Naunton 2003.

Study characteristics
Methods	Aim of study: to evaluate pharmacist‐conducted post‐discharge follow‐up at home of high‐risk elderly patients on various outcomes (including adherence) Study design: RCT (unit of allocation: individual) Number of arms/groups: 2
Participants	Description: both patient/consumer and carer Geographic location: Australia Setting: home (post discharge) Inclusion criteria: ≥ 60 years, ≥ 2 chronic medical conditions requiring medication (≥ 1 of HF, IHD, COPD, or DM), ≥ 4 prescribed regular medications Exclusion criteria: lived in domiciliary care facility or beyond greater Hobart area, were to be visited at home by a community nurse within 5 days of discharge, had terminal malignancy, were unable to provide informed consent Number of participants randomised: 136 Number of participants included in analysis: 121 (57 and 64) (unclear as number of participants alive at 90 days: 54 vs 59) Age: median (range): 74 (65 to 90) vs 77 (60 to 91) Gender: female: 56% vs 69% Ethnicity: not specified Number of medications: regular medications on discharge: median (range): 8 (3 to 15) vs 8 (3 to 16) Frailty/Functional impairment: nursing assistance at home: 28% vs 21% Cognitive impairment: not specified Comorbidities: chronic medical conditions: median (range) = 5 (2 to 9) vs 5 (2 to 13)
Interventions	Group 1 ‐ Pharmacist post‐discharge home visit: 5 days after discharge, patients visited at home by study pharmacists. Objective of visit was to educate, answer any queries, optimise medication management (e.g. Dosette or Webster, if necessary), improve compliance, detect DRPs, and improve liaison with community‐based health services. Brief letter composed in the patient's home was given to the patient to present to the doctor. Study pharmacist also called GP and community pharmacy to inform of study Group 2 ‐ Usual care (no specific post‐discharge follow‐up for this group) Co‐intervention: 89% in both groups were seen by hospital pharmacist before discharge Provider: pharmacist Where: home When and how often: once, 5 days after discharge Intervention personalised: yes ‐ based on adherence assessment at home visit; specific strategies were offered such as compliance aids, carer assistance, community nursing, etc.
Outcomes	Timing of outcome assessment: 90 days post discharge Medication adherence (subjective) : self‐reported missing doses: compliance defined as 'never miss medication'. Non‐compliance ‐ therefore any self‐reported missed doses (from 'rarely' to 'once a day'). Self‐reported 'never' forget to take their medication. 1 question ‐ "How often would you say you miss taking your pills?", with 7 response options from never to once a day. Presented as dichotomous variable adherent or not adherent. ? Not validated Satisfaction with intervention (subjective): satisfaction survey ‐ intervention group only Adverse clinical health outcomes (objective): deaths: % patients who died within 90 days of discharge. Retrospective medical record review and contact with family and/or GP Adverse clinical health outcomes (objective): unplanned hospital re‐admissions: % patients with 1 or more unplanned re‐admissions within 90 days of discharge (patients asked and retrospective medical records checked)
Notes	Trial registration: N/A Consumer involvement: not specified Funding source: Abbott Australasia Pharmacy Research Grant, through SHPA Dropout: 2 withdrew (1 I, 1 C); 13 were lost to follow‐up (3 C died, 3 I and 2 C were uncontactable, 2 I and 1 C were admitted to nursing home, 2 I was admitted to intensive nursing care)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised by study pharmacist using a computer‐generated list of random numbers
Allocation concealment (selection bias)	Low risk	Randomisation occurred after discharge from hospital and collection of baseline data
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not blinded
Blinding of outcome assessment (detection bias) All outcomes	High risk	Not blinded post randomisation
Incomplete outcome data (attrition bias) All outcomes	Low risk	Good breakdown of excluded patients
Selective reporting (reporting bias)	Low risk	As described in methods
Other bias	Low risk	None apparent