Process-Based Measures of Quality: The Need for Detailed Clinical Data in Large Health Care Databases

Abstract

The Institute of Medicine defines health care quality as increasing “the likelihood of desired health outcomes” using “services … consistent with current professional knowledge.” This definition implies that quality measures can be based on either achieving health care outcomes or completing processes that experts agree have been shown by scientific evidence to improve outcomes. Process-based measures are especially suitable when the user needs to know how to improve quality, when provider comparisons show equivalent outcomes but all providers should improve processes, when measures are needed to evaluate health care that is intended to improve long-term outcomes, or when the contribution of individual providers (especially providers who have a small number of cases) needs to be defined. However, many different process-based measures are needed to comprehensively assess quality, and many process-based measures require detailed clinical data currently found only in medical records. Therefore, the expense of abstracting records is a barrier to process-based measurement. Fortunately, large-scale process-based measures are becoming more feasible because the required clinical data are being included in large databases. The merging of existing inpatient and outpatient databases with pharmacy and laboratory databases is an important step toward obtaining data that link all patient admissions, appointments, diagnostic procedures, and prescriptions with diagnoses and test results. Other data that are valuable for process-based measures must still be obtained by abstracting data from records, including clinical findings, patient preferences, and medical and family history. In the future, such data may be added to large databases to create computerized medical records.