Rahe 1979.

Methods	Study design: parallel RCT Recruitment: October 1971‐June 1972 (January‐July1973 17 additional post‐MI patients were referred for treatment) Allocation: not reported Blinding: not reported Randomisation: no method described Follow‐up(s): 18 months, 3‐4 years Description: a randomised treatment group and a non‐randomised group of volunteers (not included in the meta‐analysis) received group therapy
Participants	Baseline characteristics Intervention group 1 (randomised) Mean age (SD): 50.9 Sex (male %): 85 Number of participants randomised: 22 Working before CHD: 17 Intervention group 2 (non‐randomised volunteers, not included in the analysis) Mean age (SD): 51.5 Sex (male %): 100 Number of participants: 17 Working before CHD: 13 Control group (CG) Mean age (SD): 55.2 Sex (male %): 94 Number of participants randomised: 22 Working before CHD: 12 Inclusion criteria MI survivors First MI (unequivocally documented) < 60 years Patients eligible to return to work Resided in the San Diego area and planned to remain there for at least 3 years Exclusion criteria: not reported Baseline imbalances: average age: intervention group 1 50.9 years; control group 55.2 years Physically demanding work: unknown Severity of CHD: severe
Interventions	Intervention characteristics 6 group therapy sessions once every 2 weeks, beginning 1 month following hospital discharge Spouses were invited to attend the 2nd session (topic: the contribution of physical and psychological risk factors to CHD) The progression of material: life stress and the onset of MI the contribution of physical and psychological risk factors to CHD coronary‐prone behaviour home problems RTW the beginning of each session often included a didactic presentation of educational material; followed by an active discussion where participants were encouraged to report their experiences with the topic Duration of intervention: 6 sessions; period not clear Providers: the senior study author, with training in both psychiatry and internal medicine, first‐year residents in internal medicine, two hospital corpsmen, one medical student, chief cardiologist Control group The control participants received a regular outpatient medical treatment for post‐MI participants.
Outcomes	Proportion at work at < 6 months (short term): 3 months Proportion at work at 6–12 months (medium term): 6, 12 months Participants working full‐time after > 12 months to < 5 years: 4 years Clinical anxiety (general and cardiac‐specific) Adverse events (mortality, reinfarction, bypass)
Identification	Sponsorship source: by the Naval Medical Research and Development Command, Department of the Navy, under Research Work Unit ZF 51.524.002.‐5020 Country: USA Setting: Naval Regional Medical Centre/US Naval Hospital Possible conflicts of interest: no information provided Ethics committee approval: not reported
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "Twenty‐two patients were randomly assigned to the treatment group and 22 to the control group." No further information provided
Allocation concealment (selection bias)	High risk	No allocation concealment was described, and the study authors report that, "17 additional post‐MI patients who met the research criteria were referred for treatment." Although these participants should have been randomly allocated to the study arms, the study authors explain that these participants joined the study expecting the intervention (and examined as a separate treatment group). The results of these non‐randomised people are nevertheless excluded from the quantitative synthesis of results.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Due to the nature of the intervention (group therapy), blinding of participants and personnel would not have been possible.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No blinding of outcome assessors is described, and it is unclear how exactly RTW was assessed. The study authors state that standardised interviews and research questionnaires were used at the follow‐ups. However, no validated questionnaire or validation with independent (unbiased) occupational records is described.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Loss to follow‐up was low across all groups. Three people died before the 3‐4‐year follow‐up, but the study authors do not describe to which of the three study groups these participants were allocated.
Selective reporting (reporting bias)	Unclear risk	Although the study authors do mention a study protocol, it does not appear to have been published.
Other bias	Unclear risk	None identified