Human immunodeficiency virus prevention and testing strategies among men who have sex with men in the UK: the PANTHEON research programme including the SELPHI RCT

Summary of alterations to the programme

Over the 6 years of our programme of applied research, we completed all planned work. We modified certain aspects the programme as follows:

1. extended the process evaluation (PE) to cover not just the initial phase of the RCT (WS1), but the whole trial to provide more in-depth interpretation of outcomes and to assess what works well about the intervention and for whom

2. extended the follow-up time of the trial (WS2) in a no-cost extension to achieve 2 years follow-up of all participants in randomisation B

3. expanded the qualitative work in WS2 to contextualise trial quantitative data in key populations to allow maximal impact and benefit from HIVST; the groups that we conducted further qualitative work with were (i) trans people, (ii) Asian, black and Latin American MSM and (iii) those who reported any harms arising from the RCT

4. estimated the healthcare costs of caring for people diagnosed with HIV using data from a single English HIV treatment centre instead of linking to national Hospital Episode Statistics.

Ongoing human immunodeficiency virus transmission, low rates of testing and late diagnosis in men who have sex with men and the trans community in the UnitedKingdom

In 2014 the HIV epidemic in MSM continued to grow, with 3000 new infections occurring per year despite intensive prevention efforts. 1,2 This was probably due to modest increases in condom-less anal intercourse (CAI)among MSM since the late-1990s, coincident with realisation of the health benefits of antiretroviral treatment (ART)3–5 and as HIV infection was seen as less of a threat to health. 6 This was also against a background from 2011 of HIV-positive people being shown to be dramatically less infectious if using effective ART. 6

Uptake and frequency of HIV testing among MSM in the UK remained low (estimated 30% never tested, 75% not in past year). 7 Approximately 25% of HIV-positive MSM were unaware of their HIV infection and disproportionately contributed to onward transmission (60–80% of new HIV transmissions come from people not diagnosed) as well as presenting late with consequent increased risk of death. 8,9 An estimated 14% of diagnoses made in MSM in the UK were diagnosed late in the course of their HIV infection with advanced immunodeficiency. 10

Additionally there were no reliable data on HIV prevalence, incidence or HIV testing among trans women or trans men in the UK in the UK as historically gender identity had only been captured as male or female, although the UK Health Security Agency (formerly Public Health England) were expanding this categorisation.

While most HIV tests were conducted in genito-urinary medicine (GUM)clinics,11 some initiatives had offered HIVSS, which involves an individual taking their own blood sample, which they post back to the relevant laboratory for testing, and they are subsequently contacted with the result. A further approach was to offer HIVST where the person not only takes the sample but also immediately processes it themselves, so only they are aware of the result. The potential advantage of the self-testing method was that the associated privacy and convenience could increase initial and repeat testing rates and therefore diagnosis. Self-testing was made illegal in the UK in 1992 because of concern, among other factors, that a person could discover they had HIV without ready access to counselling, which may lead to distress or self-harm. With the new climate that effective ART had engendered, the Department of Health (DoH) announced that HIVST would become legal from April 2014, meaning that an immediate assessment of the feasibility, efficacy and cost-effectiveness of providing this technology free to MSM through the NHS was required. Despite theoretical potential, there was an absence of evidence on the potential impacts of self-testing approaches,12 and limited qualitative data to inform how such interventions would be perceived, used and experienced among MSM.

Evaluating the cost-effectiveness of human immunodeficiency virus-prevention intervention options

In addition to increasing HIV testing, other HIV prevention strategies in addition to self-testing include PrEP (where high-risk HIV-negative people take HIV drugs regularly to prevent infection), immediate treatment for all people with diagnosed HIV (not standard of care at the time this programme was planned), behaviour-change interventions and interventions prioritising partner notification of newly diagnosed cases. However, it was unclear which offered the greatest benefit for cost. Most of the relevant existing cost-effectiveness studies were either non-UK focused and/or excluded MSM. A search of PubMed and the NHS Economic Evaluation Database identified only one directly relevant study published in the 5 years preceding 2014. 13

In practice, assessing the cost-effectiveness of a range of prevention interventions can only be achieved by modelling, by synthesising the relevant evidence in order to quantify the long-term impact of reduced HIV population spread. A key part of our proposal was that a ‘comprehensive’ assessment of all relevant prevention activities using a single model would allow prevention activities, singly and combined, to be compared in a coherent and consistent manner. This was done by adapting our existing individual-based simulation model, which had already been used to explain why the incidence of HIV in MSM had been increasing in the UK despite the increase in ART coverage. 1 The model simulates the UK population of MSM from the start of the HIV epidemic, tracking in detail levels of condom-less anal sex (CAS)with long-term and casual partners, risk of HIV acquisition, ART use and risk of HIV disease progression. 1

Outputs

The overall primary anticipated outputs were to provide:

1. an estimate from a RCT of the effectiveness of offering free self-testing for HIV, with reminders to test, in increasing HIV diagnosis compared to standard of care.

2. a comprehensive evaluation of the cost-effectiveness of combinations of potential HIV prevention strategies.

Workstream 1: overview

The main objective of this WS was to consolidate the existing evidence on HIVST and gain an understanding of the values and preferences, and the potential barriers, relating to HIVST among MSM and trans people in England and Wales. This initial work would directly feed into the design and implementation of the RCT planned in WS2. To consolidate the existing evidence on HIVST we planned to undertake a series of systematic reviews and produce a systematic map, in collaboration with the World Health Organization (WHO), in order to update HIVST policies and guidance. Simultaneously we ran focus groups and interviews with key stakeholders, including MSM and healthcare professionals, to gain insight into the relatively new concept of HIVST, using the COM-B (capability, opportunity, motivation, behaviour) framework for understanding behaviour. 14 A further objective was to develop and undertake a process evaluation to assess the feasibility and acceptability of the intervention (HIVST) used in the RCT in WS2, including using data from the pilot phase of the study.

Workstream 1: introduction

With pending legalisation of HIVST in 2014, individuals would soon have the ability to perform unsupervised a rapid diagnostic test at home. This innovation was not without complications; in particular, there were no means to ensure that those with a reactive result presented for care or support and that self-test kits were used free from coercion or violence. Additionally, the impact of self-testing on risk behaviours, other sexually transmitted infections (STIs), and engagement in HIV and STI prevention initiative was unclear.

Workstream 1: research aims and objectives

Workstream 1: discussion

The systematic reviews and the systematic mapping process that were conducted as part of WS1 demonstrated that HIVST increased HIV testing among key populations including MSM, and among MSM HIVST increased the mean number of tests taken over follow-up time. The reviews also demonstrated that although there was a legitimate concern over HIVST and linkage to care after a positive result, this was not the case for MSM and trans people [although not overall when female sex workers (FSW) were included]

Our FGDs demonstrated the diversity of patient preferences for self-testing interventions, and potential adaptations for subgroups. 20,28 The discussions highlighted the importance that MSM placed on the accuracy and window period of the HIVST, the convenience of the distribution of the test, and how HIVST may complement existing testing strategies that MSM employ, in particular to maintain regular testing when not deemed to be at particularly high risk. 28

The process evaluation of the pilot phase of the RCT demonstrated that it was feasible and acceptable to recruit MSM online to a HIVST trial. 32 The usability of the HIVST kit (detailed in WS2 Methods) was found to be high and the experiences of the participants were good. As a result of valuable data collected from the process evaluation within the trial pilot phase, and the identification of subgroups that may particularly benefit from HIVST, the process evaluation was extended to cover the whole RCT, further detailed in WS2, Studies 2B, 2C, 2D and 2E.

Workstream 2: overview

Formative data from WS1 were used to inform the design and implementation of the RCT in WS2. The main objective of WS2 was to conduct a RCT to measure the impact of the provision of free HIVST on new confirmed HIV diagnoses linked to clinical care. The SELPHI RCT was co-ordinated by the Medical Research Council’s Clinical Trials Unit at UniversityCollegeLondon and is detailed in Study 2A. A further objective of WS2 was to undertake the extended process evaluation to cover the whole RCT. Specifically, the process evaluation explored, through qualitative data collection, the experiences and utilisation of HIVST in Study 2B, the impact that HIVST might have for specific groups of participants such as trans people (Study 2C) and Asian, black and Latin American MSM (Study 2D) and finally the types of harms that may have arisen from the SELPHI RCT (Study 2E).

Workstream 2: introduction

Although several randomised controlled trials had assessed the impact of providing HIVST on rates of HIV testing,33,34 SELPHI was designed to evaluate whether this intervention resulted in increased rates of HIV diagnoses that link to clinical care. Linking to care is the critical public health outcome enabling access to ART, which results in individual health benefits and reduction in HIV transmission.

Workstream 2: research aims and objectives

Study 2A: randomised control trial (SELPHI: a HIV Self Testing Public Health Intervention)

This study was designed to answer the overarching question of the programme grant, whether the provision of free HIVST increases rates of diagnosis among MSM, and to specifically meet objectives 1, 2 and 3 of this workstream.

Methods

Human immunodeficiency virus Self Testing Public Health Intervention was an open-label parallel-group randomised controlled trial with a two-stage simple randomisation aiming to enrol 10,000 participants. The SELPHI protocol is published and methods are described. 31 To recruit participants, advertising campaigns placed on social networking websites and via mobile phone applications designed to facilitate sexual and social contact, such as Facebook, Grindr, Hornet and community web pages were used.

Randomisation A took place at enrolment, with participants randomly allocated (in a 3 : 2 ratio) to the offer of a free baseline HIVST (BT) versus no offer of a free baseline HIVST (nBT). An unequal allocation ratio ensured a majority of those agreeing to participate would receive a free self-test and increased the numbers eligible for randomisation B. This second randomisation occurred at month 3 after enrolment and was open only to participants who were initially allocated to the BT group in randomisation A, completed the 3-month survey, remained HIV-negative; reported CAS with ≥ 1 male partner in the previous 3 months (indicating higher risk of incident HIV infection) and were interested in using HIV self-test kits in the future. Eligible participants were randomised (1 : 1) to receive the offer of regular (immediately and every 3 months thereafter) free HIV self-tests + testing reminders (RT) versus no such offer (nRT). Both groups were sent 3-monthly surveys over 2 years of follow-up. A flow chart depicting the study randomisations is shown in Figure 2.

FIGURE 2.

Flow chart of the randomisation process for the SELPHI RCT. a Surveys include questions on sexual behaviour and HIV testing.

Human immunodeficiency virus self-testing kits

In the UK, HIVST was legalised in April 2014, and the first CE-marked kit (BioSURE HIVST, BioSURE, UK) was released to the UK market in April 2015. The BioSURE HIVST kit is classed as a second-generation test (an antibody immunoassay detecting HIV 1/2 antibodies from approximately 28 days after infection), used a whole blood sample and retailed at £30–35 at that time.

The HIVST kit used in the study was the BioSURE^® HIV Self-Test, which had an estimated sensitivity of 99.7% (95% CI 98.9 to 100). 35 In addition to written information provided with the test kit, an online video providing instructions on kit use (produced by BioSURE) (https://youtu.be/N4CAqsmN_6g) was also promoted to participants on joining the study and was available on the study website (www.selphi.org).

Primary outcome measure

The primary outcome for randomisation A was a confirmed new HIV diagnosis within 3 months of randomisation (binary outcome). The primary outcome for randomisation B was the time from randomisation to a confirmed new HIV diagnosis (time-to-event outcome). For both randomisations, the date of diagnosis was defined as the date of the first confirmatory HIV test at clinic. For randomisation B, the censoring date (31 December 2019) was based on the date of the last UnitedKingdomHealthSecurityAgency (UKHSA)linkage exercise.

A HIV diagnosis primary end point is a key feature of SELPHI and distinguishes it from other randomised trials of HIVST, which have all used testing as the primary outcome. HIV diagnoses were primarily obtained from linkage to the national HIV surveillance database, the HIV and AIDS Reporting System (HARS), which is maintained by UKHSA.

Key findings

Randomisation A

Ten thousand seven hundred and ninety-one participants were randomised between 16 February 2017 and 1 March 2018. Of those, 648 were later deemed ineligible (details in Figure 3, 9 asked for all of their data to be withdrawn, and 24 were trans women whose data are reported elsewhere. 36 This left 10,111 participants in the analysis data set, 6049 of whom were allocated to BT and 4062 to nBT. Two hundred and sixty-two (3%) participants subsequently withdrew or unsubscribed from further contact but were assessed for the primary outcome.

FIGURE 3.

Consolidated Standards of Reporting Trials diagram for randomisation A.

Baseline characteristics

Participants’ baseline characteristics and previous HIV testing behaviour have been described. 37 Median age was 33 years [inter-quartile range (IQR) 26–44], 9000 (89%) participants were white, 8118 (80%) were born in the UK, and 4706 (47%) were university-educated (Table 1). Only 81 (1%) participants were trans men. One thousand six hundred and ninety-five (17%) participants had had a HIV test in the 3 months before enrolment, and 1537 (15%) had never HIV tested. The most recent HIV test was conducted at a sexual health clinic for 5089 (61%) of the participants who had tested before, using a self-sample for 1380 (17%), a self-test for 556 (7%), and other modalities for 1291 (16%). In terms of numbers of CAI partners in the 3 months before enrolment, 3330 (33%) of participants reported 1 partner only, 2943 (29%) reported 2–4 partners and 1009 (10%) reported 5 or more partners. At the time of enrolment, 389 (4%) were taking PrEP. Baseline characteristics were reasonably balanced over the two groups.

Survey response rates

Three thousand eight hundred and ninety-five (64%) BT participants completed the 2-week survey, of whom 167 (4%) reported not having received the kit, 110 (66%) of whom later reported receiving the kit. The 3-month survey was completed by 4041 (67%) participants in the BT group and 1566 (39%) in the nBT group. Completion rates were higher among participants who were older, more highly educated, those who had HIV tested more recently and those reporting more CAI partners (data not shown). The final survey was completed by 1695 (28%) participants in the BT group and 1069 (26%) in the nBT group.

Human immunodeficiency virus testing

Five thousand nine hundred and ninety-six (99%) participants allocated to BT accepted the offer of a free HIV self-test kit. Of these, 4263 (71%) participants in the BT arm reported having used the SELPHI self-test kit by the 3-month survey: 530 at the 2-week survey and 3733 at the 3-month survey. Of the 182 remaining BT participants who completed the 3-month survey, 97 (68%) had accessed another HIV test in the previous 3 months. Additionally, linkage with UKHSA data indicated that eight participants HIV tested who did not respond to either survey. Thus overall, 4368 (97%) of BT participants had evidence of any HIV test before 3 months, significantly higher than the proportion in the nBT group (670, 43%) (p < 0.001). Of the 3722 BT participants who had used the SELPHI self-test kit and responded to further questions, 892 (24%) reported having had an additional HIV test after the self-test.

Self-tests and linkage to care

Four thousand four hundred and forty-nine (76%) participants reported having used the self-test kit at either the 2 weeks, 3 months, or final survey. Of these, 4378 (98.4%) participants obtained a non-reactive result, 14 (0.3%) a reactive result, and 57 (1.3%) obtained no result, that is, no lines appeared or there was another problem with the test. Of the 14 participants with a reactive result, 10 either reported a positive confirmatory clinic result or linked to the UKHSA database (implying a positive confirmatory result in a clinic), and one reported a negative confirmatory clinic result. There were no reported false negatives; however, this would have been difficult to ascertain through the surveys. A study clinician attempted to contact the remaining four participants, but was not successful. Median time between enrolment and linkage to care for those reporting using the self-test kit was 9 days (IQR 6–12).

Confirmed human immunodeficiency virusdiagnoses

A total of 34 (0.3%) participants had a confirmed HIV diagnosis within 3 months, the primary outcome for randomisation A (Table 2). There was no evidence of a difference between the two groups (p = 0.64), with 19 (0.3%) BT participants versus 15 (0.4%) nBT participants diagnosed (risk difference 0.0%, 95% CI −0.2% to 0.3%). This finding was unchanged in sensitivity analyses that explored looser and stricter criteria for a HIV diagnosis (data not shown).

Baseline characteristics

Baseline characteristics of randomisation B closely mirrored those of randomisation A. The median age was 34 (IQR 27–44), the majority were of white ethnicity [2052 (89%)], and nearly half were educated to degree level or above [1092 (47%)]. All participants, by definition, had at least one condom-less male anal sex partners within the past 3 months. One thousand one hundred and eighty-seven (51%) participants reported one such partner, 490 (21%) two partners, 616 (27%) three or four partners and 678 (29%) five or more partners. Two hundred and fifty-three (11%) participants had ever taken PrEP, with 148 (6%) being current users.

Survey response rates

Participants were asked to complete a survey every 3 months for 2 years. The number of eligible participants fell sharply after the sixth survey, related to the timing of recruitment. Until this point, the response rate decreased slowly but steadily in each arm (from 84% to 64% for RT, from 83% to 54% for nRT).

Human immunodeficiency virus testing

Overall, a total of 5085 self-test kits were requested by participants in the RT arm (Table 3); 78% kits were reported as having been used at the subsequent survey. The main reason for not having used the test was an intention to use the test at a later date.

Considering all types of HIV testing, 87% of participants in the RT arm reported at least one HIV test during follow-up compared with 45% in the nRT arm, giving a risk difference of 42% (95% CI 31% to 53%, p < 0.001). The proportion of participants reporting the use of a HIV test in the previous 3 months was stable over the course of follow-up, with a range of 84–87% in the RT arm and a range of 34–44% in the nRT arm. Most of the testing in the RT arm was done with a SELPHI kit; self-testing was used infrequently in the nRT arm.

Confirmed human immunodeficiency virus diagnoses

A total of 16 (0.7%) participants had a confirmed HIV diagnosis during follow-up according to the primary definition (Figure 4). There was no evidence of a difference between the two groups (p = 0.63), with nine (0.8%) RT participants versus seven (0.6%) nRT participants diagnosed (hazard ratio 1.27, 95% CI 0.47 to 3.41). This finding was unchanged in sensitivity analyses that explored looser and stricter criteria for a HIV diagnosis (data not shown).

FIGURE 4.

Kaplan–Meier plot of time to confirmed HIV diagnosis in randomisation B (n = 2308).

Study 2B: qualitative interviews with an human immunodeficiency virus Self-testing Public Health Intervention participants

This qualitative study was part of the process evaluation from WS1 which was extended to cover the RCT and aimed to:

1. explore how those utilising self‐tests experience HIVST and the implications for further intervention development and scale‐up.

2. explore how the SELPHI interventions might be experienced by, and the pathways to impact on behaviour for, different groups of RCT participants.

This study would specifically meet objective 4 of this WS.

Methods

We conducted a qualitative substudy in which 37 cis-MSM participants from the SELPHI RCT were interviewed: 10 during the pilot phase (May 2017) and the remainder during the main trial (January–October 2018). Interviews were conducted remotely (n = 17) or face to face (n = 20) based on location and geography. 38,39 Sampling was purposive and aimed for maximum diversity first based on HIV testing experience, then age and ethnicity. 38,39 Our topic guide was developed within the WS1 team and covered testing history, engagement with SELPHI, experience of the interventions and preferences for future HIVST interventions. Interviews were audio-recorded and transcribed verbatim.

Key findings

Aim 1: MSM were motivated to access the intervention because of a reduction of barriers related to stigma from clinic staff and attendees, as well as simultaneous increases in privacy and in convenience. 39 Overall, we found that individuals who had higher barriers to clinic access based on stigma, privacy concerns or geographic issues tended to describe HIVST as facilitating increased testing frequency, but with the potential to reduce STI screening by reducing incentives to access clinical services. However, some minor adverse outcomes (n = 2; fainting, relationship discord) reported by two individuals were discovered spontaneously. 39 Finally, both individuals who had positive HIVST results linked to confirmatory care within 24 hours and described very high intervention acceptability, crediting the technology with ‘saving their lives’. 39

Aim 2: this analysis identified three groups: ‘inexperienced testers’, ‘pro self-testers’ and ‘opportunistic adopters’ (Figure 5). 38 Inexperienced testers were those who had little or no testing history and did not test out of routine; they were typically early in their sexual or testing careers and tended not to be very open about their sexual orientation or practice. They had high psychosocial barriers to testing relating to the COM-B domains of motivation (reflective and automatic), opportunity (social and physical) and capability (psychological). 38 These barriers included lack of risk perception, shame, fear of stigma, privacy concerns and low self-efficacy when considering testing in a clinic. HIVST ameliorated many of these concerns, facilitating testing uptake. The intervention performed most closely to hypothesised mechanisms of action for this group, with strong or moderate evidence across all intervention components (seeFigure 5). 38

FIGURE 5.

Groups of testers in SELPHI RCT.

‘Pro self-testers’ were often at an intermediate point in their sexual careers and sometimes lived partly hidden lives in terms of their sexual orientation. This group tended to have a testing history and were somewhat motivated to access testing, but their frequency was constrained by the high barriers to clinic access they faced (seeFigure 5).

‘Opportunistic adopters’ were men who are well served by existing testing opportunities with few distinct COM-B barriers, except for some minor issues with opportunity (physical). 38 Some in this group used HIVST because of increased convenience, but the majority engaged with the intervention out of novelty and to respond to social norms around testing. The intervention largely did not perform as hypothesised for this group, with weak or moderate evidence across the majority of hypothesised mechanisms of action (seeFigure 5).

Limitations

The SELPHI RCT required informed consent for the HIVST to be delivered to an address, residential or otherwise. This likely excluded the most marginalised and/or those with greatest concerns surrounding domestic privacy. Second, despite substantial outreach, we were only able to interview two participants who reported positive results. These accounts are therefore likely not reflective of the range of experiences within this group.

Successes

The study provided important and current data on HIVST acceptability after use. Prior evidence was largely prospective and explored the perceived acceptability and values and preferences of groups for HIVST. We also provided critical data demonstrating that these tests were feasible to use by a wide range of MSM. We found that the experiences of Asian, black and Latin American MSM were likely to be different from those of white ethnicity and therefore planned a follow-on study specifically exploring these, which is detailed below. In addition, the reports by two participants of adverse events prompted further study of the experiences of those who reported harms in the main trial.

Workstream 2: discussion

The SELPHI RCT recruited over 10,000 MSM into the study and showed that, across England and Wales, the online promotion and subsequent postal delivery of free HIVST self-testing kits were highly acceptable to MSM. Overall, the numbers of new HIV diagnoses in both arms of the trial were low (34 in total) after 3 months of follow-up, which reflected the national decline in HIV diagnoses at the time that SELPHI was conducted. STI testing rates were similar in both arms. Importantly in terms of linkage to care, the majority (10 of the 14 participants) who received a reactive result in the BT arm either had a clinic-confirmed positive result or were linked to the UKHSA database (implying a positive confirmatory result). HIV testing rates were higher in the BT arm and in the subsequent RT arm with no decrease in STI testing or increases in STI diagnoses or CAI, suggesting that while HIVST may be used to mitigate HIV risk, participants were also actively managing other STI risks. However, differing response rates between arms may have introduced some bias in comparisons based on questionnaire responses. Among participants in the nBT group who completed the 3-month survey, the proportion who had a HIV test between baseline and 3 months was twofold higher than the proportion who had a HIV test in the 3 months prior to enrolment. This suggests that participation in SELPHI even without an offer of a free self-test kit may have increased testing rates. One reason for this is that participants may have enrolled in SELPHI because they were considering testing for HIV and when they were not randomised to receive a HIV self-test with the trial decided to test elsewhere.

Qualitative work with groups of MSM from SELPHI demonstrated that HIVST was highly acceptable for diverse groups of MSM, particularly marginalised groups that experience stigma and discrimination in traditional testing settings. 36

Workstream 3: overview

The main objective of this workstream was to understand the contribution of different HIV prevention and testing uptake interventions in reducing the HIV incidence among MSM so far and to estimate the value (i.e. the maximum cost for an intervention to be cost-effective) of activities going forward (including self-testing) to support HIV elimination. To do this we adapted our existing individual-based simulation model, which simulates the population of MSM from the start of the HIV epidemic, tracking levels of CAS with long-term and casual partners and hence risk of HIV acquisition, according to ongoing HIV prevalence in men having condom-less sex (CLS). After infection, we use detailed information from cohort studies to model cluster of differentiation 4 (CD4) count, viral load, use of specific antiretroviral drugs, adherence, resistance, risk of AIDS and death. Areas of uncertainty for which definitive data were needed for the cost-effectiveness analysis were on longitudinal patterns of CLS around the time of infection and as a result of diagnosis. In order to collect this information, a prospective addition to an existing cross-sectional study [the Attitudes to and Understanding Risk of Acquisition of HIV (AURAH2) study (study 3A)] used frequent low-cost brief web-based questionnaires over a three-year period to assess recent sexual activity. In addition to providing information about longitudinal changes in risk behaviour, the AURAH2 study allowed us to investigate incidence and predictors of new HIV infections, and trends in HIV testing behaviour.

To ensure that the proposed economic evaluation would be relevant to the NHS, it was important that a carefully considered set of prevention activities were identified and evaluated, which we did through a systematic review of HIV-prevention interventions (study 3B). Estimates of the UK healthcare costs associated with different stages of infection were also calculated using electronic records from a single London-based HIV treatment centre, for incorporation into the cost-effectiveness model (study 3C).

In this workstream, we set out the overall scope in terms of interventions to be evaluated and conducted a number of studies to inform any necessary changes to the model in terms of structure as well as inform parameter values. This allowed us to estimate the value of HIV prevention and testing activities, including the potential role of self-testing in study 3D.

Workstream 3: introduction

There was a substantial national decline in new HIV diagnoses among MSM in the UK that was coincident with the SELPHI study, where new diagnoses nearly halved from a peak of 3214 in 2014 to 1700 in 2019. 42 The decline was attributed to a number of policies and interventions that limited the onward transmission of HIV, including the offer of immediate ART at diagnosis for PLHIV in 2015,43 the availability of PrEP for HIV-negative people to prevent HIV acquisition,44,45 which was initially only available through the PrEP Impact trial [capped at 26,000 people until 2019 when it became freely available on the NationalHealthService (NHS)]46 or through specific websites set-up to improve access to PrEP in 2015/6,47,48 and ongoing HIV and STI screening offered by sexual health clinics and other healthcare settings. It was important to determine what policy investments might best support further progress towards HIV elimination and offer the most economically sound investment for the NHS.

Workstream 3: research aims and objectives

The aim of WS3 was to estimate the value of strategies, individually and in combination, to prevent HIV in MSM in the UK.

The specific objectives of WS3 were to:

1. provide prospectively collected data to inform parameters of the mathematical model through longitudinal assessment of:

   1. changes over time in the number of CLS partners

   2. number of CLS partners before, during and after the estimated period of primary HIV-infection, and time of HIV diagnosis.

2. assess the extent to which baseline demographic, socioeconomic, health and lifestyle factors and attitudes to HIV are predictive of subsequent levels of CLS, incident HIV infection and HIV testing behaviours.

3. conduct an updated systematic review and consult with key bodies to identify proposed HIV prevention strategies to include in the cost-effectiveness modelling analysis.

4. update estimates of HIV care costs through a resource linkage study.

5. determine the value (from an NHS perspective with outcomes expressed as quality-adjusted life-years) of strategies for preventing transmission of HIV in the UK among MSM, alone and in combination, including free self-testing provision after obtaining an estimate of the effect from our RCT in WS2.

Work package 3: discussion

The results from the AURAH2 study provided a detailed observation of a cohort of HIV-negative MSM at risk of acquiring HIV over a 3-year period and provided key information for specific parameters of the model, particularly on sexual behaviour among MSM. The national decline in HIV diagnoses from 2016 onwards was reflected in the results of the AURAH2 study, indicating that while HIV prevention strategies are working, there remains a clear need to focus prevention efforts on potential transmission risk among a small group of HIV-negative MSM with risk behaviours. The HIV costing study highlighted the importance of timely diagnoses, as the strongest predictors of cost were having a very low CD4 count and being a new patient to the hospital trust. By compiling the information from studies 3A and 3C, the model-based cost-effective analysis demonstrated that multiple prevention strategies are necessary, and, in line with the systematic review on prevention interventions, PrEP offers clear potential benefits. Additionally, increasing HIV testing through modalities such as HIVST is necessary if HIV elimination is to be achieved in the UK. However, a reduction in the cost of delivery of HIV testing and PrEP is necessary in order to provide value for money.

Conclusion

The national decline in HIV incidence can be attributed to a combination of HIV prevention strategies, with PrEP being one of the major contributors, along with high testing rates, immediate HIV treatment initiation and good retention on ART for those diagnosed. The RCT did not demonstrate that self-testing increased HIV diagnoses. However, multiple prevention strategies such as increasing the uptake of HIV testing, PrEP and condom use are needed to achieve elimination of HIV in the UK.

Publications in peer-reviewed journals

WS1:

• Witzel TC, Rodger AJ, Burns FM, Rhodes T, Weatherburn P. HIV self-testing among men who have sex with men (MSM) in the UK: a qualitative study of barriers and facilitators, intervention preferences and perceived impacts. PLOS ONE 2016;11:e0162713. https://doi.org/10.1371/journal.pone.0162713

• Witzel TC, Rodger AJ. New initiatives to develop self-testing for HIV. Curr Opin Infect Dis 2017;30:50–7. URL: https://discovery.ucl.ac.uk/id/eprint/1529770/ (accessed 30 September 2024).

• Witzel TC, Weatherburn P, Burns FM, Johnson CC, Figueroa C, Rodger AJ. Consolidating emerging evidence surrounding HIVST and HIVSS: a rapid systematic mapping protocol. Syst Rev 2017;6:72. https://doi.org/10.1186/s13643-017-0452-4

• Witzel TC, Weatherburn P, Rodger AJ, Bourne AH, Burns FM. Risk, reassurance and routine: a qualitative study of narrative understandings of the potential for HIV self-testing among men who have sex with men in England. BMC Public Health 2017;17:491. https://doi.org/10.1186/s12889-017-4370-0

• Witzel TC, Gabriel MM, McCabe L, Weatherburn P, Gafos M, Speakman A, et al. Pilot phase of an internet-based RCT of HIVST targeting MSM and transgender people in England and Wales: advertising strategies and acceptability of the intervention. BMC Infect Dis 2020;19:699. https://doi.org/10.1186/s12879-019-4247-1

• Witzel TC, Eshun-Wilson I, Jamil MS, Tilouche N, Figueroa C, Johnson CC, et al. Comparing the effects of HIV self-testing to standard HIV testing for key populations: a systematic review and meta-analysis. BMC Med 2020;18:381. https://doi.org/10.1186/s12916-020-01835-z

• Eshun-Wilson I, Jamil MS, Witzel TC, Glidden DV, Cheryl J, Noelle T, et al. A systematic review and network meta-analyses to assess the effectiveness of HIV self-testing distribution strategies. Clin Infect Dis 2021;73:e1018–28. https://doi.org/10.1093/cid/ciab029

• Jamil MS, Eshun-Wilson I, Witzel TC, Siegfried N, Figueroa C, Chitembo L, et al. Examining the effects of HIV self-testing compared to standard HIV testing services in the general population: a systematic review and meta-analysis. EClinicalMedicine 2021;38:100991. https://doi.org/10.7448/IAS.20.1.21594

WS2:

• Gabriel MM, Dunn DT, Speakman A, McCabe L, Ward D, Witzel TC, et al. Protocol, rationale and design of SELPHI: a randomised controlled trial assessing whether offering free HIV self-testing kits via the internet increases the rate of HIV diagnosis. BMC Infect Dis 2018;18:531. https://doi.org/10.1186/s12879-018-3433-x

• Witzel TC, Gabriel MM, McCabe L, Weatherburn P, Gafos M, Speakman A, et al. Pilot phase of an internet-based RCT of HIVST targeting MSM and transgender people in England and Wales: advertising strategies and acceptability of the intervention. BMC Infect Dis 2019;19:699. https://doi.org/10.1186/s12879-019-4247-1

• Wright T, Nicholls EJ, Rodger AJ, Burns FM, Weatherburn P, Pebody R, et al. Accessing and utilising gender-affirming healthcare in England and Wales: trans and non-binary people’s accounts of navigating gender identity clinics. BMC Health Serv Res 2021;21:609. https://doi.org/10.1186/s12913-021-06661-4

• Witzel TC, Bourne A, Burns FM, Rodger AJ, McCabe L, Gabriel MM, et al. HIV self-testing intervention experiences and kit usability: results from a qualitative study among men who have sex with men in the SELPHI (Self-Testing Public Health Intervention) randomized controlled trial in England and Wales. HIV Med 2020;3:189–97. https://doi.org/10.1111/hiv.12818

• Rodger AJ, Dunn D, McCabe L, Weatherburn P, Lampe FC, Witzel TC, et al. Sexual risk and HIV testing disconnect in men who have sex with men (MSM) recruited to an online HIV self-testing trial. HIV Med 2020;21:588–98. https://doi.org/10.1111/hiv.12919

• Witzel TC, Weatherburn P, Bourne A, Rodger AJ, Bonell C, Gafos M, et al. Exploring mechanisms of action: using a testing typology to understand intervention performance in an HIV self-testing RCT in England and Wales. Int J Environ ResPublic Health 2020;17:466. https://doi.org/10.3390/ijerph17020466

• Witzel TC, Wright T, McCabe L, Gabrie ML, Wolton A, Gafos M, et al. Impact and acceptability of HIV self-testing for trans men and trans women: a mixed-methods randomised controlled trial and process evaluation in England and Wales. EClinicalMedicine 2021;32:100700. https://doi.org/10.1016/j.eclinm.2020.100700

• Rodger AJ, McCabe L, Phillips AN, Lampe FC, Burns F, Ward D, et al. Free HIV self-test for identification and linkage to care of previously undetected HIV infection in men who have sex with men in England and Wales (SELPHI): an open-label, internet-based, randomised controlled trial. Lancet HIV 2022;9:e838–47. https://doi.org/10.1016/S2352-3018(22)00266-1

• Witzel TC, Nicholls EJ, McCabe L, Weatherburn P, McCormack S, Bonell C, et al. Understanding experiences of potential harm among MSM (cis and trans) using HIV self-testing in the SELPHI randomised controlled trial in England and Wales: a mixed-methods study. Sex Transm Infect 2023;99:534–40. https://doi.org/10.1136/sextrans-2023-055840

WS3:

• Sewell J, Speakman A, Phillips AN, Cambiano V, Lampe FC, Gilson R, et al. Attitudes to and understanding of risk of acquisition of HIV over time: design and methods for an internet-based prospective cohort study among UK men who have sex with men (the AURAH2 Study). JMIR Res Protoc 2016;5:e128. https://doi.org/10.2196/resprot.5582

• Sewell J, Miltz A, Lampe FC, Cambiano V, Speakman A, Phillips AN, et al. Poly drug use, chemsex drug use, and associations with sexual risk behaviour in HIV-negative men who have sex with men attending sexual health clinics. Int J Drug Policy 2017;43:33–43. URL: https://discovery.ucl.ac.uk/id/eprint/1541552/ (accessed 30 September 2024).

• Sewell J, Cambiano V, Miltz A, Speakman A, Lampe FC, Phillips A, et al. Changes in recreational drug use, drug use associated with chemsex, and HIV-related behaviours, among HIV-negative men who have sex with men in London and Brighton, 2013–2016. Sex Transm Infect 2018;94:494–501. https://doi.org/10.1136/sextrans-2017-053439

• Sewell J, Cambiano V, Speakman A, Lampe FC, Phillips A, Stuart D, et al. Changes in chemsex and sexual behaviour over time, among a cohort of MSM in London and Brighton: findings from the AURAH2 study. Int J Drug Policy. 2019;68:54–61. https://doi.org/10.1016/j.drugpo.2019. 03.021

• Hanum N, Cambiano V, Sewell J, Rodger AJ, Nwokolo N, Asboe D,et al. Trends in HIV incidence between 2013–2019 and association of baseline factors with subsequent incident HIV among gay, bisexual, and other men who have sex with men attending sexual health clinics in England: a prospective cohort study. PLOS Med 2021;18:e1003677. https://doi.org/10.1371/journal.pmed.1003677

• Hanum N, Cambiano V, Sewell J, Phillips AN, Rodger AJ, Speakman A, et al. ; for the AURAH2 Study Group. Use of HIV pre-exposure prophylaxis among men who have sex with men in England: data from the AURAH2 prospective study. Lancet Public Health 2020;5:e501–11. https://doi.org/10.1016/S2468-2667(20)30186-9

• Cambiano V, Miners A, Dunn D, McCormack S, Ong KJ, Gill O, et al. Cost-effectiveness of pre-exposure prophylaxis for HIV prevention in men who have sex with men in the UK: a modelling study and health economic evaluation. Lancet Infect Dis 2018;18:P85–94. https://doi.org/10.1016/S1473-3099(17)30540-6

• Sewell J, Fakoya I, Lampe FC, Howarth A, Phillips A, Burns F, et al. Effectiveness of interventions aimed at reducing HIV acquisition and transmission among gay and bisexual men who have sex with men (GBMSM) in high income settings: a systematic review. PLOS ONE 2022;17:e0276209. https://doi.org/10.1371/journal.pone.0276209

• Miners A, Lampe FC, Cambiano V, Schwenk A, Rodger AJ, Sadique Z, et al. Estimating the hospital costs of adults diagnosed with HIV in England using routinely collected data. HIV Med 2023;24:1115–25. https://doi.org/10.1111/hiv.13529

• Cambiano VC, Miners A, Lampe FC, McCormack S, Gill N, Hart G,et al. The impact of combination prevention strategies on HIV incidence among gay and bisexual men in the UK: a model-based analysis (Erratum in: Lancet HIV). Lancet HIV 2023;10:e713–22. https://doi.org/10.1016/S2352-3018(23)00204-7

Presentations at conferences

WS1:

• Witzel TC, Rodger AJ, Burns FM, Weatherburn P. Values and Preferences of HIV Self-testing Options in UK Men Who Have Sex with Men (MSM): Barriers, Facilitators and Intervention Preferences. World Health Organisation HIV self-testing guideline development meeting. Durban, South Africa; 2016.

• Witzel TC, Weatherburn P, Burns FM, Rodger AJ, Rhodes T. What Role Does HIV Self-testing Have for Men Who Have Sex with Men in the UK? Testing Needs, Social Norms and Biological Citizenship. In: 21st International AIDS Conference. Durban, South Africa; 2016.

• Witzel TC, Gabriel MM, McCabe L, Speakman A, Gafos M, Pebody R, et al. Feasibility of Recruiting Men Who Have Sex with Men and Transgender People Online to a Randomised Controlled Trial Evaluating HIV Self-testing: Results from the SELPHI Pilot. European Clinical AIDS Society Conference (poster presentation); 2017.

• Witzel TC, Guise A, Parkes L, Weatherburn P, Burns FM, Mandal S, et al. Working across epidemics: comparing the acceptability of self-led testing options for HIV and hepatitis C virus (HCV) among men who have sex with men (MSM) and people who use drugs (PWUD). HIV Med 2018;19:S66. BHIVA/BASHH Joint Annual Conference (poster presentation).

• Parkes L, Burns F, Rodger A, Weatherburn P, Witzel TC. The acceptability of different modalities of HIV testing amongst men who have sex with men (MSM) in the UK. HIV Med 2018;19:S142–3. BHIVA/BASHH Joint Annual Conference (poster presentation).

• Witzel TC, Burns FM, Roger AJ, Bonell C, Lampe F, Speakman A, et al. Using Social Science Methods to Understand Pathways to Impact: A National HIV Self-testing Intervention Targeting Men Who Have Sex with Men and Trans People in England and Wales. In Symposia Session: Innovations in Social and Behavioral Science: Frameworks, Methods and Measures. AIDS 2018 (invited speaker presentation).

• Witzel TC, Bourne A, Burns FM, Rodger AJ, McCabe L, Gabriel MM, et al. HIV Self-testing Intervention Experiences and Kit Usability: Results from a Qualitative Study Among Cisgender Men Who Have Sex with Men (MSM) in the SELPHI RCT in England and Wales. AIDS Impact (oral presentation).

• Witzel TC, Burns FM, Bonell C, Rodger AJ, Weatherburn P. What Are the Perspectives on Key Informants on the Implementation of HIV Self-testing in England? A Qualitative Study of Barriers, Facilitators and Anticipated Impacts. BASHH conference 2017; 93:A42.

WS2:

• Rodger A, Dunn D, Phillips AN, McCabe L, Weatherburn P, Lampe F, et al. Sexual Risk and HIV Testing Disconnect in MSM Recruited to an HIV Self-testing Trial. CROI: Seattle, WA; 2019.

• Rodger A, Dunn D, Phillips AN, McCabe L, Weatherburn P, Lampe F, et al. Does Provision of Free HIV Self-testing Increase HIV Diagnosis in MSM? CROI: Boston, MA; 2020.

• McCabe L, Rodger A, Dunn D, Phillips AN, McCabe L, Weatherburn P, et al. Can the Offer of Regular HIV Self-testing Kits Reduce Time to HIV Diagnosis in MSM? Results from the SELPHI RCT. 4th HIV Research for Prevention conference. 2021.

• The Present and Future of HIV Epidemiology. 9th German Hepatitis/HIV Workshop, 17–18 April, Cologne, Germany.

WS3:

• Cambiano, V Is PrEP for HIV Prevention Cost-effective in MSM in the UK? Imperial College London, 15 June 2015.

• Cambiano V. Is Pre-exposure Prophylaxis for HIV Prevention Cost-effective? Seminar in the series ‘Global Health and Development’ at London School of Hygiene and Tropical Medicine, 17 June 2015 on 18 September 2015 in a webinar organized by EATG.

• Cambiano V. Is PrEP for HIV Prevention Cost-effective in MSM?’15th European AIDS Conference, 21–24 October 2015, Barcelona, Spain.

• Cost-effectiveness of ART as Prevention. Plenary Lecture, 22nd Annual Conference of BHIVA, 19–22 April 2016, Manchester, UK.

• Cambiano V. Overview on Cost-effectiveness of PrEP: What Evidence is Available? 27 April 2016, ‘Pre-exposure prophylaxis in the EU/EEA setting: challenges and opportunities meeting’ organized by the European Centre for Disease Prevention and Control, Stockholm, Sweden.

• Cambiano V. Evaluation of the Cost-effectiveness of Pre-exposure Prophylaxis for HIV Prevention Among Men Who Have Sex with Men in the UK. 15 July 2016, 10th European Conference on Mathematical and Theoretical Biology, Nottingham, UK.

• Cambiano V. Brief Overview of Cost-effectiveness of PrEP. 26 October 2016, 13th International Congress on Drug Therapy in HIV infection, Glasgow, UK.

• Sewell J, Cambiano V, Speakman A, Lampe FC, Phillips A, Stuart D, et al. Changes in Chemsex and Sexual Behaviour Over Time, Among a Cohort of MSM in London and Brighton: Findings from the AURAH2 Study. Presented at 2nd European Chemsex Forum, Berlin, 22–24 March 2018.

• Sewell J, Cambiano V, Speakman A, Lampe FC, Phillips A, Stuart D, et al. Changes in Chemsex and Sexual Behaviour Over Time, Among a Cohort of MSM in London and Brighton: Findings from the AURAH2 Study. National HIV Nurses Association Conference 2019. Manchester, UK.

• Hanum N, Cambiano V, Sewell J, FC Lampe, Rodger AJ, Speakman A, et al. Substantial Decline in HIV Incidence Between 2015 -2018 Among a Prospective Cohort of Men Who Have Sex with Men in England. Presented at 25th annual British HIV Association Conference; April 2019; Bournemouth, UK. 2021. URL: www.bhiva.org/file/5ca73250ec605/P106.pdf (accessed 8 April 2024).

• Hanum N, Cambiano V, Sewell J, LampeFC, Rodger AJ, Speakman A, et al. Substantial Decline in HIV Incidence Between 2015–2018 Among a Prospective Cohort of Men Who Have Sex with Men in England. Presented at AIDS Impact International Conference, 29–31 July2019, London, UK.

• Hanum N, Cambiano V, Sewell J, Phillips AN, Rodger AJ, Speakman A, et al. Use of HIV Pre-exposure Prophylaxis Among Men Who Have Sex with Men in England: Data from the AURAH2 Prospective Study. Presented at AIDS Conference 6–8 July2020. URL: https://programme.aids2020.org/Programme/EPoster (accessed 8 April 2024).

• Hanum N, Cambiano V, Sewell J, Phillips AN, Rodger AJ, Speakman A, et al. Use of HIV Pre-exposure Prophylaxis Among Men Who Have Sex with Men in England: Data from the AURAH2 Prospective Study. Presented at British HIV Association Conference 29–31 November 2020. URL: www.bhiva.org/file/5fd20424663ab/O20NadiaHanum.pdf (accessed 8 April 2024).