542 zyxwvutsrqp COMPARATIVE MEASUREMENT EFFICIENCY AND SENSITIVITY OF FIVE HEALTH STATUS INSTRUMENTS FOR ARTHRITIS RESEARCH MATTHEW H. LIANG, MARTIN G. LARSON, KAREN E. CULLEN, and JULIE A. SCHWARTZ zy Five health status instruments were administered in random order to 50 arthritis patients before and after total joint arthroplasty. Relative efficiency and sensitiv- ity in measurement of change in pain, mobility and physical function, social role and social activity, and global health were assessed. The instruments had highly correlated scores, but had differences in certain dimen- sions. Inter-instrument differentials were larger for social and global outcomes than for pain or mobility. No single instrument consistently outperformed the others. A method for determining relative efficiency is de- scribed. Over the last 10 years, the measurement of outcome in chronic disease has been pursued actively. For the rheumatic diseases, specific instruments mea- suring quality of life or health status have recently been developed, building on a strong tradition of concern with the accurate measurement of musculo- skeletal function. Evaluating the results of health care in its multiple dimensions and from the patient’s perceptions represents an important shift in health From the Departments of Medicine, Kheumatology/lmmu- nology, and Orthopedic Surgery and the Robert B. Brigham Multi- purpose Arthritis Center, Brigham and Women’s Hospital, Harvard Medical School, and the Department of Biostatistics, Harvard School of Public Health. Boston. Massachusetts. Supported by NIH Multipurpose Arthritis Center grant AM-20580 and Biomedical Research Support grant RR-05669. Matthew H. Liang, MD, MPH; Martin G. Larson, ScD; Karen E. Cullen, MEd; Julie A. Schwartz, BS. Address reprint requests to Matthew H. Liang, MD, Robert B. Brigham Multipurpose Arthritis Center, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 021 15. Submitted for publication March 27, 1984; accepted in revised form October 23, 1984. care assessment which has posed new questions. Among these are whether the new instruments have sufficient sensitivity to measure clinically important changes (1,2) and which instruments are most efficient for detecting these changes. We present the results of a study of the relative measurement sensitivity of zyxw 5 widely-used health status indicators as applied to patients with end-stage joint disease who were undergoing primary total joint re- placement. This population was chosen because total joint replacement is a unique pulse intervention in arthritis management, resulting in nearly certain im- provement in pain and function (3). Patients predict- ably attain much of their expected improvement by 3 months after discharge. Thus, the population provided a high likelihood of change on which the instruments could be tested. These results should assist the investigator who is planning a study in which improved functional performance is likely and who must choose among the many available instruments. Our premise is that, all other things being equal, the statistically most power- ful instrument, with highest probability of demonstrat- ing significant change, would be the most desirable. PATIENTS AND METHODS Five health status instruments which fulfill the mini- mum requirement of measurement reliability and validity were studied: the Functional Status Index (FSI) (4), the Health Assessment Questionnaire (HAQ) zyx (3, the Arthritis Impact Measurement Scales (AIMS) (6,7), the Index of Well-Being (IWB; also referred to as the Quality of Well- Being Instrument) (8), and the Sickness Impact Profile (SIP) (9). The first 3 instruments were developed for patients with rheumatoid arthritis and have been used extensively. The Arthritis and Rheumatism, Vol. 28, Nu. 5 (May 1985)