Abstract

A number of interesting arguments have been made about the utility of the hospital standardized mortality ratio (HSMR) as a "big dot," broad-based safety indicator that is useful for initiating change and measuring change over time. This rejoinder examines these arguments. Section one briefly revisits issues of measurement related to the HSMR. Section two discusses whether and how big dots should be used. Section three addresses problems with using HSMRs to track changes in mortality for a facility over time. The final section describes issues with the indirect effects of HSMR and engaging staff.