This review assessed interventions to improve medication taking in elderly patients prescribed multiple medications and concluded that no particular intervention was favourable because of inconsistent methodology and findings across studies. The authors' cautious conclusions reflected the evidence presented. Potential for missing studies and the uncertain quality of the included studies should be borne in mind.

To assess the effectiveness of interventions to improve medication taking in elderly patients prescribed multiple long-term medications.

MEDLINE, EMBASE, CINAHL, PsycINFO and IPA were searched for English-language studies to 2006; search terms were reported. Additional studies were sought through searches of references from published reviews.

Controlled studies with at least 60 patients in each group that focused on community-living elderly patients and assessed interventions to improve medication adherence were eligible for inclusion. Patients were aged at least 55 years of age (mean or median age of at least 65 years) and received at least three (or mean/median of four or more) long-term medications. Interventions that sought to increase medication adherence were delivered to the patient or their carer. Control groups could comprise either no intervention or usual care. Patients had to be followed for at least four weeks after the intervention. Adherence to all oral, long-term, solid-dose medications were measured at baseline and study conclusion.

Included studies used educational (verbal or written information) or behavioural (dose administration aid, special medication packaging or labelling, reminder chart or calendar, home visit, follow-up or regimen simplification) interventions or medication review. All studies tested the role of patient education in combination with behavioural strategies and/or provider-focused strategies such medication review. None of the studies assessed a single adherence-enhancing intervention. Interventions were delivered by hospital, community or primary care pharmacists (or a combination). Adherence was measured using self-reported medication utilisation, Morisky scale and its modifications, refill date, prescription claims data and pill count. Relative changes in adherence were estimated from the difference in mean adherence at the conclusion of the study and baseline divided by the baseline adherence. Mean age ranged from 68 to 84 years. The proportion of females was zero to 83%. Studies were conducted in Canada, USA, Australia, UK and one study in seven European countries.

Two reviewers independently selected studies for inclusion in the review. Disagreements were resolved through review by a third reviewer and consensus between all reviewers.

The authors did not state that they assessed validity.

Two reviewers independently extracted adherence data for intervention and control groups at baseline and endpoint as reported in the included studies. Authors assessed whether there was any significant difference in adherence between intervention and control.

Disagreements were resolved through review by a third reviewer and consensus between all reviewers.

The studies were combined in a narrative synthesis.

Eight studies (n=4,140, range 135 to 2,454) were included.

Interventions resulted in a significant improvement in adherence in four studies, all of which used individualised patient education provided by a pharmacist and supplemented by at least one behavioural strategy either with or without medication review.

Relative change in adherence in the intervention groups varied from -13% to 55.5% (mean 11.4%; eight studies). In the control arms, relative change in adherence ranged from -54.1% to 36.7% (mean -7.5%; eight studies).

Medication review by pharmacists with a focus on regimen simplification was found to be effective in two studies.

Inconsistent methodology and findings across the eight studies precluded firm conclusions in favour of any particular intervention.

The review question was clear and supported by appropriate inclusion criteria. Several relevant databases were searched. There was potential for language and publication biases as only published English-language articles were eligible for inclusion and it was unclear whether unpublished studies were sought. Two authors undertook study selection and data extraction, which minimised risks of error and bias in the review. There was no reported validity assessment, which made it difficult to judge the reliability of the included studies. A narrative synthesis was presented; this was appropriate given the variability across studies.

The authors' cautious conclusions reflected the evidence presented. Potential for missing studies and the uncertain quality of the included studies should be borne in mind.

Practice: The authors stated that combinations of educational and behavioural strategies should be used to improve medication adherence in the elderly until evidence was available from single-intervention strategies.

Research: The authors stated that reliable measures for assessing adherence were required, as were innovative strategies for enhanced medication adherence in the elderly.

None stated.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.