Objective: To identify factors associated with peritoneal dialysis-related infections at one center.

Design: The study was a retrospective study of a 3-year time period with relatively stable treatment patterns.

Setting: A single center experienced academic peritoneal dialysis program.

Patients: Patients (N = 163) receiving peritoneal dialysis (PD) from January 1989 to December 1991 who had received treatment at home for at least one month.

Interventions: None.

Main outcome measures: Catheter-related infection and peritonitis were the main outcome measures. Variables affecting infection rates that were assessed included age, time on PD, prior end-stage renal disease (ESRD) therapy, diabetic status, catheter type, exchange device, nasal carriage of S. aureus, and prophylactic rifampin therapy. Data were analyzed with univariate as well as with a fixed-effects and a mixed-effects gamma-Poisson multiple regression model.

Results: Variables associated with an increased risk of new peritonitis included age under 20 years (p < 0.009; rate ratio 4.54) and nasal carriage of S. aureus (p < 0.04; rate ratio 1.75). Decreased new peritonitis risk was associated with the ULTRA Set exchange system (p < 0.008; risk ratio 0.38) and intermittent prophylactic rifampin therapy (p < 0.001; rate ratio 0.99 for each 1% time on therapy). Catheter-related infections were increased in patients who had double-cuff catheters (p < 0.003) and nasal carriage of S. aureus (p < 0.04; rate ratio 1.82). Decreased catheter-related infections were noted in older patients (p < 0.02; rate ratio 0.983/year) and increasing months of study follow-up (p < 0.03; rate ratio 0.97/month).

Conclusion: In our program nasal carriage of S. aureus increased the risk of peritonitis and catheter-related infection. Prophylactic rifampin significantly decreased peritonitis, as did use of the ULTRA Set. Single-cuff opaque catheters had the lowest catheter infection rate. Analysis of the relationships between clinical and demographic variables and peritoneal dialysis-related infection rate can identify significant contributing or protective variables and allow peritoneal dialysis programs to develop preventive strategies to minimize the risk of infection.