990. Prophylaxis against spontaneous bacterial peritonitis: too much or too little?

Abstract

BackgroundProphylaxis against spontaneous bacterial peritonitis (SBP) is a guideline-recommended strategy; there are limited data on rates of concordance with guideline recommendations. We sought to evaluate rates of concordance, hypothesizing that antibiotics would overprescribed for prophylaxis against SBP.MethodsThis retrospective cohort study included all patients at the Boston Veterans Affairs Medical Center who underwent paracentesis between 1/1/2014 and 12/31/2018. Exclusion criteria included absence of cirrhosis and hepatic transplantation, either prior to enrollment or during the study period. Manual review was used to capture demographic data, guideline concordance, microbiology results and healthcare utilization within one year of enrollment. Descriptive and analytical statistics were performed.ResultsOf 259 patients eligible for analysis, 181 (70%) met inclusion criteria; 65 patients (25%) were excluded as cirrhosis was not confirmed. Small numbers of other patients were excluded for other reasons [Figure 1].Incorrect antibiotic utilization was noted in 80 patients (44%) [Figure 2]. Among 93 patients meriting antibiotics, 65 (70%) did not receive them. Conversely, among 90 who did not have an indication for antibiotic prophylaxis, 15 (17%) received it (p = 0.03, chi-squared test).Receipt of SBP prophylaxis was not correlated with gastroenterologist involvement, infection by antibiotic-resistant bacteria or development of illness due to Clostridioides difficile. No difference in hospital readmission rates was observed between groups receiving guideline-concordant and guideline-discordant prophylaxis.Figure 1 Figure 2 ConclusionWe expected to find overprescription of SBP prophylaxis. In fact, we found that the largest error in prescribing was underprescribing, which may be equally as harmful as inappropriate use of antibiotics. SBP prophylaxis may be an important target for antibiotic stewardship and education.Disclosures All Authors: No reported disclosures