Changing Realities - VR, AR and Old Practice
This post is a written version of a talk by Dr Andrew Weatherall at the College of Intensive Care Medicine Annual Scientific Meeting, held in fairly sunny and definitely cold Hobart, Australia in May 2025.
It is always an excellent thing to have a chance to think about things that are cool. Penguins. The classic Sesame Street pinball number count. This autonomous soft-legged inflatable robot. Wait, that last one might be terrifying.
And listening to a range of people describing the ways they have made cool things happen in their joint speaks to all the wondrous bits about actually meeting up with people. It is an excellent bit of conferences.
But my job here is to talk about really old things. Specifically the new worlds on offer with virtual and augmented reality (VR and AR) and the mission to bring them to our clinical space.
A few minutes on whole new worlds. Should be easy.
This must be the ticking clock
My Actual Job
To understand why we might be interested in VR and AR, maybe a moment reflecting on my actual job would be useful.
While I will let the casual reader insert an easy joke or two right here about what anaesthetists do ….
[No, I’ll wait…]
[Feels like you must have one more ….]
[OK, hope you feel great, let’s get back to it….]
… on a good day my job is about a lot more than pharmacology, analgesic plans and exercising a few technical skills.
On most days I get to build worlds with kids to hopefully make their experience at the start of an anaesthetic great. It might be an underwater expedition, or we need to use their dragon breath to cook some food with that mask. Or perhaps together we’ll show their parent exactly how you go off to space, not with a rocket but with an amazing balloon.
And we do this because we have that deep sense, with a little bit to back it up, that an experience changes how you feel about the world afterwards.
This is a very old concept. It is older even than the great classics that showed us that you can take the sort of nasty old rich guy who would walk across the street to kick a dog and shave the eyebrows off the kid with them and totally transform their world view. All you need is one night, three ghosts and a weird trip and that individual can be transformed into the sort of person who would help feed the dog and help the kid glue their eyebrows back on.
Or maybe the story goes that they kicked the kid and shaved the eyebrows off the dog.
Anyway, enough about A Muppet Christmas Carol.
Our efforts are directed at making a potentially tough experience kind of great, very much with the firm view that feeds into everything that follows.
So when VR and AR loom into view offering an entirely different way to create great experiences, we’re interested.
Particularly when wearing it makes you look so cool.
This human really reinvigorated the world of VR. And now he has a company that focuses on military technology and autonomous drones. Hard to say where the photo fits into this narrative arc.
But what we really need of course is evidence that it might just be useful.
Before, During, After and Forever
Before
A really core part of business for an anaesthetist is the perioperative experience. So it is just as well there is a bit of evidence that patients given the chance to have an immersive VR experience to familiarise them with the trip through the big doors display lower levels of anxiety at the time of induction.
Ryu et al showed this in a (not particularly large) RCT in 2017 recruiting children between the ages of 4 and 10. With 34 patients in the intervention arm and 35 getting standard care, they gave the first group the chance to view the experience of Pororo the Penguin, in giant mascot form, as they went into the operating theatre. They then checked in on the children 30 minutes prior to anaesthesia. The children who had spent time with Pororo had a median score on the modified Yale Preoperative Anxiety Scale of 31.7 while those who were bereft of penguin time scored 51.7. This is on a scale where 100 is peak anxiety and no anxiety is 22.33. Because 'maths’ I guess.
When it came to induction the children in the intervention arm also displayed greater compliance with induction (and a far higher number of ‘perfect’ compliance scores) and lower scores on the procedural behaviour rating scale. This again suggests a much more relaxed child.
Oh, the VR video was 4 minutes long.
That is a pretty good result in 4 minutes.
The group has more recently shown that the optimal timing of this VR intervention is in the hour prior to surgery, with little impact if the children see it a few days prior. Recency makes a difference.
Another group published work on the value of an AR experience in 2024 which also suggested lower anxiety scores on m-YPAS at induction, along with very high satisfaction scores. These findings are a little harder to interpret because while 37 children were analysed in the AR intervention group, this was after 9 children didn’t actually complete the experience because they removed the headset and 10 had their experience interrupted because their surgery happened early. They weren’t analysed. (4 also had a missed m-YPAS.)
But these are not the only studies out there and on balance it appears that this is an option that makes a difference.
During and After
The critical bit for the work of a paediatric anaesthetist might be whether VR/AR offer something novel for procedural distraction. This would be ideal given there are plenty of patients who find all types of procedures, whether it’s induction, vascular access, or dressings to be a bit of an ordeal. There are also a range of procedures where we have a tendency to default to general anaesthesia, things like lumbar punctures and bone marrow aspirates, where there are probably options to avoid that entirely.
And happily the evidence around procedural distraction actually extends back decades to when VR systems cost tens of thousands of dollars and were even less cool to wear on your head than today.
To point to just a couple of examples, Jeffs et al demonstrated in a cohort of 30 patients across the ages of 10 to 17 that those provided with VR distraction with the classic Snow World experience during burns dressings reported significantly less procedural pain than those watching a movie (23.7 mm on a standard Visual Analogue Scale). A publication by Hoffman et al from 2015 noted that burns patients utilising highly immersive VR reported a reduction in pain scores during procedures of 35-50% and that the better the immersion, the better the result.
These relationships hold for severe pain in small studies, even though it had been theorized that at some critical threshold the distraction would be overwhelmed by the intensity of the stimulus and distraction would become rapidly ineffective.
This observation of a reduction in pain intensity comes with some evidence of a biological correlate. As covered in that review article, imaging with functional MRI during pain testing showed again a reported reduction in pain intensity while using VR. The associated imaging showed a 50% or greater reduction in intensity of signal in five regions of the brain examined. This included the anterior cingulate cortex, insula, thalamus, and the primary and secondary somatosensory cortices.
And while burns therapy has been studied quite extensively, it is not the only relevant area. There have been explorations of its usefulness during vascular access procedures. Interestingly in work by Caruso et al looking at use of VR at the time of cannulation the reductions in reported pain intensity and anxiety did not reach significance and yet the satisfaction ratings provided by patients, clinicians and carers were extremely positive.
Positive reviews all round.
As a clinician, if the patient leaves the procedure feeling it went great and up for using that headset thing again, even though the pain scores reported are about the same, that is probably still a win. It certainly seems like they’ll carry that experience with them. That impact in the ‘after’ bit of the procedure probably has some value, though the work to pin down how significant it is doesn’t seem easy to find.
But perhaps the real demonstration of usefulness relates to whether these modalities can help with the trickiest of beasts. Anxiety.
Well not really forever. That might be a bit much.
There are examples of some pretty profound work in the area of phobias though. For example way back in 2002, Garcia-Palacios et al came up with a test that seems like it would be the stiffest of all challenges.
Across an average of 4 treatment sessions using VR, 83% of subjects went from ‘actually I think I won’t go into that room with the tarantula in a glass case on the other side thanks’ to ‘yes of course I will pick up this virtual spider wearing a haptic glove that makes it feel like virtual spider seem like it is moving in my hand’. Kind of disgusting, but powerful.
Freeman et al showed similarly impressive effects in dealing with fear of heights. Participants with at least a moderate fear of heights had either VR-supported care or no intervention. With an average of a bit over sessions and an average total treatment of 124.43 minutes, the reduction in anxiety compared to those with no treatment utilising the relevant scale was 68%.
Put another way, 51% of participants had a reduction in anxiety scores of 75% or more, while 78% had a reduction of 50% or more. Those changes were largely still present 4 weeks later. Big impacts with relatively little time.
But maybe that’s not enough to suggest there is something profound on offer.
It turns out that there is a little evidence that by including the use of embodiment, a VR experience where the user rapidly adopts an avatar in virtual space as something they control, you might be able to influence some broader thoughts.
Peck et al developed an experience where the user entered a virtual world and adopted control of an avatar projected into that space in front of them. They also did baseline testing of implicit racial bias and for users spending their 12 minutes in this space with an avatar with a different skin colour, post-intervention testing suggested a decrease in implicit racial bias. It was still significant a week later on repeat testing.
That might be something.
So if it works then, what was our plan?
Getting on with it
At this point it’s pretty obvious. We decided we had to bring commercially available VR and AR to the periprocedural space. We figured we’d lazily create a preoperative preparation experience. And then obviously we figured we’d develop a platform that empowered kids to co-design the experiences they wanted to see.
Easy.
The most pertinent of these was the work in the procedural space. So we developed a library of already available experiences. These experiences had to fit a few key characteristics.
1. Direction of Action
You can do anything in a virtual world. But procedures are hard if the patient wants to climb over the back end of the bed because something cool is happening. So things have to mostly happen in front of them.
2. Procedural Needs
A VR or AR experience that helps you work on your light saber skills is probably not ideal for distraction during cannulation. So the experience has to allow you to still physically do the procedure as a clinician.
3. Include the clinician
To make the most of this tech we felt it was really important for the clinician to know what the patient was seeing so they could be part of the interaction and know where the patient was up to. So that required casting the experience to an external device. Hmmm.
4. Empower the patient
It seemed sensible to give the child some choice in which experience would do the job rather than telling them how it is.
So in that context let’s meet an adolescent, PB, approaching the end of their leukaemia treatment. They had reached the point where the idea of an anaesthetic itself was a source of days of anxiety. Because no matter what we tried, there was nausea, vomiting and general sense of awfulness for days afterwards.
So we were asked if VR distraction might be an option. And after a meeting, an introduction to the system and a chance to choose something we all thought it was an option.
And two days later this patient had a lumbar puncture, bone marrow aspirate and insertion of central venous access with only local anaesthetic as another agent. The whole procedure took a bit over an hour and at the end the best bit was the sense of empowerment so evident in PB.
And it was good enough that they chose the same for their next outpatient procedure. But this time they were so confident they didn’t fast, had everything done, and chomped on an apple as they walked out a few minutes later.
Giving Kids the Space to Build a World
At the same time we managed to do that little bit of research on how kids might design in VR. Teaming up with a crew from the School of Design at The University of Sydney we managed to conduct 16 workshops and learn a little about how children might go about designing their own VR or AR experiences.
We met explorers who voyaged over a whole world before choosing to start adding to it. We met artisans who carefully looked at every available asset before embarking on, and sharing in real time, a narrative they were creating. And we met planners who mostly went quiet as they started with an entirely blank canvas and got going, only stopping to clue us in at the end.
At which point it was all fireworks and success.
Just one of the worlds built in those workshops
And you might be wanting a few specifics on making choices to make this happen. Which I would be extremely happy to share.
But the thing is …
This is the point in the story where things get way more satisfying than me just saying ‘we did great’.
This is the bit where I have to mention an inconvenient truth: we pretty much failed.
If you go to our joint this week it won’t be at all an easy thing to make distraction with one of these new realities happen.
Because it doesn’t matter if the theory behind a change is good. It only matters if the change happens. And it has only really happened when it is so embedded it has disappeared from view.
So the question becomes who do I turn to if I want to introduce a change?
The answer will shock you.
Those Nerds Again
Look I just wanted to try out that clickbaity last line.
But it is a little shocking that the actual people who are experts at changing things in a way that matters are the same clinicians who I thought were so incredibly boring at university that leather patches on their elbows would have felt a bit too spicy for their liking.
Fast forward a decade or two and they have proven through their heroic efforts during a lazy old pandemic that actually they are the coolest even though their intuitive understanding of the word ‘stat’ is different to mine.
And it should be obvious they are really good at introducing an intervention to the point that it becomes just how things are done. There is even a concept in the literature for this: normalisation process theory. That’s the fancier way of saying make it disappear from view and you can read more about it here and here and here.
It is most likely this will actually feel pretty intuitive but an extreme simplification suggests there are 4 key things to do:
1. Make it make sense
The intervention needs to make sense to those participating. So from a practical point of view you need to be able to describe your change in a distinct and easy way that is clearly about one thing.
This was just one area of failure in trying to bring VR/AR into routine practice. On reflection we were so enthused by the endless possibility we probably didn’t give clinicians something relevant to them in a way that made them grab hold. First lesson.
2. Use engagement to create collective action
In the text this is referred to as ‘cognitive participation’. Basically you can’t just say why the change should happen. That version of why has to be relevant to the clinician. So it has to be a bit specific. Engagement is also helped by having trusted messengers far and wide who model the change.
Another fail here as only a handful of folks were ever fully in the mix. This led to ups and downs in availability and delivery. We just never had critical mass.
3. Make action the easy choice
This is critical and something we have done better with subsequent projects. Ideally the change needs to actually be something that is the easier choice so it just kind of makes sense. In a general sense that means removing the barriers to doing the change and facilitating the choice.
In the context of the VR/AR work this one is thorny. And it’s actually a design problem.
Even though if we get through a procedure only with distraction the episode of care itself is quicker, for the clinician it still has more friction in delivery than its very elegant competition – the anaesthetic machine.
Stacked up against a machine that can have a patient asleep in under a minute, a process that requires switching things on, hooking up to Wifi, onboarding, drawing boundaries and positioning is just clumsy. This one is going to be tricky.
4. Give good feedback
This one is straightforward. Give people lots of feedback on what is happening and the wins along the way, along with the challenges you’re solving as you go. Ultimately we probably never had the critical mass of cases to do this well. Another lesson.
Not nailing the above steps led to the equivalent of this rapid unscheduled disassembly that is a very useful way of visually depicting our failure.
But then again …
Of course now that we have the wisdom of epidemic-slaying (wait, epidemic-ameliorating is probably a more acceptable term for those punters) public health physicians we kind of wanting to get back to it.
But this time we’ll start with a very snappy purpose that is directly relevant to clinicians. We’ll engage a broader group to work together. We’ve already upgraded some equipment so that the clumsy initiation of care is just that much easier. And we’ll share our successes.
And with all that onboard we hope to be sharing much better rocket rides with our patients soon enough.
Or we’ll kidnap Pororo from that Korean crew.
The References Plus A Little More
That first paper on preoperative preparation by Ryu et al is this one:
And their follow-up is this one:
The more recent AR one is right here:
The first of the burns papers mentioned is this one:
And the second one that includes the mention of fMRI is here:
That paper from the Stanford crew relating to vascular access is here:
There is also a good review on VR in procedural distraction here that I didn’t directly refer to. But it is worth the time:
That not at all exciting sounding paper on the arachnophobia treatment is this one:
The heights one can be found here:
That really interesting thing dealing with embodiment is right here:
That collaboration looking at kids’ design is open access and right here:
It is worth the time looking into normalisation process theory. You can do that here:
And here:
As another example of using this in change, we included some of these ideas in a paper all about sustainability in the anaesthetic context:
And did you get all the way down here? In that case maybe take a break and chill out to the music created by this entirely extraordinary, marble-based instrument.