The impact of initial antibiotic therapy (linezolid, vancomycin, daptomycin) on hospital length of stay for complicated skin and soft tissue infections

BACKGROUND: Empiric therapy of inpatient skin and soft tissue infections (SSTIs) generally require methicillin resistant Staphylococcus aureus (MRSA) coverage. Limited data are available to directly compare the effect of initial antibiotic choice on treatment outcomes and length of stay (LOS). OBJECTIVE: To assess potential differences in length of hospital stay when inpatients with complex skin and soft tissue infections (SSTIs) were initially treated with either vancomycin, linezolid, or daptomycin. METHODS: A retrospective review of 219 patients diagnosed with inpatient SSTI who received linezolid, vancomycin, or daptomycin for >48 hours was performed. Data collected included demographics, comorbidities, microbiologic/laboratory data, additional management (surgical, non-study antibiotics), hospital LOS, treatment outcome and morbidity/mortality. RESULTS: The three groups evaluated were linezolid (n = 45), vancomycin (n = 90) daptomycin (n = 84). There was no difference between the three groups with respect to gender, age, comorbidities, leukocytosis, fever, antibiotics prior to admission, site of infection culture results and surgical intervention. One death was recorded, not associated with diagnosis of SSTI. No significant difference in LOS was found (P = 0.525) between the 3 groups. The mean LOS in entire cohort was 4.5 days (SD ± 2.5); thirty patients had prolonged LOS for non-SSTI reasons; reanalyzing the data without these 30 patients did not produce any difference in the mean LOS between the 3 groups. Switching vancomycin just prior to discharge to facilitate outpatient therapy was common but did not impact LOS. CONCLUSIONS: No difference was detected in hospital length of stay with respect to the initial choice of antibiotic (linezolid, vancomycin, or daptomycin) for SSTI. The three antibiotic regimens were equally effective in treating SSTIs as judged by LOS, irrespective of age, gender, comorbidities or baseline severity of SSTI. Given the large standard deviation in LOS, this result should be confirmed by larger studies.