Effectiveness of cold HD for the prevention of HD hypotension and mortality in the general HD population

Abstract

ABSTRACTBackgroundCold hemodialysis (HD) prevented intradialysis hypotension (IDH) in small, short-term, randomized trials in selected patients with IDH. Whether this treatments prevents IDH and mortality in the HD population at large is unknown.MethodsWe investigated the relationship between dialysate temperature and the risk of IDH, i.e. nadir blood pressure <90 mmHg (generalized estimating equation model) and all-cause mortality (Cox's regression) in an incident cohort of HD patients (n = 8071). To control for confounding by bias by indication and other factors we applied instrumental variables adjusting for case mix at facility level.ResultsTwenty-seven percent of patients in the study cohort were systematically treated with a dialysate temperature ≤35.5°C. Over a median follow-up of 13.6 months (interquartile range 5.2–26.1 months), a 0.5°C reduction of the dialysate temperature was associated with a small (–2.4%) reduction of the risk of IDH [odds ratio (OR) 0.976, 95% confidence interval (CI) 0.957–0.995, P = .013]. In case-mix, facility-level adjusted analysis, the association became much stronger (OR 0.67, 95% CI 0.63–0.72, risk reduction = 33%, P < .001). In contrast, colder dialysate temperature had no effect on mortality both in the unadjusted [hazard ratio (HR) (0.5°C decrease) 1.074, 95% CI 0.972–1.187, P = .16] and case-mix-adjusted analysis at facility level (HR 1.01, 95% CI 0.88–1.16, P = .84). Similar results were registered in additional analyses by instrumental variables applying the median dialysate temperature or the facility percentage of patients prescribed a dialysate temperature <36°C. Further analyses restricted to patients with recurrent IDH fully confirmed these findings.ConclusionsCold HD was associated with IDH in the HD population but had no association with all-cause mortality.