Educational interventions aiming at improving adherence to treatment recommendations in type 2 diabetes

Review

Educational interventions aiming at improving adherenceto treatment recommendations in type 2 diabetesA sub-analysis of a systematic review of randomisedcontrolled trials

Johan Wens a,*, Etienne Vermeire a,b, Hilary Hearnshaw c, Antje Lindenmeyer c,Yves Biot a, Paul Van Royen a

aDepartment of General Practice, Interdisciplinary Health Care and Geriatrics, University of Antwerp, BelgiumbDepartment of Nursing and Midwifery, University of Antwerp, BelgiumcCentre for Primary Health Care Studies, University of Warwick, UK

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8

a r t i c l e i n f o

Article history:

Received 13 February 2007

Accepted 16 June 2007

Published on line 23 July 2007

Keywords:

Type 2 diabetes mellitus

Patient education

Adherence

Self-management

Systematic review

a b s t r a c t

Objective: In the management of type 2 diabetes, a complex interaction takes place between

medical professionals’ treatment goals and patients’ health beliefs about the disease and its

treatment options.

The contribution of self-management education to adherence in general or even more

specifically to medicine taking is not known. We assessed educational interventions aimed

at improving adherence to medical treatment recommendations, other than lifestyle advice.

Study design: Systematic literature review.

Setting: This paper represents an analysis of eight articles describing an educational inter-

vention as a subgroup of a Cochrane Review [E. Vermeire, J. Wens, P. Van Royen, Y. Biot, H.

Hearnshaw, A. Lindenmeyer, Interventions for improving adherence to treatment recom-

mendations in people with type 2 diabetes mellitus, Cochrane Database of Systematic

Reviews 2005, Issue 2, Art. No.: CD003638, doi:10.1002/14651858.CD003638.pub2] on inter-

ventions to improve adherence to treatment recommendations in people with type 2

diabetes.

Results: Four studies reported interventions using face-to-face education, two reported on

the effects of group education and two on distance education by telemedicine. Due to poor

quality of study designs, a variety of heterogeneous outcome measures in different time

intervals, unclear definitions of adherence, and difficulties in evaluating different aspects of

education performed, general conclusions could not be drawn.

Conclusion: Consistent conclusions about the effectiveness of educational interventions on

adherence to treatment recommendations were hard to be drawn. There is an urgent need

for well-designed intervention studies on the effect of different aspects of education on

adherence to treatment recommendations.

avai lab le at www.sc iencedi rec t .com

journal homepage: www.e lsev ier .com/ locate /d iabres

# 2007 Elsevier Ireland Ltd. All rights reserved.

* Corresponding author at: University of Antwerp, Faculty of Medicine, Department of General Practice, Interdisciplinary Health Care andGeriatrics, Universiteitsplein 1, 2610 Wilrijk, Antwerp, Belgium. Tel.: +32 3 820 25 29; fax: +32 3 820 25 26.

E-mail address: [email protected] (J. Wens).

0168-8227/$ – see front matter # 2007 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.diabres.2007.06.006

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8378

Contents

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378

2. Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378

2.1. Systematic literature search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378

2.2. Study selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379

2.3. Data collection, quality assessment and analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379

3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380

3.1. Methodological quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380

3.2. Kind of interventions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380

3.3. Quantitative synthesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386

4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386

5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387

Conflicts of interests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387

1. Introduction

The global burden of diabetes has been estimated at many

occasions [2,3]. WHO estimated 135 million people to be

affected in 1995, their number reaching 299 million by the year

2025 [4]. Over the last 30 years, type 2 diabetes has changed

from being considered a relatively mild ailment associated

with ageing, to one of the major contemporary causes of

premature mortality and morbidity worldwide [5]. Diabetes is

ranked among the leading causes of blindness, renal failure

and lower limb amputation [6], and is one of the major causes

of cardiovascular death.

To date there is strong evidence that the vigorous

treatment of type 2 diabetes can decrease morbidity and

mortality [7–10]. However, treatment outcomes in many

people with diabetes remain unsatisfactory. Since these

outcomes may be improved by a number of – usually complex

– interventions, full benefits of medication cannot be realized

at currently achievable levels of adherence [11]. Non-adher-

ence with therapeutic regimens therefore has been a

continuing problem for health care providers and a major

concern for public health [12,13].

Rather than the health care provider, the person living with

the disease seems to be the key to successful management

[14]. In addition, interventions focusing on modifying the

behaviour of people living with diabetes seem to have better

outcomes than those focusing on modifying clinician’s

behaviour [15].

Earlier quantitative research on the use of formal diabetes

education as a therapeutic tool, to promote self-management

in adults with type 2 diabetes, did not produce consistent

results in terms of improved metabolic control and of reduced

risks for cardiovascular disease. Positive effects on patient

knowledge, self-care behaviour, metabolic control, and psy-

chological outcomes were reported in reviews on the effec-

tiveness of education in type 2 diabetes [16–19]. However, in

these reviews, a complete and adequate description of the

interventions often is lacking. In general, comprehensive

interventions combining cognitive, behavioural, and educa-

tional components are more effective than single-focus

interventions [20]. Self-management education in patients

with type 2 diabetes mellitus, especially interventions invol-

ving active participation, may have short-term effectiveness

for improving diabetes knowledge, self-monitoring of blood

glucose levels, self-reported dietary habits, and glycemic

control [21]. However, the long-term outcomes of diabetes

education, aiming at the promotion of self-management and

adherence in adults with type 2 diabetes still have to be

established.

Diabetes self-management appears to be among the most

difficult of all chronic illnesses’ self-management regimes

[22,23]. Where quantitative studies failed to, the findings of

qualitative studies contributed to a deeper understanding of

the context in which diabetes education should be applied, in

order to be successful in promoting self-management beha-

viours in adults with type 2 diabetes [24]. By setting self-care

goals themselves, people may be more likely to adhere to

treatment plans [14,25–27]. Therapeutic patient education

may contribute to the adherence process by providing people

with the information they need for priority setting and

problem solving, assisting them in identifying realistic targets

for behavioural changes, and providing ongoing emotional

support and encouragement in their self-management plans.

We performed a systematic review of published articles to

identify the effectiveness of interventions aimed at improving

adherence to medical treatment recommendations in people

with type 2 diabetes mellitus, not lifestyle changes, diet nor

exercise [1]. This paper reports a synthesis of studies of

educational interventions that are a subgroup of studies

included in the Cochrane Review.

2. Materials and methods

2.1. Systematic literature search

A systematic literature search [28] was conducted to identify

interventions containing information on improving adher-

ence to medical treatment recommendations, not oriented to

other aspects such as exercise, attendance at clinic nor diet.

Search terms consisted of the Cochrane Metabolic and

Endocrine Disorders Group search strategy for type 2 diabetes

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8 379

mellitus and a combination of MESH terms: patient compli-

ance, health behaviour, health education, self-care, patient

education, patient satisfaction, educational status, patient

dropouts, physician–patient relations and delivery of health

care, all with subheadings and categories included. In

addition, truncations of compliance, concordance and adher-

ence were used besides word searches on self-care or self-

management. The full search strategy is described in the

original Cochrane publication [1].

We searched electronically the Cochrane Library including

the Cochrane Controlled Trials Register (CCTR) and the

Database of reviews of effectiveness (DARE), the Metabolic

and Endocrine Disorders Group Specialized Register, MEDLINE

(1966 to November 2002), EMBASE (1980 to November 2002),

Psych Info (1987 to November 2002), CINAHL (1982 to November

2002), Dissertation Abstracts (1961 to November 2002), ERIC

(to November 2002), and furthermore Sociological Abstracts,

Digital Dissertations Proquest, the National Library of Medicine

Gateway and Current Controlled Trials. Besides these databases

Internet search engines were used exhaustively. The reference

lists of relevant articles were investigated too.

2.2. Study selection (see Plate 1)

Studies in primary care, outpatient, community settings and

hospital settings were included. Randomized and quasi-

randomized controlled trials, controlled before and after

studies, observational studies and cohort studies were

appropriate for inclusion. Two researchers (E.V. and J.W.)

independently selected articles for review based on abstracts.

Plate 1 – Progress of th

Full articles were retrieved for further assessment if the

information given suggested that the study:

1. i

e s

ncluded patients with type 2 diabetes mellitus,

2. a

nor to diet,

3. m

enhance adherence,

4. a

5. u

design.

For the purpose of this study, adherence was defined in a

broader way than WHO that considers adherence as the extent

to which a person’s behaviour corresponds with agreed

recommendations from a health care provider [29]. Also

articles mentioning compliance, concordance or related

aspects in title or abstract were included. For the purpose of

this analysis the included papers were re-examined (Plate 1).

2.3. Data collection, quality assessment and analysis

Two teams of reviewers (E.V. and J.W., P.V.R. and Y.B.)

independently assessed all trials identified for inclusion.

Three teams of two reviewers (H.H. and A.L. in addition to

the two previous teams), assessed trial quality and extracted

the data required. Researchers were not blinded about

information on authors and journals. The quality of an

individual trial was assessed by scoring a list of nine topics:

randomisation, concealment of allocation, patient blinding,

tudy selection.

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8380

blinding of administrator of treatment, blinding of outcome

assessment, description of losses to follow-up/withdrawals,

intention to treat analysis, similarity of groups at the start of

the study and groups equally provided of care. Quality

assessment was done individually by every individual

researcher by scoring on these questions where answers

were compared afterwards. Overall quality was graded as A

(good), B (medium) or C (poor) by mutual agreement.

Whenever reviewers disagreed, consensus was obtained

through discussion or by a referee’s decision (PVR).

The extracted data were entered in a structured Excel sheet

adapted from the Cochrane Metabolic and Endocrine Disorder

Group data extraction sheet. Studies with comparable inter-

ventions were grouped and the results were synthesised in a

narrative way. HbA1c was considered a common outcome

measure which allowed performing a meta-analysis to

calculate pooled weighted mean differences (WMD). Because

of clinical heterogeneity a random effects model was used. For

this purpose the Cochrane review manager (RevMan) software

package version 4.2 was used.

3. Results

This paper summarizes the results of eight educational

interventions aimed at improving adherence to medication.

A total of 4387 references were identified of which only 74

studies met selection criteria and were assessed in full text of

which 21 articles were include in the review. The selection

process was done by two teams or reviewers assessing

studies independently. The inter observer agreement of the

first team (P.V.R. and Y.B.) expressed as a kappa was 0.526

(S.E. 0.085), that of the second team (E.V. and J.W.) was 0.827

(S.E. 0.032).

The studies in this sub-analysis contain data of 772

patients. A summary of the study characteristics, methodo-

logical quality and outcome is shown in Table 1.

Of the included studies six were randomized controlled

trials (RCTs) (ordered alphabetically Jaber [30], Krier [31],

Mease [32], Piette [33], Rachmani [34], White [35] and two

studies were studies controlling before and after (CBAs)

(Coast-Senior [36] and Jiang [37]).

3.1. Methodological quality

The overall quality of the included studies was medium to low.

The most frequent shortcomings were found in the conceal-

ment of allocation and the blinding procedures. Randomiza-

tion was adequate in all six RCTs while concealment of

allocation and blinding procedures were estimated as ade-

quate in only one study [33]. In seven trials, groups were

similar at the start of the study, in one study [34] data were

missing. Intervention and control groups were equally

provided of care in four studies [31,32,34,35], in one study

[30] this was not the case and data were missing or unclear in

three other studies [32,36,37]. Two studies did not describe

losses to follow-up [34,37], six adequately described these

losses. Only in four studies, it was clear that the authors

performed an intention-to-treat analysis [32–35].

It is striking that not one author mentioned details about

the calculation of the statistical power of the study. Moreover,

not one of the studies included assessed any economical

aspect of the interventions.

3.2. Kind of interventions

The different educational intervention studies were grouped

by the kind of intervention: (1) face-to-face education, (2)

group education and (3) telemedicine education.

(1) Face-to-face education

Two studies [30,36] assessed the effect of pharmacist-

led interventions (pill count, Micro-Electro-Mechanical

Systems (MEMS), comprehensive care, treatment adjust-

ments or prescription refill reminders) on self-reported

adherence, medication prescription refill, and metabolic

parameters. The latter are indirect measurements of

adherence. One paper [30] studied the impact of a

pharmacist-delivered comprehensive care model on

under-served urban African–Americans. Patients in the

intervention group were offered diabetes education,

medication counselling and evaluation, plus adjustment

of their hypoglycaemic regimen. A significant improve-

ment of HbA1c levels, from 11.5 � 2.9% to 9.2 � 2.1%

(p = 0.003), was obtained in the intervention group. Renal

function parameters, fasting blood glucose, blood pressure

and lipids were also outcome measures, but it remains

unclear, based on the published data, whether differences

were found between intervention and control groups. More

hypoglycaemic episodes, reported as mild to moderate,

occurred in the intervention group than the control group

(17 versus 2, p = not mentioned). Health status and Quality

of Life were not significantly improved in the intervention

group. The second paper [36] reported the outcomes of a

program of direct pharmacist involvement in initiating

insulin treatment in the care of older veterans. A change

from 11.1 � 1.6% to 8.9 � 1.4% (p = 0.00004) in glycated

haemoglobin was reported after 12 months in the inter-

vention group of 23 veterans.

The effect of quarterly visits of a diabetes educator was

assessed versus usual care after 3, 6 and 9 months on

HbA1c, different metabolic parameters, blood pressure and

self-reported compliance with diet and medication [31]. A

multivariate analysis using repeated measures technique

showed no differences on HBA1c, on weight and adherence

values using the initial value as a covariate. The compar-

isons over time and between groups showed no significant

differences. Barriers were the small sample population

(n = 39) and the poor adherence among both groups due to

poor motivation and to lack of financial support to afford

their medications.

The effect of standard consultations was compared

with a patient participating program in which patients

shared therapeutic responsibilities with a follow-up of 48

months [34]. HbA1c, blood pressure, LDL and the number of

cardiovascular events (total mortality, cerebrovascular

mortality, non-fatal vascular events, non-fatal acute

myocardial infarction, non-fatal stroke and coronary

Table 1 – Overwiew of study results

Author Subheading Descriptor

Coast-Senior [36] Aims To determine the impact of clinical pharmacists, involved in direct patient

care, on the management of patients with type 2 diabetes who require insulin

Methods Patients were referred to clinical pharmacist for management of diabetes

Controlled before and after study

No blinding

No concealment of allocation

Adequate description of losses to follow-up

Groups similar at the start of the study

Single centre study in an urban area in the USA

Participants Twenty-three type 2 diabetes patients

Patients recruited in primary care services

Sex: male 23

Mean age: 65 years, S.D. 9.4

HbA1c: 11.1%, S.D. 1.6

Fasting blood glucose: 219 mg/dl, S.D. 45 mg/dl

Random blood glucose: 236 mg/dl, S.D. 72 mg/dl

Duration since diagnosis: 8.8 years, S.D. 4.2

Treatment modality: diet + oral hypoglycaemic agents + insulin: 23

Interventions Interview

Initiating insulin

Adjusting insulin

Follow-up by appointment or phone call

Outcomes HbA1c

Fasting blood glucose (FBG),

random blood glucose

Notes Participants were veterans with multiple co-morbidities (mean 4, range 2–8)

and on multiple medications (mean 5, range 2–10)

Quality Medium

Jaber [30] Aims To determine the impact of a comprehensive pharmaceutical care model

(pharmacist provided care) versus standard care on treatment outcomes

of type 2 diabetes urban African-Americans

Methods Randomised controlled trial

Adequate randomisation

No blinding of patients and administrators, unclear blinding of outcome assessment

Concealment of allocation: unclear

Losses-to-follow-up adequately described

No intention-to-treat analysis

Groups were similar at the start of the study

Groups were not equally provided of care

Single centre study in a city in the USA

Participants Outpatients

Thirty-nine type 2 diabetes patients (intervention 17, control 22)

Sex: intervention (female 12, male 5), control (female 15, male 7)

Mean age: intervention (59 years, S.D. 12), control (65 years, S.D. 12)

Ethnicity: intervention and control groups: 100% African-American

Mean weight: intervention (93 kg, S.D. 22), control (88 kg, S.D. 19)

BMI: intervention (34, S.D. 7), control (33, S.D. 7)

BMI > 27: intervention (13), control (18)

Systolic blood pressure: intervention (140 mmHg, S.D. 20), control (143, S.D. 23)

Diastolic blood pressure: intervention (82, S.D. 10), control (88, S.D. 9)

Hypertension: intervention (14), control (17)

Lipid abnormalities: intervention (10), control (11)

HbA1c: intervention (11.5%, S.D. 2.9), control (12.2%, S.D. 3.5)

Fasting blood glucose: intervention (11.1 mmol/l, S.D. 4), control (12.7 mmol/l, S.D. 4.7)

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8 381

Table 1 (Continued )Author Subheading Descriptor

Duration since diagnosis: intervention (6.8 years, S.D. 6.5), control (6.2 years, S.D. 4.8)

Treatment modality: diet + oral hypoglycaemic agents: intervention (17), control (22)

Exclusion criteria: type 1 diabetes, renal dysfunction, history of non-attendance

Interventions Evaluation and adjustment of doses

Education regarding diabetes and its complications

Training on recognition of hypoglycaemia and hyperglycaemia

Instructions on diet and exercise

Training in self-monitoring of blood glucose

Treatment was titrated to targets

Outcomes Blood glucose control

Self-recorded adherence

Self-monitoring blood glucose logs

Notes Self-monitoring mentioned, but not specified and no data were presented

Quality Medium

Jiang [37] Aims To assess diabetes self-care after a five-section diabetes education program

including basic knowledge, dietary control, blood glucose monitoring,

management of hypoglycaemia, medication compliance, foot care and exercise

Methods Controlled before and after study

Data are missing on patient, administrator of treatment, and outcome

assessment blinding

Concealment of allocation: not used

Description of losses to follow-up: missing

No intention to treat analysis

Similarity of groups at the start of the study: not mentioned

Groups equally provided of care: data missing

Multicentre study in urban areas in Taiwan.

Participants Two hundred seventeen type 2 diabetes patients in an outpatient setting

Inclusion criteria: aged between 35 and 70 years old, able to read, HbA1c

level �8.0% and with stable metabolic control

Sex: intervention (female 61, male 69), control (female 49, male 38)

Age: intervention (52.3 � 6.7), control (52.9 � 7.4)

Body mass index: intervention (25.2 � 3.5), control (25.6 � 3.2)

Duration of disease since diagnosis: intervention (8.0 years � 8.0),

control 6.7 years � 5.3)

Treatment: diet and oral hypoglycaemic agents: intervention 121, control 87

Interventions A total of 121 attended four to five sections and were called the intervention

group, the 87 who only received the basic section were called the control group

Outcomes Fasting plasma glucose

HbA1c

Total cholesterol

Triglycerides

Systolic and diastolic blood pressure

Body weight

Waist–hip ratio

Notes

Quality Poor

Krier [31] Aims To evaluate the effect of quarterly visits of a diabetes educator

Methods Randomised controlled trial

Randomisation adequate

Concealment of allocation: unclear

Patients are blinded, but not the administrator of treatment,

Blinding of outcome assessment: data missing

Losses to follow-up described

Intention-to-treat analysis: unclear

Groups were similar at the start of the study

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8382

Table 1 (Continued )Author Subheading Descriptor

Groups were equally provided of care

Single centre study in a city in the USA

Participants Thirty-nine type 2 diabetes patients

Twenty-one people were allocated to the intervention group and 18 to the control group

Sex: intervention (female 9, male 5), control (female 10, male 4)

Age: intervention (54.2 � S.D. 5.6), control (56.2 � S.D. 9)

Ethnicity: intervention (six Afro-Americans), control (seven Afro-Americans)

Smoking: intervention (3), control (3)

Weight: intervention (97.4 kg � 23.9), control (93.8 � 19.4)

Locus of control: intervention (4.5 � 1), control (4.0 � 1.1)

Knowledge test: intervention (11.3 � 3.8), control (14.2 � 3.6)

HbA1c: intervention (9.6% � 1.9), control (10.3% � 2.3)

Treatment modality: oral hypoglycaemic agents (intervention 8, control 7),

insulin (intervention 5, control 5), combination (intervention 1, control 2)

Interventions Quarterly visits of a diabetes educator versus usual care

Outcomes HbA1c, four-point Lickert scale on compliance with diet and medication

Notes

Quality Medium

Mease [32] Aims To determine whether telemedicine can improve self-care for type 2 diabetes

Methods Randomised controlled trial

Randomisation: adequate

Blinding of patients, administrators or outcome assessors: unclear

Concealment of allocation: unclear

Adequate description of losses to follow-up

Intention-to-treat analysis

Groups were similar at the start of the study and were equally provided of care

Single centre study in USA

Participants Twenty-eight type 2 diabetes patients with HbA1c > 8.0% recruited in primary care

services

Interventions Education classes and weekly telemedicine ‘visits’ versus education only

Outcomes HbA1c

Weight

Notes Data on adherence are not available, although the intervention included ‘reinforcement

of medication compliance’

Quality Medium

Piette [33] Aims To evaluate automated telephone disease management with telephone nurse follow-up

as a

strategy for improving diabetes treatment processes and outcomes

Methods Randomised controlled trial

Randomisation and concealment of allocation: adequate

Blinding of patients and administrators: adequate

Blinding of outcome assessment: adequate

Description of losses to follow-up and intention-to-treat analysis

Groups were similar at the start of the study

No information on control group care

Participants Type 2 diabetes patients recruited in an outpatient setting in a mixed rural and urban

area

Participants: intervention (132), control (140)

Sex: intervention (male 126), control (male 138)

Mean age: intervention (60 years, S.D. 10), control (61 years, S.D. 10)

Ethnicity: intervention (black 32, hispanic 18, other 11), control (black 17, hispanic 16,

other 15)

BMI: intervention (31, S.D. 7), control (31, S.D. 6)

HbA1c: intervention (8.2%, S.D. 1.7), control (8.1, S.D. 1.7)

Serum glucose: intervention (188 mg/dl, S.D. 94), control (168 mg/dl, S.D. 68)

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8 383

Table 1 (Continued )Author Subheading Descriptor

All complications: intervention (1), control (0.7)

Vascular symptoms: intervention (0.7, S.D. 0.8), control (0.6, S.D. 0.8)

Treatment modality: insulin: intervention (39), control (31)

Interventions Telephone nurse follow-up versus usual care

Outcomes All diabetes related symptoms

Hypo- and hyperglycaemic symptoms

Vascular symptoms

HbA1c

Serum glucose

Self-monitoring frequency

Foot inspection

Weight monitoring

Podiatry visits

Ophthalmology visits

Diabetic clinic visits

Cholesterol tests

Notes Patients were veterans aged younger than 75, with a life expectancy of more than

12 months having a touch tone telephone a home. Newly diagnosed type 2

diabetes were not included

Quality Medium

Rachmani [34] Aims To examine whether sharing therapeutic responsibility with patients will improve

outcome

Methods Randomised controlled trial

Randomisation adequate

Concealment of allocation: unclear

Blinding of patients or administrators of treatment: data missing

Blinding of outcome assessors: unclear

Losses to follow-up: data missing

Intention-to-treat analysis

Groups were similar at the start of the study and were equally provided of care

Single centre study in Israel

It is not clear whether patients were recruited in rural or urban areas

Participants One hundred forty-two type 2 diabetes patients: intervention (71), control (70)

Sex: intervention (female 35, male 36), control (female 37, male: 33)

Mean age: intervention (57.4 years), control (56.8 years)

BMI: intervention (28.4), control (28.7)

Systolic blood pressure: intervention (162 mmHg, S.D. 7.3), control (160 mmHg, S.D. 6.9)

Diastolic blood pressure: intervention (96 mmHg, S.D. 2.4), control (95 mmHg, S.D. 2.0)

High density lipoproteins: (intervention 38 mg/dL, S.D. 3), control (39 mg/dL, S.D. 4)

Trigycerides: intervention (236 mg/100 ml, S.D. 42), control (243 mg/100 ml, S.D. 36)

Low density lipoproteins: intervention (146 mg/dL, SD 10), control (146 mg/dL, SD 9)

HbA1c: intervention (9.5%, S.D. 1.6), control (9.6%, S.D. 1.9)

Retinopathy: intervention (11), control (10)

Treatment modality: diet alone (intervention 18, control 16), diet + oral hypoglycaemic

agent (intervention 43, control 46), diet + insulin (intervention 10, control 8)

Interventions Involving patients in the process of their management by providing them with tools to

monitor and supervise the effects of therapy in comparison to a standard consultation

Outcomes Total mortality

Non-fatal vascular events

Cerebrovascular death

Non-fatal acute myocardial infarction

Non-fatal stroke

Coronary artery bypass surgery

HbA1c, BMI

Albumin/creatinine ratio

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8384

Table 1 (Continued )Author Subheading Descriptor

Blood pressure

Retinopathy

Glomerular filtration rate

Notes

Quality Medium

White [35] Aims To compare the effect of group management versus an advice-educational technique

Methods Randomised controlled trial

Randomisation: adequate

Concealment of allocation: inadequate

Blinding of patient, administrator of treatment, and of outcome assessment: data

missing

Losses to follow-up clearly described

Intention-to-treat analysis

Groups were similar at the start of the study

Groups were equally provided of care

Participants Thirty-two type 2 diabetes outpatients

Inclusion criteria: less than satisfactory control (FBG > 140 mg/dl), infrequent

hypoglycaemic reactions (<1/mol), no history of ketoacidosis, body weight >15%

above the mean value for height, no history of alcohol abuses or severe personality

disorder, and no current use of glucocorticoids

Sex: men only

Age: intervention (62.4 years � 5.5), control (60.7 years � 6.4)

Percentage of weight excess: intervention (36.3% � 21.0), control (44.3% � 21.0)

Duration since diagnosis: intervention (10.2 years � 12.9), control (13.6 years � 9.6)

Interventions Group management versus an advice-educational technique

Outcomes HbA1c

Notes

Quality Poor

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8 385

artery bypass surgery) were evaluated in both groups.

HbA1c improved by from 9.5 � 1.6% before to 8.2 � 1.5%

after the patient participation program (p < 0.05 versus

standard care group).

(2) G

Group education was compared with a usually used

management technique, based on advice and education

over 6 months in 32 veterans [35]. The advice-education

control group format was designed to minimize subject

interaction by having a nurse and a dietician give lectures

on the disease and is management. The experimental

management group was divided in three smaller groups of

less than eight members which met separately with a

clinical psychologist experienced in small group manage-

ment. An endocrinologist served as a consultant on call for

the psychologist during and between experimental group

sessions. Subjects in the experimental group were encour-

aged to interact and assessed their own and their peers’

progress towards managing their diabetes by sharing

ideas, advice, and support. There was a 10% decline in

HbA1c levels (p < 0.05) during the first 3 months of the

study in both groups, The effect, however, did not persist

until the end of the study.

The effect of a five-section education program inter-

vention over 4 months was compared with a basic course

(control) by measuring of fasting blood glucose, HbA1c,

serum cholesterol, triglycerides, blood pressure and body

weight [37]. The educational program covered the defini-

tion of diabetes mellitus, dietary control and meal plan,

urine and blood glucose testing, action of insulin and oral

hypoglycaemic agents, symptoms and treatment of hypo-

glycaemia and uncontrolled hyperglycaemia, foot care and

exercise. Those who attended only Section 1, the basics of

diabetes mellitus, were considered the control group.

Those who attended four to five were designed as the

experimental group. In both the experimental and the

control groups, the decline in HbA1c levels was statistically

significant (from 9.4 � 1.2% to 8.7 � 1.4% p < 0.001, versus

from 9.3 � 1.4% to 9.0 � 1.5%, p = 0.008). In the experi-

mental group, the decline in fasting plasma glucose, total

cholesterol, systolic blood pressure, body weight and

waist–hip ratio was also statistically significant. The effect

on drug adherence, though assessed by a questionnaire,

was not reported separately.

(3) T

The comparison of the effectiveness of education

classes plus a weekly nurse telemedicine ‘home visit’

versus usual care showed a statistically significant reduc-

tion in mean HbA1c level of 14% (from 9.5 (8.1–12.6)% to 8.2

(5.7–10.2)%, p < 0.05) over a period of 3 months [32]. The

mean weight reduction was limited and there were no

significant changes on a Diabetes Quality of Life scale or on

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8386

the Medical Outcome Health Survey SF-36 scale. Some

metabolic parameters such as micro-albuminuria, serum

creatinin and serum lipids did not improve during the

study period. Physicians and case managers considered

telemedicine to have a high benefit, but technological

problems were a major obstacle.

A nurse-led automated telephone management inter-

vention, with weekly follow-up nurse calls about self-care,

medication adherence and symptoms, showed a small but

statistically significant (p = 0.04) lowering of glycated

haemoglobin for patients with an initial level of �8%(33).

Patients in the intervention group also reported more

frequent glucose self-monitoring and foot inspections and

a 10% reduction in diabetes related symptoms. However,

there was no significant lowering of glycated haemoglobin

levels for the entire intervention group. Finally there were

no differences between groups in the proportion of

patients reporting one or more medication adherence

problems.

3.3. Quantitative synthesis

Studies assessing educational interventions ranging from

offering education individually [30,31,33,36] over group educa-

tion [35,37] to education on distance [32,33] were pooled

statistically. Since comparable assessments on medication

taking were not available in the different studies, the HbA1c

was used as a common outcome measure. The weighted mean

difference (WMD) of the studies evaluating the effect of

education offered individually on the short (3–6 months) and

medium term (9–12 months) was calculated. It was not

surprising that the I2 test exceeded 70%, indicating consider-

able levels of heterogeneity [38] thus making statistical pooling

almost unacceptable.

4. Discussion

This subgroup analysis of a Cochrane Systematic Review could

not draw clear and consistent conclusions on the effect of

educational interventions aimed at improving adherence in

medication taking in people living with type 2 diabetes. A

slightly positive overall effect is shown, but this leaves us with

the question whether a small statistical significant effect

entails clinical relevant effects. Moreover, even when an effect

has been shown it remains unclear which intervention or

which part of a complex educational intervention contributed

the most to the perceived effect. A second conclusion of this

review is that there is an urgent need for high quality

interventional and observational studies on the relation

between education in people living with type 2 diabetes and

adherence to medicine taking or even self management as a

whole, assessing effectiveness and efficiency of the performed

interventions. Finally, authors should clearly define adher-

ence and should search for valid methods to measure the

adherence as they define it.

Though the review process was performed according to the

Cochrane Metabolic and Endocrine Diseases Collaborative

Review Group recommendations, there are some limitations.

First, on the one hand a comprehensive search in different

electronic databases and by hand searching was performed,

but on the other hand one may assume that in this area of

research many studies may be unpublished. Unfortunately,

funnel plots could not be calculated to assess publication bias

because of the high clinical heterogeneity. Second, only eight

studies could be included in this analysis. The most important

reason for exclusion was the low quality of retrieved studies

which main shortcomings were the unclear or missing

concealment of allocation and the blinding procedures.

Moreover, the lack of appropriate operational definitions of

adherence contributed to a high exclusion rate. In a majority of

articles, although compliance or adherence was mentioned to

be the topic of research, it merely was diabetes care in general

or self-care in particular that was assessed. Many authors

even did not feel the need to define adherence nor did they

make no clear distinction between compliance, adherence or

concordance [12]. The authors’ lack of definition of adherence

was the most disturbing factor in this reviewing process. At

least authors could lay upon Glasgow and Anderson [25] who

state that the constructs of ‘compliance’ and ‘adherence’

might be replaced by ‘collaborative goal-setting’ and ‘on-going

self-management support’. Third, as a result of the lack of

definition, valid measurement instruments and valid outcome

measures are missing. Since HbA1c was shown in most of the

included studies this outcome measure was used as a proxy

for adherence. Fourth, this review was not able to explore

which educational interventions or which part of complex

interventions contributed most to the overall effect size. Fifth

is the sample size that may limit the synthesis and its

conclusions. Most of the studies included rather small

numbers of participants limiting statistical procedures and

hence preventing from doing comparisons between groups

and subgroups. Sixth, no economic aspects were studied. This

is quite surprising because, although better diabetes outcomes

are aimed at by governments, managed care organizations,

caregivers and by patients themselves, no economic con-

sequences of interventions were considered.

Seventh, the literature review was completed in November

2002 and trials published since then may differ in their results,

although they would be unlikely to substantially change the

conclusions drawn from the eight trials included in this

review. An update of the review will be performed in 2008.

Besides the mentioned limitations and weaknesses, a

number of strengths are to be stressed. The review’s

conclusions are consistent with a UK Health Technology

Assessment systematic review [39] which concludes that

education, as part of an intensification of treatment, produces

improvement in diabetic control in type 1 diabetes, which has

not consistently been shown in type 2 though. For people with

type 2 diabetes, a meta-analysis assessing the role of

education shows overall modest improvements in glycaemic

control too [40]. The problems that were faced in the reviewing

process seem to be common to reviewing adherence issues in

medical care in general. Authors of different systematic

reviews of adherence to treatment for blood pressure lowering

medication [41], lipid lowering drugs [42] or even medication

adherence as a whole for chronic health problems [11]

described exactly the same problems and pitfalls. In general

clear conclusions were hard to be drawn.

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8 387

The importance of education on improving adherence to

medication has been stressed earlier [43]. The relation between

adherence to medications (based on pharmacy claims) and

clinical outcomes has been shown for type 2 diabetes, since a

10% decrease in adherence to metformin and statins was

associated with an increase of 0.14% of HbA1c [44]. It has to be

stressed that clinical outcomes are not always consistently

related to the magnitude of adherence results [45].

The complexity of adherence has prevented the develop-

ment of a gold standard method of measurement [46]. Direct

and indirect measurements are being used, but serious

obstacles remain in generating valid and reliable estimates of

the extent of adherence [12]. Aiming at understanding better

and improving more adherences, a focus on reliability and

validity of adherence measurement is urgently needed [47].

Moreover, it is challenging to decide which outcomes of

educational interventions should be considered. Adherence to

treatment recommendations is to be found in physical

activity, eating, medication taking, monitoring of blood

glucose, problem solving as for sick days, reducing risks for

diabetes complications and psychosocial adaptation [48].

Further research into this area is desperately needed. It is

striking that in such an ‘epidemic disease’ with a high human

and financial burden so little high quality research on

improving adherence has been executed. Though everybody

claims that education is an essential part of the management

of diabetes, the effectiveness of educational interventions on

adherence and the relation between adherence and clinical

outcomes still need clarification. Also, the efficiency of

different educational interventions such as group sessions,

education on distance and the use of new information

technology need to be studied more profoundly, not omitting

economic assessments [21,49,50].

5. Conclusions

The question whether educational interventions aiming at

improving adherence to medication in people living with type

2 diabetes are effective remains unanswered.

In spite of a high number of abstracts fitting the search

strategy only eight studies could be included in this subgroup

analysis of a systematic review. Though the term ‘adherence’

is often mentioned in titles or abstracts, adherence itself is not

the subject of the research presented.

It is important though to state that this review did not show

evidence that educational interventions may be ineffective or

even harmful. Finally, this review process point at how

trustworthy research in the future should be conducted.

Conflicts of interests

None.

Acknowledgement

The research project was made possible by an unconditional

grant of Primary Care Diabetes Europe ivzw/inpo.

r e f e r e n c e s

[1] E. Vermeire, J. Wens, P. Van Royen, Y. Biot, H. Hearnshaw,A. Lindenmeyer, Interventions for improving adherence totreatment recommendations in people with type 2 diabetesmellitus, Cochrane Database of Systematic Reviews 2005,Issue 2, Art. No.: CD003638, doi:10.1002/14651858.CD003638.pub2.

[2] International Diabetes Federation, Triennial Report (1991–1994) and Directory, Brussels, 1994.

[3] A. Amos, D.P.Z. McCarty, The rising global burden ofdiabetes and its complications: estimates and projectionsto the year 2010, Diab. Med. 14 (1997) S1–S85.

[4] H. King, R.E. Aubert, W.H. Herman, Global burden ofdiabetes, 1995–2025: prevalence, numerical estimates, andprojections, Diab. Care 21 (9) (1998) 1414–1431.

[5] G. Roglic, N. Unwin, P.H. Bennett, C. Mathers, J. Tuomilehto,S. Nag, et al., The burden of mortality attributable todiabetes: realistic estimates for the year 2000, Diab. Care 28(9) (2005) 2130–2135.

[6] American Diabetes A, Standards of medical care indiabetes—2007, Diab. Care 30 (Suppl. 1) (2007)S4–S41.

[7] UK Prospective Diabetes Study Group, Intensive bloodglucose control with sulphonylureas or insulin comparedwith conventional treatment and risk of complications inpatients with type 2 diabetes (UKPDS 33), Lancet 352 (1998)837–853.

[8] UK Prospective Diabetes Study Group, Effect of intensiveblood glucose control with metformin on complications inoverweight patients with type 2 diabetes (UKPDS 34),Lancet 352 (1998) 854–865.

[9] UK Prospective Diabetes Study Group, Tight blood pressurecontrol and risk of macro vascular and micro vascularcomplications in type 2 diabetes (UKPDS 38), BMJ 317 (1998)703–713.

[10] UK Prospective Diabetes Study Group, Efficacy of atenololand captopril in reducing risk of both macro vascular andmicro vascular complications in type 2 diabetes (UKPDS 39),BMJ 317 (1998) 73–720.

[11] R.B. Haynes, X. Yao, A. Degani, S. Kripalani, A. Garg, H.P.McDonald, Interventions for enhancing medicationadherence. Cochrane Database of Systematic Reviews 2005,Issue 4, Art. No.: CD000011, doi:10.1002/14651858.CD000011.pub2.

[12] E. Vermeire, H. Hearnshaw, P. Van Royen, J. Denekens,Patient adherence to treatment: three decades of research.A comprehensive review, J. Clin. Pharm. Ther. 26 (5) (2001)331–342.

[13] M. Nagasawa, M.C. Smith, J.H.J.R. Barnes, J.E. Fincham,Meta-analysis of correlates of diabetes patients’compliance with prescribed medications, The Diab.Educator 16 (3) (1990) 192–200.

[14] E. Vermeire, P. Van Royen, S. Coenen, J. Wens, J. Denekens,The adherence of type 2 diabetes patients to theirtherapeutic regimens: a qualitative study from the patient’sperspective, Pract. Diab. Int. 20 (6) (2003) 209–214.

[15] H. Van Dam, F. Van der Horst, B. Van den Borne, R.Ryckman, H. Crebolder, Provider–patient interaction indiabetes care: effects on patient self-care and outcomes. Asystematic review, Patient Educ. Couns. 51 (2003) 17–28.

[16] S.E. Ellis, T. Speroff, R.S. Dittus, A. Brown, J.W. Pichert, T.A.Elasy, Diabetes patient education: a meta-analysis andmeta-regression, Patient Educ. Couns. 52 (1) (2004)91–105.

[17] S. Brown, Effects of educational interventions indiabetes care: a meta-analysis of findings, Nurs. Res. 37(1988) 223–230.

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 7 9 ( 2 0 0 8 ) 3 7 7 – 3 8 8388

[18] D. Padgett, Meta-analysis of the effects of educational andpsychosocial interventions on management of diabetesmellitus, J. Clin. Epidemiol. 41 (1988) 1007–1030.

[19] S. Brown, Meta-analysis of diabetes patient educationresearch: variations in intervention effects across studies,Res. Nurs. Health 15 (1992) 409–419.

[20] D.L. Roter, J.A. Hall, R. Merisca, B. Nordstrom, D. Cretin, B.Svarstad, Effectiveness of interventions to improve patientcompliance: a meta-analysis, Med. Care 36 (8) (1998) 1138–1161.

[21] S.L. Norris, M.M. Engelgau, K.M. Venkat Narayan,Effectiveness of self-management training in type 2diabetes: a systematic review of randomized controlledtrials, Diab. Care 24 (3) (2001) 561–587.

[22] International Diabetes Federation, Diabetes Atlas/DiabetesEducation, second ed., IDF, Brussels, 2003, Chapter 6, p.208.

[23] S.A. Brown, Interventions to promote diabetes self-management: state of the science, The Diab. Educator 25 (6Suppl.) (1999) 52–61.

[24] P. Corabian, C. Harstall, Patient diabetes education in themanagement of adult type 2 diabetes: Centre for Reviewsand Dissemination. Database of Abstracts of Reviews ofEffects (DARE); 2001.

[25] R.E. Glasgow, R.M. Anderson, In diabetes care, moving fromcompliance to adherence is not enough. Something entirelydifferent is needed, Diab. Care 22 (12) (1999) 2090–2092.

[26] R.M. Anderson, M.M. Funnell, Compliance and adherenceare dysfunctional concepts in diabetes care, The Diab.Educator 26 (4) (2000) 597–604.

[27] Olivarius NdF, H. Beck-Nielsen, A.H. Andreasen, M. Horder,P.A. Pedersen, Randomised controlled trial of structuredpersonal care of type 2 diabetes mellitus, BMJ 323 (7319)(2001) 970.

[28] J. Wens, E. Vermeire, P. Van Royen, H. Hearnshaw, Asystematic review of adherence with medications fordiabetes: response to cramer, Diab. Care 27 (9) (2004), 2284-.

[29] World Health Organisation (WHO), Adherence to long-termtherapies. Evidence for action, World Health Organisation(WHO), Geneva, 2003, Chapter 1, Defining adherence, pp. 3–7.

[30] L.A. Jaber, H. Halapy, M. Fernet, S. Tummalapalli, H.Diwakaran, Evaluation of a pharmaceutical care model ondiabetes management, The Ann. Pharmacother. 30 (1996)238–243.

[31] B.P. Krier, R.D. Parker, D. Grayson, G. Byrd, Effect of diabeteseducation on glucose control, J. La State Med. Soc. 151(1999) 87–92.

[32] A. Mease, W.L. Whitlock, A. Brown, K. Moore, H. Pavliscsak,A. Dingbaum, et al., Telemedicine improved diabeticmanagement, Mil. Med. 165 (8) (2000) 579–584.

[33] J.D. Piette, M. Weinberger, F.B. Kraemer, S.J. McPhee,Impact of automated calls with nurse follow-up on diabetestreatment outcomes in a department of veterans affairsHealth Care System, Diab. Care 24 (2) (2001) 202–208.

[34] R. Rachmani, Z. Levi, I. Slavachevski, M. Avin, M. Ravid,Teaching patients to monitor their risk factors retards theprogression of vascular complications in high-risk patients

with Type 2 diabetes mellitus. A randomized prospectivestudy, Diab. Med. 19 (2002) 385–392.

[35] N. White, J. Carnahan, C.A. Nugent, T. Iwaoka, M.A.Dodson, Management of obese patients with diabetesmellitus: comparison of advice education with groupmanagement, Diab. Care 9 (5) (1986) 490–496.

[36] E.A. Coast-Senior, B.A. Kroner, C.L. Kelley, L.E. Trilli,Management of patients with type 2 diabetes bypharmacists in primary care clinics, The Ann.Pharmacother. 32 (1998) 636–641.

[37] Y.-D. Jiang, L.-M. Chuang, H.-P. Wu, S.-J. Shiau, C.-H. Wang,Y.-J. Lee, et al., Assessment of the function and effect ofdiabetes education programs in Taiwan, Diab. Res. Clin.Pract. 46 (1999) 177–182.

[38] Available from: http://www.cochrane-net.org/openlearning/HTML/mod13.htm.

[39] E. Loveman, C. Cave, C. Green, P. Royle, N. Dunn, N. Waugh,The clinical and cost-effectiveness of patient educationmodels for diabetes: a systematic review and economicevaluation, Health Technol. Assess. 7 (22) (2003) 1–202.

[40] T.L. Gary, J.M. Genkinger, E. Guallar, M. Peyrot, F.L. Brancati,Meta-analysis of randomized educational and behavioralinterventions in type 2 diabetes, The Diab. Educator 29 (3)(2003) 488–501.

[41] K. Schroeder, T. Fahey, S. Ebrahim, Interventions forimproving adherence to treatment in patients with highblood pressure in ambulatory settings, Cochrane Databaseof Systematic Reviews 2004, Issue 3, Art. No.: CD004804,doi:10.1002/14651858.CD004804.

[42] A. Schedlbauer, K. Schroeder, T.J. Peters, T. Fahey,Interventions to improve adherence to lipid loweringmedication, Cochrane Database of Systematic Reviews2004, Issue 4, Art. No.: CD004371, doi:10.1002/14651858.CD004371.pub2.

[43] G. Rutten, Diabetes patient education: time for a new era,Diab. Med. 22 (6) (2005) 671–673.

[44] M. Pladevall, L.K. Williams, L.A. Potts, G. Divine, H. Xi, J.Elston Lafata, Clinical outcomes and adherence tomedications measured by claims data in patients withdiabetes, Diab. Care 27 (12) (2004) 2800–2805.

[45] S. Kripalani, X. Yao, R.B. Haynes, Interventions to enhancemedication adherence in chronic medical conditions: asystematic review, Arch. Intern. Med. 167 (6) (2007) 540–549.

[46] K.C. Farmer, Methods for measuring and monitoringmedication regimen adherence in clinical trials and clinicalpractice, Clin. Ther. 21 (1999) 1074–1090.

[47] M.R. DiMatteo, Variations in patients’ adherence to medicalrecommendations: a quantitative review of 50 years ofresearch, Med. Care 42 (3) (2004) 200–209.

[48] American Association of Diabetes E, Standards foroutcomes measurement of diabetes self-managementeducation, The Diab. Educator 29 (5) (2003) 804–816.

[49] D.C. Klonoff, D.M. Schwartz, An economic analysis ofinterventions for diabetes, Diab. Care 23 (3) (2000)390–404.

[50] R.M. Kaplan, W.K. Davis, Evaluating the costs and benefitsof outpatient diabetes education and nutrition counseling,Diab. Care 9 (1) (1986) 81–86.