OBJECTIVE: To determine the cost of treating an intensive care unit (ICU) patient with severe sepsis who expires while hospitalized at The Johns Hopkins Hospital (JHH), prior to the introduction of drotrecogin alfa (activated) for the treatment of severe sepsis. METHOD: The study utilized a cost-of-illness methodology. Data were collected through retrospective chart review. Patients with sepsis were identified based on their discharge summary in an integrated patient care database at JHH. Patient selection included those hospitalized between October 1, 2000 and September 30, 2001, in an adult ICU, who expired prior to hospital discharge. A random sample of these patients was chosen for chart review. Chart review identified patient status based on the PROWESS study inclusion criteria for the diagnosis of severe sepsis. Cost-of-illness calculation included direct medical costs (medications, hospital days, and ICU days). The cost to the hospital was calculated using the cost-to-charge ratio specific for JHH. Indirect costs were not included. RESULTS: A total of 60 patients were included in the study. The mean total hospital, non-ICU, ICU, and sepsis-related lengths of stay were 24, 8, 16, and 18 days, respectively. Patients spent a mean of 67% of hospitalized days in the ICU. The mean total hospitalization cost for a patient with severe sepsis who expires was approximately $54,000 and the mean sepsis-related hospitalization cost for one of these patients was approximately $40,000. The mean total costs per day (medication costs per day in parentheses) for the entire hospitalization, non-ICU stay, ICU stay, and sepsis-related stay were $2270 ($313), $1512 ($210), $2649 ($364), and $2245 ($339), respectively. CONCLUSION: Sepsis is a condition that has a significant impact on patient mortality and hospital costs. The results of this study will be used to monitor the effect of the use of drotrecogin alfa (activated) for the treatment of severe sepsis at JHH.