Begley 1997.

Study characteristics
Methods	Aim of study: to evaluate the influence of domiciliary pharmacist visits on medication management in a sample of elderly people recently discharged from hospital to their own homes Study design: RCT (3 hospitals; unit of allocation: individual) Number of arms/groups: 3
Participants	Description: patient/consumer Geographic location: UK Setting: community (post discharge) Inclusion criteria: ≥ 75 years, ≥ 3 prescribed drugs, ≥ 2 doses/d, under care of participating consultant, consented to participate, discharged to own home Number of participants randomised: 222 (A: 74, B: 75, C: 73) Number of participants included in analysis: 190 (A: 61, B: 63, C: 66) Age: median (range) = A: 84 (75 to 94), B: 81 (75 to 96), C: 82 (76 to 92) Gender: female A: 61%, B: 65%, C: 56% Ethnicity: not specified Number of medications: mean (SD) prescribed: A: 4.6 (1.8), B: 4.8 (1.6), C: 5.5 (1.9); mean (SD) OTC: A: 2.6 (0.7), B: 4.1 (1.4), C: 2.2 (1.8) Frailty/Functional impairment: not specified Cognitive impairment: abbreviated mental test mean (SD): A: 8.4 (1.5), B: 8.8 (1.1), C: 7.9 (1.3) Comorbidities: not specified
Interventions	Group 1 ‐ Domiciliary pharmacy medication management visit: interview consisted of 6 sections: patient information, drug knowledge, patient dexterity, abbreviated mental test, medication management, compliance. Following interview, intervention group (A) received structured counselling on correct use, storage, and compliance (including simplifying regimen, emphasising importance of compliance, positive reinforcement) Group 2‐ Group (B): control, with home interview but no counselling Group 3 ‐ Group (C): control with no home visit (i.e. usual care) Co‐intervention: N/A Provider: pharmacist (investigator) Where: home (post discharge) When and how often: home visits for A and B occurred at baseline, 2 weeks, 1 month, 3 months, and 12 months Intervention personalised: yes ‐ counselling tailored to needs of the patient
Outcomes	Timing of outcome assessment: 12 months Medication adherence (objective): percentage of medications with which participant is compliant using pill count. Researcher counted remaining tablets and measured volume of liquid. To improve reliability and accuracy, patients were asked to retain all used medicine containers for removal by the investigator Medication‐taking ability (objective): dexterity medication management. 5‐task dexterity test (e.g. opening child‐resistant closure) assessed at baseline and at 12 months; 1 point awarded for each successfully completed activity Knowledge about medicines (subjective): drug knowledge. Percentage of correct answers. Patients asked about name, purpose, dose, dosage frequency, and side effects. Accuracy compared to hospital discharge or GP instructions
Notes	Trial registration: N/A Consumer involvement: not specified Funding source: not specified Dropout: 4 withdrew (refused), 28 were lost to follow‐up (7 death, 7 hospitalised, 10 nursing home, 4 moved out of area) For this review, group A vs group C was considered under Comparison 1 ‐ Intervention vs control; group A vs group B was considered under Comparison 2 ‐ Intervention vs intervention
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Recruitment staff blinded to identity of groups; allocated patients consecutively to group A, B, or C. No random sequence generation
Allocation concealment (selection bias)	Unclear risk	Allocated sequentially by blinded recruiter
Blinding of participants and personnel (performance bias) All outcomes	High risk	Unable to blind participants; patients knew what they were getting (respondent bias)
Blinding of outcome assessment (detection bias) All outcomes	High risk	Intervention delivered by same pharmacist taking outcome assessment
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Lost to follow‐up; reported only completed patients; rates seem similar across groups but breakdown of reasons not available for each group
Selective reporting (reporting bias)	High risk	Some outcomes not reported at follow‐up (e.g. dexterity). Unclear whether results for adherence are patient‐reported or pill count; gives only 'difference' ‐ not raw scores for both
Other bias	Low risk	Required sample size achieved (61 per group)