Medicines adherence: involving patients in decisions about prescribed medicines and supporting adherence

lived in residential care

received dose-dispensed medicine from a pharmacy

had medicine dispensed by a home nurse

had terminal illness

had conditions that precluded patient interview

lived outside the region of Southern Denmark

were treated with more than 1 drug within a drug class.

collaborative care

medication review

adherence counselling including motivational interviewing and telephone follow‑ups with patients for up to 6 months.

composite MPR at 3, 6 and 9 months

adherence and persistence at 12 months to:

a 12‑months combined end point of:

Median composite MPR was 0.91 (interquartile range [IQR] 0.76 to 0.98) in the control group, compared with 0.93 (IQR 0.82 to 0.99) in the intervention group (p=0.02).

A significantly greater percentage of control group patients (30.2%) were non-adherent compared with intervention patients (20.3%) (p=0.01).

At 12 months, a greater proportion of control patients were classified as non-adherent for lipid-lowering agents (33.0% versus 23.4%; p=0.04) and plain renin-angiotensin agents (25.1% versus 16.2%; p=0.05).

There were no statistically significant differences in 12‑month adherence rates for the remaining drug classes analysed.

There were no statistically significant differences in non-persistence for any of the drug classes.

There were no statistically significant differences for any of the clinical outcomes.

Physicians accepted 58 (64%) of 91 recommendations made by pharmacists on drug-related problems. However, there were no significant differences between the proportions of patients in each group with medication changes (49% of the intervention patients versus 46% of the control group [p=0.54]).

The study was adequately powered to demonstrate significant differences between groups.

The study enhanced the evidence base for multifaceted pharmacist interventions, including motivational interviewing, in the secondary care setting.

The risk of bias was low due to appropriate randomisation, concealment, objective outcome measures and intention to treat analysis.

The study included a 12‑month follow‑up period to assess sustained effects after the intervention had ended.

The study used 5 extensively trained pharmacists at 3 different outpatient clinics.

Objective and established measures of adherence were used.

Although data were obtained from 3 different outpatient clinics, the study was conducted at a single site in Denmark, and only patients from that specific region were eligible.

Blinding of patients and pharmacists was not possible due to the nature of the intervention, but the risk of bias was reduced by the use of objective outcome measures.

A longer-term follow‑up study may be required to fully assess the prognostic impact on clinical outcomes.

The study was focused on patients with hypertension. Larger studies across multiple conditions, which also include cost-effectiveness data, may be required to establish an impact on NICE guideline CG76.

were under the age of 18 years

lived outside of England

had difficulty reading or understanding English or understanding the details of the study

were prescribed drugs for dementia or showed signs of cognitive impairment

had substantial hearing or sight impairment

had medications ordered by a caregiver or family member

received a first prescription of a medication included in the study.

Two tailored telephone consultations with a pharmacist, between 4 and 6 weeks apart. The first followed a semi-structured interview format and the second was to follow this up with a review and a discussion of any new or outstanding issues.

A written summary of the discussion, following the first telephone consultation.

A personalised medicines reminder chart setting out the patient's prescribed medications, including purpose of the medicine, dosage and timing.

lipid and glycaemic control, measured by a self-administered finger prick test

patient satisfaction, measured by 4‑week follow‑up self-report questionnaire.

An additional objective measure of adherence was used to reduce the risk of bias inherent in the subjective measure of self-reported adherence.

Incomplete outcome data was explained by the reasons for withdrawal and addressed by intention to treat analysis.

The study built on earlier research that led to the development of the New Medicines Service (NMS), by focusing on established medicines over a longer period.

The study was reported to be the first of its kind to investigate the impact of a pharmacist-led intervention in mail order pharmacy.

In view of the chronic nature of the conditions, the follow‑up for 6 months was too short to assess fully the clinical impact on symptoms.

Blinding of participants and pharmacists was not possible, although the risk of performance bias was lessened by presenting the intervention to participants as an opportunity to access advice from a pharmacist, without mentioning adherence.

The authors acknowledged that the study would have benefitted from a pilot and feasibility stage, which may have helped to improve the initial response rate of invited participants.

A single mail-order pharmacy was used for recruiting participants. It is possible that the results may not be generalisable to the whole UK setting, where patients are able to obtain prescriptions from any pharmacy.

A single pharmacist was used to deliver the intervention.

Cost-effectiveness data were not included in the analysis.

The number of respondents completing the satisfaction questionnaire was not clearly reported.

A total of 128 (37.6%) patients in the intervention group either did not receive the intervention or were lost to follow‑up. Only 64 (19%) patients in the control group were lost to follow‑up. Although intention to treat analysis was employed to address this imbalance, the authors did not discuss the impact of the high attrition rate.