Long-Term Predictors of Mortality in Peritoneal Dialysis Patients

Abstract

Objective: Peritoneal dialysis (PD) is one of the essential treatments for end-stage renal disease. Cardiovascular disease and peritonitis are the most common causes of death among PD patients. This study aimed to explore the long-term mortality predictors of PD patients. Methods: One hundred forty-one PD patients were retrospectively included in the study. Patients were selected among cases admitted to the PD outpatient clinic between January 2015 and November 2020. Clinical and laboratory findings obtained at the first visit were recorded. Prevalence of all-cause mortality and associated prognostic factors were analyzed. Results: The mean age of the patients was 52.5 +/- 15.2 years (range 19-86). Thirty-two patients (22.7%) died in the followup. Age, hemodialysis (HD) history, diabetes mellitus (DM), ultrafiltration, urine volume, glomerular filtration rate, and normalized protein catabolic rate in modified peritoneal equilibrium test were significantly different between survived and non-survived patients. Non-survived patients had higher ferritin, C-reactive protein (CRP), and CRP to albumin ratio (CAR) (P:.003, P:<.001; P: <.001; respectively). The Cox regression analysis revealed that the presence of DM HR 95%CI: 3.755 (1.703-8.280), P:.001, HD history HR 95%CI: 2.843 (1.291-6.263), P:.010, and higher CAR ratio HR 95% CI: 4.235 (1.857-9.662), P:.001 were independent predictors of all-cause mortality. ROC curve analysis demonstrated that CAR > 1.94 predicted all-cause-mortality with 73% sensitivity and 71% specificity (AUC = 0.789, 95%CI: 0.697-0.881). Conclusion: The present study demonstrates that DM, HD history, and high CAR are independent predictors of all-cause-mortality among PD patients.