Incremental Hemodialysis Practices and Impact on Survival: Systematic Review and Meta-Analysis

Background: Starting hemodialysis (HD) incrementally may help reduce the burden of treatment in patients with established kidney failure. A systematic review has been conducted to describe variations in the practice of incremental HD worldwide and to study its impact on mortality. Study Design: Systematic review. Setting & Population: Patients with established kidney failure starting HD. Selection Criteria for Studies: Medline and Academic Search Premier were searched from inception to July 20, 2020, for observational and interventional studies comparing incremental or twice weekly treatments with conventional treatment, with mortality as one of the outcome measures. Predictor: Incremental (or twice-weekly) HD therapy versus thrice-weekly treatments. Outcomes: Mortality. Results: Fourteen studies were included (combined 91,928 participants; 5,075 [5.5%] in the intervention, 86,853 [94.5%] in standard treatment groups). Large variations in the practice of incremental HD were noted with treatments differing in HD frequency, treatment goals, monitoring schedules, duration of the incremental program, and cointerventions. Larger studies with lowest risk of bias demonstrated equivalent survival between incremental and conventional treatment groups. Meta-analysis of mortality hazards showed an overall hazard ratio of 0.97 (95% CI 0.76-1.19). Centers which screened patients for adequacy of residual kidney function at baseline and pursued prespecified treatment goals have demonstrated better outcomes in incremental HD recipients compared with conventional treatment. There is evidence of publication bias in the literature. Limitations: Studies from diverse settings, searches limited to English language. Conclusions: There is wide variation in the interpretation of incremental HD treatments and therefore primary studies of incremental HD must be examined in the context of their settings, population, and available resources. Optimal method of implementing incremental HD remains controversial; available data indicate that incremental HD is at least noninferior to conventional HD.