A randomized crossover trial of regional anticoagulation modalities for intermittent haemodialysis.

The optimal regional anticoagulation (RA) of dialysis filters in patients at risk of bleeding remains elusive. Inducing hypocalcaemia within the filter by using a calcium-free dialysate has emerged as an easy-to-use heparin-free RA, including in critically ill patients, but comparative studies are lacking. We conducted a multicentre, randomized, crossover trial to compare the efficacy and tolerance of two RAs [heparin-coated membrane (HCM) and calcium-free dialysate (CFD) with calcium reinjection according to ionic dialysance] in patients requiring haemodialysis and at risk of bleeding. During the study period, each patient received two dialysis sessions (one with each RA in a randomly assigned order). The primary endpoint was the proportion of dialysis sessions completed (≥240 min). A total of 94 patients were included in the intention-to-treat analysis, including 16 critically ill patients (17.0%). Coagulation and inflammation parameters, as well as haemodynamic status at baseline, were balanced between groups. Premature coagulation of the filter occurred in 19 (20.9%) HCM sessions compared with 3 (3.2%) CFD sessions. In half of the sessions with premature termination, coagulation occurred before 180 min. The proportion of patients who completed the CFD session while failing to complete the HCM session (n = 17) was significantly higher than the proportion of patients who completed the HCM session while failing to complete the CFD session (n = 1; P < .001). Haemodynamic and metabolic tolerance were not different between groups. In individuals at risk of bleeding, RA with CFD significantly reduced the incidence of premature dialysis termination compared with HCM without safety concerns. ClinicalTrials.gov, NCT03842657.