Thirteen studies with 17,643 participants (range 40 to 4,739) were included in the review: 10,304 (58.4%) who received preoperative statin therapy and 7,339 (41.6%) without statins. Three studies were RCTs, two were prospective observational studies and were eight retrospective studies. Two RCTs (one of which was not a placebo-controlled or blinded trial) scored 3 on the Jadad scale and one scored 5. The average Downs and Black score was 21.9 (range 18 to 28); four studies were rated low quality (scored <21).
The incidence of any type of atrial fibrillation was statistically significantly lower in patients who received preoperative statin therapy compared with controls (unadjusted OR 0.78, 95% CI 0.67 to 0.90; 13 studies). There was evidence of significant statistical heterogeneity (p=0.0010, I2=63%). The number needed to treat was 18.2. Reductions in new-onset atrial fibrillation were reported with statin therapy compared with controls (unadjusted OR 0.66, 95% CI 0.51 to 0.84; nine studies). There was evidence of statistical heterogeneity (p=0.0002, I2=73%). The number needed to treat was 25.1. The results did not alter significantly when only studies that reported adjusted odds ratios were included.
Relative risk reductions were reported in the review.
Investigation of statistical heterogeneity suggested that use of β-blocker and angiotensin-converting enzyme inhibitors were statistically significantly associated with improved statin treatment estimates for new-onset atrial fibrillation, but not for all types of atrial fibrillation. Sensitivity analyses of only studies of low quality significantly altered the results and showed no significant reduction for developing any type of atrial fibrillation (OR 0.82, 95% CI 0.57 to 1.16; four studies).
There was evidence of publication bias for both outcomes.