Public health practitioners are increasingly being asked to partner with people in nonhealth sectors of the community to develop policies and build environments that will support health-promoting nutrition and physical activity behaviors (1-7). For example, Preventing Childhood Obesity (2), a recent report by the Institute of Medicine, recommends that state and local governments work with communities to expand and support opportunities for physical activity and access to healthy foods. Programs for successful obesity prevention and reduction need to be multilevel, community based, and sustainable (2,8). Populationwide changes in behaviors require interventions that address policies that affect nutrition and physical activity environments (5,9).

Critical factors for successful policy and environmental change include collaboration, support from community decision makers, and data that favor the intervention, in addition to funding, other resources, and skilled staff (1). Successful collaborations and community support require durable planning structures and social capital or social readiness (10,11). The elements of social capital — social relationships, social networks, social norms and values, and trust — reflect dynamic processes and interactions (12). Meaningful assessment of social capital accounts for its dynamic nature and is more complex than simply quantifying how many people are involved and which agencies are represented.

Community partnerships are part of social capital. Understanding characteristics of a community partnership provides insight into how an active partnership is functioning and how this functioning affects project implementation. Public health practitioners typically evaluate community partnership projects by identifying actions and outputs that measure whether a project is meeting its objectives and goals. Equally important but perhaps more difficult to evaluate is the nature of the partnership itself — the ways that members come together and interact and how the work of the project is accomplished. Public health practitioners need to understand community partnerships as social capital to provide guidance for setting goals and objectives and to give appropriate technical assistance. In this article, we describe the evolution of methods and tools that were used to identify and understand the characteristics, structures, and processes of a community partnership for health improvement in Washington State.

The Washington State Department of Health (DOH) has lead efforts to prevent obesity and overweight on both the state and community policy levels. Community efforts have been organized as Healthy Communities projects. With funding from the Centers for Disease Control and Prevention (CDC), DOH recruited the city of Moses Lake in 2002 as the first pilot community to test implementation of the Washington State Nutrition and Physical Activity Plan (13). Table 1 provides a timeline of the Healthy Communities projects.

Initial planning for the Moses Lake pilot project was conducted by an advisory committee composed of leaders and representatives of civic organizations, city and county agencies, businesses, and interest groups as well as community residents. Members of the advisory committee identified three initial strategies from the state plan and created and formalized an action plan for their implementation. In the next phase, activity shifted from the advisory committee to the three project teams: 1) trails and paths, 2) breastfeeding, and 3) community garden. Many members of the advisory committee served on a project team and recruited new members. During this phase, the advisory committee as a whole met less frequently. The leadership group (planning team), including the project coordinator, advisory committee, three team leaders, and DOH staff, continued to meet to coordinate the subprojects and monitor overall progress.

The evaluation team for Healthy Communities included staff from the University of Washington (UW) and DOH. Because our initial tools did not adequately explain the Moses Lake pilot project experiences, we searched for a theoretical framework that would include constructs that describe factors we identified as barriers to or enhancers of the success of Healthy Communities Moses Lake. We conducted a literature review of available frameworks and identified the community health governance (CHG) model developed by Lasker and Weiss (14) as a best fit for existing data from Moses Lake. The CHG model served as the basis for an evaluation tool that was applied to the second Healthy Communities pilot project, the city of Mount Vernon, from its beginning.

Mount Vernon was selected as the second pilot city in 2003. The Mount Vernon project's assessment, planning, and project team organizational phases were similar to those of Moses Lake, although the projects and composition of committee partnerships were different. DOH and the pilot project communities also had distinct objectives and deadlines according to their interests.

Data collected in the first 2 years of the Healthy Communities project were used for qualitative and quantitative process evaluation. These data were collected through two telephone interview surveys of advisory committee members and a survey of project team members from each project: the trails planning team, the Breastfeeding Coalition, and the community garden team. Supplementary data included observations during meetings and events, meeting evaluations and minutes, and debriefing of project staff (Table 2). The initial evaluation plan for Healthy Communities Moses Lake was developed in partnership with action-oriented stakeholders. It examined some aspects of collaborative processes but was not based on an explicit overarching theoretical model.

The UW evaluation staff invited Healthy Communities participants to respond to the surveys. The surveys were designed to be as brief as possible while providing sufficient detail on partnership functioning, issues of interest to advisory committee members, processes of program planning and implementation, and use of technical assistance to inform the public health practitioners. Survey results were compiled and reported to DOH staff, who shared the information with the respective community leadership group that in turn reported to its advisory committee.

The Healthy Communities project outlined in the Washington State Nutrition and Physical Activity Plan was based on the social–ecological model (5), and development of the intervention and evaluation methods were guided by understanding of obesogenic environments (5) and values of participatory research and empowerment (22-24). We find that the CHG model (14) includes most of the identified elements and may prove to be a useful resource to public health practitioners who are working with communities to facilitate policy and environmental changes.

Integral to the CHG model is the assumption that processes of community participation are crucial to effective solutions for community health problems (14). The model is rooted in the belief that sustainable solutions to many adverse health outcomes will be found only when people and organizations come together to address the social, economic, political, and environmental determinants of these outcomes. In the CHG model, the role of governmental agencies such as DOH changes from a director exerting control to a community advisor that actively participates in and drives the process of positive change. This change in role for governmental agencies represents a new paradigm for public health practitioners in community-based health promotion projects.

Similarly, the CHG model affects program evaluation. Rather than focusing only on behavior change and long-term health outcomes as measures of health promotion interventions, the CHG model provides a framework for examining the intermediate processes of the partnership as proximal outcomes. Use of the CHG model to organize information can aid in assessing processes in which individuals and organizations work together to identify and address health problems at the community level. The model can also help to guide public health workers in project management. A well-designed process evaluation guides the use of limited staff time and project funding and thus can improve efficiency and effectiveness.

The CHG model shows how the dynamic and complex interactions of community partners using community resources can lead to improved community health. The model identifies markers that can reinforce partnership activities. The markers, such as representation of community diversity and mechanisms for accountability within a collaborative process, are not ends in themselves, but the presence and strength of these markers become guides for public health practitioners in promoting community-based interventions.

The CHG model served as a guide for evaluating the Healthy Communities pilot projects in Washington. The model was used to develop a telephone survey tool to assess community partners' perceptions of a Healthy Communities project. Results from the survey provided feedback for DOH staff, university partners, and community leaders. In addition to scaled questions that addressed elements of the CHG model, several open-ended questions were important for identifying issues raised by the scaled questions. These open-ended questions provided the opportunity for respondents to voice their opinions and provided essential information for translating survey findings into meaningful directions for actions.