Preliminary Tests of a 6-Item General Health Survey

A Patient Application

Published in: Measuring Functioning and Well-Being: The Medical Outcomes Study Approach / edited by Anita L. Stewart and John E. Ware, Jr., (Durham, N.C.: Duke University Press, 1992), Chapter 17, p. 291-303

Posted on RAND.org on January 01, 1992

by John E. Ware, Eugene C. Nelson, Cathy D. Sherbourne, Anita Stewart

To experiment with the extremes of the trade-offs involved in using short-form and full-length scales, the MOS staff developed a short-form survey to be administered in less than 2 minutes. The trade-offs include the choice between breadth (comprehensiveness and content validity) and depth (precision) of measurement. To achieve breadth, the MOS selected six health concepts: physical, social, and role functioning; bodily pain; mental health in terms of psychological distress; and general health perceptions. Depth of measurement was limited to the precision that could be achieved using a single questionnaire item defining five or six levels for each concept. In physical functioning, a noteworthy number of physical limitations may have been missed by the single-item measure. In social functioning, 52 percent of those surveyed indicated that they had no limitations. Role limitations were assessed using an item that asked about difficulty in performing daily activities as a result of physical health or emotional problems, and a second item was administered in the MOS to explore whether any difficulties reported were due mostly or entirely to physical or emotional causes. Both single-item measures must be presented in tandem. As to pain, the size of the intervals between item levels should be taken into account when scoring those levels. The most clinically and socially relevant item is psychological distress. The general health item achieved the closest approximation to a normal distribution of responses observed for any of the single-item scales. The coarseness of single-item measures will limit their usefulness. Preliminary results for these scales support their use in studies based on group-level analyses (rather than of individual patients).

This report is part of the RAND Corporation external publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.

The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.