UTILIZATION, COSTS AND REIMBURSEMENT OF INPATIENT AND AMBULATORY TREATMENT OF ACUTE BACTERIAL SKIN AND SKIN STRUCTURE INFECTIONS AMONG THE MEDICARE FEE-FOR-SERVICE POPULATION

OBJECTIVES: Recent clinical trials of antibiotics targeted to ABSSSI have included large proportions of patients treated in the ambulatory setting. An objective of this study of real-world patient transaction data was to analyze differences between Medicare patients treated in the inpatient versus the ambulatory setting. METHODS: Analysis used Medicare 5% Limited Data Set (LDS) files containing claims for 27,607 patients age ≥18 with a principal diagnosis of ABSSSI between Q4, 2007 – Q3, 2010 from a 5% sample of Medicare beneficiaries, including patients hospitalized for ≥2 days (hospitalized patients; HPs), or using emergency room (ER) or observation for 1 day (emergency room patients; ERPs). Reimbursements were based on claims and inflated to 2010 USD; costs were derived from 2010 Premier data. Net reimbursement was analyzed by MS-DRG and length of stay (LOS). The risk of all-cause hospitalization and factors correlated with LOS were determined using regression modeling. RESULTS: Across all study years, the median age was 71 for HPs, 65 for ERPs. Median Charlson Comorbidity Index (CCI) was 4 for HPs and 2 for ERPs. HPs had more cellulitis on the leg or surgical infection; ERPs had more cellulitis on the face, trunk, or arm. Median HP LOS was 4 days; 33% of patients had LOS >6 days. Age, race, and history of bacterial infection were correlated with LOS. Median all-cause, ABSSSI-related and index event costs during the index quarter for patients with a principle ABSSSI diagnosis were $9,930, $6,123 and $5,604 for HPs and $1,127, $108 and $100 for ERPs respectively. CONCLUSIONS: HPs were older and sicker than ERPs, with more cellulitis on the leg and post-operative infections. Further research and analysis may help determine whether treating some hospitalized patients with less comorbidity in the ambulatory setting might result in lower costs per patient after controlling for these factors.