Timing of Dialysis Initiation, Duration and Frequency of Hemodialysis Sessions, and Membrane Flux: A Systematic Review for a KDOQI Clinical Practice Guideline

Background In 2006, NKF-KDOQI (National Kidney Foundation-Kidney Disease Outcomes Quality Initiative) published clinical practice guidelines for hemodialysis adequacy. Recent studies evaluating hemodialysis adequacy as determined by initiation timing, frequency, duration, and membrane type and prompted an update to the guideline. Study Design Systematic review and evidence synthesis. Setting & Population Patients with advanced chronic kidney disease receiving hemodialysis. Selection Criteria for Studies We screened publications from 2000 to March 2014, systematic reviews, and references and consulted the NKF-KDOQI Hemodialysis Adequacy Work Group members. We included randomized or controlled clinical trials in patients undergoing long-term hemodialysis if they reported outcomes of interest. Interventions Early versus late dialysis therapy initiation; more frequent (>3 times a week) or longer duration (>4.5 hours) compared to conventional hemodialysis; low- versus high-flux dialyzer membranes. Outcomes All-cause and cardiovascular mortality, myocardial infarction, stroke, hospitalizations, quality of life, depression or cognitive function scores, blood pressure, number of antihypertensive medications, left ventricular mass, interdialytic weight gain, and harms or complications related to vascular access or the process of dialysis. Results We included 32 articles reporting on 19 trials. Moderate-quality evidence indicated that earlier dialysis therapy initiation (at estimated creatinine clearance [eCl_cr] of 10-14 mL/min) did not reduce mortality compared to later initiation (eCl_cr of 5-7 mL/min). More than thrice-weekly hemodialysis and extended-length hemodialysis during a short follow-up did not improve clinical outcomes compared to conventional hemodialysis and resulted in a greater number of vascular access procedures (very low-quality evidence). Hemodialysis using high-flux membranes did not reduce all-cause mortality, but reduced cardiovascular mortality compared to hemodialysis using low-flux membranes (moderate-quality evidence). Limitations Few studies were adequately powered to evaluate mortality. Heterogeneity of study designs and interventions precluded pooling data for most outcomes. Conclusions Limited data indicate that earlier dialysis therapy initiation and more frequent and longer hemodialysis did not improve clinical outcomes compared to conventional hemodialysis.