WHO 1983.

Study characteristics
Methods	Study design: Multicentre RCT (24 sites; 12 centres accepted for meta analysis.) Country: Multiple European countries Dates participants recruited: 1972 to 1974 Maximum follow‐up: 3 years Participants randomised on discharge from hospital.
Participants	Inclusion criteria: Men < 65 years with first or consecutive MI. Exclusion criteria: NR N randomised: total: 3184; intervention: 1655; comparator: 1529 Diagnosis (% of participants): MI: 100% Age (years): intervention: 52.3; comparator:53.5 Percentage male: intervention: 100%; comparator: 100% Ethnicity: NR
Interventions	Intervention: Comprehensive programme dependent on local provision. Physical training was not compulsory but was strongly recommended. Components: exercise, education and psychosocial support. Setting: centre. Exercise programme modality: NR Length of session: NR Frequency: NR Intensity: NR Resistance training included? NR Total duration: 6 weeks. Co‐interventions: The intervention had to be at the highest possible level available locally. It had to be comprehensive, with the aim of improving health and reducing IHD risk. It comprised treatment of heath failure, arterial hypertension etc, risk factor modification, weight loss and improving physical working capacity. Comparator: usual care. Co‐interventions: none described.
Outcomes	Total mortality, CVD, CHD & sudden death. Fatal & non‐fatal re‐infarction.
Source of funding	WHO Regional Office for Europe and the Ministries of Health of the participating member states.
Conflicts of interest
Notes	Methodological problems with the execution of the study allowed only death and re‐infarction to be successfully used as end points.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	"Patients were randomised at admission.....by means of random number tables". However, only "12 centres out of the 24 seemed to have achieved proper randomisation in their groups of R and C patients”
Allocation concealment (selection bias)	Unclear risk	Not reported.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Blinding not described.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	No description of withdrawals or dropouts. Varied greatly from site to site.
Selective reporting (reporting bias)	Low risk	All clinical endpoints were reported for 12, 24 and 36 month follow‐ups.