Peritonitis in an urban peritoneal dialysis program: An analysis of infecting pathogens

We have previously found that race, level of education, and peritoneal dialysis system are factors that significantly and independently influence peritonitis rates in our patient population. We now extend these observations by assessing the pathogens responsible for peritonitis in these subgroups. Between January 1, 1981, and May 15, 1993, 248 peritoneal dialysis patients underwent dialysis at our facility. The rate of peritonitis by pathogen was determined in these patients using the fixed effects Poisson model. Total peritonitis rates in black patients (1.89 episodes/patient-year) were significantly greater compared with white patients (1.11 episodes/patient-year; P < 0.0001). Increased infection rates in black patients were significant for Staphylococcus epidermidis, Staphylococcus aureus, and gram-negative pathogens. The level of education had a negative correlation with peritonitis rates (≤ 8 years, 2.00 episodes/patient-year; 9 to 12 years, 1.64 episodes/patient-year; and ≥ 13 years, 1.24 episodes/ patient-year) with patients having ≥ 13 years of education at the start of dialysis demonstrating a significantly lower total peritonitis rate compared with patients with 9 to 12 years (P = 0.001) or ≤8 years (P < 0.001) of education. This was accounted for by a significant decrease in infection rates for S epidermidis, polymicrobial, and gram-negative organisms. Finally, patients on automated peritoneal dialysis had significantly lower total peritonitis rates (0.59 episodes/patient-year) compared with patients on either a connect (2.11 episodes/patient-year) or disconnect (1.46 episodes/patient-year) system. A significant decrease in infection rates for every category of pathogen except fungal infections was found in patients on automated peritoneal dialysis in comparison to the other two systems. Overall, significant differences in S epidermidis infection rates was the single most important factor accounting for the differences in peritonitis rates observed between the various subgroups of patients evaluated. This is best demonstrated in the evaluation of the peritoneal dialysis systems in which S epidermidis infection rates were only 0.20 episodes/patient-year for automated peritoneal dialysis and 0.55 episodes/patient-year for the disconnect system compared with 0.85 episodes/patient-year for the connect system. Since S epidermidis infections generally are attributed to touch contamination, identifying patients at greatest risk allows the potential for intervention with closer supervision, adjustments in the peritoneal dialysis training/education approach, or a change in the peritoneal dialysis system.