Abstract
Introduction: The clinical outcomes of starting peritoneal dialysis (PD) in kidney failure patients according to different break-in periods are not well established. Our aim was to assess whether the strategy of PD initiation interferes with clinical outcomes over the initial 180 days. Methods: This retrospective study included incident kidney failure patients starting PD at a single center (November 2016–July 2022). Patients were divided into three groups: (1) Urgent-start (US-PD), initiated within 3 days after catheter insertion without prior hemodialysis (HD); (2) Early-start (ES-PD), initiated between 3–14 days, including those with ≤30 days of prior HD; (3) Planned-start (Plan-PD), initiated after 15 days without prior HD. Mechanical and infectious complications, hospitalizations, mortality, and time on PD were compared at 180 days. Patient dropout was defined as the discontinuation of PD due to death or transfer to HD. Results: A total of 211 patients were included: 118 (55.9%) US-PD, 46 (21.9%) ES-PD, and 47 (22.2%) Plan-PD. Among ES-PD patients, 15 (32.6%) had prior HD (<30 days – median time 19 days). Catheter insertion was mostly performed by nephrologists (60.6%) using the modified Seldinger technique (59.2%). Early complications included catheter dysfunction, which occurred in 12.7% of the overall cohort (17.8% in US-PD vs. 4.3% in ES-PD vs. 8.5% in Plan-PD; p = 0.04), and leakage, observed in 7.1% of the overall cohort (9.3% in US-PD vs. 6.5% in ES-PD vs. 2.1% in Plan-PD; p = 0.26). Later complications, hospitalizations, mortality, and time on PD did not differ significantly between groups. Peritonitis, poor education, and hospitalization were associated with dropout. Conclusion: Although initiating PD within 72 h of catheter insertion was associated with more mechanical complications in our study, it resulted in similar clinical outcomes to Planned-start PD patients within the first 6 months of therapy, making it a viable option for urgent dialysis initiation in kidney failure patients.

This publication has 20 references indexed in Scilit: