Ultrafiltration rate and mortality in patients undergoing extended-hours haemodialysis

Clinical guidelines recommend maintaining ultrafiltration rate (UFR) below 10-13 ml/h/kg for patients undergoing conventional haemodialysis to reduce adverse outcomes. However, achieving this target may compromise nutritional status, particularly in underweight patients. We aimed to explore the association of the interaction between UFR and body weight with mortality.

We analysed the LIBERTY cohort data to assess the association between UFR (net ultrafiltration volume divided by session length) and mortality. We employed three multivariable Cox regression models: baseline (long-term effects), time-dependent (short-term effects), and time-averaged (cumulative effects). Contour plots examined the UFR and post-dialysis weight interactions. Dialysis and laboratory parameters were electronically collected and averaged quarterly.

Among 614 patients (mean age 62 years; 65% male) undergoing extended-hours haemodialysis, the median net ultrafiltration volume was 3.0 (IQR, 2.4-3.7) kg/session, with dialysis session length of 21 (18-24) hours/week. The median scaled and unscaled UFR were 7.4 (6.0-9.1) mL/h/kg and 434 (334-531) ml/h, respectively. Over a median follow-up of 6.2 (3.5-10.1) years, 225 patients died. The association between UFR and mortality was significantly modified by post-dialysis weight. Higher-weight patients showed increased mortality with higher UFR in baseline and time-averaged models, while lower-weight patients showed poorer outcomes with lower UFR in time-dependent models. These findings were particularly pronounced in older patients.

UFR-mortality association is significantly modified by body weight in extended-hours haemodialysis patients. These findings highlight individualized UFR management needs and potential nutritional intervention benefits for underweight patients.