Bivalirudin versus heparin in cardiac surgery remains an important question when conventional heparin and protamine cannot be used. A systematic review and meta-analysis by Cadd and colleagues brings together three randomized trials involving 407 adults. The pooled results detected no statistically significant differences in reexploration for bleeding, major thromboembolic events, mortality, or transfusion requirements. However, the small evidence base and wide confidence intervals leave substantial uncertainty about comparative benefits and harms.
What the Review Examined
The authors searched electronic databases and additional literature sources from January 1990 through January 2026. The review was registered with PROSPERO (CRD420251274233) and reported using PRISMA guidance. Eligible trials compared bivalirudin with unfractionated heparin in adults undergoing cardiac surgery, with or without cardiopulmonary bypass.
Three randomized studies met the inclusion criteria. Two studied elective off-pump coronary artery bypass surgery, while the remaining trial included cardiac operations performed with cardiopulmonary bypass. Participants were predominantly men, and mean ages across trials were approximately 64–65 years. This mix matters: pooled findings across on-pump and off-pump procedures should not be interpreted as equally precise evidence for every perfusion setting.
Bleeding and Reexploration
The primary outcome was reexploration for hemorrhage. The pooled risk ratio for bivalirudin compared with heparin was 1.38 (95% confidence interval 0.42–4.48). The interval crosses the null value and is wide enough to include both a reduction and a substantial increase in risk. The result therefore indicates that a difference was not demonstrated; it does not establish equivalent bleeding safety.
Postoperative blood loss could not be pooled because the included studies reported medians and interquartile ranges. The individual trial findings also depended on the assessment time. In the on-pump trial, bivalirudin was associated with greater blood loss at two hours, although the authors reported no significant difference at 24 hours. This early signal deserves attention even though the pooled reexploration endpoint was not statistically significant.
Transfusion, Thromboembolism, and Mortality
The review did not detect statistically significant differences in the reported pooled transfusion outcomes, including packed red cells, platelets, and fresh frozen plasma. It likewise found no significant differences in the reported major thromboembolic outcomes or mortality. Rare clinical events and the limited number of participants constrain how confidently these results can exclude clinically important differences.
Transfusion findings also require caution because postoperative transfusion triggers were not standardized. Intraoperative blood-product reporting was incomplete in two studies, and one trial reported different use of antifibrinolytic drugs between groups. Such differences complicate interpretation of a direct anticoagulant effect on bleeding or transfusion.
Relevance for Perfusion Practice
The paper supports continued consideration of bivalirudin as an alternative when heparin is unsuitable, rather than establishing a reason to replace routine heparin–protamine anticoagulation. Its clinical discussion focuses on circumstances such as heparin-induced thrombocytopenia, heparin resistance, and intolerance of heparin or protamine. Those indications should be distinguished from the populations and outcomes directly tested in the pooled randomized trials.
The authors emphasize practical differences between the anticoagulants, including the absence of a specific reversal agent for bivalirudin and variability in dosing and anticoagulation monitoring. Their review is not a validated dosing protocol. For perfusion teams, the unresolved questions concern how best to standardize administration, monitoring, and management at separation from bypass while accounting for the individual clinical setting.
Limitations and Interpretation
All three included trials were judged to have some concerns for risk of bias. Two were open label, and treatment crossover after randomization occurred in two studies. The small number of trials prevented the planned publication-bias assessment and leave-one-out sensitivity analysis. Low statistical heterogeneity cannot resolve these limitations or the imprecision of the pooled estimates.
The most useful conclusion is therefore restrained: available randomized evidence has not demonstrated a clear clinical advantage or disadvantage for bivalirudin compared with heparin in the studied adult cardiac surgical populations. Larger, well-designed trials are needed to clarify bleeding and thrombotic risks, define consistent monitoring strategies, and distinguish results for on-pump and off-pump procedures.
Original Article
Cadd M, Green S, Puntis D, van Huellen H, Seavill M, Bullard S. Bivalirudin Versus Heparin and Protamine in Adult Patients Undergoing Cardiac Surgery: A Systematic Review and Meta-Analysis. Annals of Cardiac Anaesthesia. 2026;29:494–502. Published October 12, 2026. doi:10.4103/aca.aca_28_26.
The authors reported no financial support or sponsorship and no conflicts of interest.
