Evolution of HIV prevalence and behavioral factors among MSM in Togo between 2011 and 2015

Introduction The aim of this study was to assess sexual behavior and measure HIV prevalence among MSM in 2015, in Togo. Methods We conducted a cross-sectional study from February to March 2015 in nine major cities of Togo. The respondent-driven sampling method was used to recruit MSM. Behavioral data were collected by interviewer-administered questionnaires. The blood tests were then carried out among MSM to assess their HIV status. Data were inputted into an Epidata database and exported to STATA® 9.0 for analysis. Qualitative variables were compared using the chi-2. Results A total of 496 MSM were involved in this study, with 43.35% in the capital, Lome. Over the past 12 months, 88.9% of MSM had had sexual intercourse with men and 24.1% of them had had sex with women. The last sexual intercourse was with a casual partner among 52.9% of MSM. During the last 30 days preceding the survey, 68.5% of MSM had regularly used a condom during active anal intercourse and 71.9% had used it during passive anal intercourse. The national prevalence rate of HIV among MSM was 13.0%. The factors associated with HIV infection were age of MSM OR = 5.30 [1.85-15.1], HIV testing history OR = 2.63 [1.18-5.87] and the city of residence of MSM OR = 5.56 [2.90-10.64]. Conclusion This study confirms that HIV prevalence among MSM is five times higher than in the general population (13% vs 2.5%). Thus, the need to rethink HIV sensitization and prevention strategies targeting hidden and stigmatized populations such as MSM.


Introduction
Globally, HIV infection rates among men having sexual intercourse with other men (MSM) are far higher than among the general population [1]. Sero-epidemiological surveys conducted in Western European countries reveal an increase of more than 50% of new HIV positive cases among MSM between 2001 and 2004 [2]. In sub-Saharan Africa, reliable data are scarce and difficult to obtain because of stigma and socio-cultural taboos linked to homosexuality [3]. However, some studies conducted in Africa in this regard reveal an HIV prevalence rate 3 to 4 times higher than that noticed among the general population [4]. Thus, in South Africa, in 2008, HIV prevalence was estimated at 49.5% among 204 MSM screened in Johannesburg and 27.5% among 81 MSM screened in Durban [5].
In Senegal, two surveys were conducted among MSM and revealed an HIV prevalence rate of 17.5% in 2004 and 13% in 2007, while it was only 1% in the general population at the same period [6,7].
Finally, in Nigeria, a national survey of 879 MSM revealed a 1.1% prevalence rate in Cross River State, 9.3% in Kano and 17.4% in Lagos [8]. Like Togo, many countries do not yet have a national surveillance system for most-at-risk populations. The only surveillance data available are sentinel sero-surveillance results obtained from pregnant women aged 15 to 49 years during antenatal consultations. Latest estimates revealed a prevalence rate of 2.5% in 2013 [9], indicating a significant decrease both by age groups as well as by geographic zones. However, the spread of the epidemic among highly exposed populations such as MSM, sex workers (SWS) and injectable drug users is still unknown. In 2011, a prevalence study conducted in six major localities in Togo among 758 MSM recruited through the snowball method revealed an average HIV prevalence rate of 19.6% [10]. The aim of this study, four years after the previous one, is to assess sexual behavior and measure HIV prevalence among MSM in Togo, in 2015, and to assess the impact of interventions implemented among this population since 4 years in the country.

Methods
Type and period of study: Tt was a transversal study with a descriptive and analytic orientation, conducted from February to March 2015, in Togo, among a sample population recruited using the Respondent Driven Sampling (RDS) method. The   survey population basically comprised of males who openly stated   having had sexual intercourse with other men, aged 18 and above, having lived in Togo for more than 3 months, regardless of nationality. Sexual ambiguities or transgender persons were excluded from this study. people with sexual ambiguities that are difficult to identify as male or female sex and trans-sexual people with disorders of sexual identity and who identify themselves as the opposite sex to their bodies were excluded.

Choice of sites:
The survey was conducted in nine major cities of Togo. Major cities are those with at least 50,000 inhabitants and all cities that share borders with one of Togo's neighbors (Benin, Ghana and Burkina-Faso), regardless of their population. There are nine major cities in Togo. The following cities were chosen for this survey, from north to south: Sinkance, Dapaong, Kara, Sokode, Atakpame, Kpalime, Aneho, Tsevie and Lome. These cities were identified during a mapping study as having meeting hotspots for MSM, with an estimate of the national population of MSM and a per city estimate of MSM [11].

Recruitment of MSM:
The RDS method was used to recruit MSM.
It is a sampling method based on referral by MSM peers, known for its efficiency in HIV behavioral and biological surveillance among most-at-risk groups that are difficult to reach using classical methods. To start with, 25 resource persons from the MSM community were chosen as first interviewees ("seeds"). The choice of these "seeds" was motivated by their representation of the MSM social sub-groups in the study cities that is: insertive (tops) and receptive (bottoms), versatile, bisexual and gay. After their first interviews, these "seeds" were assigned to each recruit three other MSM from their social network using a coupon with a unique code.
Thus, those recruited by the "seeds" were the first wave of participants and who, in turn, recruited a maximum of three (03) interviewees. The procedure continued until the ideal sample population required for the study was attained. Ethical and statutory aspects: The ensuing protocol and amendments were submitted by bioethics committee of the Health Minister of Togo. Their approval has obtein before the implementation of the study. A written and verbal informed consent was obtained prior to the survey that started with an interpersonal interview with each respondent. Respondents were duly informed on how the study will be conducted by NGOs working in partnership MSM. Participants were informed that blood swab testing for HIV will be conducted. Study staff ensured that the anonymity of participants in this sero-epidemiological survey was respected. All surveillance data was kept in a safe place. Each participant was given an identification number. In order to safeguard anonymity, all the blood samples, data collection tools and administered questionnaires were identified just with a code and an identifier of the center. This made it possible to match the interview questionnaire with the results of the laboratory tests. The specimens were destroyed immediately after the study. There was no register bearing names or other personal details. All databases were protected with strong passwords.
Data entry and analysis: Data was inputted into an Epidata database and exported to STATA® 9.0 for analysis. The results were presented in proportions with a confidence interval of 95%.
Qualitative variables were compared using the chi-2 or Fisher's exact tests and the mean or median was compared using Student's t-test or variance analysis or the Kruskall Wallis nonparametric or Wilcoxon tests. Univariate and or multivariate logistic regression analyses were also conducted in order to study the relationship between the dependent variable (HIV infection) and the explanatory variables (age, education level, and risky behavior) so as to study the determinants of HIV infection. In this model, the variable, center or region, was systematically incorporated as an imposed variable. Ethics approval and consent to participate: The study protocol was submitted and had the approval by bioethics committee of the Health Minister of Togo. A written and verbal informed consent was obtained of respondent prior to the survey that started.

Results
Respondents were duly informed on how the study will be conducted by NGOs working in partnership MSM. Participants were informed that blood swab testing for HIV will be conducted.

Discussion
Our survey allowed us to assess the behaviors of MSM and measure the prevalence of HIV among them in Togo. The results of this survey show a decrease of HIV prevalence after 4 years of intervention among MSM ( Table 3). The average age of MSM involved in our study was 23 years and the majority of them, 97.4%, were Togolese; this is close to the average age of 24 of MSM, with a proportion of 90.3% among Togolese in 2011 in Togo [10]. In Cameroun, the average age of MSM was also 24 [12] while in Tanzania, the average age of MSM was 23 [13]. The MSM population in Togo is, therefore, a young adult population. Only This shows that MSM in Togo are more aware of their HIV status. In Malawi, 44.3% of the MSM had never done an HIV test [14].
The HIV prevalence rate among MSM involved in our study was 13%. In 2013 in Tanzania, the HIV prevalence rate among MSM in two major cities was 30.2% while in [13] Malawi, the prevalence rate among MSM was 15.4% in 2013 [14]. In Togo, the previous national HIV prevalence study among MSM that was conducted in 2011 revealed a prevalence rate of 19.6% [10]. This study, Advanced age, HIV screening history and the city of residence were connected to HIV prevalence among MSM in our study. HIV screening history and residence in the city of Lome were also linked to HIV prevalence among MSM in 2011 in Togo [10]. In Tanzania, a high HIV prevalence rate among MSM was equally noticed among MSM in the biggest city of the country, Dar es Salaam [13]. In Malawi, the fact of being more than 25 years old was associated with HIV prevalence among MSM [14] and in Kenya in 2007, advanced age was also associated with HIV prevalence among MSM [16]. More efficient means need to be sought to decrease the HIV prevalence rate that is currently high among MSM in Togo.
Measures to reduce their social vulnerability, fight stigma and discrimination also need to be designed in order to allow them live their sexuality without risks.

Conclusion
This survey confirms the high HIV prevalence rate among MSM in Togo, with a prevalence rate that is 5 times higher than the national. It also shows a decreasing trend in new infections in this group since (19.6% in 2011 vs 13.1% in 2015) after strengthening the implementation of a prevention and care program. But these results should be consolidated by effectively fighting against stigma and discrimination that limit access to prevention and care services for these HIV-vulnerable populations.
What is known about this topic  HSH are vulnerable populations for HIV;  HIV prevalence is higher in MSM than general population;  MSM are hard-to-reach and marginalized population.  Table 1: Socio-demographic characteristic of MSM