All-or-none measurement raises the bar on performance.

THE PURSUIT OF EVIDENCE-BASED MEDICINE IS NOW AT the core of the agenda for improving health care in the United States. All major quality measurement systems use science-based indicators of proper processes of care, such as the ORYX measures of the Joint Commission on Accreditation of Healthcare Organizations, the Health Employer Data and Information Sets measures of the National Committee on Quality Assurance, the measures used by the Quality Improvement Organizations under contract with the Centers for Medicare & Medicaid Services, and at least 70 of the 179 measures in the 2004 National Health Care Quality Report from the Agency for Healthcare Research and Quality. Often, several individual performance measures are used to assess care of the same condition. For example, a recent summary of data on the Joint Commission on Accreditation of Healthcare Organizations’ standardized performance measures included 9 measures for acute myocardial infarction, 4 measures for congestive heart failure, and 5 measures for pneumonia. The Joint Commission on Accreditation of Healthcare Organizations’ standard set for pneumonia measures performance with respect to oxygenation assessment, pneumococcal vaccination, blood cultures, smoking cessation counseling, and mean time to initiation of antibiotics. The first 4 of these are discrete measures that indicate the presence or absence of the item in the medical record. The time to initiation of antibiotics is a continuous variable measured in minutes. There are at least 3 different options for calculating performance on multiple, discrete measures for the same condition.