You are developing a study to analyse the incidence of sexual behaviour between men in order to better understand the spread of HIV/AIDS in a small rural community. Sexual intercourse between men in this area is highly stigmatised and you are concerned that you may not find an appropriate number of participants for the study. What kinds of data gathering methods could you use to recruit study participants? Describe specific characteristics or adaptive techniques that could be used to gather valid and reliable data. Provide examples of social methods that are aimed at the collection of data for stigmatised or illicit behaviour.
This study aims to understand the sexual behaviours, attitudes to risk, and levels of health literacy of men who have sex with men (MSM) in a small rural community in sub-saharan Africa, in relation to the spread of HIV/AIDS. Research objectives include:
- To gather data to improve the understanding of the links between behaviour and the spread of HIV/AIDS in MSM in this location.
- To improve health literacy and awareness of HIV/AIDS transmission.
- To identify existing resources and community structures that could be utilised in health interventions.
Contextual considerations for the study will be explored, then study design itself, analysis and interpretations, followed by limitations and conclusions.
MSM, and individuals with HIV/AIDS have always been stigmatised: “What society judged was not the severity of the disease but the social acceptability of the individuals affected with it…” (Shilts, 1987). Studies show that while unprotected sex between MSM is commonplace in Africa, knowledge of the risk of HIV/AIDS is poor. Stigma, violence, and a lack of safe social and health resources is widely reported (Smith, 2009).
As with any public health issue, it is critical to recognise “the importance of taking into account the characteristics of the relationships and interactions in which risk takes place” (Van Campenhoudt et al., 1997, from Dulor and Hubert, 2000). In this case, the stigmatised and illicit nature of sex between men is one in a complex array of factors which influence attitudes to risk and decision making.
MSM are particularly vulnerable to HIV/AIDS due to transmission via unprotected sex (Jaffe, 2007, Smith, 2009). Although this has prompted effective interventions to reduce transmission, it is still difficult to access and identify MSN for the purpose of needs assessment and other health-focused research. Reasons include that homosexuality in some regions is illegal or stigmatised, sex may occur under coercion and force, or MSM may not identify as gay (McKenna, 1996, from Boyce et al, 2004).
This stigmatisation and marginalisation means that these groups are hard to reach, or hidden, and may be under-served by health research and interventions. Sydor (2013) describes this type of group as:
- Hard to reach: a population that is difficult to access.
- Hidden: a population with no defined limits, such that its exact size cannot be known.
- Sensitive subject: a subject that some people prefer not to discuss publicly.
Not only are these groups difficult to identify and access, but people who engage in illicit or stigmatised behaviour may be reluctant to provide personal information to researchers or health practitioners. MSM may perceive threats to their livelihoods, social status or lives, should their sexual behaviour be revealed.
A lack of understanding of research, mistrust in western medicine and a distrust of authority increases friction. Prevention and control programs for HIV/AIDS in sub-saharan Africa have seen patients refuse surgery or other medical treatment unless their traditional healer sanctions it. (Liverpool et al, 2004)
Another contextual factor to consider is that communities such as this may present low educational attainment (Williams, 2005); in one study, rural students in sub-saharan countries scored significantly lower in reading tests than their urban counterparts (Zhang, 2006). This low literacy alongside poor understanding of health risks can result in reduced condom use, low rates of HIV testing and higher incidence of HIV/AIDS, but crucially it also presents challenges in study design and obtaining informed consent.
Finally, exploring heuristics is important to this study in order to understand how people who may not possess the knowledge or information to make fully-informed, rational decisions make “short-cut” decisions instead: “Individuals use heuristics to rationalize uncertainty” (Bailey and Hutter, 2006). For instance, the perceived risk of HIV/AIDS may be lowered by knowing the sexual partner (Skidmore and Hayter, 2000), believing their partner is faithful, or sex occurring between consenting adults rather than being paid for (Vanlandingham and Trujillo, 2002). There may even be a belief that sex between men (opposed to women) does not pose a risk for HIV. (Zulu and Zulu, 2006)
The study design chosen for this research is based upon the Rapid Assessment and Response (RAR) model of cross-sectional community-based qualitative research. Through this, the study will focus on the settings and contexts of the sexual behaviours, health literacy and attitudes to risk of MSM in this location.
The study will adopt a Social Determinants of Health (SDH) approach, recognising that individual behaviours are not always an individual choice; health and choices are influenced by life circumstances and a complex web of social, political and economic conditions (Lins et al. 2010). As well as targeting individuals, interventions to control HIV should aim to change aspects of cultural and socio-economic context that increase the vulnerability of people and communities to HIV (Buve et al, 2001, from Marmot and Wilkinson, 2005). An SDH approach allows examination of systemic “causes of the causes” to better facilitate effective health interventions.
The study population is men aged 16+, living within this community who have engaged, or engage, in sex with men, even if they do not identify as being MSM.
Participants will be identified and accessed by working with existing community groups, aid and outreach workers, and medical professionals. Recruitment posters will be placed in suitable locations. Screening questions will be used to ensure participants meet the criteria for the research. Whilst social media may provide an avenue for promotion, the small geographic study size in this case makes this approach unlikely to prove worthwhile.
Snowball sampling will be employed to increase the size of the survey group in this study, as it is an effective method for sampling hidden populations (Atkinson, 2001). Participants will be asked to recruit friends and acquaintances who also meet the criteria. It is recognised that selection bias, gatekeeper bias and the risk of only recruiting participants from a sub-group (Becker 1963) may limit the validity of the samples. However, as qualitative research, this study does not require significant representativeness for generalisability, and as such these potential issues are not of great concern (Bonevski et al, 2014).
Small payments for time and expenses will be provided to participants. The size of this payment will ensure that participants are not put at a financial disadvantage through their involvement, whilst avoiding the potential for acquiescence bias.
Focus groups of six to ten individuals will be used to build the exploratory information base, while providing a platform for improving the health literacy of participants. Then private semi-structured, questionnaire-driven interviews will be conducted to dive deeper into beliefs and contexts.
Questionnaires will be pre-tested for time taken, ease of use and any emotionally overly-burdensome questions, using volunteers from the same region with similar socio-economic and literacy status to the participants.
Contextual questions include socio-demographic characteristics such as estimated birthdate, education and work status and multivariate poverty assessment. Sexual behaviour, spousal status, and sexual history will be discussed and recorded, alongside health literacy, including understanding of the transmission and effects of HIV/AIDS and other STIs.
The use of substances including drugs and alcohol will be investigated as although findings are inconsistent, studies have shown that substance use can be correlated with high risk behaviours such as unprotected sex. (Myers et al, 2004)
The interviews will be written and tested to reduce bias related to question framing, anchoring and recency/recallability. Self-completion or electronic questionnaires could be used, and have shown to be useful when surveying sensitive topics (Gnambs and Kaspar, 2015, Johnson et al, 2001), but due to the variable literacy level of the target population, this is not considered viable in this case.
As in research by Maher et al (2009), text messaging will be offered as an effective way to keep in touch with participants and remind them of interview sessions and focus group meetings.
Building rapport with the participants is essential to generate deep insights given the sensitive nature of the issues, and to obtain interactive informed consent, particularly when participants may have low levels of literacy (Cortes et al, 2010). Researchers will be encouraged to initiate conversations with participants about the study itself, and about unrelated topics if they arise. Participants will also be encouraged to initiate contact with the researchers themselves: this provides participants with a degree of autonomy and control over the process that may aid communication. To further build rapport and trust, interviews will be carried out in the participants’ native language, by men of the same ethnicity as participants.
The venue will be easily accessible and unrelated to the diagnosis and treatment of HIV/AIDS: the study is focussed on context, understanding and behaviours, and perceived “medicalisation” of the study could be detrimental. Similarly, although participants will be offered voluntary HIV tests, it will be made clear that testing is not compulsory or part of the research, as concerns about compulsory testing could reduce the sample size.
It is important to measure the MSM population size in this region in order to determine representativity of the study sample, and establish effective health interventions. A capture-recapture method as in Mastro et al (1994) would take too long, and other methods require data that is not available for this region. As the size of the general population is known, the network scale-up method as described by Bernard et al (2010) is a suitable approach. Each participant will be asked to estimate how many people they know, and how many of those are MSM. This method is subject to inaccuracies and will tend towards an underestimate as MSM may be hidden from other MSM in the same population, but accuracy is improved by increasing the number of respondents.
Reassuring participants that their data will be safeguarded is paramount in order to minimise loss to follow up. Researchers will demonstrate measures to safeguard data by visibly entering paper records and recordings at the conclusion of an interview into a lock box for transport. Data will be anonymised and participants will be offered the opportunity to provide pseudonyms.
The establishment of a longitudinal cohort as a by-product of this study is tempting, but requiring participants to agree to long term involvement is likely to reduce participation, and as such has been discarded.
Analysis and Interpretation
Questionnaire data will be analysed using SPSS and R. Regression analysis will be conducted to examine the association between variables such as health literacy and condom use. Level of significance will be set at a probability of less than or equal to 0.05. Metrics such as health literacy will be quantified by converting to a scale of 1-5 upon normalisation of recorded responses.
Study data and findings in the form of a comprehensive report will facilitate health interventions intended to reduce the incidence of HIV/AIDS, including education programs, public health communication, social and medical provision alongside influencing local and regional policy.
The insights gained from this study will steer proactive testing programs and improved services for those who are living with HIV/AIDS.
A limitation of this study includes being unable to establish any causal links between associations: only correlation, not causation, can be demonstrated due to a lack of any temporal aspect to the study. Multiple confounding factors may be present; a further longitudinal cohort study may be able to demonstrate causal relationships between the factors examined in this research and eliminate confounding variables. This study is carried out in a specific temporal, spatial and cultural context, and as such, the findings will not be representative of other contexts.
This study will generate rapid, actionable insight into health literacy and sexual behaviours of MSM in this region, while improving awareness and helping men make more informed decisions. Existing resources available for utilisation in further research and investigations will be identified.
Whilst policy makers and agencies may prefer quantitative research that is seen as more repeatable, rigorous and robust for decision making (Tierney and Clemens, 2011), the results of this study can be used to quickly inform potential health interventions and local decision making, along with directing future research.
Word count: 1999
Atkinson, R. and Flint, J. (2001) Social research update. Department of Sociology, University of Surrey, (33).
Becker, H. S. (1963) Outsiders; studies in the sociology of deviance. London, Free Press of Glencoe.
Bernard, H.R., Hallett, T., Iovita, A., Johnsen, E.C., Lyerla, R., McCarty, C., Mahy, M., Salganik, M.J., Saliuk, T., Scutelniciuc, O. and Shelley, G.A., (2010). Counting hard-to-count populations: the network scale-up method for public health. Sexually transmitted infections, 86(Suppl 2), pp.ii11-ii15.
Bonevski, B., Randell, M., Paul, C., Chapman, K., Twyman, L., Bryant, J., Brozek, I. and Hughes, C., (2014). Reaching the hard-to-reach: a systematic review of strategies for improving health and medical research with socially disadvantaged groups. BMC medical research methodology, 14(1), p.42.
Boyce, Paul, Aggleton, Peter and Malcolm, Anne (2004) Rapid assessment and response adaptation guide on HIV and men who have sex with men. Technical Report. World Health Organization, Geneva.
Buvé, A., Lagarde, E., Caraël, M., Rutenberg, N., Ferry, B., Glynn, J.R., Laourou, M., Akam, E., Chege, J., Sukwa, T. and Study Group on Heterogeneity of HIV Epidemics in African Cities, (2001). Interpreting sexual behaviour data: validity issues in the multicentre study on factors determining the differential spread of HIV in four African cities. Aids, 15, pp.S117-S126.
Cortés, D.E., Drainoni, M.L., Henault, L.E. and Paasche-Orlow, M.K., (2010). How to achieve informed consent for research from Spanish-speaking individuals with low literacy: a qualitative report. Journal of Health Communication, 15(S2), pp.172-182.
Delor, F. and Hubert, M., (2000). Revisiting the concept of ‘vulnerability’. Social science & medicine, 50(11), pp.1557-1570.
Gnambs, T. and Kaspar, K., (2015). Disclosure of sensitive behaviors across self-administered survey modes: a meta-analysis. Behavior research methods, 47(4), pp.1237-1259.
Jaffe, H.W., Valdiserri, R.O. and De Cock, K.M., (2007). The reemerging HIV/AIDS epidemic in men who have sex with men. Jama, 298(20), pp.2412-2414.
Johnson, A.M., Copas, A.J., Erens, B., Mandalia, S., Fenton, K., Korovessis, C., Wellings, K. and Field, J. (2001) ‘Effect of computer‐assisted self‐interviews on reporting of sexual HIV risk behaviours in a general population sample: a methodological experiment’, AIDS 15, 111‐115.
Lins, N. E., Jones, C. M., & Nilson, J. R. (2010). Commentary New frontiers for the sustainable prevention and control of non-communicable diseases (NCDs): a view from sub-Saharan Africa. Global Health Promotion, 27–30.
Liverpool, J., Alexander, R., Johnson, M., Ebba, E.K., Francis, S. and Liverpool, C., (2004). Western medicine and traditional healers: partners in the fight against HIV/AIDS. Journal of the National Medical Association, 96(6), p.822.
Maher, J.E., Pranian, K., Drach, L., Rumptz, M., Casciato, C. and Guernsey, J., (2010). Using text messaging to contact difficult-to-reach study participants. American Journal of Public Health, 100(6), pp.969-970.
Marmot, M. & Wilkinson, R. (2005) Social Determinants of Health. OUP Oxford. ISBN 0191578487, 9780191578489
Mastro, T.D., Kitayaporn, D., Weniger, B.G., Vanichseni, S., Laosunthorn, V., Uneklabh, T., Uneklabh, C., Choopanya, K. and Limpakarnjanarat, K., (1994). Estimating the number of HIV-infected injection drug users in Bangkok: a capture–recapture method. American Journal of Public Health, 84(7), pp.1094-1099.
Mercer, C.H., (2010) Measuring sexual behavior and risk. London UK: Health Protection Agency.
Myers, J.P. Aguinaldo, D. Dakers, B. Fischer, S. Bullock, P. Millson & L. Calzavara (2004) How Drug using men who have sex with men account for substance use during Sexual Behaviours: Questioning assumptions of Hiv Prevention and Research, Addiction Research & Theory, 12:3, 213-229, DOI: 10.1080/16066350310001640161
Shilts, R. (1987) And The Band Played On. Politics, People and the AIDS Epidemic. St. Martin’s Press. ISBN 0-312-00994-1
Skidmore, D. and Hayter, E. (2000) Risk and sex: ego-centricity and sexual behaviour in young adults. Health, Risk and Society, 2, 23–32.
Smith, A.D., Tapsoba, P., Peshu, N., Sanders, E.J. and Jaffe, H.W., (2009). Men who have sex with men and HIV/AIDS in sub-Saharan Africa. The Lancet, 374(9687), pp.416-422.
Sydor A (2013) Conducting research into hidden or hard-to-reach populations. Nurse Researcher. 20, 3, 33-37.
Tierney, W.G. and Clemens, R.F., (2011). Qualitative research and public policy: The challenges of relevance and trustworthiness. In Higher education: Handbook of theory and research (pp. 57-83). Springer, Dordrecht.
Vanlandingham, M. and Trujillo, L. (2002) Recent Changes in Heterosexual Attitudes, Norms and Behaviour Among Unmarried Thai Men: A Qualitative Analysis. International Family Planning Perspectives, 28, 6–15.
Wensink, M., Westendorp, R.G. and Baudisch, A., (2014). The causal pie model: an epidemiological method applied to evolutionary biology and ecology. Ecology and Evolution, 4(10), pp.1924-1930.
Williams, J.H., (2005). Cross-national variations in rural mathematics achievement. Journal of research in Rural Education, 20(5), pp.20-5.
Zhang, Y., (2006). Urban-rural literacy gaps in Sub-Saharan Africa: The roles of socioeconomic status and school quality. Comparative Education Review, 50(4), pp.581-602.
Zulu K, Bulawo N, Zulu W. (2006) Understanding HIV risk behaviour among men who have sex with men in Zambia. AIDS 2006–XVI International AIDS Conference; Toronto, ON, Canada; Aug 13–18, 2006: abstr WEPE0719.