Abstract
BACKGROUND AND OBJECTIVES: As a result of improved clinical and quality-of-life outcomes compared with conventional hemodialysis, interest in nocturnal home hemodialysis (NHD) has steadily increased in the past decade; however, little is known about the flow of patients through NHD programs or about patient-specific predictors of mortality or technique failure associated with this modality. This study addressed this gap in knowledge. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: This study included 247 NHD patients of the Canadian Slow Long nightly ExtEnded dialysis Programs (CAN-SLEEP) cohort from 1994 through 2006 inclusive. The association between program- and patient-specific variables and risk for adverse outcomes was determined using uni- and multivariable Cox regression. RESULTS: A total of 14.6% of the cohort experienced death or technique failure. Unadjusted 1- and 5-year adverse event-free survival was 95.2 and 80.1%, respectively. Significant predictors of a composite of mortality and technique failure included advanced age (P < 0.001), diabetes (P < 0.001), central venous catheter use (P = 0.01), and inability to perform NHD independently (P = 0.009) and were adjusted for center effect. Weekly frequency of NHD was not predictive. Age and diabetes remained significant with multivariable analysis (hazard ratio 1.07 and 2.64, respectively). Unadjusted 1- and 5-year technique survival was 97.9 and 95.2%, respectively. Only age was a significant predictor of technique failure. CONCLUSIONS: NHD is associated with excellent adverse event-free survival. This study underscores the importance of modality-specific predictors in the success of home hemodialysis, as well as favorable baseline characteristics such as younger age and the absence of diabetes.