FAQ IT! : A venue-based HIV prevalence study among Men who have Sex with Men (MSM) in Flanders, Belgium: On methods en ethics Wim Vanden Berghe HIV/STI Epidemiology and Control Unit Department of Microbiology Institute of Tropical Medicine, Antwerp Methods in health survey research Brussels, 17th December 2010
Overview presentation Background Research questions Methods and data collection procedures Ethics Results and conclusions Dissemination of the results
Background : Epidemiology of HIV (Belgians) Source: Belgian InSTItute of Public Health
Background : Research objectives A rise in confirmed HIV diagnoses among MSM since 1999, continues until today - Not only in Belgium but in most Western countries Reasons: a better uptake of HIV testing but also a rise in sexual risk behaviour Funding : Flemish ministry of well-being A need for more targeted HIV prevalence estimates among MSM Evidence-based prevention
Research questions What is the prevalence of HIV among different kinds of MSM venues in Flanders? How many HIV infections are not (yet) detected? What is the frequency of sexual risk behaviour and is there an association with HIV serostatus?
Methods: Time-space sampling Pioneered by the CDC, US Hard to reach groups => no offical listing (of MSM) Urban health research Location-based populations (museums, polling, clubs/bars) Probability-based method => A random sample!!! Subcultural venues or spaces have to exist for the target group otherwise respondent driven sampling
Methods: Time-space sampling Mixed methods = qualitative and quantitative research Mapping and etnography of the venues: Discussion on research and framework with venue owners Description of settings and public => List of eligible settings
Methods: Time-space sampling Multi-level randomization: Time (day of week, time of day) Space/Venue Individuals within venues? => research showed not necessary
Methods: Substrata of venues Three substrata of 7 venues : N1: A selection of cruising venues (n=4) N2: A selection of regular gay (dance)clubs or bars (n=2) N3: A selection of party s, clubs targetting younger MSM (<25 years old)(n=3)
Methods : Sample size Sample size for each substratum based on: HIV prevalence estimates from neighbouring countries Self-report HIV prevalence estimates
Methods: Community-based research Interviewers (n=8) : community volunteers Principles of community-based research: Community partners should be involved at the earliest stages of the project, helping to define research objectives and having input into how the project will be organized. Research processes and outcomes should benefit the community. Community members should be hired and trained whenever possible and appropriate. Community members should be part of the analysis and interpretation of data and should have input into how the results are distributed. This does not imply censorship of data or of publication, but rather the opportunity to make clear the community's views about the interpretation prior to final publication. Productive partnerships between researchers and community members should be encouraged to last beyond the life of the project. This will make it more likely that research findings will be incorporated into ongoing community programs and therefore provide the greatest possible benefit to the community from research.
Methods: Social Marketing Frequently = Asked QueSTIons
Data collection procedures Self-administered behavioural questionnaire in 3 languages Interviewer-assisted fingerprick : blood sample on filterpaper
Data collection procedures: Inclusion/exclusion Inclusion: 18 years old Sexual contact in the last 12 months Exclusion: Mentally/physically fit Already participated
Data collection procedures: Setting flowchart Interviewer A : Information Participate? Inclusion check Self administered : Behavioural survey : 20 questions Interviewer B : Informed consent Biosample support Interviewer B : Feedback on the research Incentive (free drink) 1 or 2 interviewer(s) A : entrance 1 or 2 interviewer(s) B : table near the entrance or other fixed point Shift = 3 hours mainly between 10 pm and 4 am After each shift: storage!
Ethical aspects Informed consent (no name required) Anonymity (number) Non-disclosure of HIV status Storage + analysis samples at AIDS reference laboratory, Institute of Tropical Medicine, Antwerp : tested for HIV and confirmed Ethical clearance by IRB of ITM and EC of UHA
Info/prevention leaflets and incentives Info leaflet on the research project was given to the respondent with info and locations for HIV/STI testing Did you know leaflet => everyone present, passing by Incentive respondent: free drink at the venue Distribution of free condoms/lubricant => everyone present, passing by
Results
General sample characteristics 6 months : september 2009 march 2010 Total respondents : 658 Substrata : - N1 (n=3) = Cruising venues (Antwerp) = 167 - N2 (n=2) = Regular gay venues (Antwerp) = 219 - N3 (n=3) = Younger MSM venues (Antwerp, Ghent)= 263 Participation rates : N1 58%, N2 75%, N3 70% No bloodsample : 8.5%
HIV prevalence
20,0% 15,0% 10,0% 5,0% HIV status : bio sample -Wat is de prevalentie van HIV 14,5% infectie bij MSM in Vlaamse MSM uitgaanssettings? 4,9% 1,8% HIV+ Disc (n=4),0% N1=152 HIV + : 3,2%-31,8% N2=205 N3=221 Steekproef HIV + : 0,9%-7,1%
Hiv status : self report 30,0% 25,0% 26,0% 20,0% 15,0% 10,0% 5,0% 16,1% 16,1% 5,8% 17,5% 3,3% HIV + Niet gekend,0% N1=161 N2=206 N3=246 Steekproef
Percentage undetected HIV infections 14,3% of HIV + men were unaware of their status or thought they were - Five men reported being HIV + but were - Four respondents had discordant results = HIV-/?
Socio demographics
Age (in years) Gem. N1 = 161 38,5 N2 = 215 30,6 N3 = 255 27,5 Totaal 31,4
Nationaliteit 100,0% 80,0% 60,0% 40,0% 20,0% 65,0% 29,4% 77,8% 83,7% 16,7% 13,6% Belgisch Nederlands Frans Andere,0% N1=163 N2=216 N3=258 Steekproef
Woonplaats Antwerpen Oost-Vl West-Vl Limburg Vl-Br Wallonïe Brussel 5,4 6,2 7,2 5,2 10,2 19,6 46,1,0 10,0 20,0 30,0 40,0 50,0
Main conclusions HIV prevalence study First venue-based study in Belgium for estimation of HIV prevalence among MSM Use of time-space sampling and blood sampling through finger prick => feasible and well accepted N1 : 14,5%, N2 : 4,9%, N3 1,8% => range, age specific! 14,3% of HIV positive men was unaware of their status +/- 50% HIV test, +/- 60% STI test in the last 12m Effect mobility within the European context Less condom use among HIV positive men at cruising venues (harm reduction!) High rates of Co-infections HIV/STI and Drug use
Limitations of the study Time-space sampling => results are representative only for MSM present at the settings that were visited => no generalization is possible Reasons why certain venue owners did not want to participate are unknown Reasons for not giving a blood sample are unknown What role do incentives play?
Dissemination of results Output for scientific journals Community reporting: General press :
Dissemination of results Community reporting: Community press www.faqit.be
Dissemination of results Evidence-based prevention:
Next steps Community day planned for march 2011: Interpretation of results by the target group Group discussions with different stakeholders New research questions? Output : direct prevention activities European HIV prevalence project among MSM (SIALON)
Thank you! Wim Vanden Berghe HIV/STI Epidemiology and Control Unit Department of Microbiology Institute of Tropical Medicine Nationalestraat 155 B-2000 Antwerp Belgium wvandenberghe@itg.be 03 247 0712 www.itg.be www.faqit.be