Virtual reality and augmented reality smartphone applications for upskilling care home workers in hand hygiene: a realist multi-site feasibility, usability, acceptability, and efficacy study

Abstract Objectives To assess the feasibility and implementation, usability, acceptability and efficacy of virtual reality (VR), and augmented reality (AR) smartphone applications for upskilling care home workers in hand hygiene and to explore underlying learning mechanisms. Materials and Methods Care homes in Northwest England were recruited. We took a mixed-methods and pre-test and post-test approach by analyzing uptake and completion rates of AR, immersive VR or non-immersive VR training, validated and bespoke questionnaires, observations, videos, and interviews. Quantitative data were analyzed descriptively. Qualitative data were analyzed using a combined inductive and deductive approach. Results Forty-eight care staff completed AR training (n = 19), immersive VR training (n = 21), or non-immersive VR training (n = 8). The immersive VR and AR training had good usability with System Usability Scale scores of 84.40 and 77.89 (of 100), respectively. They had high acceptability, with 95% of staff supporting further use. The non-immersive VR training had borderline poor usability, scoring 67.19 and only 63% would support further use. There was minimal improved knowledge, with an average of 6% increase to the knowledge questionnaire. Average hand hygiene technique scores increased from 4.77 (of 11) to 7.23 after the training. Repeated practice, task realism, feedback and reminding, and interactivity were important learning mechanisms triggered by AR/VR. Feasibility and implementation considerations included managerial support, physical space, providing support, screen size, lagging Internet, and fitting the headset. Conclusions AR and immersive VR apps are feasible, usable, and acceptable for delivering training. Future work should explore whether they are more effective than previous training and ensure equity in training opportunities.


Background and significance
Hand hygiene is a simple infection prevention and control (IPC) behavior often practiced incorrectly by healthcare workers. 1,24][5] This is concerning given that appropriate hand hygiene is one of the most simple but effective means of preventing the transmission of harmful microbes and healthcareassociated infections, including antibiotic-resistant bacteria and COVID-19. 6,7Previous hand hygiene interventions have focused on acute care [8][9][10] and have overlooked care workers. 3,113][14][15] In VR, users wear a headset to feel immersed in a digital environment, while AR requires an overlap between the real world and the digital.
AR and VR apps for hand hygiene training afford learners opportunities that traditional learning approaches do not.For example, VR enables learners to interact with and problemsolve in context, given that it employs a three-dimensional simulated reality. 16AR hand hygiene apps can also facilitate psychomotor learning through repetition and testing, and encourage self-directed and purposeful practice. 15Lastly, when delivered on smartphone apps, the training can be easily accessible and use behavior change techniques like prompting practice (through push notifications) or personalization. 12xisting AR and VR hand hygiene technologies have been evaluated in hospitals 17,18 and universities 15 but not in care homes (which may include residential or nursing homes).The development of other VR training is also ongoing. 193][24] This highlights a need to understand how, for whom, and in which contexts the technologies work.Furthermore, AR/VR hand hygiene interventions and implementing them in care homes may be complex as they will be delivered to learners with varying experience and in organizations with different policies and infrastructure.Lastly, implementation and uptake require careful consideration to determine which technology (AR or VR; immersive or non-immersive) is appropriate for each care home and the learners within them, which requires awareness of the available infrastructure and personal preferences. 25e are undertaking a program of work that seeks to understand how, for whom, and in which contexts or circumstances AR/VR can be used to upskill care home workers in hand hygiene practice.The program takes a theory-based perspective, whereby evidence reports on the processes that can lead to outcomes and prerequisites and contexts/circumstances for the intervention are considered.The process is iterative, consisting of developing, testing, and refining a program theory.We take a realist approach, 26 which is theoretically driven, based on the principles of critical realism and considers that some interventions are complex and do not work for everyone equally.Realist evaluations therefore focus on understanding the mechanisms that are triggered by interventions and required to lead to outcomes.
Previous phases included a systematic review of 90 hand hygiene apps to identify high-quality existing apps that use AR/VR 12 and a realist review of 80 papers to form a program theory on AR/VR training for healthcare workers which was tested and refined with 46 empirical studies. 27Lastly, interviews with 25 care home staff helped to refine the program theory further and understand implementation considerations. 25Through an iterative and systematic process various mechanisms have been identified as crucial for learning hand hygiene.These include repeated practice, task realism (ie, perceived relevance to work), feedback and reminding, and interactivity.
The current study forms the next phase in our program, where we implemented the training in a sample of care homes to address questions regarding intervention content and delivery, acceptability to learners and training completion, and fit within care homes.The purpose is to justify undertaking a full-scale evaluation 28 and, while not the main focus, gather evidence on short-term/immediate outcomes to suggest potential effectiveness. 28Additionally, the work will fill a gap in research on hand hygiene training in care homes and on the feasibility, usability, acceptability, and real-world experiences of implementing AR/VR training in care homes.It will also help to explain how the technologies promote learning, by providing evidence for the mechanisms.

Objective
This study aimed to assess the feasibility and implementation, usability, acceptability, and efficacy of VR and AR smartphone apps for upskilling care home workers in hand hygiene.
A secondary objective was to explore mechanisms for learning triggered by the apps.

Design
A realist mixed-methods study with pre-and post-test measures was conducted.The University of Manchester's Proportionate University Research Ethics Committee approved the study (Reference: 2022-14903-24944).Participants provided written consent.The study aligns with the Good Reporting of a Mixed Methods Study checklist 29 and the Realist And Meta-narrative Evidence Syntheses: Evolving Standards guidelines. 30

Setting, sample, and recruitment
We recruited a sample of care homes in Northwest England, United Kingdom with consideration of varying characteristics.The homes were identified from their websites and from our previous study 25 in which we identified managers through networks of colleagues and the Care Quality Commission website.Managers recruited care workers (assistants or seniors) who spoke English and were older than 18 years.Participants were compensated with a £20 voucher.
We aimed to recruit up to 50 participants, with a minimum of 12 using AR and 12 using VR, which is considered sufficient for pilot and feasibility studies. 31

Procedure
Data collection occurred in-person at the care homes across 9 days between October and December 2022.The researchers helped care home managers choose which technology to deliver the training (AR or VR) by considering their preferences and access to the Internet and a computer.Participants first gave informed consent and completed the baseline and pre-test measures (Figure 1).This included providing demographic information (age, gender, ethnicity) and their role, experience working in care homes, previous hand hygiene training, and AR/VR use.They also answered 2 questions about hand hygiene compliance and 2 common reasons for poor practice, inspired by Baier et al. 32 Participants then completed the training, followed by post-test and acceptability measures.Finally, we conducted semi-structured interviews and made observations throughout.

Measures and analysis
Informed by our interview study, 25 the primary outcomes were feasibility and implementation, usability, acceptability, and efficacy of the AR/VR training, while the secondary outcomes were the mechanisms of repeated practice, task realism (ie, perceived relevance to work), feedback and reminding, and interactivity.Table 1 shows how these were measured.
We analyzed the quantitative data descriptively using IBM SPSS (v.27) (see Table 1 for further detail).The 10-item System Usability Scale (SUS) 33,34 measured the usability of the technologies.Similar methods to a paper by Dowding et al. 35 were used to analyze the SUS data.Responses were first converted to range from 0 to 4, with 4 indicating the most positive response.They were then summed and multiplied by 2.5, to give a score out of 100.Descriptive statistics were generated.The scale scores ranged from 0 to 100, with scores below 50 indicating poor usability, over 70 indicates good usability, and scores over 85 reflect excellent usability. 36he videos of the hand hygiene technique (before and after the training) were assessed for alignment with WHO best practice guidelines 37 using an 11-item checklist based on the 6 poses, the length of the practice (minimum of 30-40 s) and whether the hand surfaces were covered with enough sanitizer.Each item was scored 1 if present and 0 if absent.The items were summed, with higher scores indicating better technique.One researcher (N.G.) first rated 20 videos (21%).A second rater who was blinded to the technology/intervention groups rated the same videos.Inter-rater reliability was calculated using a Cohen's Kappa on SPSS and indicated substantial agreement between the raters: j ¼ 0.79 (95% CI, 0.711-0.871),P < .001.Consensus was reached through discussion, and this informed the rating procedure for the remaining videos.
Similar to our previous study, we used NVivo (version 12.7.0)and a combined deductive and inductive approach to analyze the qualitative data, consisting of the interviews and observations. 25The mechanisms identified in our previous work were used for the deductive approach.New themes were identified from the data pertaining to implementation considerations and reactions to the technology.The researcher who collected the data (N.G.) analyzed all the data by adding, changing, and moving codes while a second (D.D.) double-coded the data for 10 participants (21%).Disagreements regarding interpretation and the codes were discussed until consensus was met.The final themes were then determined and represented by interview quotes or observational • Overall compliance (0%-100%) and a scorecard of missed instances was recorded by taking screenshots of the VR app or computer screen.

Intervention (training)
Informed by our previous interview study, 25 the intervention/ training was a short "refresher session" of hand hygiene best practice.A similar approach was taken when using VR to train hospital staff. 17The intervention functions included: • Education: increasing knowledge/understanding of when infections spread and of best practice guidelines.This may influence psychological capability and reflective motivation.
• Training: imparting skills that align with best practice hand hygiene which may impact a learner's physical and psychological capability, physical opportunity, and automatic motivation (as practice becomes subconscious).• Modeling: providing examples for imitation to impact automatic/subconscious or reflective/conscious motivation.
As the training needed to be suited to the context, 2 technology options were used: (1) Tork VR Clean Hands Training app in a headset, 13 with a non-immersive option and (2) Sure-Wash. 14The apps were selected from our review of 90 hand hygiene apps as they were among the highest scoring. 12The training took approximately 15 minutes to complete.
A researcher with behavior change expertise gave feedback on the intervention, measurement tools, and study design.A member of the public tested the AR intervention and outcome measures, giving feedback on understanding and readability.A care home worker, a registered nurse with formal IPC training and an individual who provided home care provided feedback on the interventions, specifically the care and IPC aspects.

VR training
The VR groups were screened for health problems or preference to determine whether the headset was suitable.If unsuitable, non-immersive training was completed on a laptop on a stand (for ergonomic optimization).Learners navigated the game via the Bluetooth mouse (Figure 2).For the immersive training, a smartphone with the Tork VR Clean Hands Training 13 app was placed in the DESTEK V5 headset.Participants were seated to avoid bumping into objects or tripping (Figure 2).They wore a disposable VR mask under the headset and used a Bluetooth remote and "swiveled" to navigate by moving the head and upper body.
Both options delivered the same training.The app demonstrated the WHO's 6 steps for hand hygiene. 37Learners then played a nurse and cared for 3 patients in the game, identifying moments for hand hygiene from the WHO 5 Moments for Hand Hygiene. 37A patient alarm prompted learners to move to the next patient.Feedback included reminders for missed hand hygiene moments, when to use gloves and how to dispose of them and medical waste.A compliance score and summary of the missed instances were presented upon completion.

AR training
We used the premium SureWash 14 (GLANTA) app.Participants read the educational content (lesson), which explained the importance of hand hygiene, WHO 5 Moments for Hand Hygiene and the correct technique.The smartphone was placed on a table, and learners practiced the technique above the device, which used a motion-sensing algorithm to detect movement (Figure 3).Level 1 guided learners through the WHO 6 steps 37 and gave feedback on switching hands.Level 2 taught how to build a smoother flow by putting the poses together.This was completed twice to record changes over repeated attempts.Level 3 asked learners to perform the poses in under 40 seconds.After each level, the app gave a pass/fail and reported challenging poses and the overall time to complete the poses.

Feasibility and implementation
Care home recruitment and technology choice Five care homes participated in the study.All were privately owned, with 3 belonging to a chain organization (see Table 2).Three were in neighborhoods with the highest levels of deprivation (Indices of Multiple Deprivation 40 rating of 1), and 2 were in less deprived neighborhoods (7 and 9, respectively, of 10).One home experienced Internet issues during the study, but all owned computers and iPads.In one, staffs were allowed to use smartphones during quieter days for training.Another only allowed seniors to use smartphones to communicate with management.
Two care homes chose VR to try a different approach to training, while 3 chose the AR training for convenience and infrastructure reasons, as it does not require a computer or Internet.

Uptake, session and training completion, and engagement
Fifty staff members were enrolled in the training, but 2 did not complete the session due to time constraints so were ultimately excluded from the analysis (see Figure 4).Of the 48 participants, 41 were interviewed.On average, the interviews lasted 4.23 min (61.16;range 2.08-8.20).There were approximately 36.5 hours of observations.
As summarized in Table 2, the 48 participants had an average age of 38.10 years (613.57),ranging from 18 to 67 years, with 15% male and 25% identifying as Black, Asian, or mixed ethnicities.This is similar to England's adult social care workforce, where the mean age is 45 years, 18% are non-female, and 23% have Black, Asian, or minority ethnicity. 41On average, participants had 5 years and 8 weeks experience working in care homes (range: 3 weeks to 22 years).The majority (56%) had previous experience of VR in their personal or professional lives.
Ninety percent had received hand hygiene training and 63% had their hand hygiene monitored informally.On average, participants estimated hand hygiene compliance at their care homes at 89% (612.6,range: 60-100%).The 2 most common reasons for non-compliance were forgetting and lack of knowledge of best practice.
Uptake was defined as the proportion of participants who completed the training session they were offered.There was 100% uptake for the AR training, as all 19 staff who were offered the AR training completed it.Twenty-nine participants were offered the immersive VR training, however only 21 participants completed the immersive version, as the remaining 8 were unable to use the VR headset due to health reasons.Uptake for the immersive VR training was therefore 72%.All 8 participants completed the non-immersive training after being offered it (100% uptake).

Training engagement and implementation considerations
For the AR training, 47% (n ¼ 9) of participants were observed skim-reading the lesson content, while 26% (n ¼ 5) used the "help" button to get assistance, and 21% (n ¼ 4) stood up for the training.For the immersive VR training, 29% (n ¼ 6) read the instructions aloud, spoke to the patient characters, or narrated their gameplay.Four (14%) held the headset to stabilize it.A stress/panic response was observed during the patient alarms, whereby 48% (n ¼ 10) for the immersive VR training and one in the non-immersive training moved quicker or verbalized surprise.One yelled, "I'm coming patient 3." Support by managers and senior carers was crucial for recruitment and scheduling.For the immersive VR training, the space was vital for safety reasons, as some participants moved forward and needed to be moved away from furniture.
Interruptions were frequent, and many sessions were paused due to staff, residents, and other activities (eg, COVID-19 vaccines for staff).
From the interviews (Table 3), participants identified characteristics that may limit the uptake of the training, including limited English, health issues (eg, nausea), claustrophobia or poor eyesight, and limited digital skills.Participants suggested encouraging potential non-users, providing step-by-step instructions, health pre-screening, and a training exercise for the VR headset.
Technology-related issues were observed whereby some participants commented on the small screen size and attempted to make the AR content bigger by zooming.One suggested using an iPad instead.Some participants struggled to fit the headset due to their hairstyles or smaller head sizes.When the Internet disconnected, the non-immersive VR training lagged and lost audio.

Usability
The immersive VR and AR training had the highest mean SUS scores at 84.40 (68.66, median: 85) and 77.89 (615.62,For the AR training, 9 participants (47%) needed prompting to keep their hands in the frame and use the app (eg, changing poses).Four (21%) were observed struggling with the fingertips pose, which the app struggled to detect: I was doing the correct motion, but the app was struggling to pick it up. ..But overall, I feel like the app works well.(Participant 36, AR training).
For the immersive VR training, 11 participants (52%) needed prompting to navigate the training and unfamiliar tasks (eg, wound care).For the non-immersive training, 7 participants (88%) had issues navigating the game and commented on the sensitivity and difficulty moving around in the digital hospital.Most overcame these problems as they progressed through the game: I think when you first start off it is really hard to get the gist of how it's working, like, when I thought the arrow, I didn't know it was the blue dot we were following, I thought it was the arrow really. ..by the third patient I did get the hang of it but it took me three patients to get there.(Participant 14, non-immersive training).

Acceptability: learner satisfaction and tolerability
Participants were satisfied with the training, with 90% (n ¼ 43) agreeing or strongly agreeing that they would support further work of the technology in their training.Scores were highest for the immersive VR (mean: 4.67 of 5, 60.58) and AR training (mean: 4.37, 60.60), whereby 95% agreed or strongly agreed with the statement.Scores were lowest for a Nursing and residential homes provide 24/7 care and support.Nursing homes provide higher levels of medical care as qualified nurses are available on-site.
b Care assistants are unregistered staff who provide care-related tasks such as dressing, washing, and physical support.They may also perform domestic duties like food preparation and emptying bins.
c Senior care staff delegate tasks and have other responsibilities (eg, administering medications and record-keeping).Abbreviation: SD, standard deviation.
the non-immersive VR training; only 63% agreed or strongly agreed with the statement (mean: 3.75, 61.04).
We observed 3 participants in the AR group displaying discomfort from holding their hands above the smartphone, evidenced by shaking their arms and commenting on minor discomfort.For 3 others, dry skin was observed on the device and table from rigorous rubbing.One participant reported mild dizziness during the immersive VR training from moving between rooms in the game.
After the training, no-one could recall all 5 Moments for Hand Hygiene.Seventeen participants mentioned hand hygiene around other tasks, such as before food preparation, feeding and laundry, and after emptying bins.Some mentioned after sneezing or coughing (n ¼ 5) and using the toilet (n ¼ 7).Only 52% (n ¼ 25) correctly stated that hand hygiene should be performed before and after wearing gloves.However, 73% (n ¼ 35) identified the ideal length of hand hygiene compared to 56% (n ¼ 27) before the training.Majority (85%, n ¼ 41) also identified healthcare workers' hands as the main route of cross-transmission of potentially harmful germs between patients.

Hand hygiene confidence
Figure 5 shows the average scores per group for the confidence questions, compared to baseline.Almost everyone (96%, n ¼ 46) agreed or strongly agreed that they felt

Considerations Supporting quotes from interviews
Learners with health issues or limited digital literacy Certain people won't be able to use that training due to either illness or different reasons like nausea or sickness and things like that.So obviously that's got to be a challenge trying to roll that out to the people.And obviously people that are not very good with technology may struggle a bit.

(Participant 11, immersive VR training) Language barriers
There will be a little bit of language barriers as well if you, you know, get in contact with staff who probably are not like 100%, you know, there is some language barriers, then you may struggle a little bit.So, it will probably take longer than you would expect There was no change to the number of participants who agreed or strongly agreed that they understood how and why infections spread (98%, n ¼ 47 before/after training).Only the AR group scored slightly higher after the training.
Lastly, 81% (n ¼ 39) agreed or strongly agreed that they could improve their hand hygiene technique compared to 85% (n ¼ 41) after the training.Perceived improvement was highest in the AR group.

Hand hygiene skill
Overall, average skill scores increased from 4.77 (of 11) (61.80) to 7.23 (62.31) after the training.We noticed improvement in 79% (n ¼ 38) of participants.These were highest in the AR group (AR: 4.58-8.32;immersive VR; 4.90-6.71;non-immersive VR: 4.89-6.71).Of the 10 participants who did not improve, 4 did not change their scores, and 6 had lower scores after the training.Six of the non-improvers were in the immersive VR group, 3 were in the non-immersive VR group, and 1 was in the AR group.
Figure 6 shows the number of learners achieving the skills for each intervention group.Everyone used enough sanitizer to cover their hands before and after the training.Major improvements were seen in the fingertips and the thumbs poses, whereby an additional 20 and 17 care staff, respectively, performed these poses after the training (fingertips: 4 before, 24 after; thumbs: 11 before, 28 after).The specific length of the technique improved in 77% (n ¼ 37), increasing from 22 (69.40) to 26 seconds (610.41) on an average.

Mechanisms
Perceived task realism (relevance to work), interactive learning, feedback and reminding, and repeated practice were all considered important mechanisms for learning by participants, which the training triggered (see Table 4).
Ninety-six percent agreed or strongly agreed that the task in the learning tool was relevant to their work.However, some of the content (not related to hand hygiene) in the immersive VR and non-immersive VR training was less relevant (eg, hemoglobin device and nursing tasks).
Ninety-four percent agreed or strongly agreed that the learning tool was more interactive than previous training with the immersive VR training perceived most interactive and the non-immersive VR group least interactive.Participants One hundred percent of participants agreed or strongly agreed that the learning tool gave them feedback and reminded them about good practice.However, scores were lower in the non-immersive VR group.This mechanism also facilitated reflection on poses that are easily forgotten or on usual technique.
In the interviews, everyone stated that repeated practice is essential to correcting hand hygiene technique, helping those who have forgotten and ensuring that skills are up-to-date.Skills can then become "routine" or subconscious "muscle memory."The AR app recorded how long it took to perform the 6 poses.Ninety-five percent improved at the second attempt but only 47% improved at the third attempt.For the immersive VR and non-immersive version, compliance scores were recorded for each patient.Fifty-five percent improved at the second patient and 86% improved at the third with 83% scoring 100%.

Discussion
This study has explored VR/AR smartphone apps for upskilling care home workers in hand hygiene.The immersive VR and AR training had good usability and acceptability, being perceived as more interactive than previous training and resulting in high learner satisfaction with support from 95% of learners.However, the non-immersive VR training was less acceptable, less interactive, and had borderline poor usability.Regarding efficacy, after the training, there were minimal improvements to knowledge.However, hand hygiene technique improved (especially for the AR group).Importantly, AR/ VR triggered mechanisms including repeated practice, task realism, feedback and reminding, and interactivity, which are essential for learning.Lastly, we noted implementation considerations, including managerial support, the space available, support for potential non-users, screen sizes, lagging due to poor Internet connections, and fitting the headset.
Leadership and collaboration are crucial for facilitating the implementation of AR/VR training, [42][43][44][45] with care home managers and senior carers playing an important role in training engagement and choosing which technology is best suited to their setting. 25However, it is vital that the considerations regarding using the VR headset are clearly outlined to managers as it requires more infrastructure, funding, and technical support.It is especially important that the headset is fitted properly and Internet issues are resolved for learner comfort, as a poorly fitting headset, flickering screen, or lagging/ latency may induce cybersickness. 46These additional requirements may encourage managers to use the AR training instead.
Ideally, short refresher sessions for hand hygiene training should be introduced, which can be repeated for new and temporary staff.Self-directed training may be appropriate given that staff may already own devices that can be used; with smartphone penetration estimated to reach 95% of the UK population by 2025. 47Short sessions that can be repeated also account for the transient nature of the care staff workforce as staff may work in multiple care homes or on shifts where training and monitoring are unavailable.In 2021 and 2022, 24% of England's adult social care workforce was employed on zero-hours contracts, and employers struggled to recruit and retain their staff, resulting in competition and movement. 41ur results regarding improved technique and learner satisfaction were consistent with research on AR/VR hand hygiene training for other learners.In one study, the SureWash AR system was implemented in an Irish hospital with poster reminders, an adenosine triphosphate monitoring system, auditing through observation and verbal feedback. 18Sure-Wash demonstrated the correct technique and assisted with auditing.Overall, this approach significantly improved technique and compliance.SureWash alone also led to compliance Table 4. Supporting data and interview quotes for the mechanisms triggered by the hand hygiene training.

Mechanisms
Quantitative results

Supporting quotes from interviews
Perceived task realism (relevance to work)

Immersive and non-immersive VR
• Everyone agreed that the hand hygiene content was relevant, but most were unfamiliar with the hemoglobin device and nursing tasks (eg, wound care).• Participants identified differences to their work like that they always wear gloves when touching a resident, wear spitguards, and that residents are mobile.
• Some suggested scenarios like food preparation, care, and managing bed sores better reflect their work.

AR
• Participants explained that they physically practiced the poses rather than looking at images or videos.• They enjoyed the game component, feedback and said it felt more human and thorough, as it was not rushed and was conducted one-to-one inperson, unlike training on Zoom.
• One participant explained that while it was fun, learning at a sink or using the UV machine was more interactive.

Immersive VR
• Participants described it as interesting, novel, modern, and fun.They felt like they were actually caring for patients.• Everyone preferred it over reading, listening, and watching.

Non-immersive VR
• Some likened it to Zoom, online modules, or videos.• Others said it was more interactive and fun, which made it better than listening to others describe hand hygiene. in 81% of 47 university students after an average of 24 training sessions. 15Another study compared VR training to lectures in German hospitals, finding that 69% of the 81 healthcare workers preferred VR. 17 However, it did not result in higher compliance with disinfectant use.While AR/VR may deliver effective hand hygiene training, it remains unclear whether this translates to longer term impacts and how many sessions are required for optimal learning.• The AR app also gave feedback on the length and challenging poses.• Participants valued the final report card in the VR training and that they could not continue until they had followed the correct procedure.

Immersive and non-immersive VR
• Compliance scores were recorded for each patient cared for in the game.The non-immersive VR training was less usable, acceptable, and had lower efficacy compared to the other training modalities.It is possible that digital literacy skills influenced the usability of the technologies, while the novelty and motivation of the AR and immersive VR technologies may have influenced the outcomes.For example, a previous RCT with 66 participants learning neuroanatomy found that while using a novel immersive VR system was just as effective as educational textbooks, it was significantly more engaging, enjoyable, useful, and motivating for learners. 48The ARCS Model of Motivational Design further explains the importance of motivation for learners, stating that this consists of attention, relevance, confidence, and satisfaction. 49Specifically, learners must be interested in the topic, understand the relevance to their personal goals/motives, and feel confident in their ability to learn.The construct of "relevance" aligns with our mechanism of "perceived task realism (relevance to work)" which is important for hand hygiene training in care homes. 25It may be possible that the novel technologies were more engaging and interactive, thus facilitating the other constructs of motivation and resulting in better outcomes, compared to the computer-based version.
The lower usability and efficacy of the non-immersive VR training may result in inequitable training opportunities if only the immersive VR training was rolled out, as 28% of the VR participants could not use the headset for health reasons.However, offering the AR training as a non-immersive alternative (rather than the non-immersive VR training) would ensure that staff who cannot use VR headsets could still engage in effective and acceptable training.Differences in the training also helped explain reflection as an essential part of the feedback and reminding mechanism and potentially the reason for adverse outcomes.Participants in the immersive VR group had less confidence in their technique, and those in the AR group were more likely to perceive a need for improving their technique.As suggested in the qualitative data, learners may have realized what they were doing wrong by reflecting on their usual practice during the training (education and modeling), resulting in less confidence.
Our findings clearly indicated that future development of VR/AR training should focus on applying the WHO 5 Moments for Hand Hygiene to care homes and re-educating staff on wearing gloves.The 5 Moments only partially translate to care homes due to environmental differences.Teesing et al. 50explained that hand hygiene should be performed after touching a patient's surroundings, but this is difficult as residents are mobile, so the entire home may be their surroundings.It was therefore unsurprising that none of the learners could recall all 5 Moments and instead named other care tasks.Regarding gloves, only 52% correctly stated that hand hygiene should be practiced before and after wearing gloves, although most learners wear them to touch residents.This may be because healthcare workers sometimes practice hand hygiene only to protect themselves. 3,51Literature also explains that gloves may reduce hand hygiene adherence 52 and are often worn incorrectly (eg, layered or sanitized) 53 by care and nursing home staff.In a study of 20 nursing homes in Norway, compliance dropped by 30.8% when wearing gloves and was not performed correctly in 64.7% of instances. 51Similarly, gloves were appropriately used in just 16.8% of observations by personal care staff in Hong Kong, compared to 54.7% by professional staff (primarily nurses). 52

Implications
Care home managers should consider rolling out AR/VR hand hygiene training, supported by workforce training policies.Further experimental research should determine whether AR/VR training is more effective than previous training, any long-term impacts, and how many sessions are required for optimal learning.As the non-immersive VR training was unacceptable, usability improvements are needed, such as simplifying navigation.Other technical improvements include amending the VR scenarios to be care home related and completing the AR training on an iPad/tablet to minimize screen size issues.

Limitations and strengths
There were only 3 repetitions within the training, so our ability to determine how many repetitions are required to support and sustain learning is limited.Another potential limitation is selection bias, as participants may have been more comfortable with new technologies or digitally literate.Lastly, we did not formally assess cost-effectiveness, which is an important factor to consider regarding the feasibility and future implementation of the training.
Methods, analyst, and data triangulation were a strength.The questionnaires, interviews, and observations explored many of the same phenomena.Two researchers individually coded the qualitative data which were collected from different care homes and learners.The representativeness of our study sample also strengthens the generalizability of our findings to other care homes in the United Kingdom.

Conclusion
AR and immersive VR smartphone apps are feasible, usable, and acceptable technologies for delivering more engaging and potentially effective hand hygiene training in care homes.Future work that explores whether hand hygiene training delivered via AR/VR is more effective than previous training is warranted.There is also a need to improve the nonimmersive VR and ensure roll-out of the training includes the AR training as the non-immersive alternative, to provide equitable training opportunities.

Figure 2 .
Figure 2. Images showing the seated position for the immersive VR training (top left and middle), non-immersive VR set-up (top right), and screenshots from the game 13 (middle and bottom row).

Figure 4 .
Figure 4. Flowchart presenting care home and participant recruitment process.
, you know, to complete the task.(Participant 30, AR training) Supporting implementation through a training exercise Give them practice with it. . .instead of bringing them in for the initial training, just say okay, here's like five, ten minutes to have a play around, get the feel for it and then we'll begin.(Participant 26, immersive VR training) confident in their hand hygiene technique after the training, compared to 85% (n ¼ 41) before.The AR and nonimmersive VR groups reported the highest scores after the training.Unlike the AR group which reported the most improvement in confidence after the training, the immersive VR group reported less confidence.In fact, their confidence decreased from baseline.The majority (79%, n ¼ 38) agreed or strongly agreed that their hand hygiene practice is up-to-date and aligns with best practice, increasing to 94% (n ¼ 45) after the training.Mean scores after the AR and non-immersive VR training improved, while the scores after the immersive VR training decreased slightly.

Figure 5 .
Figure 5. Clustered bar graph showing mean confidence scores at baseline and after the training.

Figure 6 .
Figure 6.Radar charts showing how many of the learners met the hand hygiene technique criterion before and after the training, per intervention group.

Table 1 .
Outcomes, mechanisms, and details of measures used.

Table 1 .
(continued) a Outcomes.b Presented in the Supplementary Material as the data do not pertain to the whole participant sample.c Mechanisms identified from our previous work.

Table 2 .
Summary of the care homes and participants.

Table 3 .
Supporting quotes for individual-level implementation considerations for AR and VR hand hygiene training.