Peritonitis: Episode Sequence, Microbiological Variation, Risk Factors and Clinical Outcomes in a North China Peritoneal Dialysis Center

Abstract

Background/Aims: This study investigated peritonitis episodes with regard to time sequence, microbiological variation, factors associated with peritonitis and clinical outcomes in peritoneal dialysis (PD) patients. Methods: This single-center cohort study enrolled all incident patients who met the inclusion criteria at our center from June 1, 2012 to June 30, 2015 and who were followed until June, 2017. Clinical, biochemical characteristics and detailed data on peritonitis episodes, and hospitalizations were recorded. Results: A total of 218 episodes of peritonitis corresponding to a rate of 0.27 episode per patient-year were recorded. Gram positive bacteria, identified in 115 (52.8%) episodes, were the most common pathogens. The occurrence of enterococcus peritonitis increased from 15.1% of the first to 27.3% of the later episodes. Multivariate logistic regression showed that the presence of cardiovascular disease (CVD, odds ratio [OR] 2.177, 95% confidence interval [95%CI] 1.214–3.903, P=0.009), age≤ 55 (OR 2.282, 95%CI 1.062–4.906, P=0.035), non-independent operator (OR 0.440, 95%CI 0.206–0.938, P=0.034), lower values of potassium (OR=0.671,95%CI 0.472–0.954, P=0.026) and higher values of calcium-phosphate product (OR 1.410, 95%CI 1.065–1.868, P=0.017) were associated with peritonitis. Besides CVD (risk ratio [RR] 2.591, 95%CI 1.893–3.543, P< 0.001) and non-independent operator (RR 0.583, 95%CI 0.439–0.776, P< 0.001), a lower level of education (RR 0.641, 95%CI 0.487–0.842, P=0.001) was associated with higher peritonitis rates in log-linear analysis. Spearman analyses indicated that the time to the 1st episode was negatively related to the peritonitis rate (r=-0.291, P=0.001). Time-dependent Cox regression showed no association between the time to the 1st episode and patient survival (P=0.151). Patients with a high peritonitis rate (HPR) demonstrated worse technique survival (P< 0.001). Conclusion: The present study has revealed several center-based features and modifiable risk factors for peritonitis. The presence of CVD and the need for assistance with PD operation not only increased the odds of peritonitis but were also associated with more peritonitis episodes. Time to first peritonitis was related to the peritonitis rate but not associated with patient survival. Patients with HPR had worse technique survival.