Reprocessing dialysers for multiple uses: recent analysis of death risks for patients

Background. Reprocessing dialysers is a common cost-saving, practice in the USA. It began when patients were treated with bio-incompatible cellulosic membranes that were associated with medical complications, but has continued for economic reasons despite the current use of more biocompatible non-cellulosic membranes. A dialysis services and product provider using primarily its own non-cellulosic membranes recently embarked on a staged programme to stop reprocessing dialysers. Approximately a quarter of 71000 patients had been switched from reuse to single use by July 1, 2001. The transition offered a unique opportunity to re-evaluate death risk associated with the reuse practice. Methods. Patients were classified as reuse or single use as of July 1, 2001. Survival time measurements started on that date (Lag0) and at four 30 day intervals after it (Lag30, Lag60, Lag90 and Lag120). Thus, patients must have been treated in their reuse group after Lago for at least 30, 60, 90 or 120 days, respectively. Survival time was evaluated during 1 year following the lag date using the Cox method in unadjusted, case mix-adjusted and case mix plus other measure-adjusted models. Results. All analyses suggested favourable survival advantage among patients treated with single use dialysers. The differences were statistically significant at all lag times in the unadjusted models but became significant only at later lag times in the case mix-and case mix plus other ineasure-adjusted models. For example, single Lise/reuse hazard ratios in the case mix-adjusted models at Lag0-Lag120 were 0.96 (NS) 0.96 (NS) 0.94 (P=0.02) 0.93 (P=0.02) and 0.92 (P=0.01), respectively. Conclusions. A risk benefit appears associated with abandonment of the dialyser reuse practice, although the benefit may lag behind the change. In the USA, the relative risk burden associated with the reprocessing of dialysers may have changed over time with the evolution of clinical practice.