Extracorporeal support modalities are highly prothrombotic. Anticoagulation is frequently used for patients receiving Continuous Renal Replacement Therapy (CRRT), Molecular Adsorbent Recirculating System (MARS), and Extracorporeal Membrane Oxygenation (ECMO). The objective of this systematic review and meta-analysis is to determine if prostacyclin-based anticoagulation strategies are effective compared to other anticoagulation strategies, in critically ill children and adults who needs extracorporeal support, such as continuous renal replacement therapy.

We conducted a systematic review and meta-analysis using multiple electronic databases and included studies from inception to June 1, 2022. Circuit lifespan, proportion of bleeding, thrombotic, and hypotensive events, and mortality were evaluated. Out of 2,078 studies that were screened, 17 studies (1,333 patients) were included. The mean circuit lifespan was 29.7 hours in the patients in the prostacyclin-based anticoagulation series and 27.3 hours in the patients in the heparin- or citrate-based anticoagulation series, with a mean difference of 2.5 hours (95%CI -12.0;16.9, p=0.74, I²=0.99, n=4,003 circuits). Bleeding occurred in 9.5% of the patients in the prostacyclin-based anticoagulation series and in 17.1% of the patients in the control series, which was a statistically significant decrease (LogOR -1.14 (95%CI -1.91;-0.37), p<0.001, I²=0.19, n=470). Thrombotic events occurred in 3.6% of the patients in the prostacyclin-based anticoagulation series and in 2.2% of the patients in the control series, which was not statistically different (LogOR 0.97 (95%CI -1.09;3.04), p=0.35, I²=0.0, n=115). Hypotensive events occurred in 13.4% of the patients in the prostacyclin-based anticoagulation series and in 11.0% of the patients in the control series, which was not statistically different (LogOR -0.56 (95%CI -1.87;0.74), p=0.40, I²=0.35, n=299). The mortality rate was 26.3% in the prostacyclin-based anticoagulation series, and 32.7% in the control series, which was not statistically different (LogOR -0.40 (95%CI -0.87;0.08), p=0.10, I²=0.00, n=390). The overall risk of bias was low to moderate.

In this systematic review and meta-analysis of 17 studies, prostacyclin-based anticoagulation was associated with fewer bleeding events, but with similar circuit lifespans, thrombotic events, hypotensive events, and mortality rates. The potential benefits of prostacyclin-based anticoagulation should be explored in large randomized controlled trials.