Comparative hospital performance: new data, borrowed methods, more targeted analysis?

Abstract: What is the purpose of comparing hospital performance? To identify benchmark performance, and motivate and inform performance improvement at non-benchmark institutions? In line with the value-based pricing of new technologies, decisions to intervene to improve hospital performance should reflect the expected costs of improved healthcare provision, and the value of improved patient outcomes.

Although not routinely available to date, the linkage of de-identified individual-level hospital and mortality data provides longitudinal data on the whole populations of patients treated at alternative hospitals. In Australia, bottom up costing to inform casemix funding provides individual-level cost data for most inpatient episodes. On a small cohort of four hospitals, we analyse these data to inform:

a) priority areas for action to improve performance, through analyses of the incremental costs and effects of care provided by alternative hospitals for specific conditions,

b) targets for investigation, as points of variation in the clinical pathways of patients treated at alternative hospitals.

Such methods provide a link between funding decisions for new technologies and for improving the use of existing technologies, and may highlight greater potential for improving population health at low incremental cost from the latter. Remaining issues include the timeliness of data access, how to use the outputs to motivate performance improvement, and the applicability of the process to a much larger set of hospitals.