The study used a qualitative and quantitative method to generate data. The study used both methods to generate data for triangulation of findings. The qualitative data helped to corroborate the findings from the quantitative data. It also provided context relevant explanation to some of the findings from the quantitative data. Qualitative data were collected using in-depth interviews (IDI) and key informant interviews (KII) conducted with key opinion leaders in the MSM community, and focus group discussions (FGD) with MSM. Quantitative data were collected through interviewer administered structured questionnaire.
Study participants were MSM resident in urban and rural Nigeria. Study participants had to be 18-years-old or older and self-identified as a MSM. The identified seeds had served as peer educators in structured HIV prevention services programs for MSM. MSM involved in the development of the study protocol were excluded from study participation.
Study participants were recruited from two states in Nigeria - Rivers and Kaduna States. These states had ongoing donor (PEPFAR, Global Fund and World Bank) funded programs that supported the access of key populations to HIV prevention services. The two states were selected to enhance the geographical diversity of study participants thereby increasing the representativeness of the data. Rivers State is located in Southern Nigeria while Kaduna State is located in Northern Nigeria.
Recruitment for quantitative study
A convenient sample of 300 MSM (150 per state was proposed for the quantitative study: 100 (50 per state) from rural area and 200 (100 per state) from urban areas. In River State, study participants were recruited from the riverine Bonny Island (rural area) and Port Harcourt (urban area). In Kaduna State, study participants were recruited from Samineka and Kanfanchan (rural areas), and Kaduna (urban area).
The snowball approach was used for recruitment. Initial study contacts were identified from civil society organizations working with MSM in the target study sites. Five initial contacts in the rural areas and 10 initial contacts in the urban areas were identified as seeds. These contacts were peer educators enlisted on the Heartland Alliance peer education programs implemented in the target states. The seeds invited two MSM each in the first wave of recruitment. The MSM identified in the first wave of recruitment invited two peers each for the second wave of recruitment. Efforts were made to ensure the selected seeds were identified from various hotspots highlighted in the epidemic appraisal in the target states and from different age groups . This helped ensure geographical diversity of participants, and prevented recruitment of respondents from a single cluster. The use of snow balling technique is appropriate for the recruitment of hard to reach populations , and the use of appropriate strategies enhances the diversity of study participants .
Recruitment for qualitative study
Ten FGD were conducted (five per state: two in the rural area, two in the urban area, and one with MSM who had not publicly disclosed their sexual orientation resident in Rivers State). Each FGD included 10 participants. Six KII (three from each state) were conducted with MSM who had accessed the MPPI through an MSM intervention program implemented over the last three years preceding the conduct of the study. Thirty IDI (15 in each state) were conducted with MSM who were not willing to participate in FGD but were identified as key opinion leaders in the community. In total, 136 MSM were recruited for the qualitative study. These participants were not included in the quantitative study.
Recruitments for the FGD and IDI were made through contacts with non-governmental organizations working with MSM in the target States. Participants were randomly (every 5th person) selected from the list of clients at the organization, and invited by the organization to participate in the FGD. Those that declined to participate in FGD were invited for IDI. Recruitment of study participants continued until the number of participants required for the FGD and IDI were reached. Invitees for the KII were purposefully selected. They were MSM who had had accessed the MPPI through an MSM intervention program implemented over the last three years preceding the study. They were also identified to be key opinion leaders in the community.
Field workers were selected based on their competence and experience working with MSM, and trained on the study protocol and use of the data collection tools. The IDI solicited the participants’ perspectives about structural issues that affect the access of MSM to HIV prevention services. This included discussions on challenges with operationalization of the MPPI for MSM, suggestions for improving its operationalization and how to optimize uptake of the MPPI by MSM. For the KII, each participant was interviewed about their experiences and perspectives on enablers and barriers for HIV prevention service access and uptake, and how the services could be improved. FGD were conducted to seek the views of discussants on the importance of provision of HIV prevention services for MSM and how to improve access of MSM to HIV prevention services. All interviews and FGDs were conducted in a private room in the offices of the organizations that provided contacts. Participants who came for the FGD were reimbursed N2000 (approximately $11.50) for transport, and were provided refreshment. All participants were provided with written information about the project.
For the IDI and KII, the field worker greeted the participant, explained the purpose of the interview and obtained consent before proceeding with the data collection and audio-recording the sessions. At the end of the IDI and KII, participants were thanked for their time, and reimbursed for transportation. Immediately following each IDI and KII, the interviewers wrote a detailed debriefing note and filled out a one page debriefing form that listed some basic statistics about the session and a summary report of the interview. The FGD also took the same format except for the presence of a note-taker who wrote the debriefing notes. Participants were encouraged to keep all information shared at the session confidential.
The study questionnaire was divided into eight sections. The first section generated information on participants’ profile (age and level of education, history of use of psychoactive substances and history of sexual intercourse). The second section generated information on the HIV sexual risk behavior – history of sexual abuse, age of sexual debut, multiple sexual partners and use of condom at last sexual intercourse. The third section generated information on willingness to use HIV prevention services. The fourth, fifth and sixth sections generated information on the perceived barriers to access of HIV prevention services by MSM when using public, private and peer-led facilities respectively. The seventh section enquired about willingness of MSM to use services that provided structural interventions. The eight section elicited information on perception of the availability of structural intervention services. The tool used for the study had been used in a prior study conducted to generate similar information from MSM, female sex workers, and people who inject drugs resident in four States in Nigeria .
Sixteen questions were asked on the willingness of respondents to use HIV prevention services. Responses options to each question were ‘very willing’, ‘neutral’ or ‘not willing’. Table 2 highlights each of the 16 questions asked.
Ten questions were asked on perceived barriers to access of HIV prevention services when using public, private and peer-led facilities respectively. Respondents were asked to respond to a ‘yes’ if they perceived listed factors would serve as a barrier to access of HIV prevention services in public, private or peer-led facilities, or a ‘no’ if the factors would not serve as a barrier to accessing HIV prevention services in those institutions. These questions are highlighted in Table 3.
Questions were asked on willingness to use services that provided eight structural interventions and respondents’ perceived availability of these eight services. Response option for willingness to use services ranged from ‘very willing’, ‘willing’, ‘seldom willing’ and ‘not willing’. Response option on perceived availability of services ranged from ‘always available’, ‘available’, ‘seldom available’ and ‘not available’. For analysis purposes, the responses ‘very willing’ and ‘willing’ were collapsed into a single response (willing) and the responses ‘seldom willing’ and ‘not willing’ were collapsed into a single response (not willing). Similarly, the responses ‘always available’ and ‘available’ were collapsed into a single response (available) and the responses ‘seldom available’ and ‘not available’ were collapsed into a single response (not available). These questions are highlighted in Table 4.
The study instruments were developed in English. All the interviews and FGDs were conducted in English. The questionnaires were also administered in English. While it might be useful to translate entire questionnaires into local languages, this would be impractical given the multiplicity of languages in Nigeria. Instead, key words and phrases, especially sensitive ones, were translated in the languages of each selected community and collated as a list generated during the training of interviewers. Interviewers used this material as a reference when in the field. A similar technique was successfully used past studies on sexual and reproductive health in Nigeria [25,26,27]. Participants for the FGD were encouraged to share their opinions in languages they felt comfortable with.
Data analysis for quantitative data
The proportion of respondents who engaged in HIV sexual risk behaviors, those willing to use HIV prevention services and those who identified specific barriers to access of HIV prevention services in public, private and peer-led facilities were analyzed. Also, the proportion of those who were willingness to use different structural intervention services and those who perceived those structural intervention services were available were analyzed. The association between HIV sexual risk behavior and willingness to access structural interventions was also determined. Pearson’s chi-square and Fischer’s Exact test were used to test significance of associations where appropriate. Statistical significance was established at p ≤ 0.05.
Data analysis for qualitative data
Transcripts from audio recorded and summary notes from the IDI, KII and FGDs were analyzed using a grounded approach. Inductive thematic analysis was conducted to identify salient themes. Themes that emerged from the interviews and discussions on perception of need for HIV prevention services, challenges MSM encounter in accessing HIV prevention service, ways to address the challenges identified, and proposals for improving the current service delivery models for the MPPI were highlighted. The findings of the survey and the qualitative data were triangulated.
The study protocol received ethics approval from the Jos University Teaching Hospital Health Research Ethics Committee (JUTH/DCS/ADM/127/XIX/6261). Written consent was also obtained from all study participants.