ABSTRACT
The usefulness of an action-research model is demonstrated in the evaluation and improvement of ambulatory medical care in a variety of settings: solo office practice, prepaid capitation multiple-specialty group practice, and medical school hospital-based outpatient clinic practice. Improvements in the process of medical care are found to relate directly to the intensity and duration of planned interventions by the study group and are demonstrated to follow organizational changes in the participating sites--primarily managerial and support services initiated by policy decisions in each study site. Improvement in performance approaching one standard deviation results from the most intense intervention, about one-half standard deviation at the next level of intervention, and virtually no change from a simple feedback of performance measures. On the basis of these findings and other operational and research efforts to improve physician performance, it is unlikely that simple feedback of performance measures will elicit a change in behavior. However, noncoercive methods involving health care providers in problem identification, problem solving, and solution implementation are demonstrated to be effective.
