Men Who Have Sex With Men in Kisumu, Kenya: Comfort in Accessing Health Services and Willingness to Participate in HIV Prevention Studies

Abstract

Men who have sex with men (MSM) are a crucial and marginalized at risk population for HIV in Africa but are poorly studied. Like other areas of Africa, homosexuality is illegal in Kenya. We assessed MSM comfort in accessing health services and willingness to participate in HIV prevention research in Kisumu, Kenya—an area of high HIV prevalence. We conducted a two-phase formative study with individual interviews (n = 15) and a structured survey (n = 51). Peer contact or snowball method (n = 43, 84.3%) was the primary recruitment strategy used to locate MSM. Exact logistic regression models were used for survey data analysis. Over 60% (32/51) of survey participants were not very comfortable seeking health services from a public hospital. Almost all MSM (49/51; 96.1%) reported willingness to be contacted to participate in future HIV research studies. Efforts to provide facilities that offer safe and confidential health services and health education for MSM is required. Continued community engagement with the MSM population in Kenya is needed to guide best practices for involving them in HIV prevention research.

Recent research has focused on the important role that men who have sex with men (MSM) are playing in the HIV epidemic in Africa (Geibel, Tun, Tapsoba, & Kellerman, 2010) after decades of considering AIDS in Africa a heterosexual disease (Smith, Tapsoba, Peshu, Sanders, & Jaffe, 2009). Homosexuality is illegal in many African nations, and MSM report rapes, evictions, and alienation from friends and family members and even worse (Arreola, Hebert, Makofane, Beck, & Ayala, 2012; Gettleman, 2011). Early studies on MSM in Africa found that rampant homophobia, discrimination, and criminalization of MSM undermined efforts to bring the HIV epidemic under control (Smith, Tapsoba, Peshu, Sanders, & Jaffe, 2009).

Appropriate health care and prevention services are vital to bringing the HIV epidemic under control in Africa (Global HIV Prevention Working Group 2007). Despite South Africa’s constitutional protections and visible gay communities in major cities, an interview study of 32 Black MSM in Soweto and elsewhere in the Gauteng province found all reported witnessing or experiencing homophobia from health care workers (Lane, Mogale, Struthers, McIntyre, & Kegeles, 2008). Some said they use word of mouth to find more welcoming clinics and prefer younger providers because they tend to be less homophobic. Similarly, Rispel and colleagues (Rispel, Metcalf, Cloete, & Reddy, 2009) found MSM reluctant to visit general medical clinics in South Africa and also that a minority of MSM in Malawi, Botswana, and Namibia had revealed their sexuality to a health care worker (Baral et al., 2009; Ehlers, Zuyderduin, & Oosthuizen, 2001).

Perhaps because of this rejection of homosexuality in Africa, research has been delayed compared to in North America and Europe; yet the prevalence of HIV in sub-Saharan Africa is two to four times higher among MSM than in the general population (Baral et al., 2009). A recent report focusing on most at-risk populations in Kenya estimated the HIV prevalence among MSM living in Nairobi to be 18.2% (Kenya Ministry of Health, 2012), which is higher than the overall prevalence in Nairobi of 8.8% (National AIDS and STI Control Programme, 2009). The same report estimated the HIV prevalence among MSM living in Kisumu to be 11.1% (Kenya Ministry of Health, 2012). Kisumu is one of three major cities in Kenya and is situated in Nyanza Province, where HIV prevalence is the highest in the country: 14.9% among persons aged 15–64 years compared with the national prevalence estimate of 7.1% (National AIDS and STI Control Programme, 2009). While there is a growing lesbian, gay, bisexual, transgender, intersex (LGBTI) movement in Kenya that is helping to make Kenya somewhat less dangerous for these minority groups than in other nearby countries (Migiro, 2013), there is a dearth of information about MSM in Kenya, especially in Kisumu Kenya and especially with regard to health services.

Efforts to decrease risk behaviors and increase HIV testing among MSM must include the assessment of structural factors, especially access to health services, as barriers or facilitators. Moreover, with the high HIV prevalence (Sanders et al., 2007) and incidence among MSM in some parts of Kenya (National AIDS Control Council, 2009), their participation in HIV prevention trials will be essential. However, in the context of homosexuality being illegal, recruiting MSM may be difficult. Examining self-reported barriers and motivators to participation in HIV prevention studies is useful in designing studies and recruitment strategies that can best accommodate their concerns and needs. Thus, the purpose of this study was to (1) describe the proportion and characteristics of MSM not comfortable seeking health services at a public hospital and (2) describe MSM motivators and barriers for participating in HIV research studies, including their willingness to be contacted for an HIV study and acceptability of fingerprinting for study identification purposes.

METHODS

The study was conducted in Kisumu, Kenya in Nyanza Province, which has a population of approximately 500,000 residents (Central Bureau of Statistics, 2000) who are predominately of Luo ethnicity (Bailey et al., 2007). We conducted a two-phase, formative study that involved individual interviews and a structured survey. The inclusion criteria were men, aged 18 to 64 years of age, who self-identified as MSM. To confirm that men were truly MSM, we enlisted the help of five self-identified MSM known to the staff over a period of several months from different local MSM support groups and organizations. These five MSM had interacted with the staff at MSM gatherings, activities, and other studies that recruited MSM. We required that the potential MSM participant was acknowledged to be MSM by at least two of these five MSM known to the staff. The reason for the confirmation is because during an earlier cohort study focusing on persons at high risk for HIV, some men posed as MSM in order to be recruited into the study and receive the transport reimbursement. When this was discovered, the men were dropped from the earlier study.

All participants received information about the objectives of the study and were informed that the information they provided would be kept confidential, that they could choose not to participate, and that they would not be identified when the information was reported. Written consent was obtained from each person agreeing to participate in the interviews or survey. Approximately five U.S. dollars were given to each participant at each visit for their transportation expenses. The study was approved by the Kenya Medical Research Institute Scientific Steering Committee and the CDC Institutional Review Board.

Procedure

Phase 1

Individual interviews

A purposive snowball sampling approach (Patton, 2002) was used to identify 15 MSM for individual interviews. Interviewers had extensive experience working with local MSM groups and had undergone specific training for this study. To obtain more information on the best ways to recruit this relatively hidden population, interview participants were asked, “What would be the best way to find MSM to ask them about being in an HIV research study?” For this analysis specifically, we focused on the following questions: “The last time you were tested for HIV, how were you treated by the health care provider?,” “Did you reveal to the health care provider that you have sex with men?,” and “How did learning your HIV test results influence your future sexual behavior?” Participants were not asked to disclose their HIV test results but could do so voluntarily. Other questions were asked, though not analyzed here, about topics such as MSM support groups. Interviews lasted approximately 60 minutes and were audio-taped. Interviewers completed a form for each interview that documented the strategies reported for contacting individuals at high risk and supplemented the forms with their own notes. The reported strategies were compared and discussed by the interviewers at regular meetings during the data collection period. Interviewers scored each strategy based on perceived potential success in locating MSM then ranked them.

Phase 2

Survey

The three highest ranked strategies to contact MSM were selected based on the individual interviews and implemented to recruit 51 men (none of the 51 also participated in the interviews) for the survey portion of the study. The number of men recruited was not based on a power calculation. In the survey phase of the study, the three strategies used to locate MSM were peer contact or snowball method (n = 43, 84.3%), use of link persons or people who knew individual MSM—for example, bar attendants and hotel proprietors (n = 5, 9.8%), and direct contact where the study staff directly contacted those already known to them (n = 3, 5.9%). Staff attempted to use all three strategies; however, it was quickly determined that in Kisumu, Kenya, where homosexual behavior is illegal, MSM were most comfortable being recruited through other MSM they knew.

Participants were asked about their demographic characteristics and about health services and sexual identity issues. For the questions regarding acceptability of fingerprinting to identify persons in research studies, participants were first educated about fingerprinting technology, specifically that the images of the fingerprint would be stored in the form of numbers and could not be transformed back to images. The survey interview lasted approximately 30 minutes and was conducted by the staff in English, Dholuo, or Swahili using a handheld computer.

Measures

The dependent variable was “How comfortable are you seeking health services at a public hospital?” (very, somewhat, not at all, do not seek services). For the purpose of this analysis, any response other than very was coded as indicating some discomfort with seeking services at a public hospital. Independent variables included age, education level, income in the past 30 days, ethnic group, and religion as well as the perception people were staring at them when they went to the health clinic (“I feel like people are staring at me when I go to the health clinic”), receipt of useful information from public hospitals, and questions about self-identity (sexual identity: homosexual/bisexual/heterosexual; transgender status: yes/no; and gender identity: man/woman).

Analysis

Qualitative

Interviews were transcribed verbatim and translated from Dholuo or Swahili to English. A codebook was developed by two of the coauthors (DO and DG) based on a priori categories, specifically the interview questions. Coding was carried out by one coauthor (DG) using Nvivo version 8, a QSR international qualitative data analysis program and analyzed using grounded theory. Quotes were selected to represent typical responses by respondents regarding their perceptions of health services, whether they revealed to the provider that they have sex with men and how learning their HIV test results influenced their future sexual behavior.

Quantitative

Initial comparisons were made using contingency table analysis and chi-square tests of significance. To address the issue of sparse data, variable categories were collapsed where appropriate. Specifically, age was collapsed into 18–23 years and 24+ years; education was collapsed into less than college and college/university; income was collapsed into 5,000 Kenya shillings (KES) or less, 5,001–10,000, and more than 10,000; and religion was collapsed into Catholic and Protestant/Muslim/Other. Associations between each outcome and the factors of interest were measured using exact logistic regression models.

RESULTS

Demographic characteristics of the individual interviews (n = 15) and survey respondents (n = 51) are presented in Table 1. Although our age eligibility was from 18–64 years, the oldest participant was 34 years of age.

TABLE 1.

Demographic characteristics of individual interview and survey participants, MSM in Kisumu, Kenya

	Interview n(%) n = 15	Survey n(%) n = 51
Gender
Male	15 (100)	51 (100)
Age (years)
18–20	3 (20)	20 (39)
21–23	9 (60)	13 (25)
24–26	3 (20)	10 (20)
27–34	0 (0)	8 (16)
Tribe/Ethnic Group
Luo	15 (100)	44 (86)
Other	0 (0)	7 (14)
Income KES (past 30 days)‡
<1,000	3 (20)	10 (20)
1,001–5,000	3 (20)	18 (35)
5,001–10,000	6 (40)	13 (25)
>10,000	2 (13)	10 (20)
Marital Status
Single	15 (100)	49 (96)
Not married but living as married	0 (0)	1 (2)
Separated/divorced	0 (0)	1 (2)
Religion
Catholic	4 (27)	16 (31)
Protestant	9 (60)	25 (49)
Muslim	1 (7)	10 (20)
Other	1 (7)	0 (0)
Education of respondent
Primary/vocational	0 (0)	5 (10)
Secondary	4 (27)	33 (65)
College/university	11 (73)	13 (25)
Sexual Identity
Homosexual	8 (53)	21 (41)
Bisexual	7 (47)	20 (39)
Heterosexual†	0 (0)	9 (18)

Note. One interviewee declined to provide information on income and one survey respondent refused to answer the sexual identity question. DK = don’t know; OR = odds ratio; CI = confidence interval.

^‡The currency conversion at the time of the study was 81 KES = 1 USD.

Percentages are rounded to the nearest whole number, thus the total for each category may not add up to 100.

Health Services at a Public Hospital

Survey

Of the 51 survey respondents, 32/51 (62.7%) had some discomfort seeking health services from a public hospital (very comfortable: 37.3%; somewhat comfortable: 39.2%; not at all comfortable: 17.7%; do not seek services: 5.9%).

Factor associated with reporting some discomfort seeking health services

Only one factor was associated with reporting some discomfort in seeking health services at a public hospital. Odds of having some discomfort seeking health services were greater for men who felt like people were staring at them at the health clinic (OR = 9.23; CI = 1.73–95.45). No demographic variables were significantly associated with reporting some discomfort seeking health services (Table 2).

TABLE 2.

Factors associated with discomfort seeking health services from a public hospital among MSM in Kisumu, Kenya, N = 51

	Not very comfortable n(%)	Very comfortable n(%)	OR (95% CI)
Demographic Factors
Age (years)
18–23	20 (62.50)	13 (68.42)	Ref
24+	12 (37.50)	6 (31.58)	1.29 (0.34–5.32)
Education
Less than college	22 (68.75)	16 (84.21)	Ref
College/university	10 (31.25)	3 (15.79)	2.38 (0.50–15.65)
Income KES (past 30 days)‡
5,000 or less	16 (50.00)	12 (63.16)	Ref
5,001–10,000	8 (25.00)	5 (26.32)	1.20 (0.26–5.91)
more than 10,000	8 (25.00)	2 (10.53)	2.92 (0.46–33.16)
Ethnic group
Luo	28 (87.50)	16 (84.21)	Ref
Other	4 (12.50)	3 (15.79)	0.77 (0.11–5.90)
Religion
Catholic	8 (25.00)	8 (42.11)	Ref
Protestant/Muslim/other	24 (75.00)	11 (57.89)	2.15 (0.55–8.65)
I feel like people are staring at me when I go to the health clinic
Agree	17 (53.13)	2 (10.53)	9.23 (1.73–95.45)*
Disagree/DK	15 (46.88)	17 (89.47)	Ref
Receipt of useful health information from a public hospital
Yes	16 (55.17)	7 (38.89)	Ref
No	13 (44.83)	11 (55.17)	0.52 (0.13–1.99)
Sexual Identity†
Heterosexual	7 (21.88)	2 (11.11)	Ref
Bisexual	12 (37.50)	8 (44.44)	0.44 (0.04–3.22)
Homosexual	13 (40.63)	8 (44.44)	0.48 (0.04–3.44)
Transgender status
Yes	3 (9.38)	4 (22.22)	Ref
No	29 (90.63)	14 (77.78)	2.7 (0.40–21.04)
Gender identity
Man	29 (90.63)	17 (89.47)	1.13 (0.09–10.97)
Woman	3 (9.38)	2 (10.53)	Ref

Note. One respondent refused to answer the questions on transgender and sexual identity and four respondents chose “not applicable” for the question regarding receipt of useful health information from a public hospital.

^‡The currency conversion at the time of the study was 81 KES = 1 USD.

^†All men reported having sex with other men, though some still self-identified as heterosexual.

^*significant; p < 0.05.

Individual Interviews

The major themes for the interaction with a health care provider question were (1) comfort with MSM-friendly providers, (2) stigma, and (3) disclosing sexual orientation. Thirteen of 15 MSM interviewed reported they received good treatment from a health care provider the last time they were HIV tested, and two said they did not. However, 5 of the 13 were tested at a facility where the provider was “MSM-friendly” and 7 of the 13 said they did not reveal their sexual orientation to the provider.

“The healthcare provider responded positively because he knew about MSM, that is why I went to the one who knew that.” (23 years of age, college education)

“The people who tested me knew my sexual orientation so it just went well.” (24 years of age, college education)

“The last time … I went with my parents, we went and I was very scared … At first it was difficult because going to see someone who is not MSM friendly. To him, it was like I was not OK and that I was mad. (21 years of age, college education)

“… You cannot reveal something unless the healthcare provider is somebody who is confidential and can keep information.” (20 years of age, college education)

“… I saw it (my sexual orientation) as a bit confidential … Yah to me and he did not ask me.” (22 years of age, college education)

“They have never asked me so there is no need for me to tell them.” (23 years of age, college education)

There was just one theme for the question about how learning your HIV test results influenced your future sexual behavior (more careful about using protection), given that of the 13 individual interview participants who responded to the question, all 13 responded that it made them want to be more careful in protecting themselves.

“I think it is so good the moment you realize you are negative; you get the zest to become more careful …” (21 years of age, university education)

“When I got the result, I decided it was high time that I continued to use a condom.” (21 years of age, college education)

“Getting the results helped me to learn to have one partner and to talk to your partner to be faithful.” (22 years of age, college education)

“Now that I have come to realize that I am HIV negative, I learned that I should use condoms to protect myself.” (22 years of age, college education)

Participation in Research Studies

Survey

Reasons that would motivate and inhibit MSM from participating in an HIV research study

The main motivator cited for participating in an HIV research study for MSM was getting HIV education that could help prevent HIV (35.3%), followed by receiving HIV testing and counseling (17.7%) and helping to find a cure for HIV (17.7%; Table 3). The main barrier was having personal information made available to others in the community, accounting for nearly two thirds of responses (64.7%), followed by no barriers (9.8%) and taking too much time away from your job (5.9%).

TABLE 3.

Self-reported motivators and barriers to taking part in an HIV research study, MSM in Kisumu Kenya, N = 51

Motivators	N (%)	Barriers	N (%)
Getting HIV education could help you prevent HIV	18 (35.3)	Having your personal information made available to others in the community	33 (64.7)
HIV testing and counseling	9 (17.7)	Nothing	5 (9.8)
Helping find a cure for HIV	9 (17.7)	Taking too much time away from your job	3 (5.9)
Getting information on how to take care of yourself if you become HIV infected	8 (15.7)	Losing your job	2 (3.9)
Getting incentives for taking part	4 (7.8)	Moving out of the area	2 (3.9)
Getting free HIV treatment and care	1 (2.0)	Having to deal with new medical costs not covered by the study	2 (3.9)
Being a part of a social group that gets together to talk periodically	1 (2.0)	Fear of testing positive for HIV	1 (2.0)
Having friends enroll	1 (2.0)	Sexual partner/spouse refusal	1 (2.0)
		Taking too much of your blood/taking blood	1 (2.0)
		Refuse to answer	1 (2.0)

Willingness to be contacted for an HIV study and acceptability of fingerprinting for study identification purposes

The vast majority (49/51; 96.1%) of MSM reported they would be willing to be contacted to participate in future HIV research studies; only 2 (3.9%) said they were not sure. When asked about the acceptability of using fingerprints for identifying persons who participate in research studies in Kisumu, 49 out of 51 (96.1%) said yes, it would be acceptable in Kisumu. Furthermore, when the 49 MSM who thought it would be acceptable were asked how they thought participants would like it, 57.1% (28/49) reported they thought participants would like it very much, 38.8% (19/49) thought persons would find it acceptable with few reservations, and 4.1% (2/49) thought participants would disapprove of it.

DISCUSSION

Nearly two thirds of MSM in Kisumu Kenya reported having some discomfort when seeking health services at a public hospital. The only factor in our analysis associated with these men was a report of feeling people were staring at them. In a similar vein, the main barrier to participating in HIV research was having their personal information made available to others in the community. Receiving education that could help prevent HIV infection was the main motivator to taking part in an HIV research study.

That a large proportion of MSM in Kisumu had some discomfort (62.7%) when seeking services at a public hospital agrees with reports of MSM from other areas of Africa. Among MSM in South Africa, all reported witnessing or experiencing homophobia from health care workers (Lane et al., 2008; Ehlers et al., 2001; Baral et al., 2009). Among MSM in Malawi, Namibia, and Botswana, there was an association between fear of health care services and ever having experienced discrimination (Fay et al., 2011). Homophobia and provider stigma were noted as barriers to obtaining condoms, lubricants, HIV testing, and HIV treatment in a global MSM health and rights study (Arreola, Hebert, Makofane, Beck, & Ayala, 2012). In our study, 13 of 15 MSM interviewed reported they received good treatment from a health care provider the last time they were HIV tested, and just two said they did not. However, five of the 13 were tested at a facility where the provider was “MSM-friendly” and seven of the 13 said they did not reveal their sexual orientation to the provider. In a Kenyan study, MSM reported preferring to receive medical care for HIV/STI and counseling in private clinics because of the perception of increased confidentiality. MSM do not seek advice from health care providers because they are afraid that they will be turned in for same-sex behavior, which is illegal, or that the providers will in some way discriminate against them (Onyango-Ouma, Birungi, & Geibel, 2009). In a study of MSM in Nairobi, health care providers did not ask about same-sex behavior and thus did not offer prevention information (Geibel et al., 2010). This contributes to both a lack of appropriate health education and promotion for MSM and lack of quality health care services.

In efforts to understand the African HIV epidemic over the past two decades, the role of same-sex behavior has been, for the most part, left out. The mention of same-sex behavior as a risk factor for HIV was noted in just 14 of 118 studies of risk factors for HIV in men in sub-Saharan Africa between 1984 and 2007 (Smith, Tapsoba, Peshu, Sanders, & Jaffe, 2009). However, there has been a recent increase in attention and study (Kenya Ministry of Health, 2012). In fact, the Kenya National AIDS Strategic Plan notes that MSM “need to be recognized as significantly contributing to HIV incidence” (National AIDS Control Council, 2009). This would suggest the need to better understand risk factors of MSM in Kenya and their inclusion in HIV prevention studies. The vast majority of MSM surveyed for our study reported they would be willing to be contacted to participate in future HIV research studies and that use of fingerprinting for identification purposes would be acceptable in Kisumu, though some thought there would be reservations. Including MSM in innovative HIV prevention research is needed due to the many factors that influence effectiveness of current prevention programs (e.g., seroadaptive strategies; Grossman 2011). Onyango-Ouma and colleagues (Onyango-Ouma et al., 2009) stated that including MSM in research is feasible if correct procedures are followed and issues are addressed such as identifying and involving key stakeholders and negotiating around rules regarding how researchers and participants interact with one another.

One important finding from the qualitative data in our study was that MSM who reported they tested negative for HIV said that the result made them more careful in protecting themselves so that they remained HIV-negative in the future. In the first phase 2 HIV vaccine trial, VAX004, 56% of participants reported joining the trial to reduce their risk behavior (Colfax et al., 2005), and participants showed a reduction in engaging in the highest HIV risk behaviors over the 36-month trial (Bartholow et al., 2005). Risk reduction counseling offered as part of clinical trial participation has been associated with reduced HIV risk behavior and low HIV incidence in Kenya (Kaul et al., 2002). These results suggest that HIV testing and participation in an HIV research study where risk reduction counseling is offered may be effective HIV prevention interventions in and of themselves (Padian, McCoy, Balkus, & Wasserheit, 2010).

Several limitations should be considered in the interpretation of our results. First, there may have been response bias in that some men may have posed as MSM to be part of a study and thus reimbursed for their time and transport, or modified their responses due to social desirability bias. However, we guarded against the first possibility by requiring that two known MSM verify the sexual orientation of the potential participant and by using the contact strategies favored by the local MSM. We guarded against the second possibility by using strategies suggested by MSM so that they had a level of comfort that our goal was to learn more about them while keeping their identity confidential. Second, because specific strategies suggested by the 15 MSM interviewed were used to contact the MSM for survey administration, the survey sample was not probabilistic and the survey respondents cannot be presumed to be representative of the MSM in Kisumu. Third, our sample size was small, so we were limited to bivariate analyses. The sample size may also have been too small to detect differences when they, in fact, existed. Results might be different with a larger sample of MSM. Fourth, we did not have a non-MSM group to compare discomfort seeking health services. Fifth, we did not collect systematic information on the HIV status of participants. Sixth, we included “do not seek services” in the analysis because the number of respondents was small and because not seeking services at a public hospital could be related to avoidance due to anticipated negative treatment. However, it could also be because the respondent had no need to seek services at a public hospital. Seventh, the oldest survey participant was 34 years of age, so we did not have an MSM sample covering the full adult age distribution. Future studies should give special effort to recruiting older MSM. Finally, important factors associated with discomfort in seeking health services at a public hospital may not have been included in our survey. The strength of this study is that it offers one of the first insights into the attitudes related to health services and HIV research study participation among MSM in Kisumu, Kenya.

Our study found that a large percentage of MSM in Kisumu Kenya had some discomfort when seeking health services at a public hospital; a feeling of being stared at was the only factor associated with this group. Similarly, the main barrier to participating in HIV research was having personal information made available to others in the community. Other studies of MSM in Africa have noted the importance of confidentiality (Geibel et al., 2010). Until the discriminatory laws change, as suggested by Lane and colleagues (Lane et al., 2008), community-based organizations can help MSM receive needed health services by identifying and developing “MSM-friendly” networks of health care providers. There was some evidence in our qualitative data that such a network was available to some MSM in Kisumu.

Given that MSM in Kenya are being recognized as a key group in the HIV epidemic in Kenya, efforts to provide for safe and confidential facilities to offer health services and health education and promotion to MSM in all parts of Kenya is required. Furthermore, efforts to begin or continue community engagement of the MSM population in Kenya are needed to help MSM feel comfortable participating in HIV prevention research. Based on recent events in Kenya, such as the National Gay and Lesbian Human Rights Commission (NGLHRC) collaborating with the Kenya Human Rights Commission in honoring lesbians, gays, bisexuals, transsexual, intersex, and queer (LGBTIQ) individuals for their contributions to the Kenyan society (Silva, 2012), Kenya is ready for these initiatives.