Predicting and Evaluating Patient Outcomes

Excerpt

Three methods can be used to improve medical science: observation, experimentation, and measurement. In an article in this issue of the Annals, Dubois and Brook (1) describe how indirect observation of clinical practice and explicit measurement of patient risk factors may identify patients whose deaths could have been prevented.

This study and related reports (2, 3) would have implications for physicians who treat acutely ill hospitalized patients regardless of when they were published. Coming at a time when this country is beginning to compare hospital performance through the public release of nationwide mortality rates, the insight and lessons of this