Study setting
As part of a larger study to evaluate the impact of a community-based implementation of HIVST delivery strategies among MSM in Kenya (for details see [31]), we undertook a socio-sexual network study to characterize and identify patterns of connection between different types of service using (or avoiding) MSM in each of three study sites (Kisumu, Mombasa, Kiambu counties). The study was implemented by the University of Manitoba and Partners for Health and Development in Africa (PHDA), in partnership with the National AIDS and STI Control Programme (NASCOP), as well as community-based partners G10 (an MSM research network in Kenya), and three community-based organizations (CBOs) in Kenya: Mamboleo Peer Empowerment Group (MPEG in Kiambu), Men Against AIDS Youth Group (MAAYGO in Kisumu), and the HIV & AIDS People’s Alliance of Kenya (HAPA Kenya in Mombasa).
In terms of population composition, Kiambu has a slightly higher population (1.6 million), compared to Kisumu and Mombasa (1 million each). HIV prevalence in the general population in Kisumu is 16%, and 4% in both Mombasa and Kiambu [32], compared to the self-reported HIV prevalence among MSM of 13%, 19% and 23% in the same three sites [33]. According to size estimation studies of physical hotspot spaces where MSM meet sexual partners, there were an estimated 2,492 MSM in Kisumu, 2,855 MSM in Mombasa, and 1,664 MSM in Kiambu [34]. However, these counts underestimate the population size, as internet-based mapping has revealed that 25% of MSM seeking sexual partners do not visit physical hotspots [35]. According to Kenya key populations program data, 53% of the estimated population of MSM living with HIV were known and registered in key population programs at the end of 2018 [36].
Data Collection
Following a community-based research methods approach [37], training of community researchers (CRs) took place over two days in March 2019. Four CRs were recruited from each study site. The training included sessions on HIVST, research ethics [38], and the process for administering the socio-sexual network survey. During the training, the CRs reviewed and finalized the data collection tools.
Data collection occurred between March and April 2019, for approximately two-and-a-half weeks per site. The CRs in each site selected eight seed respondents (N=24) to complete a demographic form and short network surveys for 15 of their sexual and 15 of their social network members (Appendix Table 2). Sexual network members were defined as partners (either male or female) with whom the seed had sex in the past 12 months. Social network members were defined as MSM contacts with whom respondents had communicated with in the past 30 days. Each CR recruited two seeds, including someone who had accessed services from the programs and someone who had never accessed services from MSM programs. Following the survey, the seeds then selected three individuals from their sexual network contacts to participate in the study: 1 young (18-29 years) service user; 1 older (30 years and older) service user; and 1 “unreached” MSM who has never accessed services. The three new respondents each similarly identified three respondents from their network; thus, in addition to the initial seeds, there were 2 waves of recruitment. All participants were 18 years and above, identified as MSM, and had anal or oral sex with another male in the previous 12 months.
The survey included questions on age and social-economic status; the location where a respondent met and/or had sex; whether the person was “out” or “closeted” in their community (i.e., disclosed to family, friends, CBO-based health care providers, married to a woman); means of connecting with other MSM (i.e., social media, cruising/hotspots, CBO events); and whether they were enrolled in an MSM program. For the purpose of the study, MSM program enrollment was self-reported and included enrollment in any MSM program.
Analysis
The network visualizations were depicted to understand how network members share characteristics and to explore whether network approaches might be a useful strategy to reach those individuals not connected to programs and services more effectively. The basic idea behind the visualizations were to understand whether the network members share common characteristics with the seeds and whether seeds can play a role in implementing the HIVST program in the country. The analysis focused on presenting the profile of respondents, visualization and characterization of networks, and uptake of services across three different sites. As HIVST was a relatively new intervention at the time of the study, we further analyzed data related to health services access and HIV testing. For analysis, RDSAT 7.1(Cornell University Ithaca, NY), and Stata 15.0 (Stata Corp, Texas USA) were used. Network diagrams were created using NetDraw 2.1 (NetDraw Software for Network Visualization, Lexington, NY) to understand the size of the MSM network, identify patterns of connection between different sub-populations of MSM, and understand how these connections pivot on characteristics such as age, gender, sexual identity, disclosure, program enrollment, and meeting places for sex. In the bivariate analysis, t-tests were used to test for differences in outcomes across sites.
Ethics
Ethics approval was obtained from the institutional review boards of the Kenyatta National Hospital – University of Nairobi, Kenya (P557/08/2018) and the University of Manitoba – Health Research Ethics Board, Canada (HS22205).