High Acceptance of HIV Self-testing Among PrEP Recipients before COVID-19 Era: A Cross-sectional Analysis from PrEP Demonstration Project of MSM in China


 Background: Many Pre-exposure prophylaxis (PrEP) users have difficulty attending the quarterly facility-based HIV testing, which leads to the potential risk of drug resistance in the context of breakthrough infection with low drug compliance. We explored the acceptance of HIV self-testing (HIVST) service among PrEP recipients. Methods: MSM were recruited for the PrEP demonstration in four major cities in China from December 2018 to September 2019, provided with regimens of both daily and on-demand PrEP. Facility-based HIV testing was provided quarterly at clinic visits. Previous HIV testing history and acceptance of free HIVST kits to use between each quarterly clinic visit was collected. Correlates of levels of acceptance were analysed using multivariable ordinal logistic regression. Results: We recruited 1,222 MSM. among which 48.5% preferred daily PrEP and 51.5% preferred on-demand PrEP. There was 26.8% (321/1222) had never been to any facility-based HIV testing previously, and the self-reported major reason was that they had already routinely used HIVST. A quarter of the participants (74.5%, 910/1222) had used HIVST previously. There were 1184 MSM (96.9%) accepted to use HIVST between each quarterly clinic visits during PrEP usage, composing 947 ( 77.5%) very willing to, 237（19.4%）willing to, 29 (2.4%) unwilling to, and 9 (0.7%) very unwilling to. Participants preferred daily PrEP (vs. on-demand PrEP, aOR=1.8, 95% CI:1.3-2.4) and had less than 2 times of facility-based HIV testing in the past year (vs. ³2, aOR=1.4,95% CI:1.1-1.9) were more likely to have higher level of acceptance of HIVST.Conclusions: MSM had high acceptance of HIVST, especially among those preferred daily PrEP and with less facility-based HIV testing in the previous year. Offering HIVST services PrEP recipients is feasible and necessary. Above result is of great significance for promoting HIVST among PrEP users during COVID-19, improving awareness of their HIV infection status and ensuring compliance with medication. Future study should exam the impact of HIVST on HIV testing frequency among PrEP users.Trial registration: ChiCTR1800020374 on 27th Dec 2018. http://www.chictr.org.cn/searchproj.aspx


Background
World Health Organization (WHO) and Center for Disease Control and Prevention (CDC) (1) recommend users of oral pre-exposure prophylaxis (PrEP) to test for HIV quarterly, to avoid resistance to antiretroviral therapy (ART) by breakthrough infections (2). Medicine compliance to PrEP is barely 60% among users worldwide (3). Mutations that impact the e cacy of rst-line ART can develop within 2 weeks of PrEP being administered every day to HIV-infected individuals (4). Although the number of cases of resistance to anti-HIV-1 drugs during PrEP is small (5)(6)(7)(8) it remains one of the main concerns for HIV key populations for PrEP acceptance (9,10).
In real world settings, even testing HIV every 3 months at facility-based HIV testing service is implausible among PrEP recipients. There were 75% of PrEP recipients missed their facility-based testing after PrEP initiation (11) and had di culty attending one during PrEP (12) (13). Moreover, among HIV key populations such as men who have sex with men (MSM), there is only 38.0-65.1% had tested for HIV in the past 6-12 months worldwide, and only 47.0-83.6% had ever been tested for HIV in their lifetime (14)(15)(16)(17)(18)(19). It is crucial to explore innovative HIV testing strategies to pair PrEP implementation.
In recent few years, HIV self-testing (HIVST) has changed the course of HIV testing among HIV key populations by reaching more rst-time testers (18,20) and increasing the HIV-testing frequency (21,22) among them, especially in resource-limited settings (23,24). HIVST allows people to take an HIV test and discover their result in their home or other private location (25). In 2016, the WHO published the rst global guidelines on HIVST, in which HIVST was recommended as an additional approach to HIV-testing services (26). Seventy-seven countries have adopted HIVST policies, while many others are currently developing them currently (27).
HIVST has been offered to the HIV-negative partners of HIV-discordant couples and women at risk who were on PrEP to use between quarterly clinic visits (28) or biannual clinic visits (29). HIVST has demonstrated high acceptance, high use, and ease of use among PrEP recipients in the settings mentioned above (28,30). Higher preference for HIVST than facility-based HIV testing during PrEP among female sex workers has been documented in one setting (31). However, HIVST uptake among MSM and its correlates are not known.
COVID-19 is still in a pandemic stage. Social distancing is still a primary to preventing the spread of SARS-COVI-2, causing a severe impact on maintaining the routinely traditional clinical-based HIV testing (32). Therefore, the need to develop HIVST in the COVID-19 era has become increasingly prominent. The CDC recommends that HIVST services be provided to PrEP users when other options are not available (33). But so far there is no report of PrEP MSM's willingness to apply HIVST. This study is of great signi cance for the promotion of the application of HIVST among PrEP users.
We aimed to assess acceptance of HIVST among PrEP candidates and explore its correlates, including demographics, HIV-testing history, preference for PrEP regimen, HIV-related behaviour. This study will shed light in the utility of HIVST in the era of PrEP, and provide evidence for paired HIV-testing strategy with PrEP implementation. Participant recruitment was conducted o ine and online using convenient samples and snowballing (i.e., chain referral sampling) by: (i) distribution of posters and brochures in the HIV VCTs at the four study sites for MSM who arrived for counselling and testing; (ii) postings in MSM chat groups and on the study sites' o cial WeChat™ and Tencent QQ™ accounts (two of the most widely used social media applications in China); (iii) outreach events run by community-based organisations (CBOs) in bars, public baths and parks that provided study information, and MSM peer referral.

Design, setting and procedures
All individuals who were interested in the recruitment advertisement of CROPrEP online and o ine, needed to registered for and attend to an on-site screening appointment for eligibility of the CROPrEP trial. This on-site screening provides free testing for HIV infection and syphilis for adult MSM and transgender women. After providing written informed consent, participants willing to enrol underwent consultations with physicians and laboratory testing. The information of demographic, HIV related risk behaviour, HIV testing history and acceptance of HIVST was collected.

Measures
Data were collected through self-administrated questionnaires on each participants' smartphone.
Participants were surveyed for acceptance to use free HIVST kits between each quarterly clinic visit, previous experience of facility-based HIV testing, and HIVST (frequency of using each type of testing service, as well as the reasons for using and not using them), their preference for PrEP regimens (daily or on-demand), demographics (age, education level, marital status, monthly income), HIV-related risk behaviour in the past 3 months (frequency of homosexual anal intercourse, and existence of condom-less anal intercourse). The levels of acceptance of HIVST were de ned as "very willing to accept", "willing to accept", "unwilling to accept" and "very unwilling to accept".

Statistical analyses
We conducted statistical analysis by SPSS v20.0 (IBM. Armonk, NY, USA). Demographics, preference for PrEP regimen, history of HIV testing, and level of acceptance of HIVST among all participants were described using numbers and percentages. Univariable and multivariable ordinal logistic regression analyses was used to assess the predictors of an increased level of acceptance to use HIVST between each quarterly clinic visit. Variables with P < 0.2 in the univariable analysis were included in the multivariable model. Variables in the nal model were selected with an entry procedure. Model tness was assessed by the goodness-of-t statistics. Odds ratio and adjusted odds ratios were calculated by a general linear model. P < 0.05 in the multivariate analysis was the cut-off for a signi cant difference.

Results
A total of 1222 men participated in on-site screening and completed the relevant screening programs. Among these men, the median age was 30-year-old (Interquartile rang, IQR 25-36) and 79.4% (970/1222) had education level as undergraduate and above. There was 68.3% (835/1222) of them had unprotected anal intercourse during the previous 3 months. Among all participants, 48.5% (592/1222) preferred daily medication and 51.5% (630/1222) preferred event-driven medication (Table 1).   The multivariable ordinal logistic regression analysis revealed the signi cant and independent predictors of higher acceptance of HIVST use during PrEP. Participants who preferred daily PrEP were 1.8-times (95% CI: 1.3-2.4) more likely to have higher acceptance of HIVST than those who preferred on-demand PrEP (P < 0.001). Compared with those with an education level of senior high school or below, men having an education level of undergraduate and above were 2.6-times (95% CI: 1.1-2.1) more likely to have higher acceptance of HIVST (P = 0.023). Compared with people who had monthly income lower than 858 USD, men with monthly income above 858 were 0.7-times (95% CI: 0.5-0.9) more likely to have higher acceptance of HIVST (P = 0.017). Compared with people who had ≥ 2 facility-based HIV tests in the previous year, those who did not were 1.4-times (95% CI: 1.1-1.9) more likely to have higher acceptance of HIVST (P = 0.014). Compared with people who had ≥ 3 HIVSTs in the previous year, those who did not were 0.3-times (95% CI: 0.2-0.4) more likely to have higher acceptance of HIVST (P < 0.001) ( Table 3).

Discussion
We exam the acceptance of HIVST among PrEP candidates and its correlates in a multicentre oral PrEP demonstration project in China. The median times of HIVST and facility-based HIV testing in the past year of these candidates were 3 and 2 respectively. Nearly all (96.9%) of the PrEP candidates accepted HIVST. Men with a preference for daily PrEP and had used facility-based HIV testing less than twice in the previous year were more likely to have higher level of acceptance of HIVST. This result showed the necessity and feasibility of offering HIVST service during PrEP.
The rate of acceptance of HIVST is high among PrEP candidates of MSM, which is in accordance with previous studies of sero-discordant couples and females using PrEP in Africa (28)(29)(30)(31)34). This high rate can be explained by the fact that HIVST had already became a main way of HIV testing among these MSM. Among these PrEP candidates, the median number of HIVST use in the previous year had already outnumbered that of facility-based HIV testing. There were three-quarters of the PrEP candidates had used HIVST previously and nearly all of them have heard about HIVST before. The self-reported major reasons of this acceptance were good privacy, easy to use and getting testing results quickly. Also, this study was conducted before the COVID-19 pandemic, which indicates possibility of a higher acceptance of HIVST during the pandemic. This high acceptance of HIVST also suggests a method to maintain this ongoing PrEP trial, since the quarantine against COVID-19 is still required. Above results suggest feasibility to provide HIVST to MSM PrEP users.
This study also indicates necessity to provide HIVST to MSM PrEP users. The WHO recommends general MSM test for HIV every half year (35) and PrEP users test HIV at a facility-based testing service at least every quarter (36). However, there were more than a quarter of the PrEP candidates had never used any facility-based HIV testing before, which indicates potential hardship in maintaining their attendance to facility-based HIV testing during PrEP. Missed attending facility-based testing and di culty in attending clinical visits leads to interruption and discontinuation of PrEP (12) (13). Fortunately, PrEP candidates with less than two facility-based HIV testing in the past year were more likely to accept HIVST. Providing PrEP users with HIVST can serve as additional options for HIV testing.
Preference for daily regimen of PrEP is also correlated with higher acceptance of HIVST during PrEP. This study allowed PrEP candidates to choose their regimen between daily and on-demand (37). Real-world studies have shown that, compared to those who chose on-demand PrEP, MSM chose daily PrEP were having more HIV-related risk behaviour (38,39). Resistance to ART is more likely to occur with a higher selection pressure of the drug (4). Considering the higher exposure to HIV infection and generally low adherence to PrEP among MSM, with daily administration of PrEP to a HIV-infected person, mutations that affect the e cacy of rst-line ART can develop in 2 weeks (4). Providing HIVST for daily users of PrEP could meet their needs. Nevertheless, challenges in supervision for on-demand PrEP must be addressed because non-daily use of PrEP becomes more common outside of controlled research settings (8).
This study has both strengths and limitations. This study explored acceptance of HIVST among MSM PrEP users for the rst time, revealing the high necessity and feasibility to paired PrEP with HIVST. However, since fourth generation of HIVST, which could can detect both HIV antibodies and p24 antigens, we provided the third generation HIVST to PrEP candidates. The high acceptance of HIVST reported in this study indicates better acceptance of fourth generation of HIVST among PrEP users in the future. Also, this study is a cross-sectional design, which cannot provide actual use in the follow-up and impact of HIVST usage on behaviour of PrEP users. The data of actual usage of HIVST and its impact on behaviour among PrEP users will be reported soon.

Conclusions
The

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.