Getting Better at Measuring Hospital Mortality

The call for quality measurement to focus on outcomes rather than care processes is loud and persistent. However, robust consideration of problems with the validity of outcome measures is often lacking. In this issue of JAMA Internal Medicine, Silva and colleagues make an important contribution to that discussion. They used data from the Veterans Health Administration to examine trends in mortality 30 days following hospital admission for veterans with heart failure or pneumonia. Risk-adjustment models that used only comorbidity data to assess risk of death overestimated declines in risk-adjusted mortality compared with models that used measures of severity of illness. This finding is important by itself, but it also has broader ramifications.
Source: JAMA Internal Medicine - Category: Internal Medicine Source Type: research