Thirty-six RCTs (n=4,360) were included in the meta-analysis. Fourteen RCTs were judged as high quality (Jadad score of at least 3).
Compared with controls, biocompatible circuits were significantly associated with a reduction in packed red cells transfusions (OR 0.8, 95% CI 0.69 to 0.93; 20 RCTs), atrial fibrillation (OR 0.76, 95% CI 0.61 to 0.93; eight RCTs), intensive care unit stay (SMD -0.25 days, 95% CI -0.37 to -0.14; 13 RCTs) and hospital stay (SMD -0.55 days, 95% CI -0.68 to -0.42; nine RCTs). No significant differences between the two groups were observed for resternotomy, perioperative myocardial infarction, pulmonary complications, stroke and hospital mortality.
Subgroup analyses of high-quality studies showed that results of a reduction in atrial fibrillation and intensive care unit stay remained significant; biocompatible circuits were no longer significantly associated with a reduction in packed red cells transfusions compared with controls.
Statistically significant heterogeneity (p<0.1) was observed only for the outcomes of mechanical ventilation time (I2=91%), intensive care unit stay (I2=89%) and hospital stay (I2=90%). The authors reported that publication bias was found only for the outcome of mechanical ventilation time. Funnel plots and results of Egger tests were not presented.
Other outcomes that were not included in meta-analyses were reported descriptively.