Long-term mortality and technique survival in peritoneal dialysis patients: a 25-year retrospective analysis in a single center.

Abstract

Background: We comprehensively assessed patient survival and the duration of peritoneal dialysis (PD) treatment over a 25-year period within our PD unit.

Methods: We retrospectively evaluated 497 PD patients who initiated PD between 1996 and 2021. The cohort was divided into three distinct periods based on pivotal events, such as the introduction of more biocompatible dialysis solutions and the initiation of remote monitoring technologies. Kaplan-Meier survival assessments, Cox proportional hazards model and Gray subdistribution hazard model were employed to evaluate patient survival and PD-to-hemodialysis (HD) transfer.

Results: The use of PD as the initial treatment increased significantly over the years. Mean age was 63.5 ± 15.7 years; 61% were male, and 61% had three or more comorbidities including hypertension (81%), dyslipidemia (66%), cardiovascular disease (56%) and diabetes (16%). The five-year mortality rate was 40%. Risk factors for mortality included continuous ambulatory peritoneal dialysis (CAPD) [hazard ratio (HR) = 2.63, 95% confidence interval (CI) 1.76-3.93; P < .001], older age (HR = 2.96, 95% CI 1.98-4.43; P < .001), cardiovascular disease (HR = 1.96, 95% CI 1.31-2.95; P = .001) and the use of renin-angiotensin-aldosterone system inhibitors (RAASi) (HR = 1.81, 95% CI 1.22-2.70; P = .004). At 5 years, 48% of patients remained on PD. In the Cox model, risk factors for PD-to-HD transfer included CAPD (HR = 1.62, 95% CI 1.21-2.16; P = .001). RAASi use (HR = 0.66, 95% CI 0.46-0.94; P = .02) and female sex (HR = 0.70, 95% CI 0.51-0.96; P = .03) were associated with longer PD duration.

Conclusions: The study provides insights into the changing landscape of PD. Advances in PD solutions and remote monitoring have contributed to changes in PD outcomes and its increased adoption over the years. Given the observational nature of the study, caution is warranted in interpreting the association of both CAPD and RAASi with mortality.