Hemodiafiltration versus High-flux Hemodialysis and Risk of Mortality
Giovanni F.M. Strippoli, Giovanni Tripepi, Bernard Canaud, Stefano Stuard, Franklin W. Maddux, Len A. Usvyat, Paola Carioni, Matteo Savoia, Germaine Wong, Carmine ZoccaliBackground:
Randomized trials suggest that high-volume post-dilution hemodiafiltration (HDF) may improve survival compared with high-flux hemodialysis (HD), but evidence from routine clinical practice and from regions underrepresented in trials remains limited.
Methods:
We emulated a target trial comparing HDF with high-flux HD using data from EuCliD®, a multinational registry of dialysis patients treated in eight European countries. Adults receiving thrice-weekly in-center dialysis between 2014 and 2019 were eligible. Follow-up began 91 days after dialysis initiation. Sustained treatment strategies were defined as receipt of the assigned modality for at least 90% of sessions. Inverse probability weighting was used to emulate randomized treatment assignment. The primary outcome was all-cause mortality, with kidney transplantation treated as a competing event.
Results:
Among 19,539 eligible patients at day 91, inverse probability weighting created a weighted pseudo-population of 19,758 patients (8,641 HDF; 11,117 HD). During a median follow-up of 16 months (interquartile range 6–32), 4,282 deaths occurred. Hemodiafiltration was associated with a lower risk of all-cause mortality compared with high-flux HD (hazard ratio 0.72; 95% confidence interval 0.67–0.77). At 2 years, the weighted cumulative incidence of death was 20.6% in the HDF group and 22.3% in the HD group, corresponding to an absolute risk reduction of 1.7 percentage points. Results were consistent across sensitivity analyses, including analyses accounting for country, competing risks, informative censoring, protocol adherence, and an ITT-like exposure definition. The association was broadly similar across prespecified subgroups, with only a stronger relative benefit observed among patients with pre-existing cardiovascular disease (interaction p<0.001). Higher delivered convective volumes were associated with greater survival benefit; however, these findings should be interpreted cautiously, as higher convective volumes may reflect patient stability and center expertise rather than a causal dose–response relationship.
Conclusions:
In this large multinational target trial emulation, sustained high-volume post-dilution hemodiafiltration was associated with lower mortality than high-flux hemodialysis.