Counterpoint: public disclosure of process and outcome measures.

Bad data is a toxic substance. In the release of process and outcome measures in the field of health care, numerous examples exist of published bad data. In 1986, the Healthcare Financing Administration released 14 volumes of data concerning Medicare mortality rates which, on analysis, were misleading and unrelated to quality of care. Good data on outcome measures need to follow accepted, rational, and scientific procedures. Such procedures include appropriate definitions of the process or outcome to be measured and careful description of the population being observed, risk adjusted for severity of illness. When this is done, the data can be published with some confidence that it will have value.