From 2002 to 2004, the 11 rural cancer coalitions in Pennsylvania and New York completed 1369 coalition initiatives. Of these initiatives, six of every seven were community interventions with one in seven being coalition development activities. Development activities are necessary for long-term sustainability of a cancer coalition because it is through these activities that coalitions recruit, educate, and equip coalition members. However, the appropriate ratio of community interventions to development activities for effective, efficient, and sustainable cancer coalitions has not been estimated. The ratio of community interventions to development activities for the 11 coalitions in this study was substantial and ranged from 1.3:1.0 to 16.8:1.0; the overall ratio of community interventions to development activities for this study was 5.2:1.0.
Within community interventions, the number and percentage of screening interventions (and education interventions generally) increased during the study period, whereas the number and percentage of outreach-only interventions decreased during the study period. In fact, outreach-only activities decreased from two in every three interventions to one in every two interventions. This appropriate increase in proportion of interventions may be partially attributed to the participatory partnership between the coalitions and the academic researchers. The research team provided monthly technical assistance, support, and training that strengthened coalition efforts, focus, and direction. We are encouraged by these changes because education and screening interventions are more likely to have a direct impact on cancer prevention and control than are outreach-only interventions. Additional years of data may establish that the temporal trend for these 11 coalitions was statistically significant.
It is difficult to associate specific outcomes, such as a community change in cancer risk, to outreach-only interventions because outreach-only interventions tend to be nonspecific and difficult to measure. Interventions that are more individualized and easier to measure would include those with an education or screening objective. However, outreach-only interventions may be an important component in a comprehensive mix of coalition interventions. Although the ideal ratio of outreach-only to targeted interventions is not known, the range for the 11 individual coalitions was 0.5:1.0 to 9.0:1.0, and the ratio for all coalitions was 2.1:1.0.
The measures and observed results of coalition effectiveness in this study, including the increasing trend in interventions, the substantial number of completed screenings, and documented community changes, are indicators of the success of the 11 coalitions over the 3-year study period. The CCAT model (20
) posits that the combined resources and intervention strategies of coalition members and their partners can improve health outcomes and lead to sustainable community change. The 11 coalitions and their academic partners in this study were able to achieve more through their participatory partnership than any one of the coalitions or the research team alone could have achieved, which is a principal tenet of community-based participatory research (36
). Interventions were possible because of the long history and trusted relationships between the coalitions and their communities. The researchers added scientific rigor and a system for documenting processes and impact of the coalitions' development activities and interventions. Thus, the partnership of coalitions and academic researchers greatly enhanced the potential to reduce the cancer burden in this rural Appalachian population.
This study is limited in several aspects. First, the validity and reliability of this coalition data system have not been formally tested. However, extensive training of field staff, their ongoing use of data manuals, as well as quality-control procedures and oversight of the project director improved reliability and reduced potential misclassification bias. In addition, this data system was developed from an established community-based intervention model. Second, a primary impact, community change, was infrequently reported. Many coalitions rarely focused on community change as a primary objective and focused instead on community education, screening recruitment, and cancer prevention. Anecdotal reports indicate that community changes occurred more frequently than were recorded in the data system. This is partially because a substantial amount of time frequently lapsed between an intervention and the resulting sustainable community change. This time delay may be responsible for the lack of recorded data on community change in the data system. Therefore, we believe that we have underestimated the true number of community changes that resulted, at least in part, from these 11 rural coalitions. Third and finally, because the population was mostly rural and white, the results cannot be generalized to other populations. Further studies are needed to establish external validity to nonrural and diverse racial and ethnic groups.
Despite these limitations, this study has several strengths. First, this study documents the activities and impact of 11 community cancer coalitions over a 3-year study period. These coalitions have been in existence for more than a decade, so these results represent coalitions with an extensive history. More importantly, these data were captured through a model- and Web-based approach for coalition-driven, community interventions in cancer prevention and control. The Web-based data system is mutually beneficial to both researchers and communities, particularly those in rural, less accessible areas. By allowing remote data entry by regional field staff who reside near coalitions that are dispersed across a wide geographical area, the data system enhances timely submission of coalition data, provides technical assistance and communication with the coalitions, and reduces travel costs. In turn, coalitions receive data for strategic planning, new member recruitment, and program evaluation and report these data back to their communities. The Internet has been used to assist women in rural areas with day-to-day management of breast cancer (37
), provide electronic support groups for breast cancer patients (38
), and improve quality of life for low-income women with breast cancer (39
). However, to our knowledge, this data collection system is relatively unique for evaluating community-based cancer prevention and control initiatives. Finally, this study provided estimates of a mix of various development activities and community interventions, aggregated and compared across 3 years, against which comparisons with this network or other networks of community coalitions may be made. Analyses of future coalition outcomes examined in comparison with the current study will help establish the long-term sustainability of these coalitions and provide recommendations for an appropriate mix of coalition development activities and interventions that achieve measurable reduction in cancer burden for underserved populations.
This study demonstrates that community cancer coalitions in Pennsylvania and New York reached rural residents through cancer prevention and early detection education and screening interventions. Additional research is needed to determine the appropriate mix of development activities and community interventions for coalitions to achieve their cancer prevention and control goals. In addition, evidence-based interventions are needed for coalitions to meet the cancer-related needs of specific populations. Finally, improved measurement of cancer screenings and community change will help coalitions document progress toward their overall goal of reducing cancer disparities in rural Appalachia.