Qualitative evaluation of digital vending machines to improve access to STI and HIV testing in South West England: using a Person-Based Approach

Abstract

Objectives

To report the development, implementation, acceptability and feasibility of vending machines offering HIV and sexually transmitted infection (STI) testing kits.

Design

A qualitative study using the Person-Based Approach with patient and public involvement workshops and stakeholder involvement and interviews with machine users, sexual health service (SHS) staff, venue staff and local authority sexual health commissioners. Transcripts were analysed thematically.

Setting

Bristol, North Somerset and South Gloucestershire (BNSSG).

Participants

15 machine users, 5 SHS staff, 3 venue staff and 3 local authority commissioners.

Intervention

Four vending machines dispensing free HIV self-testing and STI self-sampling kits in publicly accessible venues across BNSSG were introduced to increase access to testing for groups at higher risk of HIV and STI infection who are less likely to access SHS clinic testing services (young people, people from black communities, and gay, bisexual and other men who have sex with men).

Results

Machine users reported the service was convenient, easy to use and accessible; however, concerns regarding privacy related to machine placement within the venues and issues of maintenance were raised. Promotional material was inclusive and informative; however, awareness of the service through the promotional campaign was limited. Vending machines were acceptable to venue staff once clear processes for their management were agreed with the SHS. SHS staff identified challenges with the implementation of the service related to the limited involvement of the whole SHS team in the planning and development.

Conclusions

The codeveloped vending machine service was acceptable, addressing some barriers to testing. Resources and protected staff time are needed to support greater involvement of the whole SHS team and service providers in venues. Adopting a similarly robust coproduction approach to the implementation of the machines could avoid the challenges reported. The placement of the machines to assure users privacy and repeated, targeted promotion could encourage service use among target groups.

Strengths and limitations of this study.

• Person-Based Approach ensured meaningful involvement of representatives from the main target groups in the design of the service and promotional material.

• The implementation and evaluation occurred in parallel allowing researchers to capture the process of implementation and provide suggestions for improvements in real time.

• Inclusion of the perspectives of machine users, the sexual health service and venue staff provided a diverse understanding of the implementation process, feasibility and impact.

• Due to the recruitment method the study does not include the views of people unable to locate or use the vending machines.

• Most of the interviewees were of white ethnicity, gay, bisexual and other men who have sex with men or men who have sex with women, and had previous sexually transmitted infection and HIV testing experience.

Introduction

Sexually transmitted infections (STIs) and HIV are unequally distributed in England with higher rates of infections among black African and black Caribbean heritage groups (black communities), gay, bisexual and other men who have sex with men (GBMSM) and young people (<25 years).^{1 2} Barriers to testing and contributors to delayed diagnoses include individual concerns about stigma (related to negative perceptions about the routes of transmission), concerns relating to confidentiality and anonymity as well as service-related factors such as not being offered a test, long waiting times for services and test results, worsened by the COVID-19 pandemic.^3–6

Testing options outside of sexual health service (SHS) clinics such as self-testing (individuals collect and test a sample, eg, saliva or blood, receive and interpret the results) and self-sampling methods (samples taken by an individual sent to a laboratory for processing and results returned to the individual) could help increase access to testing by increasing confidentiality, privacy, convenience and have the potential to reduce healthcare costs because they require less healthcare professional time.^7–12

Research in Brighton and Hove (BH) suggests vending machines containing HIV self-testing, and STI (gonorrhoea, syphilis, chlamydia) and HIV self-sampling kits could help increase access to testing by increasing confidentiality, privacy and convenience.^7–15 However, to date the use of vending machines has been limited to their placement in settings serving GBMSM with limited evidence on the implementation, acceptability and use of vending machines in publicly accessible settings and with other key underserved populations.^{10 11 13}

In 2022, seven vending machines were installed in BH and four machines were installed by Unity Sexual Health (Bristol, North Somerset and South Gloucestershire’s (BNSSG) integrated SHS in the South West England).¹⁶ In these areas, STI diagnosis rates (excluding chlamydia) are greater than the English national average and HIV testing coverage is lower. Both areas are members of the Fast-Track Cities initiative to eliminate HIV.^17–20 The aim of introducing the vending machines was to increase access to testing for those at higher risk of HIV and STI infection and who are less likely to access SHS clinic testing services, namely young people, people from black communities and GBMSM (referred to as target user groups throughout).^17–20 The service in BNSSG, including the interface on the machines, was adapted from the service delivered in BH originally targeted towards GBMSM. A mixed methods evaluation of the machines in both areas found that uptake was high and over half of the people using the machines self-reported infrequent or never testing for HIV or STIs.²¹ In BNSSG, 1395 testing kits were dispensed over a 6-month evaluation period with 43.8% of people aged under 25 years using the machines, 58.3% identified as male (including trans-male), 20.9% were GBMSM and 38.4% used venue A.¹⁶ The proportion of people using the machines that reported they never tested/tested more than 12 months ago for HIV and STIs was 70.4% and 60.0%, respectively.¹⁶ This qualitative study reports the service development, implementation, acceptability and feasibility of using vending machines in BNSSG to improve access to HIV and STI testing.

Methods

The Person-Based Approach (PBA) was used in the planning and optimisation (phase 1) and implementation (phase 2) of the vending machines.²² PBA is an iterative approach to intervention development, implementation and evaluation combining stakeholder and patient and public involvement (PPI), behaviour change theory and mixed methods research. It aims to systematically understand and address the key behavioural and contextual needs of target users.²²

The full process from obtaining test kits through to receiving test results is described in detail elsewhere.¹⁶ Briefly, machine users answered six questions (age, gender, gender of sexual partners, place of residence and time since last STI and HIV test) before choosing a kit to test for STIs or HIV appropriate for their body (eg, penis pack, vagina pack) and sexual partners. A verification code sent to the user’s mobile phone was required before the kit was provided by the machine. STI kits were posted back to the SHS for processing and results were received within 21 days (average 5–10). HIV test kits contained an oral swab that provided results in 20 minutes. Machine users under 16 or 18 years of age in BNSSG and BH, respectively, with symptoms or with a recent history of condomless anal sex with a man and in need of postexposure prophylaxis were signposted to the SHS and to not use the machine.

Phase 1: intervention planning and optimisation

PBA recommends drawing on relevant literature to understand the barriers and facilitators to intervention engagement. A review of relevant qualitative and mixed methods research was conducted by MG to identify barriers and facilitators for people engaging with STI and HIV testing (online supplemental table 1).

While the vending machines were publicly accessible to anyone, it was important to coproduce the service with target user groups to ensure its design met the needs of those less likely to test through sexual health clinics. Input was therefore sought from stakeholders (SHS clinicians and local authority commissioners) and target users through meetings and four online PPI workshops to understand barriers and facilitators to vending machine testing and attitudes towards introducing the machines, to determine the attributes of the optimal machine locations, to refine the machine interface (designed for and in use in Brighton) and to design a promotional campaign. The PPI workshops involved sharing images of the existing vending machine interface to illustrate the process and experience of using the machine. Workshop participants were asked to discuss which aspects of the process and interface were suitable to the needs of the users and which would need to be changed. Participants were invited to share ideas or suggestions of changes as part of the discussion. Participants were also provided with examples of sexual health campaign material to facilitate discussion around acceptable and effective imagery, tone and language. This process created guiding principles (intervention design objectives and intervention features that achieve each objective) and a logic model underpinning the intervention (online supplemental tables 2 and 3).

The four PPI workshops facilitated by MG and JMK, lasting approximately 90 minutes each, were conducted in May and June 2022: two for young people and one each for GBMSM and people from black communities. Recruitment of workshop participants used social media targeted at local community groups and the National Institute for Health and Care Research Applied Research Collaboration-West Young Person’s Advisory Group. The workshops included 9 young people (6 female), 4 GBMSM and 6 from black communities (all female). None of the participants were from North Somerset. Contributors were paid for their time.

Stakeholders included local SHS commissioners and SHS consultants who were members of the project steering committee.

All feedback (both positive and negative comments) about the elements and features of the vending machine interface and promotional material was collated and tabulated into a PBA ‘table of changes’ (online supplemental table 4). The ‘table of changes’ enabled systematic collation of feedback on the interface and promotional material and prioritisation of potential changes. Comments were analysed, based on their frequency and significance, to identify potential changes to materials. A six-point coding framework was used to decide the importance of each proposed change and required consideration of whether the change was important to behaviour change; easy to implement and not controversial; repeatedly stated in the feedback; supported by the experience of either the users, SHS or reported in the literature; did not contradict the logic model, experience or guiding principles; or not to be implemented, for example, because it was unfeasible. Changes were then prioritised using the MoSCoW framework (Must do; Should do; Could do; Would like to do) based on their importance to behaviour change and ease of implementation.

The vending machine interface was adapted by the manufacturers and the promotional material was designed by an artist from a young person-led creative agency. The interface adaptation and promotional material were codeveloped iteratively with workshop participants and stakeholders to optimise acceptability via workshops and email correspondence using a bespoke feedback form.

Phase 2: intervention implementation and process evaluation

A 6-month evaluation of the service was conducted following installation of four machines in September 2022 (1 October 2022 to 31 March 2023). Throughout this period, ongoing feedback was provided to the SHS clinical team and sexual health commissioners to support implementation improvements. This paper describes the findings from one-to-one semistructured interviews with vending machine users, staff working in venues where the machines were installed, staff working in Unity Sexual Health and sexual health commissioners which gathered feedback on experiences of the service. Quantitative findings describing users, usage of the service and survey findings are presented separately.¹⁶

Patient and public involvement

Members of the public were involved in the intervention planning and optimisation as described above. Three of the PPI workshop attendees (described above) also reviewed participant-facing materials such as the participant information sheet. Findings from the study will be shared with PPI contributors and interview participants.

Recruitment

People who used the vending machines were invited to complete an online survey regarding their views and experiences using the machines. The link to the survey was sent via text message as part of the vending process and was also included on posters attached to the machines and social media posts. Details of the survey and data collection process are reported separately.¹⁶ At the end of the survey participants were invited to provide their email address to receive details about participating in an interview. All participants who expressed an interest in being interviewed were emailed by MG with a participant information sheet and invited to arrange a time to be interviewed. Anyone who did not respond to the invitation email was sent two reminders. The recruitment process was identical across the venues.

Venue staff involved in machine installation and maintenance, staff from the local SHS and sexual health commissioners working in BNSSG were invited to an interview via email by MG and provided with a participant information sheet.

All interview participants were offered a £20 high street shopping voucher.

Data collection, management and analysis

Interviews lasting between 30 and 45 minutes were conducted online or by telephone by MG using topic guides specific to the participant group (online supplemental table 5). Service users were asked their past experience of testing for HIV and STI, views on self-sampling and self-testing for STIs and HIV, motivation and experience of using the machines, views on the promotional campaign and how the service could be improved. Stakeholder interviews explored perceptions and experiences of the service, effects on workload and organisational operations, communication and service support provision, service improvements and considerations for future implementation. All interviews were audio recorded then transcribed verbatim, anonymised and stored in compliance with the General Data Protection Regulation (2018) and University of Bristol’s data protection policies.

The sample size was determined following the guiding principle of information power, suggesting that the more information the sample provides, the smaller the sample size needs to be, and vice versa.²³ Analysis, sampling and participant recruitment were conducted in parallel to allow for the continuous assessment of the suitability of the information within the sample with regard to study aims. Anonymised interview transcripts were imported into QSR NVivo software and analysed thematically.^24–26 MG and JMK separately coded a small number of vending machine user transcripts line by line and discussed the coding approach which was initially deductive and adapted from the BH’s team interview coding based on the research questions. MG led the coding of the remaining vending machine user and staff interviews, adding new codes inductively. Themes were generated by reviewing and grouping connecting codes. Coding and theme generation were regularly discussed by JMK and MG and early findings were discussed with the wider team which included non-clinical qualitative and behavioural science researchers, clinicians and commissioners who offered different reflections on the data. Feedback from the interviews was collated into a second table of changes highlighting further areas for optimisation.

Results

Guiding principles

Workshop feedback and the scoping review of the literature identified that people felt that STI and HIV testing was difficult to access and highly stigmatising and that the vending machines should provide a service that was: (1) private and confidential, (2) quick, easy, convenient and accessible to use and (3) did not stigmatise or single out any population group. These guiding principles informed decisions regarding the promotional campaign, design of the interface and machine locations (online supplemental table 2).

Promotional campaign

Based on feedback from the workshops and stakeholders and in line with guiding principle 3, campaign material was designed to be graphic with eye-catching images and informative, positive and empowering text that did not directly identify any target groups (eg, no photography) and avoided risk-based messages which could be perceived as scaremongering and accusatory and not be effective for behaviour change to encourage people to test.

The promotional campaign consisted of two posters, three social media squares and an animated video walk-through of the machine interface available only in English. Addressing guiding principles 1 and 2, the promotional materials highlighted that the machines are ‘Free, Fast, Easy, Private’ and used the phrase ‘Free sexual health testing that works for you’ (online supplemental table 6). Posters with details of the machine locations were put up across central locations in Bristol, the Bristol bus station and three major train stations in the BNSSG area in two phases; phase 1: 2-week campaign across all sites in October 2022, 1 month after the service was launched, and phase 2: 1-month campaign in the bus and train stations in December 2022. Posters detailing the tests offered and the location were also put up in Unity Sexual Health clinic in Bristol and the venues where the machines were installed and further disseminated by the local authorities’ public health teams and local community groups accessed by the target user groups.

The social media promotional materials and video walk-through were posted by stakeholders and local community groups on Twitter, Instagram and Facebook repeatedly following the launch of the service. A boosting social media campaign on Facebook and Instagram aimed at black African men and women aged 18–65 years in BNSSG and Snapchat aimed at 16–18 year-olds in BNSSG ran between January and March 2023 in response to data on vending machine usage among these groups.

All the material was also posted on the Unity Sexual Health website (available at https://www.unitysexualhealth.co.uk/our-services/unityvendingmachines/).

Interface design

The interface of the vending machines was streamlined to reduce the number of pages needed to click through and increase speed of use. Text was simplified to clearly explain all test options so that users could quickly select the most appropriate test for their needs (guiding principle 2). All text was written in English only.

Location

Workshop feedback recommended that to avoid stigmatising (guiding principle 3), the vending machines should be located in venues that were not specific to any one of the target groups, with a high footfall, attracted people from a large geographical catchment, were publicly accessible and with extended opening hours in the evenings and weekends. Local authorities engaged with a range of venues (including pharmacies, libraries, shopping centres) that met these criteria. The machines were ultimately placed in an arts centre with cinema and café in a city centre—venue A; a coworking community enterprise with a café—venue B; and two shopping centres—venues C and D. Three locations identified and trained venue staff to restock and monitor the machines and liaise with the SHS for stock or maintenance issues. The machine in venue B was monitored and restocked by SHS staff.

Evaluation

97 machine users completed the survey and of these, 41 participants requested more information about the interviews. Emails were sent to 39 participants who provided valid email addresses and interviews were subsequently conducted with 15 vending machine users. The majority of the machine users interviewed were over 25 years old, of white ethnicity, GBMSM or men who have sex with women (MSW) and had been tested for HIV and STIs within the last year and used the machine at venue A (table 1).

Table 1.

Characteristics of service users

	Interview n (%)
Total number of users	15 (100)
Age group
<18*	0 (0)
18–25	2 (13.3)
26–35	8 (53.3)
36–45	4 (26.7)
46–55	0 (0)
56+	1 (6.7)
Gender
Male	8 (53.3)
Female	5 (33.3)
Trans-male	0 (0)
Trans-female	1 (6.7)
Non-binary	1 (6.7)
Ethnicity
White*	11 (73.3)
Mixed	2 (13.3)
Asian/Asian British	1 (6.7)
African/Caribbean/Black British	0 (0)
Other	1 (6.7)
Sexual orientation
MSW	4 (26.7)
GBMSM	4 (26.7)
WSM/WM	5 (33.3)
WSW	0 (0)
Other†	2 (13.3)
Previous HIV test
≤3 months	4 (26.7)
3–12 months	7 (46.7)
>12 months ago	2 (13.3)
Never	2 (13.3)
Previous STI test
≤3 months	6 (40)
3–12 months	4 (26.7)
>12 months ago	4 (26.7)
Never	1 (6.7)
Machine location
Venue A	8 (53.3)
Venue B	3 (20)
Venue C	1 (6.7)
Venue D	3 (20)

*White includes English, Welsh, Scottish, Northern Irish, Gypsy, any other white background.

†Other includes 1 trans-female sex with men, 1 non-binary sex with men and women.

‡

§

GBMSM, gay, bisexual and other men who have sex with men; MSW, men who have sex with women; STI, sexually transmitted infection; WSM/WM, women who have sex with men only or women and men; WSW, women who have sex with women only.

Five of the 10 SHS staff invited to be interviewed participated in the study and included one SHS consultant, two nursing staff and two members of the administrative team. All three local authority sexual health commissioners and three of the four venues participated in the study with one staff member interviewed from each venue.

Three themes are presented from the interviews with illustrative quotes (box 1): expectations and experience of vending machines, service delivery and lessons identified.

Box 1. Quotes.

Experience and expectation of users

Because obviously you’ve gotta wait quite some time now for, to go to the clinic and stuff now, so to do it [use a vending machine], it was very convenient um and the results were pretty much instant, 20 minutes or whatever it was, so it was, yeah, a lot, it was a lot easier and it was also a lot more private as well, I didn’t have to you know, like go down so I didn’t have to ring up or anything, so it was great that there was, yeah. (Service user, White, MSW, 26–35 years old, Venue A, Interview 3)

Yes, it was easy. It was convenient. So yes, I think that like I'm gonna switch into just using the machine, because it’s just also right close to the city centre, close to where I work. So it’s just, like, easier to just go there and pick it up. (Service user, White, GBMSM, 26–35 years old, Venue A, Interview 7)

I don’t like the way it’s positioned to be honest in the venue because when you type your information in, there’s like people behind you can see. I mean they probably don’t look but still I felt a bit nervous about selecting all those different things when I was you know, with my back to the rest of the room. (Service user, White, WSM, 36–45 years old, Venue A, Interview 8)

Service delivery

[Vending machine] adds an extra couple of little points to our promotion of being the heart of the community…. it’s the right place for it to be. But beyond that it doesn't make any kind of massive impact on us, positive, negative or otherwise really (…) we’re more than happy for it to stay here for as long as the project sees it’s relevant. Perhaps it’s me but I don't really understand why any venue that was approached would say no, to be honest. It was communicated clearly. (Venue D, Interview 17)

So we found like a bit of a process, but it was quite clunky because we didn't really know where our responsibilities were, where your [SHS] responsibilities were… So yes, I think knowing whose roles and responsibilities were (…) was getting ironed out at that time. And if you guys had that maybe in advance, then that would be helpful for new people.(…) I just gave them feedback in an email about that person who came in who was distressed and then they were like, oh well, you know, you can tell people to do the posting kit and I was like, it would have been helpful to have known. (Venue B, Interview 16)

But the limitation is, or has been, firstly our management of it as a team, because we didn’t have the staff infrastructure initially to pack the machines or to create the boxes—create the kits to go in the machines, so that was a limiting factor which probably meant that at the start of the service they were out of stock more frequently than they should have been. (SHS staff, Interview 21)

We just got told that it was happening. And then when they were gonna be set up, we were just told like, you know, (…) we’re gonna be expecting the kits to come in. So it wasn’t really, which was a shame (…) we felt like we should have had more involvement in it. Because it was something that we were taking on. We would take on the work and we were doing the work that was gonna come in. So we just felt that, you know, it would have been nice if we were more involved in getting involved in the vending machines when they were been set up really, have a bit more say into it. (SHS staff, Interview 20)

It makes us sound a bit useless. We have got patients who are using the testing kits phoning up the (SHS staff) and we don’t have answers for them. We don’t know when they are going to be restocked. We don’t know who is responsible for restocking them. No, we don’t know why it is not working. No, we don’t know how accurately these test kits are. It is just a difficult conversation to have when people phone up asking about them. (SHS staff, Interview 22)

Lessons identified

But in both venues the screen is, it’s quite a huge screen that’s sort of facing out into the lobby of the venue and so it’s sort of asking you questions like, how old are you, who do you normally have sex with, all these things (…) and I just thought, surely these could be placed in a more discreet way so that they’re—the screen is, like find a little corner in a venue where you can sort of put the machine round the corner or something so that like it’s very unlikely that there’s gonna be just people standing around seeing what you’re typing in. (Service user, White, GBMSM, 36–45 years old, Venue A, Interview 10)

They need to have a very clear way of making sure that they can cope with the capacity for kits or any surges that might happen and that the venue is able to stock them up, you know, as required. You know, try and keep it as slick as possible so that we don’t have people being, you know, at that teachable moment or at that moment where they might get a test and then can’t. (Commissioner B, Interview 25)

I think all it would need is someone to be given designated hours to commit to the vending machine (….) at the beginning of the project I had to stop the machines because we didn’t yet have the infrastructure, or the connections with the venues. So then in that situation if there could be a vending machine administrator for example, who wouldn’t need that many hours a week, but would need to have time build into their job to address the vending machine issues promptly; for example, to be able to get there on the day it dysfunctions, call the vending machine team and get the problem troubleshooted quickly. (SHS staff, Interview 21)

[Installation] could have been streamlined where (SHS staff) would have been there for the delivery and (…) could have helped (…) position it and talk us through or even, not just (SHS staff) but just somebody just to be there and be like, cool, so let me train you have to stock up the machine. Here is some stock, we’ll be sending you another box of stock next week or in the next two weeks just to know when stock is then coming in. (Venue B, Interview 16)

I think actually probably what would be a good idea (…) is maybe having like a bit of like a script or something for users(…) But maybe if there was somebody who’s a bit more nervous, who was like, don’t really know what to advise. And you know maybe from this and from information that you've had from people being like asking particular types of questions, you could have that little script there and be like if somebody asks. (Venue B, Interview 16)

I think having a meeting with the teams who manage the results and actually take calls from patients would have been good because then we would have had the opportunity to ask questions about the actual vending machines themselves and how they work, what tests are offered, what the window periods on the Oraquick [HIV self test] are, what the accuracy of those kits are, how they work, because that way if we get users phoning us, and we do get users phoning up asking about them, we can then advise those users. (SHS staff, Interview 22)

Expectations and experience of vending machines

For most machine users interviewed, there was a history of recent and repeat STI and HIV testing and the decision to use the machines was influenced by convenience, accessibility and privacy offered by the service which circumvented challenges such as long waiting times to get a clinic appointment and for postal kits to arrive encountered in accessing in-person or postal testing. Interviewed users stated that the machines would now be their main route for accessing testing, including if symptomatic and/or following high-risk behaviour. For some users, the ready access to testing offered them the option to collect kits for future use when required. Users’ choice of test was informed by their perceptions of risk and history of testing with most users stating a preference for a full STI screen; however, for some users, choice was limited by the availability of kits within the machine.

Interviewed machine users valued the prompt access to testing, at a time and place that worked for them with minimal disruption to their daily routine and without the need for specific planning and time beyond locating the machine. They also felt that the ability to obtain tests from machines without the need to interact with anyone or disclose their sexual health risk could avoid the shame and embarrassment other people may experience when accessing SHS or receiving postal tests.

The experience of most users was positive; the promotional material shown during the interview was easy to understand and informative, the machines were quick and straightforward to use and users valued the requirement for minimal personal information to obtain a kit.¹⁶ Negative experiences related to issues of maintenance and stock with some users expressing frustration and disappointment when they encountered machines that were out of stock or not working. Users approved of the public locations of the machines acknowledging the convenience offered and appropriateness of the selected venues. Balancing privacy and not being overlooked while using the machine in accessible, convenient, high footfall areas was discussed by several participants. For some, the vending machine was situated in a sufficiently private setting, while for others a more private location was preferred.

Service delivery

Venue staff noted that the machines aligned with their organisation’s values and staff in two venues volunteered to act as champions for the service. Venue staff reported that limited information regarding the installation and operation of the machines resulted in initial uncertainty regarding their responsibility. Once these issues were resolved the machines were easy to manage (including restocking) with little to no impact on staff or the operation of the venues. However, venue staff felt ill equipped to advise or signpost users with questions about the machines or who encountered problems using them, for example, when out of order or stock (when out-of-order machines displayed QR codes which provided information signposting to the local SHS).

SHS staff commented that limited staff capacity and resources to establish the service along with the rapid nature of the service introduction (8 months after decision to install) affected the involvement of wider SHS staff in the development and initial service delivery. Additionally, the unanticipated high demand for the service further contributed to challenges encountered with stock production (test kit boxes were packed by SHS staff), processing of returned kits and wider involvement and interactions with the venues.

SHS staff reported that the initially limited and ad hoc communication about the implementation of the machines impacted on their ability to support users and handle queries. SHS staff raised concerns regarding the potential reputational impact on the SHS and venues while these early-stage challenges related to stock production, machine maintenance and support to users were resolved.

Lessons identified

Most machine users had not seen the promotional campaign and only became aware of the machines through online searches to find faster alternatives to postal or clinic testing. Users emphasised the need for targeted and repeated promotion of the machines in settings accessed by priority groups to ensure that people currently not accessing any SHS know about the machines and are encouraged to use them.

Users recommended that steps should be taken to improve privacy such as through the choice of locations within the venues, changes to the position of machines/screens so that they face away from people within the venue or changes to the machines to include a privacy screen.

Commissioners identified access to testing as a priority for their local populations and were keen to see more machines in locations that are easily accessible by the general public such as pharmacies, hospitals, supermarkets and community centres.

SHS staff and commissioners identified the need for appropriate staff capacity, oversight and management as well as time to ensure that the service is integrated into the wider STI and HIV testing. SHS and venue staff also flagged the importance of early engagement and regular communication to ensure a shared understanding of the service, expectations of staff and help to identify and feed back any issues with the vending machine.

Discussion

This study demonstrates the vending machine service overall was acceptable, easy to use, convenient and accessible to interviewed users and had limited negative impacts on venues. However, concerns about privacy related to the placement of machines and lack of stock may have influenced people’s decision to use them. The positive user experience is contrasted with the experience of the SHS where limited resources and engagement with staff along with the rapid implementation and high service demand impacted negatively on the initial delivery of the service and relationship with venues. The insights from the interviews, combined with insights from the survey, vending machine and clinic data (presented separately), indicate support for the proposed intervention guiding principles and logic model. For example, the interview finding that the vending machines were easy and straightforward to use relates to self-efficacy to use the machines, and finding the machines through online searches of testing options supports the intervention function: intrinsic motivation to test.

The vending machine service was identified as an important addition to the SHS, with the potential to reach high-risk groups who may be less likely to attend sexual health clinics or use online postal testing services.^{16 27} Quantitative findings, reported elsewhere, found that during the evaluation, 70% and 60% of the vending machine users accessing testing for HIV and STI had not tested within the last 12 months for HIV and STI, respectively,¹⁶ suggesting the machines were reaching people who are not routinely using SHS. The promotional campaign was intended to engage these groups; however, most interviewees only encountered the material when searching online for options for instant access to testing. Given the reported convenience and public availability of the machines, in the absence of a more effectively targeted campaign, an unintended consequence of the service is that it could risk becoming a replacement testing option for people who do not experience barriers to accessing testing through other routes.

Machines dispensing HIV test kits in targeted settings were popular among GBMSM and have the potential to overcome barriers to testing encountered by other population groups.^{3 4} However, similar results among other target groups are not guaranteed, and prior to implementation ‘local adaption’ of the service involving coproduction is necessary to improve the cultural fit of an intervention and achieve positive outcomes for the different target groups.^{28 29} PBA provided a flexible, systematic and iterative approach to coproduction of the current service, including the machine locations, interface and promotional materials, that could be adapted to the available resources and timescales of the project.³⁰

Innovations in the delivery of SHS occur in the context of limited resources, including lack of protected staff time and additional funding.^{31 32} Service development and implementation work were added to existing high workloads and the rapid introduction of the vending machines made involvement of wider staff challenging. The involvement of SHS staff is an important component of local service redesign to understand the impact on wider service delivery and identify potential blocks.³³ Insufficient inclusion of staff in the development of new services such as described in this study has been previously described as a major barrier to their implementation. Concerted efforts are required to engage all staff groups and services involved in service delivery in strategic discussions about the design and implementation of the new service.^{34 35} However, staff engagement with service innovation is increasingly challenging as SHS funding (via government public health grant) has been steadily cut resulting in high staff pressures and understaffing while service demand increases.³⁶

The main strength of the study was the involvement of public contributors in the design of the service which ensured that the vending machine locations, interface and promotional material designs were appropriate, relevant and acceptable to target users. The inclusion of the perspectives of the SHS and venue staff provided an understanding of the feasibility and impact on both the SHS and venues. Implementing and evaluating the vending machines in parallel captured the process of implementation in real time. The evaluation provided ongoing feedback and suggestions for implementation improvements. However, the findings from the service users may not reflect the experience of all vending machine users, and due to the recruitment method the study did not capture the views of people unable to locate or use the machines. Most of the interviewees were of white ethnicity, GBMSM/MSW and had previous STI and HIV testing experience.

Conclusion

Vending machines dispensing HIV and STI testing kits were convenient, accessible and easy to use and could be a useful addition to SHS. Robust approaches to coproduction, such as using PBA, involving both service users and providers, are important to ensure vending machines overcome the barriers encountered through other testing routes and are sustainably integrated into the SHS. Additionally, for service optimisation, steps to assure users privacy and targeted, repeated promotion of the service have the potential to encourage use among people not regularly testing for STIs and HIV. To support continued service optimisation and uptake, further understanding of service experience and acceptability and partnership working with community organisations particularly among people from black communities, young people and those who have not previously tested is warranted.

Data availability statement

Data are available upon reasonable request. Data are available on application at the University of Bristol data repository, data.bris, at https://doi.org/10.5523/bris.13sdqcehce0qp2g3hocwdhxwsa. Data access is restricted to bona fide researchers for ethically approved research and subject to approval by the University’s Data Access Committee.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s).

Ethics approval

This study involves human participants and was approved by NHS Health Research Authority (IRAS ID 306738). Verbal consent was sought from all participants before any interviews or focus group discussions were conducted and participation in the study was entirely voluntary. Participants gave informed consent to participate in the study before taking part.