Carson 1982.

Study characteristics
Methods	Study design: single‐centre RCT; participants randomised 6 weeks post‐admission Country: UK Dates participants recruited: NR (recruited over a 3.5 year period) Maximum follow‐up: 3 years
Participants	Inclusion criteria: MI; diagnosis based on ECG changes and/or elevation of serum glutamic oxaloacetic transaminase or lactic dehydrogenase taken on three consecutive days. Exclusion criteria: > 70 years; heart failure at follow‐up clinic; cardio‐thoracic ratio exceeding 59%; severe chronic obstructive lung disease; hypertension requiring treatment; diabetes requiring insulin; disabling angina during convalescence; orthopaedic or medical disorders likely to impede progress in the gym, personality disorders likely to render participant unsuitable for the course. N randomised: total: 303; intervention: 151; comparator: 152 Diagnosis (% of participants): MI: 100% Age (mean ± SE): intervention: 50.3 ± 0.65; comparator: 52.8 ± 0.67 Percentage male: intervention: 100%; comparator: 100% Ethnicity: NR
Interventions	Intervention: participants attended the hospital gym twice weekly for 12 weeks. They were supervised by a doctor and physical educationalist and full resuscitative equipment was available. The exercises were arranged on a circuit basis and pure isometric exercise was avoided. Components: exercise only Setting: centre Exercise programme modality: exercises arranged on a circuit basis Length of session: NR Frequency: twice per week Intensity: NR Resistance training included? No Total duration: 12 weeks Co‐interventions: none described Comparator: did not attend gym Co‐interventions: none described
Outcomes	Total mortality, non‐fatal MI at 5 months, 1 year, 2 years and 3 years after MI (mean follow‐up 2.1 years)
Source of funding	Department for Health and Social Security Grant
Conflicts of interest	NR
Notes	There appears to be a reduction in mortality in exercise participants with inferior MI.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"randomly allocated"
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	High risk	21% lost to follow‐up; no description of withdrawals or dropouts
Selective reporting (reporting bias)	Low risk	All outcomes were reported at all time points described in the methods.