Combining Community-Based Participatory Research (CBPR) with a Random-Sample Survey to Assess Smoking Prevalence in an Under-Served Community

Antoine Messiah,Noella A Dietz,Margaret M Byrne,Monica Webb Hooper,Cristina A Fernandez,Elizabeth A Baker,Marsha Stevens,Manuel Ocasio,Recinda L Sherman,Dorothy F Parker,David J Lee

DOI: https://doi.org/10.1016/S0027-9684(15)30030-4

Abstract:Acknowledgments: The authors would like to thank Laura McClure for her help with the manuscript submission, the Liberty City Community Health Advisory Board for its collaboration on this study, as well as the survey interviewers, and the survey participants. Introduction: Underserved communities might lag behind Healthy People 2010 objectives of smoking reduction because of smoking behavior disparities. This possibility was investigated through a random-sample survey conducted in a disenfranchised community in Miami-Dade County, Florida, using a Community-Based Participatory Research (CBPR) framework. The survey was triggered by our finding that this community had higher than expected incidence of tobacco-associated cancers. Methods: Survey methods, resulting from a dialog between the Community Advisory Board and academic researchers, included: (a) surveying adult residents of a public housing complex located within the community; (b) probability sampling; (c) face-to-face interviews administered by trained community residents. 250 households were sampled from 750 addresses provided by the county Public Housing Agency. The completed surveys were reviewed by the academic team, yielding 204 questionnaires for the current analysis. Results: Of the 204 respondents, 38% were current smokers. They estimated the percentages of smokers in their household and among their five best friends at 33% and 42%, respectively, and among adults and youth in the community at 72% and 53%, respectively. Conclusions: A mix of state-of-art methodology with CBPR principles is seldom encountered in the current literature. It allowed the research team to find a high smoking prevalence in an underserved community, twice the statewide and nationwide estimates. Similar or higher levels of smoking were perceived in respondent's entourage. Such disparity in smoking behavior, unlikely to result from self-selection bias because of our rigorous methodology, calls for community-specific tobacco control efforts commensurate to the magnitude of the problem.