Carroll 2006.

Methods	Setting: Boston, San Francisco, USA. Recruitment: Subjects were over age of 65 years, unpartnered (i.e. widowed, divorced, never married and not in a relationship), had a telephone in their home and were able to speak and understand English Randomisation: RCT 3 arm. There was no clear definition of peer support, however study met the inclusion criteria of using peers that shared one or several key characteristics of the target population. Peer training: Training was based on the self‐efficacy model.
Participants	132 older adults (over age of 65 years) who had experienced a myocardial infarction.
Interventions	Intervention 1: Peer advisor intervention group. The peer advisor was a 'graduate' of a local cardiac rehabilitation program. The peer advisor was encouraged to share personal experiences and information with subjects during telephone contact but was warned to avoid sharing clinical information or health advice. Subjects assigned to the peer advisor group received a telephone call from the peer advisor once a week for the 12 weeks after discharge from the hospital. Intervention 2: Advanced Practice Nurse (APN) group. The APN provided additional patient information. Subjects assigned to the APN group received a telephone call from the APN once a week for the 12 weeks after discharge from the hospital. Both intervention groups received standard care. Control group: Standard care consisting of discharge instructions provided by the clinical nurse. Discharge instructions included a review of medications, diet, physical activity, symptom management and follow‐up appointments. Caller: Calls made by peer supporters.
Outcomes	Physical health outcomes: health status (Medical Outcome Survey (SF‐36) ‐ self report (validated)). Psychological health outcomes: mental health (Medical Outcome Survey (SF‐36) ‐ self report (validated)). Behavioural health outcomes: Self‐efficacy (Jenkins Self Efficacy Scales ‐ self report (validated), Duke Activity. Status Index Scale ‐ self report (validated). Social health outcomes: none reported. Impact on participants: none reported. Impact on peer supporter: qualitative data. Method of assessing outcome measures: phone survey
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Low risk	Table of random numbers.
Allocation concealment?	Unclear risk	B ‐ Unclear.
Blinding? Participants	High risk	Not done.
Blinding? Providers	High risk	Not done.
Blinding? Outcome assessors	High risk	Not done.
Blinding? Data analysts	High risk	Not done.
Incomplete outcome data addressed? All outcomes	Low risk	Intention‐to‐treat analysis.
Free of selective reporting?	Low risk	All measures reported.