health informatics- umkc, presentation

45
Using SMS to Provide Continuous Assessment and Improve Health Outcomes for Children with Asthma By NIKHIL KASSETTY SANDEEP KESHETTI

Upload: nikhil-kassetty

Post on 21-Feb-2017

110 views

Category:

Healthcare


0 download

TRANSCRIPT

Page 1: Health Informatics- UMKC, Presentation

Using SMS to Provide Continuous Assessment and Improve Health Outcomes for Children with Asthma

ByNIKHIL KASSETTYSANDEEP KESHETTI

Page 2: Health Informatics- UMKC, Presentation

Introduction

→ The World Health Organization estimated in 2008 as People suffering from asthma across the world is 300 million People suffering from asthma in U.S is 17.3 billion individuals

including 5 million children

→ Asthma is the most prevalent chronic respiratory disorder in the U.S

→ Most school absences caused by chronic conditions of asthma in many children

→ Long term data shows that both prevalence and morbidity of asthma in the U.S are on the rise

What is the problem?

Page 3: Health Informatics- UMKC, Presentation

Got a Solution!

→ Managing Chronic health conditions with the adoption of mobile phones and text messaging

Why this solution??→ There are billions of mobile phones exchanging tens of billions of text/SMS messages

daily→ A survey showed that 27% of North American teens used Text messages at least once a

month→ It is increasing likely that a child living with asthma can have access to SMS→ The child’s mobile/phone use can be leveraged to help with asthma management

Page 4: Health Informatics- UMKC, Presentation

How it works ??!!!

→ There are several known factors that influence effective management of a chronic condition including

Open communication between patient and health

care providers

A patients' awareness of symptoms and knowledge of her/his condition and level of adherence to medical

regiments

Effective communication may prove as important as

proper medication choice in long term success of asthma

control

Page 5: Health Informatics- UMKC, Presentation

Why we need effective communication??!!!

→ Medical care for pediatric patients that have moderate to severe asthma includes regular visits to their pulmonologist every 3 to 4 months

→ This communication is vital and there is no communication between patient and physician between these visits unless there is an emergency situation

→ There are some questionnaires such as Asthma Therapy Assessment Questionnaire(ATAQ) were developed to assist physicians in identifying children at risk

→ These questions are only administrated when pediatric patients visit their doctors and not all doctors use them

Effective communication between patients and doctors positively influences chronic health management, as does increased patients awareness of their

symptoms and general knowledge of the condition

Page 6: Health Informatics- UMKC, Presentation

What we are going to develop??

→ Design and evaluation of the system that leverages SMS messaging to facilitate continuous communication between a child with asthma and a physician who treats that child based on Health Belief Model(HBM)

Our findings !!

The SMS system provided

physicians with continuous

assessment of their patient’s

asthma condition

The simple act of

communicating a child’s asthma

status interspersed

with knowledge about asthma

via SMS

With this leads to improved pulmonary

outcomes for pediatric patients

Page 7: Health Informatics- UMKC, Presentation

Related Work

Page 8: Health Informatics- UMKC, Presentation

Mobile applications for Health Management

→ The ubiquity of mobile phones, as well as this platform’s increasing computational and sensing capabilities, has made it an ideal experimental platform for mobile research

→ These are helpful to support general wellness or to track important objective indicators of chronic condition

→ To avoid additional hardware and software requirements, researchers have opted to use the short message service (SMS) that is readily available on mobile phones

→ This study offered patients simple SMS reminders about use of medication, which helped patients comply with daily medication regiments

→ The Centers for Disease Control (CDC), has used SMS to provide information about HIV test centers, and to deliver weekly flu maps across the United States

Page 9: Health Informatics- UMKC, Presentation

Communication between Stakeholders in Chronic Care→ Communication is one of the significant elements of social influence and a critical

process to support behavior change

→ It is particularly relevant for patients suffering from chronic diseases, because better communication leads to better health outcomes

→ Communication tools to support patients and health care providers include:1. Telemonitoring to monitor patients at a distance2. Teleconsulting to enable consultation between geographically separated

individuals3. Patient education

→ The dominant technologies that support communication include web applications, email communication, or mobile phone(SMS, voice mail)

→ The challenge is that the communication mechanism needs to fit into the daily routines of clinical practitioners

→ We chose SMS on the pediatric patient’s mobile phone because of highly adaption rates within the teen/youth population

Page 10: Health Informatics- UMKC, Presentation

Health Belief Model

→ Health Belief Model(HBM) is one of the most widely used conceptual frameworks in health behavior research

→ The HBM explains change and maintenance of health related behaviors and guides the framework for health behavior interventions

→ This model includes several elements that predict when people will take action to prevent or control illness

→ Two main components include1. Symptom awareness2. Knowledge acquisition

→ These two factors are crucial to individuals engaging in proactive behavior to improve their condition

Page 11: Health Informatics- UMKC, Presentation

System Design

Page 12: Health Informatics- UMKC, Presentation

System Design

→ System is design with two parts1. An SMS service as an asthma survey and education tool for children with asthma2. A Physician’s Dashboard as an interface to review patient’s status (the web based

visualization tool)→ The system explicitly supports

→ Communication between the patient and physician→ Between the physician and the other clinical support staff

→ The system implicitly supports→ The out of hand communication between the parents or other caregivers→ Between the doctor’s office and the family

Figure 1 to be included here

Page 13: Health Informatics- UMKC, Presentation

SMS Service

→ The SMS service sent questions about asthma management and questions about asthma knowledge directly to pediatric patients in a predetermined fashion

→ The patient decided what time of day they would like to receive the queries

→ For asthma management, we developed SMS-formatted questions based on a standard instrument used during doctor visits, the ATAQ (Asthma Therapy Assessment Questionnaire)

→ In addition, we formulated SMS-formatted asthma knowledge questions

→ The child would be asked whether a statement was true or false and would be sent the right answer regardless of his or her response

→ Our SMS gateway was instrumented so that we could monitor the response rates of the participants

Page 14: Health Informatics- UMKC, Presentation

SMS Service - Versions

→ We built two versions of the SMS service for the study1. Query version2. Query and knowledge version

→ Patients were randomly assigned to one of these groups or to third control group that receive no SMS messages

Page 15: Health Informatics- UMKC, Presentation

Query Version

→ In this version, one of 15 yes/no questions about the patient’s asthma symptoms and management was sent out every other day

→ Patients data was aggregated and visualized in the physicians dashboard

Page 16: Health Informatics- UMKC, Presentation

Query Version

→ After patients had responded to the 15 questions, the system calculated the “rolling ATAQ scores” that are sum of the points for a given response set

→ This score categorized patients into three zones1. RED: needs attention soon (a score of 3+ points)2. YELLOW: might need attention eventually(1-2 points)3. GREEN: ok (0 points)

→ In addition, based on questions that indicated that the patient may be trouble or headed for trouble, the system automatically categorized the patient into the red zone

Page 17: Health Informatics- UMKC, Presentation

Query and Knowledge Version

→ This version is the same as the query version with the addition of fifteen true/false questions about general asthma knowledge on days that they did not receive queries

→ E.g., Asthma is a psychological condition F=False, T=True

→ The system separately calculated the rolling ATAQ scores and the knowledge scores

→ In order to increase the patients knowledge about asthma, patients were told whether or not they correctly answered the question, and were then given information related to the specific question

→ The patients response to the knowledge questions were also aggregated in the physician dashboard and the system calculated the rolling knowledge scores whenever it had a new answer

Page 18: Health Informatics- UMKC, Presentation

User Interaction

→ The layout of the screens varied based on the specific mobile phone’s text messaging capabilities

→ When a child with asthma received a message, he or she could reply to the message by entering ‘y/n’ or ‘t/f’

→ For knowledge questions, the SMS service sent the correct information regardless of the user’s answer

Page 19: Health Informatics- UMKC, Presentation

Physician’s Dashboard→ Physician’s Dashboard was populated with their patients’ SMS responses

→ This allowed physicians to monitor patients’ status and to interact with their care team for administrating care outside of the physician’s office

→ The web interface helped highlight patients in at least three different zones1. RED: needs attention soon (a score of 3+ points)2. YELLOW: might need attention eventually(1-2 points)3. GREEN: ok (0 points)

Page 20: Health Informatics- UMKC, Presentation

Physician’s Dashboard→ This Dashboard contained all of the relevant care information for each patient in the

study, including→ Summaries of asthma management status → And/or knowledge based on patients responses to the SMS-facilitated surveys,→ Relevant identification information to link to the electronic medical record system

used in the doctor’s office

→ Occasionally, the physician would be sent email alerts to encourage them to view the dashboard

→ There were two situations that would trigger the sending of such an email alert1. When a patient answered “yes” to any of a set of specific “Red zone” questions

This was intended to inform the physician that one of their patients had entered the red zone and may need immediate attention

2. When a fresh rolling ATAQ score had been calculated for a given child This was intended to remind the physician to review information on this

child

Page 21: Health Informatics- UMKC, Presentation

Physician’s Dashboard

→ If, for any reason a physician is concerned about the status of a patient when viewing his or her detailed information in the dashboard

→ Am email could be sent to that physician’s nurse directly from the dashboard

→ An option to customise the message should be provided to the physicians to indicate actions the nurse should take

Page 22: Health Informatics- UMKC, Presentation

Study Design

Page 23: Health Informatics- UMKC, Presentation

Hypotheses

H1 •Pediatric asthma patients answering regularly- administered questions about their asthma management via SMS will demonstrate better health outcomes than a control group of asthma patients, as measured both by a quality of life questionnaire and a pulmonary function test.

H2 •Pediatric patients answering regularly- administered questions about their asthma management via SMS and receiving regularly- administered information via SMS intended to increase their knowledge about asthma will demonstrate better health outcomes, as measured both by a quality of life questionnaire and a pulmonary function test, compared to those receiving only regularly- administered SMS questions and controls.

H3 •The rolling ATAQ score calculated by the answers of the SMS queries will provide valid assessments of pediatric asthma management on a more frequent basis when compared to the original ATAQ score.

Page 24: Health Informatics- UMKC, Presentation

Participants→ Conducted this study at a private practice in a large metropolitan city.

→ Physicians recruited: 11→ Children : 30→ Caregivers: 3

→ Children had to be ten years age or older who have their own phone and able to read at a 5th grade level.

→ In addition, children had to be regular patients who had met health care providers at least once before they participate in the study.

→ Children randomly assigned to one of three groups→ Control group→ Query group→ Query and knowledge group

The age and sex of participants

Page 25: Health Informatics- UMKC, Presentation

Method→ The duration of the study was the time between two scheduled visits to the doctor’s

office, typically 3-4 months.

→ Patients that were assigned to the intervention group received SMS messages every day(Query & knowledge group) or every other day(query group)

→ Physician participants could access the dashboard to review the status of their participating patients and received alerts via email when a patient was in the red zone.

Research tools used in each phase

Page 26: Health Informatics- UMKC, Presentation

Method - Phases

First doctor’s visit

•Consent form for caregivers•Need to fill out a set of surveys about technology usage and the PACQLQ (Pediatric Asthma Caregiver Quality of Life Questionnaire)•Physicians was scheduled for an introductory survey

During the study

•Our system logged SMS data transactions and the usage of the physician’s dashboard

Follow up visit

•After the follow visit, patients were administered a number of surveys including•Asthma knowledge test•PAQLQ•ATAQ• and pulmonary function test•Interviews

Page 27: Health Informatics- UMKC, Presentation

Results

Page 28: Health Informatics- UMKC, Presentation

Results

→ The section consists of three parts1. Characteristics of communication technology usage in pediatric asthma patients

and their caregivers2. Intervention results3. Utilization of our intervention

In all cases where we compare differences betweengroups, we will present both the p value and the effect size. Theeffect sizes, r, calculated based on Cohen's formula were interpreted according to Cohen’s guidelines:r < 0.2 (Small effect size)r = 0.5 (Medium effect size)r > 0.8 (Large effect size)

Thus, if we have medium or large effect size, the statistical outcomes are still valuable.

Page 29: Health Informatics- UMKC, Presentation

• Cohen's d is defined as the difference between two means divided by a standard deviation for the data

X1- mean of group 1X2- mean of group 3S- Pooled standard deviation(Both Groups).

Here S^2 indicates variance.

Page 30: Health Informatics- UMKC, Presentation

When test value “t” and degree of freedom “df” values are found then,

Cohen’s d and effect size “r” can be calculated as follows:

t(df)= some value.

t(8)=3.11

Here t= 3.11 and df= 8.

r = sqrt( ( t2 ) / ( ( t2 ) + ( df * 1) ) )

d = ( t*2 ) / ( sqrt(df) )

By using these formulae , we can calculate the “r” value and “d” value.

r = sqrt( (3.112 ) / ( (3.112 ) + ( 8 * 1) ) )

r = 0.74

d = ( 3.11*2 ) / ( sqrt(8) )

d = 2.20

Page 31: Health Informatics- UMKC, Presentation

Characteristics of Participants- Communication Technology Usage

Description Count

Total families participated 30

Internet access at home 29

Pediatric patients who accessed internet 6

Care givers who accessed internet 9

Families who used SMS 2

Description Count

Total families participated 30

Mobile service usage count 30

Families has data plan 1

Families had public insurance 6

Caregivers who had data plan 21

Children who have data plan 13

Internet

Mobile

Page 32: Health Informatics- UMKC, Presentation

Health Outcomes1. Baseline at the first doctor’s visit

→ Our result indicate that the three invention groups did not differ on any meaningful clinical or psychological characteristics

→ We did not find significant age, asthma knowledge, quality of life perception (PAQLQ, PACQLQ), or pulmonary function differences across all participants

→ In terms of the scores of asthma knowledge test, we found that age was significantly correlated with the scores

→ Caregiver’s quality of life score(PACQLQ) was significantly correlated with their child’s quality of life score

→ Caregiver’s quality of life score (mean: 5.4) was higher than pediatric patients quality of life score(mean: 4.97)

→ Thus, we replicated a result from the literature, which caregivers tend to overestimate quality of life when compared to the child's own perception

Page 33: Health Informatics- UMKC, Presentation

Health Outcomes2. Outcomes

→ Data for 15 pediatric asthma patients and 6 physicians.→ The progress of participants during the course as shown below.

Enrollment table of the pediatric participants through the study and the age, sex, insurance type of participants at the follow up visits.

Page 34: Health Informatics- UMKC, Presentation

Health Outcomes2. Outcomes

→ Clinical and psychological outcomes

To test hypothesis H2,Compare the pre-post differences of PAQLQ and FEF25-75% between the control and the other two groups.

Thus patients answering regularly- administered questions about their asthma symptoms and management did not demonstrate better health outcomes, as measured both by quality of life questionnaire and pulmonary function test, as compared to control. Hence, we reject hypothesis H1.

Page 35: Health Informatics- UMKC, Presentation

Thus, pediatric patients answering regularly- administered questions about both asthma management and information via SMS demonstrated better pulmonary function and increased their knowledge about asthma compared to those receiving only regularly- administered SMS questions or no SMS questions.

To test hypothesis H2, we compared the pre- post differences of PAQLQ and FEF25- 75% between the Query group and the Query& Knowledge group, and the Control group and Query & Knowledge group

PAQLQ improvement in the Query & Knowledge group was not statistically different from the querygroup or the control group.

Statistically significant differences were found for the FEF25- 75% in the Query & Knowledge group compared to the other two groups.Query group- t(8)=3.11, p=0.007, Cohen’s d= 2.20, effect-size r=0.74)Control group- t(9)=2.04, p=0.036, Cohen’s d= 1.36, effect-size r=0.56)Query and - t(13)=2.15, p=0.026, Cohen’s d= 1.19, effect-size r=0.51)Knowledge group

Page 36: Health Informatics- UMKC, Presentation

Utilization of the Intervention

The (original) ATAQ score and the rolling ATAQ score at the follow up visits.

The rolling ATAQ score calculated by the answers of the SMS queries provided valid assessments of pediatric asthma management on a more frequent basis when compared to the standard ATAQ score

Based on the statistics :Total scores in the original ATAQ at the follow up visits (r=0.91, p< 0.001)Total scores in the rolling ATAQ at the follow up visits (r= -0.89 , p< 0.001)

Page 37: Health Informatics- UMKC, Presentation

Utilization of the Intervention

Daily response rate

The total number of alerts sent via email to eachphysician and the number of logins by each physician

Page 38: Health Informatics- UMKC, Presentation

Discussion

→ We successfully deployed a new intervention for pediatric asthma management in an ecologically valid setting using a widely adopted mobile technology, SMS

→ The new intervention was found to have some positive effects on health outcomes and asthma knowledge of the patients

→ Patients readily adopted it and maintained a high rate of response throughout the intervention period

→ Thus, our system can provide the physicians constant assessment of their patients asthma between visits

→ Some of the physicians responded positively to the use of SMS and dashboard technology

→ In particular, they altered their practise such that they would access the dashboard when they received email alerting them that their patients has entered the red zone

Page 39: Health Informatics- UMKC, Presentation

Health Improvement OutcomesReceived partial support for the hypothesis.

Found that the pediatric asthma patients that both answered questions about their asthma status and received information about asthma showed improved health outcomes, as evidenced by lung function analyses compared to the query only and control groups.

Some participants expressed their words : 1. “It [knowledge] helps me become more aware of what causes asthma.” 2. “If I forgot my medicine in the morning, and I got the text, remembered what I had to do[take medicine] after I got the text.”

Found that pulmonary function was not correlated significantly with quality of life in patients with moderate to severe asthma. This correlation was only found in patients with mild asthma.

Since patients have different severities of asthma, it could have affected the results.

Page 40: Health Informatics- UMKC, Presentation

SMS as a Communication Tool in the Clinical Setting

Characteristic Previous study Our StudyResponse rate 69% >87%

Age of participants Median age 13 years Median age 38.5 years

Duration of study 2 months 3-4 months

Questions per day 4 1

The rolling ATAQ scores can provide time-sensitive and unique information about the child’s asthma symptoms during the span between two regularly scheduled visits and thus improve current clinical practice.

SMS for question and answer exchanges may be a cost-and labor-effective.

Page 41: Health Informatics- UMKC, Presentation

Impact on Clinical Practice

Improve communication between the patients and the caregivers(specifically physicians)

Physicians should present various terms in the context of SMS queries such that the doctor can confirm that the patients are aware of the distinctions between different types of medicines.

“One participant got confused about oral steroid as compared with ProAir [rescue bronchodilator medication.”

With the help of SMS system doctor can know the patient’s asthma status even though doctor did not contact the patient between visits.

One participant mentioned that “He(doctor) already knew what I did, so he talked about what I was supposed to do.”

SMS system indirectly mediates communication between patients and their parents. One participant mentioned “ I had to ask mom to tell me what it means… I didn’t understand the question… ‘have you taken controller medication everyday in the past four weeks?’”

Page 42: Health Informatics- UMKC, Presentation

Limitations

LIMITATION GOAL

Study with sample size (N=15)which is small. Replicate the study with a larger size.

Samples were mostly people with middle Income.

Study with more economically diverse Sample.

SMS questions sent to children must be appropriate to someone with a 5th grade level of literacy.

To maintain an appropriate terminology and make the SMS questions easily understood by the children.

Page 43: Health Informatics- UMKC, Presentation

Conclusion and Future Work

SMS is ubiquitous, easy to use and inexpensive.

SMS system can be built in such a way that healthcare providers could customize the database of text messages created for their patients , from this we can say that SMS usage can be embedded into the clinical practice.

Future systems may provide a channel for caregivers to understand their child’s needs. For example, such channel can be called as parent’s dashboard. This can help the parents to know what the child is doing in the SMS system.

Finally, study indicated that engaging a child’s social network may lead to better asthma management.

Page 44: Health Informatics- UMKC, Presentation

THANK YOU !!!

Page 45: Health Informatics- UMKC, Presentation

QUERIES !!!