Lim 2004.

Study characteristics
Methods	Aim of study: to evaluate the impact of a pharmacist consult clinic on health‐related outcomes of elderly outpatients in a local setting Study design: RCT (randomised in blocks of 2 participants) Number of arms/groups: 2
Participants	Description: patient/consumer Geographic location: Singapore Setting: outpatient clinic (geriatric medicine hospital outpatient clinic) Inclusion criteria: participants who required drug therapy monitoring, evidence of polypharmacy (> 3 regular meds or > 9 doses per day), documented non‐compliance, self‐administered drugs that require psychomotor skill and co‐ordination, on nasogastric tube feeding, > 1 doctor managing care OR hospitalised within last 6 months Exclusion criteria: stable on follow‐up, cognitive impairment and no caregiver to participate, life expectancy < 6 months, medications supervised by other healthcare personnel Number of participants randomised: 136 (68 and 68) Number of participants included in analysis: 126 (64 and 62) Age: mean ± SD: 79.6 ± 7.7 vs 80.5 ± 8.1 Gender: female: 60.9% vs 69.4% Ethnicity: Chinese 73.4% vs 83.9%, Malay 6.3% vs 6.5%, Indian 12.5% vs 6.5%, Other 7.8% vs 3.2% Number of medications: regularly scheduled medicines: median (range): 6 (3 to 16) vs 7 (3 to 10) Frailty/Functional impairment: ADL independent: 50.8 % vs 40.3% Cognitive impairment: impaired cognition: 20.3% vs 21.0% Comorbidities: not specified
Interventions	Group 1 ‐ Pharmacist consult in clinic (10 to 30 minutes) ‐ evaluate patients for MRPs by reviewing medical records and medication list and by interviewing patient and caregiver. Recommendations to simplify, reduce ADEs, decrease cost, etc., discussed with primary physician, and accepted recommendations implemented. Pharmacist also counselled on medication knowledge, administration, etc. Group 2: assumed usual care Co‐intervention: N/A Provider: pharmacist Where: outpatient clinic When and how often: once at baseline Intervention personalised: yes ‐ Individualised based on medications and MRPs
Outcomes	Timing of outcome assessment: baseline and 2 months Medication adherence (subjective) : self‐reported compliance: patients asked if they 'forgot to take medication as directed'. Then categorised as compliant or not. Participants then classified as least compliant (compliant at base, not at 2 months), not compliant (not compliant at base or 2 months), compliant (compliant at 2 months) Knowledge about medicines (objective): composite % knowledge of dose (D), frequency (F), and indication (I), reported as percentage correct. Adverse clinical health outcomes (subjective): reported ADRs. Asking patients if they experience side effects or unwanted reactions with their medications. Patients asked to name medication involved; this was assessed by primary care physician to ascertain whether symptoms were indeed ADRs of the implicated medicine
Notes	Trial registration: N/A Consumer involvement: not specified Funding source: National Healthcare Group research grant NHG‐RPR/01027 Dropout: 10 excluded before intervention (4 and 6), 9 withdrew (5 and 4), 17 lost‐to follow up (8 and 9) Further information required: raw data on adherence and medication knowledge at follow‐up (email correspondence with trial author successful, but further data not available)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomly assigned via computer‐generated numbers in blocks of 2
Allocation concealment (selection bias)	Unclear risk	Randomisation carried out before consent (Zelen design)
Blinding of participants and personnel (performance bias) All outcomes	High risk	Unblinded
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not specified
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Figure 2 shows study profile; ITT concluded patients only
Selective reporting (reporting bias)	High risk	Raw values for outcomes not listed; 90% CI?; no sample size calculation
Other bias	Low risk	None apparent. Sample size 60/arm "to achieve a power of 80% to detect a 10% difference between the 2 groups in the knowledge outcome"