The Evolution of IT and Healthcare... Where Do We Go from Here?

Brian Urban (00:22):

Yes, this is the Healthcare Rethink podcast. I'm your host, Brian Irvin, and today we have ourselves a special guest, the Chief Medical Officer of Health Gorilla, Dr. Steven Lane, joining us today. And if you don't know about Health Gorilla, they're helping to change interoperability within the whole healthcare ecosystem through data exchange and speed. We'll get to know a little bit more about them, but a lot about our guests. So Dr. Lane, welcome to our little show.

Dr. Steven Lane (00:51):

Thank you, Brian. It's great to be here with you.

Brian Urban (00:54):

Well, with each episode, we like to get our guests a little bit more familiar with our audience and vice versa. And what better way to do that than to get to know who Dr. Steven Lane is before the MD and how you got to be the Chief medical Officer of Health Gorilla and everything in between because you've had a fascinating story. We've got to know each other a little bit before the show here. And I want to start there. So where you're from, how did you get into medicine, and tell us a little bit more about Steven.

Dr. Steven Lane (01:25):

I love it. No, it's been a fun ride and I don't mind telling the story. So I'm a native Californian, actually grew up in San Francisco, went to public schools, humble beginnings. Ended up going to... Really in high school, really loved thinking, knowledge, people, the arts, putting it all together. So very much a holistic view of the universe and my place within it. Ended up very excited about human behavior, the biological basis of our experience, our relationships. So went to UC Berkeley, where I wanted to study things like psychology and behavioral science and neuroscience. And actually even as a 17 year old, I jumped right into a behavioral endocrinology lab and was learning about the endocrinological basis of sexual behavior and beagles. I mean, it was great stuff and I just pursued that. At Berkeley, I ended up getting a degree in cognitive science before that was a thing. I actually got the first degree ever awarded in cognitive science, and now I understand it's one of the most popular majors at that university.

(02:49):

But really learned a lot in that space, very interested in health and healthcare. And then decided to go to medical school. I started out in an MD PhD program, studying neuroscience, thinking I was going to be an academic neurosurgeon. That sounded cool. But then partway through, I decided I liked people more than sea slugs and rats and the individuals I was interacting with in the lab. So I dropped the PhD, went back to Berkeley, got a master's in public health, which was a fascinating experience, had a great time there. Had a chance to do some extended travels internationally. Got to see healthcare in other countries. I was really excited about that. So I planned to go into academia still and had to choose a specialty. And at that point I decided I really liked family medicine. I liked doing everything. I liked young people, old people, sick people, well people, people of all stripes and colors. So family medicine was a good fit, found a great program, again, near home down in Salinas, California.

(04:06):

It was a public health hospital at the county level serving an indigent population, learned a ton of medicine there, and then came back to UCSF on faculty. But while I was a resident, I was doing a lot of moonlighting, trying to make some extra money and see more patients, and ended up working as an ER doc at the same hospital where I was born at Kaiser, San Francisco. And they had a little homegrown EMR there. One of the docs had brought a couple of the first generation Macintosh computers and stuck an ethernet cable between them, created a little network and the docs would sit at a station and we'd write down our notes. And it was amazing because when a patient came in, you could go find their last note and you could read it. And it really got me thinking like, wait, if we're writing this down in a computer, we could analyze the data, we could actually provide decision support, we could understand what's going on and maybe provide help in making diagnoses.

(05:10):

So I was 28, 29 at the time, got very excited about the possibility for computers in medicine. And then that ended up really impacting my direction. So at UCSF, I helped them tweak up the hospital information system to make it more helpful for clinicians. And then when I left there after a few years and went to work at Sutter Health down in Palo Alto, found a few other hacker docs who were interested in using computers to improve their processes. And we got together and looked across the landscape of health IT found a few vendors who were making electronic medical records. This was the 1990s, there wasn't much of a market at that point. And then we met this nice lady named Judy who had this nice little company on a farm in Wisconsin, and we were one of her first customers and built out one of the first systems at about the same time that Harvard was building an EMR and the VA was building an EMR. And it took a while, I mean, I had to build it myself, nights and weekends, configure the system.

(06:24):

But then we finally went live in the late nineties and had one of the first EMRs in the country. It was really very exciting. We hired some folks onto our team and we then implemented the first patient portal. Actually, Danny Sands at Harvard was implementing one, but his was standalone, ours was actually tethered inside of Epic. So almost all of your listeners have probably used the MyChart product. We implemented MyChart the first time and it was great. So at one point, I think Danny at Harvard and I had the two largest online practices on the planet and really learned a lot building, spreading, proselytizing, EMR, patient portal, patient access. So fast-forward, we got to the point in the implementation of EMR that the early adopters, the enthusiasts had all gone live, and we were left with the people who were challenged by the whole process, and we're still in that phase, but that was about 15 plus years ago.

(07:34):

And I said, I want to just keep innovating, dragging people along, kicking and screaming into the world of EMR was not really what I wanted to do. So at the time, I got very excited about interoperability. So the sharing of information between, really it was between providers. I worked in a clinic that didn't have a hospital attached to it. My patients, when they got admitted, they'd go across the street to Stanford. Stanford also became an Epic customer, and EPIC was really working hard at putting together the tools that would allow their customers to share data. So it was really the very beginning of EMR level interoperability as opposed to just writing interfaces between disparate systems. So worked hard with my colleagues in my region to put that together. Then we spread that across the state of California, got involved in governing that process at the national level.

(08:31):

And really that got me into more national level EMR and Health IT initiatives. Got involved in the certification commission for Health IT, the precursor to the current ONC Health IT certification program. Met some nice people there, got involved with implementing the eHealth Exchange connectivity for our system. And then we still had this problem like, yeah, Epic using organizations could exchange data, but when people went to a hospital or a health system that was using a different vendor, it was really hard to share data like next to impossible. So got involved with some friends and we put together Care Equality, which was really the first nationwide framework that allows the various networks to exchange data amongst them. So the Commonwealth Network, the EpicCare Everywhere network, the eHealth Exchange, can all exchange using this nationwide framework. So I helped to bring that up. The first patient exchanged on Care Equality was one of my patients. It was great fun.

(09:37):

And then worked with that at one point, got interested in contributing to the federal work that was going on. One of my friends was involved in one of the federal advisory committees, so ended up getting appointed to the first iteration of the ONC's Health IT Advisory committee seven years ago now. And I've been super involved there with ONC, their initiatives, driving forward, the 21st Century Cures Act, information blocking, fire, all sorts of things. And then part of that of course was TEFCA. So TEFCA is a component of the 21st Century Cures Act, final rule that really promises to take everything that we did in building a health exchange, Care Equality, et cetera, and take that to another level. Bring that into new technology, new web services, and really push it in the direction of serving a broader population of users.

(10:40):

So not just providers exchanging data for treatment purposes, but really moving towards getting patients more access, more access than just portal access, which we did 20 plus years ago. But really being able to use apps, modern APIs so patients can access aggregate, utilize their data, and then also more use cases. So again, treatment, most providers across the country now have EMRs. They're connected up to the framework or one of the networks. They've got some largely good interoperability to support treatment use cases. There are clearly lots of problems there. But beyond that, there's exchange with payers, there's exchange with public health, which we saw was a huge problem in the context of the pandemic. There are other purposes of use, when you think about home care, dental, complimentary care. I mean, the list goes on and on of folks who have had no benefit from all the work we've done in interoperability over the past 15 plus years.

(11:42):

So the opportunity to work with an organization that was really pushing the envelope on interoperability, really pushing to get TEFCA alive, that was exciting to me. So after 30 years in my practice doing family medicine, I decided to switch to a little bit of same daycare and really go all in with my informatics practice. And that brought me to Health Gorilla and it's been exciting. I've been here about nine months and we're doing a lot of really interesting work on TEFCA connectivity, data quality, compliance, data privacy, things that have been near and near to my heart for a long time, and now getting to focus on that from an industry perspective as opposed to the [inaudible 00:12:31].

Brian Urban (12:31):

Yes. And that's an amazing journey that you've had, Dr. Lane from... A lot of firsts I heard there as well, cognitive science degree, of course, standing up the infrastructure of the EMR, I'll say universe now. And then also now going through your Kaiser experience to having one of your patients actually be an actual record on one of the first EMRs that you've worked with and you continue to build out. And you mentioned compliance, you mentioned privacy and security, and then also access for patients. That's huge in where we're going now because it's a shared information space in which patients need to be aware of their treatment path and what they need to be educated on and what they can outreach for help on and how it's a shared space. So it's gone to this maturity that's almost like this civilization arc with inside EMR technology.

(13:33):

And you've seen all the milestones along the way, and we're going to get deeper into that. But one thing you mentioned being a primary care physician. And you've juggled that between your love and passion for data exchange and IT and healthcare delivery. And I'm curious, with patient engagement today, primary care or other lenses that you can look through from a health practitioner perspective, what's the biggest challenge with staying connected with patients? Is it health systems not having the right tech? Is it patients not knowing that? Is it health plans getting in the mix here with maybe too many patient or member engagement tools? What's the biggest challenge right now to keep patient engagement solidified with a doctor and a person?

Dr. Steven Lane (14:26):

So I think all the things you just mentioned are all challenges, and we could unwrap each one of those in turn. I mean, as you say, I'm a doctor through and through. And I take care of patients, patients like you, patients like your mother, patients like your neighbor, and really caring for people. I mean, really caring for them is hard. And engagement is all about the relationship. It's about trust. It's about sharing bi-directionally. So what I've found in the 20, what is it now, three years since we launched the first MyChart, is that patients get it. Patients love the idea of having a secure safe channel to their care team. And if you provide that, they will use and support it. They will use it. Now, I've heard the notion of portalis. I don't know if you've heard that term, but the idea true me, I like it.

(15:33):

Patient portals are not the panacea. I mean, lots of patients receive care in multiple institutions on multiple health IT systems, and if you haven any moved from one to the other, it's a major pain, a lot of loss of fidelity. So I think one of the big opportunities is providing a single channel. Because none of us as individuals and we're educated, capable, et cetera. But as you get to people who have more challenges accessing technology, utilizing it, et cetera, we need to have solutions that leverage technology, but that are very flexible. I mean, some people have trouble with video, some people have trouble with broadband. Some people want to have an old-fashioned flip phone and it may be appropriate for them. Some people have full service broadband everywhere all the time. So I think coming up with solutions that can serve the whole population, if we really want to address the challenges of health equity and health data equity, we need to be thinking broadly.

(16:41):

And then having them really also bring together the entire care team. That's the promise of interoperability, is that patients should be able to choose the tools that work for them. And again, in an app ecosystem that we've all become very familiar with, you can choose Uber or Lyft or what have you. You even can choose your bank's app, et cetera. You should be able to choose the app that you use to interact with your healthcare team. So that's easy to say, hard to do. So we're working on that. Similarly, when you do that interaction, it should include all the workflows. You should be able to message with your care team. You should be able to do your refills, check your results, you name it, pay your bills and bring that all together. I mean, it's great if you're a giant vendor, like an Epic, and you can try to do the whole thing soup to nuts.

(17:38):

But if you're talking small office practice, you're talking people who are working with multiple providers, doing that in a way that is simple, understandable, is a huge lift. So the promise of the new fast healthcare interoperability resources, the fire standard, the API requirements that are being driven by ONC and CMS and others, the promise of all that is that over the next five years or so that we're going to see this thousand flowers bloom and an API ecosystem where these things are actually going to start to work. I mean, it's really cool when you see APIs work. I mean, wow, this app went to that app, "I logged in with these credentials and boom, I've got my car rental or my money transferred or my groceries ordered, or whatever it is." But there are challenges of course, because healthcare data is not groceries, it's not even money.

(18:39):

I mean, there's all kinds of sensitivities and need for privacy that are unique to the healthcare and health data ecosystem. And so why does our industry lag behind so many others in digital transformation? That's why, because it's not easy. And lots of really smart people with lots of money have tried all sorts of things over the years. All the big tech companies have tried this and that. And some of them have made real progress with patient data access with tools. I mean, my company now supports a lot of digital health companies, and we're really starting to see those stick and get traction and survive in the marketplace. So really, but it's not happening overnight. There's a lot of foundational work. I mean, we had to digitize medicine, we had to install EMRs, we needed the internet. We need a lot of things to really make this work.

Brian Urban (19:40):

And I love where you took us because it's human condition, it's human health. It's not just a transactional thing that we're talking about when you're looking across the healthcare ecosystem. Sometimes it is, but a lot of it's very complex. We're talking about it human health here, not getting your groceries, not exchanging dollars, things like that through a Venmo. So it's beyond that. And I love where you were going too in terms of how data is being more accessible to help decision making. And you said health equity, you said data health equity as well. And as a primary care physician, you've seen a lot of different life and health scenarios come into your setting. And in an it, you've seen how that data is tough to share from one entity to another, and how ecosystem is starting to be built out in a more sophisticated way.

(20:41):

So this makes me think of social health needs or social determinants of health data. And there's a lot of primary care, there's a lot of other settings, ERs, ambulatory, et cetera, that are collecting these types of needs through SDOH screenings. And NCQA is supporting this, CMS is starting to require that these assessments are being done and starting to be tracked. There's no incentives or disincentives yet. But when do you think we're going to get past screening and we're going to start to see an integration of socioeconomic or nonclinical data into EMRs? I mean, do you think that's on the horizon? It's going to start in small waves. I mean, do you see it building out? When's this going to happen from your [inaudible 00:21:32].

Dr. Steven Lane (21:32):

I thinking it is happening? It's happening in little baby steps, but just like everything else we've talked about, there's a lot of foundational work that needs to be done. So I like to use the SDOH acronym to refer to social drivers of health. I'll just put in a little plug here. My friend Medell Briggs-Malonson at UCLA taught me this, is that when you talk about determinants, it sounds very deterministic. That's it, you're stuck with it. But these are social drivers. And when you think about it, we all have social drivers to health. From where do you live? What's your educational level? What's your nutritional situation? What's your job situation? And these things change sometimes throughout a day. I mean, you lose your job, your refrigerator's empty, somebody turns off your power. So the social drivers of health are important for everybody.

(22:32):

And I think we do have this misperception that this is only... We're only talking about the unfortunate, we're only talking about poor people who don't have insurance, don't have this, don't have that. But everybody has social drivers of health. I mean, we all know that everyone has psychological drivers of health. And what's your psyche determined by if not your social situation? Your relationships, et cetera. And when your belly is empty, you're not going to be in a good place. So I think the social drivers of health SDOH are critical. We knew this 100 years ago. This is not new, but there is this renewed focus on it. And as you said, the focus is on the data. It's one thing if you have a provider that you've known your whole life or for a long time and they know you personally, and you do all this stuff on an intuitive humanistic level.

(23:28):

But if you're moving around the system, I mean, you need to write it down somewhere. This person has this issue. Historically, we used the problem list and that's a fine place for it. And they've created these diagnosis codes, Z codes, et cetera. So you can write them down, you can say, "Well, here, this person has this social driver. I've figured it out. I'm writing it down, so it's going to be there the next time." That's okay. But to me, I mean, the problem list is always changing. The social drivers are always changing. I mean, I think about the SDOH list, if you will, as more like the medication list. Things come on it, they go off it, they change, et cetera. So when we think about SDOH, we have to think about it very dynamically and very holistically.

(24:14):

But then we also need to have a way to write it down. So the Gravity project, which I'm sure you and most of your listeners are aware of, is an HL7 fire accelerator, a bunch of my friends work there. And it's really been focused on laying that foundation of data standards. What are the social determinants dimensions? What are the key issues within each of those dimensions? And then as you were saying, what are the questions that we can ask? How can we standardize these instruments or these methods of collecting this data so that if I collect a piece of data in my clinic, it's going to mean something to the food bank or some other user social worker down the line who has an interest in the same sorts of issues. So standardization, definitions, and then we have to figure out how to share that data.

(25:08):

So it's a lot of work to say, "Well, here's this new class of data. Here's this new data elements. Here are the definitions." Now you have to bake it into your systems. You have to figure out how to exchange it. So we're involved in piloting the use of SDOH data. We provide that data to our network participants where we get that from a vendor who collects it with publicly available data. We're also working with a number of partners to build out the ability to bring other sources of SDOH data together. Some of that is geographic data just based on where you live, work, play worship. Some of that is going to be data that's collected in questionnaires. But of course, I mean, if you ask someone a really sensitive question about their housing or their transportation or this or that, who's asking? In what context, with what linguistic and social sensitivity is going to impact the question or the answer.

(26:08):

So suddenly you've got the need to understand the provenance of that data and the timing of that data. So how do you bring together this SDOH data in a way that's accurate, that's meaningful, that's interoperable? And then as you were saying, how do you integrate it into the other healthcare workflows? Whether it's in the electronic medical record's being used at the doctor's office or the hospital, whether it's in the social care record that might be used by a community-based organization, et cetera. So it's really important stuff. And when you're a provider and you're trying to make the best decision for your patient, having that information there along with their allergies and their med list and their family history is going to be critical. We are not there yet. We're not even close, but we're on the path to go there.

Brian Irvin (27:01):

And I love the honesty because I think there's an allure in the tech world to say, "We're doing this, we're doing that. We're going to achieve this at this speed." And you hit it on the head, we are not there. We're at stage maybe between stage zero and stage one, we're capturing information. We're learning how trust and engagement is built as a foundation that people can validate their needs. And then downstream, how is that connected into an additional service of social, behavioral health, psychological needs, et cetera on those dimensions you mentioned? So it's good to know that we have a realistic viewpoint from a physician that's also a tech leader that's coming out of Palo Alto. So you bring a real humility and reality to it, so I appreciate you saying that. But we have so much more work to do. And I'm thinking about patient safety related to social determinants of health. And you touched on a little bit, but how much patient safety is put into your philosophy of your data exchange and your tech that, maybe is that with Health Gorilla or maybe from other experiences?

(28:21):

So I guess how do you envision patient safety being involved in SDOH, how we collect it, how it's validated, how it's used for good, not for being able to position all kinds of maybe unnecessary needs in a health setting. So I'm curious about your take on patient safety with SDOH.

Dr. Steven Lane (28:45):

It's a fascinating set of questions. Each one of your questions includes all sorts of sub questions that we could-

Brian Irvin (28:53):

I happen to do that. It's by accident.

Dr. Steven Lane (28:54):

No, it's all good. So I mean, it's interesting when you think about patient safety, you want to prioritize what are the issues? I mean, in what ways are patients unsafe? Those are the things you want to focus on first. I mean, in healthcare, we do all kinds of things to individuals that are unsafe. I mean from surgeries to medications, to you name it, dropping the ball for follow up, et cetera. So I think SDOH data can inform pretty much every healthcare decision. And you can quote me on that. Never said that before. I think that's really true.

(29:37):

I mean, whether you're talking about prescribing a medicine, scheduling a surgery, recommending a referral, those decisions, if somebody doesn't have transportation, can't read, doesn't have food in their fridge, doesn't have the money to pay for their meds, any of those things. Or as we were saying earlier, they're just so depressed that they're not going to follow through with the recommendation. All of those things can be impacted by social drivers of health. So bringing that data into the decision-making process and so that we can meaningfully inform decisions is key. And coming back to safety. I mean, if I am seeing a patient and I say, "You have got to go to the ER. So I want you to get in the car and get down there because you're having whatever, chest pain, bleeding, et cetera.

(30:35):

There's a lot of situations where you don't call 911. You just go to the ER, you go to urgent care. If that person doesn't have transportation and I don't know that, and they're not comfortable sharing that with me for whatever reason, that creates an unsafe situation. Similarly, if there's a medicine I feel needs to be prescribed and somebody can't afford to pick it up or doesn't have the transport to get to it or doesn't have the linguistic access to get to it, then that's a patient safety situation. And then also, it's not specific to SDOH, but the whole notion of just having incomplete data generally, whether it's clinical data, social data, personal data. If you have an incomplete picture of somebody's situation, you're much more likely to make the wrong recommendation, where again, downstream safety impacts. So I think that to take optimal care of people, you need to know as much as you can about them, but you can't get so lost in all the data that you're frozen.

(31:44):

I mean, so one of the big challenges of interoperability generally is finding that sweet spot. We don't want people drowning in data. Because then basically you're paralyzed. You don't know what's important. And we don't want so much data that you can't find the thing you're looking for. And these are real challenges. When we were just implementing EMRs, like, "Oh, this is great, I can find my last note." Or we're just implementing interoperability, it's like, "Oh, great, I can find the x-ray report from across the street." But it gets way more complicated now that millions and millions of transactions are going, that you're getting the same data back from multiple sources, the lab data from the lab, the lab data from the hospital, the lab data from the doctor's office, and you have to deduplicate it and you have to then errors creep in along the way. So there are real patient safety issues related to the digitization of healthcare generally, and interoperability in particular that we're still trying to sort out.

Brian Irvin (32:50):

I love the EMR side of your brain because it really speaks to fundamentally how are we able to keep track of patient needs, coordinate care, share information. And you said something that I want to go into for a moment. The EMR space has blown up so much over the course of the last 15 plus years, and you've seen that and you see a lot and you hold a lot of positions on compliance and regulatory reviews. So are we getting to this place where it's getting really complicated that if we would try to integrate SDOH data, it might be too much, it might start to come to a breaking point. I mean, for instance, like you said, what's the sweet spot? We would never want to put in front of a physician, "Hey, this guy just had a DUI, he lost his spouse last month." A laundry list of things that would be very difficult to, in a empathetic way address. Or is it a score? Is it a measure? Is it a direction?

(33:55):

I think about the overwhelming amount of information that vendors could throw in an EMR that is SDOH related. Do you see a big danger happening if we go down that direction? Or do you see it still, whoever makes the first most meaningful baby steps into a helpful action for a physician today will help be the next generation of the EMR? I'm curious of what you think there, because we touched on patient safety, but I feel like there's a big danger to just plugging in potentially a wave of new data that's not clinical, and it could really maybe wreck the patient physician relationship if we're not careful.

Dr. Steven Lane (34:41):

Well, I don't think that the challenge that you've identified is specific to SDOH. We will go back there, but you can think about it more generally as you bring in new types of data. Think about wearable data, think about genomic data, think about all just getting claims data back to providers. I mean, there's a lot of different sources of data. I mean, think about your online behavior and your shopping. I mean, that matters to me as a provider. What food are you buying when you go down to the grocery store? What kind of supplements are you putting in your body? So back in the day, I used to say, "I just want all the data. Just give me all the data and I'll just manage it cognitively." Well, I know as a cognitive science that I can't do that.

(35:31):

So regardless of which of these classes of data we're talking about, SDOH being a perfect example, we need cognitive support. We need augmented intelligence. And lo and behold, right as we're hitting this tipping point of drowning in data and solving the interoperability problem, along comes augmented intelligence in a meaningful way. So I think that what our future path is going to show us is that we're going to use the machines to help us manage the data. To help us to surface the signal in all of that, we can call it noise, but I mean, truly it all matters. I mean, all this data matters, but some of it matters more. And some of it matters more for this person today in this situation. And I think as a physician, you try to do that in your head, but it's like if you haven't had your coffee or it's the 12th patient of the morning, it ain't going to happen.

(36:41):

But the good news is that, the computer doesn't need coffee and it doesn't get tired. And so [inaudible 00:36:47] with AI, we really have an opportunity to find that diamond in the rough, needle in the haystack, find that bit of information. And then what we need, you were saying what about the EMR process? I mean, EMR is just the physician's window on the data. I mean, there was a time when the EMR was the only health IT application available to the physician, and it was everything. And I think a lot of our workflows really are built on that assumption. I'm a Cerner user, I'm an eClinical works user, et cetera, and they define the arc of my day.

(37:24):

But truly, like in other parts of our life, this app ecosystem, the APIs should allow us to be more sensitive to the needs of an individual clinician, be it a nurse, a social worker or a dentist, a physician, to bring that portion of the patient's holistic data to their brain, their workstation, their screen, in a way that will support them in utilizing their special knowledge to make the decisions that the AI isn't going to be able to make for them.

(38:01):

So I think when you talk about SDOH, it's the same thing. Yes, we should be capturing this SDOH data. We should be maintaining it, we should be curating it, we should be sharing it, but a lot of decisions aren't going to need it. So don't flash it up in front of my eyes when it's not relevant. And we haven't figured that out yet. We haven't figured out... We've done a ton of work. I think I may have written the first EMR based clinical decision support rule, way back in 2002 or something. But it's really hard to get that stuff right and everybody has a different opinion about it. You talk to any clinician and they'll say, "I want to see this alert, but not that one. I want it this color and not that one." And at this point in the workflow, and it's so hard to do that really well. So bring SDOH into the mix as an input to those systems and it doesn't get any easier.

Brian Irvin (38:56):

It's a whole nother potential mess. And you're right, in terms of the messaging and alerts, we've just, I think as a healthcare society become more aware that we need to reduce that noise, that distraction, that non-helpful alert or message to a physician to let them do their work with their patients.

Dr. Steven Lane (39:21):

That's a tricky balance to strike. Because, I mean, I've got physician colleagues who say, "I don't even want to touch the computer. I don't want to go [inaudible 00:39:29], I want to scribe someone else. But then you miss all the opportunity to provide decision support. So you want alerts because we make all kinds of mistakes. We choose the wrong drug, we cut off the wrong leg. We do all kinds of things. It's like we need that oversight. We need to augment our intelligence, but it's not always with the same tool.

Brian Irvin (39:54):

And I was waiting for the AI bomb that you were going to drop too, which is timely in our evolved state of technology and data exchange and diluting data as well. So it's about time that we have a really strong middleman that is supportive and being able to tell us what priorities makes sense for what patient. And we're on the cusp of it, and it's just going to get better. I know that there's concerns about AI, so I got to ask you this. I feel like I could just keep you on the line for hours. I just love learning from you. But I got to hit this question before we get to the close of our conversation. Dr. Lane. AI in healthcare, I say those words, does that scare you or does that get you excited on what can happen because there's opposing views all over the place.

Dr. Steven Lane (40:47):

I think most clinicians will tell you it's a double-edged sword. It's both our salvation and the potential solution to a lot of the problems of our current healthcare situation. And it's incredibly risky. And this isn't just true in healthcare. I mean, you've got people across the spectrum saying AI could really help and save our civilization and our species, and it could be the end of it. And I don't think healthcare is any different that way. I think a lot of really smart people are coming together trying to figure out how we can go down this path safely. What are the opportunities to automate routine tasks to improve the reliability of things that a computer could probably do better than a human being? I mean, this interesting study came out a few months ago where they took unfiltered patient requests that came in across the portal and they gave it to a bot or to a group of physicians, and then they rated the responses in terms of patient satisfaction and thoroughness, and the bot did a better job.

(42:00):

I mean, why have me up at 11 o'clock at night responding to my patient's online messages when a bot could do that? And so I think there's a lot of opportunity. So automating routine tasks, I think that's pretty easy. I think we're going to get a long way, we're going to hopefully save some money along the way. And because of course, one of our biggest challenges in US healthcare is that we spend too much and we don't get enough for it. But then going on to the clinical processes, how can we insert AI into the diagnostic process, the therapeutic process, the follow-up process, and how do we do this in a way where we're still getting at the human part of healthcare? How do we train doctors in this modern world so that they can still do the humanistic part of medicine with the assistance of the augmented intelligence that we don't lose everything along the way?

(43:01):

I mean, two of my daughters are going into medicine and I watch them and I see their path, and it's so different than mine. In some really good ways and in some ways that really make me worry. How do we train the next generation of clinicians? Given what's going on today? And I'm sure the same concern was raised 100 years ago, but it's a concern today too.

Brian Urban (43:28):

I feel part two coming onto this episode, perhaps with some guests that share some views and have some opinions, but I have to thank you so much, Dr. Steven Lane an encyclopedia of knowledge across Healthcare, IT and beyond. I really have loved our time here today. And for more exciting excerpts and insights, please visit finthrive.com. Thank you, Dr. Lane.

Dr. Steven Lane (43:56):

Thanks, Brian.