Proportional Costs in Trauma and Acute Care Surgery Patients: Dominant Role of Intensive Care Unit Costs

BackgroundControlling inpatient costs is increasingly important. Identifying proportionately larger cost categories may help focus cost control efforts. The purpose of this study was to identify proportionate patient cost categories in trauma and acute care surgery (TACS) patients and determine subgroups in which the largest opportunities for cost savings might exist.Study DesignAdministrative data from our academic, urban, level I trauma center were used to identify all adult TACS patients from FY07 through FY11. We determined, on average, what proportion of the whole each cost category contributed to patients' total costs and examined the same proportions for subgroups of patients.ResultsWe identified 6,008 TACS patients. Trauma patients (n = 3,904) made up 65% of the cohort (mean Injury Severity Score 13.2). Payers were: 22% government (Medicare, Medicaid, Champus), 27% private, 43% self-pay/indigent, 3% other, and 5% workers compensation. Nontrauma (general surgery) patients (n = 2,104) made up 35% of the cohort. Payers were: 44% government, 24% private, 29% self-pay/indigent, 2% other, and 1% workers compensation. Total inpatient costs were $141,304,993. Per patient costs rose from $17,245 in FY07 to $26,468 in FY11. In the aggregate, supplies, ICU stays, and ward stays represented the largest proportionate cost categories. On a per patient basis, however, ICU stays were by far the largest cost. Patients with ICU stay greater than 10 days were only 7% of all patients but accounted for 41% of total costs.ConclusionsTrauma and acute care surgery patients represent a significant and increasing institutional cost. Per patient ICU costs were the largest single category, suggesting that cost control efforts should focus heavily on critically ill patients. Nontrauma patients who require critical care have the highest per patient ICU costs and may represent a previously underappreciated opportunity for cost control.