Home Hemodialysis and Mortality Risk in Australian and New Zealand Populations

Abstract

There is a resurgence of interest in home hemodialysis (HD), especially frequent or extended forms involving unconventionally frequent (>3 times/wk) and/or long (>6 hours) treatments. This resurgence is driven by cost containment and experience suggesting lower mortality risk compared with facility HD and peritoneal dialysis (PD). We performed an observational cohort study using the Australia and New Zealand Dialysis and Transplant Registry, using marginal structural modeling to adjust for time-varying medical comorbidity as both a source of selection bias and an intermediary variable on the causal pathway to death. All adult patients starting renal replacement therapy in Australia and New Zealand since March 31, 1996, followed up to December 31, 2007. The main predictor was dialysis modality (conventional facility HD, conventional home HD, frequent/extended facility HD, frequent/extended home HD, and PD). We adjusted for the confounding effects of patient demographics and comorbid conditions. Patient mortality. We analyzed 26,016 patients with 856,007 patient-months of follow-up. Relative to conventional facility HD, adjusted mortality HRs were 0.51 (95% CI, 0.44-0.59) for conventional home HD, 1.16 (95% CI, 0.94-1.44) for frequent/extended facility HD, 0.53 (95% CI, 0.41-0.68) for frequent/extended home HD, and 1.10 (95% CI, 1.06-1.16) for PD. The apparent benefit of home HD on mortality risk was less for patients who were nonwhite, non-Asian, and older. Potential for residual confounding from the limited collection of comorbid conditions (no collection of cognitive or motor impairment, depression, left ventricular volume or structure, or blood pressure/fluid volume status) and lack of socioeconomic, medication, and biochemical data in analyses. Our study supports a survival advantage of home HD without a difference between conventional and frequent/extended modalities. Suitably designed clinical trials of frequent/extended HD are needed to determine the presence and extent of mortality benefit with this modality.