Variation in Peritoneal Dialysis–Related Peritonitis Outcomes in the Peritoneal Dialysis Outcomes and Practice Patterns Study (PDOPPS)

Abstract

Peritoneal dialysis (PD)-associated peritonitis is a significant PD-related complication. We describe the likelihood of cure after a peritonitis episode, exploring its association with various patient, peritonitis, and treatment characteristics. Observational prospective cohort study. 1,631 peritonitis episodes (1,190 patients, 126 facilities) in Australia, New Zealand, Canada, Japan, Thailand, the United Kingdom, and the United States. Patient characteristics (demographics, patient history, laboratory values), peritonitis characteristics (organism category, concomitant exit-site infection), dialysis center characteristics (use of icodextrin and low glucose degradation product solutions, policies regarding antibiotic self-administration), and peritonitis treatment characteristics (antibiotic used). Cure, defined as absence of death, transfer to hemodialysis (HD), PD catheter removal, relapse, or recurrent peritonitis within 50 days of a peritonitis episode. Mixed-effects logistic models. Overall, 65% of episodes resulted in a cure. Adjusted odds ratios (AOR) for cure were similar across countries (range, 54%-68%), by age, sex, dialysis vintage, and diabetes status. Compared with Gram-positive peritonitis, the odds of cure were lower for Gram-negative (AOR, 0.41 [95% CI, 0.30-0.57]), polymicrobial (AOR, 0.30 [95% CI, 0.20-0.47]), and fungal (AOR, 0.01 [95% CI, 0.00-0.07]) peritonitis. Odds of cure were higher with automated PD versus continuous ambulatory PD (AOR, 1.36 [95% CI, 1.02-1.82]), facility icodextrin use (AOR per 10% greater icodextrin use, 1.06 [95% CI, 1.01-1.12]), empirical aminoglycoside use (AOR, 3.95 [95% CI, 1.23-12.68]), and ciprofloxacin use versus ceftazidime use for Gram-negative peritonitis (AOR, 5.73 [95% CI, 1.07-30.61]). Prior peritonitis episodes (AOR, 0.85 [95% CI, 0.74-0.99]) and concomitant exit-site infection (AOR, 0.41 [95% CI, 0.26-0.64]) were associated with a lower odds of cure. Sample selection may be biased and generalizability may be limited. Residual confounding and confounding by indication limit inferences. Use of facility-level treatment variables may not capture patient-level treatments. Outcomes after peritonitis vary by patient characteristics, peritonitis characteristics, and modifiable peritonitis treatment practices. Differences in the odds of cure across infecting organisms and antibiotic regimens suggest that organism-specific treatment considerations warrant further investigation.