In an effort to understand the disparity in cancer rates, we recruited unofficial community leaders who were concerned about cancer to join the work groups. These community leaders helped us select work group participants with vital community connections and a personal interest in cancer. Credible cancer communication partners, especially during recruiting, are needed to conduct CBPR in Appalachian and other rural communities (8). We identified and recruited people who had personal experience with cancer (eg, survivors, caregivers) and other volunteers who were able to understand and discuss opinions about research (eg, educators, community advocates, church leaders). We intentionally did not select participants who were health professionals. All participants were willing to interpret and present research findings within their communities. The work groups were successful because the mix of community members represented a blend of diverse voices. Careful recruitment of the 24 participants (14 in Tennessee and 10 in Virginia) ensured diversity of professional backgrounds, educational levels, and demographic characteristics. Participants ranged in age from 30 to 65 years and included preachers, a funeral home director, current and retired school teachers, a librarian, a lawyer, housewives, and community activists. White, African American, and Hispanic community members participated, as did men and women of all ages. Work group moderators (K.A.D. and S.P.H.) reminded participants that they were not representing any particular demographic group, organization, or program. Instead, the moderators asked participants to focus on their understanding of cancer as a community problem and blend a variety of viewpoints from their communities.

During the work group meetings, 1 common question participants raised was, "What makes us [Appalachians] unique?" In addressing this question, both researchers and participants examined qualitative and quantitative research and identified possible characteristics, such as personal behaviors, literacy levels, and health care access. The work group approach provided a shared time and place for participants and researchers to discuss these factors within a regional framework.

First, the research team encouraged a "co-learning" climate. The project title, Community Cancer Research Review Work Groups, indicated the importance of participants' roles. We designed each work group session to present cancer research in an understandable way and allowed sufficient time for participants to question and interpret the studies.

Second, the tone set during the work group sessions ensured that all voices were heard. Moderators created an environment where community members freely examined the research and offered interpretations. For example, participants explained that the belief that "Cancer is a death sentence" (voiced in earlier focus groups of the Rural Appalachian Cancer Demonstration Program) (9) was held largely on the basis of Appalachian experiences associated with late-stage cancer diagnoses common in rural communities. Participants challenged the idea that "Appalachian fatalism" was fading or perhaps was being replaced by fatalism over an inability to effectively access the health care system. These types of alternative explanations and interpretations are an example of empowerment evaluation (10), an approach that directly involves communities in improving researchers' findings and translating findings into effective programs.

Third, the research team recognized that community members often cannot commit to long-term action and, therefore, required only a short-term commitment from participants. This respect of participants' busy schedules may have encouraged high rates of attendance (20 hours, on average, over the course of 3 months, including the five 2-hour work group sessions). The Table outlines the content of the 6 work group sessions. Sessions 1 through 4 were 2 hours and were facilitated by ETSU faculty members (K.A.D. and S.P.H.). For session 5, participants used information from sessions 1 through 4 to design cancer education events or activities in their communities, which they then organized and implemented at times convenient to community members shortly following sessions 1 through 4 during May 2006. For session 6, the work groups met together to report their activities and to further discuss the research from an Appalachian community perspective.

Community members knew that cancer dramatically affects their families and neighborhoods, but most were unaware of their actual "cancer numbers" (eg, incidence and mortality rates for their communities and national statistics on survivorship by type of cancer) (11). We presented cancer data in a variety of ways. Guest speakers began the first 4 sessions, presenting research findings during the first hour. During the second part of sessions 1 through 4, moderators (K.A.D. and S.P.H.) facilitated a discussion of work group member interpretations of the research. Participants learned new ways of questioning data, examining individual and collective community experiences, and generating group interpretations that began to create an Appalachian cultural lens for cancer.

In Session 1, a researcher (B.B.) presented color-coded maps from disparities studies and localized data tables, so participants could visually compare cancer rates in their region with state and national rates. Data were displayed by cancer type and by sex and age group of people diagnosed with cancer. Participants were able to examine objectively whether their intuition about cancer in their communities was accurate and to increase their understanding of Appalachian cancer disparities. Session 2 introduced a series of quantitative studies from the literature, the Appalachian Regional Commission, and the Rural Appalachian Cancer Demonstration Program qualitative studies (1,5). These findings reinforced participants' concerns about barriers to effective communication between patients and health care providers, such as talking about cancer with physicians, including physicians from foreign countries. Participants cited their own communication experiences during stressful events, such as receiving a cancer diagnosis. Participants said health professions schools and health systems should do a better job of teaching and using culturally appropriate communication.

Session 3 introduced basic research principles and types of study designs. B.B. explained the institutional review board approval process for the work group study, including issues of human subjects' protection. During this session, participants expressed concern about cancer research being fragmented, with many organizations competing for funds. Additionally, participants highlighted the need to share new cancer research findings in a comprehensive manner and stressed the need for improved communication about research with community members.

Participants were unaware of state cancer plans. They suggested this might be because of the distance between mountain counties and state capitals, where, they perceived, health care decisions are made. In session 4, State Comprehensive Cancer Control Plan representatives from Nashville, Tennessee, and Richmond, Virginia, explained their state plans. Participants asked questions, clarified information, and expressed concern that they did not know who from their communities had contributed to the state plans or served as local members of statewide coalitions. They explained that a lack of involvement in statewide cancer coalitions might reflect the Appalachian sense of distrust toward government and the regional history of communities "taking care of our own" rather than relying on state help.

Work group members agreed to formulate their own cancer control messages and initiatives on the basis of what they learned from sessions 1 through 4. Session 5 provided participants an opportunity to design and lead local education activities in their communities. Each participant designed and delivered cancer control messages by using his or her preexisting social networks. Messages were disseminated in various ways: through their professions (eg, schools, libraries, funeral homes), their personal contacts (eg, civic groups, social groups, churches), and local community media (eg, newspaper editorials, articles). The most frequently communicated message was about high cancer mortality rates in Appalachia. Participants enthusiastically communicated messages through creative grassroots community approaches: talking about cancer disparities with a community quilting group, writing letters to their physicians to support hiring patient advocates, designing posters that target men at a local lawnmower repair shop, distributing pamphlets at employment offices, producing radio segments, and publishing letters to editors. Participants reported feeling like informed messengers, and stated that work group participation gave them something to take back to communities to potentially improve quality of life.