Originally Published: Feb 2, 2023
YouTube Video: https://youtu.be/eVPVnTJKWUQ
Since the advent of the electronic medical record, there has been tension between the many advantages technology can bring to the patient care process and the fact that more and more clinician time is spent facing the computer rather than the patient. As technology improves, we hope we reach the inflection point where technology can give time back to the clinician by automating repetitive non-core tasks. As healthcare faces the biggest staffing crisis ever, it’s not a matter of people or machines in healthcare but both working together. Listen as panel members share how their organizations and healthcare stakeholders are evolving processes, technology, and coordination of tasks to focus clinician expertise on patient care and alleviate unnecessary administrative burdens – ultimately improving the patient experience and clinician job satisfaction.
John Chelico, MD, National System Chief Medical Information Officer, CommonSpirit Health
Yan Chow, MD, MBA, Global Healthcare Leader, Automation Anywhere
Kenrick Cato, PhD, RN, CPHIMS, Professor, University of Pennsylvania/Children's Hospital of Philadelphia
The Great Resignation Meets Clinical Automation Technology At Once the Problem And the Solution
Host: We are gonna have our first panel today on a very important topic, right? It's called the Great Resignation Meets Clinical Automation. Really what we're gonna be talking about is the fight, right? The fight between technology and humanity, particularly at the clinician level. I think since the introduction of the EMR many years ago with all of its promise, a lot of unintended consequences have resulted and they've really bubbled up to the top in the past few years particularly with Covid, and layering on, you know, doctors and clinicians don't wanna be, you know, tapping out for whatever hours a day on a computer. But add to that, just the level of dissatisfaction with all of the administrative burden with insurance and not being able to get your patients exactly what they want, when they want it, and when you need it.
And so, not to be so negative, but we are gonna be talking about where things are today in 2023, and possibly how technology as Gayle was saying earlier, right, how, how can we make technology bring delight and joy back to the practice of medicine? And so we've got a phenomenal panel. I'm going to briefly introduce them and then they'll tell you a little bit about themselves.
Then we'll get started. So from my left, we have Dr. Yen Chow, who is the global healthcare industry leader at Automation Everywhere. Next to Dr. Chow, we've got Kenrick Cato who is also a PhD and a professor at Penn. Now. He recently left New York and he was recently at NYP. And now he's at the University of Pennsylvania and chop.
And next to Dr. Cato, we've got Dr. John Chelico, who's the National Systems C M I O for Common Spirit Health. I've known John for many years when he was in New York at Northwell. He's still in New York but now at Common Spirit. So thank you all. Would love to start with you, Dr. Chow. Just a brief intro, and when I say brief, I mean like one to two minutes.
Yan Chow: Okay. Only one to two minutes. It's hard for the older people. .
Host: Hey, I'm, I'm pretty old too. All right. So, yeah.
Yan Chow: So I'm a pediatrician. Yeah. I spent 32 years at Kaiser Permanent. Having a happy pediatric life. But the last two eight years I was a national director for Innovation Advanced Technology, where we looked at over 2000 startups in Healthtech just to inform the Kaiser strategy and a very fun time.
We had a big innovation center and a big program, and then I left Kaiser to become a Chief Innovation Officer in, in Washington, DC at a consulting. working with the VA and the dod. At the time, they were looking at the choice of EHRs, which is a really interesting time. And then after that I spent two and a half years at Amgen as their digital medicine lead working on digital trials and sightless trials remote sensors, things like that.
And all topics that are very timely today actually. And three and a half years ago, I got a call from Automation Anywhere launching a greenfield initiative for healthcare. . I couldn't, I couldn't pass that by . So it's great, great great prospects in automation. So I joined three and a half years ago. Today we have a dedicated team and seeing a lot of response, mostly pushed by Covid.
So, lot of issues of course, but I love being on the cusp of the transition to next generation healthcare where I see myself.
Kenrick Cato: Hi, so good morning everyone. My name's Kenrick Cato. So I am. A nurse in my nursing area was in emergency nursing and oncology nursing. Also before I became a nurse, I was a software engineer and a data engineer.
Had a little segue in the army. Spent 10 years as an infantryman. And what I spend my time doing is basically trying to figure out how we. , take the data that's in the EHR and repurpose it to help support clinical decision making. So I spent a lot of time first at, at Columbia New York Presbyterian and now at Penn and Chop just basically mining data and then building software on top of, of the, those data to help support clinical decision making.
And I'll talk a little bit more about it today, but one of the, the things that I've been really involved in, in. About three years now is a national initiative called 25 by Five, where a whole group of folks came together and said, there's a lot of people talking about documentation reduction documentation, burden reduction, and everyone's doing their own little part, but we wanted to get everybody in the same boat rolling in the same direction.
And so, we created a, a symposium series around it a few years ago, and. , the American Medical Informs Association of which I'm a board member is sponsored a national effort where we have all of the vendors and a lot of leading physician groups and also some policy folks as well.
Trying to make sure that we can reduce today's documentation burden by 75% in the next five years.
John Chelico: I'm John Chelico. I'm I guess a software developer to begin with in the 1990s. Worked for two Health IT startups during that time. Ended up going to medical school at, in sun Downstate. Ended up doing formal training in internal medicine, followed by an informatics fellowship.
I sent, I did an informatics fellowship when it wasn't cool, but now it's cool. , . But followed by some amazing opportunities for me to be. It's Chief Medical Information Officer of Bellevue Hospital here in the city. I was the director of event of informatics for the NYU Lang. Epic implementation in early with the early stages and past 10 years.
I was at Northwell Health, both as the C M I O for North Shore University Hospital, long Island Jewish Medical Center. And then for the last six years, the chief informatics Innovation Officer for the health system where we started the Center for Research Informatics and Innovation at Northwell and Tour the Einstein Research Institute.
About a year ago, I broke out of my New York bubble and joined common Spirit Health which is I went from the largest healthcare. in New York State one of the largest healthcare providers in the country. And for those of you who don't know Common Spirit Health we're two health systems two faith-based health systems Catholic Health Initiatives and Dignity Health that came together in 2019.
We are in 22 states, or 150 hospitals, 1500 practice locations. We have about 25,000 providers that we work with across the country and about 25 million patients that we care for. in some of the most populated parts of the country, but some of the most less least populated parts of the country as well.
So really our mission is to really use technology. to really bring care to for those that that need it and, and where, where they can't have it. And I think that's really the, the, the goal of Common Spirit, I think as a platform. I think when Successful a joke we had from the other day when successful, I think you know, we'll have the big largest operating integrated delivery network across the country, which really leads well into value-based care and other.
Host: We have three panelists who are truly unique in, in healthcare, right? You, you each really embody this combination of clinical expertise and technical prowess and expertise. The vast majority of the healthcare system is not like that, right? So just to set the stage, In each of your institutions or previous institutions, what is, what's the temperature level like now with both physicians, nurses, techs, just in general, what's it like and what's driving that?
Do you wanna start, John?
John Chelico: Yeah. No, I mean, I think it's, it's much like what's happening in the rest of the country. I think the, the title of this, the Great Resignation, has led to significant staffing shortages. I think we're all across the country sort of looking at changes in the way we, the volumes of patients that we took care of or the different dynamics of the patients.
I think there's a different ex sort of expectation from our patients and our, even our providers really post covid. . You know, I wanna say it's, it's a, it's a difficult time, but, you know, I, I haven't had a year that in healthcare for 25 years, that hasn't been a difficult time . But to be hon to be honest I, I mean, I think there's unique opportunity.
I think there's a u unique opportunity to use our technology and people are embracing technology and I think Covid has sort of, you know, got us to where we wanted to be. You know, we, you know, five years from now in some places in telehealth and other things. But I think the expectations is that we can do a lot.
sort of asynchronously that we would've otherwise had to do in person and other things.
Yan Chow: Maybe that's, yeah. I think in the organizations that we talked to, and we've talked to tons of them, the number one concern is labor shortage and therefore labor shortage leads to burnout. it leads to, I mean, I saw one statistic that 60% of the healthcare organizations in this country have over 100 physicians in their hospitals open that they cannot fill.
And you just imagine. Burden that places on the people that are still there, that they basically don't wanna be there. And so, there's a lot more openness to technology because we don't have a choice essentially. Covid sort of brought that on us. We knew that was a problem going forward for many years, but Covid really brought all this to the floor.
And so like I was saying, I think there's a lot of opportunity.
Kenrick Cato: Yeah. Yeah. I, I mean, I, I, I would add to that, you know, when I, especially, I talked to a lot of nurse leaders in a lot of organizations, and quite frankly, everybody knows what I like to say is that the wheels have fallen off the bus in nursing, basically.
You know, it was a kind of perfect storm where there a lot of nurses, the average age of a nurse right now is 52, I believe. And when Covid hit a lot of nurses were planning to retire. The average age of nurses were actually higher than they stayed on for a year or two, and then they left. I can't remember the statistics, but something around 25 to 30% of new nurses leave the profession after a year of working.
nursing has some of its unique challenges to, to deal with, I think. But um, you know, similar to what the other panelists. I like to think of abundance instead of scarcity. And so this is an opportunity. One of the things to deal with and, and nursing as well, is just innovation in how we do things.
And so this is an opportunity. People are a little bit more open to innovation because of the challenges that.
Host: Okay. Well, that, that's a really positive thing that folks are open to innovation because I think what many of us heard for years was the physicians were getting really tired, right. Of the old technology.
So can you give us some examples of what's actually working and making a positive impact in automation and, and in that opportunity that we're talking.
John Chelico: Yeah, I mean in, in every facet of your life, whether you're calling a card, delivering your groceries, I mean, there's an expectation that we can do a lot more sort of, you know, independent of.
You know, showing up at a store, you know, getting in the car or whatever else. And I think that expectation has really kind of led itself to, to healthcare and I think it will cross common spirit. I think one of the things, you know, we have staffing shortages. We can't fill spots that are vacant in our clinics across the country.
And I think, , that has really kind of said, well, what, what? Burdens can we sort of remove from front end staff? I mean, I think that automation has sort of led its its place to actually create the virtualized environment where you can actually check into your visit, pay your copays, do all your insurance.
It's almost like a very low bar compared to everything else we do in technology. But it's something that like we are sort of saying as the standard and we're doing that in all our EHR platforms and everything else across the country. And I think, you know, it's little things like.
wanting ourselves to sort of the earlier, our earlier speaker is that is the sense is that, is that that sort of opportunity lets you sort of get to the doctor's office more efficiently, get all of the sort of the front matter out, get all your consents out of the way and really kind of jump into whatever else.
And that's sort of even now getting to another level is that, you know, we've actually sort of, Hey, you're here for a headache. You're here for abdominal pain, you're here for whatever, you know, let's start the. early, we know what we're gonna ask you. So let's just ask all those things ahead of time. So, so having really an intelligent sort of clinical intake of, of a patient patients are expecting it.
I think the opportunity for us to sort of start the visit at a, at another level where we only have limited time with you is the opportunity really for our providers, our, our staff, and others to really kind of, you know, take things up to a notch. that is really only lend itself well through, you know, automating some of those tasks.
Yan Chow: Yeah. And I think just to add to what some of you talk, talked about doctors hating the old technology. What happened was that when, when the government pushed everybody into electronic health records and they just were certifying records like crazy and, and everybody was buying records, what they did was move the old paper processes into the web
That's what they did. They didn't think. Is this the right process? Should we do anything with it? So, so we have done that. And so of course on the web, on the software now we get a lot more data, right? So now the problem is a lot worse. What we used to do with paperwork is a hundred times worse.
So no wonder they hate it, right? So the next step is to like, think about healthcare. Like what should, what should it look like, right? I mean, I think we haven't done that work yet. I think that's a hard, hard thing to do because with the new technology, healthcare could look very. , why should we humans be doing machine work?
And so, but I think it's gonna take a while. Cause everything takes a while. in healthcare.
Host: Well, can I ask on that what are the incentives for health systems to, to do that? Right. It's so clear that we need it.
John Chelico: Yeah.
Kenrick Cato: Yeah. So I, I think I'm forgetting her name right now, but she she just left , hca, she was the chief nursing executive.
And one of the one of the things that, that they did at HCA in the last couple years around nursing is they quantified how much it actually cost to replace a nurse. And that actually happened just before the pandemic. And it was a real powerful kind of cost benefit analysis because it, it allow, It, it allowed the leadership to free up money to invest in automation.
They had started documentation reduction work three years ago before the pandemic because they, they re recognized that it actually costs more money for people to leave because they find their work burdensome than to, to fix some of the things and retain them. So I think that's a really good example of.
Yan Chow: you know, what's important and also just I'm
John Chelico: sorry. No, I, I mean, I, I was gonna say that, I mean, I think there's things that are pushing us in that direction as health systems. I think during Covid there was sort of a lax in the, you know, in how we bill and how we document. So some of those things have sort of, sort of had that tail end in the sense that, do we need to record a review of systems?
Do we need to do other things on every visit? I think some of those things are now, you know, we had, we had the opportunity to. take care of patients without sort of having all the documentation burden. Some of those rules change, some of those things are still sticking. Some of the changes in telehealth, obviously with renewal of how we bill for telehealth allowing us to do that.
So the drivers are really yes to health systems. The drivers are some of the EHR burden, but some of the drivers of what EHR really are are billing systems, right. Billing revenue cycle capture systems, right? I mean, as a primary care provider or as a hospitalist in the hospital. I mean, I would rather just block my note rather than, you know, you know, really what I want to tell my next provider or next to all the patient is everything's the same.
When I raise your blood pressure medication, I'm done like, but I have to write a three page note to drop a bill. Huh. And, and like those are the things I think that were sort of pushing us in the, in, in the, and that those things are, have changed a little bit during Covid, but I think under, under sort of the current sort of fee for service, again, borrowing from our current, our, our last speaker.
I mean, I think once we get to push to value and the opportunity for us to sort of unburden ourselves from sort of proving ourselves that you, we, we did all this during a visit is really gonna change the dynamic. And I think that's where, you know, I.
Yan Chow: They're going and, and to the point of value-based care.
So all these organizations have signed into that arrangement have to maintain the quality of care with fewer people, . And so, so that is a driver. The other thing is there are a lot of new regulations that have come in the last couple of years, like visibility to peers, providers that we come in have come no surprises, act, information, blocking rule.
All those things require a lot of resources, which organizations don't. And so we're finding that for many reasons, it's driving organizations to really look outside their traditional solution, which is to hire more people and even hiring more people to the nursing point. I've talked to organizations where contract nursing is so expensive and they're not devoted to your organization.
They, they work and then they leave and they get to retrain the next one. So, there's a lot of pressures that are making. in a desperate situation and people need to change.
Host: On the provider side, what are some of the tools for automation that you've seen actually, that they like and that are being embraced and are working the way they should in, like, in other parts of our life, right?
John Chelico: Yes. Uh, Where things help us.
Kenrick Cato: Yeah. So, one of the things in both the org, the organization I just left, you know, Presbyterian and chop, where I just land. , they've gone in really big and secure chat. And that's something that I think has really changed, you know, in the inpatient setting, at least being able, it's, it's a simple thing, but being able to communicate with your care care team has really been helpful, I think, so that, you know, obviously not every institution has the resources to spend the, the big bucks for that.
that is a technology that I feel has, has been a game changer. And it, it actually from the hospital innovation side, it, it's driving the ways that we think people can do medicine. You know, because we have people that are doing all kinds of interesting things right, on their iPhones. And then it, it makes us think, oh, they don't actually need to go on the EHR to, to do that task or get that inform.
And so I think because of Covid, we haven't exploited it as well as much as we can, but I think that's something that has been working and we'll continue to
John Chelico: change things. You know, we, we often talk in on the provider side, sort of everyone practicing up to their license. I mean, oftentimes we have providers or, you know, docs, apps, and.
PAs, nps doing things that they, otherwise filling out paperwork and other things even from our nursing staff or our MAs. I think, you know, we'd see a lot of promise using robotic process automation to do sort of those redundant, mundane things. I mean, if a provider can have a chart reviewed and, and have the question answer with the provider, with a patient prior to their visit about their cancer care or cancer, Right.
And really at the time of the visit really just review that or just tee up a colonoscopy or a mammogram or whatever else. I think we've seen significant improvements in, in a lot of the, the, the redundancy or the other things like that a provider would need to do five minutes before every visit. And I think that's I think where we can automate things and again, doesn't, does a, does a provider need to do that?
Does an MA need to do that? Does a staff, an administrative staff person need to. Probably none of the above. We can really kind of get that to the, into the pa, into the patient's hands, and then again, free up the providers to sort of work to the top of their license to do the best things that they can do, or the nurse or the ma or whatever the case may be.
I think all of them can do more and want to do more sort of meaningful things. versus sort of the things that are sort of, you know, asking every patient if they smoke or not. I mean, like, or asking every patient when they got their last colonoscopy. Easy peasy things that you could really do and expect to do sort of prior to your visit.
So I think that's where I think we're, we're definitely you know, seeing the promise.
Yan Chow: Yeah. Yeah. I think I think physicians are, and nurses are in a place where it doesn't take a lot to make them delighted to be, to be honest. And the the biggest success stories are like, just a little thing that they just hate doing and you automate it, it.
Oh, we love you forever, .
Kenrick Cato: One of the things is that where all of life
Yan Chow: Yes, yes. Start small medicine is target rich. I mean, it's target rich. There's so many things. You can't automate everything, right? So, so you look at it and you go, okay, what makes biggest difference? So financially, morale wise, whatever.
And one example is that in the, in the UK know we worked with the NHS National Health Service. So in the middle of Covid when patients were flooding into the. nurses had to check the oxygen tank levels for Covid, obviously. And so they, they would take bit of, little bit of time, but they check for every patient.
So they wrote a bot in 12 hours to check it for them. A hundred percent accurate. End of the year, it saved 1500 hours of nursing time. That's huge in the middle of Covid. Right. So that's the kind of stuff you don't need a, a big thing, just a little thing, you know,
John Chelico: it works.
Host: I, I love that insight by the way.
We often in on the text side think, you know, like go for the bells and whistles and all the cool stuff and we will talk about chat, G B T. Don't worry. But it's often the little stuff like as a patient just getting a text reminder Right. That my appointment is tomorrow. Yeah. Excellent.
John Chelico: You know? Exactly.
Kenrick Cato: You're gonna say something. Can I just add something? Yeah, sure. So John had talked. Offloading things to patients. I think it's really important, especially in the tech area when we think about it, to make sure that we meet patients where they're at. Mm-hmm. and understand patients. You know, I had, I had a experience, I had volun when, when the Covid vaccine first came out, I volunteered to to give injections.
And I was at the armory up in Washington Heights, giving in. And this woman came in and she was clearly not that healthy. And she only spoke Spanish. And I started talking to her and she had a caregiver with her and I found out that she'd been trying to, she'd had a stroke and she'd been trying to schedule follow up and she couldn't because they were only calling her in English.
And no one, no one in our house spoke Spanish, I mean, spoke English. And it was actually, it was a good example of secure chat. You know, I used Secure Chat to talk to all of our clinic. And, and it also turned out that her, one of the clinicians she was trying to contact had just left the organization.
But I just think, you know, we, we really need to make sure that the technologies that we're using for patients and, and the things we're asking patients to do to understand what they, what they, their technical capability is, and making sure they can actually do it as well.
John Chelico: Yeah, I would say, I would say the same of our providers.
I mean, I. Different pieces of our providers sort of uptake, sort of different pieces of technology more than others. I mean, and I think that's something where you really have to sort of take into consideration of people's preferences Yeah. And how they like it. You know, we were talking the other day about pajama time.
Right. You know, we said, oh, providers, You know, have increased pajama time. They go home and document stuff, and we have some providers and, and I, you go to them and be like, well, why do you have so much pajama time? He's like, well, I wanna make the soccer game . I want to go home and I'm fine. Sort of going home and taking, you know, making myself a cup of tea and finishing my nose.
you know, but, you know, and I, I think that there's different sort of preferences in how people want to work, and I think that flexibility is something where many of us are now working from home are, are it changes in sort of the dynamic. Yeah. But people need to sort of understand, you know, how, how that sort of technology could be best used by that person.
Host: How are you finding the adoption of this hybrid virtual care? Right. Obviously, COVID you. Virtual care, digital care like exploded. And you were, you know, earlier you were talking about how we just, we had paper records and we just digitized them. Right? We didn't think about the process. Many are saying the same about virtual visits, right?
You're just putting a doctor or nurse online and not thinking about the process. How are you finding your clinicians reacting to it? Adoption. Where, where has it gone since Covid?
Yan Chow: I, I would say it's probably early . Okay. A lot of skepticism and even a lot of uncertainty, you know, with Covid, people had to do telemedicine, remote sensing.
But I was talking to my doctor who I know and, and he said you know, he had one patient, for instance, that was a woman who complained about chest pain or something. And, you know, typical didn't sound very serious, but somehow he had the sixth sense to make her. and it turned out she had a problem in her abdomen, which is pretty serious, which you can't tell from tele.
So things like that, and the fact that there's really no best practices defined yet for Telemis to make it kind of risky, you know, so you're kind of on your own as a doctor, you're licensed on the line, you know you're practicing that way. But I think it is a future, we'll have better technology in the future to understand patients better.
And I don't know that younger patients don't prefer the virtual I. To in, in person. It's fairly interesting right now.
John Chelico: That's true. It's a li It's a little bit of the information overload. The fact that you can automate stuff and have that asynchronous visit. I mean, if a provider doesn't have sort of a time set aside for his day to review all of that information coming in, you know that patient who said they have chest pain or the patient says, you know, Feeling well, or on a behavioral health survey or something to that effect.
And, and it may be something, some sign, something more imminent. Those things I think are, are, you know, again, they could be just now be fed into the chart without really, and, and that's concerning for us. Mm-hmm. , I mean, they could be fed into the chart. , there's documentation that that person had chest pain and never was addressed.
And I think that's always been sort of the, the, the, the piece. And I think there has been sort of a couple steps back, especially in, in, in some things around behavioral health and others and where you, you, you really have to have the conversation. You really have to have the one-on-one. And I think that, that, you know, you can automate so much, but I think.
There are some things where you, it's a feeling. But I will say in, in sort of my prior health system, I think we played a lot with early warning scoring systems, and I think that's something where you could really take a lot of surrogates for, for some of that stuff. You know, in like an 800 bed hospital, we, you know, you can't have a nurse in every bed.
You can't have a, a doctor at every bed. But if you have an MA going around taking vital signs and taking a, a couple other things where as a, as a nurse or a doc, you walk in the room, you're like, that patient looks sick. Well, what are you saying? The patient's, the hip, the patients, you know, di maybe you know, a little bit.
Di di di diaphoretic or whatever. And, and I think you see that, or you, you kind of interpret that. But, you know, could we put all those things together and ring the bell early? Where you have, you are collecting vitals on a sort of ongoing basis. Well, could we put that all together and ring the bell early?
And I think there's some hope in, in really kind of putting, piecing that together. But what's the version of that for someone who's feeling depressed and, and alone at home? Like, is there, are there other things that. Take into consideration from other data that we have that can maybe clue you in that this person needs help, that person may not, or you know, I, I often say just simple things, but like the sensor in your house saying, Hey, mom didn't open the refrigerator today.
What's going on? Like, that's all I need to know. Like if the refrigerator was open today, like someone went into the kitchen, she got up, everything was fine. But like, and I think those are things we have to kind of think about. .
Yan Chow: Mm-hmm. . Also, the other thing is talking about decision support. I think the augmentation of physician nurses is great, you know, so, so, warning science scores that prioritize your patient, this as you round on patients, things like that.
But when it comes to the area of decision support Kaiser actually did an experiment years ago where they had a very complex mathematical approach to recommend what to do with the patient at the point of. And they rolled it out in Hawaii and to, to universal , to universal skepticism. And doctors didn't want to use it because the patient would say, why do you say that?
And they couldn't explain it. Right. It's the explainability issue. And yes, the AI is looking at 300 factors. But you're never gonna explain it to the patient. So the human part is really important. You know, the doctor has to be able to say, this makes sense, and it's why, why did it recommend this or not that?
And it's, it's you need to do that for the patient as well.
John Chelico: Now, one other, one other thing comes to mind I think, is that you can sort of have the normal conversation with your patients. Mm-hmm. , but really kind of put it in context. , you know, how could the computer help me document this in like you, you know, if you just put the computer aside, have a conversation with your patient, you as a doctor now have to spend that extra time after the conversation to document what you just heard.
And I think there's now, you know, this sort of push towards ambient sort of Yeah. Documentation where, you know, the bot or whatever else can sort of be listening in the background, really interpreting. Filling in the notes, doing all those check boxes you need for your insurance companies . But I think that there, I, I, I haven't, I I, we're starting that kind.
We're starting that, but I, I can't say it's entirely there, but like, could you have Yeah. The good old, traditional thing and, and, and, and not burden the doctor to rush to write his note before he sees the next patient. And I think that's something to think about or. not having him do it later that evening or whatever the case may be.
Mm-hmm. . So again, there's hope, there's s of, of an opportun of many opportunities, but you know, what mixture of things that are gonna make it a good sort of scenario.
Yan Chow: Even years ago, I, I thought that the ideal EHR is where you walk into the room with a patient, there's a video camera and a microphone, and you just do your thing and all of a sudden the stuff magically shows up in the chart.
Yep. almost done. And you just have to reprove it. Right.
Kenrick Cato: But yeah, that would be ideal.
Host: The technology, when you're recording right a session or anything, you can get it , you can get it transcribed. Yeah. Within about three minutes with multiple services. Right. It's incredible.
So I can't see how this would be far behind where you can, so that brings us to ai, which of course everyone loves to talk about AI and now, We've got chat, G p T and generative ai which is the buzzword for 2023. Just if you, if you don't remember anything today, just remember generative ai, you will hear it a lot.
So do you see, what do you see for that in your organizations?
Kenrick Cato: Yeah, so I mean, I think AI is one of those things. I spend a lot of time thinking about it because I do that kind of research and I also hate it at the same time because I, if anybody's old enough reminds me of, you know, the internet in the ni early nineties.
You mean the interweb The interwebs, yeah. . In the sa in the same sense that we really don't know what we don't know. At the end of the day, that's what I think about ai. First of all, it's not new. You know, a lot of the methods we use have been, people have been doing it since the late sixties, but it, it really I think a couple of things from my perspective are really important, especially in nursing.
One is that that clinicians are involved in from the beginning of the development of any products that we're talking about. You know, we find in engineering there's a saying that, you know, a solution in search of a problem, and we find a lot of that. I think in artificial intelligence Jan mentioned, you know, the, the explainability part of it as well.
The explainability part is really important because the AI can be it can, it can tend towards something that the, that's important to a machine, but it's not important to a human right. Like it can generate inflammation. And so I think all of those things are really important. And, and then at the end of the day, just like the inter.
I think it's really important to understand where we are and what we can actually do with it, as opposed to what, you know, whoever's trying to make a lot of money is, is selling us that we can do. But, but there, there is obvious power in ai. It just, I think people just need to be very knowledgeable and there needs to be clinician expertise infused in it at every step of the way.
John Chelico: Yeah, I mean the, the one sort of fear factor. of it all really is the fact that, you know, no one place has all the information necessary to sort of, you know, feed these algorithms. In the end of the day, we haven't solved the core pieces that we don't have all your data in one place or all in one ehr.
You know, it may sort of suggest you may need a mammogram, but you may have gotten your mammogram somewhere else. That's not documented in the chart or, or whenever the case may be. This technology, any technology can only be best. The be the best it can be is, is to kind of have access to, you know, all the data that it has.
I think, I think that's the problem. That is the problem. I mean, it's old technology, but we haven't solved the problem that we don't have one ubiquitous record of your life or, or everything. And I think we don't have our information sharing together very well between healthcare systems, whether you call it a technology.
Culture and politics problem. But I think that is the ness of it, is that these are not, you know, intel, they're only as intelligent as, as the information you give them. Yeah. And I
Kenrick Cato: also want to, you know, so I've done some work on de deter on using machine learning to predict deterioration. And one of the really eyeopening things I found was that the machine basically replicated bias in the clinician.
So we found that, for example, white patients are, they're clinicians are paying much more attention to white patients that are deteriorating the non-white patients. Mm-hmm. and something like that. You know, of course we adjusted for it in our calculations, but something like that, it's important to understand how the models are working so that you're not replicating those kinds of things.
So that's what, when I point out about really being knowledgeable about how it's working. and having clinicians involved is, is really important because you will, you know, anybody who's ever worked, if you've ever let a chatbot off go on the internet and try to learn, you know, it comes back as a horrible, racist chatbot, right?
Like, and so, and that's just what the internet looks like. So I think it's important for people to, to really understand the limitations, really understand how it's working, so that you can produce some beneficial outcome.
Yan Chow: I'm just kind of curious how many people have tried Chad g p t? Just
John Chelico: raise your hands,
There
Yan Chow: you go. It's great. I mean, I've tried it and and then now you know, there's a Princeton student that wrote a program has zero G P T to find out if it's really rich by Chad, P G P T. So basically have, AI is fighting each other, right. So I'll have my AI in the morning . So I think it's a really interesting time and the issue.
One of the issues chat gpt, just like when, when you talk to a doctor, when you talk to your doctor, you trust that he or she has your interest at heart. Mm-hmm. , they've been trained, they're certified, the organization says it's certified. They
Kenrick Cato: have experience. When you
Yan Chow: hear something from chat G P T, you don't know where it came from.
I, that's the issue. What's the proven of the data that it used? I mean, you may not agree. And so I think if you present that as a patient interface or a doctor interface, we need to make that. So people understand the limitations.
Host: Well, I'd love to hear from the audience if there are any questions.
So the question was from, from tools that ask patients to fill in information before the visit, and I've done these how much is the doctor willing to give up Right. Before the visit? And I'd like to add to that. How much does the doctor actually look at it before the.
John Chelico: Well, they don't have the time.
I think that's a good question, . I mean, to, to be totally honest, I think most of the stuff that's automated are the, sort of the mundane it's, it's your registration, your insurance, your copay, your consents for treatment and other things, right? I mean, doctors have no issues and front, front test staff.
We'll hug the patient when they come in instead of ask them all those questions. So none of that. I think we are having issues with sort of some of the pre-visit planning as we call it. A lot of the, you know, the question and answer, right? I mean, . Again, it's sort of, very robotic and I think we can only do decision trees as best we can to, to say, say, we may call it ai, but at the end of the day, it's like choose your own adventure , and it may sort of take you down the wrong path in, in questioning or whatever else.
So oftentimes, is it wasting time with the provider that they have to review now something and now rejigger the, the. Conversation you just had with tech Bott online? So I, again, I, I think, I think for some things it works very well. For some things. I think we need a little bit more like I said, like in behavioral health, we need a little bit more work.
So I think that, that there's, there's, there's certain just things that are simpler to automate than others. So I, I mean, most of our providers are receptive in the fact that like this is helping them and it can help them, you know, if we can incorporate into their documentation, which is burden. You know, have that hpi that, that, that pop into their hpi part of their note is, is a key, key thing that I think people are welcome to, but they have to get used to
Yan Chow: it.
The other I'd like to answer that too. The other issue is when I was trained as a physician no matter how many pre-visit things that they went through, you were trained to, you still have to ask the same questions. And that's one of the issues about patients, you know, why didn't I tell this to the computer or the nurse?
Because when you ask it yourself, you can detect other things, right? Like, did they really intend to say what they put down? Like every patient wants to put on what you think they, what they think you want to hear. And so you can tell as physicians, so you ask it to make sure so that when you do your recommendation, you're actually recommend recommending on truth.
And so, that's an issue, you know? I mean, that's an issue.
John Chelico: It's like I took my mom from the doctor last week. Significant spinal surgery over the past couple 60 years with 23 screws in her back and everything like that. She complains to me every day we go to the neurologist like. Tell 'em, oh, I'm fine
And I'm like, no, you're not fine. . Oh, yeah. Or like, it's like, it's, it's like, it, it's like I, I, I'm like, it's funny, I can't, my mom can't go to the doctor without my dad or me in the room because it's like, she's just from the old school of like, I, I, I'm fine. I'm like, but you're not fine. . Like, they're here to help you.
And I think that it's, it's funny, it's like, try to read that between the lines,
Audience Member: Okay, so, you guys talked a lot about the problem and the, a little bit about the automation. Maybe I missed it, but I'd like to hear about your favorite things that you've seen in your systems and your experience lately. For where automation or AI or. Chatbots or what have you have really just helped and had a really positive impact on reducing, you know, the fourth of the quadruple aims.
Kenrick Cato: Yeah. It's so great question. Yeah. It's not in my system, but a colleague of mine is who's in St. Louis university of Washington. They have in their hospital. Basically somebody had the great idea to put an Alexa in the. Supply room. And if you've ever worked as a nurse, which I have, you spend a lot of your time running around getting supplies and you know, clinicians are, you know, spend a lot of time and, and basically it's a closed system that they built with a, with a, with a Amazon smart device where basically they tell it that they need a new supply.
It goes it and it sends it down to to the supply. Supply room and, and, and, you know, so in real time, instead of going to the next unit and stealing their supplies, , you can just tell the smart speaker that you're running low on whatever, and it helps, you know, make the, the supply chain a little tighter.
So that's something that's relatively simple, but and they've, you know, they've done the metrics on it. It saved lots of time, but it makes people very
John Chelico: happy. I would say one of the goals in Common Spirit Health is we have, you know, again, hospitals in 22 states, some in the most sort of heavily populated areas, large academic centers, you know, but we have, we have care in places where it's like a desert.
I mean, you know, 10 bed hospital, none of care site for 400 miles. And I think you know, the opportunity and one of the missions of Common Spirit is to bring. To where PA patients are physically, mentally. And, and I think that's where I think we're, we're using technology to bring the special. To the, to the patients.
And I think that's something to really kind of really think through is that how could we have that expertise of you walking into an academic medical center? And I think that's been the biggest change of me leaving New York where I can throw a stone and hit another Ivy League medical school,
Kenrick Cato: John, before you, can you explain, can you just paint the picture of what some of these hospitals look like? Like in
John Chelico: North Dakota? Think . Think of a 10 bed hospital in North Dakota with. with 10 beds. Two are ICU beds and one doctor covering the urgent care, the urgent care of the ICU , the 10 beds, PA patients in the ho in the hospital and the primary care clinic.
It, it's really that. I mean, that is where we are and we're not so farfetched. I mean, you take. , some of our practices in Dig, dignity, health, and you go 20 miles out of San Francisco, you're in rural America and, and you have these sort of deserts of care. And I think that's where, you know, we may take it for granted or we may think, oh, that doesn't happen in this country.
But to be honest, like. You know, using technology and, and where we can to sort of, again, ring the bell when someone needs help or, or, or ring the care or the, the tele provider to, to that, to the, to the foot of the patient. You know, may, you know, may ultimately sort of save them a 400 mile track to somewhere else.
Or even getting in a plane to get care. Huh? You didn't actually.
what I mean to, to, to that, to that point. I think that's where I see, I see huge successes in how we introduce technology.
Kenrick Cato: Yeah.
Yan Chow: An example. Yeah. So, so what was the question again? I'm sorry. Big example.
Kenrick Cato: Tell us about the innovations you've seen.
Yan Chow: Yeah, actually probably in, in my time at Kaiser and was innovation director, so. 2,500 startups and probably one of the most interesting was there are two that are really interesting that stuck still stick in my mind. One was a optimi optimization startup where they would look at all the hospital processes and automate, you know, where people should be and should be doing what and so on and so forth.
And they did in San Francisco at one of the community hospitals, it got them an extra hour of or time. That's huge for a community hospital. The problem was surgeons hated it. . Cause there's no more coffee time . You have to be fully working all the time. Yeah. And that's the human part of it, right?
John Chelico: Well we, we've used things like, I mean, some of our primary care clinics have, you know, RFIDs on all the patients, all the doctors, all the equipment.
And I think it's, it's been really, you know, things like this patient's been, hasn't interacted with somebody for 10 minutes. Yeah. And having someone say, Hey, the doctor's. or things like that have, have really kind of mm-hmm. felt, and people have really felt a sort of a, an understanding of what's happening.
No one ever intended that to happen, but at least be able to sort of catch some of those things and whatever
Yan Chow: else. And the, the other thing I saw that was really interesting was that not a tech innovation, but a business model innovation that was supported by tech. And the idea was for the start, it was that you have these home health aides that go in and see people at home every.
for almost every day. They know what the people look like. They know any changes in the skin, the, the, the temperament and so on and so forth. So why not empower them with an iPad that they can just check a checklist. This is patient having a skin rash like this. In fact, it doesn't, it wasn't even in English, it was just pictures.
And so the check to check, cuz these folks are like, they're from other countries, you know, they, the home health is, they're lowly paid. , it really improved the pickup of the early signs so that they could prevent, because anything they check is gonna go to nursing. Case manager is gonna call the patient.
John Chelico: It bring, it brings to mind sort of, we're always talking about technology innovation and things like that. Mm-hmm. , at Long Island Jewish Medical Center in Long Island at, at, at Northwell Health, we, we really had a brand new, built the emergency room in the sort of the borderline of Queens long Island and.
we built this, this emergency room to handle 80,000 visits a year. Right. But the way things are going, you know, we knew that that emergency room needed to sort of have 120,000 visits a year in the upcoming years. We weren't gonna rebuild the emergency room. So we kind of did, we did something around something called split flow.
So if you actually come into the, to come into the, the emergency room walking , you kind of go down this path where it is like sort of a, a process where we, we. , we ask the right questions. We do everything. Tests are ordered, blood work is ordered. You go to the waiting room, but instead of calling it the waiting room now, which was this beautiful, you know, cathedral ceiling and lit up place, you're not just waiting for you to see your doctor.
Mm-hmm. , you're, we called it the results waiting room, and the waiting room was now only us. Six chairs at the entrance of the ED and that dynamic of allowing us to sort of change their, that waiting room, which now like, I don't know, 60 people in it to beds. I mean, that was the technology innovation where that.
That chair became a bed. And so if you came in lying down , you went down the traditional path in the emergency room. But the ability for us to sort of change that dynamic of like, these are all the things we're gonna do anyways, once you see the doctor to this sort of what we called split flow, allowed that emergency room that was built for 80,000 visits a year to sort of go to 120 visits a year.
The only thing that came to me as C M I O was, was, can you make this chair a bed? ? And we're in the, in the ehr. And I think that's sort of, you know, thinking a little bit differently about how we just normally
Yan Chow: do things, you know, in another area that's kind of related during covid people might wonder, how do we get these covid vaccines so fast?
That was actually a very good lesson in terms of how you re-engineer an industry. One of the lessons is that they did as much as they could in parallel mm-hmm. , where they used to do sterile, they did parallel processes that use the same machine in parallel. And, you know, we haven't done that kind of thinking healthcare, you know, that's sort of process re-engineering, but it's
Kenrick Cato: a good lesson.
Host: Absolutely. So last question I'll ask and then we're gonna go to lunch. , we've talked a lot about challenges. What, what makes you optimistic in the next five years, five to 10 years, that things really will change? I'll start with you and just go down the line.
John Chelico: I mean, I think the fact that we see, you know, move towards value, I think the opportunity for us to change some of our documentation burdens with our insurance payers and things like that has really changed the dynamic of us really trying to do the things.
I think everyone is faced with the. Something's broken and we need to fix it. And the, the op, the, the move towards more openness with our EHR platforms and data being shared, all of these things are really kind of, really putting us in a good spot where, you know, I've spent now, you know, a good 25 years in healthcare IT and informatics.
And I think I've seen sort of now the promise of, of, you know, this is where we need to be. There's no other way to fix it. .
Kenrick Cato: Yeah. And so for me I'm hopeful because, you know, we've talked a lot about clinician burden today, but the leaders are burdened as well. And that's just been my experience.
And I think that, that there's a lot of, there's been a lot of complaining in my, you know, 20 some odd years in healthcare. But now that the decision makers are really burdened people are open to, to trying innovative things and, and, and thinking outside the. , which was an issue in healthcare before.
You know, healthcare is a, the way you're trained as a clinician, you do what you know works. You don't, you don't go out and do something wild and crazy, right? Mm-hmm. . And and not that innovation is wild and crazy, but but I think we're at a place where everybody's ready for innovation from the top down, bottom up.
So it makes me helpful.
Yan Chow: What I'm seeing is that I think the healthcare industry is opening. It's becoming more, data's becoming more liquid and so the government is really pushing hard and I think we're finally getting to a place where we may have free exchange of healthcare data, maybe after all the lawsuits,
So, so we have no surprises out information block and rule. We have all kinds of things that fire API and things like that. So I'm hopeful that healthcare will become much more free. and where people will get value for the dollar they're spending. Yeah. And we'll have this, we need do that. And as a result, the competition would be much better for both providers and patients and for patients.
So that's, for me, that's optimism. It took a long time to get there. Okay.
Host: Well, in, in five years we'll come back to this very place and we will check each one of those sounds. Meanwhile, thank you so much for a really interesting conversation and. Thank you all.