Improving Healthcare Outcomes with Simulation

Ashley K:

Welcome to the latest episode of Resoundingly Human, the podcast brought to you by Informs, the leading association for professionals and students who apply science, math, technology, and analytics to make smarter decisions for a better world. I’m your host Ashley K, and thank you for joining me to explore how Informs members are saving lives, saving money, and solving problems.

MUSIC:

(Upbeat instrumental music)

Ashley K:

I’m so pleased to introduce Tom Stevenson, director of services and strategic partnerships at Simul8 Corporation, who joins me today to take a look at how simulation is being leveraged to help improve healthcare outcomes in a wide variety of applications.

Tom, thank you so much for joining me.

Tom Stephenson:

And thanks for having me.

Ashley K:

So I’d love to kick things off by having you share a little background on Simul8 and your role there.

Tom Stephenson:

Absolutely. So Simul8 is a software company, and what we do is we build a virtual replica of the real system. And when you have that virtual replica in software, it gives you a platform to test a lot of different changes. And I’ve been with Simul8 for 12 years now, just over 12, and I’ve worked in various industries where this applies, but my main focus for at least the first eight years of my career was healthcare. So looking at how you can build virtual replicas of hospitals or healthcare systems, how you can test the changes that you might want to make to those healthcare systems and seeing what the impact’s going to be on for the patient and for the provider.

Ashley K:

In preparing for our conversation today, I’ve had the opportunity to read a number of case studies from Simul8 on a fascinating range of simulation applications within the healthcare industry. But before we dive into these, I’d love to take a broader look at how simulation is helping to improve healthcare outcomes.

Tom Stephenson:

So I think there’s three broad areas where we see a lot of simulation work. One is directly trying to improve how providers or hospitals mainly can treat patients. We also see a lot of need for population health type work. So I think across the world we’ve got populations that are aging, there’s perhaps obesity, there’s perhaps more people living with or managing long-term conditions. There’s a lot of treatment out there that can keep people healthy, but every population’s a bit different and they’re all changing, and that creates a challenge for healthcare providers.

So we do a lot of work in terms of projecting forward. “What is my situation going to be like. How’s it going to change, and how can I make sure I provide the best service as it does change?” Some of the work that we’ll do as well will be with pharmaceutical companies as well. Maybe there’s a new medical device or a new drug that’s going to change a patient pathway, and we want to have a way of analyzing the impact of that change. So yeah, all of them have their unique challenges, but they’re all really interesting in terms of building simulations for them.

Ashley K:

Tom, I’d love to have you share a little background on some of the earliest applications that you’ve seen and how these have grown over the years in the healthcare setting.

Tom Stephenson:

Yeah, so I think what’s interesting for me is that the problems that I’ve been solving over the past 12 years haven’t necessarily changed that much. We still have patients who are waiting a long time for care. We still have patients who are maybe not receiving care in the correct place because it’s maybe not available to them. And I think that’s true across the world, really. So it’s not a sign that the healthcare systems are failing because we’re still solving the same types of problem. It’s just that populations are changing, it’s getting harder and harder to keep that population healthy and treat them.

So I think early on, just worked with a lot of hospitals, and emergency departments always need help. You need to make sure when people need that urgent care, you can see them in a reasonable amount of time. But more and more there’s people preparing for surgical procedures and that type of thing. And I think especially now with the impact of COVID. Waiting lists, we’re still catching up from a lot of that. So some of them have just got harder and harder to manage because of population change as well as this big global pandemic and what we’ve all had to deal with.

Ashley K:

Now you work with organizations in the UK and the US. Could you share some of your own personal experiences and how providers differ in terms of their readiness and application of simulation and digital twins?

Tom Stephenson:

First off to say, we do separate simulations and digital twins. So what we mean when we talk about “digital twins” is being able to have a recurring need to use a simulation. So it might need to work with live data, recreate the live situation and run forward for the next eight hours or the next 12 hours or something like that. And to do that, that means we need to have the information to hand about how our system looks right now and what people are in that environment. And I think hospitals are definitely getting more and more equipped to do that. So we see electronic health records being utilized and leveraged everywhere now. So I think the technology that we’ve got in the healthcare setting is definitely improving. And also I think the people are coming out with a lot of analytical skills. Now, some of them I see coming from different sectors. So it might be you work with a lot of people who have been in manufacturing, let’s say, and they can see how the same approach to removing bottlenecks can also apply in healthcare.

I think around the world as well, people are solving the same problems. It might just be that there’s certain things that are a bit more important to show in a simulation. So in the UK it might be less important to have revenue as a target because it’s a government-run healthcare system largely. Whereas in the US that might be a more important metric to show. In either case, we’re trying to make sure we can get the best result for the patient. But it’s just maybe we want to highlight different things a little more from the simulation.

Ashley K:

I think that’s just so interesting. Before we get into the specifics, it might be a good opportunity for you to explain why simulation works so well in healthcare and what are the opportunities that you want to make providers aware of?

Tom Stephenson:

Yeah, okay. I think there’s a few reasons why Simul8 is a good tool for healthcare. So there’s a lot of tools that are good for healthcare, but I think one thing is that healthcare systems are generally complex and they are related. So an example, my son and my nephew playing a couple of nights ago. My nephew had a little fall, had a cut on his cheek, nothing too bad, but it was like seven P.M. at night. And we phoned the non-urgent advice line and they kind of said, “There’s no primary care available, which might be a good place to go. We can’t go to a minor injury unit because he’s too young. He’s only three. So you’re just going to have to go to the emergency room.”

So we go to the emergency room almost feeling a bit bad, like are we taking the place of somebody who really does need this urgent care? But when you get there, you see that there’s lots and lots of other parents in the same situation, and there’s young toddlers there who have had a mild allergy, but there’s never with the kind of service available at that time. So you find more people come to the emergency department. And it’s that type of link between different healthcare providers that can be important to include. You see it in a lot of other places. If a lab is only open at certain hours of the day or they only do ECGs at a certain time of the day, patients aren’t going to be able to move on in their journey until that’s happened. If a pharmacy isn’t available and can’t get the drugs that you need to get out of bed, you’re not going to be able to move forward.

So all of those relationships are what Simul8 copes with really well. Because it’s a connection, parts that are connected in a flow. So if you change one part, it’s going to have a knock on impact elsewhere in real life. And that’s something that you’d also see in a simulation. And that complexity can be quite hard to figure out, especially if you’re using static Excel-based tools. A similar thing is the variation that we see in healthcare. So you could have 10 people and nine of them might have a length of stay that’s four hours, five hours, four-and-a-half hours, three hours. But then you suddenly have one that’s 72 hours or a hundred hours in there. And that one person is really important to kind of factor in, because they’re the people who are going to create problems for you. And you might not get them all the time, but you are going to sometimes get them.

So what a simulation does is looks at the variation that exists and feeds that exacted variation that happens in reality into the simulation. You’ll see the variation happening, not in terms of just length of stay, but also when people arrive. It’s going to be more common at certain times of the week. There’ll be variation in terms of weekends, you might have less discharges, that type of rule. So all of that, again, is something that you can factor into a simulation quite easily.

And I guess those two things, what’s key in terms of including them is that it’s going to allow you to get a model that is really similar to the real-life scenario. And that means if you want to play around with different parameters and you see an impact you know that it’s going to be the great impact that you’d see in real life. So I think they’re really important things, but also it’s very visual. You can see your simulation. You can test a lot of things and you can test them really quickly as well. So that’s key because we see a lot of people with an alternative of running a pilot in healthcare, but that could be costly. It could take a lot of time. Whereas if you can have a virtual representation that you trust matches your reality, it means you can just test a lot of these what-ifs quite quickly.

Ashley K:

I think it’s really incredible the number of factors that simulation has to take into account. And I think the example you share with your nephew is very relatable and I’m glad that he’s all right.

Tom Stephenson:

Okay. It’s okay. He has actually pulled his stitches out already, so not the best outcome one day in. But yeah, so he has been back since today. But yeah, he’s okay.

Ashley K:

It’s a good story to tell, too.

Tom Stephenson:

I know. But it does just show that everything, any service you receive is heavily dependent on what’s happening elsewhere, what other services are available. And that’s not just within a geographical area, that could be within a hospital or anything else as well.

Ashley K:

Well, Tom, now I’d love to dive into some of the specific examples of the work that Simul8 has done. I’d love to take a look at how simulation helped to improve cardiac surgery flow at Guy’s and St. Thomas’ NHS Foundation Trust. How is simulation leveraged there?

Tom Stephenson:

Yeah, absolutely. So cardiac surgery is quite hard to manage because you’ve got these different flows of people. So you’ll have people who are scheduled. They’re booked in for their surgery that’s going to happen at a certain amount, at a specific time and day. But on top of that, you’re going to have emergencies that you need to deal with right away. You imagine heart attacks and that kind of thing where people need surgery. So it’s quite a tricky balance to manage that, because you need to make sure there’s enough space for those people who are going to need it when they have an emergency. But you also don’t want there to be too much waste in the system either. So you don’t want to leave too many operating rooms and recovery beds free, because then that’s wasteful. But if you don’t need enough, don’t leave enough, then you’ve got the potentially opposite problem.

So it was how to find that balance that they wanted to do, especially when, as we say, the population of London, which is where this hospital is, was changing. So what we could do is try lots and lots of different tests to understand what might work well. And the obvious ones would be, can we increase capacity? Can we have more beds? But then that’s going to cost more money and it might not necessarily be feasible. But if we can create a business case, we’ll maybe do that.

And some of the more interesting ones was more around ring-fencing beds. So if you have cardiac surgery, you will want to recover in a particular type of bed. But that bed base is shared. So some people from other areas of the hospital might be occupying those beds. And if you come in and you’ve got a surgery that’s scheduled, you could have to be canceled because an emergency patient needs the bed, needs the operating room at that time. But also it could just be that all your recovery beds are full as well. So, you can’t start a surgery if there’s no recovery bed. So all of these kind of blinked elements were things that we had to take into account. And what they wanted to do to kind of ensure that their service would work was look at ring-fencing those recovery beds, and ring-fencing some of the surgical rooms, to just make sure that the flow could deliver what it needed for those patients.

Ashley K:

And ultimately, can you share? What was the impact of these simulation applications?

Tom Stephenson:

Yeah, so in this case, we were looking at making sure we could reduce the amount of cancellations that would happen, and also making sure that we can meet our utilization targets. So we want to hit a certain utilization. If we try and hit too high a utilization, we’re going to have a lot of cancellations. It was just about finding that balance really. And those were the main metrics that we looked at. We looked at how long patients need to wait and how utilized our rooms would be and how many cancellations there were. And that meant that we could configure a service that would give us a nice balance between those metrics.

Ashley K:

All right, so now the final application we’ll discuss today, and that’s with IVF. This is an application that I will share I’m personally excited to explore with you, as it has benefited both my family and a number of my friends. This is obviously a very emotional and complex area of medicine. How have simulation been leveraged here to improve both the patient experience and outcomes also at Guy’s and St. Thomas’ NHS Foundation Trust within their assisted conception unit?

Tom Stephenson:

So first of all, I’m pleased to hear that it’s had a positive impact for you. I think there’s a lot of people in that situation, and that probably shows in the growing demand that they’ve had in London and all around the UK. I think that their demand for IVF is growing and growing. With that increased demand, they needed to make sure that they could offer the service that was required. And what I found out about IVF that it’s quite interesting is just how precise the timings need to be. If you are out with the window where it’s all going to work from the kind of biological point of view, you might as well not do it. The chances of it actually being a success are so much lower. So that was key. It was making sure that those time windows that were crucial to hit, they could do that. Especially when they have increased demand coming in and maybe needed to grow.

But in the same way as the cardiac surgery. And the simple solution to that would be, let’s just hire lots and lots of people, have lots and lots of rooms, make sure that even with all this variation, we’re going to have enough spare capacity that that’s never a problem. But then, that’s going to be very wasteful. So again, it’s the perfect point to give the patients what they need from this service.

And in the UK, you can have a service on the NHS to start with. I think maybe you get one or two attempts on the NHS, but there was also a private kind of flow as well of people who were paying for the service. I mean, they’re all treated in the same way. But yeah, I mean just managing all of that flow, all of those different people coming in. And not everybody’s flow was the same, as well. So we had to cope with that. Some people would have a sperm frozen and that type of a thing, or they might be… There’s a lot of nuances there that you don’t even consider unless you’re involved in the process. So we had to include all of that kind of variation to make sure that, again, we could trust the simulation and trust the changes that we were testing with it.

Ashley K:

And within these applications, what have been the outcomes for patients as well as medical professionals?

Tom Stephenson:

Yeah. I mean, I think you always want patients at the heart of what you do. So reduced waiting time is always a big win. In a hospital, it’s going to mean that you’ve got a better chance of getting the care that you need. There’s reduced chance of problems, complications, and that sort of thing. Or if you’re on a waiting list, it’s vital that you get the care at the time that you need it. So I think that’s really key from simulations, making sure that people can get the right care at the right time.

But then from the provider point of view as well, they also want that for their patients, but they also don’t want their staff to be so burnt out and so pressured that they can’t give care in the safe way that’s required as well. So there’s a lot of that kind of making sure we’ve got humans working alongside the project as well. Bringing that perspective and making sure that whatever the kind of optimized hospital is saying is actually going to work for staff and everybody else. But really it’s better outcomes that we’re looking for, better outcomes for everybody. And the primary goal is always how does that impact the patient? But secondarily, there might be things like, how have we reduced our cost without impacting the patient? Or, have we freed up the time that nurses have to actually care for patients rather than doing admin tasks. And all that sort of thing, really.

Ashley K:

Well, I’m a very proud and very grateful auntie thanks to IVF. So I was really looking forward to talking with you about that today. So thank you.

So we’ve talked a lot today about different applications for simulation that are currently in use. Anything coming down the pipeline that you’d like to share? Maybe something new you’re really excited about?

Tom Stephenson:

Yeah. I mean, I think the thing that I want to come back to is the situation that hospitals are in, where they have got data. They have got electronic medical records. They have got data that refreshes in, basically, real time. That gives us such a good opportunity to help with decision-making in real time as well. Because often if you make the wrong decision right now, you’re going to see in 12 hours time the situation’s going to be much worse. But what we can do, because we’re able to pause the live situation that the data is telling us and run forward is we can almost have that crystal ball that say, “This is what a situation’s going to be like in 12 hours if we do what we’d normally do. Let’s try some things at the start of the shift and let’s see what impact that has.”

It’s not just hospitals that are able to do this now. We are looking at cancer waiting lists as well, where we can see, “These are the people who’ve had their initial test that they’re waiting to get on treatment. This is how many appointments we need to do to move them towards that safe treatment target that we have.” So in all areas, I think just having that wealth of data available. But a lot of people maybe have the data available but aren’t leveraging it. They’re not able to visualize it. They’re not able to make decisions on it. I think that’s where simulation can really come into its own, and especially over the next couple of years, I can see it being more embedded in that regard.

Ashley K:

Tom, thank you again so much for joining me. It’s really been such a pleasure to feature these incredible applications of simulation in healthcare and just to talk with you today.

Tom Stephenson:

Yeah. Well, thanks very much for having me. And yeah, great talking to you, too.

Ashley K:

If you’d like to learn more about today’s episode and guest, visit resoundinglyhuman.com and check out our show notes. The podcast is also available for streaming and download on Amazon Music, Apple Podcasts, Google Podcasts, and Spotify. Wherever you listen, please be sure to leave a five-star review to help others find and enjoy the podcast. Until next time, I’m Ashley K, and this is Resoundingly Human.