Bethell 1990.

Methods	Study design: Single centre RCT Country: UK Dates patients recruited: 1 December 1979 to March 1984 Maximum follow up: 5 years
Participants	Inclusion criteria: < 65 yrs post MI; history of chest pain typical of MI, progressive ECG changes, rise and fall in aspartate transaminase concentrations with at least one reading above 40 units/ml. Exclusion criteria: Medical or orthopaedic problems that precluded their taking part in the exercise course; insulin‐dependent diabetes mellitus; atrial fibrillation; on investigator's personal general practice list. N Randomised: total: 200; intervention: 99; comparator: 101 Diagnosis (% of pts): MI: 100% Age (mean ± SD): intervention: 54.2 ± 7.2; comparator: 54.2 ± 7.2 Percentage male: intervention: 100%; comparator: 100% Ethnicity: NR
Interventions	Intervention: Treatment patients entered a three‐month course of three times a week circuit training. Components: exercise only Setting: centre Aerobic exercise: Modality: 8 stage circuit aerobic training. Length of session: NR Frequency: 3 times a week. Intensity: 70‐85% predicted HR max. Resistance training included? weight training. Total duration: 3 months. Co‐interventions: NR Comparator: Patients were given a short talk on the sort of exercise that they might safely take unsupervised. Co‐interventions: NR
Outcomes	Total mortality, CHD mortality, non fatal MI (11 year follow up published in 1999. 5 year follow up data from unpublished material used for meta analysis.)
Source of funding	British Heart Foundation and Wessex Regional Health Authority
Conflicts of interest	NR
Notes	229 patients were randomised; 14 in the intervention group and 15 in control dropped out before the first exercise test due to death, refusal or other problems. Therefore 200 took part in the study. Cardiac mortality of 3% pa, once patients survived to be in the trial. Suggests more severely affected patients were not included. Significant predictors of cardiac death were pulmonary oedema on admission, complications during admission, one or more previous infarcts, increasing age and low initial fitness.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random letter sequence.
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	16% lost to follow up, no description of withdrawals or dropouts.
Selective reporting (reporting bias)	Low risk	All outcomes were reported at all time points.
Groups balanced at baseline	Low risk	“The two groups were comparable in terms of age, presence of Q waves on the electrocardiogram, aspartate transaminase concentration, presence of pulmonary oedema, presence of complications, initial Vo2 max, and time to return to work”.
Intention‐to‐treat analysis conducted	High risk	No.
Groups received same treatment (apart from the intervention)	Low risk	“Control patients were given a short talk on the sort of exercise that they might safely take unsupervised. Treatment patients entered a three‐month course of three times a week circuit training at Alton Sports Centre.”