Impact of diagnosis-related group-based reimbursement for treatment of rheumatic diseases in a teaching hospital.

The impact of Medicare's diagnosis-related group (DRG)-based reimbursement system was examined for care given to 734 rheumatic disease patients discharged from a teaching hospital during a 2-year period. The analysis accounted for length of stay "outliers" as defined by Medicare and distinguished costs from charges. Excluding outliers, DRG reimbursement would result in net revenues to the hospital of +1,126 per DRG 240 patient and $1,794 per DRG 241 patient. The difference between DRGs in cost per patient was significant, indicating that DRGs clearly identify 2 groups of rheumatic disease patients. After excluding outliers, the coefficients of variation in costs for DRGs 240 and 241 were 72% and 80%, respectively, which although high, were average for DRGs at our institution. Mean total charge per patient was different for groups defined by their primary rheumatologic diagnosis in DRG 240 but not DRG 241. For rheumatoid arthritis and systemic lupus erythematosus patients, the total charge per patient did not differ, but the types of services did. The cost of treating outliers would create an average loss per outlier of $18,400 and $16,500, respectively, in DRGs 240 and 241. Outliers accounted for 34.5% and 21.8% of the 2 DRGs' total costs, respectively, but only 6.4% and 3.6% of the total number of patients. Under current DRG reimbursement rates, the cost of care for rheumatology patients would be adequately reimbursed in our hospital: losses from outliers would be offset by net revenues from inliers as long as current Medicare adjustments for capital and medical education costs were continued.