One hundred and seventeen studies were included in the review: 106 RCTs (range 18 to 2,329 patients; 63.2% of RCTs enrolled less than 100 patients) and 11 observational studies (range 438 to 9,598 patients). Of the included RCTs, 35 had adequate allocation concealment and were double-blinded. The observational studies were judged to be high-quality for measures to protect against selection and detection bias.
The results for mortality (82 RCTs, 11 observational studies; 41,350 patients) showed that compared with aprotinin, tranexamic acid was associated with a reduced risk of mortality (OR 0.64, 95% CrI 0.41 to 0.99); the 95% credible intervals for other comparisons were wide and inconclusive. Tranexamic acid was estimated to have a 73.4% probability of being the lowest risk of treatment, followed by epsilon-aminocaproic acid (24.0%). When cohort studies were incorporated into the analysis, both tranexamic acid (OR 0.71, 95% CrI 0.50 to 0.98) and epsilon-aminocaproic acid (OR 0.60, 95% CrI 0.43 to 0.87) were associated with a reduced mortality risk compared with aprotinin.
For myocardial infarction (67 RCTs, eight observational studies; 26,694 patients) and stroke (40 RCTs, 10 observational; 32,006 patients) the 95% credible intervals were wide and inconclusive for all treatments when compared with aprotinin. Epsilon-aminocaproic acid was associated with the lowest risk for each of these outcomes, followed by tranexamic acid for myocardial infarction and aprotinin for stroke. When cohort studies were incorporated into the analyses, the credible intervals narrowed, but results remained inconclusive and the clinical interpretations of the results remained unchanged.
For renal failure/dysfunction (28 RCTs, nine observational studies; 31,559 patients), the 95% credible intervals were wide and inconclusive for all treatments when compared with aprotinin. Epsilon-aminocaproic acid was associated with the lowest risk, followed by tranexamic acid. When cohort studies were incorporated into the analyses, there was a significant risk reduction with no treatment (OR 0.66, 95% CrI 0.45 to 0.88), tranexamic acid (OR 0.66, 95% CrI 0.48 to 0.91) and epsilon-aminocaproic acid (OR 0.65, 95% CrI 0.45 to 0.88).
Results of sensitivity analyses were also reported.