What determines HIV self-test acceptability and uptake within the MSM community in Nairobi and its environs? A cross sectional study

Background: Human Immunodeciency Virus self-test (HIVST) refers to a process where a person collects his or her own specimen (blood or oral), performs a test and interprets the results. The interpretation of results can either be done in private or through support of a trusted partner. Self-test should be seen as screening and conrmatory test should be sort. To determine facilitating factors for HIVST acceptability and uptake among MSM community. Methods: The researcher employed a cross-sectional exploratory study design, targeting men who have sex with men (MSM) in Nairobi. Adult men (aged 18-60 years) that reported to be actively engaging in anal or oral sex with men were eligible for the study. The researcher used purposive sampling to identify the sites where data was collected, snowballing technique was then employed to reach the respondents. Data was collected between July 2018-June 2019. A total of 391 MSM respondent were recruited to ll the self-administered questionnaires, 369 MSM completed the questionnaires (response rate = 94%). Results: There was a signicant association between self-test and frequency of testing for HIV/AIDS at P-Value of 0.011. The respondents who had a positive result from the previous test were 16% and 80% of the respondents would seek conrmatory test, if tested HIV positive through HIVST. There was also a signicant association between respondents who had never tested using self-test kits and the need for a testing “buddy” with a P- Value of 0.014. Fifty nine percent of the respondents would prefer blood sample test kits. High cost of the self-test kits and inadequate knowledge on the use of HIV self-test kits were the main hindrances to HIVST uptake. Conclusions: A signicant number of MSM community in Kenya are willing to use HIVST and are likely to seek for care within 30 days and this is a good indicator of linkage. Blood sample kits are more preferred. Key facilitating factors for HIVST uptake include; awareness creation on HIVST, training on usage of kits, emphasis on testing ‘buddy” for rst time testers and reduction on cost of the kits. the test. A number of the respondents indicated that they would go for counselling after positive results at 169 (46%) with 85 (50.3%) among the respondents having used a self-test kit. Eighty nine percent (89%) 325 of the respondents agreed that they would use condom with their partners despite a negative HIV test result and 155 (47.7%) indicated that they had used a HIV self-test kit. There was a statistically signicant Chi-Square relationship between HIV self-test kit use and the respondent’s condom use with partners even after a negative HIV test result at P. Value =0.040. The respondents who tested negative and who would use condoms with their partner were 1.338 times as likely to test using a HIV self-test kit. Eighty six percent (86%) 315 of the respondents indicated that they would use protection/condom after a positive test with the partner and among them 152 (48.3%) had used a self-test kit. There was a statistically signicant relationship between self-test kit use and the respondent’s protection/ condom use after a positive test with their partners at P. Value =0.019. The respondents who use protection or condoms with their partners were 1.338 times as likely to test use a HIV self-test kit. A majority of the respondents at 169 (52.8%) indicated that they would prefer NGO facility/drop in center’s as distribution point, 70 (21.9%) indicated they would prefer to pick the test kits from the nearest government facility, 36 (11.6%) indicated that their point of preference to pick the kits to be Private hospital/Private clinic, the respondents who indicated that they would prefer to pick the kits from a retail or community pharmacy were 34 (10.9%). About (45%) 142 of the respondents indicated that they would prefer the hotline to be handled by a drop in centre, and 91(29%) indicated that they would prefer the information from the hotline relayed from a public hospital. a statistically signicant Chi-Square association between respondents who had never used a HIV self-test kit and the respondents need of a treatment partner on rst time use of HIV self-test kit at P. Value Thirty three percent (33%) 104 of the respondents indicated that cost would prevent them from using the HIV oral self-test kit with 40 (38%) of the respondents having used a HIV self-test kit. Thirty two percent (32%) indicated that lack of knowledge would prevent them from HIV oral self. Inaccessibility and fear of the results were also some of the other reasons that were mentioned. Twenty nine percent (29%) 89 of the respondents indicated that cost would prevent them from using the HIV blood sample self-test kit with 27 (30%) of the respondents having used a HIV self-test kit. Lack of knowledge of how to use the kits, fear of the results and inaccessibility of the kits were some of the other hinder from using the blood sample test kits. About sixty two percent of the respondents indicated that the estimated current market cost of the self-test kit was expensive with 120 (56.6%) having used the self-test kit. There was a statistically signicant Chi-Square relationship between cost of the test kits and use at a P- Value <0.0001. The respondents who thought that the current cost was expensive were 1.591 times not likely to test using a HIV self-test kit. to the test results what to do on the results and for Pre-test counseling data can also be used for the and for follow-ups. The A study with a

African context [25]. We conducted this research to explore acceptability and factors that would facilitate the uptake of HIVST among the MSM community in Nairobi.

Study Population
The study population was all adult men aged 18-60 years living in Nairobi, who self-reported having engaged in anal or oral sex with men. The age requirement for over 18 years intended to preclude any ethical issues surrounding the interviewing of MSM or MSW minors. Sampling Technique This exploratory study employed a cross-sectional design in peri-urban settings of Nairobi and Kiambu counties in Kenya. We recruited a total of 391 MSM respondents to complete self-administered questionnaires, some of the respondents were men sex workers (MSW). Only 369 MSM completed the questionnaires (response rate = 94%) out of which 172 were MSW. Only self-reported MSM aged above 18 years of age were eligible to participate in this study. Since, the MSM and MSW community is a hidden population and di cult to reach, especially in Kenya where homosexuality is criminalized, we purposively sampled eleven data collection points from where the MSM frequent. The areas include the drop-in centers, bars, hotels and massage parlors.
Snowballing was then employed to reach the respondents.
Data collection: All the MSM respondents completed paper based self-administered structured questionnaires between July 2018-June 2019 after obtaining oral informed consent. These structured questionnaires consisted of three sections. The rst section had questions on demographics of the respondents, the second section consisted of questions on HIV risk behaviors and the nal sections had question on HIVST preferences, acceptability and uptake.

Data Analysis & Management
The questionnaires were serialized and data entered in a SPSS Version 23.0 data-base for analysis (IBM Corp, 2015). Using descriptive analysis, we summarized and presented data in tables. We cross tabulated data on factors associated with HIV self-test uptake among the MSM community. Further analysis of inferential statistics was conducted using Prevalence Odds Ratio (POR) and Prevalence Ratio (PR) tests. This estimated the contribution of each of the risk factors in the outcome of HIV self-test uptake among the MSM community.

Ethical Considerations
We ensured full compliance to the research ethics codes set out in the Helsinki Declaration. Ethical clearance was obtained from the University of Ghent Approval number (PA 2016/009) and the Mount Kenya University Ethics Review Committee (Approval number: MKU/ERC/0463). Informed consent was obtained from all study participants in a language that they could understand and were informed that their participation was voluntary. To ensure con dentiality, all potential identi ers in the data were omitted and each record was anonymized using unique identi ers during data entry and analysis.
Access to the data was restricted to only those researchers responsible for analysis in password protected databases and computers. All investigators received extensive training on research ethics at the beginning of the study.

Results
The results are presented in two sections. The rst table and sections contain the demographics, social-economic and HIV risk for the respondents, the second sections data on the results for HIVST.

Demographics, Socio-economic & HIV Risk characteristics of participants
The below data is available on (Table, 2 87% of the respondent's monthly income range between USD 0 -250. Half of the respondents at (50%) of the respondents identi ed themselves as men sex workers (MSW). Most of the respondents were homosexual followed by bisexual. About sixty eight percent (68%) of the respondents were ever married to a man and (28%) had been ever married to a woman. However, the ever married to men was more of cohabiting and lacked legal documentation, since its illegal in Kenya for a man to marry a man. A majority of the respondents indicated that they had multiple sexual partners at (39.9%), those who indicated that they had two sexual partners in the past six months were (33.6%), the respondents who indicated to have had only one sexual partner in the period were (21.1%) with (5.4%) of the respondents indicating that they had no sexual partner in the last six months. Most of the respondents were versatile, they preferred being either on top or bottom. (11.6%) indicated that their point of preference to pick the kits to be Private hospital/Private clinic, the respondents who indicated that they would prefer to pick the kits from a retail or community pharmacy were 34 (10.9%). About (45%) 142 of the respondents indicated that they would prefer the hotline to be handled by a drop in centre, and 91(29%) indicated that they would prefer the information from the hotline relayed from a public hospital.

HIVST uptake facilitators & barriers
About sixty seven percent (67%) 207 of the respondents indicated that they would prefer a "testing buddy" on the rst time of self-test kit use with 108 (52%) of the respondents having used the self-test kit. There was a statistically signi cant Chi-Square association between respondents who had never used a HIV self-test kit and the respondents need of a treatment partner on rst time use of HIV self-test kit at P. Value =0.014. Thirty three percent (33%) 104 of the respondents indicated that cost would prevent them from using the HIV oral self-test kit with 40 (38%) of the respondents having used a HIV self-test kit. Thirty two percent (32%) indicated that lack of knowledge would prevent them from HIV oral self. Inaccessibility and fear of the results were also some of the other reasons that were mentioned. Twenty nine percent (29%) 89 of the respondents indicated that cost would prevent them from using the HIV blood sample selftest kit with 27 (30%) of the respondents having used a HIV self-test kit. Lack of knowledge of how to use the kits, fear of the results and inaccessibility of the kits were some of the other reasons that would hinder the respondents from using the blood sample test kits. About sixty two percent of the respondents indicated that the estimated current market cost of the self-test kit was expensive with 120 (56.6%) having used the self-test kit. There was a statistically signi cant Chi-Square relationship between cost of the test kits and use at a P-Value <0.0001. The respondents who thought that the current cost was expensive were 1.591 times not likely to test using a HIV self-test kit.

Discussion
This study majored on HIVST among MSM and MSW in Nairobi and the peri-urban areas neighboring. The ndings of this study signi cantly contributes to HIV self-test Preferences, Acceptability & Uptake The study explicitly shows there is a signi cant association between HIV self-test and frequency of testing. MSM who used HIV self-test kits had a higher frequency of testing. Our ndings are in concurrence with MSM randomized to HIVST access Vs. standard clinic-based testing in Seattle, the mean number of HIV test and quarterly testing increased signi cantly among those in the HIVST, with no increase in risk behaviors [27]. The study demonstrated that access to HIV self-testing at no cost increased testing frequency among high-risk MSM in Seattle and had no effect on STI acquisition or sexual risk behaviors. Kenya's National HIV testing guidelines recommends re-testing of HIV negative MSM every three months [28]. To achieve the above target, both clinic-based setting and HIV self-test options should be scaled up within the MSM community. Our study ndings show that most of the MSM preferred blood sample self-test kits as compared to oral self-test kits. These ndings are in line with a study conducted in South Africa among the MSM community that showed higher preference for ngerstick tests over oral uid tests among the participants [21]. Our ndings differ from study ndings conducted in the US emergency department, where most of the respondents prefered oral uid testing [29]. Given the diversity of preference for either blood sample or oral test kits and the ovewhelming support for HIVST, we would highly recommend the availability and distibution of both types of self kits to the MSM community in Kenya.

Preference of Pre / Post HIVST Counselling services
A high proportion of the respondents at (80%) indicated that if they tested for HIV positive through HIV self-test, they would go for counselling and con rmatory test. Our ndings are consistent with ndings from a cross-sectional study done in Kenya, that showed that 74% of the respondents would seek counselling services, con rm results or seek medication after a positive HIVST [30]. Pre-test counselling should be offered before dispensing the HIV self-test. This would provide an opportunity for client to get all the information they would need before testing. The provision of information before testing would be crucial in guiding the clients on how to use the test kits, so as to reduce invalid results also the clients would know what to do depending on the results and where to seek for the healthcare services. Pre-test counseling data can also be used for monitoring the patients and for follow-ups. The respondents would prefer the hotlines or helplines to be handled by a drop-in center and the public hospital. A study done with a cohort of MSM in Nigeria showed less usage of the hotline or helpline. A low number of the participants called through the helpline. This was expected, since most the respondents found the instructions easy to use to understand and use [31]. Other studies have also shown signi cantly low usage of the hotlines [32] [24]. Though the hotlines or helplines showed less usage, they should be retained in helping to reach out and provide more information on the use of HIV self-test kits and link MSM to the health care systems.

HIVST Linkage to Care & Treatment
A major concern for HIVST is whether self-testers will seek care and treatment depending on the results. Linkage to HIV care is seen as accessing a health care provider through a clinic at different stages, the stages include i) enrolment into care and treatment after diagnosis, ii) determination of antiretroviral therapy eligibility, iii) initiation of ART and iv) ART adherence [33]. Being able to link to care within a week is considered optimal behavior given that linkage to care is de ned as "having visited a health care provider within 30 days of being diagnosed with HIV [34]. A signi cantly high proportion of the respondents at (92%) would go for con rmatory test within one month of testing. Our ndings are slightly higher than a study conducted in Kenya in 2014, [30] that reported 61% of the general population sampled and 40 % of MSM would go to a clinic for a con rmatory test. This demonstrates that the MSM who self-test are willing and will still link to the healthcare system. During the implementation of HIV self-test, the following mixed approaches can be considered in linking the MSM to the healthcare services. Home visits by the healthcare workers, MSM visiting MSM friendly clinics, calls and short message service (MSM), since it's a very mobile sub-population. SMS was the least preferred, since it limits someone from expressing their concerns, feelings and it was di cult to establish a relationship.
Most of the respondents felt it was important for them to be able to express their concerns and feelings after receiving a positive result. This would be very di cult using the short message service. Scienti c evidence indicates that being sick is a strong determinant to linkage to care as one seeks medical attention [33]. Hence the MSM who feel generally "healthy" though sero-positive but with no sign and symptoms are likely not to link to care. A study conducted in Nigeria among the MSM community showed a 100% linkage to HIV and treatment [31]. The high linkage in that study was likely due to follow-up calls after HIVST distribution, the ease of participant's access to the opinion leaders and also the linkage to a well-trusted MSM-friendly facility that offers HIV prevention services. Studies conducted in other areas reported both fairly low [35] and high [36] linkage to HIV care and treatment after self-testing.

HIVST Preferred Distribution Points
The NGO facilities/drop in centres were the most preferred distibution points for the HIV self-test kits. Other distibution points include the private sector and the public/government facilities.
Our ndings differ slightly with a study conducted by Okal et al., [30], that's showed that most of the respondents preferred public health care facilities but for the general population. The variance in the ndings can be attributed to the fact that the MSM community feel stigmatized and also due to the illegal nature of homosexuality in Kenya, hence the MSM would want to stay far away from the public/Government healthcare providers. Site preference is also largely based on proximity and cost. A study conducted within the general adult population showed that "easily available" as the strongest reason for which ever pick up point [16]. A research conducted in Nigeria a rms the above preference of the MSM to pick the kits from the MSM friendly drop in centers, a high percentage (42.2%) chose to go to the community health clinic (CHC) for the HIVST kits [22]. Future HIVST distribution should consider drop in centers and community pharmacies as options of distribution sites. Peer educators or Key Opinion Leaders (KOL) are still an option to be considered, though this approach would have high cost implications.

HIVST Uptake Facilitators & Barriers
A signi cantly high number of MSM indicated that they would prefer a testing "buddy" on the rst time of self-test kit use. There was a statistically signi cant association between respondents who had never used a HIV self-test kit and the respondents need of a testing partner on rst time use of HIV self-test kit.
In the year 2013, Kenya conducted the rst HIVST validation studies that reported a higher rate of invalid HIV results (37/ 239 = 15.5%) [16]. The ndings are also consistent with another study done among Chinese MSM, that showed signi cant errors during the process of conducting HIVST that rendered almost half of the test results invalid [37]. According to that study, failure to follow the manufacturers instructions was the main cause of invalid results both for oral self test and blood sample test kits. For the nger prick users, most of the errors occurred during the stage of collecting specimen and for oral uid users made most of the errors during the stage of testing the collected specimen. A testing partner or "buddy'' should be advocated for in helping the MSM rst time testers get accurate results. The above ndings explicitly demonstrate the need for assisted testing for rst time users to reduce the invalid tests and also the need for hotlines which are accessible 24 hours and are trusted and friendly to the MSM community to be able to seek help. Further to this, MSM should be encouraged and awareness created for anyone who gets an invalid result to immediately seek con rmatory test. The high cost of the test kits, lack of knowledge on correct usage of the test kits, fear, stigma and inaccessibility of HIV self-test kits were the main hindrances on the use of the kits. Most of the respondents indicated that the estimated current market cost of the self-test kit was expensive at 8-100 USD. The preferred cost of the HIVST kit was about half to one USD. Lack of appropriate knowledge on the use of HIVST kits would prevent the MSM from using the kits. This correlates to studies showing several ways in which errors would happen, this would include during sample collection using the swab, also the handling of the swab and following the procedures [38], [16] .
This study has some limitations. Despite the fact that the respondents, were recruited through methods designed to generate a representative sample, the sample is unlikely representative of all MSM in Kenya, since most of the respondents were young and of low and medium economic status. Demonstrating that the older and MSM with higher economic status were not represented. The respondents were also drawn from Nairobi and the neighboring county, where the MSM community has been receiving considerable support from the NGOs as compared to MSM in other regions of the country. Future studies should prioritize the elder generations and also the MSM with higher income status and also from other regions of the country with few HIV prevention interventions.

Conclusions
HIVST in Kenya will signi cantly contribute to the rst United Nations 90-90-90 targets as many MSM remain unaware of their HIV sero-status [39]. Availability of data and materials The datasets supporting the conclusions of this article are included within the article and its additional les. The attached questionnaire was developed purposely for this study.

Competing interests
The author declares no competing interests.