Mortality After Home Hemodialysis Treatment Failure and Return to In-Center Hemodialysis.

Abstract

RATIONALE OBJECTIVE\nPatients on home hemodialysis (HHD) may eventually return to in-center hemodialysis (ICHD) for clinical, technical or psycho-social reasons. We studied the mortality of patients returning to ICHD after HHD comparing it to the mortality experience among patients receiving HHD and patients receiving ICHD without prior treatment with HHD.\n\n\nSTUDY DESIGN\nRetrospective cohort study.\n\n\nSETTING PARTICIPANTS\nAll patients represented in the Australia and New Zealand Dialysis and Transplant (ANZDATA) Registry who commenced HD during 2005-2015 and were treated for >90 days.\n\n\nEXPOSURES\nICHD and/or HHD, and clinical characteristics at study entry.\n\n\nOUTCOMES\nMortality and cause of death.\n\n\nANALYTICAL APPROACH\nA time-varying multivariate Cox proportional hazards analysis with shared frailty was implemented to explore the association between patient treatment states and mortality. Patients were censored at the time of transplantation and change in treatment modality to peritoneal dialysis.\n\n\nRESULTS\nA total of 19,306 patients initiated HD and were treated for >90 days. The mean age of patients was 60.8y (SD=15.4y), 62% were male and 49% had diabetes. After HHD treatment failure, adjusted mortality was increased compared to continued HHD at 0-30 days (HR 3.93, 95% CI 2.09-7.40 p<0.001), 30-90 days (HR 3.34, 95% CI 1.98-5.62, p<0.001) and >90 days (HR 2.29, CI 1.84-2.85, p<0.001).\n\n\nLIMITATIONS\nCovariates recorded at dialysis initiation. Residual confounding underlying successful initiation of HHD treatment.Observational data lacking detail on cause of HHD treatment failure.\n\n\nCONCLUSIONS\nHHD treatment failure is associated with a significant increase in mortality compared to continued HHD. This risk was present in both the early (first 30 days and 30-90 days) and late (>90 days) periods after HHD treatment failure. Further investigation into the specific causes of treatment failure and death may highlight specific high-risk patients.