Mortality rates as a quality indicator: a simple answer to a complex question.

Clinical, administrative, and information issues have been reviewed that may impact on the sensitivity of mortality rates as a quality indicator. It is clear that current technology cannot necessarily identify high-risk institutions through the use of abstracted discharge data. Furthermore, even if this screening capability is improved, verification of quality problems still requires detailed chart review. Quality of health care delivery will continue to be scrutinized by various groups, only some familiar with clinical risk adjustment and the actual process of care. In order to promote the accuracy and fairness of the review process, objective, outcome-based criteria for high-quality care must be developed for many clinical situations, and clinicians must continue to be involved in quality assessment.