MEDICAL CARE Volume 38, Number 9‚ Supplement II, pp II-184–II-191 ©2000 Lippincott Williams & Wilkins, Inc. A Critique of Current Uses of Health Status for the Assessment of Treatment Effectiveness and Quality of Care S HERRIE H. K APLAN , MPH, P H D,* R ICHARD L. K RAVITZ , MD, MSPH,† AND S HELDON G REENFIELD , MD* The great advances in the measurement of health-related quality of life (HRQOL) have yielded numerous brief, precise, and psychometrically sound instruments. However, despite their widespread use in the relatively narrow arena of randomized con- trolled clinical trials, 1– 4 a strong and consistent rela- tionship between patient-reported health status and health care has yet to be demonstrated. In fact, the applications of general health status measures to the evaluation of the “effectiveness” and “quality” of medical care have yielded a mixed and confusing picture. 5– 8 There have been few directional hypothesis-driven empirical studies that declare and evaluate the mechanism through which the process of medical care might affect change in HRQOL. 3 If not addressed, this absence of theoretically directed research will continue to restrict its use to research, without diffusion into practice. The core thesis of our comments on the excellent and carefully thought out report by Testa et al 9 in this issue of Medical Care is that unless there are clinically plausible and a priori specified links between medical care and HRQOL in a given study, any observed associations are likely to be spurious. It is one thing to assess the health of a community or population, the purpose for which general health status measures were originally in- tended. It is very different problem to investigate the relationship of an individual’s perception of his or her health to structural features or specific processes of medical care. With respect to “minimally important differences” in HRQOL, the perspective of the user of effective- ness or quality of care data directly affects the magnitude of these differences. For the patient, for example, the minimally important difference is the increment in health status that is “noticeable” as improvement or worsening. For the clinician, impor- tant differences are the amount of change in health status measures that would warrant a change in a patient’s or group’s treatment plan. For the policy maker, minimally important health status differences may be those unconfounded differences between populations who are cared for by different types of providers or different systems of care that would warrant a policy change, such as limiting HIV care to HIV specialists or subsidizing the primary care train- ing of family physicians over general internists. It could be argued that for policy purposes, the mag- nitude of observed health status differences needs to be quite large due to the array of competing expla- nations for and extraneous variables that contribute to differences in HRQOL between populations. The only way to assess the magnitude of mean- ingful health status differences for use in applied health policy purposes is to focus on specific, clinically plausible hypotheses. Suppose the hypo- thetical research findings in Table 1 were observed between 2 groups of patients undergoing a surgi- cal procedure in different systems of care. In each case, there is a 10-point net difference between health maintenance organizations (HMOs) and fee-for-service (FFS) organizations. Even with the assumption that the groups are otherwise equal with respect to patient characteristics, which, if any, differences are meaningful? We could argue that true variations in care would themselves more directly affect pain than physical functioning and *From the New England Medical Center, Primary Care Outcomes Research Institute, Boston, Massachusetts. †From the University of California, Davis, Center for Health Services Research in Primary Care, Sacramento, California. Address correspondence to: Sherrie H. Kaplan, MPH, PhD, Primary Care Outcomes Research Institute, 750 Washington St, NEMC #345, Boston, MA 02111. E-mail: skaplan1@lifespan.org Interpretation of HRQOL Study Results II-184