BackgroundDefinitive therapy for osteomyelitis (OM) is thought to be superior to empiric antimicrobial therapy; however, identifying causative pathogens is difficult.MethodsThis retrospective cohort study included patients treated with either definitive or empiric antimicrobial therapy for OM at VA St. Louis HCS between 1 January 2010 and 1 January 2018. Definitive antibiotic therapy was defined as a regimen tailored to susceptibilities of an organism(s) cultured from bone or deep tissue. The primary outcome was treatment failure, defined as a need for unplanned surgical intervention or re-initiation of antibiotic therapy for OM of the same anatomical site within 6-months after initial therapy was discontinued. Secondary outcomes included the incidence of acute kidney injury (AKI), Clostridium difficile-associated diarrhea (CDAD), and thrombocytopenia. Surgical intervention as part of initial therapy, presence of peripheral vascular disease (PVD), creatinine clearance < 50 mL/minute at initiation of therapy, receiving antibiotics at an extended care facility, age > 60 years, and receiving definitive antibiotics were included in a univariate analysis with variables with a P < 0.2 included in a multivariate logistic regression.ResultsThere were 301 patients included; 179 in the definitive therapy group and 122 in the empiric therapy group. Baseline characteristics were similar among groups; however, more patients receiving definitive therapy had a bone biopsy compared with those treated with empiric therapy (58.1% (104/179) vs. 36.8% (45/122); P < 0.05). 33 percent (60/179) of patient treated with definitive therapy failed compared with 45% (55/122) treated with empiric therapy (P = 0.109). No significant differences were observed in secondary outcomes; however non-CDAD diarrhea occurred more in the empiric therapy group than definitive therapy group (3.9% (7/179) vs. 8.2% (10/122); P > 0.05). Receiving definitive therapy (OR 1.43, CI 0.89–2.313; P = 0.138) and presence of PVD (OR 1.34; CI 0.823–2.197; P = 0.238) were included in the multivariate logistic regression, but neither were independently associated with failure.ConclusionDefinitive antibiotic therapy was not associated with a significant decrease in treatment failure.Disclosures All authors: No reported disclosures.