AI is reshaping personalized healthcare

<p>hello everyone this is Cole from the American Journal of healthc care strategy and we have a very special guest today Christine please go ahead and introduce yourself for us hi Christine calling in from Boulder Colorado I have spent the last decade equipping myself with a skill set needed to design and build care models that are using emerging technology I'm Master prepared in health administration and I'm excited to talk more about my career Journey with [Music] you so can you just take us back to start with what was the the beginning of your career like what was your undergraduate and graduate Pathways why did you choose those and how did you transition to where you are now sure it's certainly non-traditional I went to school at Clemson University in South Carolina and I started studying biology I was very interested in becoming a doctor that was my initial game plan however the first couple of tests in chemistry and biology indicated that that might not be the best path and so I actually transitioned into public health and I was excited because that meant that I could in influence patients on a larger scale at like a population level and so I studied public health and when I was in my freshman year I went around to my professor's doors and knocked around and was like is there any opportunities for me to do research with you in any capacity I just want to do some type of research and I ended up getting brought under um by one of my professors who was exploring biof feedback monitoring in Alzheimer's patients and family caregivers it was a very lucky unfolding of events where I got to partner with her for three and a half years of my undergrad to lead research uh including developing the IRB the protocol the interventions and then actually collecting and analyze analyzing data using one of the first wearables in um in the market and so we were really operating in outpatient and Hospital settings and that's where I started to get exposed to Administration to hospital settings in general and I did a few other internships throughout my undergraduate studies in in the health systems um one was at Greenville Health System in clinical trials management and the other was at MedStar Health in quality and patient safety in DC and yeah so that's kind of where I I got started and after college I was looking at getting into Consulting right away but I didn't have a tremendous amount of experience and so I thought it would be best to go and get potentially another degree to be able to make myself a stronger applicant I was exploring programs and I ended up talking to a lot of Hospital administrators at that time and they encouraged me to look at VCU and look at the Masters of Health Administration check and so I did and I ended up at VCU and spent two years in the classroom and one year as an administrative fellow and I worked a number of internships there as well some of them were impatient uh quality and safety some were in patient experience largely I was getting experience in operations and managing a hospital and typically sat alongside the executive teams and really observed them making decisions and supported where I could how impactful was that that Fellowship for what you ended up doing next so the the experience that I got understanding a hospital leadership is is very crucial to what happens next in my journey and I guess what I learned during my mha was that I was gravitating towards startup events in in the community I was grabbing pizza with Founders and kind of King their brains around how they were starting companies and I just was very interested in Innovation and technology and that wasn't getting like tickled in in hospital Administration and I think the closest I got was during like Capital requis requisition processes where I was looking at like who we were investing in but in terms of actually developing the solutions or bringing in Innovation I I was really craving that and I wasn't really getting that opportunity in in the hospital walls right so after my fellowship I moved to Boston I knew Boston had a lot going on and I got the opportunity to join Phillips which was running Global accelerators for startups and they had different um focal points for for their startup programs the first one was actually Ai and so that's where I really got excited about Ai and product development co-creation models um business model Innovation and really where I started to understand how this knowledge that I had from the hospital realm could actually support with maybe the development of products and also the business model for some of these emerging startups in healthtech now did you have direct reports when you at that time when you got there no so we were a team of there was 10 or 11 Innovation leads and these were folks that had expertise in like design research business modeling and really um techn ology validation for the internal arm and so then we also had um supportive roles my my role was very supportive in nature where I was helping with the marketing and operations of the accelerator program and so that looked like creating a a database or a CRM of folks that might want to have a conversation with these startups to validate their solution and so we had pitch nights investor networking nights and flew these startups to Hospital Systems had them engage with Hospital Executives really the goal was to figure out how to drisk these solutions for Phillips to potentially uh work with them or partner with them in some way how does that all fit in did you enjoy your time working in at Phillips you know what was that like for you yeah it it was an interesting transition because you're coming out of this very like glorified internship where you're a fellow and you're working day in day out with the CEO and then when you step outside of that you you kind of take back on that in individual contributor hat and so it can be a bit of a shift like a mindset shift to okay I'm not playing in the leagues anymore you know I've got to cut my teeth and and prove my worth here and so in terms of moving outside of the hospital walls I'd say it's it's very important to have Champions at wherever you're going um to work and having people that can advocate for you to get back up to where you're comfortable playing in the most efficient manner and and so I I would say that a lot of this was exploratory for me and I took on a couple of different roles because I knew I didn't want to play hospital administrator but I wasn't exactly sure what was next so it was either going to be some sort sort of like public health angle like research um maybe design as very interested in product development uh I was also just have a general knack for marketing and and building and so so I viewed a lot of these almost kind of like back to that undergrad experience where I had done this like independent research study to really like figure out what I wanted to do I viewed each of these roles as independent research and the research or the problem statement was like what should Christine's career be and I love you know these trials for me were typically like a year and and so with regards to the operation and marketing at Phillips I recognized that rather than playing a supportive role kind of on the sidelines supporting the action that I actually wanted to be the Innovation lead I wanted to be the person that was advising the startups and bringing through ideation workshops and um interfacing with the investors I didn't know how I was going to make the leap from hospital Administration to like investor and so I thought that product would be the next natural stepping stone for me and that's what took me to CVS in your skills based on the LinkedIn right this what you were involved with at CVS seems to be quite a bit different than what you were involved with at Phillips right so can you tell me was that scary at all was that a huge learning curve what was that like and also how did you get that position because that seems like a fairly prestigious position for someone who's a year or so out of their mha program so I I I guess I skipped a bit so in between during graduate school and before Phillips I had done a number of startup engagements and I supported folks in the community that didn't have a sense of like documenting research protocols or communicating their Vision say through a pitch deck and so I did a lot of work with one one of the guys I worked with developed a music intervention app for children with autism and he was looking to scale that across Virginia and so I supported him for um a semester in developing research protocols and then a pitch deck to go support uh secure like approval of funding with the phasin school which was a Alternative Learning school that had multiple sites across Virginia I think those experiences and being able to speak to those experiences of kind of like self-drive curiosity interest in improving Public Health really it was that Foundation that I got at Clemson that really supported me with executing in this role the role was really focused on social determinance of health and how to improve access for the disenfranchised and so a lot of our interventions were around digital modalities or interventions that you know this experience that I got in graduate school and a few other internships or independent studies kind of built on allowed me to be a strong candidate for that role yep so that makes a lot of sense so you sought out these experiences early on of things that you were just genuinely interested in from the sounds of it didn't seem like you had like a master a master plan right you were just really doing things that you felt had impact that you wanted to do and you wanted it seems like you look for things that had the most impact it seems like that's why you and correct me if I'm wrong but that's why you went with the startup space is because you wanted to innovate you wanted to create a big impact yes I early in my career wanted to develop skills that would position me to be in a place where I could design and build like the future of healthcare and so I always gravitated towards what was going to be like the thing in like 2 to three years and I tried to get in early and get experience handson building those products and so like at CVS they were talking a lot about like Omni Channel and that's kind of like you know haptic experiences which we all crave this human element but also having this data driven digital interface and so a lot of what I've gravitated towards has been how do we position ourselves to age gracefully in our homes using technology like that has been the North Star throughout all of these experiences and how did that cause you to transition from CVS to is it pronounced Linus Linus yes Linus So Co hit right and I was driving to Rhode Island every day uh and and that was beginning to take a toll and so for my own like lifestyle when Co hit I had a lot of opportunities to get introspective and think about where I was going not only in my career but personally and in my lifestyle Boston is incredible for for industry connections I would say that it did not cater to a healthy lifestyle for me personally and so when I had the opportunity to move to Boulder I was like yes that I can see myself cycling and skiing and going to yoga and really taking care of myself kind of back to that whole work life balance aspect I had been go go go for so many years you know through graduate school and then through all of these competitive environments that I needed to take a step back and take care of myself before I could continue in a way that was going to be sustainable and so how you I assume you've liked Boulder Colorado because you've stayed there four or five years right three three or four years I think yes um so can you tell us about I guess one of the things that really interests me personally maybe this a selfish question I don't want to derail things too much but you went out freelance um for over two years now right yes um I guess what I want to know is you didn't have the 20 years of experience I see a lot of people have who go out freelance but it seems like you've been successful because you're still doing it yes what what was the motivation like how how did you manage this I would love nothing more than to to be freelance freelance consultant so uh is it is it as good as you've hoped would you do it again would you do it earlier you know just tell tell me your thoughts on what that's been like yeah I think there's tradeoffs and it's kind of like what is your risk appetite would be my first question because it is a high-risk uh environment where you know some of your core clients if they were to go away then you may be in in a difficult position however in terms of being able able to craft your your lifestyle and craft your workload and decide which problem areas you want to focus on or who you want to work with um all of these aspects make it worth it for me um you know if I decide everyone's going on a big ride um up the mountain on Wednesday at lunch and I want to go on that I can structure my week um in a way that allows me to to go for that and I think that at this juncture in my career that makes sense um for for the long term it's it's an interesting question of like is this going to support the skill sets that I need to be developing and product to be able to continue to be advising Folks at the Forefront of innovation um but for the past two or three years it's been a a dream for me to be able to to step out and to work with like 30 to 40 different startups um on on their problems on supporting them with growth and I spent a lot of time in strategic planning branding and positioning digital transformation market research product ideation and and business model Innovation yeah that's what I wanted to go into as well is based on you know the site here you have a lot of different skills that you're offering to these startup organizations but you've also really specifically it seems like focused on AI that's been an area that you really are are passionate about out of all the startups that you support how many of them are involved in AI if you had to give a percent is it you know 100% is it no it's not it's I would say that the the focal point for me or where things kind of intersect is like Health technology care delivery and like that's kind of the and Care delivery could be like a food and beverage solution that you get at the grocery store like it's not contained to what we traditionally think of in terms of like Hospital operation arms you know inpatient outpatient so you know in terms of AI I've worked on um largely on a a cardiac wearable dashboard for um a you you'd know them but it's it's proi but you know looking at kind of the revenue model behind um patient care and so looking at cardiac dashboards in that way and then in terms of some of the other startups that I've worked for it's been more on the personalized medicine front so looking at actual like genomics or life sciences assays and solutions there where maybe it's a diagnostic and the diagnostic is actually using uh language learning models to improve um you know its ability to match a sample to a database and generate meaningful output one of the things that you mentioned in the um you know in the discussion with us is you sent us a a slide and it has kind of the a the healthcare AI Market map on it and that that's really fascinating to me because you you wrote you know how many how many side K or how many co-pilots does one need right you're going to end up with with a hundred of these and and I've struggled with that myself especially when you you're talking well I'm sure you've used them a lot of these personally you know you end up having 10 or 20 $20 a month subscriptions for all these different AI Services right and yes yeah and you end up with a real problem not just because of the the cost but also it actually becomes more inefficient so what what are you doing in that situation where do you think we're going to go yes integration and interoperability is not a new challenge it's something that we've been trying to tackle for for decades and so you know you look at things like fire like the new standards for uh standardized documentation like that's that's just the first step and like that's been a huge bottleneck you know like we need to get the data structured in a way where you can then use it to train a model currently that is not how things are structured and and and there's part of that that's intentional you know and so you have to look systemically like how are we being incentivized to make these Solutions work why are there silos in data structures you know are those something that is contributing to a bottom line or like what is the The Narrative there and a lot of the data structuring and processing in my opinion is where we're very weak right now I think that's a huge issue for AI being successful today and how much of this is due to those cognitive distortions that you're talking about and do is that something that you help people to overcome as a consultant because you know reading through a lot of them it seems like a lot of the issues we do face with data processing with that interoperability comes from a lot of these cognitive uh distortions that you're coming from right so when you're talking about cognitive distortions so humans are lazy we take mental shortcuts and these are typically biased and so when you look at at these and how they're impacting the system of healthc care you can see it everywhere I mean the shiny object syndrome is like all of the Venture Capital right now is is under that and you can see like with all of the startup Investments that are happening in like Precision medicine medical coding um clinical trials neuros stimulation devices Pharmacy tele medicine like these are all great potential use cases however like a lot of these uh LL or language learning models they're they're in their infancy right and so they haven't been tested at a systemic level for health outcome so a lot of it still hearsay you know if you put AI on a napkin and you and you're at a coffee shop with an investor and you're like this is my solution and then Circle AI like you're going to get funding and that's just because you know there's this like like you mentioned this fear of missing out you know this desire to adopt the next new thing for fear of being left behind there's also yeah the shiny object syndrome which is like being drawn to new flashy Technologies without considering their true value when we don't talk about AI when we're going to investors they want to see a slide deck they want to see a pitch deck they want to see our business plan they want to understand the company when we talk about AI it's here's the money send me the pitch deck when you get a minute it's a very like different reaction and I wonder how we're actually going to change from that you know where do you see us going in five years because it does worry me that this is just going to be an ongoing I mean an issue that we're gonna be facing for you know 20 years or so it's it's kind of reminds me of um tele medicine right it's like is AI the new tele medicine like a lemon disguised as a Ferrari or will it change healthare forever and that's kind of the question that I think we'll be answering over the next five years but right now today there there are a few other biases coming up which is action bias which is preferring action over inaction even when the benefits are unclear um uninformed optimism overestimating the benefits of AI in healthcare without solid evidence and then Game Theory and that's coming up in in the business modeling so making decisions based on competition rather than on patient outcomes yep and we tend to stick to past behaviors even if they're not working and so a lot of these folks that are putting their stake in the ground for AI like they're not going to backtrack on that they're going to try and figure out how to make it work and so you've seen Google coming in Microsoft and Oracle like there's a lot of big Heavy Hitters coming into the clinical settings and then what I've been most excited to see is that there's a New Consortium of healthcare leaders that are creating like a trustworthy and responsible AI Network and they you know a lot of the larger health systems or organizations have bought into this but I think you know when you're looking at the next five years there's a lot of pitfalls you know and there there's a lot of things that I think are coming up as dilemmas that I would like to see overcome what's that Consortium you're referring to yeah so it's called um train t r a i n trustworthy and responsible AI Network making safe and fair AI accessible to every Healthcare organization is there technology out there that can actually make an AI actually safe right now that's unclear and you know we're all aware of the bias uh uh that's at play right now um racial bias gender bias you know there's General like in healthcare if you think about it like a lack of consent for for patients to even know if there's AI being used in their treatment plan there's like the liability if something goes wrong does the doctor get in trouble does the software developer get in trouble there's like privacy and Trust concerns but I think what you brought up is it my biggest concern is the general purpose base models that a lot of these AIS are being trained on so I'll give you an example there's this W Aid um CA legal case where it was it was screening um folks who came in for shoplifting using a general purpose Ai and and and making some assumptions that were inaccurate and so they looked at okay so why did this occur that you know they go into the code and they see that this base model that it was trained on had gender and racial bias at play so because this is such a new area when you're actually in the code as a software developer you're using a lot of things to speed up your workflow that are already kind of in place so say you grab a general purpose B base model and then you start um using that to to build whatever you're building if that has a ton of bias in it which it does right now like that's going to be overfitting to everything that you train it on in in regards to like a health system or a patient population and so so the way to combat that is to take the approach that a lot of these genomics AI folks have taken which is to create repositories or open model networks this is a huge challenge in healthcare right you can't have an open model of Phi and so then you can't interrogate one another the whole it's kind of like you put on your like your good hacker hat and and you go and you say is this trustworthy or ethical because of the nature of healthc care like you can't really lift up the hood and let people look at the parts look at the engine and say and and then when you think of like the business model too it's like it's all like patented or siloed in ways where you can't get to the core of of what it's doing that is my biggest concern today I want to ask kind of a closing question question on this is how can someone like myself or someone who's in hospital operations how can they learn to to kind of be like you how can they learn these skills learn about these things I mean I'm assuming not all of it came from the education you received at your mha program or just from your experience so are there you know books we can read you know communities we can get involved with absolutely there are there are a lot of courses coming out right right now that uh I can link after this to kind of direct folks to but there's a lot of courses that can get you upskilled in just understanding kind of the lay of the land and the technology wa if you want to start getting into these pitfalls and ethical dilemmas and really the ins and outs of patient rights or model training data or updates General updates um substack is becoming one of my go-to sources there are AI experts on substack who write specifically about Healthcare and I really appreciate when they give actual examples of you know like a 2019 United Health AI algorithm prioritized care for healthier white patients over sicker black patients impacting over a 100 million patients like there is evidence there's a news article and a link and those stories for me storytelling is always so crucial to my understanding when I can see examples of of these things playing out in actual you know siga AI algorithm automatically denies claims Sigma doctors concur without even looking like when I see a headline like that I'm like okay what are we looking at there and I want to know more a lot of it for me is just curiosity and kind of following the Rabbit Trail seeing what I find I think that's uh very good advice to just be have that sense of curiosity it seems like that's something you've had throughout your career you've had a real sense of wanting to learn and in curiosity in the decisions that you made on where to go to school what to go to school for what jobs to get it seems like that's led to your success quite a bit yes and and just to kind of tie the bow on AI like I I don't think AI is like all bad I think that AI as a companion a co-pilot like a way to augment human versus like replace them there's so much opportunity to improve productivity decision making and outcomes and it can be like an additional Observer assistant or an eye like ensuring that important details are not overlooked and this is very important in like image analyses or Diagnostics or genomics um so the the proposition here is that like a system or process would involve both Ai and humans but it would be human designed first you know it wouldn't be an AI LE model ever would be my recommend recommendation like there will always be a human involved and it'll be designed in a way that it isn't Force fitting Solutions based on what this technology is capable of but rather you're stepping back and you're saying okay what are the unmet needs or jobs to be done or problem areas here and then you go to your toolbox and AI is just like Excel it's one of the things you might pick up to solve for this problem but it's not something that you're going in like how can we use AI to get on this train that's like everyone's moving on like I would strongly discourage that approach Y and I'm a big fan of AI as well you know I I'm not I think it can be very useful and it can definitely speed things up as well I would love to talk for uh a few more hours but of course we have to stay within time so hopefully you'll consider coming back on I couldn't be more appreciative for having you on this afternoon yeah thank you and and just one last thing because you mentioned a hospital administrative students or early careerists maybe entertaining the thought of doing something about this about these problems absolutely and I I think I'm seeing like this opportunity to become a patient experience designer or someone that is being the glue between the tech and the human and we call it human computer interaction and design that's like the technical phrase for it but I think someone that understands the operations and can wear the hospital admin hat but then be designing these Solutions with like a patient experience lens and maybe upskill in in some of the the basics of uxui and some of those like technology Concepts that is aner merging career path</p>