CLINICAL PRACTICE GUIDELINE IMPLEMENTATION DEMONSTRATION

临床实践指南来源于有效性、结果和

Clinical practice guidelines derived from effectiveness, outcomes, and appropriateness research are the focus of intensive research and policy activity within the U.S. health care system. Large-scale, well-funded efforts are underway throughout this system to develop, test and refine guidelines for a wide variety of medical conditions and procedures. The ultimate impact of these efforts on clinical practice remains unknown, however: information concerning processes of guideline dissemination, implementation, and use is scarce -- and evidence from previous efforts to change physician behavior suggests that, absent active dissemination and implementation efforts, guidelines may fail to have much influence on medical practice. The Guideline Implementation Demonstration study proposed here represents a rigorous, large-scale demonstration and experimental evaluation of two alternative guideline implementation strategies. Each strategy is grounded in behavioral science theory and health services research, and each has shown considerable promise in previous studies and demonstrations. Both alternatives represent potentially effective means of changing clinical practice patterns through guidelines. The proposed study aims to test these strategies in a randomized trial using 12 physician groups in 6 community hospitals -- settings similar to those in which the majority of complex medical care is delivered in the U.S. Two comprehensive, credible practice guidelines produced using the RAND appropriateness method will be examined: coronary angiography and hysterectomy. Implementation strategy effectiveness will be evaluated through explicit review of 2400 medical records for patients treated before and after the guideline implementation interventions; additional background data will be collected on hospital and physician characteristics and behaviors. Random assignment to intervention or control groups and other research design features will ensure validity and generalizability of the results.