We propose to conduct a randomized controlled trial comparing alternative strategies for implementing national guidelines related to the care of hypertension and depression at the largest HMO in Western Washington (Group Health Cooperative) and at two hospitals affiliated with the U. of Washington (Harborview Medical Center and the Seattle V. A. Medical Center). As part of a factorial design, 12 small groups or """"""""pods"""""""" of primary care physicians will be assigned to standard care versus two intervention arms: (1) a """"""""physician-oriented"""""""" program of education and computerized feedback, led by a recognized opinion leader as part of an academic detailing program; and (2) the physician-oriented program plus a """"""""system-oriented"""""""" program already in use at the Cooperative which incorporates principles of CQI to engender changes in the delivery infrastructure. Within each pod the care afforded 460 diagnosed hypertensives and 150 diagnosed depressed patients on average will be examined. Given that the ultimate purpose of guideline implementation is the improvement of patient outcomes, the trial's primary end points will be two disease- specific intermediate outcomes: (1) the percentage of hypertensive patients controlled (diastolic < 90mm Hg) and optimally controlled (diastolic 80-90mm Hg); and (2) scores on the SCL-90-R, Depression Scale. Supplementary outcomes will include measures of patient general health status, and patient satisfaction. To better understand the process of how interventions are translated into observed changes in outcomes, additional data will be collected to determine the interventions' effects on: (1) compliance with guideline- recommended case finding, and pharmacologic and non-pharmacologic therapeutic modalities; and (2) overall and disease-related patient resource utilization. Thus, the trial's three major objectives are: (1) to provide data on the relative effectiveness of approaches to quality improvement that emphasize individual-physician and/or organizational responsibility for patient outcomes; (2) to conduct a preliminary comparison of the direct and time costs of implementing these interventions; and (3) to increase our understanding of why these approaches differ in their ability to obviate barriers to practice pattern change, if differences in effectiveness between them are found.
