The centre-calculated cutoff value is better for identifying fast peritoneal solute transfer of patients on peritoneal dialysis than the traditional value: a retrospective cohort study

Abstract

Background The mean 4-h dialysate to plasma ratio of creatinine (4-h D/Pcr) is a vital cutoff value for recognizing the fast peritoneal solute transfer rate (PSTR) in patients on peritoneal dialysis (PD); however, it shows a noticeable centre effect. We aimed to investigate our centre-calculated cutoff value (CCV) of 4-h D/Pcr and compare it with the traditional cutoff value (TCV) (0.65). Methods In this study, we enrolled incident PD patients at our centre from 2008 to 2019, and divided them into fast or non-fast PSTR groups according to baseline 4-h D/Pcr–based CCV or TCV. We compared the efficiency of the fast PSTR recognized by two cutoff values in predicting mortality, ultrafiltration (UF) insufficiency and technical survival. Results In total, 1905 patients were enrolled, with a mean 4-h D/Pcr of 0.71 ± 0.11. Compared with TCV (0.65), CCV (0.71) showed superiority in predicting mortality of PD patients [hazard ratio (HR) 1.27, 95% confidence interval (CI) 1.02–1.59 vs HR 1.24, 95% CI 0.97–1.59]. The odds ratio (OR) of the fast PSTR in centre classification was slightly higher than traditional classification in predicting UF insufficiency (OR 1.67, 95% CI 1.25–2.24 vs OR 1.60, 95% CI 1.15–2.22). Additionally, the restricted cubic splines 4-h D/Pcr has an S-shaped association with mortality and UF insufficiency, and the inflection points of 4-h D/Pcr were 0.71 (equal to CCV). Conclusions The CCV of 4-h D/Pcr for identifying fast PSTR was 0.71. It was superior to TCV in predicting mortality and UF insufficiency.