304 - Predictors of mortality in carbapenem-resistant Enterobacteriales bacteremia

Background: Carbapenem-Resistant Enterobacteriales (CRE) bacteremia is associated with significant morbidity and mortality. CRE were assigned a threat level of “urgent” in the 2019 CDC report on antibiotic resistance in the United States. We attempted to identify predictors of 30-day mortality in patients with CRE bacteremia.

Methods: We performed a chart review of 146 patients with CRE bacteremia from January 2010 - July 2019.  CRE was defined using the current CDC definition. Electronic medical records were reviewed to obtain clinical characteristics and outcomes including prior antibiotic use, comorbidities, prior location, treatment, hospital course, microbiological data and outcomes including in-hospital mortality.

Results: Of 146 patients included for analysis, the overall 30-day mortality rate was 36.3%. Patients admitted from a healthcare facility including outside hospitals, rehab, nursing homes, and LTACs had a 49.1% (29/59) 30-day mortality rate compared to 27.5% (24/87) for those admitted from home (RR=1.78, 95% CI 1.16-2.73, p=.0082). Patients with a Pitt bacteremia score ≥ 4 had a greater 30-day mortality rate (42.6%, 26/61) compared to those with a Pitt bacteremia score < 4 (17.6%, 15/85)  (RR=2.92, 95% CI 1.40-4.16, p=.0015). Patients that received inactive empiric therapy had a 30-day mortality rate of 36% (36/100) compared to 36.9% (17/46) in those that received active empiric therapy (RR=.9741, 95% CI .6155-1.59, p=.9109). Patients with isolates determined to have a meropenem MIC ≥ 4 had a 30-day mortality rate of 40.2% (37/92) while those with an MIC < 4 had a 30-day mortality rate of 30.2% (16/53) (RR=1.33, 95% CI .8250-2.1513, p=.2408). A pulmonary source of bacteremia was associated with an increased risk of 30-day mortality (64.3%, 9/14) compared to all other sources of bacteremia (34.8%, 31/89) (RR=1.85, 95% CI 1.39-2.99, p=.0129). No other infection source was associated with an increased 30-day mortality rate.

Conclusion: Admission from a healthcare facility, Pitt bacteremia score ≥ 4, and pulmonary source of bacteremia were associated with increased risk of 30-day mortality. Interestingly, administration of active empiric therapy was not associated with a decreased mortality risk. Meropenem MIC was not predictive of 30-day mortality.