In a clinical setting, designing with the user in mind is often an overlooked, yet vital step towards better health outcomes. Using principles of behavioral economics to align the right choice with the easy choice is paramount to ensuring a successful experience for clinicians and patients alike. Deriving insights and experiences from other industries can enhance every logistical and medical pathway in a health system.
On today’s episode of In Network's Designing for Health podcast feature, Nordic Chief Medical Officer Craig Joseph, MD, talks with Kit Delgado, MD, emergency room physician and director of the Penn Medicine Nudge Unit. They discuss his role in the Penn Medicine opioid task force, the impact of small, technology-based solutions on the patient experience, and what medicine can learn from the world of auto insurance. They also talk about the integration of behavioral economics into clinical workflows, utilizing technological solutions to decrease ER visits, and how to ask difficult questions as a physician.
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
[09:54] The Penn Medicine opioid task force
[20:04] Injury prevention research
[21:49] Lowering distracted driving rates
[27:55] Injury and death prevention
[34:20] Asking tough safety questions of parents
[36:04] Dr. Delgado’s favorite well-designed thing
[38:12] Outros
Transcript:
Dr. Craig Joseph: Kit. Welcome to the pod.
Dr. Kit Delgado: Hi, Craig. Good to be here.
Dr. Craig Joseph: Well, it's great to have you here. You're a practicing emergency medicine physician at Penn Medicine. You're the director of the Penn Med Nudge Unit. It also says that you're working on world peace and also a new rocket program for NASA. Is that all accurate, or did I misread some of that?
Dr. Kit Delgado: I think the last two are maybe not accurate, but I am involved in a lot of things. And it is true. I'm a research scientist by day and probably emergency doctor by night. I did work last night, so I apologize if I'm a little weary today.
Dr. Craig Joseph: Wow. All right. Well, I completely understand and will not take advantage of that by asking you leading questions that, you know, could get into some juicy information. So let's talk about your background. How does one go from being you know, attending medical school and then kind of, with a vision of maybe emergency medicine or maybe something else to practicing emergency medicine and leading the Nudge Unit at an Ivy League medical school. How does that arc flow?
Dr. Kit Delgado: Yeah, it's been an interesting journey. I went through medical school as the guy who went through every clinical rotation, liking a little bit of everything: surgery, pediatrics, cardiology. And I realized I didn't want to do just one thing. And I liked having a general, clinical, practice. The other thing that attracted me to emergency medicine is that emergency departments are really a window into the community.
Dr. Kit Delgado: Every failure in the health system, every failure of public health or the social safety net, results in people coming to the emergency department. And so, as someone who is gravitating towards public health and bigger picture issues, I saw the clinical practice, which is interesting in its own right, as an opportunity to meet people and take care of people who could tell me about what's going on in the community or the things that were going on, so I trained in emergency medicine, but I had an eye on doing research and affecting practice and policy and therefore a pursuit of research, fellowship after residency in health services research. So, this is when you learn how to analyze data, to compare whether treatment A versus treatment B makes a difference. Analyze outcomes. And I got great training at Stanford in doing that and joined the faculty here at Penn, initially as a very traditional health services researcher, focusing on how we should be structuring our emergency care systems or trauma systems, to better take care of patients. And it was after several shifts working in the emergency department, you see people coming in after distracted driving crashes or a drunk driving crash, people who are not taking their medications consistently and winding up with flares of their asthma or congestive heart failure. People who are struggling with drug use. And I became much more in trust instead of putting a Band-Aid on the problem. And treating people in the emergency department and sending them out, thinking more systematically about how we can develop solutions, to both help clinicians and patients, improve health behaviors and reduce risky behaviors. And being at Penn is really an amazing place because there have been some thought leaders here, like Kevin Volpp, who directs the Center for Health Incentives and Behavioral Economics, David Asch, who directed the Innovation Center here, and Mitesh Patel, who was the founding director of the Nudge Unit in 2016. And, with this ecosystem here, the nudge unit was set up to apply these insights from behavioral science to changing practice within the health system. And the first thing that they did was they sent out a call for proposals to frontline clinicians about ideas where there is a best practice or a thing that we should be doing, but there's something that makes it hard or is nudging people in the wrong direction. And how can we change that? And so, I saw this, and I thought to myself, this seems like a really interesting opportunity. And I was thinking back to one of my first years of attending practice at a different institution. And the resident physicians who are supervising kept on bringing me opioid prescriptions, for 30 tablets every time. So, ankle sprain, 30 tablets, fracture, 30 tablets. I thought, well, maybe the ankle sprain didn't need the opioid prescription, but why are they getting the same amount? And it turns out that the default number in the electronic health record, when you typed in the order for the opioid prescriptions, is 30 and you had to go in and reduce it to a more reasonable number every single time. And I said, we should be introducing default options for medication amounts that are much more in line with guideline recommended amounts. And so, I proposed that and it was selected as one of the first three projects that the Nudge Unit was supporting. Long story short, we implemented a default option for ten tablets in our emergency department, which was consistent with guidelines, basically double the number of prescriptions, for ten tablets to reduce the overall number of tablets prescribed. And it was a very strong proof of concept that, when the clinicians lack strong preferences and things are aligned with best practice guidelines that manipulating these defaults in the electronic health record could be really effective. And that really started my journey into leveraging behavioral insights and nudges for improving healthcare. And then there's a longer series of a longer journey into running a health system opioid task force and applying these insights to changing care.
Dr. Craig Joseph: Well, I love it. And I love that it kind of you, you went from being one of the one of the first folks to put out a suggestion, you know, to respond to that, that nudging at the beginning. And now, just a few years later, you're running the place. And so, kudos to you, sir. So, you know, you've described your work as an effort to redesign choice environments to make the safe choice the more attractive choice. And I love that quote. I, as you may know, I co-wrote a book about design and healthcare, and I'm not as bright as you. And so, that same concept, I, you know, we have a chapter called ‘Make it Easy to Do the Right Thing.’
Dr. Kit Delgado: Yeah. I like that chapter.
Dr. Craig Joseph: Thank you. and the $5 is in the mail for that. You know, make it easy to do the right thing. And, I often add, in, in a quieter voice and slightly more difficult to do the wrong thing. Right? And, and in this example, the right thing is to give a small number of opioids if you need to give any, but give a small number with no, make sure no refills. And but it's still possible to do the quote unquote wrong thing. And, and the reason that it is possible is that sometimes the wrong thing is exactly what you want to do. And so there are times where patients should have, something that departs from whatever the standard of care is, and that makes complete sense. And that's why physicians are allowed to do that. But as you point out, if it if you put even a little bit of friction in there, it makes it much less likely that someone's going to do that without intent. And, and I think that's what you're describing. Every resident coming to you with, refills of 30, they were not thinking that through. That just is what the electronic health record kind of popped in. So, let's talk about if you, if you're willing, some more about that, Penn Med opioid task force. This is one example of something that you've done. Are there other things where you've learned from behavioral economics and, and your emergency management practice that, you know, help you to, to make opioids safer? It's been a scourge in this country and sounds like it's, we know it's getting better, but I think it's getting better because of, in large part to work that you and others in this area are making.
Dr. Kit Delgado: Yeah. Yeah. This is work that we started back in 2016, which seems like a lifetime ago or several generations ago in the opioid epidemic. At that time, it was the diversion and misuse of prescribed opioids from physicians was the big problem. And actually, emergency departments weren't the biggest issue there because we were prescribing relatively low amounts. And one of my roles was looking at the broader health system in terms of prescriptions that were coming out of the system and looking at our electronic health record data. And when I looked at the data, it was kind of interesting as an emergency physician and sort of my clinical area, I've been focused in the emergency department. But when we're looking at the broader system where we're actually finding that for new opioid prescriptions, by far the biggest amount were coming from our surgeons and people like orthopedic surgeons. And so I used the data, and this is sort of goes along with one of our general approaches. But you're using data to figure out where we should focus. And I reached out to one of the top prescribing surgeons who's actually a really nice guy and very well intentioned, who did a lot of ACL repairs and simple knee surgeries, he said, our OR prescribed 30 for all my knee scopes and 50 pills from all my ACL repairs. I'm sort of thinking, okay, we were just doing ten pills for in the emergency department, and it was very well-intentioned. He asked him why and he said, well, I don't want my patients to be in pain. I don't want them to go to my competitors or call in the middle of the night. And we're doing these surgeries to relieve pain. And so he had all these well-intentioned ideas and using the behavioral insights lens to realize, okay, I could be in one room in the emergency department prescribing opioids to take care of pain from a bad fracture, and in the next room. Someone's coming in after an overdose. So I have that availability bias of seeing the consequences firsthand. But he's doing six or seven surgeries a day, one follow up appointment just to make sure the wound looks good and then probably never sees those people again. And so that was sort of the insight, in sort of well-intentioned addressing pain. And we said, how can we better understand this and realize that they didn't actually have a guideline for how much needs to be prescribed. So, we had a research assistant call all those patients a week after surgery and ask them, how many pills did you actually take? And the data were shocking. It's basically 30% took none. The average number taken was three. And we showed that data and he was like, wow, like this is really bad. Yeah. And so I said, well, we're doing hundreds of different procedures. So, we’ve got to come up with the guideline for everything, how we're going to do this. And so, in collaboration with one of my colleagues here, Anish Agarwal, MD, is really an expert in digital health. We pilot-tested a text message chat bot. Basically it started as another research assistant pretending to be an automated chat bot. And we refined a texting script that would basically, we texted patients after surgery, asked them how they're doing with their pain, asked them how many pills they used, and how many they had left over. Once we worked out all the kinks, we coded it and integrated it into our electronic health records such that every patient having surgery in the system was getting these texts. And then we aggregated like 20,000 data points across all these surgeries, published guidelines about how much should be defaulted for all these procedures. And we said, okay, we solved the problem. Now we just got to put in some order sets that create these order sets, an electronic health record that defaulted by mild, moderate to severe different departments. And we're like, oh, we won the battle. Just kind of done three months later. Like, let's pull the data on this. And there was like 0% usage of these order sets. So, I'm happy to report that these are design fails. Because what I've sort of later come to discover and really made part of our practice is that you really have to talk to the stakeholders in the specific clinical context. I really understand the specific clinical context. And it turns out that, particularly for hospital discharge prescriptions, it's a totally different process that doesn't use order sets. And so one big insight in our work is that user-centered design is so key, particularly in deploying nudges within a health system or electronic health record that you can come in flying in and they're going to take this shiny tool out of our toolbox and implement it, and it's going to work across every context. But in this case, it didn't work because the workflow is completely different. And what we now do sort of consistently as part of our process is very user-centered design, where we co-design these things with the stakeholders we wouldn’t have spent all that time developing those order sets. So we went back to the drawing board with our surgeon, and he said, you know, if you show people the data that people do just as well managing their pain, if they got an amount within the guideline as opposed to, outside of the guideline, that'd be very helpful. So we created some feedback reports that we sent to all the residents, the nurse practitioners, the PAs, is showing that for how much they prescribe relative to their peers and that if they prescribe within the guideline that patients do just as well with managing pain as it is, those who got above that guideline, they do just as well.
We implemented that; it affected over 20,000 patients. We reduced the number of pills prescribed pretty dramatically. The ability to manage pain remained the same. And guideline adherence shot up from 40% to 70% and stayed that way even though we turned it off. So it just goes back to that user-centered design and listening to what their primary motivations and sort of thinking about, okay, this was the availability bias. And now we addressed it and solved the problem.
Dr. Craig Joseph: Yeah. Well, I think the thing that shocked me the most was, I probably should have known this, but the number of people or the number of pills that people are actually taking, it's shocking. Because I would have thought a lot more, as did your orthopedic surgeons. Right. And then to actually have someone call them and just say, hey, just how many pills, do you have left over at this point? Yeah, I think that probably takes you a long way to kind of telling physicians where these guidelines, are coming from a source of kind of evidence and truth. And these are your patients. These are not made-up patients. Shocking.
Dr. Kit Delgado: Yeah, I think part of it's using a low-tech method first for refining that and then really listening to what people's barriers are and directly addressing those, I think were the keys to success. And then certainly going in were, thinking, you know, what the right thing to do is without really understanding, you never walk into a patient's room and say, hey, I think you need antibiotics or without sort of actually taking a history and doing a physical exam. And the same thing is true in designing behavioral interventions, for improving healthcare.
Dr. Craig Joseph: This story is making me think back to when I was, I was a new medical student. And I actually had dental insurance for the for the first time, and I had, wisdom teeth that just needed to come out at some point but wasn't an emergency. Got my wisdom teeth pulled out. It was fun. And, the dentist, someone from the dentist, the oral surgeon's office called me, a few days later to check on me, and I said, well, pain wise, I'm doing all right, but I'm not really feeling that great. And they were like, why? And I said, oh, I'm, you know, I've got nausea and I'm a little, I'm a little dizzy. And they're like, are you, are you taking the pain medicine? I'm like, yeah. I just remember the I remember the nurse clear as day. Stop taking the pain medicine. She didn't say the word dummy, but it was implied. And yeah, there's side effects there that you don't think about, and I didn't even prescribe. I didn't even ascribe to the right thing. So, yeah, I totally get it. And so workflow matters. Having a brilliant idea doesn't matter if you can't execute, a change and getting the right, in this case, users getting the right people to talk about how they do things so that you can work with them to enact changes to make it easy to do that right thing once they agree with you on what that right thing is. So you're not only working with, electronic health records in the hospital, but you're doing things outside, and I know, one thing that you mentioned to me is we were preparing, for this podcast is some injury prevention work with, with an insurance company. Tell me, tell us more about that.
Dr. Kit Delgado: Yeah. one of my areas of interest working in the emergency department is we take care of a lot of people who are injured, and one of the major injuries you see every day or people injured in car crashes, the prototypical one that I think is super common is, the person sitting at the traffic light minding their own business, and then a car slams into the back of them, which is these rear-end collisions are probably about anywhere between 30 and 40% of car crashes. And they're all caused by distracted driving, pretty much, but usually are unmeasured because people don't report it, they don't come out in the police reports. And I've been interested in distracted driving as this prototypical behavior in which if you do a survey of teens or the general population, everyone knows that they shouldn't be texting and driving on the phone or using their phone while driving. But people still do it consistently. And, it is a prototypical impulsive, automatic behavior where I thought, hey, maybe this is an area where we can maybe make some headway. The other kind of thing that enabled us to do this work is that we can actually measure distracted driving on a phone, because their apps there are commonly being used right now to measure how you drive. And insurance companies are promoting these apps because it's a win-win situation. For the customer, you sign up for insurance and you get invited to download this app. You can get a major discount for driving safer, you watch any pro sports game, now you'll probably be inundated with multiple insurance companies saying, talking about their apps and discounts for safer driving for the insurance companies rather than saying, hey, Craig, you're a guy who lives in this zip code and, and you've had this many prior crashes before, and we're going to make a prediction about your likelihood of getting a crash and then pricing your insurance accordingly so that they make a small margin. On average, they can look at your driving data and make a much better and more accurate prediction. And so their loss ratios go down. So what you're having now is 10% of insurance policies in the country are using these apps that track how much you're using your phone while driving, how much you slam on the brakes, and your total mileage. And the use of these apps is exploding. So you have this opportunity for scale. And right now, the current approach to doing this from a discount perspective is that you sign up, you get measured. You can look in the app, you'll get some big feedback about how you're doing. And then at the end of your six-month period, you get a discount, safe-driving discount. And we looked at that model and said, you know, behavioral science tells us that delaying rewards isn't that effective and that providing more frequent feedback, and particularly for this behavior where there's an instant, you know, if you get a ping on your phone and you tend to that right away, and creating a more instant feedback loop, we thought could make a difference. So we partnered with an insurance company that had an app that measured distracted driving. We recruited about 2,000 customers, and they were randomized to a couple of different conditions. One is the standard approach where they get a discount at the end. The second approach was instead of that discount at the end, why don't we tell you you're going to get paid in weekly allotments? But if you use your phone above a certain threshold, we're going to withhold some of that funding that that money and thereby leveraging loss aversion, but also making those rewards more frequent. And we're going to run a tally of how much money you're losing over time. And what we found was when we leverage those behavioral insights, that was twice as effective for reducing phone use while driving, compared to the standard approach. We’ve then gone on to other future research to look at how do we build these habits over time. And one of the one of the insights is that people don't change immediately right away. And so you can't say ‘use your phone not at all’ if they're using their phone ten minutes per hour. So we created an intervention where we grouped people together based on their average amount of phone use. So, people who use their phone eight minutes per hour all into one group. And we created a weekly competition to see who could reduce their phone the most and then we also, instead of people telling people not to use their phone, we provided tips and tricks for how to create sustained change and make it easier not to use the phone. So like here's how to make it easy to turn on Do Not Disturb while driving. Create a commitment plan for how you're going to not use your phone or if you need to use maps or music, how to do that. And kind of make it easy to do the right thing. And when we test that, compared to usual care, we found that that intervention reduced phone use by 30%. And those changes were actually sustained when we turned everything off. A lot of this stuff is coming out soon, but we think that we can make a lot of headway by applying behavioral insights to this emerging field.
Dr. Craig Joseph: Yeah. You're like, you and your colleagues are like wizards, you know, taking a little bit from here, like, hey, we're going to let you see the money that you're saving, but we're going to give it to you weekly. And not only are we going to tell you, hey, it's gone from 0 to $10, but we're going to tell you ahead of time. You're going to get $10, and then we're going to start taking money away from you. For behavior that you've agreed is, is, is not good. And, and most humans, myself, amongst us would we deal much worse with taking our money away once we've decided that's our money than we do with not getting free money? Or money that we didn't kind of expect or feel that we earned. So, boy, got a lot of a lot of different tricks in there and I look forward to kind of reading about your successes and the specifics of how you do that. You're also working with, not only, injury prevention from, from motor vehicle accidents, but also firearm, injuries. Tell us more about that work.
Dr. Kit Delgado: Yeah, we've done some work on the epidemiological side, highlighting the unrecognized burden of unintentional firearm injuries. If you just look at what's published out there, most of the data has to do with deaths and, of course, yeah, suicides. And then, assaults are the most, most common cause of death. But what you don't see is that a major cause of firearm injuries coming to the emergency department is unintentional injuries. Most people survive, but can be quite devastating, particularly for kids who pick up an unsecured, firearm. So this kind of came up as an opportunity. There's been a lot of concern in health systems about workplace violence and people coming in to hospitals or medical offices upset, bringing firearms with them, and taking it out on medical staff. And it is a countermeasure. What you're seeing increasingly in workplaces, including hospitals, is putting weapons detectors in the front of hospitals and a lot of proactive messaging that firearms and other weapons are not allowed inside the hospital. And so one of our hospitals implemented this weapon screening. And we provide a prompt that you can't bring your weapons inside. And so, the security guards, were noticing that people are coming in with weapons or they have to be turned away. And for a short time, we were able to actually provide storage of weapons within the security office while people are going to their healthcare appointments. And when I heard about this, I saw it as an opportunity, knowing that in behavioral interventions, to the extent that you can layer on something to, to something that exists in a normal operation, it's one of the most scalable approaches for delivering and nudge intervention. And so we partnered, we reached out to our hospital leadership, who is very forward thinking, and our security officers. And they said, yeah, we would love to be able to do something to promote safety. So we created a brochure and some training and actually started stocking cable locks that are able to secure guns and actually had the security guards engage these patients and visitors in promoting firearm safety. So distributing firearm brochures and promoting cable locks and in a couple of months time period, we distributed hundreds of cable locks through this approach. And it's something like 30 or 40% of households in America have a firearm, and many people are traveling around with them. But we know from data that 30 to 40% of people also don't have their firearms secured at home. So this is like, a not controversial area where a lot of people get hung up on the politics, but this is one area I think we can make a lot of headway in terms of promoting safe storage and using these touch points where people are being screened or sort of thinking about this. And now we're, we're piloting some other approaches in terms of just making it easy, having posters with QR codes that you can snap and just have a gun safety device mailed to your home, and even putting it on web pages and things like that. So just making it easy to access these safety devices and we'll be starting to see where other people are actually using them and what's more effective.
Dr. Craig Joseph: Yeah, you're, this warms the cockles of a pediatrician's heart. This is one of the things that, and you're right, there's some political overtones with it. There shouldn't be, but there are. It's a public health issue, of course. And, that was one of my standard questions that I would ask about once a year for my well-child checks is, is there a firearm in the house? And, and I would often get side-eyes from some parents, and I'd say, well, you know, this is a major cause of death for children. So, if you have an unsecured firearm, that's a major cause of, concern. And I think most people take that in the right way. And I think that's a that's a great idea. While they're at the hospital, promoting clearly they're not there thinking about, kids at their house, but they most likely do either have their own children or have, family members that, you know, bring kids over and explaining to them that this is a helpful thing, that's a usable, certainly usable intervention.
Dr. Kit Delgado: Yeah, it's funny you mention that because I had an annual visit with my family medicine physician, and I asked him, do you talk to your patients about it because it is the leading cause of death for children now, unfortunately. And, he said, yeah, it's always uncomfortable. There's always some people that give a lot of pushback within that and within the visit. I still think it's the right thing to do. I think it's a challenge, though, to get every primary care physician to do it consistently, and having these high volume touch points where it's sort of within the context of like coming in and out of a place or something else that may be a little bit separate and outside of the healthcare encounter, but still providing that resource. I don't know, I think it's an opportunity and makes it easier for people, to get what they need.
Dr. Craig Joseph: Yeah. Well for sure. And, you know, the way that I approached it and I think this is not, this is just the way I was kind of trained at a children's hospital. This is how we think. to me, this is part of my safety. My safety list of questions. Right? And so, for older kids, who could ride a bike, it would be, you know, does the child, does your child wear a bike helmet? and, you know, other kind of safety questions do you have if it's a baby, is the baby secured in the car? They have the car seat. let me ask you even more questions about that. Is it in the back? Is it facing backwards? You know, like, these are kind of very specific questions. Then the third question I'd throw in there, right in the loop is any guns in the house? And so clearly, we're talking about safety and when I, I don't practice now, but boy, when I was practicing, I assure you that, firearms were not the number one cause of death of children, but they are now in the United States. Let's go through the list of all the other, countries where this is the number one cause of death in children... Okay, we're done with that list. That list is very short. It's got one country on it. So, yeah, I think kind of leveraging all of these design tricks to make it easy to do the right thing. The right thing is to keep kids safe and do what you can. And, that makes a lot of sense. Well, we are nearing the end of our time together. And I wanted to ask you the question that I ask all of our, interviewees on the, on the podcast. And, we like to talk about design in healthcare. And, boy, we've hit both of those targets, quite well today. Are there any things or workflows or processes, or devices or apps that are so well-designed that they, they bring you joy? or happiness, just whenever you're doing it, you're like, wow, this is a well-designed thing that I'm participating in. And it warms the cockles of your heart as I like to say.
Dr. Kit Delgado: Yeah. I had to think a little bit about this, but, for me, it's not a specific app, but it's a class of apps. It's the parking apps. I just loathe carrying around change for parking meters. And I also, I hated reading the meter and I just love that you can drive, anywhere now where these exist and type in a code, it just taps into your credit card and your parking is done. So it makes it easy from a payment standpoint, but then you're off and doing your thing and you can just look at your phone and see how much time is left on your meter and then refill it. It just makes me really happy.
Dr. Craig Joseph: All right. Paying, parking fees is what makes you happy. It's actually the parking fees.
Dr. Kit Delgado: It just makes it easier. And I think it's a win-win. They probably make a lot more money in their revenue. But for me, the pain of pulling change out of my pocket or finding change or going into a convenience store to find quarters to fill the parking meter, I mean.
Dr. Craig Joseph: That's impossible. I think it's impossible nowadays. No, your point is, is very well taken. I have a big thing of change. in my car for just such a purpose. And I'm convinced that, it hasn't really been touched in years. and the only time it's going down is maybe some valet guy is taking a quarter or 50 cents sometimes, but I'm certainly not using it any other way.
Well, Dr. Delgado, thank you so much for spending time with us and talking about the Penn Med Nudge Unit and all of your work, with it and outside of it, it's very much appreciated. Look forward to your scholarship in the future and helping us, be safer and better users of technology and healthcare in general. So once again, thank you.
Dr. Kit Delgado: Thanks, Craig. And just a very quick plug. It is that the Penn Medicine Nudge Unit is hosting our annual symposium September 26th and 27th. Last year, we had over 220 people from many institutions across the country and across the world. I think this year will be even better. We have some great keynote speakers. And if you just Google the Penn Medicine Nudge Unit, you can find the conference page and registration on our site.
Dr. Craig Joseph: And I will attest that it's a fine, a fine institution and a fine meeting I went last year for the first time. It was terrific. I learned a lot. It was kind of shocking to be surrounded by hundreds of people talking about the behavioral implications that can be gained for healthcare. So definitely, would recommend that to anyone that's even remotely interested. Definitely, you'll learn a lot and might meet some new colleagues as well. So, thank you for that shout-out. Definitely appreciate it.
Dr. Kit Delgado: Thanks, Craig. Appreciate it.