Peritonitis in children on chronic peritoneal dialysis; experience from a paediatric nephrology centre

Abstract

Aim To determine variation in clinical manifestations, antibiotic sensitivity, patterns of causative organisms and the outcomes of bacterial peritonitis in children in our institution. Methods This retrospective study population included all those children who developed chronic peritoneal dialysis (CPD) related peritonitis during October 2005–February 2008 at Children Kidney Centre at UHW, Cardiff. The parameters collected were age, presenting symptoms, presence of exit site infection/tunnel infection, nature of peritoneal effluent (PE) at presentation and at 72 h after treatment, type of organism grown, antibiotics sensitivities, response to treatment and outcome. Results Out of 15 patients who were on CPD during 3 years of the study period, 10 patients developed a total of 15 episodes of peritonitis. Risk factors for peritonitis were presence of gastrostomy, use of diapers and exit site infections. Main presenting features were PE with WBC count >100 (87%), abdominal pain and fever (53%), turbid PE and vomiting (40%). Organisms were grown from PE in 13 (86%) episodes and culture was negative in two. Gram positive (GP) organisms accounted for 47%, gram negative (GN) for 27% of episodes and mixed organisms and candida found in one episode each. 86% of GP organisms were sensitive to a glycopeptide and 75% of GN organisms were sensitive to ceftazidime/gentamicin. 10 episodes showed good response to therapy, four episodes improved following removal of PD catheter and one patient died due to Staphylococcus aureus sepsis. Discussion and Conclusions The children with gram negative peritonitis were younger than those who developed gram positive peritonitis. Abdominal pain was common in peritonitis caused by staph aureus, streptococcus and Gram negative organisms. Fever and turbid PE were more likely with Gram negative peritonitis. With appropriate antibiotic therapy, supportive treatment and PD catheter removal where indicated, good clinical improvement was seen in all patients except one (93%). Peritonitis due to staph aureus, pseudomonas and candida were more likely to need PD catheter removal. Those who ultimately need PD catheter removal are more likely to have associated exit site infection, more symptomatic, are culture (effluent) positive and show unsatisfactory response after 72 h of therapy. Strategies to prevent CPD related peritonitis include early detection and treatment of exit site infection, appropriate care of gastrostomy site, optimum placement of PD catheter and avoiding nappy area for catheter exit site in infants. Early detection and prompt initiation of appropriate therapy is essential for a favourable outcome.