Case mix, costs, and outcomes. Differences between faculty and community services in a university hospital.

To gain insight into the possible consequences of prospective payment for university hospitals, we studied 2025 admissions to the faculty and community services of a university hospital, measuring differences in case mix, costs, and mortality in the hospital. The faculty service had more of the patients with costly diagnoses, but even after adjustment for diagnosis-related groups (DRGs), costs were 11 per cent higher on the faculty service (95 per cent confidence limits, 4 to 18 per cent). The percentage differential was greatest for diagnostic costs. The differential was particularly large--70 per cent (95 per cent confidence limits, 33 to 107 per cent)--for patients with a predicted probability of death of 0.25 or greater. The in-hospital mortality rate was significantly lower on the faculty service after adjustment for case mix and patient characteristics (P less than 0.05); the difference was particularly large for patients in the high-death-risk category. Comparison of a matched sample of 51 pairs of admissions from the high-death-risk category confirmed the above results with respect to costs and in-hospital mortality, but follow-up revealed that the survival rates were equal for the two services at nine months after discharge. The effect of prospective payment on the cost of care will be closely watched; we conclude that is will also be important to monitor the effect on outcomes, including hospital mortality rates.