Methods of Sampling

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 3, p. 27-47

Posted on on January 01, 1992

by William H. Rogers, Elizabeth A. McGlynn, Sandra H. Berry, Eugene C. Nelson, Edward Perrin, Michael Zubkoff, Sheldon Greenfield, Kenneth B. Wells, Anita Stewart, Sharon B. Arnold, et al.

The basic MOS sample design involved a five-stage process: (1) identify appropriate geographic sites in which to conduct the study; (2) select health maintenance organizations and multispecialty group practices within those sites; (3) select physicians and other providers within each practice type; (4) select patients within those practices for the cross-sectional survey; and (5) select patients for the longitudinal panel. The MOS had a large cross-sectional component. In implementing the design, it was necessary to balance the most generalizable sample to maximize the external validity (see Chapter V-18) and cases that best met the criteria for the tracer conditions---diabetes, hypertension, depression, and heart disease---so that a comparison of outcomes could be made over time. As well, it was important to obtain patient and provider cooperation and maintain it over time. This chapter provides an overview of each stage of the process, detailing selection criteria and important considerations in each stage. For example, the average age of participating clinicians was 40, and a patient may have been excluded for one tracer condition but included for another. The chapter concludes with a summary of the characteristics of the sample of clinicians and patient participants, by type of health care system and site.

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.