ABSTRACT
Introduction: Gram-negative peritonitis (GNP) is associated with significant morbidity in children receiving long-term peritoneal dialysis (PD) and current treatment recommendations are based on limited data. Methods: Analysis of 379 GNP episodes in 308 children (median age 6.9 years, interquartile range [IQR]: 3.0-13.6) from 45 centers in 28 countries reported to the International Pediatric Peritoneal Dialysis Network registry between 2011 and 2023. Results: Overall, 74% of episodes responded well to empiric therapy and full functional recovery (FFR) was achieved in 82% of cases. In vitro bacterial susceptibility to empiric antibiotics and lack of severe abdominal pain at onset were associated with a good initial response. Risk factors for failure to achieve FFR included severe abdominal pain at onset and at 60 to 72 hours from treatment initiation (odds ratio [OR]: 3.81, 95% confidence interval [CI]: 2.01-7.2 and OR: 3.94, 95% CI: 1.06-14.67, respectively), Pseudomonas spp. etiology (OR: 1.73, 95% CI: 1.71-4.21]) and in vitro bacterial resistance to empiric antibiotics (OR: 2.40, 95% CI: 1.21-4.79); the risk was lower with the use of monotherapy as definitive treatment (OR: 0.40, 95% CI: 0.21-0.77). Multivariate analysis showed no benefit of dual antibiotic therapy for treatment of Pseudomonas peritonitis after adjustment for age, presenting symptomatology, 60 to 72-hour treatment response, and treatment duration. Monotherapy with cefazolin in susceptible Enterobacterales peritonitis resulted in a similar FFR rate (91% vs. 93%) as treatment with ceftazidime or cefepime monotherapy. Conclusion: Detailed microbiological assessment, consisting of patient-specific and center-specific antimicrobial susceptibility data, should guide empiric treatment. Treatment "deescalation" with the use of monotherapy and narrow spectrum antibiotics according to susceptibility data is not associated with inferior outcomes and should be advocated in the context of emerging bacterial resistance.
