1364. Microbiology of Sepsis in a Combat-Trauma Military Population

Abstract Background There are limited data on sepsis in combat casualties. We examined characteristics of sepsis, specific infections, and associated microbiology in a complex combat trauma population. Methods The Trauma Infectious Disease Outcomes Study collected infection-related data from military personnel wounded during deployment (2009-2014). Medevac patients transferred to participating US military hospitals with sepsis or septic shock based on the Sepsis-1 SIRS criteria were analyzed for associated potential sources and infection-associated clinical microbiology. Results Prevalence of sepsis was 24.7% (667 of 2699 patients) with 93 (14%) patients meeting septic shock criteria. There were 1013 sepsis/shock episodes (SSE) among 667 subjects. Infections attributed to SSE were identified in 996 (98.3%) of 1013 episodes, primarily being bloodstream infections (BSI) +/- other infections (29.5%), skin and soft tissue (SSTI)/osteomyelitis (35.3%), pneumonia (12.1%), and multiple concurrent infections (14.2%). At least 1 organism was identified in 96% of SSE and 53% were polymicrobial. Gram-positive organisms (GP) were identified in 54% of SSE: 16% with multiple GP, of monomicrobial infections 4.1% were S. aureus, 15.8% other staphylococci, and 13% Enterococcus spp. Gram-negative bacilli (GN) were identified from 74.5% of SSE: 34% with multiple GN, of monomicrobial infections 11% were Pseudomonas spp., 8% E. coli, 6% Enterobacter spp., and 6% Acinetobacter spp. Mycobacterial species were uncommon (0.9%). Yeast, mold, and anaerobes were identified from 19%, 22%, and 12.5% of SSE, respectively. Compared to sepsis, septic shock infections were more often polymicrobial (p< 0.001), and had more infections with ESKAPEE pathogens, only Mucor spp., and only Bacteroides (p< 0.05). More infections with only Pseudomonas spp. and only non-lugdunensis coagulase-negative Staphylococci were identified among sepsis patients (p< 0.05). Conclusion Sepsis rates, using the Sepsis-1 criteria are sensitive but lack specificity supporting reclassification using updated Sepsis-3 criteria. In a complex trauma population, sepsis is common with most frequent infections related to SSTI/osteomyelitis, as well as BSI and multiple concurrent infections with a diverse spectrum of microbiology. Disclosures Dana M. Blyth, MD, Nothing to disclose