Solomon 2004.

Methods	Provider RCT Follow up: providers: DONE patients: DONE Blinded assessment: NOT CLEAR Baseline: DONE Reliable outcomes: NOT CLEAR Protection against contamination: NOT CLEAR Overall quality: LOW
Participants	21 rheumatologists at 1 large academic arthritis practice Country: USA Proportion of eligible providers who participated: 100% Hospital, outpatients Academic/Teaching setting: university‐based Type of targeted behaviour: PRESCRIBING (glucocorticoid‐induced osteoporosis) Complexity of targeted behaviour: LOW
Interventions	1. CME: 1.5 hr educational dinner meeting, mixed format (feedback material discussed) + list of patients with rheumatoid arthritis sent to doctors (feedback) 3 weeks later + reminder guideline 2. No intervention control
Outcomes	Professional practice: % of patients having medication for osteoporosis Patient: none Seriousness of outcome: LOW
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Unclear risk	B ‐ Unclear