2102. Does Monitoring Procalcitonin Levels in Septic and Septic Shock Patients Decrease the Use of Antibiotics and Predict Length of Hospital Stay?

Abstract

BackgroundElevated levels of procalcitonin (PCT) reflect systemic inflammation associated with bacterial infection (BI). Compared with other acute-phase reactants, PCT levels more rapidly rise with BI and decline quickly as BI improves. A PCT protocol was implemented at University Medical Center New Orleans (UMCNO) in November 2017 that guided discontinuation of antimicrobial therapy in septic and septic shock patients if clinically improving with declining levels of PCT.MethodsWe performed a retrospective chart review of UMCNO patients 18+ years with a diagnosis of sepsis and a PCT level between January 1st, 2018 to July 31st, 2018 compared with those with sepsis and no PCT level. ICD-9/10 codes were used for diagnoses of sepsis and septic shock. The baseline characteristics including age, gender, body mass index, race and Charlston Comorbidity Index (CCI) data were collected. The primary objective was to compare the total days of antibiotic therapy (DOT) between the two groups. Secondary outcomes were broad-spectrum antibiotic DOT, patient length of stay (LOS), and all-cause 28-day mortality. SPSS was used for data analysis. P < 0.05 indicated statistical significance.ResultsThere were 44 patients in the PCT group (PCTg) and 35 in the non-PCT group (nPCTg). The demographics are outlined in Table. The mean DOT was 6.25 days in PCTg and 10.74 days in nPCTg (P = 0.006). LOS was 7.5 days in PCTg and 14 in nPCTg (P = 0.006). The mean CCI was 2.4 in PCTg and 4 in nPCTg (P = 0.007). The all-cause 28-day mortality was 11% in PCTg and 23% in nPCTg (OR 0.4; 95% CI 0.128–1.466). On bivariate analysis, LOS was significantly associated with CCI (P < 0.05) and total DOT (P = 0.000). On multivariate analysis, LOS was only significantly associated with age (P = 0.015) and total DOT (P = 0.000) but not CCI (P = 0.811) nor PCTg (P = 0.250).ConclusionDOT was significantly shorter in the PCTg than in nPCTg. The LOS was 50% less in PCTg than in nPCTg; however, PCT monitoring did not contribute to LOS in multivariate analysis. Although the nPCTg were sicker, CCI did not correlate with LOS either. However, age and total DOT therapy remained positive predictors of LOS. Monitoring PCT levels decreased antibiotic use in septic patients. LOS, however, was not significantly affected by PCT monitoring. Disclosures All authors: No reported disclosures.