LAUSR

Early studies in peritonitis showed no difference in outcomes between anuric and non-anuric patients, despite pharmacokinetic studies showing the effect of residual renal function (RRF) on antibiotic clearance. Recent work identified a relationship between preserved RRF and treatment failure in gram-positive and culture-negative peritonitis. Our centre uses empiric vancomycin and gentamicin dosed by weight (both drugs) and by RRF (gentamicin). We sought to examine whether RRF is associated with treatment failure in our population. We retrospectively identified all episodes of PD peritonitis between January 2014 and July 2019 including demographic and clinical information. The RRF measured closest to the peritonitis episode was used. Treatment failure was defined as death, catheter removal or relapse. 189 peritonitis episodes occurred in 128 patients (43% female, mean age 61.3±17.9). 80 episodes were caused by gram-positive bacteria, 49 gram-negative, 51 culture-negative, 5 fungal and 4 polymicrobial. 21, 88 and 61 episodes occurred in patients with a creatinine clearance <0.5, 0.5-5, and >5ml/min respectively. Creatinine clearance data was unavailable in 21 patients. Treatment failure occurred in 72 cases (38%) (19 relapses, 52 catheter removals, 7 deaths). Treatment failure for all bacterial peritonitis episodes was not affected by RRF as compared to anuric patients: 0.5-5 (OR 0.41, 95% CI 0.12-1.35, p 0.14) or >5 (OR 0.90, 95% CI 0.22-3.67, p 0.88). There was no difference in outcome in sub-group analysis by type of organism. Antibiotic concentrations did not differ between the RRF groups. Although limited by small numbers we found no association between the degree of RRF and outcome. We hypothesise this is because we monitor antibiotic concentrations and therefore dose antibiotics appropriate to the level of RRF. In PD peritonitis regimes that adjust for RRF there is no impact of RRF on treatment outcomes.