LGBT Appalachians and tobacco use: the role of social connectedness

Social network analysis applied to health behaviors is an innovative technique that is quickly becoming a major focus of health systems research. I plan to recruit and survey via internet social networking sites, which is also an emerging technology and methodology that shows promise in achieving more representative sampling of populations considered "hidden" or hard-to-reach by traditional research methods. (Dr. Jeff Jones at the University of Kentucky is completing such a study surveying men who have sex with men [MSM] across the state. Advisors Riggle and Rostosky used Facebook to recruit participants for the research in their latest book22. Dr. Jose Bauermeister at the University of Michigan is also running several studies using this methodology in an urban setting. ) These methods can also inform the feasibility of including other innovative techniques such as text-message reporting in real time for intervention studies. Additionally, the focus on rural Appalachian Lesbian, Gay, Bisexual & Transgender (LGBT) is new and greatly needed, as even the sparse existing literature on LGBT people in the South or Appalachian region tend to recruit participants from relatively urban centers. Preliminary Data As a new area of research both in general and for myself as a researcher, I do not have preliminary data to contribute. It will, however, draw on work of my mentors and advisors. Dr. Schoenberg has demonstrated the ability to recruit and retain over 600 Appalachian residents in a community-based smoking cessation intervention. Drs. Riggle and Rostosky have found examples and models of resiliency among rural LGBT persons. Dr. Stoops has demonstrated the feasibility and short-term effectiveness of internet-based smoking cessation techniques among Appalachian residents, and Dr. Havens is currently performing a study that has successfully recruited difficult-to-reach intravenous drug users in Appalachia for a social network analysis. This group of mentors, advisors, and consultants has amassed a considerable amount of data and methodology expertise that will form the foundations and support structure for my study. Proposed experimental approach Stage 1 - Staged mixed methods: qualitative exploration and quantitative estimation of burden Because so little is known about LGBT Appalachians and their smoking behaviors, preliminary data are needed to formulate an appropriately targeted and conducted social network analysis. The purpose of the first stage is to provide those data. The hypotheses specifically being explored in Stage 1a are 1a,b; 2a,c; 3a-b. In Stage 1b are 1a,b, 2b, 3a. (See pg. 1-2.) 1A. Qualitative: Interviews - This portion will consist of one hour long, semi-structured interviews of key informants in order to gather perceptions of Appalachian LGBT social networks, including sense of isolation from local culture and other LGBT persons, impressions of smoking prevalence and how social contacts influence smoking. Recruitment will occur through advertising in the Gay and Lesbian Service Organization newsletter, which is widely distributed and read across Kentucky, and at a bar that caters to LGBT patrons in Hazard KY. Further recruitment will proceed via respondent-driven, or "snowball," sampling. Participants will be 18 years of age and over, currently living in Kentucky Area Development District 9-13, a grouping of counties organized for economic development that conveniently define Appalachian Kentucky24, and self-identify as a sexual minority. Target number of participants is 10-15 and each will be paid $25 for his or her time. Interviews will be audio taped and transcribed. Transcripts will be analyzed by 2 members of a research team using Grounded Theory approach to identify themes and subthemes, as well as perceived relationships between social networks, substance use, and health system factors. This data will be used to inform development of the survey in 1B as well as the social network analysis of Stage 2. 1B. Quantitative: Facebook survey - Although it is not possible to determine true population of LGBT-identified persons in Appalachia or smoking prevalence without a population-based study and random sampling, recruitment through widely-used social networking internet sites allows for participation from members of stigmatized groups who might not be willing to identify themselves or due to isolation might not be reached through typical recruitment or research efforts. This approach allows for a link to an anonymous survey with waiver of documentation of consent, making the sharing of potentially sensitive information more likely. Survey research using Facebook is commonly used by market researchers and becoming a viable academic research vehicle as well. Prior research has demonstrated the feasibility of avoiding multiple entries by the same participant, and allowing for a mechanism to receive an incentive such as a gift card22. Therefore, the purpose of the Facebook survey is to estimate smoking prevalence in LGBT Appalachians and to correlate smoking behaviors with measures of social isolation, psychosocial distress, health care access, and simple measures of social network position (e.g. number of LGBT and non-LGBT friends visited in the last month). There are a number of validated scales from which questions can be taken or adapted. The qualitative data in Stage 1A will be used to choose which questions and measures are most appropriate. Inclusion criteria are self-identified LGBT 18 years and older, currently living in ADD 9-13, and ability to read and write in English. The survey will be designed to take no longer than 20 minutes and will be anonymous. We will invite participants to fill out a separate contact form for recruitment in the social network analysis in Stage 2. Participants will be offered a $20 gift card obtained through a link separate from the research site. Because these are pilot data, the target number is as large as possible; however, we will perform a power calculation to determine a minimum number