Ep 13: Harnessing AI for Seniors - Interview with Cheryl Field

[00:00:22] Oh, thank you, Rosa, for having me. Oh, I'm happy to tell that story to your listeners, Rosa. And really, in fact, I would have bet that I would not have become a nurse. I was quite a tomboy growing up. However, I always was that person who sat with my grandmother. I was really a good companion. And in high school, I even had an opportunity to do companion care in the home of a senior.

[00:01:27] So as I was heading off to college, really kind of thinking that I would head into medical school, I knew I wanted to do something in the healthcare industry. And through different experiences of meeting nursing students and really learning the difference between cure and nursing care and the whole person and person-centered care, I was, you know, gently but very, very excited to change my major into

[00:01:57] nursing. And I knew that I wanted to pursue a career where I had an opportunity to really look at that whole person in a way that I had seen as a companion. You know, I saw living in someone's home and coping with chronic illness or disability in your home for the rest of your life, and the way you had to adjust what you were doing. And I saw that there was creativity and opportunity and really a tremendous amount of vibrant ideas in

[00:02:27] nursing. I ended up in nursing. I ended up in nursing. I ended up in nursing. I ended up in rehabilitative nursing from my first job, right out of nursing school, which again, was just a wonderful space for me to connect. It was a perfect blend of, you know, acute illness and learning anatomy pathophysiology. I worked in acute rehabilitation. I worked in oncology rehabilitation.

[00:02:51] And started to become what I started to become what I thought of as a little bit of a misfit nurse because I liked the money. I liked the business. I liked understanding how the healthcare organization functioned, mostly because I wanted to manipulate that to enhance care outcomes.

[00:03:08] And so that was just an area and attract where I moved into that space, made a transition into post-acute care. And I was in direct care for about 17 years before I had an opportunity to apply that knowledge into a new software organization, a startup that really wanted to digitize and automate some of the data on which the skilled nursing space is really built.

[00:03:36] It was an opportunity for me to apply what I learned, what I learned, what I learned, what I knew around the workflow and merge that with technology in a way that I could bring software as a service to the hands and the fingertips of nurses, my colleagues who were, you know, touching keyboards for the first time at 35 years old.

[00:03:57] And so it was just an exciting career path to transition from direct care to software as a service. And I've been in clinical informatics now for 25 years.

[00:04:13] No, that startup grew. I was there for 17 years. They were purchased. I moved into a chief product officer role for another startup organization.

[00:04:22] And then I moved into a analytics role with a very large electronic medical records organization, again, with the vision that I could empower and change the course of the day for nurse leaders through the tools that I was helping to design and to build, which was, you know, really all of the aggregate analytics that you need to understand.

[00:04:49] And is your business thriving? You know, how are you doing as an operator of skilled nursing from, you know, a revenue quality, you know, a regulatory and compliance perspective.

[00:05:03] And so I was working in software in those spaces where that those metrics were really important.

[00:05:11] But what was really unique about this opportunity was actually going to tie to my question later is really about care coordination.

[00:05:18] And so one of the things that, you know, if you kind of think back to the meaningful use dollars, which was, you know, dollars infused into health care organizations, hospital networks, physician groups to help them to connect, to be interoperable, to pass information back and forth, you know, really freely.

[00:05:43] And skilled nursing, and skilled nursing, and skilled nursing, and skilled nursing, independent living, assisted living, a lot of small physician practices were left out of those meaningful use dollars.

[00:05:51] There was no dollars to build that infrastructure.

[00:05:53] And so the organization that I work for really saw that as a problem and invested in coordination of care through information exchange.

[00:06:03] And so I came to that organization because of the great work and the great vision that they were doing in communicating between a hospital system, a skilled nursing facility where a human might go for two to four to six to eight weeks, and also communicating to the payer who's really a partner in that, you know, in that union, you have a provider, a patient, and a payer.

[00:06:31] And that's an important partnership.

[00:06:34] And the easier we can make information move back and forth between those three, of course, grounded on regulations and compliant with all those rules, really makes a big, big difference to the outcomes of care that seniors receive.

[00:06:50] And I saw that as, I saw that in direct practice, I knew what those problems were, care coordination, care transition, huge gaps.

[00:07:00] But then I started to see it as a family member of my own loved ones.

[00:07:05] And that actually, that gap and closing that gap is the reason that I wrote, prepared a healthcare guide for aging adults.

[00:07:12] I wrote, it really was a combination of, on a career where I knew what should be happening.

[00:07:31] But then I was in the healthcare environment as an advocate, you know, again, for my parents, for my loved ones.

[00:07:38] And I was not seeing what was supposed to be happening.

[00:07:42] Or I was seeing these, you know, the gaps that exist in care coordination, you know, a piece of paper gets literally left out of the packet.

[00:07:52] And your medications, your wound care, everything is missing.

[00:07:55] And it's the sole source of information.

[00:07:57] And it's easy to lose under the pillow of an ambulance.

[00:08:00] And that is still, sadly, how we transmit our most vulnerable patients today around this country is with, you know, manila envelopes of information under pillows.

[00:08:13] And I just knew we needed to do better.

[00:08:15] And I wanted to be able to actually teach the senior themselves and their family members who didn't have what I had.

[00:08:22] And I had a background.

[00:08:23] I was an expert.

[00:08:23] I knew it was.

[00:08:24] And I could advocate in a gentle way.

[00:08:27] And so prepared really teaches them what to expect, what questions to ask, what's going to happen.

[00:08:34] Checklists, you know, to go through each phase of that transition that happens when you have a sudden change in health and you have to go to an emergency room and to a hospital and to a rehab center and to get back home.

[00:08:47] And each one of those transitions is really fraud with error.

[00:08:52] And if you have a recipe for success, if you take up, you know, prepared with you and just kind of follow along as you go, it's like being guided by an expert who knows the system and knows what's going to happen and the kinds of things you should be thinking about.

[00:09:04] So I wrote that book to really help close some of the gaps.

[00:09:09] But I think some of these gaps need I know they need to be closed with technology and automation and artificial intelligence.

[00:09:15] And that's where I get to blend my passion with my professional work and really help to move the industry forward with nurses being infused into projects that are usually responsibly using artificial intelligence to help solve complex problems.

[00:10:33] Well, certainly depending upon your comfort and willingness to experiment with technology.

[00:10:41] You know, one thing I do reinforce to seniors is that, you know, dialing a rotary phone is hard for today's five-year-olds.

[00:10:49] If you put it in front of them, they can't figure it out.

[00:10:52] They try to punch it.

[00:10:53] They think it's going to work on a touch screen.

[00:10:55] So I do believe that our seniors can learn how to utilize technology and technology has been designed so well for seniors, for seniors with dementia.

[00:11:07] You know, the utilization, the human centered design experts have done a great job designing software and designing applications for your phones that are actually quite easy and quite intuitive to use.

[00:11:23] Behind the scenes, there's a lot of technology.

[00:11:26] But for the experience, it's quite nice.

[00:11:29] You know, talking to...

[00:11:31] Sorry, go ahead.

[00:11:38] Even something as simple, you know, my father uses Facebook's portal.

[00:11:43] You can use Alexa, right?

[00:11:45] You can use reminders on your phone if you ask your phone to set a reminder.

[00:11:50] That's really simple technology.

[00:11:52] And I think one of the pieces that is so exciting about what we call artificial technology,

[00:11:58] but if we really look at that definition, it's algorithms that are running,

[00:12:04] but the human being is allowed to communicate with that algorithm in a way that humans communicate.

[00:12:13] So we're talking to our phone and we're talking to Alexa and it's speaking back to us in language that we understand.

[00:12:23] Multiple languages, different, you know, you can change your languages.

[00:12:26] There's translators on phones that are just doing amazing.

[00:12:30] Now, behind the scenes is artificial intelligence.

[00:12:33] It's language models that are, you know, that are running.

[00:12:37] And in the healthcare space, especially being responsible.

[00:12:43] And what that really means is that there's a tremendous amount of transparency.

[00:12:47] This is not something to be afraid of.

[00:12:50] To be very honest, when I teach this to nurses, I just remind them, you know, Florence Nightingale was an observer of patterns.

[00:12:59] She noticed, you know, the surgeons who were washing their hands, patients weren't infecting.

[00:13:04] Surgeons not washing, infections.

[00:13:07] She recognized patterns in data.

[00:13:11] And because we have so much clinical data that's digitized now, we've stopped writing on paper.

[00:13:19] We started putting it in a computer.

[00:13:21] We can use that data and we can set up a training environment.

[00:13:25] And we just look for patterns with algorithms that can look at patterns and recognize patterns that result in certain outcomes.

[00:13:33] And so I think it's actually the most important role for nurses to be in, a role like mine, where what it means to be responsible is understand what data went in to making this algorithm.

[00:13:47] You know, if Florence Nightingale wasn't observing surgery, if she was observing some other data set, we might have had a different approach to hand washing.

[00:13:57] Or maybe we had gone another 50 years without really discovering how important hand washing was in infection prevention and control.

[00:14:06] So training data matters.

[00:14:08] Having enough data to train your algorithm is an important part.

[00:14:13] And we have to be transparent.

[00:14:15] You know, if we're going to put a model into the market, nurses should know what data was trained, what that model is good at, what it's not good at.

[00:14:29] And, you know, our Office of National Coordination and, you know, groups that have some regulatory input in the United States talk about, you know, here's what it means when we are transparent and we can build trust.

[00:14:43] We're going to be transparent with the clinicians who are actually going to be transparent and we can build trust in care.

[00:14:50] And for a long time, we've had what we would call an evidence-based system, right?

[00:14:55] You have an EKG machine and it's or a monitor in the build on the bed and there's an alarm.

[00:15:00] And the alarm is going to be sounding at some set points that are really just rooted in evidence.

[00:15:08] There's been 50 research papers that show, you know, that if your heart rate is more than 220, that's concerning.

[00:15:16] And we need a nurse to come into you.

[00:15:18] And so those kinds of evidence-based models have been integrated in our workflow for many, many years.

[00:15:27] The kinds of models we're bringing forward today are more of the models you hear like, again, having a conversation with a chatbot and asking a medical question, getting medical advice.

[00:15:40] And these kinds of models have to have clinicians part of the design, the development, the testing, so that we know if that advice is in fact based on best practice and the model is reinforcing best practice and making suggestions to say, here's what I know about your, you know, you just got a cut.

[00:16:02] But certainly first aid, people can take care of first aid at home.

[00:16:06] At what point is that cut too deep when I might need stitches?

[00:16:09] This is a question I get all the time as a nurse.

[00:16:12] Don't you get pictures?

[00:16:13] You get, you know, you get these crazy pictures that come to you on your phone.

[00:16:16] What do you think about this?

[00:16:17] Should my friend go in?

[00:16:18] And I'm always saying super glue, super glue, super glue.

[00:16:22] But you know that there's this idea that maybe 80% or 85% of some of our medical questions, our health change questions all over the world could be answered by an expert.

[00:16:39] Without having to go to that expert, if that intelligence can be built into these models.

[00:16:45] And that's the really exciting promise of responsible AI and clinical informatics.

[00:16:49] So that vision for me looks like this.

[00:17:16] A patient arrives in any healthcare setting and a nurse can go into their technology and say, hey, tell me everything important I need to know about Cheryl Field.

[00:17:33] She just showed up and I don't know her.

[00:17:34] And if you could just imagine integration, think of a data exchange that's enhanced between hospitals, physicians' offices, a convenient MD.

[00:17:47] Maybe I slipped into a boutique convenient MD or an urgent care a few days ago.

[00:17:54] I picked up a prescription in a pharmacy across town.

[00:17:57] All that information spanning any health setting, like that to me is utopia.

[00:18:03] So that that clinician knows instantly and in real time what they need to know about that human.

[00:18:11] And then they can best care for them.

[00:18:13] We lose people every day in emergency departments because they come in with no information.

[00:18:18] And what we don't know does hurt them.

[00:18:20] And so I would be spending that billion dollars on nurses, giving them advanced training and education in AI technology so that they could be in there could be more of me in pivotal roles building these United Health Records.

[00:18:38] Where, again, artificial intelligence is spanning and we're able to get this real-time access.

[00:18:46] And it's not an easy problem to solve.

[00:18:48] We have interoperability challenges, even with consistent ways for sending messages between providers that, you know, groups like that use like HL7.

[00:19:00] All right.

[00:19:00] The Office of National Coordination has said you should send messages like this.

[00:19:05] There are still errors.

[00:19:06] It's not perfect care coordination.

[00:19:10] But I would be investing those billion dollars to connect our skilled nursing, our independent living, our assisted living, places where our seniors, those people who are, you know, probably not carrying their entire medical record on a thumb drive.

[00:19:26] Right?

[00:19:26] They probably don't have a barcode you can scan that's syncing to the cloud with everything about them on it, although those products are out there.

[00:19:34] They could have primary record.

[00:19:41] They will be fine.

[00:19:42] And these kinds of things are here and they are coming.

[00:19:46] But even primary record today hasn't had the billion dollar infusion to make all those connections happen.

[00:19:54] And so I think that I think that there was a big sector left out of that investment.

[00:20:00] And it's not free.

[00:20:01] This kind of technology, the kind of investments that organizations need to make would really be well served by that billion dollar grant and would be well led by nurses, advanced nurses who have right kind of training,

[00:20:16] who also understand the importance of the care coordination so that you can have an integrated care plan.

[00:20:23] And we don't miss, we don't miss a beat with that human.

[00:20:26] We don't make a mistake with that human.

[00:20:29] I know we set people back when we don't know what they, we don't know how they walk.

[00:20:34] So we assume, hey, you should stay in bed.

[00:20:36] Don't get up by yourself.

[00:20:38] And or the opposite, people who should have assistants who are just, you know, coming, coming out on their own.

[00:20:45] And so I think that this, this kind of enhanced unified data set would allow for comprehensive and dynamic automations and alerts.

[00:20:57] So it becomes the data on which, on which artificial intelligence really runs and augments that workflow for nurses, clinicians, physicians in these decision, you know, decision making roles.

[00:21:10] I think that becomes, that would be step one for me in real time, get quick access to information.

[00:21:17] Then at a population level, we have advanced analytics and predictive tools, you know, a longitudinal view of a human being with really the pattern recognition is so critical to care outcomes.

[00:21:33] And so I'd be investing there.

[00:21:35] I think it would not only save, not only save lives, but it would enhance the quality of life for, for our seniors and reduce those feelings they have of fear and uncertainty at a time when their health is changing, they're losing control and give them confidence that, you know, one of the things I hear all the time from seniors, because they'll experience a med reconciliation, right?

[00:22:05] You go into the emergency department and the first nurse asks you about your medications.

[00:22:08] The doctor asks you about your medications.

[00:22:10] You go up to the floor and the first nurse asks you to do a med reconciliation.

[00:22:13] The next morning, they ask you to do a med reconciliation.

[00:22:17] And you know what that senior says to me?

[00:22:19] These guys don't talk to each other.

[00:22:21] They've all been asking me the same questions for two days.

[00:22:26] And it's so, yes, we're doing med reconciliation verbally because we know that medication transparency and coordination is the biggest risk, right?

[00:22:36] So we're doing it verbally, but to our clients, it's like, we don't know what we're talking about.

[00:22:41] We don't know what we're doing.

[00:22:41] Imagine solving that with, you know, care with an integrated health system that was really error free.

[00:22:49] And when there was an error, there was an AI running to say, hold on a second.

[00:22:54] Cheryl was at the pharmacy yesterday.

[00:22:55] I'm not sure that you have the right cardiac medication because it was a new medication added to the plan of care yesterday.

[00:23:02] And this integration of a care plan in real time to providers, I think, gives us confidence that we have all the information that we need to do well for our seniors.

[00:23:13] And they're so vulnerable.

[00:23:15] And, you know, that's my space.

[00:25:07] Absolutely.

[00:25:07] It made me think of a few things.

[00:25:09] First of all, you know, the polypharmacy, the multiple medications.

[00:25:13] And I ask that question differently.

[00:25:14] I ask people, what do you put in your mouth every day?

[00:25:18] Like what's going in your body?

[00:25:20] Not just your medications, I think, that they take out of a prescription bottle, but to your point, supplements, protein, collagen.

[00:25:26] You know, what else are you putting in your body?

[00:25:28] Excuse me, every day.

[00:25:30] And those medication lists are quite lengthy.

[00:25:32] So it's very easy to forget one.

[00:25:34] I'm not going to remember all my stuff.

[00:25:37] And when you put that into, you know, the combination and you mentioned about augmented artificial intelligence.

[00:25:44] These are expressions that the American Nurses Association, the American Medical Association, they've put forth in their guidance to be very clear to clinicians.

[00:25:52] These are to support your decision making in workflow.

[00:25:57] And in fact, you know, the regulations make it required that organizations who are developing these models, one, they have to make sure that the user, they know they're using one.

[00:26:10] You know, you have to know you've turned on something that is either evidence-based or predictive in nature.

[00:26:16] It's, you know, and then you have to actually show that source.

[00:26:21] Like if we're using a certain type of criteria to decide that, you know, this is tachycardia, this is a high heart rate.

[00:26:28] This is what evidence was that, you know, was 50 years of cardiac studies.

[00:26:33] Which one did you cite?

[00:26:34] Or if you're using training data like Florence Nightingale did, and she said, hey, I predict that guy's going to have an infection because doc doesn't wash his hands.

[00:26:43] If you're building a predictive model, same thing.

[00:26:47] You are transparent about how, what data you use and what the source of that information is.

[00:26:52] And because clinicians are always going to be the human in the loop who are using these tools in real time in their clinical decision-making in the American Nurse Association, American Medical Association, they reinforce that.

[00:27:06] It's not replacing us.

[00:27:07] To your point, it's making us better.

[00:27:09] I love the idea of supporting a new graduate nurse, a tired nurse, an agency nurse, a nurse who's really probably really exhausted from a double shift who could easily, human error, miss that from a cognitive load perspective.

[00:27:26] But instead, in their workflow, they're going to have information coming from the model that suggests they want to go back and look at something and calling it to their attention.

[00:27:41] And I just love that.

[00:27:42] I think, to your point, we know human error is real and fatigue is real.

[00:27:48] But this is a great way for us to invest our billion dollars into tools that actually support and reduce the cognitive load and the stress level for clinicians in those workplaces.

[00:28:03] And give them that support that technology can, in fact, deliver.

[00:28:09] And technology doesn't get tired.

[00:28:11] It's running 24-7, and it has monitoring 24-7.

[00:28:15] That's part of responsible AI.

[00:28:17] So we know that the model is predicting accurately as we had designed it to do, and we monitor that every single day.

[00:28:48] So Prepared is everywhere where you buy your books if you want to grab a copy of it there.

[00:28:53] If you wanted to get a personalized copy for someone that you love or you want to write a little note inside to encourage someone that's on a healing journey,

[00:29:03] then you can actually go to my website, which is www.cherylfield.com, just like you see in the podcast here.

[00:29:11] And we can put in the show notes.

[00:29:13] And you can order one directly from me anywhere in the United States.

[00:29:17] I'll mail that out to you at no charge and encourage your loved one to continue on their healing journey.

[00:29:24] It's kind of a book you don't want to share with a friend, though, because it's like your recipes.

[00:29:29] You don't give away your original recipe card in case you want to make that dish.

[00:29:34] So you keep it on the bookshelf because if anyone that you love ends up with a sudden change in health and you're right back headed to the emergency room,

[00:29:42] you need your checklist and you need your questions list and you need to know what to bring and what's going to happen.

[00:29:48] And so you want to pull that book right out again and put in your bookmark and head off to, you know, head off to that institution.

[00:29:55] And you don't want it to be lent out to your friend when it's I have a friend who keeps lending hers out and keeps buying new copies from me because she hasn't gotten them back.

[00:30:09] That's right. That's right.