Twenty-five RCTs (7,703 participants) were included. One study had 3,241 participants and others ranged from to 28 to 591 participants. Three trials were primary prevention and 22 were secondary prevention.
Study quality was limited. Only four trials reported adequate methods of randomisation and allocation concealment. Only six trials reported blinding of outcome assessors for clinical outcomes. Bias because of high or unbalanced losses to follow-up were unclear or high in more than half of the studies. Losses to follow-up ranged from zero to 44%. Eight trials reported more than 60% attendance on available sessions.
Tests did not suggest publication bias, but may have been inconclusive given the clinical variation in the included studies.
Seven trials with moderate or high intensity intervention and three out of 13 trials with low or very low intensity intervention reported significant improvements in self-reported risk behaviour in at least two aspects.
Clinical event rates were low. Compared to control, the intervention was associated with a lower risk of acute myocardial infarction (RR 0.53, 95% CI 0.35 to 0.80, I2=0%,; six trials) and cardiovascular events (RR 0.68, 95% CI 0.50 to 0.93, I2=42%; seven trials). There was no statistically significant effect on mortality (I2=0%; 12 trials), PCI (I2=23%; seven trials) and coronary artery surgery rates (I2=10%; nine trials).
Compared to control, the intervention was associated with no statistically significant or small changes in laboratory and surrogate outcomes (results reported in full).
Subgroup analyses failed to show any clear indications that results were affected by intensity of intervention, disease status and components of the programme.