ETIOLOGY AND THE IMPACT OF REFRACTORY PERITONITIS ON CLINICAL OUTCOMES OF PATIENTS ON PERITONEAL DIALYSIS-12 YEARS' SINGLE-CENTER EXPERIENCE FROM TURKEY

Abstract BACKGROUND AND AIMS The outcomes of refractory peritonitis in peritoneal dialysis (PD) patients have been reported to be inferior to those of solitary peritonitis. The current study aimed to examine the factors associated with treatment failure in PD patients experiencing refractory peritonitis. METHOD This single-center retrospective study included all episodes of refractory peritonitis in adult PD patients in Marmara University Hospital, Turkey, between 2009 and 2020. Patient characteristics, microbiological data, outcomes and factors associated with refractory peritonitis were analysed. The primary outcome was peritonitis-related catheter loss. Secondary outcomes were hospitalization and peritonitis-related death. RESULTS Overall, 236 episodes of refractory peritonitis occurring in 135 patients were included. Gram-positive, gram-negative and fungal infections accounted for 44.1%, 20.4% and 2.4% of all peritonitis episodes, respectively. Forty-seven patients (34.8%) needed catheter removal, 2 patients (1.5%) died due to peritonitis complications and 59 episodes (25%) needed hospitalization. Mean fifth day PD fluid cell count was significantly greater among patients who required PD catheter removal (3621.3 $ \pm $ 3144.1 versus 1589.4 $ \pm $ 2316.6 P < 0.001). Furthermore, patients with >1000/mm3 cell count on the fifth day had higher rate of catheter removal (72.3% versus 37.6%, P < 0.001) as compared to patients with cell count under 300/mm3. Treatment failure was more common in peritonitis episodes caused by gram (−) organisms (31.9% versus 14.8%, P:0.012). Pseudomonas and fungi-associated peritonitis were also significantly correlated with catheter loss (P:0.001 and P: <0.001) (Table 1). When peritonitis episodes with more and <1000 cells/mm3 on the fifth day were compared, there were more episodes with gram (−) bacteria (29.7% versus 12.9%, P:0.003) and hospitalization (41.9% versus 11.4%, P: <0.001) in the group with more than 1000 cells/mm3. In the multivariate analysis, factors associated with catheter loss were a cell count of >1000 on the fifth day and hospitalization, while presence of gram (+) bacteria related peritonitis was inversely correlated with catheter loss (Table 2). CONCLUSION Our study shows that the PD cell count on the fifth day of peritonitis can be used as a prognostic tool to determine the prognosis of refractory peritonitis episodes. Although we were unable to show the adverse prognostic effect of gram (−) bacteria related peritonitis, gram (+) bacteria related peritonitis was associated with better outcome. Prospective studies are needed to assess the risk factors for adverse outcomes of patients with refractory peritonitis, as the evidence in this area is sparse.