RESEARCH ARTICLE

Retrospective examination of selected outcomes of MedicinesUse Review (MUR) services in New Zealand

Ernieda Hatah • June Tordoff • Stephen B. Duffull •

Claire Cameron • Rhiannon Braund

Received: 26 March 2013 / Accepted: 12 January 2014 / Published online: 15 March 2014

� Koninklijke Nederlandse Maatschappij ter bevordering der Pharmacie 2014

Abstract Background Poor adherence to medication can

lead to suboptimal outcomes, and is reported to occur

frequently. Pharmacists in some countries are funded to

support the appropriate use of medications in patients and

enhance medication adherence, by providing services such

as Medicines Use Review (MUR). Objective To describe

and investigate factors that may influence patients’

knowledge and perceptions of and adherence to medica-

tions as determined during MUR. Setting Community

pharmacies in a locality in New Zealand. Method Fol-

lowing consent from five MUR service providers, records

of patients’ MUR consultations conducted between

November 2007 and December 2011 were retrospectively

reviewed for information on patients, services, and out-

comes. Using multilevel mixed-effects logistic regression,

factors that predicted the providers’ score of the patients’

medication knowledge, and the patients’ score of their

adherence to and perceptions of medications, were inves-

tigated. Main outcome measure patients’ knowledge, per-

ceptions and adherence scores. Results A total of 353 MUR

patients’ records were evaluated. The median (IQR) age of

patients was 73 (63–81) years. About 41.1 % of patients

were Maori. A total of 204 (57.8 %) patients had two MUR

consultations and only 53 (15 %) had four. The mean score

of patients’ knowledge, perceptions of, and adherence to

medications were found to increase in each visit which

suggests that adherence support by pharmacists might

improve patients’ outcomes. Females had higher medica-

tion knowledge scores than males (OR 3.09, 95 % CI

1.29–7.44). There was some evidence to suggest Maori had

lower scores for knowledge of medications than non-Maori

(OR 0.092, 95 % CI 0.02–0.36). In addition, longer dura-

tion in the program predicted better scores for medication

knowledge, adherence and perceptions of medications.

Conclusions MUR was found to have the potential to

improve the scores of patients’ knowledge and perceptions

of and adherence to medicines, and factors such as gender,

ethnicity and longer duration in the services were found to

predict these outcomes.

Keywords Community pharmacy services � Medication

adherence � Medication knowledge � Medication review �New Zealand � Pharmaceutical care services � Pharmacists

Impact of findings on practice

• When providing the MUR adherence based services,

pharmacists should consider factors that may influence

patients’ outcomes such as gender, ethnicity and longer

engagement with the service.

• Where possible, the influence of gender, ethnicity and

longer engagement with the MUR service should be

discussed with patients, using specific counselling and

educational strategies.

E. Hatah (&) � J. Tordoff � S. B. Duffull � R. Braund

School of Pharmacy, University of Otago,

PO Box 56, Dunedin 9054, New Zealand

e-mail: erniedamdhatah@yahoo.com

E. Hatah

Faculty of Pharmacy, Universiti Kebangsaan Malaysia,

Kuala Lumpur, Malaysia

C. Cameron

Preventive and Social Medicine, Dunedin School of Medicine,

University of Otago, Dunedin, New Zealand

123

Int J Clin Pharm (2014) 36:503–512

DOI 10.1007/s11096-014-9913-1

Introduction

The prevalence of poor adherence to medications is high. It

is estimated that within developed countries, around 50 %

of patients with chronic diseases have poor adherence to

their medication treatment plan [1, 2]. Poor adherence to

medications can lead to negative outcomes on patients’

health such as suboptimal clinical benefit, hospitalisation

and death [3, 4]. It also carries a large economic burden to

healthcare systems [3]. In a number of countries pharma-

cists conduct medication reviews to decrease drug related

problems (DRPs) and to improve adherence to medications

[5, 6]. One example of this type of funded medication

review service to improve and support adherence to med-

ication is MUR [5, 6].

Medicines Use Review services in New Zealand are

defined as a structured and systematic review of all

patients’ current therapy [7]. MUR aims to improve the

patient’s understanding of their medication-related health

outcomes by identifying access, adherence, and day-to-day

management issues and setting goals with the patient to

resolve these issues [7]. The service involves patients

spending time with the pharmacist to discuss the best way

to manage and get the most out of their medication. It

comprises educational activities, counselling on current

medications and medical conditions to improve adherence,

removal of out of date medications, aligning dispensing of

medication, and reporting of medication adverse effects

[7]. Similar to the MUR service in the United Kingdom

(UK), MUR services in New Zealand usually take place

without pharmacists having access to clinical information.

MUR is founded on techniques to improve medication

administration techniques and adherence and it does not

include assessment of the appropriateness of medication

therapy [7, 8]. Patients can engage the services by self-

referral or referral from prescribers, pharmacists, primary

health care nurses or nurse practitioners [7]. In New Zea-

land, MURs can be conducted either at the pharmacy, at the

patient’s home or by telephone. MUR pharmacists are

funded based on a ‘‘fee-for-service’’ model by their District

Health Board (DHB) for up to four MURs per patient per

year. The fees vary between localities but they are usually

paid between $100 and $150 for three MUR consultations,

to $181–$200 for four MUR consultations per patient per

year [9]. In fee-for-service MUR, the services are carried

out by pharmacists in their usual settings where they may

have more limitations or barriers in implementing the

services to patients e.g. be subject to time, staffing and

resource constraints. This differs from a service being

provided in a highly research controlled environment. In

New Zealand, GPs are not reimbursed for being involved in

medication review services, as occurs in Home Medication

Review services in Australia.

Funding for this service in New Zealand commenced in

2007 [6]. During the first year of the service, MUR was

reported to have a low-uptake [9]. Of 20 DHBs, only five

(25 %) were funding their community pharmacists to

provide the service [9]. The service has steadily increased

in the past few years, with 11 DHBs in 2011 reported to be

funding their community pharmacists for MUR services

[10]. Common barriers to the uptake of MUR include; no

current contract with funders, a complex claim process,

insufficient time, and general practitioners and patients

lacking interest in the service [9].

While these services are available internationally there

are a limited number of papers relating to the outcomes of

the services, most have focused on patient perspectives. A

recent qualitative paper by Latif et al. [11] reported that

patients were found to be satisfied with the service despite

the majority believing that their knowledge about medi-

cations was not improved by the service. To our knowledge

no study, however, has evaluated the outcomes of MUR

services quantitatively in relation to its aims of improving

patients’ knowledge, perceptions of, and adherence to

medications. In addition, no study has investigated factors

that may influence patient outcomes in a fee-for-service

MUR.

Aim of the study

This study aims to: (1) describe the fee-for-service MUR

characteristics provided in a locality in New Zealand (2)

quantify the MUR service outcomes on patient’s knowl-

edge and perceptions of, and adherence to their medica-

tions and (3) investigate factors that may influence those

outcomes.

Ethical approval

Ethical approval for this study was granted by the New

Zealand Northern Regional Ethics Committee, reference

number NTX/11/EXP/182.

Method

The study was conducted retrospectively through exami-

nation of MUR patients’ consultation records documented

by pharmacists in one locality in New Zealand. The current

study design was chosen to capture the effect of fee-for-

service (funded service) MUR by community pharmacists.

Although randomised controlled trials (RCTs) are gener-

ally considered the best design for this type of study, they

are often difficult to carry out when the service that is being

504 Int J Clin Pharm (2014) 36:503–512

123

explored is part of current practice. In addition, the highly

structured setting of an RCT does not necessarily reflect

current practice. The locality was selected because a MUR

service had been in place since 2007. At the time of the

study, there were seven MUR providers in this locality.

Providers’ were located in two geographical areas. Six

providers were in the city area with a population of

approximately 80,000 and one in a rural area with a pop-

ulation of almost 9,000 [12]. All MUR providers were

invited to participate in the study and an invitation letter

and information sheet about the study was mailed to them.

Participation in the study was voluntary and providers who

agreed to participate were asked to return their consent

form to the researcher, EH. Providers were not involved in

collection, collation or analysis of the data.

Data collection was performed by EH at the providers’

workplace (pharmacy or office). All MUR consultation

records for each visit (e.g. visit 1 up to visit 4) were scanned

and relevant data were extracted from the forms. Inclusion

criteria were those records that had complete documenta-

tion and were within the dates of initially receiving funding

until December 2011. Data extracted consisted of patients’

age, gender, ethnicity and health issues (as documented by

the pharmacist with data usually gained from patients’

interviews), reasons for MUR, number of visits, information

about the medication used and DRPs identified (defined as a

circumstance related to a patient’s use of a drug that actu-

ally or potentially prevents the patient from gaining the

intended benefit of the drug). Since the results on the types

of DRPs found during MUR were worthy of a detailed

investigation and discussion, the authors have reported the

findings in a separate paper [13]. The providers’ score of the

patients’ knowledge, adherence behaviour and perceptions

of medications were also collected. The score, that rated by

pharmacists based on patients’ response, was allocated on a

scale of 1–4, and are presented in Table 1. Each item was

scored separately with the lowest score reflecting the

poorest performance (for example a ‘1’ for knowledge

represented no knowledge about medications). This scoring

system was developed by the service’s funders and was

implemented by the MUR providers. The use and imple-

mentation was not influenced by this study.

Data were analysed descriptively with Microsoft Excel

and inferentially using IBM SPSS statistics version 20 and

Stata, version 12 [14]. Multilevel models were used to

allow for correlated observations within pharmacy (indi-

viduals) and individual (multiple observations over time).

Within that framework, logistic regressions were used to

model a binary outcome for each of knowledge, perception

and adherence to medications. The score of 1 and 2 were

categorised as ‘poor achievement’ (e.g. poor knowledge,

poor adherence and poor perceptions about medications)

and score of 3 and 4 were categorised as ‘good

achievement’ (e.g. good knowledge, more adherent, and

more positive about medications).

The fixed effects included in the models were age,

gender, and ethnicity, number of medications e.g. short

term and long term medications and duration in the pro-

gram (up to 6 months and more than 6 months). Prior to

the construction of the models, univariate tests were used

to determine which variables might be important in the

modelling of each measure: age, gender, ethnicity, number

of reason for MUR, number of long and short-term medi-

cations use, number of co-morbidities, total number of

medications and duration in the service. Only variable with

p values \0.25 will be included in the multivariate model

(results of univariate analysis available in ‘‘Appendix 1’’).

Following the common practice in health research, vari-

ables such as age and gender were always included. In the

multivariate model, a p value of \0.05 was considered

statistically significant.

Results

Of the seven MUR providers, five agreed to participate in

the study. Four providers were community pharmacists and

one provided a local domiciliary pharmacist support ser-

vice. Four providers were in the city area and one provider

was located in a rural area, 146 km from the city.

A total of 353 MUR patients’ consultation records were

reviewed. Patients’ characteristics according to the service

providers are presented in Table 2. The median age (IQR) of

patients was 73 (63-81) years. There were 199 (56.4 %)

female participants in this study. A total of 131 of the par-

ticipants were Maori (n = 131, 41.1 %). The median num-

ber of comorbidities was two (range between 1 and 11) and

patients used a median of eight medications (range 2–19).

Table 1 Scoring system of patients’ medication knowledge, adher-

ence and perceptions

Type of score Question asked Score details

Knowledge Why were you on this

medication?

1-No knowledge

2-Some knowledge

3-Good knowledge

4-Superb knowledge

Adherence How often do you miss

a dose of this medication?

1-Always

2-Often

3-Seldom

4-Never

Perception How well do you think this

medication is working?

1-Not working at all

2-Some

3-Good

4-Great perception

Int J Clin Pharm (2014) 36:503–512 505

123

Between November 2007 and December 2012, a total of

709 MURs were conducted. All patients (n = 353) in this

study attended visit one, half of the patients (n = 204,

57.8 %) returned for visit two, 99 (28 %) continued to visit

three and 47 (13.3 %) patients had four visits. Reasons for

not returning for MUR or no rescheduled visit were not

recorded. Regardless of the number of visits, a total of 254

patients attended the service up to 6 months and 99 were

followed-up for more than 6 months (range from 6 to

41 months).

In visit one, the majority of MUR services were con-

ducted at the patients’ home (n = 185, 52.4 %) or at the

pharmacy (n = 126, 35.7 %). Others were conducted at the

GPs’ surgery (n = 21, 5.9 %) or in other settings (n = 5,

1.4 %) such as the patients’ workplace or a relatives’

home. The setting was not recorded for 16 patients (4.5 %).

During visit two, three and four, most follow-ups were

conducted at the pharmacy (n = 139, 39 %) or by tele-

phone (n = 122, 34.3 %).

The most common reasons for MUR, besides having at

least one chronic disease, were taking five or more medi-

cations (n = 319, 90.4 %), having DRPs (n = 147,

41.6 %), taking medication with high risk for adverse

effects (n = 127, 35.9 %) and having a recent hospital

admission (32.0 %). Although enrolment in the service did

not require referral, 53 patients (15 %) were referred to

pharmacists for the services. Of these, 37 were referred by

GPs. Other referrals were from nurses (n = 3), other

pharmacists (n = 5), or independent organisations such as

a Hospice or the Salvation Army (n = 8).

The mean scores for patients’ knowledge, perceptions of,

and adherence to medication for each visit (rated by phar-

macists based on patients’ response) is presented in Table 3.

When knowledge, adherence and perception scores of

patients who continued the services were compared to

patients who did not have a further visit (using Mann–

Whitney U test), no significant differences were found

except during the third visit where more patients with lower

adherence score did not return to the program, p \ 0.001

(see ‘‘Appendix 2’’). Table 4 provides the odds ratio (OR)

of potentials factors of being a good achievement (score 3 or

4) in medication knowledge, perceptions and adherence.

Based on a multivariate analysis, the findings indicated that

women were three times more likely to attain a good

achievement in knowledge of medication compared with

men (OR 3.09, 95 % CI 1.29–7.44). In addition, patients

with Maori ethnicity appeared to be less likely than non-

Maori to attain good achievement in knowledge of medi-

cation score (OR 0.092, 95 % CI 0.023–0.362). Patients

who engaged with the services longer were more likely to

Table 2 Summary of MUR patients’ characteristics

MUR service provider Total 1 2 3 4§ 5

Number of patients, n (%) 353 (100) 17 (4.8) 36 (10.2) 134 (38) 151 (42.8) 15 (4.2)

Age* 73 (14–97) 63 (35–85) 75 (30–89) 81 (18–97) 68.5 (14–93) 73 (51–89)

Age categories, n (%)

\40 10 (2.9) 1 (5.9) 2 (6.5) 1 (0.8) 6 (4.0) 0 (0)

40–64 87 (25.3) 11 (64.7) 8 (25.8) 11 (8.4) 51 (34.0) 6 (40.0)

65–74 88 (25.6) 3 (17.6) 5 (16.1) 17 (22.1) 61 (40.7) 2 (13.3)

[75 159 (46.2) 2 (11.8) 16 (51.6) 102 (77.9) 32 (21.3) 7 (46.7)

Gender, n (%)

Male 154 (43.6) 8 (47.1) 17 (47.2) 44 (32.8) 78 (51.7) 7 (46.7)

Female 199 (56.4) 9 (52.9) 19 (52.8) 90 (67.2) 73 (48.3) 8 (53.3)

Ethnicity, n (%)

Non-Maoria 188 (53.3) 15 (88.2) 21 (58.3) 111(82.8) 28 (18.5) 13 (86.7)

Maori 131 (37.1) 1 (5.9) 6 (16.7) 14 (10.4) 108 (71.5) 2 (13.3)

Not recorded/specified 34 (9.3) 1 (5.9) 9 (25) 9 (6.7) 15 (9.9) 0 (0)

Number of health issues*? 2 (1–11) 3 (1–11) 2 (1–5) 2 (1–5) 3 (1–7) 2 (1–4)

Number of medications*? 8 (2–19) 8 (2–14) 9 (4–17) 8.5 (3–17) 7 (2–16) 8 (4–13)

Long term medications 343 17 (5) 36 (10.5) 121 (35.3) 150 (43.7) 15 (4.4)

Short term medications 281 13 (4.6) 32 (11.4) 117 (41.6) 106 (37.7) 13 (4.6)

Non-prescription items 67 2 (3.0) 4 (6.0) 34 (50.7) 27 (40.3) 0 (0)

* Median (range)§ located at a district area, others are at city area? per/patienta with European descent, Pacific Islanders, Asian and others

506 Int J Clin Pharm (2014) 36:503–512

123

Ta

ble

3S

um

mar

yo

fp

atie

nts

’sc

ore

ov

erti

me

Su

mm

ary

of

sco

reK

no

wle

dg

eab

ou

tm

edic

atio

ns

(nu

mb

ero

fp

atie

nts

,%

)A

dh

eren

ceto

med

icat

ion

s(n

um

ber

of

pat

ien

ts,

%)

Per

cep

tio

ns

tom

edic

atio

ns

(nu

mb

ero

fp

atie

nts

,%

)

Vis

it1

Vis

it2

Vis

it3

Vis

it4

Vis

it1

Vis

it2

Vis

it3

Vis

it4

Vis

it1

Vis

it2

Vis

it3

Vis

it4

Ov

eral

l

To

tal

nu

mo

fp

atie

nts

35

32

04

99

47

35

32

04

99

47

35

32

04

99

47

Mea

nsc

ore

2.2

2.5

2.9

3.3

2.8

3.4

3.6

3.7

2.7

3.0

3.2

3.5

Sco

reC

31

27

(36

)9

5(4

6.6

)4

4(4

4.4

)2

8(5

9.6

)2

02

(57

.2)

15

0(7

3.5

)6

7(6

7.7

)2

5(5

3.2

)2

00

(56

.7)

13

8(6

7.6

)6

5(6

5.7

)2

6(5

5.3

)

Sco

re\

31

98

(56

)7

9(3

8.7

)2

1(2

1.2

)3

(6.4

)1

21

(34

.2)

27

(13

.2)

8(8

.1)

1(2

.1)

12

0(3

4)

32

(15

.7)

6(6

.1)

2(4

.3)

No

tre

cord

ed2

8(8

)3

0(1

4.7

)3

4(3

4.3

)1

6(3

4)

30

(8.5

)2

7(1

3.2

)2

4(2

4.2

)2

1(4

4.7

)3

3(9

.3)

34

(16

.7)

28

(28

.3)

19

(40

.4)

Mao

ri

To

tal

nu

mo

fp

atie

nts

13

16

81

42

13

16

81

42

13

16

81

42

Mea

nsc

ore

1.9

2.2

2.2

2.0

2.8

3.0

3.0

3.0

2.8

2.9

2.9

3.5

Sco

reC

31

5(1

1.5

)2

0(2

9.4

)4

(28

.6)

1(5

0)

51

(39

)5

1(7

5)

9(6

4.2

)1

(50

)9

5(7

2.5

)5

3(7

7.9

)9

(64

.2)

2(1

00

)

Sco

re\

31

14

(87

.0)

48

(70

.6)

10

(71

.4)

1(5

0)

78

(60

)1

7(2

5)

4(2

8.6

)0

(0)

32

(24

.4)

13

(19

.1)

2(1

4.3

)0

(0)

No

tre

cord

ed2

(1.5

)0

(0)

0(0

)0

(0)

2(1

.5)

0(0

)1

(25

)1

(50

)4

(3.1

)2

(2.9

)3

(21

.4)

0(0

)

No

n-M

aori

To

tal

nu

mo

fp

atie

nts

18

88

95

22

41

88

89

52

24

18

88

95

22

4

Mea

nsc

ore

2.4

2.8

3.0

3.3

2.8

3.6

3.7

3.7

2.6

2.9

3.2

3.4

Sco

reC

36

6(3

5.1

)5

6(6

2.9

)4

0(7

6.9

)2

3(9

5.8

)1

04

(55

.3)

80

(89

.9)

47

(90

.4)

21

(87

.5)

85

(45

.2)

69

(77

.5)

46

(88

.4)

22

(91

.6)

Sco

re\

31

02

(54

.3)

30

(33

.7)

10

(19

.2)

1(4

.2)

61

(32

.4)

9(1

0.1

)4

(7.7

)1

(4.2

)7

9(4

2.0

)1

5(1

6.9

)3

(5.8

)1

(4.2

)

No

tre

cord

ed2

0(1

0.6

)3

(3.4

)2

(3.8

)0

(0)

23

(12

.2)

0(0

)1

(1.9

)2

(8.3

)2

4(1

2.8

)5

(5.6

)3

(5.8

)1

(4.2

)

Int J Clin Pharm (2014) 36:503–512 507

123

Table 4 Summary of multivariate analysis of potential factors that predict the odds ratio of attaining a good achievement (score 3 or 4) in

knowledge and perceptions of, and adherence to medications

Odds Ratio SE z P [ |z| 95 % CI

Knowledge score (num of patients)

Age categories

\40 (n = 10) 1 – –

40–64 (n = 87) 0.09 0.16 -1.36 0.174 0.003 2.9

65–74 (n = 88) 0.11 0.197 -1.24 0.216 0.003 3.63

C75 (n = 159) 0.044 0.079 -1.74 0.081 0.001 1.48

Sex

Male (n = 154) 1 –

Female (n = 199) 3.09 1.38 2.52 0.012* 1.29 7.44

Ethnicity

Non-Maori (n = 188) 1 –

Maori (n = 131) 0.092 0.064 -3.42 0.001* 0.023 0.362

Short term medication (n = 281) 1.33 0.174 2.21 0.027* 1.033 1.723

Duration in the service

At the beginning (n = 353) 1 –

Up to 6 months (n = 251) 6.28 2.83 4.07 \0.001* 2.59 15.2

More than 6 months (n = 102) 20.1 13.3 4.53 \0.001* 5.49 73.5

Adherence score

Age categories:

\ 40 (n = 10) 1 –

40–64 (n = 87) 0.266 0.376 -0.94 0.349 0.017 4.26

65–74 (n = 88) 0.421 0.596 -0.61 0.541 0.026 6.74

C75 (n = 159) 0.285 0.397 -0.9 0.368 0.018 4.39

Sex

Male (n = 154) 1 –

Female (n = 199) 0.789 0.324 -0.58 0.564 0.353 1.76

Number of medications (n = 353) 1.09 0.076 1.2 0.231 0.948 1.25

Duration in the service

At the beginning (n = 353) 1 –

Up to 6 months (n = 251) 15.3 8.25 5.05 \0.001* 5.3 44

More than 6 months (n = 102) 13.4 8.26 4.21 \0.001* 4 44.9

Perceptions score

Age categories:

\40 (n = 10) 1 –

40–64 (n = 87) 0.216 0.422 -0.79 0.321 0.005 9.9

65–74 (n = 88) 0.575 1.13 -0.28 0.778 0.012 27.1

C75 (n = 159) 0.3 0.593 -0.61 0.543 0.006 14.5

Perceptions score cont.

Sex:

Male (n = 154) 1 –

Female (n = 199) 1.79 0.949 1.09 0.275 0.63 5.06

Long term medications (n = 343) 1.1 0.181 0.6 0.551 0.799 1.52

Number of medications (n = 353) 1.13 0.153 0.92 0.358 0.868 1.48

Number of co-morbidities (n = 353) 1.04 0.229 0.17 0.862 0.674 1.6

508 Int J Clin Pharm (2014) 36:503–512

123

attain good achievement in knowledge of medication score.

The odds of patients attaining good achievement in medi-

cation knowledge if engaged up to 6 months, are about six

times those at the start of the program (OR 6.28, 95 % CI

2.59–15.2), while it is about 20 times the odds at more than

six months (OR 20.1, 95 % CI 5.49–73.5) (see Table 4). A

good achievement in knowledge of medication was also

found to be predicted by the number of short term medi-

cations such as antibiotics or anti-diarrhoeal medicines

(OR1.33, 95 % CI 1.03–1.72).

Only length of follow-up period was predictive of

patients’ good achievement in medication adherence and

perceptions about medications (score C3) (see Table 4).

Patients who were in the program longer were more likely

to attain a good achievement in adherence score: followed-

up up to 6 months (OR 15.3, 95 % CI 5.30–44), followed-

up for more than 6 months (OR 13.4, 95 % CI 4.0–44.9).

The similar pattern was also seen with good achievement in

perceptions score: followed-up up to 6 months (OR 14.1,

95 % CI 4.42–44.8), followed-up more than 6 months (OR

33.2, 95 % CI 5.72–193). Age categories (1–39, 40–64,

65–74 and [75), gender, number of medications, number

of co-morbidities were not found to predict the scores

(p [ 0.05) (see Table 4).

Discussion

The current study provides insight into the characteristics of

fee-for-service MUR and its outcomes within the New

Zealand context. Commonly, MUR services in this study

were provided by pharmacists as face to face consultations

with patients either at the pharmacies or at patients’ homes.

The most common criteria for enrolment in MUR in this

locality were taking five or more medications, or having

DRPs. As expected (due to the criteria), the majority of

patients enrolled in MUR services were 65 years and over

(n = 247, 70 %). The current criteria for MUR in New

Zealand include patients with multiple medications and with

DRPs. Older people are more likely to meet these criteria

because they tend to consume many medications due to

chronic disease [15]. Furthermore, with age-related physio-

logical changes, older people are at increased risk of DRPs or

adverse drug events, and the consequent risk of hospitaliza-

tion [15]. Older people have been reported to commonly have

problems with day-to-day medication use such as difficulties

in remembering to take their medications, problems in

swallowing tablets/capsules or reading the labels or opening

the medication packets or bottles [16]. Since a systematic

review by Guaraldo et al., found that the prevalence of

inappropriate medication use in community-dwelling older

people is high (ranging from 11.5 to 62.5 %) [17], a MUR

service that does not include an assessment of the appropri-

ateness of medication therapy may not be the safest option for

patients. Instead, as suggested in our earlier publication,

funding a clinical medication review alongside a MUR may

be necessary to ensure patient safety [13]. The wide variation

of prevalence of inappropriate use of medication reported by

Guaraldo et al. [17] was thought to be caused by the diversity

of criteria defining inappropriate medication use.

From the results (see Table 3) the mean score of patients’

knowledge, perceptions of, and adherence to medications

was found to improve following MUR visits. This shows that

MUR could potentially improve patients’ outcomes. How-

ever, a RCT would be required to establish this. Although in

previous studies the number of co-morbidities were reported

to have a significant influence on medication adherence [18,

19], this was not observed in the current study. These find-

ings might be influenced by the quality of the information

given by the patients. Identification of the influence of co-

morbidities would be subject to confounding as an under-

standing about co-morbidities relied on the patient’s

understanding of their clinical picture and a poor under-

standing would result in limited knowledge of their co-

morbidities. Hence this study is not designed to determine

the influence of co-morbidities as an independent predictor.

In addition, previous studies also reported that the com-

plexity of patients’ medications may influence patients’

adherence [20, 21]. The effects, however, were not observed

in the current study. Since more than one third of the patients

received compliance packs that helped to simplify their

medication regimen, the true effects of medication com-

plexity would be difficult to estimate independently.

The multilevel mixed effects logistic regression model

analysis found that predictors of good achievement in

medication knowledge were being: (1) female (2) non-

Table 4 continued

Odds Ratio SE z P [ |z| 95 % CI

Duration in the service:

At the beginning (n = 353) 1 –

Up to 6 months (n = 251) 14.1 8.32 4.48 \0.001* 4.42 44.8

More than 6 months (n = 102) 33.2 29.9 3.9 \0.001* 5.72 193

* p \ 0.05

Int J Clin Pharm (2014) 36:503–512 509

123

Maori (3) having a longer follow-up duration or (4) on a

short-term medication. Although the number of short-term

medications was shown to positively predict the good

achievement in medication knowledge score, the impor-

tance of this finding may not be significant in the clinical

setting or for patients on medications for chronic condi-

tions. A specific reason for women having a better

achievement in knowledge about medications could not be

identified. However a study by Burge et al. [22] reported

better medication knowledge was demonstrated to correlate

with higher levels of patient’s satisfaction, education and

confidence that they can take medication as prescribed.

Some of these factors may have contributed to our finding

that women have better achievement in medication

knowledge score but this information was not available.

Despite the fact the women appeared to have higher

knowledge scores than men this did not translate into

improved medication adherence scores.

Also consistent with the findings by Burge et al., the

current study indicated a possible correlation between

ethnic background and medication knowledge [22]. In their

study non-Latino patients were found to have a better

knowledge of medications than Latino patients, and eth-

nicity and background education were significantly related

to knowledge [22]. Previous studies also reported that

medication knowledge and adherence could be influenced

by poor health literacy [23, 24]. An earlier study conducted

in New Zealand found that people with Maori ethnicity

have poorer health literacy skills than non-Maori [25]. It is

possible that this may have contributed to the lower med-

ication knowledge score in Maori seen in this study.

The length of engagement in the program was shown to

influence all three scores: patients’ knowledge and per-

ceptions of, and adherence to medications. The logistic

regression models showed that patients who remained

engaged for longer were more likely to attain good

achievement scores overtime. This finding appeared inde-

pendent of gender and ethnicity. Erickson et al. [26] in their

study perceived that increased contact with pharmacists, in

addition to usual care from doctors, contributed to increase

patients’ knowledge and awareness of the importance of

medications therapy such as adherence. Frequent contact

between a pharmacist and patient increase pharmacists’

opportunities to improve patients’ level of knowledge and

understanding about their medications and disease, and

allows patients to get more involved in their medication

therapy [27]. This should help improve patients’ adherence

to their mediation [28]. As knowledge and involvement

increase, patients’ anxiety about disease and medication

should decrease, and their confidence to self-manage and

monitor their disease conditions will improve [29].

Although the score of patients’ knowledge, perceptions,

and adherence to medications improved in each visit (see

Table 3), only 15 % of patients in this study had four MUR

consultations. The majority (57.8 %) of the patients had

only two visits. The reasons for not continuing the services

were unknown as the information was not recorded. It may

be important to examine the reason for not continuing the

service to understand more about the current provision of

fee-for-services MUR. Possible reasons for not continuing

could be related to pharmacy services such as an increased

workload, difficulty retaining staff, low profit margins; or

patient factors such as patients’ health had improved, they

no longer use the medications, patients were not satisfied

with the service, or personal reasons. Future studies should

be conducted to examine the possible cause for not

continuing the consultations.

There are some limitations to this study. Firstly, the data

were collected through retrospective review of patients’

MUR consultation records. The limited data available in the

records may have influenced the number of variables

available for analysis. A future study with a prospective

study design would allow more variables to be quantified.

Secondly, the assessments of patients’ knowledge, percep-

tions of and adherence to medications were subjective

assessments by pharmacists based on patients own impres-

sions and the accuracy of these is unknown. A limitation of

self-reported adherence includes that it is subject to recall

bias [30] and patients may tend to overestimate their level of

adherence. A more detailed evaluation that included objec-

tive assessments would have made the findings more robust.

A further limitation of the current study is its inability to

determine the influence of financial hardship on adherence,

as data were not collected on this during the consultations.

Finally, the findings on the influence of Maori ethnicity on

medication knowledge outcome may be confounded by

pharmacy as the majority of the Maori patients were

recruited from a single pharmacy (see Table 2). Since it is

not possible to disentangle the effect of ethnicity from the

pharmacy, generalization to a wider Maori population may

not be appropriate. Further study in a larger and more geo-

graphically wide-spread population of MUR patients is

required to identify patient groups with poor medication

knowledge. The findings would help healthcare providers to

develop strategies to assist any such populations.

Conclusion

The study provides insight into the characteristics of MUR

services and outcomes for patients. The adherence support by

pharmacists during MUR appeared to potentially increase the

score of patients’ knowledge, perceptions of and adherence

to their medications. Several factors such as ethnicity were

found to predict these attributes but further study is required

to confirm the findings in a larger population.

510 Int J Clin Pharm (2014) 36:503–512

123

Acknowledgments We would like to thank the MUR coordinator

and the community pharmacists who participated in this study and

helped us with the data collections.

Funding Funding for this study was provided by the School of

Pharmacy, University of Otago. EH was supported by scholarship

from Universiti Kebangsaan Malaysia and Ministry of Higher Edu-

cation Malaysia.

Conflicts of interest None to declare.

Appendix 1

See Table 5.

Appendix 2

See Table 6.

References

1. Ho P, Bryson CL, Rumsfeld JS. Medication adherence: its impor-

tance in cardiovascular outcomes. Circulation. 2009;119(23):

3028–35.

2. World Health Organization. Adherence to long-term therapies:

evidence for action, 2003. [cited 2013 Aug 24]. http://www.who.

int/chp/knowledge/publications/adherence_report/en/.

3. New England Health Institute. Thinking outside the pillbox: A

system-wide approach to improving patient medication adherence

for chronic disease, 2009. [cited 2013 Aug 24]. http://www.nehi.

net/publications/44/thinking_outside_the_pillbox_a_systemwide_

approach_to_improving_patient_medication_adherence_for_

chronic_disease.

4. Dunbar-Jacob J, Erlen JA, Schlenk EA, Ryan CM, Sereika SM,

Doswell WM. Adherence in chronic disease. Annu Rev Nurs Res.

2000;18:48–90.

Table 5 The results of univariate analysis of the variables that might

be important in the modelling of patients’ knowledge, perceptions of,

and adherence to medications

Coefficient SE p value

Knowledge score variable

Agea -0.322 0.013 0.015*

Gendera 0.391 0.410 0.341

Maori -2.056 0.739 0.005*

Num of reason for MUR 0.108 0.189 0.565

Num of long term meds -0.079 0.086 0.361

Number of short term medications 0.299 0.129 0.020*

Number of co-morbidities -0.154 0.162 0.339

Total number of medications 0.052 0.066 0.432

Duration in the service 0.305 0.073 \0.001*

Adherence score Variable

Agea -0.001 0.011 0.901

Gendera -0.155 0.295 0.600

Maori -0.174 0.300 0.561

Number of reason for MUR -0.164 0.141 0.246

Number of long term medications 0.112 0.065 0.083

Number of short term medication 0.107 0.094 0.257

Number of co-morbidities -0.113 0.113 0.319

Total number of medications 0.106 0.051 0.038*

Duration in the service 0.219 0.049 \0.001*

Perceptions score variable

Agea -0.019 0.016 0.252

Gendera 0.285 0.368 0.439

Maori 0.146 0.528 0.783

Number of reason for MUR 0.177 0.172 0.656

Number of long term medications 0.174 0.179 0.028*

Number of short term medications 0.098 0.123 0.428

Number of co-morbidities 0.193 0.156 0.215*

Total number of medications 0.153 0.065 0.018*

Duration in the service 0.558 0.122 \0.001*

A p values \ 0.25 (marked with a *) indicate variables that should be

included in the multivariate modela Even though the p values of variable age and gender were greater than

0.25, they were included in the model purely because it is a common

practice to include these variables in the health research models

Table 6 Comparison of scores between patients who continue the

program with who did not have a further visit

Score Group (number of patients) Mean

rank

p value*

Knowledge Return for visit 2 (n = 172) 157.33 0.244

Not returning to the program after

visit 1 (n = 149)

169.37

Return for visit 3 (n = 75) 95.46 0.059

Not returning to the program after

visit 2 (n = 97)

81.47

Return for visit 4 (n = 27) 42.17 0.183

Not returning to the program after

visit 3 (n = 48)

35.66

Adherence Return for visit 2 (n = 169) 166.6 0.275

Not returning to the program after

visit 1 (n = 149)

155.87

Return for visit 3 (n = 73) 105.34 0.0001*

Not returning to the program after

visit 2 (n = 96)

77.53

Return for visit 4 (n = 24) 39.25 0.552

Not returning to the program after

visit 3 (n = 49)

36.66

Perceptions Return for visit 2 (n = 165) 159.07 0.988

Not returning to the program after

visit 1 (n = 149)

158.92

Return for visit 3 (n = 69) 89.33 0.238

Not returning to the program after

visit 2 (n = 96)

82.01

Return for visit 4 (n = 24) 34.56 0.729

Not returning to the program after

visit 3 (n = 45)

35.99

* Mann–Whitney U test

Int J Clin Pharm (2014) 36:503–512 511

123

5. Clyne W, Blenkinsopp A and Seal R. A Guide to Medication

Review, 2008. [cited 2013 Aug 24]. http://www.npc.nhs.uk/

review_medicines/intro/resources/agtmr_web1.pdf.

6. District Health Board New Zealand. New Zealand National

Pharmacist Services Framework, 2007. [cited 2013 Aug 24].

http://www.psnz.org.nz/public/cop/documents/dhbnzpharmacist

servicesframework2007.pdf.

7. District Health Board Shared Services. New Zealand National

Pharmacist Services Framework, 2007. [cited 2013 Aug 24].

http://www.dhbsharedservices.health.nz/Site/SIG/NPSF/NPSF-

Framework.aspx.

8. Brandt T, Harrison J, Sheridan J, Shaw J, Jensen M. The evalu-

ation of auckland District Health Board’s medicines use review

pilot: the ADMiRE Report, 2009. [cited 2013 Aug 24]. http://

www.adhb.govt.nz/PHO/documents/ADMiRE%20report%20final.

pdf.

9. Lee E, Braund R, Tordoff J. Examining the first year of medicines

use review services provided by pharmacist in New Zealand. N Z

Med J. 2009;122(1293):26–35.

10. Cipolle R, Strand L, Morley P. Pharmaceutical care practice: the

patient centered approach to medication management. 3rd ed.

The United States of America: McGraw-Hill companies; 2012.

ISBN 978-0-07-179086-4.

11. Asam L, Boardman HF, Pollock K. Understanding the patient

perspective of the english community pharmacy medicines use

review (MUR). Res Soc Admin. 2013;9:949–57.

12. Statistics New Zealand, 2006 Census. [cited 2013 Aug 24]. http://

www.stats.govt.nz/Census/2006CensusHomePage.aspx.

13. Hatah E, Tordoff J, Duffull S, Braund R. Pharmacists’ perfor-

mance of clinical interventions during adherence support medi-

cation reviews. Res in Soc and Admin Pharm (article in press).

http://dx.doi.org/10.1016/j.sapharm.2013.04.008.

14. StataCorp. Stata Statistical Software: Release 12, StataCorp

LP2011: College Station, TX.

15. Barat I, Andreasen F, Damsgaard EM. The consumption of drugs

by 75-year-old individuals living in their own homes. Eur J Clin

Pharmacol. 2000;56(6–7):501–9.

16. Tordoff JM, Bagge ML, Gray AR, Campbell AJ, Norris PT.

Medicine-taking practices in community-dwelling people age-

d [ or = 75 years in New Zealand. Age Ageing. 2010;39(5):

574–80.

17. Guaraldo L, Cano FG, Damasceno GS, Rozenfeld S. Inappro-

priate medication use among the elderly: a systematic review of

administrative databases. BMC Geriatr. 2011;11(1):79.

18. Ahmad NS, Ramli A, Islahudin F, Paraidathathu T. Medication

adherence in patients with type 2 diabetes mellitus treated at

primary health clinics in Malaysia. Patient Prefer Adherence.

2013;7:525–30.

19. Krien SL, Heisler M, Piette JD, Makki F, Kerr EA. The effect of

chronic pain on diabetes patients’ self-management. Diabetes

Care. 2005;28(1):65–70.

20. Witticke D, Seidling H, Lohmann K, Send A, Haefeli W.

Opportunities to reduce medication complexity: a retrospective

analysis of patients discharged from a university hospital in

Germany. Drug Saf. 2013;36(1):31–41.

21. Lam PW, Lum CM, Leung MF. Drug non-adherence and asso-

ciated risk factors among Chinese geriatric patients in Hong

Kong. Hong Kong Med J. 2007;13(4):284–92.

22. Burge S, White D, Bajorek E, Bazaldua O, Trevino J, Albright T,

et al. Correlates of medication knowledge and adherence: find-

ings from the residency research network of South Texas. Fam

Med. 2005;37(10):712–8.

23. Kalichman SC, Ramachandran B, Catz S. Adherence to combi-

nation antiretroviral therapies in HIV patients of low health lit-

eracy. J Gen Intern Med. 1999;14(5):267–73.

24. Kalichman SC, Rompa D. Functional health literacy is associated

with health status and health-related knowledge in people living

with HIV-AIDS. J Acquir Immune Defic Syndr. 2000;25(4):

337–44.

25. Ministry of Health New Zealand. Health literacy and Maori

results from the 2006 adult literacy and life skills survey. [cited

2013 Aug 24]. http://www.health.govt.nz/publication/korero-

marama-health-literacy-and-maori-results-2006-adult-literacy-

and-life-skills-survey.

26. Erickson S, Slaughter R, Halapy H. Pharmacists’ ability to influence

outcomesofhypertension therapy.Pharmacother.1997;17(1):140–7.

27. Hawkins D, Bradberry JC, Cziraky MJ, Talbert RL, Bartels DW,

Cerveny JD. National pharmacy cardiovascular council treatment

guidelines for the management of Type 2 diabetes Mellitus:

toward better patient outcomes and new roles for pharmacists.

Pharmacother. 2002;22(4):436–44.

28. Al-Qazaz H, Sulaiman SA, Hassali MA, Shafie AA, Sundram S,

Al-Nuri R, et al. Diabetes knowledge, medication adherence and

glycemic control among patients with type 2 diabetes. Int J Clin

Pharm. 2011;33(6):1028–35.

29. Coulter A, Parsons S and Askham J. Policy brief: where are the

patients in decision-making about their own care?, 2008. [cited

2013 Aug 24]. http://www.who.int/management/general/deci

sionmaking/WhereArePatientsinDecisionMaking.pdf.

30. Fairman K and Motheral B. Evaluating medication adherence:

which measure is right for your program? J Managed Care

Pharmacy. 2000;6(6):499-504.

512 Int J Clin Pharm (2014) 36:503–512

123