Redesigning Primary Care Practice to Incorporate Health Behavior Change: Prescription for Health Round-2 Results

Abstract

When Prescription for Health was conceived by the Robert Wood Johnson Foundation (the Foundation), primary care practice–based research networks (PBRNs) and their community, practice, academic, and healthcare system partners began with hard-won local knowledge. They worked to integrate this with a platform of basic organizational and behavioral science research to provide a foundation on which to build a program for testing practical approaches in frontline clinical practice to help people change unhealthy behaviors. This intellectual foundation included the components of the chronic care model, the 5A’s model of behavioral counseling, the transtheoretical model of behavior change, and models of diffusion. It was also apparent that the high prevalence and clustering of unhealthy behaviors within individuals and populations would make addressing multiple behaviors and problems—sometimes simultaneously—a pragmatic necessity in primary care settings. It was hoped that addressing multiple behaviors would result in synergy among practice systems, patients, families, and communities, and thereby position redesigned primary care practices to contribute to decreasing premature mortality, avoidable morbidity, and escalating healthcare expenditures. After decades of limited success in incorporating health behavior issues into medical practice, a key challenge was how to engage those primary care clinicians who are in smallto medium-sized primary care practices, the dominant model of primary care delivery in the healthcare system. The strategy selected was to work with the nation’s PBRNs. Thirty years from their inception, approximately 100 of these networks of primary care practices were in existence, comprising mostly full-time clinicians working in their communities, asking and answering questions important to the health of their patients. Practices in these PBRNs are known to be similar to randomly chosen primary care practices and to be relevant sites for research. The Foundation decided to offer funding to primary care PBRNs to test their best ideas to help their patients modify their behaviors related to tobacco use, unhealthy diet, physical inactivity, and risky alcohol use. All projects were required to develop approaches capable of addressing two or more of these behaviors in the first round of Prescription for Health and all four target behaviors in the second round. The feasibility of working with primary care practices was shown with a first round of work by 17 PBRNs that was summarized in a supplemental issue of the Annals of Family Medicine. This early work confirmed that primary care practices were interested and willing to address health behavior issues when supported by grant funds and PBRN infrastructure. It established that health behavior counseling can be done in frontline primary care practices, and that doing so requires substantive practice redesign. It revealed the need and possibility of the explicit integration of primary care with public health and community resources. It also confirmed the utility and limitations of adapting existing models and theories into primary care, and that a co-evolutionary approach across projects created synergistic learning. This round of work was brief, focused on feasibility testing, and did not fund comprehensive trials. These successes inspired another round of work that is the basis for the papers in this supplement to the American Journal of Preventive Medicine. In Round 2 of Prescription for Health, ten PBRNs (five PBRNs from Round 1 and five new to the program) implemented 2-year studies that tested behavior-change counseling strategies, and again joined together with an imbedded evaluation team to learn what happened. The tested interventions included a broad mix of strategies and required a spectrum of evaluative methods, including surveys, interviews, focus groups, medical record reviews, site visits, and collaborative meetings. All PBRNs collected a set of common patient-oriented measures and provided systematic reports about pracFrom the Director, Prescription for Health, and Professor and Epperson-Zorn Chair for Innovation in Family Medicine (Green), the Department of Family Medicine (Green, Cifuentes), School of Medicine, University of Colorado Denver, Aurora; Kaiser Permanente (Glasgow), Denver, Colorado; and the Departments of Family Medicine, Epidemiology and Biostatistics and Sociology at Case Western Reserve University and the Case Comprehensive Cancer Center (Stange), Cleveland, Ohio Address correspondence and reprint requests to: Lawrence W. Green, DrPH, University of California San Francisco, 185 Berry Street, Suite 6650, San Francisco CA 94143-0981. E-mail: LGreen@ cc.ucsf.edu.