTE2; formerly the Unit of PharmacoEpidemiology & PharmacoEconomics, PE2), Department of Pharmacy, Faculty of Mathematics and Natural Sciences, University of Groningen. During her study, Lan and colleagues published several scientific papers and gave poster presentations at international congresses such as those of the International Society of Pharmaco-economics and Outcomes Research on topics related to cardiovascular diseases.

213


Zijlema WL

(Un)healthy in the city; adverse health effects of traffic-related noise and air pollution(Prof JGM Rosmalen, prof RP Stolk)

Zetstra-van der Woude AP

Data collection on risk factors in pregnancy(prof LTW de Jong-van den Berg, dr H Wang)

Mohammadi S

The intersecting system of patients with chronic pain and their family caregivers; cognitions, behaviors, and well-being(prof M Hagedoorn, prof R Sanderman, dr M Deghani)

Verbeek T

Pregnancy and psychopathology(prof MY Berger, prof CLH Bockting, dr H Burger, dr MG van Pampus

For more 2015 and earlier theses visit our website.

Curriculum vitae

Curriculum vitae Nguyen Thi Phuong Lan (1975) was born in Ninh Binh, Vietnam. She graduated from Thai Nguyen University of Medicine in 1998 with the degree of medical doctor, then worked as a lecturer and researcher at that university in the Faculty of Medicine. In 2001, she went to Thailand and gained a Master certification in Health System Development at Chulalongkorn University in May 2002. Upon her return to Vietnam she continued her work in the Thai Nguyen University of Medicine and Pharmacy.

During her working time in Thai Nguyen University of Medicine and Pharmacy, she also worked as co-operator in several international projects, such as the Vietnam-Sweden co-operation program to develop a Community Based Medical Education program, two projects with the Netherlands: “Centers of Excellence for Human Resources for Health: University-based Centers to Act as Resource and Transfer Point for Development Across the Health Sector in Viet Nam”, and “Strengthening Teaching and Research Capacity of Preventive Medicine in Vietnam” and Population Council Vietnam’s project on strengthening health services in Thai Nguyen province. Most of these works were related to medical education and capacity development of teaching staff and health services.

In September 2012, she started her PhD study at the Unit of Pharmaco-Therapy, Epidemiology & Economics (PTE2; formerly the Unit of PharmacoEpidemiology & PharmacoEconomics, PE2), Department of Pharmacy, Faculty of Mathematics and Natural Sciences, University of Groningen. During her study, Lan and colleagues published several scientific papers and gave poster presentations at international congresses such as those of the International Society of Pharmaco-economics and Outcomes Research on topics related to cardiovascular diseases.

214

Curriculum vitae

In recent years, Lan has been involved in several research projects in Health Economics sponsored by Marie Stopes International Vietnam, the Center of Excellence in Health Economics at Hanoi Medical University and by the NUFFIC project on preventive medicine.

Lan, her future will continue to be at the Thai Nguyen University of Medicine and Pharmacy, teaching and doing research, also with international organizations and research institutions.

university of groningen health economics of screening for ......paranimfs pham thu hiendidik...

Documents