Background: Controversy surrounds the role of biocompatible membrane dialyzers in treatment of acute renal failure. Studies that have shown a benefit have involved critically ill patients where renal recovery and patient mortality are influenced by other comorbid disease. The aim of the present work is to clarify this issue in a more homogeneous population of patients with acute renal failure following cadaveric renal transplantation.
Methods: All patients with delayed graft function between January 1996 and February 1998 were randomized to receive either a biocompatible (BCM, polysulfone) membrane or bioincompatible (BICM, cuprophane) membrane for dialysis until onset of graft function.
Results: Forty-one patients were randomized, 23 to receive BCM and 18 BICM. Five patients (2 BCM, 3 BICM; p = NS) with primary non-function of graft were excluded from analysis, leaving 36 cases of acute tubular necrosis (ATN). Patient and donor characteristics were similar in both groups. The BCM group had significantly longer periods of dialysis dependency compared to the BICM group (14 vs 10 days; p = 0.03). There was a tendency towards higher serum creatinine levels in the short term in the BCM group (318 vs 164 micromol/l at 1 month (p = 0.1), 190 vs 169 micromol/l at latest visit (p = 0.07)) and a greater number of acute rejection episodes in the BCM group (3.7 vs 1.7 episodes per 100 days of dialysis dependency, p = 0.1). With an intention-to-treat analysis of all 41 patients originally randomized, there was no significant difference in time to graft recovery between the 2 groups (p = 0.18).
Conclusions: In the setting of ARF posttransplantation, we have found no evidence to support the use of biocompatible membranes for dialysis. Rather, our study provides argument against a large benefit for the use of BCM in the recovery of ARF, as suggested by earlier studies.