Incident elderly patients on peritoneal dialysis: Epidemiological characteristics and modality impact on survival time

Abstract

Aging of the population and the increased prevalence of diseases such as diabetes and arterial hypertension result in an increasing need of dialysis treatment. Herein we describe a cohort of elderly patients on peritoneal dialysis (PD) and assess the influence of the modality on the long-term survival. Out of a multicenter prospective cohort of 2,144 BRAZPD PD incident patients during a period from December 2004 to October 2007, 762 elderly adults, defined as patients ≥65-year-old, were eligible for the study, 413 started on automated PD (APD) and 349 on continuous ambulatory PD (CAPD). Patients were followed until death, transfer to hemodialysis, recovery of renal function, loss to follow-up, or transplantation. Demographics and clinical data were evaluated at baseline and described as mean ± standard deviation, median, or percentage. Competing risk and time-dependent Cox analysis were performed, having dialysis modality APD] vs. CAPD as a dependent variable, as hazard ratio (HR) is not proportional throughout the therapy time. Mean age was 74.5 ± 6.8 years in APD, 74.6 ± 6.7 in CAPD, 50.8% females in APD, 54.4% in CAPD. The frequently observed comorbidities were diabetes (52.3% in APD and 47% in CAPD) and left ventricular hypertrophy (36.3% in APD and 46.1% in CAPD) whereas 93.6% presented Davies score ≥2. In Cox time-dependent analysis, HR did not show difference up to 18 months HR = 1.11, confidence interval (CI) = 0.85–1.46], but thereafter, APD modality revealed lower risk of mortality (HR = 0.25, CI = 0.0073–0.86), when compared with CAPD. After adjustment for the confounding factors, CAPD presented a higher risk of mortality (HR = 4.50, CI = 1.29–15.64). No differences in survival were observed up to 18 months of therapy; however, beyond 18 months, APD modality was a protection factor.