Season 1, Episode 5
Patient expectations continue to outpace digital transformation efforts, which means it has never been more important to make healthcare easier and more personalized. This will require leveraging the wealth of data and new technologies to create more robust and consistent patient experiences, as well as taking lessons learned from other industries and working collaboratively.
Join Chris Hemphill as they guide us through conversations with some of the leaders who see this new focus on digital transformation as a watershed moment for healthcare to keep iterating and see that this change continues far into the future.
Subscribe to receive emails when new episodes are released.
Thank you for subscribing!
Please check your inbox to confirm.
VP, Applied AI & Growth
Chief Digital Officer
Head of Industry, Healthcare
MF Probst Advisory
Chief Technology Officer
Director IT Applications
Chris Hemphill (00:02):
Consumer experiences, major disruptors in AI tech are shaping healthcare for years to come. On Hello Healthcare, we dive deep on these issues with leaders who are driving change. I’m Chris Hemphill, VP of applied AI at Actium Health. And we hope that these stories will help you to create or demand better future in healthcare.
Chris Hemphill (00:24):
Digital transformation, it’s a hot topic, especially in the healthcare, but what are we actually talking about? If you’re like me and you first saw the term digital transformation, you may have thought, buzzword alert. Digital is just ones and zeros. Even playing solitaire on Windows 3.1 in the nineties was digital. But think about the term Solitaire. Solitaire means solitary, confined to self, alone, siloed. That’s no good for a healthcare system that needs to act fast and response to complex consumer healthcare needs.
Chris Hemphill (00:57):
When we talk about digital transformation, we’re talking about demanding more of healthcare than just solitaire. It’s solitaire, for example, when we have vast data on encounters, but really communicate with patients about their healthcare needs. It’s solitaire when a consumer can’t get a clear answer on the next steps they need to engage for their heart condition or for their cancer.
Chris Hemphill (01:20):
Healthcare has been playing solitaire for a while now, not on Windows 3.1, mind you, but on powerful machines with multiple processor cores, or even distributed cloud hosted systems, still there’s change brewing.
Chris Hemphill (01:34):
A 2021 report by Accenture found that 81% of healthcare executives surveyed, say that the pace of digital transformation for their organization is accelerating. I wonder what caused that. The pandemic plays a huge role in this acceleration. Business models and technologies that were long debated or tabled, suddenly took center stage. We saw the pandemic drive enhanced online scheduling, the rise of tele-health, remote care, patient portals and digital messaging. The list goes on.
Chris Hemphill (02:06):
Crisis, though painful, can be a tremendous catalyst and accelerator for the change that we need. In this way, COVID is exposing the gaps and silos in traditional healthcare and forcing us to stop playing solitaire.
Chris Hemphill (02:21):
There are other factors too. There’s intense competitive pressure from retail and tech companies that are entering healthcare. And we covered this in episode three, titled, Will Tech and Retail Eat Healthcare?
Chris Hemphill (02:35):
So, we’re hearing about these pressures, competitive and pandemic that are driving digital transformation efforts. The $3 trillion question is whether it’s making a difference for patients? Well, we asked them, we asked about 1,200 patients, and it seems that their expectations and needs are still outpacing healthcare’s digital transformation efforts.
Chris Hemphill (03:00):
The biggest finding from the study was that half of patients expected more from their doctors after COVID-19. The biggest reason that they were losing confidence was infrequent communications. They’re expecting more active communication in general, however, only 45% were receiving general health information, and only 29% got scheduling information about preventative screenings. Even worse, only 21% were receiving information that was related to their chronic conditions. They’re hearing more from people that are looking to sell them jackets, leggings, and shoes than the people that are maintaining their health. Again, the study is in the show notes for those who’d like to dig deeper.
Chris Hemphill (03:43):
Patients point to communication as the top driver for their feelings toward their healthcare team. If time and effort go into digital transformation, but we’re not increasing patient engagement, then what are we really gaining? Maybe digital transformation isn’t enough until it measurably improves patient experiences. But how do we do that? Well, I know somebody who can help us, Edward Marx.
Chris Hemphill (04:07):
Before Ed Marx became chief digital officer at Tech Mahindra, he was chief information officer at Cleveland Clinic, Texas Health Resources and advisory board. He’s led all sorts of digital transformation efforts, including incorporating intelligent devices. We’ve been partnering with Microsoft on healthcare initiatives.
Chris Hemphill (04:27):
In 2020, he also wrote the book on digital transformation in healthcare, titled, wait for it, Healthcare Digital Transformation. So Ed, how should healthcare be looking at the patient experience?
Edward Marx (04:40):
Definitions are always tricky. And none of us consider ourselves the authority on defining patient experience. What I always tell people in organizations is that you need to define it in a way that makes sense for your organization, but everyone needs to agree to it, so that when you use the word patient experience, everyone in the organization understands exactly what that means. We believe we’re always in a patient experience. If you said, “Am I a patient?” I say, “Yeah, I’m a patient of a Dr. Mark Flesher. He’s my PCP, and has been for quite some time.” So, I always maintain that relationship because I’m all about wellness and health, and stuff like that too. So, I always think of myself as a patient.
Edward Marx (05:18):
And other people would define, or organizations would define a patient experience as an episodic way, and what’s your journey like. So, it’s a sum of, I think the one thing that everyone would agree, it’s a sum of a lot of different parts. And it’s important that each organization define it and make sure there’s buy-in to that definition, and that people own that definition. But yeah, that’s my perspective on it.
Chris Hemphill (05:41):
So, there are a number of ways to look at and define the patient experience. Ed argued that whatever the definition, it’s important to act on it. What’s the impact for organizations that don’t address the patient experience?
Edward Marx (05:54):
From a competitive point of view, it used to be the right thing to do to have a patient, a buds persons, or a customer relations departments. They were called a lot of different things. They’d been around for many, many years. We just added the word patient experience to it in the last five years or so.
Edward Marx (06:14):
So, the whole thinking has been around for a while, but you didn’t really have to do much, because if you were the hospital in town, people didn’t have alternatives, but today, everything, that has flipped as well. There’s a tremendous amount of disintermediation going on. And I know this isn’t the focus, but just to level set, why this is so important. Well, number one, it’s so important because it’s the right thing to do.
Chris Hemphill (06:36):
The right to do. Keep that in mind. We have a higher duty to our patients and consumers.
Edward Marx (06:41):
Number two, it’s really important because it helps in the healing process when you get patient experience right. But from these other sort of from a business perspective, you now have Amazon coming into your space, your city. You’ve got Walmart. You’ve got CVS. All the whole retail, high-tech. And then you have payers who now, the majority of physician-owned practices are in a controlled owned by payers. So, complete change.
Edward Marx (07:10):
So, now it’s a matter of survival. I call it survival of the digitalist. You have to get this patient experience thing right. So, from a healthcare hospital centric point of view, provider point of view, you have to nail this. So, it’s a matter of survival.
Chris Hemphill (07:25):
So Ed, when we talk about making the patient experience better through digital transformation, where do we start? What if I’m a leader in healthcare system with the chief in my title, or what if I’m in operations or working in the front lines, where can I start with digital transformation?
Edward Marx (07:44):
For me, one of the sayings I have is to be innovative, you must be innovative. And I think it’s the same way with digital. To be digital, you must be digital. So, if you need to experiment, if you just do what you normally do, you are not going to get it. You’re going to miss the train. So, you need to go outside of healthcare. And you can need to find out best practices other industries are doing. So, that’s really key.
Edward Marx (08:05):
And so, be digital yourself. So, force yourself to learn. That’s why we tell people, switch your phone every year. Try Android phone, if you’re always using iOS. Do something to shake things up. Try Clubhouse. Get on Clubhouse. Learn about Clubhouse. Be active on LinkedIn, Twitter, whatever. Whatever the latest things are, get engaged and learn from a personal experience, because then what’s going to happen? You’re going to start thinking, “Oh, what about this? Oh, we can apply this to healthcare.” That’s one of the reasons I started working in the OR, one day a week. I spent an entire shift in the OR once a week. Did I have to for my job? No, but it was there working in the OR. And you can do the same volunteering. So, maybe you can’t work in OR, but you can volunteer.
Edward Marx (08:45):
So, go volunteer someplace. And I did. That was one of the main reason I volunteered too, is just to be exposed to new things and to see the patient perspective or a clinician perspective and just say, “Oh, wow, man, I didn’t know it was that hard. That should be easy.” And so, it gave me new ideas, fresh ideas. But then at the same time, I would spend time with colleagues outside of healthcare.
Chris Hemphill (09:06):
That’s a really powerful point. Change starts with you. It’s one thing to read Ed’s book on digital transformation, but it’s another to put it down and go out, and get some hands-on experience in healthcare. What are some key themes that Ed has seen from those experiences?
Edward Marx (09:22):
One key sub-theme is about the millennials and other digital natives that you’ll be interacting with or are interacting with. And a lot of them are postponing care due to costs. And many of them don’t even have a primary care physician. So, what this really talks about is you need to come up with different ways and aligning delivery models, understanding digital natives. They do not want a relationship with a PCP. That’s been the traditional model, right? Baby boomers, we all want to have the same doctor for 30 years. They’re not interested. They are okay. And prefer, and jumping ahead a little bit here, but to have a mobile app and press a button and talk to whomever can help them at that time at the right price point.
Edward Marx (10:05):
So, understand your changing demographics. A lot of times, I think we still designed health care for baby boomers. And we need to be what I call bi-modal. Yes, of course, you’ll still do the traditional things that you’ve always done, but you have to realize the major shift that’s taking place.
Chris Hemphill (10:24):
All right. So, we see that there’s a major shift taking place. But what if I’m comfortable with the way things are right now, surely people will continue to need my health care services, right?
Edward Marx (10:34):
Then what are the consequences of inaction if we don’t engage? So, you can see, we just pick up on heart disease alone, the costs will triple in the next less than 10 years. And so, chronic disease continues to be a major, major issue all around the world, but especially in North America.
Edward Marx (10:55):
So, we really need to make bolder moves in prevention. And again, a lot of this is obvious. You’re probably thinking, “Well, that’s pretty obvious.” But we don’t do it by and large, but we’re doing it in other industries. You can see the statistics there. 30% of millennials prefer walking. And again, remember, as people grow older, more and more will be digital natives. And so, you’re going to see that trend continue. And so, you really have to think about it and offer new ways of delivery, because they’re not interested in getting in a car, driving for 30 minutes, having a five to seven minute consult with a physician, wait in a waiting room for an hour, reading old magazines, and then reversing that process. It’s just not going to work. And so, that’s something to really think about now.
Chris Hemphill (11:42):
Okay. I get it that types of experiences that people are willing to accept are changing, but are we looking at lasting meaningful change or are these just flash in the pan trends?
Edward Marx (11:53):
Roughly, about 1% of outpatient visits were virtual prior to the pandemic. During the pandemic, and it varied by institutions, but the average of average is about 60%. 60% outpatient visits were virtualized for a period of two to three months. Since that time, it has gone down to below 20% and continues to contract. Now, I don’t think it’s going to go back to where it was. Certainly, it isn’t. But I would resist the temptation to go back to the old model. And that’s what we’re seeing. So, you really have to make sure that you make this as convenient as possible, and then the seamlessness, right?
Chris Hemphill (12:31):
Right. So, far we’ve highlighted communication and convenient experiences, but personalization and digital transformation have a much deeper role to play. What’s the impact when we start thinking about how difficult it is, for example, for patients to navigate their own needs within healthcare?
Edward Marx (12:48):
Care pathways are really important and very critical for many, many reasons, but you can personalize these, so that it’s a care pathway designed for Ed, not a care pathway designed for a generic 50 year old. So, it’s all about that personalization. And it makes a difference in the outcomes. It’s amazing how much waste we have, both from actual costs and logistics, and supply chain, as well as a clinical quality digression because we don’t personalize. So, it’s really important. And again, convenience is as important to many of our patients as is the cost and quality.
Edward Marx (13:28):
I’ll give you a couple more examples. And remember, just be listening for the authenticity of the brand, telling stories and making sure that things are highly personalized. Patients want care where and when they prefer. And again, for most of them, it’s not in a brick and mortar. So, not only is it about mobile and about voice, but it’s also about the home.
Edward Marx (13:52):
So, care is transitioning to the home. You want some evidence. CMS, last fall launched their Hospital at Home program, where they’re for selected DRGs. They’re allowing you to discharge patients earlier. In some cases you won’t even admit these patients, and allowing them to be cared for at home using technology that exists today. It’s still a little bit of a hybrid model, but it’s a step towards a fully automated model. And that’s what people desire.
Edward Marx (14:23):
So, you really need to make sure that you have these bi-modal capabilities to deliver care in the setting that that patient would like it. So, that’s really critical. And how do you do that? You have to know me. You have to know the patient. If you don’t evolve, you will be completely disintermediated. And it may be too hard to catch up. So, it’s really a warning sign.
Chris Hemphill (14:50):
Disintermediated, eliminated, that is a serious and plausible threat for teams that resist change. Traditional healthcare is up against some powerful, deeply funded resilient players. Is it too late for change?
Edward Marx (15:03):
I think the provider side, because we haven’t moved quick enough, I think there’s still hope. So, it was really a call to action. It’s not a negative. But we’ve seceded. We’ve given away the whole primary care base, right? The largest primary care base now is with on the payers. And you’re seeing that growing. And you see the entrance of retail, and now Amazon, and other big tech. And it’s all about the patient experience. That’s where people are going.
Edward Marx (15:26):
So, with Amazon, you give me this great one-click experience for my telemedicine visit, and you’ve got a drone, that’s delivering my medicines shortly thereafter, guess where I’m going? So, it’s a wake up time for those on the provider side to really get moving on a lot of these technologies we’re talking about.
Chris Hemphill (15:41):
Personalized contextual experiences are no longer nice to haves. There must haves. Tech and retail companies that are good at this are swooping in. How can consumer interested in health systems invest effectively in this environment?
Chris Hemphill (15:56):
Carrie Liken, the head of healthcare industry at a technology company called Yext, joined me recently to discuss innovation strategy. Her unique background in healthcare technology and public health policy inform a rich understanding of how to address consumer needs. With a background in tech and healthcare policy, maybe she has some ideas on how to outpace the disruptors.
Carrie Liken (16:22):
I would say start with the small stuff, because I think that making these big changes to help drive patient and consumer expectations, health systems typically take a really long time to get something done. And by the time they get it done, the world has moved in a totally different direction.
Chris Hemphill (16:39):
Carrie Liken (16:40):
Just start with something really small. Find something really small, and then move on to the next thing. I think that consumers are forgiving, but they’re not going to be forgiving forever in healthcare when they’re seeing that there are so many other elements of their lives outside of healthcare. They are just easier. So, if you just start small and then you get some feedback on that small change or that small iteration, and keep on going, I think that is a really good place to start. So, whether it’s a digital change you need to make, or maybe it’s just, let’s say online appointment scheduling, just for urgent care right now. And that’s the way to dip your toe in the water. Or maybe it’s understanding value from one particular patient or subset of patients. And in researching all of the financials behind it, so you can get the lifetime value. Just start with one small thing and don’t let it deter you when it gets hard because it will get hard. But I think you’ll do a lot better if you just start small and keep moving.
Chris Hemphill (17:32):
We talk about experiences in other industries. Let’s break down one of those experiences and compare it to healthcare.
Carrie Liken (17:39):
I can think even about iBank with Capital One. And we even have a Capital One cafe around the corner from where I live. And I don’t use cash for pretty much anything anymore, but I like walking in there. If I do need to get cash, I can go to the ATM. They also have the Peet’s coffee, et cetera, et cetera.
Carrie Liken (17:56):
But what I think is the most memorable is when we bought our condo five years ago. We had a large sum of money in our account that we needed to transfer to. I forget, to escrow or whatever it was. We had to get this large sum of money out of the account. And I called Capital One and they sent me through the IVR tree. But then I was talking to someone and I told them why we needed this large sum of money. And the first thing they said was, “Congratulations. You’re meeting a goal that you had set out. Because in Capital One, you can actually name each of your accounts.
Carrie Liken (18:30):
So, I had a house account. So, we had saved the money for the condo, et cetera. And they congratulated. And they were like, “We’re going to take care of this.” I think the worst part about it was the IVR tree that I had to go through.
Carrie Liken (18:42):
So, even in financial services, which I wouldn’t necessarily say has the best consumer experience, I can say that there are certain companies out there including Capital One, where I would absolutely say, “This is wonderful. I will always patronize you. I will always be your customer because you make it so much better.” Can we say that about healthcare? Not a single time. Could we say that about healthcare? You could say it about an interaction with a doctor, just like, “Oh, Dr. Smith is such an excellent doctor. I’ll always recommend this doctor.” But the whole experience across the board to get to that doctor and then post, I would never endorse for the most part because that consumer experience is so bad.
Chris Hemphill (19:21):
Again, ouch. There’s a clear divide in how much investment training and thought goes into patient care versus a full consumer experience. Who’s getting the consumer experience part right?
Carrie Liken (19:33):
So, we work with Providence. And I think Providence does a really good job of attempting to get to this level. Although, they acknowledge and we acknowledge that it’s really hard to do this, but they’re always thinking about the consumer somewhere else, not pulling and forcing the consumer to come to them, but going out and meeting the consumer in pretty much any way, whether it’s digitally, whether it’s in a neighborhood, whether it’s on a website, whatever. And they’re trying to figure out the transactions involved as well.
Carrie Liken (20:03):
I would say CVS does a pretty good job. That’s not really a health system. That’s I think a new market entrant, I would call it. And while I normally always have something to say about MGH-
Chris Hemphill (20:15):
MGH stands for Massachusetts General Hospital, and it’s affiliated with Harvard Medical School.
Carrie Liken (20:21):
I have to say that I got my COVID vaccine through MGH. And it was the first time I felt like MGH had moved into this decade as far as technology is concerned. So, I have to give them kudos on it. The text message I received that said, “You’re eligible to receive your vaccine, click on the link.” It was a mobile optimized ability to actually schedule at a certain location for a certain time. When I got there, everything was completely digital and everything too. I was shocked. I was telling my husband today, I was shocked that they were using a QR code and using a mobile device to actually input this information into Epic. It’s fascinating. And then when it came time to get all my reminders as well, like reminder, “In two days, you have your vaccine.” Reminder, “Tomorrow, you have a vaccine.” Excellent. I think it was absolutely excellent. How do we translate that across the rest of MGH or the rest of healthcare? So, I would say those are a couple of examples.
Chris Hemphill (21:11):
Great question. How do we deliver Massachusetts General Hospital’s experience more broadly? Do we all have to go to Harvard? Probably not. Delivering consumers the access and experience they want is more than just a technology problem. It’s a people problem. Connecting cultures and aligning strategy is much more difficult than implementing and configuring the new technologies.
Chris Hemphill (21:35):
How about we speak with someone who’s tackled digital transformation at one of the biggest healthcare systems in the US. Marc Probst was chief information officer at Intermountain Healthcare before his current stint at LK. He breaks down the cultural silos that block innovation. Like Ed Marx, he has an inspiring way of viewing IT as a service provider for consumers and patients rather than as an isolated technology organization. Marc and I spoke on how to use data to drive the experiences that consumers want.
Marc Probst (22:10):
To try and prioritize that, really is trying to get into the heads of people and what’s most important to them. To a diabetic, having good information about glucose levels and how they’re living their lives and the activity that they’re involved in, that’s the most important use of data. To somebody with heart disease, well, some of the data is the same, right? Their activity and those kinds of things. But then there’s very specific data important to them.
Marc Probst (22:38):
There’s huge disparity around the world in providing health care. And a lot of that, I mean, there’s a lot of reasons for that disparity. But if we could standardize data and understand data, we could share knowledge, right? So, if we can now have data, that means the same thing in Kenya as it means in India, as it means in China, as it means in the United States, then the knowledge that’s starting to get created in each of these locations could be shared, because now it’s meaningful in each of those different areas.
Marc Probst (23:12):
And the point I got to in that particular talk was that there are literally hundreds of billions, if not trillions of dollars that could be saved, and tens, if not hundreds of millions of lives that could be saved, if we could get to those standards and share that knowledge across the world. And then we could start to see some of those disparities go away. Because it’s not necessarily the medical devices, although they’re important, but it’s really that knowledge it’s going to save the lives. So, it’s a long way around your answer, Chris, but it’s really dependent on the individual and what’s most important to them.
Chris Hemphill (23:52):
Marc spoke of the need for common data standards to understanding our patients. It was great to hear that from a global perspective, but it’s also important to think locally. Marc goes deeper into what that looks like for an individual health system and the implications that this has for artificial intelligence.
Marc Probst (24:11):
At Intermountain, we started an innovation center. This is 10 years ago, maybe a little longer. In that innovation center, when we first started it, it was all about devices and how could we get to the hospital room of the future and stuff. Pretty quickly, it came around to AI, right? And the things that we could do that around AI uses within our health plan and how we could do some of the predictive activities there and better serve our population within where we serve with Intermountain Healthcare.
Marc Probst (24:40):
But another thing that we… And this is, it’s not a very creative name and it’s probably used by a lot of other people, but at the time I used the term Doctor Google. That could be Doctor Apple. That could be Doctor Facebook, LinkedIn, or whatever, wherever data resides. Right? And the concept was, and I believe this is doable today. It’s not really something we can do because of FDA and other things. But I believe that very quickly, much of what we do in our urgent care centers now, not trauma or anything like that, but the more common otitis media or just the flu, or those types of things, the phone is going to become our primary care physician. And that’s because we can integrate data from so many different things, whether it’s the wearable that we have on, whether it’s our health record that we have on Epic or Cerner, or wherever. And we can bring all that data together. So, I think we’ve only begun to scratch the surface of what we can do with AI.
Chris Hemphill (25:35):
Great. So, we’re seeing some capabilities here that we can enable through digital transformation. But like we discussed earlier, this isn’t about playing solitaire. It’s about cultures coming together. How does Marc get buy-in from clinicians to drive this sort of innovation?
Marc Probst (25:52):
In my experience, it was showing them the value of what we were doing. Like advanced decision support, when we came out with decision support, it was this from the physicians, “We don’t want that. It’s cookbook medicine. You’re telling me how to do my job.” And we had to do two things. One was show them, “No, we are just giving you some guidelines, some edges that we want you to stay within and that’ll help you.” And we had to prove that to them, right? By staying within those guidelines, they were doing a better job. So, we not only had to show them the data on how to improve their care, we had to show them that they were improving their care by using those tools and those guidelines.
Marc Probst (26:33):
So, I think extending that to today to, “Well, okay, let us show you what we’re doing with this data, and together, how we’re working to improve the care that the patients that you have are getting.” And physicians are two things. They are one, quite competitive. And so, when they have data that they can see how they compare to their peers, that’s really helpful to them. And the other thing that they are is they truly care about their patients. And again, when you show them that the things that they’re doing, the things that we’re doing together with data is actually improving the lives of the people that they deal with, they come around pretty quickly.
Chris Hemphill (27:12):
Let’s think about this. For the digital and AI initiatives that Marc drives, he didn’t mention mandates and strong arm tactics to get people on board. There’s a collaboration here. Recall Carrie Liken’s advice.
Carrie Liken (27:26):
Just start with something really small. Find something really small and then move on to the next thing,
Chris Hemphill (27:31):
Starting small gives you results that you could learn from, share and iterate before making the next move. Sharing these results early on is key to Marc strategy. You want to invest more broadly in the AI for patient engagement, show that it can help a population of cardiology patients? You want to get clinicians on board with the new approach, work with them and show that it’s benefiting their patients?
Chris Hemphill (27:53):
For digital transformation efforts, there’s still something more important though, than starting small, communicating and even iterating. You also have to be able to choose the right projects. A good marketing team can make any technology sound good.
Chris Hemphill (28:09):
Look at Theranos, for example, you don’t want to be in a scenario where you’re starting small, and iterations keeps producing failure after failure. So, how do we know which AI projects have the best chance?
Joe Schmid (28:23):
There was a great study done with Boston Consulting in conjunction with The Sloan School, where they looked at a large number, over 2,500 executives who were involved with AI projects.
Chris Hemphill (28:32):
That’s Joe Schmid, chief technology officer at Actium Health. He spent the last five years developing an AI team and infrastructure to help solve healthcare’s patient engagement challenges.
Joe Schmid (28:44):
And so, there were some pretty shocking findings that we came across. The first thing that we found from the study is that a whopping 70% did not have payoff from their AI investments. So, seven out of 10, pretty abysmal results. Now, it’s not all bad news. On the flip side, that there are 30% that had some good results, and we can learn from that. So, if we dig in and we look at what are some of the key takeaways, there was one that stood out to us, and this is a great quote from the CIO at Roche.
Chris Hemphill (29:13):
The quote, “AI is not a separate agenda. It’s a subset of the tooling and capabilities we’re using to pursue strategic objectives.”
Joe Schmid (29:22):
What he captured, I think is really powerful. And it’s not that AI is some separate kind of siloed agenda. You’re doing technology for technology’s sake. It’s tooling, it’s capabilities, but it’s in support of objectives. And it’s really got to be viewed in that light. And so, we’re going to talk a little bit about how we can put that into practice with that in mind. It analogous to just like Ed talked about, it’s really about experience and the technology serving that experience. It’s AI serving that strategic set of objectives, part of which is delivering on experience, especially around digital transformation.
Joe Schmid (29:57):
So, let’s take a look at broadly, what were the types of opportunities that people pursued in this study, implementing AI? We could divide them into two simple buckets. And on the one hand, some of them were around efficiency and cost reduction. The other set of opportunities were more around growth, proactive outreach and being engaged.
Joe Schmid (30:18):
And so, interestingly enough, the results of these two areas were really different. And what the study found is that for the former, success rate of those were generally low, versus the areas that did see a great payoff were much more around these opportunities for being proactive for driving revenue.
Chris Hemphill (30:35):
To reiterate, AI initiatives tended to fail when they were just tech initiatives, not tied to strategy. When those strategies were around growth or expanding revenues, those projects tended to be more successful than those that were focused on cutting costs. So, it seems that the robots that want to take our jobs just aren’t so great at them yet. The approaches that help us grow are the ones that we should count on. If you want to read the MIT Sloan study that Joe was referencing called Winning with AI, you can find it in our show notes.
Chris Hemphill (31:08):
Let’s dig a little bit deeper on that strategy part. People like me look at AI as a set of algorithms that help make various predictions or distinctions within data. What can that possibly have to do with strategy? Maybe Joe and help us think through how to make that connection.
Joe Schmid (31:22):
What we do in working with clients is we start with their objectives and we work backwards on actions and activities that support those.
Chris Hemphill (31:30):
Well, that’s the core of it. Start from the strategy and work backwards. What does that look like?
Joe Schmid (31:34):
And we can take a simple example and just step this through. So, let’s say you’ve got some objectives around being proactive and increasing breast health, early detection. That’s great. This is a specific example. You might have a whole broader set of things. But let’s take that as one simple use case. What you want to do then is work backwards from there and talk about, well, what actions would support that?
Joe Schmid (31:56):
Personalization is such a big deal to consumers today. How can you put together personalized calls to action that would support this particular objectives? All right, if that’s going to be some of your activity, what’s going to support that? You might need some predictions from AI that are going to give you both clinical predictions, that top area around propensity for someone to need breast health services, which women are more at risk. And you would have scores from AI models in that area. But that’s often not enough.
Joe Schmid (32:25):
You’d also want models that give you predictions around what content is best for Joe, versus Ed, versus Chris. And so, now through the use of AI and training models on outreach and campaign results, we can start to make personalized predictions that said, “You know what? Joe really likes text messages. Ed really likes digital engagement in his mobile app.” And those kinds of personal, both channel outreach, as well as content that’s most effective for your consumers on a personalized basis are going to get you much better results.
Joe Schmid (32:56):
And then the last step of the process of course, is you need some data sources to be able to train models. And your data scientists, your trusted partners can help you work all the way backwards to say, “What sources do I have?”
Chris Hemphill (33:09):
To highlight what Joe just said on working backwards with AI strategy, first, start with the overall strategic objective, like growing cardiology volume by 2%. Then think about what actions consumers would need to take to make that happen such as scheduling consults or health risk assessments. Then think about what predictions would be needed to help drive those actions. And finally, do we have the data to fuel those predictions? Strategy, action, prediction, data. Again, strategy, action, prediction, data.
Chris Hemphill (33:43):
There’s still something else that’s preventing good digital and AI strategy. And that’s hype. How harmful can a little hype be? Well, think about investing heavily in a data solution to identify people with COPD risk. It’s heavily advertised, but in practice misses thousands of people who have needs.
Chris Hemphill (34:01):
Think about Epic’s approach to identify sepsis risk, which was advertised as 76% accurate, but was right only 63% of the time. We’ve linked to the Stat News article on this algorithm in the show notes. If you’re not careful, the hype in AI can lead to major pains and wasted investments on stuff that just doesn’t work. Even worse in healthcare, hype has a cost to lives and wellbeing. So, how to cut through the hype.
Joe Schmid (34:29):
There’s a lot of hype. There’s a lot of talk. And it’s hard to cut through the confusion. I think at the top level, everybody is going to tell you they have great accurate models. They’re all in production. They’re super effective. Unfortunately, you need to dig a little bit deeper. So, what we recommend is asking some specifics. And we do lots of work across different areas and things like breast cancer models. Questions that clients and prospects ask us are things like, “What kind of data was the model trained on? So, what went into it?” Vendors should be able to tell you that sort of detail, how did it perform? Unfortunately, we’ve seen a lot of history where that kind of performance measurement and metrics just aren’t shared. It’s kind of black box, secret sauce.
Chris Hemphill (35:12):
Cutting through the hype lesson one, no black boxes, no secret sauce in healthcare. If vendors refuse to tell you how well models perform or the types the data that fed them, they’re not real partners.
Joe Schmid (35:24):
They should be able to tell you details about how much lift will it get you. If you’re into lower level details, what’s the area under the curve score? Things like that. And then last but not least, what does a model use? What kind of features go into it? Things like that. That kind of visibility, transparency are things that you should be looking for.
Joe Schmid (35:40):
And then another critical question is you should be asking if there’s bias in these models. If vendors are not careful, they’re likely to not only pass along bias, but potentially amplify it. There have been a lot of high profile cases in other industries around models, unfortunately amplifying bias, where models start to use costs instead of clinical aspects and things like that. It’s something that we’ve got a lot of work to do, but that work is happening. We’re encouraged by.
Joe Schmid (36:07):
Specific questions you want to be thinking about or asking about when approaches have shown racial bias and what was done to mitigate it. You want to ask about internal review processes or these kind of one-off things that happen, or are they really baked into products and procedures. And then you want to ask about, are they working with third parties?
Joe Schmid (36:26):
As an industry, we do a great job of security audits. We expect vendors to have SOC 2s and high trust reviews from external auditors. We think we’ll see a big trend towards audits around bias and fairness. And that’s really important work that the industry is starting to do.
Chris Hemphill (36:43):
When we say demand a better future in healthcare on every episode of this podcast, this is a key example. When it comes to AI and digital efforts, we should be demanding transparency and performance from our vendors.
Chris Hemphill (36:56):
On the topic of creating and demanding a better future within healthcare, a great example of someone who’s doing an extraordinary job at creating that better future is Rebecca Wiesner, who’s director of IT applications at HonorHealth in Phoenix, Arizona. Why Rebecca? Well, she helped lead their vaccination efforts. In just a matter of weeks, they had a full scale vaccine delivery arm, delivering thousands of vaccines per day, outdoors. Her care delivery and digital transformation background are a unique intersection and perspective. We discuss what it’s like working on the front lines at multiple institutions, what’s working well, the challenges, surprises, learnings, and ultimately the role that we can all play in creating better experiences for patients. Here’s Rebecca.
Rebecca Wiesner (37:44):
Our first day we had a thousand patients. And it was very stressful. We were like, “Wow, we did a thousand patients. That’s amazing.” By the time we got to the second week, we were doing 1700 patients and thinking we might be able to do more. And then we had a day where we had a lower number than 1700 on this second round for the second dose. And we were like, “Man, it feels so slow.” We had almost gotten so efficient that it was laughable because we could do a lot more. So, we adjust our schedule as needed to accommodate for that. But we’ve just learned along the way. There’s no rule book here. There’s no playbook. It’s developed. And I think with anything like that, you really have to constantly look at the process improvement process along the way to make it even better.
Chris Hemphill (38:34):
Again, even when there’s no playbook, the iterations and collaboration among IT, leadership, clinicians was key. They met every day to course correct and steer the ship right.
Rebecca Wiesner (38:45):
We did huddle at the end of each day to address what didn’t go well, where could we improve, what adjustments do we need to make for tomorrow? And that’s discussed every single day. So, it’s a constant process improvement game we play with ourselves.
Chris Hemphill (39:02):
Let’s face it, healthcare is an industry with lots of bureaucracy, resistance to change and fixed processes. Regarding this, Rebecca offers something of a blueprint for the way we ought to look at change.
Rebecca Wiesner (39:15):
I can’t mention this enough. It’s an evolving process. We have to be meeting and discussing it daily at this point, if we want to continue to administer the vaccine in any capacity.
Rebecca Wiesner (39:28):
There’s not always an easy answer to everything. Like I said, one of our biggest challenges right now is our 75 and older population. How do we outreach to them to get them in and not leave the most vulnerable out? We don’t have a great solution as of right now. And I think you’re seeing this problem nationally, but we’re attending towards technology because we don’t really have a better way to do that. There’s not a hundred team of people ready to make calls to engage. And I think this is where we’ve leveraged our volunteers tremendously. We’re having volunteers just do calls or reach out to the community like never before. And without those volunteers, we would be done. There’s no way we could have accomplished this operational process without them. So, we have very close partnership with them.
Rebecca Wiesner (40:22):
And again, this is the time to partner up, work with your community on how to do this, because there are pockets of people who can help. We can’t silo ourselves individually as healthcare entities or the county, or the state. We have to share the information. We have to partner up and identify how we’re going to do this collectively as a nation.
Chris Hemphill (40:46):
We’re going to have to step outside of our comfort zones. It’s not in healthcare’s comfort zone to share data, but it’s time to stop playing solitaire. More from Rebecca on this.
Rebecca Wiesner (40:57):
Where we’re not experts is in traffic control. So, an example for what we did with our clinic is we don’t manage that. We know that’s outside of our realm. But what we do know is healthcare. So, I think where we went right and where we want to continue to move this is, is to try and leverage some of that space to identify, moving forward, how we’re going to help those other populations. How do we deal with them today? Can we leverage a system we already use to make that happen for the vaccine? I think it’s important that we remember that.
Rebecca Wiesner (41:36):
And then we also, don’t try to over-complicate things. In healthcare, we’re notorious for sometimes over-complicating things. And sometimes that’s related to regulatory requirements, et cetera. But I think we need to go back to the drawing board and say, how can we do this simply? We have a kind of a theme in IT to make IT or it easy. And there’s never been a time more than now that we needed to really hone in on that mantra. Is that, how do we make it easy to schedule appointments? How do we make it easy to get the vaccine? And we need to continue to think about that as we move forward. I mean, we’re just not there yet. And it’s never been more important to make it easy.
Chris Hemphill (42:19):
It’s never been a more important time to make it easy. Though, perhaps we are late to the party. The healthcare ecosystem is moving into a new digital frontier. There are cultural and technical challenges and barriers to this movement, but these are problems that other industries have faced as well or are even facing right now. It’ll be necessary to take from their lessons learned and focus on how we can collaborate and break down silos between teams. Same time, it will be necessary for us to take our lessons learned and serve an example to other industries, what healthcare can do. It might not be perfect, but we’ve seen healthcare come together, culturally, technologically in terms of business processes, in a response to this pandemic that we never seen before. And that type of change could be responsible for helping or saving hundreds of thousands of lives. Perhaps this new focus on digital transformation is a watershed moment for healthcare. Let’s keep iterating and see this change through far into the future.
Chris Hemphill (43:27):
Thanks again for tuning into Hello Healthcare. If you like what you heard, we appreciate a review on Apple, Spotify, or wherever you’re listening. You and your feedback fuel us.
Chris Hemphill (43:38):
This conversation is brought to you by Actium Health. To get the latest on what these healthcare leaders are saying, subscribe to our newsletter on hellohealthcare.com or join us for our weekly sessions on LinkedIn. Thanks. And when we see you next time, hello.
Find the Clarity You’ve Been Missing
Learn how Actium Health is driving improved quality, outcomes, and revenue for innovative health systems nationwide.