Sigma Research
TwitterFacebookVimeo

EMIS-2017: Planned international journal articles

International EMIS data for scientific articles can now be requested by members of the EMIS network including merged 2010 and 2017 data for change analyses. Data request forms are available via e-mail. Please contact coordinator@emis-project.eu.

The following data access requests for international papers have been approved by the EMIS Editorial Board.

 

Title: HIV testing behaviours and associations with anti-gay abuse among men who have sex with men in European Union Enlargement Area countries
Lead authorKristefer Stojanovski
Dataset: EMIS-2017 (AL, BA, MA, ME, RS, TK, XK)
Research question and objectives: To assess to what extent anti-gay abuse might influence HIV testing behaviours in EU Enlargement Area Countries.

  1. Hypothesis: HIV testing will be lowest among MSM experiencing higher abuse.
  2. Hypothesis: The relationship between anti-gay abuse and MSM’s HIV testing behaviours will be mediated through poor mental health.

Analytic plan: This paper will explore how experiencing anti-gay abuse and violence influence testing behaviours in EU Enlargement Area countries. The analyses will explore how the main independent variables of experiences of violence, intimidation, and assault relate to HIV testing behaviours in EU Enlargement Area countries. The analyses will explore how the anti-gay abuse and HIV resting relationship is also mediated by poor mental health. The statistical analyses will use a structural equation model to analyse the data. The outcome variables are HIV testing. The structural equation model (SEM) will be used to assess the mediating roles that intimidation/assault and mental health have on HIV testing outcomes (e.g., more intimidation and assault --> poorer mental health --> less HIV testing).
Target journals: European Journal of Public Health, AIDS & Behaviour, LGBT Health
Date of approval: 11.03.2020

 

Title: The role of policy in structurally inducing MSM’s HIV risk across the European continent
Lead author: Kristefer Stojanovski
Dataset: EMIS-2017 (48 countries)
Research question and objectives: To assess to what extent countries’ sexual and gender minority (SGM) policies, measured by the Rainbow Index, are associated with men’s sexual behavioural HIV risks as mediated through experiences of violence and poor mental health using European empirical data.

  1. Hypothesis 1: HIV risk will be highest among European MSM who live in countries with less SGM policy protections and rights.
  2. Hypothesis 2: The relationship between policies and MSM’s HIV risk behaviours and serostatus are mediated through experiences of violence, depression, and anxiety.

Analytic plan: This paper will explore how European country policies’ influence experiences of homophobic intimidation/assault, and how these experiences are related to changes in depression and anxiety, and how these relate to changes in sexual risk behaviours (e.g., use of condoms, number of sexual partners, sexual role, etc.) and serostatus. Structural-level factors for this aim are European country-level policy protections and rights afforded to sexual and gender minorities (e.g., anti-discrimination legislation, rights to marriage or civil partnership, etc.). I will use the Rainbow Index developed by the International Lesbian and Gay Association-Europe (ILGA-Europe).  The statistical analyses will use a multi-level structural equation model to analyse the data. The level one variable will be the structural context (i.e., Rainbow Index), and the level two independent variables will be the individual-level intimidation/assault, mental health, and demographic variables. The outcome variables are HIV sexual risk behaviours and HIV serostatus, which are also at the level-two variables. The structural equation model (SEM) will be used to assess the mediating roles that intimidation/assault and mental health have on HIV risk outcomes across various level 1 (i.e., Rainbow Index) situations by using a path analysis (e.g., High protection policies (Rainbow Index) --> less intimidation and assault --> less mental health issues --> more condom use --> less HIV infections). The multi-level model will explore how the pathways vary by countries' Rainbow Index with a particular focus of how the pathway changes depending on whether countries have more or fewer policy protections for SGM.
Target journals: International Journal of Epidemiology, Lancet HIV, Journal of the International AIDS Society
Date of approval: 2.12.2019

 

Title: Contextual and behavioural determinants of resurgent syphilis among MSM in Europe: cross-national analysis, patterns and variations
Lead authorJan Semenza (ECDC)
Dataset: EMIS-2010 and EMIS-2017 (28 EU countries)
Research question and objectives: Syphilis is rising markedly in western Europe since the late 1990s, albeit with marked variations across countries and over time. This trend appears to have continued through to the present. Our study will investigate the following research question: What are the contextual and behavioural determinants of the resurgence of syphilis among MSM across Europe in the last several decades?
To address this question, we will specifically assess two sub-hypotheses:
i) Has risky sexual behaviour increased / protective behaviours decreased among MSM in Europe?
ii) Have there been changes to health seeking behaviour with the spread of highly active antiretroviral therapy for HIV/AIDS, particularly HIV pre-exposure prophylaxis (PrEP)?
Analytic plan: To understand the drivers of rising syphilis, we plan to perform multi-level modelling across nations and within them at the individual level to ascertain the prevalence, change and degree of contextual and behavioural determinants and their association with syphilis risks. 
Specifically we will test two main hypotheses:
i) Rising syphilis is driven by a resurgence of risky sexual behaviours
ii) This risky sexual behaviour is in part linked to the expansion of ART for HIV, specifically PrEP.
While the final model will be developed, it will take a generalised multivariate regression model format as follows:
Syphi,j,t = α + β1SocDemi, j,t + β2Contexti, j,t+ β3SexPrefi, j,t+ β4Testi, j,t+ β5STIi, j,t + β6Riski, j,t + ηj + εi,j,t
All models will be evaluated using STATA v15.1.
Target journals: Lancet ID, Eurosurveillance, Journal of Infectious Diseases
Date of approval: 2.12.2019

 

Title: Changes of probability to be diagnosed with bacterial Sexually Transmitted Infections (STI) among MSM from 38 European countries between 2010 and 2017
Lead Author: Ulrich Marcus
Dataset: EMIS-2010 and EMIS-2017 (38 countries)
Research question and objectives: Surveillance systems for sexually transmitted infections in many European countries report increases of syphilis and other bacterial STIs such as gonorrhoea and chlamydia in recent years. MSM appear to be particularly affected by these increases. Possible reasons for increasing STI diagnoses could be: increased testing and increased diagnosis of asymptomatic infections among MSM; increasing partner numbers; changes of protective behaviours from use of barrier protections (condoms) towards pharmaceutical (antiretroviral drug-based) HIV specific protection strategies with less sero-assortative partner selection, and more HIV serosorting with less condom use among men not using pharmaceutical protection. EMIS provides data on the use of HIV specific risk management during the last event with a non-steady partner and differentiates between symptomatic and asymptomatic infections. The contribution of the different factors on the increase of symptomatic and asymptomatic bacterial STI will be analysed based on data collected in EMIS 2010 and 2017.
Analytic plan: Four outcomes will be analysed separately: diagnosis of symptomatic/asymptomatic syphilis in the last 12 months and diagnosis of symptomatic/asymptomatic gonorrhoea or chlamydia (combined) in the last 12 months. We will differentiate between diagnoses with symptoms reported at the time of testing and diagnoses when no symptoms were reported. We suggest that diagnosis rates for symptomatic infections will allow better comparability between countries than crude diagnosis rates due to the highly varying implementation of STI screening. We want to analyse differences between 2010 and 2017. We will look separately for the variables associated with symptomatic and asymptomatic infection rates. Variables for which we will control for asymptomatic infections are: age, recency of STI testing (as surrogate for frequency); number of specimens tested for STI (blood, genital, rectal); HIV testing history and HIV risk management strategy during the last non-steady partner sex event based on the hypothesis that administration of STI screening will be associated with HIV risk management. For symptomatic infections we will control for settlement size, number of (non-steady) sex partners in the last 12 months, HIV testing history and HIV risk management strategies during the last non-steady partner sex event, setting where the last non-steady partner was met (online, gay sex venue, gay social venue), and STI knowledge. For the analysis of factors associated with symptomatic/asymptomatic syphilis or gonorrhoea/chlamydia diagnosis we will use bivariate, multivariate analysis and a multilevel hierarchical multivariate regression model with countries as additional level.
Target journals: Plos Medicine; BMJ Open
Date of approval: 2.12.2019

 

Title: Factors associated with differences in diagnosis rates of symptomatic sexually transmitted bacterial infections in a large sample of men who have sex with men from 50 countries
Lead Author: Ulrich Marcus
Data sets: EMIS-2017 (50 countries)
Research question and objectives: Comparisons between countries regarding the burden of sexually transmitted infections among MSM are severely hampered by different STI testing policies and high variability between countries regarding the implementation of screening for asymptomatic infections among MSM. We intend to look at rates of symptomatic and asymptomatic syphilis, gonorrhoea and chlamydia among MSM in EMIS. We hypothesise that symptomatic infections allow better comparability between countries, because the proportion of symptomatic infections of all infections should be similar between countries, while the proportion of detected asymptomatic infections is associated with the implementation of screening policies. We intend to look as well into the extent to which STI screening has been implemented and the rates of diagnosis of asymptomatic STI among those screened. We want to analyse which behavioural factors are associated with the diagnosis of symptomatic syphilis, gonorrhoea and chlamydia.
Analytic plan: Diagnosis of symptomatic syphilis, gonorrhoea or chlamydia within the last 12 months will be used as outcome. French language of the questionnaire may be needed to be controlled for due to a known translation issue of the French language questionnaire that may have inflated the number of reported diagnoses. HIV-related sexual risk management with the last non-steady partner will be used as primary explanatory variable. Further variables which will be controlled for are: age; number of (non-steady) sexual partners in the last 12 months; recency of STI testing (as surrogate for frequency); number of specimens tested for STI (blood, genital, rectal); HIV testing history; place where the last non-steady partner was met (online, gay sex venue, gay social venue). Analysis of factors associated with diagnosis of symptomatic syphilis, gonorrhoea and chlamydia using bivariate, multivariate analysis and a multilevel hierarchical multivariate regression model with countries as additional level.
Target Journal: STIs; PLOS One; BMC Infectious Diseases
Date of approval: 12.11.2019

 

Title: Changes in sexual mixing patterns of men who have sex with men 2010-2017. Do they affect STI transmission?
Lead AuthorNicola Low
Data set: EMIS-TREND-2010-2017 (12 countries: AT, BE, CH, DK, FR, DE, IE, LU, NL, NO, SE, UK)
Research question and objectives: How have sexual mixing patterns between men with diagnosed HIV and those without diagnosed HIV changed over time? Objective one: Describe patterns of sexual mixing in 2010 and 2017. Objective two: Examine statistical evidence for a change in patterns of sexual mixing between 2010 and 2017. Objective three: Examine associations between changes in sexual mixing patterns and changes in bacterial STI diagnoses between 2010 and 2017.
Analytic plan: Objective one: Describe patterns of sexual mixing in 2010 and 2017: Proportion of serodiscordancy in sex with steady and non-steady partners; the same for condomless intercourse. Contrast last HIV-test results versus Q309 among HIV- negatives (there might be men with a last negative test who assume to be already positive). Objective two: Show trends of sexual mixing over time: compare 2010 and 2017, change per year. Objective three: If there is a trend, how is this related to and could it possibly affect STI diagnoses? Look at diagnoses of bacterial STIs, separate for syphilis, NG, CT, controlling for recency of blood test (for syphilis), and, for NG and CT, recency and comprehensiveness of swabbing. Sensitivity analysis: exclusion of French language, or controlling for French language. We suggest to work on data representing Northern/Western European countries: AT, CH, DE, LU (largely German-speaking); BE, FR (French- and Dutch-speaking), SE, NO, and DK (large-sample Nordic countries), and IE, UK (English-speaking countries), in order to minimise confounding resulting from differences in STI testing policies and/or sexual minority climate.
Target Journal: PLOS Medicine
Date of approval: 11.10.2019

 

Title: Sexual and mental health across gender identity and sex-assigned-at-birth among men-who-have-sex-with-men in Europe: findings from EMIS 2017
Lead Author: Ford Hickson
Data set: EMIS-2017 (48 countries)
Research question and objectives: MSM suffer a disproportionate burden of sexual and mental ill health, as well as drug related harms. However, health risks and precautions are not evenly distributed among MSM and health inequalities in the general population are reproduced among sexual minorities. A primary research objective of EMIS was to identify subgroups of MSM who show disproportionate unmet sexual health promotion need across a range of domains. The population of men-who-have-sex-with-men (MSM) includes people who were assigned-female-at-birth (AFB) and who now identify as men or as trans men. The current analysis compares the impact of sex-assigned-at-birth and current gender identity on a wide range of variables.
Analytic plan: The paper is descriptive. We will count gender identity and sex at birth and describe demographic variations in gender identity and sex at birth. We will then compare measures of morbidity, behaviours, needs and interventions using cross-tabs. Test for significance will include logistic regressions controlling for age, country of residence and other confounders. We will identify the largest differences between groups to help focus health promotion efforts.
Target Journal: International Journal of Transgenderism
Date of approval: 11.10.2019

 

Title: Changes of protective sexual behaviours with non-steady sexual partners among Men having Sex with Men from 11 European countries between 2010 and 2017
Lead Author: Susanne B. Schink
Data set: EMIS-2010 and 2017 (11 countries: CH; DE; ES; FR; IT; NO; PO; RU; SE; TR; UK)
Research question and objectives: There is widespread perception that sexual behaviour of MSM has changed in recent years towards less condom use and more seroadaptive and HIV-specific protective behaviours. We want to explore changes of protective behaviours based on reported data on sex with the last non-steady partner from EMIS 2010 and 2017. A set of questions that have been asked identically allows comparisons between the two samples. To check for relevant differences in the sample composition of the two samples, we will restrict the analysis to several larger country samples for which we intend to analyse sample differences regarding several key demographic variables.
Analytic plan: The main outcome will be condom use with the last non-steady partner. Depending whether there are relevant differences in the proportion of respondents from the 2010 and 2017 surveys that report no anal sex with the last non-steady partner (Q203) as alternative outcome a combined binary variable differentiating between anal sex without condom, and anal sex with condom including no anal sex could be explored in a sensitivity analysis. We want to look for differences between 2010 and 2017. Variables for which we will control for are: HIV testing history; recency of last HIV test (for those not diagnosed with HIV), undetectable viral load (for those diagnosed with HIV), HIV serostatus disclosure to the last non-steady partner, knowledge of HIV-status of the last non-steady partner, whether ejaculation into the rectum of the partner took place in case no condoms were used, setting where the last non-steady partner was met (online, gay sex venue, gay social venue), whether the respondent had sex with his partner before, use of substances during last sex with a non-steady partner, general HIV-related and STI-related knowledge. We also want to look at the following demographic variables to explore differences between the 2010 and 2017 samples: age, education level (caution: questions were different in 2010 and 2017), settlement size, partnership status, migration status, outness. Changes of the sample composition will be explored separately by country. We intend to include the following country samples: Russia, Turkey, Poland, Sweden, Norway, Germany, Switzerland, France, Spain, Italy, UK. These countries represent different patterns of sample changes between 2010 and 2017. Analysis of factors associated with condom use using bivariate, multivariate analysis and a multilevel hierarchical multivariate regression model with countries as additional level.
Target Journal: Plos Medicine; BMJ Open
Date of approval: 11.10.2019

 

Title: Structural stigma and mental health among MSM migrants to and within Europe. Results from the European MSM Internet Survey 2017
Lead author: John Pachankis
Data set: EMIS-2017 (48 countries)
Research question and objectives: Virtually no information exists regarding the influence of home-country and current-country structural stigma on MSM migrants’ mental health. We therefore propose to utilize EMIS-2017 data to investigate associations between country-level structural stigma and MSM migrants’ mental health. We propose to use a parallel approach from our EMIS-2010 migrant-related analyses, except this time we propose to model mental health symptoms as the outcome rather than HIV-prevention needs (Pachankis et al., 2017, JAIDS).  Specifically, we plan to examine MSM migrant mental health as a function of home-country structural stigma toward sexual minorities, current-country structural stigma toward sexual minorities and immigrants, length of time since migrating, linguistic minority status, migration motivations, and their interactions to predict symptoms of depression, anxiety, and suicidality.
Analytic plan: Analyses will utilize the sub-sample of respondents who indicate having moved from their country-of-origin. We will perform cross-classified models in which structural stigma in respondents’ country-of-residence and in respondents’ country-of-origin will both serve as between-country (Level 2) predictors. These variables will consist of an objective index of country-level climate relevant to sexual minorities. Number of years living in one’s current country-of-residence, linguistic status, migration motivations, and their interactions will serve as our primary person-level (Level 1) predictors.  In order to test moderation of structural stigma in one’s country-of-origin by structural stigma in one’s current country-of-residence, we will examine the cross-level interaction of these two variables.  Assuming significant associations, we will then probe the interaction to calculate simple slopes of country-of-residence structural stigma at +/- 1.0 standard deviation of our country-of-origin structural stigma index.
Target journal: Lancet Psychiatry
Date of approval: 23.08.2019

 

Title: Structural stigma and mental health among MSM in Europe. Results from the European MSM Internet Survey 2017
Lead author: John Pachankis
Data set: EMIS-2017 (48 countries)
Research question and objectives: Depression, anxiety, and suicidality represent a significant public health burden disproportionately borne by sexual minority populations. Examining associations between country-level structural stigma and these mental health symptoms would potentially establish structural stigma as a key source of adverse mental health outcomes among European sexual minority men, thereby supporting policy changes such as reductions in structural forms of discrimination and increased access to mental health services. Identifying mechanisms of this association such as sexual orientation concealment, homophobic victimization, social support, and internalized homophobia can provide researchers and clinicians with plausible targets of psychosocial interventions to improve sexual minorities’ mental health.
Analytic plan: We will create a best-fitting measurement model, including latent variables. We will control for country-level clustering by employing multilevel modelling. Structural stigma toward sexual minority individuals will serve as the between-country (Level 2) predictor and will consist of an objective index of country-level climate relevant to sexual minorities, including laws, policies, and attitudes aggregated to the country level. Sexual orientation concealment (outness), social support, and homophobic victimization will serve as primary within-country (Level 1) predictors. For those models in which mediators are associated with outcomes and structural stigma, we will test for full and partial mediation using the mediation approach outlined by MacKinnon and colleagues (2002). The size and significance of the potential indirect mediation effects of each mediator will be tested using multilevel multiple mediation analyses (Heck & Thomas, 2015)
Target journal: Lancet Psychiatry
Date of approval: 23.08.2019

 

Title: Knowledge of hepatitis A and B among men-who-have-sex-with-men (MSM) in Europe and opportunities to decrease vulnerabilities. Results from the European MSM Internet Survey (EMIS) 2017
Lead author: Sandra Dudareva (RKI) & Erika Duffel (ECDC)
Data set: EMIS-2017 (48 countries)
Research question and objectives: What is the knowledge around hepatitis A and B among MSM in Europe? Do MSM in Europe know where to get vaccinated against hepatitis A and B? Objectives: Analyse the responses provided around knowledge of hepatitis A and B and accessibility of vaccinations among MSM
Analytic plan: Descriptive analysis of questions Q135-Q139, Q142, Q144, Q140 and building of cascade of knowledge for uptake of hepatitis A and B vaccination, e.g. all vaccinated Proportion that are aware of hepatitis A and B vaccination recommendations. Proportion that is aware where to get vaccinated. Proportion vaccinated. Uni- and Multivariable analysis of factors associated with knowledge on hepatitis A and B.
Target journal: Eurosurveillance
Date of approval: 23.08.2019

 

Title: What factors are associated with the vulnerability towards hepatitis A and B among men-who-have-sex-with-men (MSM) in Europe? An update from the European MSM Internet Survey (EMIS) 2017
Lead author: Sandra Dudareva (RKI) & Erika Duffel (ECDC)
Data set: EMIS-2017 (48 countries)
Research question and objectives: Which factors are associated with the vulnerability of MSM in Europe towards hepatitis A and B and have these factors changed over time? Objectives: Analyse proportion of MSM vulnerable towards hepatitis A and B by possible explanatory factors including country, age, settlement size, educational level, outness, access to vaccination, hepatitis/STI knowledge, homosexual disclosure at last STI test and other relevant information collected through EMIS, as well as by national vaccination recommendations for MSM in their country of residence. Describe changes of hepatitis B vulnerability since 2010 by taking into account changes in explanatory factors.
Analytic plan: Update literature research on national vaccination guidelines to create variable for MSM-specific and universal vaccination programs for analyses. Investigate the legal climate for LGBTI citizens (for example Homophobic Climate Index (Lamontagne et al 2018)) and if applicable incorporate in analysis. Descriptive analysis of all variables on country level. Univariable and multivariable (multilevel model) analysis with outcome ‘hepatitis A and B vaccination history’. Multivariate model with 4 possible outcomes (“not vaccinated”, “vaccinated only towards hepatitis A”, “vaccinated only towards hepatitis B” and “vaccinated towards hepatitis A and B” will be investigated. Discuss differences in results from EMIS 2010 and EMIS 2017 (Analysis of changes over time in outcome and explanatory variables).
Target journal: Eurosurveillance
Date of approval: 23.08.2019

 

Title: Prevalence of PrEP use and modes of access in 42 European countries, Lebanon, Israel and Canada
Lead author: Peter Weatherburn
Data set: EMIS-2017 (50 countries)
Research question and objectives: Comparable country-level data on how men-who-have-sex with-men (MSM) access HIV pre-exposure prophylaxis (PrEP), and the extent to which they use it is scarce. For Europe and Israel, the European Centre for Disease Prevention and Control (ECDC) has published a map on PrEP implementation based on information received from national representatives. We aim at estimating the proportion of MSM who ever used PrEP, as well as the predominant mode of access.
Analytic plan: After excluding respondents with largely inconsistent answers (e.g. age), or respondents with diagnosed HIV, we establish the proportion of MSM reporting ever having used PrEP. All these men were asked how they accessed PrEP. We set a cut-off of 1% to determine if PrEP was implemented or not (as some men might have used PrEP while living abroad). We then look at the predominant mode of PrEP access, excluding countries where <10 men had used PrEP. Finally, in multivariable regression analysis, we determine which access modes were most strongly affected by age and financial coping.
Abstract presented at: AIDS Impact (oral)
Target journal: Lancet HIV or AIDS
Date of approval: 30.7.2019

 

Title: Men-who-have-sex-with-men HIV PrEP cascades in 43 countries (including Canada & the Philippines)
Lead authorAxel J. Schmidt
Data set: EMIS-2017 (50 countries)
Research question and objectives: Men-who-have-sex-with-men (MSM) are the subpopulation at greatest risk of HIV in Europe and Canada. HIV pre-exposure prophylaxis (PrEP) has great potential to interrupt transmission. Prevention cascades have been proposed to focus primary HIV prevention programmes on the most commonly unmet needs. We will construct a PrEP cascade for MSM and compare it across 43 countries with more than 100 respondents.
Analytic plan: To construct a five-step PrEP cascade: awareness (heard of PrEP); knowledge (key facts); intention to use (very likely to use PrEP if affordable/accessible); accessibility (tried to get PrEP); current use (daily/on demand). We applied the cascade to multiple countries with different denominator populations (100%): (A) all non-HIV-diagnosed MSM; (B) reporting two or more non-steady condomless anal intercourse partners, previous 12 months (objective risk); and (C) disagreeing with ‘The sex I have is always as safe as I want it to be’ (subjective risk). The article will be rather descriptive. 
Abstract presented at: IAS 2020 virtual (poster)
Target journal: JIAS
Date of approval: 30.7.2019

 

Title: An update on the comparison of the performance of STI screening services for gay and bisexual men across 40 European cities: Results from the 2017 European MSM Internet Survey (EMIS-2017)
Lead authorAxel J. Schmidt (Jason Doran)
Data set: EMIS-2017
Research question and objectives: Rectal manifestations of sexually transmitted infections (STIs) compromise the health of gay and bisexual men. In 2010 across 40 cities among men screened for STIs, anal swabbing was highest in London and Amsterdam with rates of more than 72%, but low across most other European cities. We apply the same methodology for comparing diagnostic procedures for MSM across the same European cities for 2017, in order to see if the gap between London/Amsterdam and other European cities has narrowed.
Analytic plan: See https://sti.bmj.com/content/89/7/575.info. As sexual healthcare for MSM in most countries is organised locally, we choose cities for comparison and focus on a sub-sample of 38,439 men living the same 40 European cities. We use the same multivariable regression model as in 2012 to compare the odds of having received anal swabbing in the 12 months, controlling for age, HIV diagnosis, and the number of sexual partners. We plan to have an online appendix where all 105 EMIS-2017 cities are included, sorted by country.
Abstract presented at: STI and HIV Vancouver (oral)
Target journal: STI
Date of approval: 16.6.2019