Nineteen studies (31,725 participants) were included: three RCTs (317 participants); three prospective observational studies (1,185 participants); and 13 retrospective observational studies (30,223 participants). Study quality was judged moderate to good: Jadad score from 1 to 5 for RCTs; Downs and Black Checklist 22 to 28 for RCTs and 17 to 23 for observational studies.
Preoperative statins lowered the risk of short term mortality (OR 0.57, 95% CI 0.49 to 0.67; no significant heterogeneity p=0.30, I2=14.6%; 10 studies), stroke (OR 0.74, 95% CI 0.60 to 0.91; no significant heterogeneity p=0.10, I2 = 45.1%; seven studies) and atrial fibrillation (OR 0.67, 95% CI: 0.51 to 0.88; heterogeneity present, p=0.003, I2=69.9%; seven studies).
There was no effect on myocardial infarction or renal failure.
When characteristics of included participants were investigated, pooled preoperative prevalence showed that people who received statins were more likely to be: younger; male; have had previous myocardial infarction, diabetes or hyperlipidaemia; and receive betablockers or aspiring before surgery. Those who did not receive statins were more likely to have had non-elective surgery and cardiopulmonary bypass.
Tests showed no evidence of publication bias.