Type 2 diabetes mellitus and clinical outcomes in peritoneal dialysis: a retrospective cohort study in a Peruvian public Hospital

Joseph V. Pezo-Medina, Nephrology Service, Department of Internal Medicine, Tarapoto’s Hospital-MINSA, Tarapoto, Peru; Faculty of Human Medicine, National University of San Martín, Tarapoto, Peru;
Isabel Saravia, Department of Medical, Vantive, Bogotá D.C., Colombia
Susan Martínez, Department of Epidemiology, EpiThink Health Consulting, Bogotá D.C., Colombia

Introduction: To examine whether type 2 diabetes mellitus (T2DM) is associated with all-cause mortality in peritoneal dialysis (PD), and to assess technique failure and first peritonitis. Material and methods: Retrospective cohort in a public Peruvian hospital (2020–2024). Mortality was analyzed with Cox models (crude, age/sex-adjusted) and a prespecified clinical Cox model additionally adjusting for cardiovascular disease and peritoneal transport status (PET; PET0 as “not available”) using complete cases. Technique failure was assessed with Fine–Gray (death as competing event), and peritonitis with Poisson regression (log person-time offset). Results: Among 123 patients (61% with T2DM), 39 deaths occurred (29/75 vs 10/48). Cox estimates were imprecise (HR 1.49, 95% CI 0.72–3.08 crude; HR 1.59, 0.71–3.53 age/sex-adjusted). In the clinical model (n = 114; 36 deaths), the HR was 0.86 (0.37–2.00; p = 0.729). No clear differences were observed for technique failure (SHR 0.71, 0.28–1.79 crude; 0.86, 0.26–2.84 adjusted) or peritonitis (RR 0.91, 0.50–1.73). Conclusion: Crude mortality was higher in T2DM, but adjusted estimates were imprecise; clinically relevant effects in either direction cannot be excluded.

Keywords: Peritoneal dialysis. Diabetes mellitus type 2. Mortality. Kidney failure chronic. Proportional hazards models. Competing risks.