Moral 2015.

Study characteristics
Methods	Aim of study: to determine whether a face‐to‐face communicative strategy based on motivational interviewing (MI), used by health practitioners (family physicians and nurses) in a primary care setting, and aimed at patients over 65 years old with a chronic disease who are being treated by polypharmacy and who have poor medication adherence can achieve better results than the usual approach based on an informative model of providing education and advice Study design: cluster‐RCT (2 arms, 16 health centres, stratified by professional) Number of arms/groups: 2
Participants	Description: patient/consumer Geographic location: Spain Setting: primary care clinic (health centre) Inclusion criteria: > 65, chronic disease, polypharmacy (≥ 5 medicines or ≥ 12 daily doses for a period ≥ 6 months), high probability for non‐adherence (Haynes‐Sackett yes, inconsistent answers to at least 1 of the 4 Morisky‐Green Qs) Exclusion criteria: serious psychiatric and neurological diseases, difficulties coping with basic daily activities (Barthel Index < 60), those who had cognitive impairment (Pfeiffer's test), those admitted to hospital at least twice in last year, patients under carer's supervision Number of participants randomised: 32 (16 and 16) health professionals, 70 vs 84 patients Number of participants included in analysis: 66 vs 81 (but included in analysis 70 vs 84) Age: 75.6 ± 5.9 vs 76.1 ± 5.8 Gender: female: 49 (70%) vs 57 (67.9%) Ethnicity: not specified Number of medications: medication consumption: 8.7 ± 2.5 vs 9.0 ± 3.1 Frailty/Functional impairment: not specified; < 60 Barthel Index ADLs excluded Cognitive impairment: cognitive impairment excluded Comorbidities: mean ± SD: chronic disease: 4.9 ± 2.1 vs 5.1 ± 2.6
Interventions	Group 1 ‐ Motivational interviewing (MI): intervention health professionals attended an additional 20‐hour workshop taught by family doctor who is expert in the field. Intervention professionals focused on motivational interviewing. Strategies of EMot are based on a collaborative, evocative style and respect for autonomy of the patient. The practice of EMot is based on 4 basic principles grouped under the acronym RULE: R (resist) resist the redirect reflex, U (understand) understand and explore the motivations of the patient himself, L (listen) listen empathically, and E (empower) empower the patient, favouring hope and optimism Group 2 ‐ Usual care: control patients received routine clinical attention based on transmission of info and persuasive advice Co‐intervention: before intervention, healthcare providers in both groups attended a 15‐hour workshop on patient safety and medication adherence. Intervention: (1) initial assessment of medication status, (2) detection of critical incidents and possible medication errors, (3) providing information, (4) developing customised action plan, (5) proposal for implementation Provider: physician or nurse (trained health professionals ‐ 16 physicians and 11 nurses) Where: patient's home or health centre When and how often: V0 baseline in healthcare setting (15 m), V1 at 15 to 20 days at home (45 to 60 m), V2 at 3 months in healthcare setting (15 m), V3 at 6 months at home (45 to 60 m) Intervention personalised: yes
Outcomes	Timing of outcome assessment: baseline and 6 months Medication adherence (objective):pill count: number of tablets presumably consumed/Number of tablets that should be consumed × 100. Adherent if average adherence > 80% and < 110% Other (objective) : average medication errors according to group. Errors including subtherapeutic dose, omission of administration, deteriorated drug, duplicate therapy, higher doses and other (reported in Perula de Torres 2014 paper)
Notes	Trial registration: NCT01291966 Consumer involvement: not specified Funding source: supported by Spanish Society of Family and Community Medicine and Andalusian Society of Family and Community Medicine research grant, and the Ministry of Health of the Government of Andalusia, Spain Dropout: 5 (3 vs 2) health professionals did not recruit any participants, 2 patients withdrew, 4 were lost to follow‐up (Perula de Torres paper says 5 lost to follow‐up) Language translation: yes ‐ Perula de Torres 2014 paper translated to English Exact ICC value not reported. Paper states that "ICC in cRCT in primary care generally less than 0.05". Thus 0.05 was used to recalculate sample sizes, 57 intervention vs 67 control
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"We performed blinded randomization to one of the two study arms (C4‐Study Design Pack; Glaxo S.A.)" 32 professionals assigned randomly and stratified by type of professional
Allocation concealment (selection bias)	Unclear risk	Allocation based on clusters and stratified by profession (nurse or physician). Unclear if/how allocation was concealed
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not possible to mask the intervention, either to patients or to providers
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear if assessors blinded. Final results were evaluated by a methodology expert of the investigation, who remained at all times blind to the status of patients
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Unclear who is in final results; states ITT
Selective reporting (reporting bias)	Unclear risk	Unclear how they assessed medication adherence at baseline (i.e. how did they do pill count). Medication adherence mean only at baseline
Other bias	Unclear risk	Sample size calculation: 78 per group not reached Recruitment bias (selective recruitment of cluster participants): high risk. Participants recruited by consecutive sampling. "Time between the training program and patient recruitment and intervention was about two weeks"; thus health professionals were aware of randomisation during participant recruitment stage