The California Hospital Outcomes Project: Using administrative data to compare hospital performance

Background: The California Hospital Outcomes Project was created by an act of the state legislature in 1991. The California Office of Statewide Health Planning and Development (OSHPD) publishes annual reports on risk-adjusted hospital outcomes for medical, surgical, and obstetric patients. These-outcomes indicators were chosen: in-hospital mortality within 30 days (acute myocardial infarction [AMI]), reported postoperative complications (diskectomy, delivery), postoperative length of stay (diskectomy), and readmission within 6 weeks (delivery). Project reports are based on discharge abstracts submitted by hospitals and edited by OSHPD. For each outcome, two risk adjustment models were used to estimate expected and risk-adjusted hospital outcome rates, along with p values representing the likelihood that the observed number of adverse but comes occurred by chance. Results: The first hospital outcomes report was distributed to hospitals in June 1993 and released to the public in December 1993. The total number of hospitals labeled as ''better than expected'' was 14 for AMI, 5 for cervical diskectomy, and 25 for lumbar diskectomy. The second hospital outcomes report was distributed to hospitals in June 1995. Response and conclusions: Letters submitted for 168 hospitals in response to the 1993 report demonstrated that hospitals had studied and used project results to evaluate their coding practices and quality of care. Media coverage of the 1993 report was balanced but sometimes critical of OSHPD's failure to identify ''worse'' hospitals. In response to providers' concerns, OSHPD has undertaken a validation study to explore whether differences in coding, unmeasured risk factors, or processes of care explain the reported differences in risk-adjusted outcome rates.