Introducing new technology to a health system can be equal parts exciting and terrifying. Freshly implemented tools can often delay workflows, confuse patients and clinicians, and can create inefficiencies. Technology for technology’s sake isn’t enough to deliver and fulfill on the promise of new advancements in healthcare. Human nature tells us that we often default to the easiest choice, so aligning that with the right choice is often the key to unlocking clinician ability, enhancing patient outcomes, and oftentimes, significant financial savings for health systems.
On today’s episode of In Network's Designing for Health podcast feature, Nordic Chief Medical Officer Craig Joseph, MD, sits down with Mitesh Patel, MD, Chief Clinical Transformation Officer at Ascension. They discuss his background in creating healthcare technology solutions, how he helped found the Penn Medicine Nudge Unit, and why he decided to attend business school before medical school. They also discuss the idea of sludges and nudges in clinical care, leveraging behavioral economics within healthcare, and how health system improvements were inspired by reality television.
Listen here:
In Network's Designing for Health podcast feature is available on all major podcasting platforms, including Apple Podcasts, Amazon Music, iHeart, Pandora, Spotify, and more. Search for 'In Network' and subscribe for updates on future episodes. Like what you hear? Make sure to leave a 5-star rating and write a review to help others find the podcast.
Want to hear more from Dr. Joseph? Order a copy of his book, Designing for Health.
Show Notes:
[00:00] Intros
[01:35] Dr. Patel’s Background
[06:30] His founding of Docphin
[10:33] The history of the Penn Medicine Nudge Unit
[14:06] Making a difference in the Nudge Unit
[24:31] The future of Nudge Units across the globe
[25:32] His current role at Ascension
[29:14] Integrating AI into wellness text messaging
[34:27] Sludge vs Nudge
[38:35] Dr. Patel’s favorite well-designed things
[40:39] Outros
Transcript:
Dr. Craig Joseph: Mitesh, welcome to the pod. You are the vice president and chief clinical transformation officer at Ascension. Is that true?
Dr. Mitesh Patel: That is correct.
Dr. Craig Joseph: That is accurate. Okay. But back in the day, back in the day, you were a lowly internal medicine resident. Is that also true? You were a lowly internal medicine resident?
Dr. Mitesh Patel: That is also true.
Dr. Craig Joseph: All right. Awesome, and you don't only spend your time at Ascension right now, but you're also an adjunct professor at the Wharton School, which I understand is a pretty good business school.
Dr. Mitesh Patel: I'd like to think so.
Dr. Craig Joseph: All right, so let the record show that you've agreed to all my statements. You know, I'm interested in kind of, how did you get from being a lowly internal medicine resident to where you are now? Was this kind of planned? Did we did you know you were going to be going in this direction, or did you say, no, I was going to open up an office and see patients and, order X-Rays and lower people's cholesterol.
Dr. Mitesh Patel: Yeah, it's a great question. I think as many people, you know, going through training, you kind of figure it out as you go and you have ideas of where you want ahead or certain directions, but there are influences or experiences that you have that really set the trajectory in one direction or another. I would say for myself, when I was going through internal medicine residency, I saw two big changes happening, kind of at the same time. One was that electronic medical records were becoming a thing and being spread more widely. And so it used, you know, when I was in residency, we documented on paper. But you could see a lot of those platforms shifting to where it was digital. And that's from the clinician perspective. From the patient perspective. You started to see iPhones come out, followed by other smartphones. And so you would go and speak with a patient and, before, you know, when you came back, you know, they had to they may have more questions, so and so forth, but now they're in between when you're going and seeing them, they're looking up these things or they're reading what other people online have said about the test result that you just gave them, or the diagnosis that they just found that they had, and they're a bit more informed and engaged. And so I saw these two technological changes happening, but I also observed that a lot of effort was being put into getting these things up and running or patients, you know, using smartphones. But there wasn't much thought into the design or the right process to help make it easy to find the right information, both for clinicians and patients. And I saw an opportunity to think about, hey, how could design fit within the shift from paper to electronic medical records and the shift for patients in the ability to engage access to information at the fingertips through their smartphones?
Dr. Craig Joseph: So you saw that kind of through the lens of design. Did you or at least were you thinking about it that way? Did you have a background in design in any way, read any books, taking any classes? Your father was an industrial designer? I just made that up. That's not accurate.
Dr. Mitesh Patel: No, you know, I didn't have any formal training in design, but I did, I did go to business school. I did an MD MBA right before I started residency. And people choose to get, you know, some clinicians who have MBAs choose to get them at different points. For me, I really wanted to get it before I went into residency, because I felt like it would give me another perspective to look at things through. So design or process flow or user interface engagement and user experience, where things that, you know, you touched upon in business school, but there were frameworks that had been put in that, that give you an additional lens to look through in addition to the medical lens. And so I think that really helped me open my eyes for thinking about processes and things that happened that other people might say, hey, that's just the way it is. And for me, I really wanted to understand and question, you know, how can it be better?
Dr. Craig Joseph: So a little bit of knowledge, can either A: take you a long way or B: be dangerous. Hopefully it was more A for you. So you got your MD MBA, did your residency, and then you got actually more training in behavioral economics and digital health and clinical trials and all kinds of fun stuff. As a Robert Wood Johnson scholar at Penn. How did that come about? So that was after you finished your residency and you had some of that more formal business training than almost all of your peers. What, why did you seek more, even more training?
Dr. Mitesh Patel: Yeah. You know, when I was going through residency and business school at Penn, I was able to meet, you know, mentors that worked in the area. And we're thinking about this idea of behavioral science and behavioral economics and how we could use that specifically to improve health and healthcare. And so Penn and a variety of other sites offered, the fellowship that you mentioned, the Robert Wood Johnson Clinical Scholars Program, which was basically two years of human capital funding to do whatever you wanted to do. And it was interesting for you. And so for me, I wanted, I had been doing research for a while, but it wasn't in the behavioral economics or behavioral science space. It's actually in medical education, you know, and other things, quality of care. And so I wanted to take some time to get to almost have, apprenticeship, you know, with my mentors, Kevin Volpp and David Asch and others. You know, how behavioral science and behavioral economics can be embedded within health and healthcare. And then I specifically wanted to learn about clinical trials. You heard about clinical trials a lot for drugs and medical devices, but less so for, you know, some of these design or behavior change applications that I really wanted to work on. And so Penn was a great place to learn as they had a lot of these things going on. And so I took that opportunity to spend that extra training to be able to set my career path on the right track. Great for what I wanted to do.
Dr. Craig Joseph: Okay, so that sounds like a good decision on your part. Around this time, you also started a company. Was it around this time you started Docphin?
Dr. Mitesh Patel: Docphin was a little bit earlier. It was towards the end of medical school and business school. I was in medical school and business school at the same time. So it was towards the end of that. It kind of took off while I was in residency and then kind of closed off when I when I started my faculty position.
Dr. Craig Joseph: So what was Docphin? What was the goal?
Dr. Mitesh Patel: Yeah. So the idea was when I was going through medical school and, and training, I used to want to keep up with the latest and greatest research that came out. Because when you're on the wards surrounded with a clinician who's an expert in their field, they're going to bring up the latest paper from JAMA or the New England Journal or, or Circulation, if you were on the cardiology rounds, and the way I did that was I had paper journals that would get delivered, you know, to my apartment, and they would stack up. And then at one time I would find an hour or two, and I flip through page by page, 10 or 15 journals, and I would find maybe three or four articles that were interesting, and I probably wouldn't have absorbed much, because I just went through a process of flipping through hundreds of articles. And around the same time, as I mentioned, things were going digital, medical journals were also going digital. And in fact, all of these, journals were now online, but they were very difficult to access. If you were at an institution where your institution paid for access to all these journals, you had to type in your username and password. Then you had to type in the journal, let's say JAMA, then the year, you know, 20-whatever, then the volume, then the issue and then the page to get to the journal, to get to the article that you want. It wasn't like flipping through it. And so what I wanted to think about was as these are going digital, how could we make it easy for you to find what you wanted? And so we created an app, first a website, and then as, iPhones and androids came out, an app that would allow you to encrypt your username and password and put it in your device so we wouldn't capture it, but the user wouldn't have to keep typing it over every time they went looking at an article. And it would give them one-touch access to all of these things. But we'll go a step further, which is once we had a network of people using this, and we had to identify it as a medical student or an attending in internal medicine or cardiology or surgery or whatever it was, we could then serve up articles to you that other people like you might be reading, sharing or discussing. And so it almost became instead a, you know, instead of you trying to find the articles that you wanted, we would present you each week with the four to five articles that we thought you might be interested in, based on what you had read in the past, what people in your field had read, and what was popular. And so this took off in medical schools because you're going on your cardiology round, you know, fellow electives for two weeks and you're able to quickly see what are the academic cardiologists from, you know, Harvard, Stanford, Penn, Michigan, etc., are reading and sharing and discussing. And you'd be up to date, but you could also, if you're a clinician at the bedside, be able to use this research at the point of care, which was really difficult beforehand.
Dr. Craig Joseph: Yeah. So, I think nowadays most people would take out almost everything you just said for granted. Yeah, that's easy. There's lots of ways of doing that. But back in the day, this sounds like it was kind of a novel idea and implementation. And so, I noted that,you won an award from this company, some may have heard of it, called Apple. And they said that you had the best new iOS app out, that year for medical professionals. And so it's not only were you, were you doing a kind of a good deed and, you know, made a lot of sense from the service that you were actually offering, but you did it in a good way, at least from a design perspective and using the technology, whatever happened to that company?
Dr. Mitesh Patel: Yeah, it was an honor to receive these awards. We had grown it pretty quickly to about 60,000 users at 500 institutions. And then, a couple of years later, it was acquired by HealthTap, a telemedicine company that wanted to kind of incorporate some of this technology platform.
Dr. Craig Joseph: That's great. And that was all very early in your career. So kind of really set you up quite nicely. So the main reason I know you is that you were associated with and helped start the Penn Medicine Nudge Unit. And so I would love to hear more about that. First of all, what is this thing called a nudge that I hear about? And, and why did you need a unit for nudges?
Dr. Mitesh Patel: Yeah. Great question. So nudges are simple ways to change information, either by looking at the choices that we offer or the information that's delivered that can have an outsized impact on one's behavior. And so some examples of nudges are changing the default to make something opt-out instead of opt-in, so the path of least resistance is that you do the behavior that's preferable. Another one is something called active choice, asking you simply a question do you want the flu shot this year? Yes or no? As opposed to waiting for you to realize that you need to make that decision? just prompting you to make the decision now and making you feel some regret if you say no as to what can happen if you don't say no or or some benefits as to what you might get if you say yes. And then there's framing information that could be through showing you transparency around the pricing of tests and treatments, or it could be around showing you how your behavior might deviate from the norm. And a lot of people want to fit in with the norm. And if you realize that, you know, 70% of people get breast cancer screening every year and you haven't been getting it, maybe that encourages you or motivates you to do that. Nudge Units are groups, behavioral design teams that focus on leveraging the science of behavior change, and specifically this idea of nudges and embedding it within real-world environments. So the first Nudge Unit started in the United Kingdom government. It was literally an experiment and written that the organization will be disbanded in two years if it was not successful. And thankfully for all of us they had many successes which translated to huge impact. And then nudge unit started taking off within governments. There are more than 50 of them in different governments or governmental organizations around the world. But never before had been within a health system. And I had been working in this space of, you know, design, behavior change, nudges, and doing some pilot work and thought, hey, this might be a good opportunity to think about whether or not a nudge unit would be something that could fit within our health system.
Dr. Craig Joseph: Terrific. And you, as you're talking about nudges, I'm thinking that I bought an airplane ticket yesterday, and I was presented with information that 5,000 people have already bought the insurance, that I chose to opt out of. So that was a little nudge from Delta, I think, trying to get me to do something that they wanted me to do. So the idea is to kind of incorporate these nudges into the real world. And I guess the opportunity to do that, at least in the healthcare setting, was much more significant. Once the electronic health record came to town. So it's a little more difficult when you have a piece of paper, and you're writing orders, and there's nothing, there's no order set. There's just a blank piece of paper. I might be dating myself, but that's how I practiced back in the day and certainly how I was trained. And so you kind of had to come up with everything, and the EHR and some clinical decision support tools really kind of expanded the area, the opportunities for you to do this. So, how did this all get started? Did you, you know, seek funding and they gave you money? And this is a great idea. Did you have to do some studies to prove something? Did you have to say nice things about the department chair? What would you have to do?
Dr. Mitesh Patel: Yeah. You know, actually, the idea for the Nudge Unit or what sparked or turned on the light bulb was actually the result of a project that I started when I was in the Robert Wood Johnson Fellowship. And the results were, and I'll share the story here in a moment. The results were pretty, you know, eye-opening to folks and got the idea of like, hey, maybe we could do this more systemically throughout the organization or systematically. The idea was this idea of generic medications versus brand name medications. Many people know generics and brand names are exactly the same chemical formulation. But brand name prescriptions are oftentimes much more expensive. And there's good data to show that patients don't adhere to brand name prescriptions as often as they do generics. And that's probably because they're more expensive. And so for you to have to refill them, you have to pay a large amount of money or a copay or whatever it might be. Whereas many generic medicines are cheaper or have no copays associated with that. And so when we looked at our data, Penn Medicine, had done a worse job at prescribing generics than all of the other peer institutions in the region, in Philadelphia, and our insurance companies were telling us that every month, and in fact, we were getting fined or losing out on opportunities for incentives and bonuses, because we were a low performing health system in terms of generics. And they tried doing all kinds of things. They tried holding Grand Rounds and educating clinicians around why it's important to order generics the same as brand name prescriptions. And of course, all the clinicians agreed and said, yes, of course, we know this makes total sense. We'll do our best to do this. They started going around and sharing data with clinicians around how their performance was. Yet nothing really changed. And so I initially had an idea, which was that what if we showed people the price difference in the medication, and put that in front of clinicians? So when they're ordering, let's say, Lipitor instead of Atorvastatin and they'll see that like, hey, there's a generic available which is X number of dollars or X percent cheaper than the brand name prescription. And I'd actually gotten a grant, a research grant, you know, just a couple of thousand dollars to implement this. And I learned when I went to the internal medicine department, they said, hey, this is a great idea. We just change the default already. And this might not work because they're never going to see that, because if they order the brand name prescription, it just now gives them the generic option. So if you type in Lipitor it just shows you Atorvastatin. And you have to click an extra button to see all of the brand names. And I thought, well, this is not what I intended to test, but maybe a better opportunity. And it ended up being a natural experiment where internal medicine was proactive and decided to do this. But the family medicine department at the organization did not do it. And so we had a natural experiment here where one department turned it on and the other didn't. And so we compared changes before and after, and we found that there was about a 5-10% increase in generic prescribing for things like statins, beta blockers and other common medicines that are similarly prescribed between family medicine and internal medicine. So we went to the health system and said, hey, and they did this national experiment. There's an opportunity here to move from 75% to 80 to 90%. We actually think you could even make the design better instead of just making, you know, instead of even, you know, not showing you and having you expand. Just make it a checkbox. Click here if you really want the brand name, otherwise it'll automatically go as generic. And that ended up sitting in a queue for over, you know, up to two years where the health system said, hey, this is a priority for us. We've had a lot of changes as we're, you know, updating our new electronic health record system. And so it went on a list and, you know, kind of stayed in touch with IT leadership. And I said, yeah, it's a priority, we're working on getting to it. And then one day, out of the blue, I got an email from an Epic analyst who I'd been partnering with in that same internal medicine clinic who said their opioid prescribing just changed, and all of a sudden, we're 99% generic, whereas opioids used to be much lower. You see 50 to 70%. Do you know if anything happened? No, I'm not aware of anything happened? I know we've been working on something. And so we started asking around, a flurry of emails initiated, and what we found out happened is an analyst who had been working for the electronic health record system update. Happened to see this on a queue and a list of things that wanted to be implemented and had already implemented this at another health system. It wasn't something that was natively there for the electronic health system. It had to be implemented instead proactively and said, hey, I can just bundle this in with another change that was going out and kind of submitted it without anyone really knowing at a leadership level and essentially had, you know, created another national experiment, but now had rolled it out to the entire health system for every specialty. And what we found was that overnight, the generic prescribing rate went from 75% to 99%. And over the course of the next two and a half years, if you take the top 100 most-prescribed medications, this saved about $32 million in unnecessary spending just by this you know, one little change that the analyst told us took about an hour to implement.
Dr. Craig Joseph: Yeah. And that analyst got what was that 10% of the savings? Is that the typical finder's fee for that?
Dr. Mitesh Patel: There should be some incentives aligned, I'm not sure that was the case in this. Yeah, it was a huge impact.
Dr. Craig Joseph: Fair enough. I feel like that might be a way of kind of incentivizing analysts around the country to be more aggressive. But that's probably got some downsides to it that I haven't really thought, completely through. So. Wow. So this change that you would wanted to happen and was kind of just being stalled or just kind of pushed down, by it for things that were higher priority got done. No one really realized it, and it made it a significant, significant change. That's amazing. And so that happened before the Nudge Unit or … ?
Dr. Mitesh Patel: Before that was really the impetus. So we went and showed this to leadership and said, hey, you know, this took two years to go from pilot to expansion, $32 million in savings. We are now the number one generic prescribing health system in the region. We are now getting bonuses instead of fines from our insurers. There is a huge opportunity to do this much more, you know, systematically and think about what the high-level opportunities are as the EHRs were going out, they would do things like list the options alphabetically or sometimes not really give as much thought to what was pre-checked and what was not pre-checked in order sets. And so there was a bunch of, you know, some low-hanging fruit that we could then go and start to tackle.
Dr. Craig Joseph: Excellent. So the Nudge Unit was born kind of based on that government organization in London, but obviously you needed to kind of transform it and configure it for healthcare. How did you do that? How does one, I completely understand the administration being like, yeah, Dr. Patel, here's money, here's lots of money. Make this happen for many, many different things. So, where does one start out, who's your team and how do they work?
Dr. Mitesh Patel: Yeah. And I would, you know, even say it wasn't like a bunch of money that was needed. It was more around, you know, making it a priority and then doing it in a more coordinated way. So I got funding for one project manager to start, we grew over time, but that's what I was I got to start, here's one project manager, but what was more important was to get leadership buy-in and to say, hey, this is a priority. And in order to adapt the Nudge Unit format for our health system, we created a kind of steering committee. And the steering committee was comprised of kind of three components we thought would be really important for us to make, you know, rapid progress and to adapt to the healthcare. One was leaders from, you know, the quality group in the healthcare system to the Chief Quality Officer, the Chief Medical Officer, and so forth, who could help with that and saying, hey, what are their priorities? How do we understand, like, what makes sense from a patient perspective? That’s also what matters. The second was the IT leaders, the CMIO folks in the IT office who could help us, you know, not sit on a long queue waiting to be implemented, but moved to the top because it was deemed a priority and get implemented. We set a timeline of going from idea to implementation in six months or less to start. So that was the goal. And the third were experts in behavioral economics and behavioral science, both from the medical school where we had a lot of strength, but also from Wharton, which is right next door. And also it has a lot of, a lot of folks that have been working in this area. So those three components, the medical and quality leadership, IT, and experts in behavioral science and behavioral economics helped steer what we were doing. But we wanted to literally kick it off, tell people, you know what we're doing it. So we decided to do a crowdsourcing tournament. We would announce this to all the members of the organization. Anyone could submit an idea for what we should work on. We would pick a couple to start with. But we made it a tournament where the steering committee were the judges, and in the first two weeks, we got 225 ideas submitted. You could be a medical student, a nurse, tenured professor, a respiratory therapist, anyone from the health system. It's a great idea. And in fact, you could also vote on other ideas, but only if you submitted an idea. So encourage people to submit more ideas. That's what we want to generate after the first round. And we got it down to about 30 ideas that we thought would be a good fit for what we wanted to focus on. And then we hand it over to our steering committee who voted, got it down to ten, and then we brought those ten groups in for a pitch day. And to make it, you know, exciting, we had a bunch of prizes to participate in the tournament. One person got, I think, an iPad at the time. Other people got gift cards. But what people really like was the opportunity to, you know, pitch their idea in front of leadership to the CEO, the CMO, all the leadership from the health system attended, and almost like a Shark Tank-style format, we're able to get feedback. And ultimately, we chose three ideas to move forward.
Dr. Craig Joseph: I just love that you were leveraging the principles of design and behavioral economics to get good ideas about design and behavioral economics. It's, kind of a meta thing, and I, I definitely appreciate that. You know, hey, you can vote on what other people do, but only if you put an idea in, and then the, you know, gamification and kind of making it, giving out prizes. So, so that's awesome. So it kind of kicked off with, three ideas that you picked from our Shark Tank, like, show contest, and, and it's taken off from there. Obviously, you're not at the organization anymore. What, where are they now? From your perspective? And have they really messed things up since you left? And you can be honest, no one's listening. It's just me and you.
Dr. Mitesh Patel: I think Nudge Units, the Unit at Penn and others are thriving. I'd say in the five years where I was there, we implemented more than 100 programs and projects across, not just Penn, but many other organizations. We also started an annual Nudges in Healthcare Symposium, where we brought in health systems from around the world who were thinking about starting one and then eventually did start one. So, UCLA and Geisinger, were ones in the United States, we've also worked with the Ministry of Health in Singapore, Saudi Arabia, and then many health systems in Australia, Europe, and South America. Implementing these ideas within their health systems or within the health portions of their government. So, the idea has really taken off. There are many Nudge Units within health systems. And after I left, the Nudge Unit has been thriving, Kit Delgado is now the current director of the Penn Medicine Nudge Unit and really impressed to see what's come out of there and excited to see what will come out of there in the future.
Dr. Craig Joseph: Awesome. Well, I'm excited as well. Let's talk more about what you're doing now. So, as we mentioned, you're at, you're one of the senior medical leaders at Ascension. Some of the things that you've been working on seem to be related to things that you have been working on in the past. So, it was really your opportunity, I think, you had discussed, when we were prepping for the show to kind of implement some of the ideas that you helped to develop and create at the Nudge Unit. So, what are some of those projects? I understand that you did a big Salesforce kind of redo, which is unusual for two physicians to be talking about Salesforce. I'll put that out there. But you seem to leverage it for a very kind of in a patient-centric way.
Dr. Mitesh Patel: Yeah. But, you know, Ascension is one of the largest nonprofit health systems in the country, very mission focused. And there was an opportunity to take a lot of the work that I've done in the scientific lens and implementation and implement it at scale, and also to reach communities that serve the poor and vulnerable, in a way that we could enable access to the care that these folks needed, that were otherwise there were just too many barriers or, you know, opportunities to make it easier for them. And so, I think that's what excited me to come over, you know, when I came over, essentially was in the process of transforming the way it communicated with patients, some of the call center data, the marketing team had moved on to Salesforce. And one of my first priorities was, how do we turn this into a clinical platform? And that meant, you know, how do we hook it up to our electronic health records? How do you make sure it's, you know, it meets all the compliance standards of HIPAA and other things, and we're keeping patient information secure. But then how do we take the science and the evidence that we used and a lot of the other settings at Penn and elsewhere and embed it within communications you might otherwise normally receive? So, for example, you know, before you see your doctor, you get an appointment reminder, hey, your visit’s coming up. Click here to confirm. We thought about, hey, what if we gave patients a care plan, a digital care plan that came from their doctor in a personalized way and had a couple of things from the record that says, hey, looking forward to seeing you. Here are a couple of things I want to talk about: flu shots to prevent you from getting sick. You know, cancer screening and reducing the risk of heart disease because a lot of times, you know, I was an internist and I practiced for ten years at the Philadelphia VA, and patients would come in and I'd bring these up with them, and it would be the first time that they hear them. They weren't primed for the discussion, so they'd have to process it, thinking they're coming in for some other reason. And so sometimes we say, okay, let's talk about it again in three months. And so, we wanted to see if there were ways that we could, we could help make that process better. We also did things around patient report outcomes, specifically around hip and knee surgery, knee replacement surgery. There was a CMS requirement that's actually coming up in July of 2024 to start tracking this in a more sustained way at health systems across the country. And so, we saw this as an opportunity to be proactive, you know, if you think about surgery, the ways to measure quality from surgery are really around, did the patient survive, which we would hope they would, obviously. But then there that's what you can see in the electronic health record databases. But what you really want to know is did the patient’s function improve. Are they able to walk up and down stairs more than they could before? Has their pain gone away? And the only way to get at that instead is to reach out to the patient and ask them. Many systems do that by phone or on paper. We saw an opportunity to leverage nudges and digital approaches to do that at scale. We're currently in the process of rolling that out in more than 60 different hospitals.
Dr. Craig Joseph: That's great. Let's talk for a second about your JAMA article that came out a couple of years ago, 2022, about nudges via text messaging. Seemed that you were very successful at Ascension to be able to really close some of the gaps that were there for vaccines and cancer screening. How did that work? What did you learn?
Dr. Mitesh Patel: Yeah, that's a great question. You know, when I came to it, it was a year and a half or so into the Covid pandemic; vaccines had been released but still were not completely taken up. Many organizations had requirements that patients get vaccinated, but they had set deadlines for far into the future. And what we wanted to try to do is understand how we could motivate our Ascension employees to get their vaccine sooner and we leveraged some research that I had done at Penn, we did a mega study. I did that with Katy Milkman, Angela Duckworth, and others. And we basically crowdsourced ideas from behavioral scientists around the world and then tested 20 to 25 different approaches to send text messages to patients, at two health systems, Penn and Geisinger, and then also, members of Walmart Pharmacy, about a million of them, from across the country. And what we found is in both studies, Walmart Pharmacy members and health system patients, this concept of psychological ownership worked the best and was really simple, as opposed to having to send them a video or a questionnaire. In some ways we tested it, even sending people a joke using humor to see if it motivated vaccination. What really worked the best was, instead of telling people a vaccine was available, was to tell them that it was either reserved for them or it was waiting for them to give them a sense of like, hey, this is theirs. They should claim it. And act on it rather than give away their opportunity to someone else or before the vaccine is gone. And in those studies, it had a 5-10% boost in vaccination. So, we wanted to try something similar here. And so, when we came in, we ran a randomized clinical trial where we rolled out to half of the folks first, and then we eventually rolled it out to the rest. And we did something very similar where we told people that it was reserved for them. And we also scheduled the date and time that they could get it and they could very easily come back and change that. And so, we went one step further. You know, we had places on-site where you could get it. So, we said, hey, this vaccine's reserved for you. We've gone ahead and made an appointment next Monday. That doesn't work for you? Just click here, and digitally, you can change it to another day of the week or another week if that works. And what we found is that it significantly boosted vaccination, both in terms of actual explanation, and also just documentation. I mean, part of the challenge, as a health system is understanding who's vaccinated, who's not. And so people, who are already vaccinated start submitting their documentation. We made it really easy to do so via an online link. It used to be you had to like, take a picture of your vaccine, and email an email to yourself, and download it, and upload it to another site. We made it so you could just do that with your phone, just click a link and take a picture and you're done. So documentation went up and then actual vaccination went up. And so, we pulled forward a lot of the vaccinations that might have otherwise been delayed towards the deadline.
Dr. Craig Joseph: Yeah, I'm continually fascinated by the little tweaks right in the language, like you said, you know, we've reserved this for you, give you kind of a sense of ownership of this thing, whether it's cancer screening or vaccine or whatever it is. And then, you know, instead of saying, hey, click here to get a date, we've given you a date, and chances are excellent it's not good for you, but that doesn't matter. We've given it to you and, you can change it if you want. And it's very easy to do that. And these little, tiny tweaks make such a huge difference. I am aware this is not the first time I've heard about the Nudge Unit. In fact, I attended the Nudge Unit conference last year, and there was a researcher who presented, they did a bunch of different kind of tweaks on the language of these text messages and other kinds of reachouts to patients, and just showed that there's not one that works for everyone, but that there are many different options. And, and I think as we get, you know, it's probably going to be very exciting for you and others as LLMs become more omnipresent to be able to say like, well, for this group of 30- to 32-year-olds who live in this area, they generally respond to this language as opposed to that group and kind of micro-target people. So, that seems like an area ripe for future research.
Dr. Mitesh Patel: Yeah. I think there's a huge opportunity. We still very much, if you look around what's being done, so I think one obstacle is just getting the science of nudges into practice in the real world. And we, you know, we're chipping away at that. But even where that's done, it's typically one-size-fits-all. We're sending that same message to a million people. Whereas I think what you're saying is right in that each of the people have different experiences, behaviors, and beliefs. And so, a different message might have more impact for some versus others. And so, with the ability to almost democratize the use of AI and prediction in a way that we haven't been able to do before, we may be able to better engage with patients by personalizing messaging.
Dr. Craig Joseph: Yeah. Awesome, well, the final article, you've got a lot of articles to go through that I wanted to talk about, was a Harvard Business Review article from 2022 talking about removing sludge from healthcare processes. And so, first of all, what is sludge? why do I want to remove it? And how did you go about, what are some of the things that one can do to remove sludge?
Dr. Mitesh Patel: Yeah. So, sludge is kind of a term maybe coined by Cass Sunstein, who coauthored the book Nudge. Sludge I kind of think of as the opposite, almost the opposite of nudge. Nudges are meant to guide and facilitate behavior. Sludges are the things that get in the way. They cause friction. They're, you know, when you click unsubscribe from an email, instead of just unsubscribing, they make you type in your email address, or they make you go through four or five clicks to do that. And so, when we started at Ascension, one of the first things we wanted to do was let's try to reduce unnecessary work or noise before we start adding, you know, new nudges and other things. And so, and we went around all the leadership and asked them, what are the opportunities to, you know, what’re the big challenges and then where are there opportunities to reduce sludge. And we found a couple of themes that came through that work. One was just reducing the number of steps in a lot of ways. And, actually, I already gave this example, which was that, documenting your vaccine, your COVID vaccine, it used to be four or five steps. So, you literally had to take a picture, email it to yourself, download it from email, upload it to another log in a website, upload to the website. And now what we would do is just serve you up a link by text message that says click this link and take a picture and you're done. So we saw that in the week that we implemented that, COVID documentation doubled, just for making it easier for people to do that. Another one that we found was making options digital that were, you know, previously done by phone or paper. And we've talked about this a little already, but the example that we wrote about in that article was we were transitioning a lot of our employees to a new pharmacy benefit platform. We actually have our own pharmacy in this country called Ascension Rx. And they were being enrolled in that, and they had to do that by a deadline. But the way they would do that is, first we asked them to call us. Not many people called, then we tried to call them. And what we found is that for every 100 phone calls we did, we converted 10% of people that we tried to reach; 10% conversion rate most of the time because we couldn't reach people. But sometimes when people picked up, they were just busy, or they didn't know who was calling on the other end. And they said, hey, I'm not going to do this right now. What we decided to do is, instead of doing it through phone, was to email them and that we actually already had a lot of the information we needed. Your name. We know that because we know who you are, the address to which to send the prescriptions. But we have an address for you. We do need you to confirm that's the right address. Did you move? Do you not want your prescription sent to this home? But you want them to send to another location and so on and so forth. And so what we did is we created an email web form, but instead of having a link, we embedded it within the email. And then we prefilled the information we had about you. And we just asked you to confirm the information or edit it within your email and hit submit. You didn't even have to leave your email to do that. And, what we found is in the two weeks after we deployed that, we started with the 1000 members who were on the most expensive drugs, and they would have gotten large bills for these had they not moved to the plan that was now covered. And so they were the highest priority to convert. And what we found is that in the first two weeks, 72% of people, 720 people out of the 1000, converted and required no phone calls, no hassle. Experience was great. And so what ended up happening was we just ended up creating an online website that you can access with single sign-on, and you could see the information we had for you. You could confirm it or change it, and then you just hit send and one click. Yeah, we have found a couple other things. One was around removing burden on patients. We did patient report outcomes and where we removed some of the unnecessary surveys. And then the other one was around doing things remotely, either through virtual care options or remote patient monitoring for things that you otherwise would have to come in person to do, but actually didn't need to be physically in person to get done.
Dr. Craig Joseph: Mitesh it’s been great. We've learned a lot. At the end of our podcast, which is where we find ourselves now, I always like to ask our interviewees if there's anything that's so well-designed that it brings them joy and happiness. Are there things that you just think, wow, someone did a really great job on this, and whenever you use it or are exposed to it, it brings you joy? Anything like that come to mind?
Dr. Mitesh Patel: Yeah, well, I'll give you two examples. Maybe opposite ends of the spectrum, one in the physical world and one in the digital world. So in the physical world, one of my favorite examples is the piano stairs. And if you've ever seen this, you can Google piano stairs. But you know, they have an escalator next to stairs to go from, kind of like a subway up to the ground floor. And what they did is they painted the stairs as piano keys, and in some places, like you can actually step on them and play music. But even without that, just gives you a sense of how like the design of those stairs encourages. And they kind of had a camera on there and showed that like a lot more people were using the stairs because they were designed as piano keys as opposed to using escalators. So, thinking about physical activity, getting people moving. In the digital, and it's something, you know, obviously that everyone's paying attention to now, you know, ChatGPT, you know, obviously like a lot of amazing things about it. But I thought one of the subtleties that was interesting was when you asked ChatGPT a question, it slowly shows you the answer. And what I learned, was that was purposely introduced by the leadership of the company to make it seem like the chatbot was thinking. It could actually spit out the whole answer. And over time, I think it's faster and cheaper, but it will kind of slowly do that to give you a sense that it's thinking and it's really being thoughtful about what it says next. And so, I thought that little subtlety really makes you feel more engaged with the robot or whatever's on the back end responding to you than if it just like you asked a very detailed question, I gave you a six-page response in like two seconds. Then you would be like, that's a bit non-human.
Dr. Craig Joseph: All right, you're freaking me out. I did not know that about ChatGPT, and that is brilliant. Of course. Wow. All right. Well, those are those are excellent examples of design. And thank you for bringing them to our attention. Well, it's been great. I really enjoyed the conversation. Thank you again. And look forward to seeing what amazing things you continue to do at Ascension and with any other research that gets published, keep it coming.
Dr. Mitesh Patel: Great. Thank you for having me on.