Sixteen RCTs (884) were included in the review. Sample size ranged from 19 to 200. Jadad score ranged from 1 to 5. Ten trials had a Jadad score of at least three. Six studies reported allocation concealment; four studies adequately described method of randomisation; eight studies were double-blind, two studies were single-blind; thirteen studies ascertained outcomes blindly; fifteen studies described loss to follow-up; and no studies used intention-to-treat analysis.
Compared with placebo, trimetazidine therapy was found to have significantly improved left ventricular ejection fraction (WMD 6.46%; 95% CI 5.20% to 7.73%; 14 RCTs; Ι²=71%). There was no significant difference in all-cause mortality between the trimetazidine and placebo groups.
Compared with placebo, trimetazidine therapy was associated with a significant reduction in hospitalisation due to cardiac causes (RR 0.43, 95% CI 0.21 to 0.91; four RCTs; Ι²=0%), left ventricular end-systolic diameter (WMD -6.67 mm, 95% CI -9.29 to -4.06; three RCTs; Ι²=95%), left ventricular end-diastolic diameter (WMD -6.05 mm, 95% CI -7.57 to -4.52; three RCTs; Ι²=56%), left ventricular end-systolic volume (WMD -0.61, 95% CI -1.14, -0.09; six RCTs; Ι²= 80%), New York Heart Association functional class (WMD -0.57, 95% CI -0.88 to -0.26; seven RCTs; Ι²=75%) and B-type natriuretic peptide (WMD -203.40pg/ml, 95% CI -308.99 to -97.81; two RCTs; Ι²=63%).
Results for left ventricular end-diastolic volume, resting systolic and diastolic blood pressure, high-sensitivity C-reactive protein, and corrected QT interval were not statistically significant.
Results for the meta-regression analysis were also reported. There was no evidence of publication bias.