Murray 1993.

Study characteristics
Methods	Aim of study: to determine the effect of unit‐of‐use packaging on medication compliance among elderly outpatients treated with complex medication regimens Study design: RCT (unit of allocation: individual) Number of arms/groups: 3
Participants	Description: patient/consumer Geographic location: USA Setting: primary care clinic (geriatric outreach centres located in urban public housing units for the elderly and disabled (people living independently)) Inclusion criteria: ≥ 60 years, ≥ 3 medications Exclusion criteria: medication pharmacologic and pharmacokinetic properties considered unfeasible for twice‐daily regimen, nursing home Number of participants randomised: 36 Number of participants included in analysis: 31 (control 1: 12, control 2: 10, intervention: 9) Age: mean (range) = C1: 71.3 (64 to 81), C2: 72.5 (60 to 87), I: 72.9 (63 to 81) Gender: female: C1: 9 (75%), C2: 8 (80%), I: 6 (67%) Ethnicity: black (not white): C1: 8 (75%), C2: 9 (90%), I: 6 (67%) Number of medications: type not specified: mean ± SD: C1: 4.8 ± 2.2, C2: 3.8 ± 1.1, I: 5.1 ± 2.1 Frailty/Functional impairment: Medical Outcomes Study General Health Survey: physical function mean ± SD = C1: 52.8 ± 34.0, C2: 43.3 ± 29.6, I: 37.9 ± 31.7 Cognitive impairment: mean ± SD MMSE: C1: 27.2 ± 2.2, C2: 28.5 ± 1.0, I: 27.7 ± 1.8 Comorbidities: not specified
Interventions	Group 1 (intervention) ‐ Unit‐of‐use medication packaging and regimen simplification: medications in unit‐of‐use packages with twice‐daily dosing intervals (morning and evening). Medications in translucent plastic cups with translucent plastic snap‐on lids. Yellow label for AM, blue label for PM Group 2 (control 1) ‐ Usual care: medications in conventional packaging and no change to dosing interval Group 3 (control 2) ‐ Regimen simplification: medications in conventional packaging but dosing intervals made twice daily (morning and night) using 2 clear plastic zip‐lock bags. Co‐intervention: all medications packaged individually by study pharmacist and dispensed monthly (33 days supply) Provider: pharmacist Where: ambulatory clinic/home When and how often: monthly (medications resupplied monthly) Intervention personalised: no
Outcomes	Timing of outcome assessment: 6 months (assessed monthly) Medication adherence (objective) : pill count: percentage compliant (note over‐adherence expressed as under‐adherence, e.g. 90% not 110%). Scale 0 to 100 Adverse clinical health outcomes (subjective): asked: "Have you had any side effects, ill effects, or any other problems caused by medications you have taken? (yes/no)"
Notes	Trial registration: N/A Consumer involvement: not specified Funding source: Health Foundation of Greater Indianapolis Dropout: 4 withdrew, 1 was lost to follow‐up
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not specified
Allocation concealment (selection bias)	Unclear risk	Not specified
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding of pharmacist who delivered the intervention and collected outcome data
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding of pharmacist who delivered the intervention and collected outcome data
Incomplete outcome data (attrition bias) All outcomes	High risk	5 people missing (16%): 3 from intervention group (25%), 1 NH, 3 returned to prior regimen, 1 disliked unit‐of‐use packaging
Selective reporting (reporting bias)	High risk	Compliance measured in 4 ways; subjective not reported at follow‐up; results in abstract not matching main paper
Other bias	Unclear risk	No power calculation. Small sample size in each group. Groups not particularly well matched (e.g. mean number of drugs). Pill counts occurred in the pharmacy ‐ patients may not have returned all meds, and this be more of an issue with unit‐of‐use packaging (a bag for empty containers was provided, but it is possible that containers were discarded and % containers returned was not reported)