Risk-adjusted hospital clinical management issue rates using data from the Victorian Audit of Surgical Mortality

Background In recent years, there has been a concerted drive for an increase in public reporting of hospital-level outcomes as a means of identifying strategies to improve patient safety. Surgical care, as a high-risk area of medical practice, has come under sharp scrutiny. This study uses data from the Victorian Audit of Surgical Mortality (VASM) in conjunction with data from the Victorian Admitted Episode Dataset to compare hospital rates of clinically identified serious clinical management issues that were definitely or probably preventable and caused or contributed to the death of the patient who would otherwise be expected to survive. Methods Cases where the date of death was between 1 July 2015 and 30 June 2017 that completed the full VASM audit process were extracted from the VASM database and combined with data extracted from the Victorian Admitted Episode Dataset, where a surgical admission occurred in the same time period. A logistic regression model was used as a method of indirect standardization to derive the probability of preventable clinical management issues, which was then used to calculate the standardized incident rate for all Victorian surgical hospitals. Hospitals were compared by plotting the standardized incident rates on three funnel plots. Results There were five hospitals (8.3%) of the 60 that deviated significantly from the state-wide rate of 0.00012. Conclusion The risk adjustment model identified several hospitals that may have a systematic issue which warrant further clinical quality assurance investigation.