The magnitude of the various quality improvement initiatives has generated high expectations for the use of valid, objective, and reliable performance measures. All of these initiatives—public reporting, quality improvement within providers’ offices, and pay for performance—depend upon the availability of an array of measures whose implementation can contribute to realizing the fundamental aims of the nation’s health care system. The committee concludes that federal leadership is necessary to ensure that performance measures address all six aims, as well as to balance private-sector initiatives with investments in quality measures for neglected areas that may lack strong constituencies. Federal leadership is also essential to provide stability, coordination, and direction when fluctuations and tensions in the health care system create an unpredictable environment for data collection and reporting. The challenge is multifaceted:
To identify the national goals that performance measures should serve.
To clarify data requirements in areas in which multiple measures have been proposed.
To identify areas in which greater effort is needed.
To build the capacity to produce, report, and analyze performance data throughout the public and private sectors of the health care system.
Quality-related efforts in all of the areas noted above—public reporting, quality improvement initiatives, and pay for performance—rely on some form of performance measurement and reporting. Components of a system that can perform these functions include the following:
Standardized performance measures—Performance measures include measures of the health care process (e.g., periodic blood and urine tests for diabetic patients), patient outcomes (e.g., 60-day survival rate for cardiac bypass patients), patient perceptions of care (e.g., experience with patient–provider communication), and organizational structure and systems associated with the ability to provide high-quality care (e.g., medication order entry systems). Standardized performance measures are those with detailed specifications (e.g., definitions for the numerator and denominator, sampling strategy if appropriate) allowing for “apples-to-apples” comparisons, sometimes requiring effective risk adjustment or stratification of results across key subgroups.
Access to patient data—Calculation of many performance measures requires access to patient-level data from administrative files and chart