Willeboordse 2017.

Study characteristics
Methods	Aim of study: to investigate effectiveness of clinical medication reviews (CMRs) for quality of life and geriatric problems in comparison with usual care in older patients with geriatric problems in general practice Study design: cluster‐RCT (22 general practices; cluster unit of allocation: GP practice) Number of arms/groups: 2
Participants	Description: patient/consumer Geographic location: the Netherlands Setting: primary care clinic (general practice clinics) Inclusion criteria: PRACTICE: inclusion = all GP practices members of Academic Network of GPs. Practice not performing CMRs on a regular basis and would not start doing if randomised to control. PARTICIPANTS: inclusion = ≥ 65, newly presented with a geriatric problem in general practice, and used ≥ 1 prescribed drug long term (≥ 3 months). Geriatric problems identified by screening electronic records and by completing screening questionnaire. Geriatric problems included mobility, dizziness, fear of falling, urinary incontinence, and cognitive impairment. Patients were included if they scored ≥ 5 on VAS scales (range 1 to 10) for geriatric problems or reported ≥ 1 fall in preceding 6 months Exclusion criteria: PARTICIPANTS: recorded dementia diagnosis, GP excluded patients who had recent CMR or deemed unable to participate Number of participants randomised: 518 (275 and 243) Number of participants included in analysis: T0 = 270 vs 239, T2 = 215 vs 211 (unpublished adherence results for 208 vs 198) Age: 77.8 ± 7.7 vs 77.8 ± 8.0 Gender: female: 177 (64.4%) vs 159 (65.4%) Ethnicity: born Dutch or other European: 91.7% vs 93.6% Number of medications: number of drugs reported by patient: 6.1 ± 3.1 vs 5.6 ± 3.2 Frailty/Functional impairment: mobility problems (≥ 5 VAS) = 57.9% vs 62.6% Cognitive impairment: cognitive problems (≥ 5 VAS) = 25.5 vs 26.9%, diagnosed dementia excluded Comorbidities: chronic diseases: 2.77 ± 1.76 vs 3.23 ± 2.19
Interventions	Group 1 ‐ Optimised clinical medication reviews (Opti‐Med): (1) preparation: info from EMRs, pharmacy, and screening questionnaire collected including drug use, medication history, potential DRPs, medical problems, recent lab results, and non‐lab measurements; (2) clinical medication review: expert team of GPs/nursing home physicians and community pharmacists performed review using adapted systematic tool to reduce inappropriate prescribing (STRIP) method; (3) pharmacotherapeutic treatment plan (PTP): PTP sent to patient's GP; (4) implementation of PTP: patients invited for consultation with GP in which PTP was discussed and was determined together with the patient Group 2 ‐ Usual care: expert team also performed CMR analyses, but GPs and patients did not receive the results Co‐intervention: N/A Provider: independent expert team (doctor and pharmacist) and patient's GP Where: primary care clinic When and how often: once ‐ baseline Intervention personalised: yes
Outcomes	Timing of outcome assessment: baseline and 6 months Medication adherence (subjective): self‐reported adherenceproblems assessed in the screening and follow‐up questionnaire Satisfaction with intervention (subjective): Medication Satisfaction Questionnaire: assessed on a 7‐point Likert scale. A 1‐point change in MSQ score was considered clinically meaningful Health‐related quality of life (subjective): SF‐12 and EQ‐5D‐3L Adverse clinical health outcomes (objective): drug‐related problems: number of DRPs per patient ‐ using the DOCUMENT checklist. Assessed by expert team at baseline and by 1 researcher at follow‐up
Notes	Trial registration: NTR4264 Consumer involvement: not specified Funding source: Dutch Organisation for Health Research and Development Dropout: 9 excluded before intervention (5 and 4), 51 withdrew (33 and 18), 42 were lost to follow‐up (27 and 15) Fidelity: 274 of 275 received CMR; 247 discussed with patient Unpublished data: adherence worsened or persisted: 65 vs 54; adherence improved or remained the same: 143 vs 144 ICC value: 0.08. Effective sample sizes were not calculated, as this study was not included in any meta‐analyses due to alternate reporting methods
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation of practices performed by statistician blinded to characteristics of practices using a computer‐generated list of random numbers
Allocation concealment (selection bias)	High risk	Randomisation done at practice level before patients recruited ‐ could have influenced recruitment of patients
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding, but blinding to treatment allocation not possible due to nature of the intervention
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Subjective outcomes unblinded; unclear if researcher blinded
Incomplete outcome data (attrition bias) All outcomes	High risk	Higher dropout and loss to follow‐up in intervention group: 51 withdrew (33 and 18), 42 were lost to follow‐up (27 and 15)
Selective reporting (reporting bias)	Unclear risk	Raw data not reported
Other bias	Unclear risk	Sample size 500 patients, not maintained during follow‐up. Based on EQ‐5D. 274 of 275 received CMR, 247 discussed with patient Recruitment bias (selective recruitment of cluster participants): high risk, randomisation carried out before patients were recruited, thus potential for selective recruitment based on practice knowledge of being in intervention or control group