Loh 2007a.

Methods	Cluster randomised controlled trial (unit of randomisation = physician).
Participants	Primary care patients newly diagnosed with depression (PHQ). Around two thirds of the participants were female. The average age of the control group was around 41 years (SD 13) and the average age of the intervention group was around 49 years (SD 17)
Interventions	Patient ‐ Decision board for use during consultation that was handed out to the patients to take away.  Printed patient information that combined evidence‐based knowledge about depression care with specific encouragement for patients to be active in the decision‐making process. Physician ‐ Modules on guideline concordant depression care. Enhancing skills for involving patients in the decision making process.  Facilitation practice, role playing and video examples of high quality decision making.  Standardized case vignettes. 5 scheduled training events over a 6 month period.
Outcomes	Patient participation doctor facilitation (PICS‐DF, validated Lerman 1990). Patient participation information seeking (PICS‐IS, validated Lerman 1990). Patient participation Man‐Song‐Hing Scale (Man‐Song‐Hing 1999). Consultation time. Levels of Depression (measured by PHQ 9 Spitzer 1999). Patient satisfaction (ZUF‐8 German version of the CSQ ‐ validated, Schmidt 1989). Patient assessment of treatment adherence (1 question on a 5‐point Likert scale). Physician assessment of treatment adherence (1 question on a 5‐point Likert scale).
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Low risk	Physicians drew blinded lots.
Allocation concealment?	Unclear risk	Randomisation was at physician level. Adequate allocation concealment at the level of the patient would not be possible for new patients, acceptable for existing patients.
Blinding? All outcomes	High risk	Providers were not blinded, participants were blind and outcome assessors and data analysts were not blind (this information was provided by the investigators).
Incomplete outcome data addressed? All outcomes	High risk	Significant loss to follow up.