Subscribe
Share
Share
Embed
Come with us as we talk to the clinicians making AI part of their daily workflows, and using it to build deeper connections with patients.
Welcome back to Trust but Verify, the Evidently podcast. I'm Andre. I'm joined by my cohost, doctor Kai Romero. If you don't know Kai, Kai is an emergency physician and palliative care doctor. And we both work for Evidently.
Andres Krogh-Walker:Evidently is a clinical data intelligence platform that ingests everything in the patient record, indexes it, finds the associations, and turns that into insight for clinicians, usually right at the point of care. And we started this podcast to explore these interesting people that we keep finding who are making AI part of their everyday workflows and finding new ways to amplify the work they're doing. In this second episode, we get to talk to Terry Shoulder. Teri is a lot of things starting with, just incredibly interesting to talk to. She was a chief quality officer at BayCare, Kettering, and a handful of other health organizations where she led teams through some pretty dramatic change, in documentation quality and how the teams execute inside, inpatient hospital care coordination.
Andres Krogh-Walker:She's also someone who always has the numbers at her fingertips, and that's a really impressive thing when you're talking about the kind of scale that she's operated at and the kind of impact she's been able to have.
Kai Romero, MD:I think the thing that really struck me when I first met Terry was something that I've watched happen to other people when they talk to Terry, which is like, you know, everyone's trying to solve this big problem, this problem of, you know, hospital readmissions, length of stay, making sure you know, trying to operate on razor thin margins. And it's kind of like when someone if you're, like, driving down the road and you're looking all around you trying to figure out how fast am I going, Terry's the person that's like, do you wanna look at the speedometer? You're like, What? There's a speedometer?
Andres Krogh-Walker:There's a number?
Kai Romero, MD:There's a number I can look at that would help me figure out how to gauge safety and my proximity to other moving objects. The person that kind of really in an incredibly friendly, low ego manner points to the solution that you've been looking for for a really long time. And it's so simple. And it's so clear. And you're like, nobody told me those large flashing numbers in front of me had meaning.
Kai Romero, MD:And now I know.
Andres Krogh-Walker:Yep. As always, let us know if you have feedback. Let us know what you want to hear more of, want to hear less of, and other guests that you think would be interesting. And without further ado, here's Teri Schoulder. Teri, thanks for spending some time with us today.
Andres Krogh-Walker:Let's start with a little bit about you. You've got a pretty interesting background.
Teri Sholder, RN MHA:I appreciate that. And I'm really honored to be on this podcast. My name is Teri Scholder. I am a registered nurse and a former chief quality officer. So for the last almost two decades, I've led some of the country's most challenging hospital turnarounds.
Teri Sholder, RN MHA:At BayCare Health System and Clearwater, we reduced readmissions by nearly twenty percent. You had a lot of folks reaching out at that point saying, Oh, what did you guys do? Also, AdventHealth system reduced length of stay by over half a day. At the same system, achieved ninety percent success on denials appeals.
Kai Romero, MD:Okay, I'm curious about what led you from your original RN degree? Where did you first practice? What were you first doing? How did you get into the whole kind of healthcare executive quality part of things?
Teri Sholder, RN MHA:Well, know, it was never intentional. I never set out to say, I'm going to be a healthcare executive. Didn't set out to be basically a leader. I naturally love being around people. That's what drove me into nursing.
Teri Sholder, RN MHA:Actually, the reason I went into nursing is I loved waitressing. I liked that interaction. I liked making somebody feel better, making somebody's evening. They took the time to go out. So anyway, that translated into nursing fairly well.
Teri Sholder, RN MHA:But I wasn't sure exactly, I loved bedside, so I spent a lot of time in ICU. Also spent time as a float nurse. And the great thing about floating is just that diverse knowledge and skill that you attain. You're going down to the emergency room and helping there, going to help with a two d echo. And so you just get this really diverse experience.
Teri Sholder, RN MHA:And with that, everywhere I would go, would think, Wow, I wonder if they did this this way, if it would be more effective or more efficient. And about that same time, of course, value based purchasing was making inroads into hospitals. So I joined the quality department as a clinical documentation. I applied for a clinical documentation improvement specialist because folks in healthcare, if you're around back in twenty ten, 'eleven, the MS DRG's were coming. The payment system was changing under the inpatient payment system under the Affordable Care Act.
Teri Sholder, RN MHA:And so, clinical documentation improvement programs were springing up all over. Well, prior to that, I had managed, I spent time as a manager of the emergency department a hospital for a while. Really wasn't my thing. I mean, of those of you who are clinical emergency nursing and ICU nursing are just completely two different animals. And so, I spent some time there.
Teri Sholder, RN MHA:But anyway, when I applied for this position, they needed a leader over the clinical documentation improvement specialist. And so, that's what I did. That's my first introduction into quality, because at that time it did live under the quality umbrella, rightfully so, in my opinion. But anyway, so made my way up in short period of time to the chief quality officer of that hospital. Within six months, made it at the system level.
Teri Sholder, RN MHA:And what I really loved about, I really learned so much about systemness and how do you align a health system around common goals? But that's really how I kind of fell into it. It was never something I envisioned. But then once I got into it and just had some early successes in looking at operating models, dissecting them, figuring out where can we gain efficiency at the same time with that patient right in the center, how can we improve that care? How can we improve the experience of that patient and create workflows and processes that drive those optimal outcomes?
Teri Sholder, RN MHA:So that's how I got there.
Kai Romero, MD:I've always said that to be a good ER doctor, you would probably also be a good waitress. You have to remember 19 item checklists as you walk by people and like avert your eyes in a non rude way when you can't stop to do what they want. I just there's so much overlap. But I think, what I'm curious about is, and you may have already alluded to this, what was, and I think what's interesting to me to think about in terms of your career, so you were doing this within the health system, you've now come out and as a consultant are kind of bringing some of these same insights in. What's the most joyful part of your work, would you say?
Teri Sholder, RN MHA:The most joyful part when I was an executive, I love leadership. And there's nothing that makes me more proud. Recently, there was a person that I worked with, actually ran into her at a conference that I worked with. And she at that point in time worked in the quality department, not in any type of leadership position, but has slowly grown, eventually was a manager. But anyway, when I saw
Kai Romero, MD:it, I said, Oh, how
Teri Sholder, RN MHA:are you doing? Of course, saw this on LinkedIn. It wasn't a surprise to me, but she's now a chief quality officer for a health system. And that's just, you know, that to me is leadership success. Oh my gosh.
Teri Sholder, RN MHA:You know, just to see these people who are working so hard and so bright, you know, just become leaders. And you know right off the bat, wow, I bet her team just absolutely loves being there. So that really what made me excited. I mean, the work itself was great, don't get me wrong, but just that leadership part of it was something that I really enjoyed. And now that I'm in the performance improvement company side and not working for a health system, seeing the problems that you see in Becker's every day, throughput's an issue, regulatory compliance, Medicaid changes are coming.
Teri Sholder, RN MHA:And with my background understanding regulatory compliance and staying very abreast on what's going on, knowing that we can patch all of that together and that we can fix that, it's that same feeling. Back when I was at some of the hospitals in my early career and I looked at sepsis and I said, Well, if we do this and this and this, then we can reduce mortality. It's that same exact feeling now through this lens. I feel the pain because I've been there as a health system leader. But to have the solution and to put that in place and seeing it work and seeing the results is just one of the most rewarding parts of doing this that I get the most excited about.
Kai Romero, MD:I definitely think that one of the things that's really interesting about your work is that you're tackling one of the very thorniest problems in medicine and kind of approaching it with a like pleasant, pragmatic and highly functional solution. And I think it's so there's something about it that's really surprising because I think often with these very complex problems, people will immediately complexify, don't know that that's a word, the solution And everything will spin out into jargon and really like impenetrable strategy in this very rapid way that makes it feel like, well, there's really no way to fix this. And I think one of the things that struck me about you was that you a seemed undaunted by the challenge, which not most people do. And b, your approach was so rooted in practical things.
Andres Krogh-Walker:That's surprisingly human.
Kai Romero, MD:Yeah. Yeah. Yeah. Absolutely. Absolutely.
Andres Krogh-Walker:I had a mentor a long time ago that said good leaders leave good leaders behind. And that definitely resonates with what you were saying before. The other thing is in the back of my mind, Terry, is you've seen this whole growth of documentation quality that I think has probably always been intertwined with technology. But I'm curious from your perspective if you think technology has made the relationship between providers and documentation quality specialists better or worse, or if it's just always been kind of a glue in the middle.
Teri Sholder, RN MHA:Yeah. It's interesting. You know, Prior to So I was around with the transition from paper to EHR. At that point in time, the clinical documentation improvement specialists would write queries to the physicians. And I had one physician, one hospitalist that would repeatedly just The response would be, Stop harassing me.
Teri Sholder, RN MHA:And so, this was somebody that I knew. When I ran into him in the hall, I would exchange pleasantries, just had a lot of respect for this individual because when I was a baby nurse, I learned a lot from this physician. And so, what I did was I actually brought him in. I said, I learned so much from you. And all I ask is that you share that knowledge with these teams.
Teri Sholder, RN MHA:If they do ask you something that's inappropriate, could you just tell them? Just say, Hey, you asked this question and that's not even the patient's problem. So, because they're learning too. And even now, there are some of the AI solutions that are built to reduce the burden and the friction of clinical documentation with the physicians. I think the adoption is extremely important, but I also think understanding the processes behind that to optimize it and leverage it, not necessarily seeing the technology as the solution as much as it is the enabler in some of those cases.
Teri Sholder, RN MHA:It's another thing too. I mean, I'm sure you've heard New York Presbyterian, the nurses walked out on strike just overworked. My And husband's a nurse, and so I get to hear the frustration of the duplicate, triplicate documentation. And so when we built this care coordination model, we built it intentionally to work around the physicians and the nurses because a lot of things, when you have barriers that crop up in discharge delays, you have a late consult, you have a physician that says, Oh, they're ready to discharge if cardiology okays, cardiology's already rounded that morning and not back till the next day. Those types of things place extra burden on the physicians and the nurses trying to care for patients.
Teri Sholder, RN MHA:And so, we built this, we built it intentionally to address those things transparently so that the nurses and the physicians do not have to. So, specifically for those care coordination teams, we were pretty successful at putting the model in place. Obviously, we've seen the results that we wanted, but I knew that there was another, that if we just had the technology to enable this, to give us information sooner, to be able to identify, wow, this patient, all of a sudden there's a new diagnosis of UTI or of a DVT that could shoot off to quality right away. If we only had the technology to basically superpower this model, what great outcomes we would see in these patients. Know, mitigating risk of delays, addressing the actual risk factors that made them high risk to begin with.
Teri Sholder, RN MHA:When you think about the EHR and the incorporation of the EHR into health systems, how
Kai Romero, MD:do you think that shift changed the way you thought about quality, quality capture, about how folks interact with with their how how aware they are of quality metrics, what needs to be measured, what doesn't, how aware institutions are of what a challenging job it is to capture those. I'm just curious about kind of how those two things intersected in your mind and in your career.
Teri Sholder, RN MHA:Well, was kind of interesting. So, like I kind of mentioned before, the EHR, knowing that the health system I was at, we were getting Epic, we just thought that it was the end all be all. It's going to save all things. Not even in the need of quality department because quality is going be so much better. I mean, there were so many good things about it.
Teri Sholder, RN MHA:I don't mean to imply that there weren't. There were a lot of good things about it. However, we over customized it, and I don't think that we were alone. I think there are a lot of health systems that over customized it. And you know why?
Teri Sholder, RN MHA:Because they had to have it fit their processes. And how many of those processes were already broken? And we know as clinicians, know, Kai, you and I, yes, lots of broken processes in these hospitals.
Kai Romero, MD:And I broke some of them, know? I'll take that on. I'm responsible for some of that breakage.
Teri Sholder, RN MHA:Can't help. Thought some of my workarounds were pretty darn clever. Right? But yeah, we found them. We found a way, right?
Teri Sholder, RN MHA:Because our eyes were on that patient and what that patient needed at that point in time, and we were going make sure that happened. And so initially, I think a lot of it was just not knowing what the EHR was going to do. Because in my mind, it was going be just like a chart. You were going to have tabs that you clicked, right? H and P, and you were just going have tabs.
Teri Sholder, RN MHA:And everybody would be able to see the same thing. And so that was our first, Oh my goodness, what are we going to do? Not everybody has the same view. Case managers can't see what quality sees, quality can't see what risk management sees. It's like, ah, we've really got And so, there was a lot of work to try and figure out, because of course, the security in HIPAA was in the back of everyone's mind as well.
Teri Sholder, RN MHA:Does risk management need to see this? Why does Terry Shoulder, the director of quality at the time, need access to the coding? Well, because I needed You know what I mean? So, it was a lot of those things that had to be worked through that I It wouldn't surprise me if a lot of quality folks back when that was happening have not experienced also. But the good thing is that once you realize that Epic is not going to act as, at that point in time, like an analytics tool, but that's where the EHR was so valuable, because you could just pull that data.
Teri Sholder, RN MHA:If it was a discrete field, you've got it, you can pull it in to your quality data. And so it made root cause analysis so much better. It made things like mortality reviews so much easier because you had access to the coding. You see that, wow, this person was at risk mortality of four out of four. Severity of illness, four out of four.
Teri Sholder, RN MHA:Probably did everything we could. So, it did make things like those reviews a lot easier. But keep in mind, they were still reactive. And so, that's another thing that, just based on my background, I thought we can solve this with this model. And that's what I had done in my previous life.
Teri Sholder, RN MHA:I brought all of those together and said, All right, how can we defragment this? And not only defragment it, but we need somebody that owns it. We need an accountable individual that's pulling this information from all these disparate areas, orchestrating it, figuring it out. It's just like with quality, for example. I mean, we use quality as an example.
Teri Sholder, RN MHA:So, a patient comes in and they develop a hospital acquired infection. Knowing that right away, using this AI to surface that and say, a hospital acquired infection. And that enables faster treatment. But the question becomes, who's responsible for that? Do you wait till the next day until infection prevention and control is rounding?
Teri Sholder, RN MHA:Now we do, right? That's when we find out. That's when the physician will see it, obviously, But that's when it's addressed internally by the infection control team, where I wanted that addressed when it happened. As soon as you see it, I want that alert or whatever going to that care coordinator so that then she can orchestrate and make sure that the right treatment is in place by collaborating with the physician. So, it's the same thing think that you're seeing.
Teri Sholder, RN MHA:Are many subspecialties within all these ancillary departments that have their own workflows. And while they work in parallel with everybody, they're not really working toward an aligned goal, you know, such as that geometric mean length of stay.
Kai Romero, MD:And I think one of the things that you and your group tends to have a ton of clarity on is like, how can we pick a North Star that captures, you know, patient like the the individual patient best outcome, quality metrics for the hospital, all these other things, and realizing like those already exist, you know, like they're out there. Everyone doesn't know what they are. Everyone doesn't hear about them or benchmark themselves against them, but they're there and they're there for a reason. I think sometimes about physicians are often encouraged to just solve the problem themselves, you know, figure it out, be creative.
Teri Sholder, RN MHA:That's one of the first things we ask when you're going to a client, who owns length of stay in your hospital? Who owns length of stay? Do you know what they say pretty much every time? Everybody. The physicians.
Teri Sholder, RN MHA:Physicians and length of stay.
Kai Romero, MD:News to us.
Teri Sholder, RN MHA:The physicians write the order. No doubt about that. The physicians write the That's true. Do your physicians call the nursing home administrator and ask why they haven't accepted the patient yet? Building the process around that so that the physician only writes the order.
Teri Sholder, RN MHA:And of course, the face to face if they're going to home health. But only does that is going to just make it much more efficient. First, we anchor it to the MS DRG, which goes beyond anchoring the care progression plan to the diagnosis, because then all of a sudden we're adding acuity. So, if it's an MS DRG that has an MCC, a major comorbidity or complication, then that has to be considered in that care progression as well. And it also gives you a different length of stay, a geometric mean length of stay target.
Teri Sholder, RN MHA:So, the way readmissions come in is that we have a risk assessment running continuously in the background that tells us, and it's not just risk of readmission, it's not just the social determinants, it's not the diagnosis itself, but it's also risk of discharge delays and barriers that could occur. That's happening in real time. And then as far as the readmission reduction goes, another thing we do is that we've number one, anticipate the discharge disposition. And then we make sure it's not about getting patients out faster. It's about making sure they're ready to discharge.
Andres Krogh-Walker:You had shared a statistic with me about the number of readmissions that are due simply to a patient not being able to afford the medication they were prescribed.
Teri Sholder, RN MHA:When medication reconciliation came out and became a thing, it was like, all we're going to have the physicians do medication reconciliation. That's all well and good, but why not the pharmacist? If physicians knew as much about pharmacy, pharmaceuticals, and if they knew as much about that, why do we have pharmacists in ICU? Right. Right?
Teri Sholder, RN MHA:And so it's putting more on the physicians. And so we did a study down at, it was in BayCare in Florida, on that very thing, on how many readmissions were coming back. Well, first we started with errors in the med rec, and it was significant. And so, what we did was we brought pharmacists in, and not for all the patients, it wasn't necessary for all the patients. It was for those high risk, high volume, the CHF patients with MCCs.
Teri Sholder, RN MHA:So, we really made it all about the highest risk patients. And we had pharmacy do those medication reconciliations, and the error rate went from about eighty five percent down to about nineteen percent when it was all said and done. Significant. And then that was when we reduced readmissions by about nineteen point five percent as well. The one thing that's really important from an operations standpoint is clarity in job descriptions and expectations, and then accountability as well.
Teri Sholder, RN MHA:One of the things that we identified early on, this was when we were creating electronic order sets. It's been a while. But with electronic order sets, what I've noticed is that the social work teams were staying six, 07:00 at night. It's like, Oh my gosh, your assignments seem appropriate. What had happened is that as they were creating these order sets, they were automatically checking social services consult.
Teri Sholder, RN MHA:Every social services consult took about twenty minutes. Even podiatry had social services consult. It's like, oh, so anyway, yeah, ended up finding 50% of their day was non value added working on these. Yeah, yeah. So that just goes to I mean, it just goes to show you that when, you know, all of these departments that make a huge impact on patient care, efficiency, effectiveness, quality, are operating in their own silos, then you're going to have a lot of fragmentation and you are going to have a lot of reactive and a lot of issues that surface closer to discharge rather than identifying those closer to the beginning.
Andres Krogh-Walker:Terry, what do you think the biggest misconception is about AI implementation or technology implementation into this world that you wish you could correct?
Teri Sholder, RN MHA:I think the biggest misperception is looking at it as the solution versus the enabler. It's not going to fix physician documentation. It's going to enable them to be much more efficient by summarizing the data, bringing it all together. Then I also depend, I think the solution, what the solution is matters too. So, can be the solution obviously in a lot of cases.
Teri Sholder, RN MHA:A chart summarization is a perfect example. My gosh, that's just And we've seen the articles, saves tons of time, and that's wonderful. But the solution to an underlying process driven outcome, such as length of stay reduction, where you really do have to have The model has to be in place, has to be effective and efficient. Or you place the AI on top of that. You have a few things.
Teri Sholder, RN MHA:Have, okay, so it's serving up the data. Who does what with it? I don't know. Who's accountable for using it? I don't know.
Teri Sholder, RN MHA:And so, I've said this digitizing dysfunction, well, in some cases, you're digitizing chaos. You're putting this And what's so frustrating about that is that the AI could be a fantastic solution, but the perception of the users is that, well, you guys promised us it would do this or that and this. Well, if you have the workflows in place, then that AI can not only maximize your processes and optimize those processes, but also just huge productivity gains to help your staff reduce the frustration and the stress. So, looking at it that way, I think that's a misperception. But I also believe that using AI, be thinking about your ROI.
Teri Sholder, RN MHA:What is the ROI you're trying to achieve from this? And if you look at that ROI, say for example, if it's length of stay, then you just got to go back, okay, who owns length of stay? So whoever owns length of stay needs to make sure that they're leveraging that AI to its maximum potential. And then you definitely get the value out of it.
Andres Krogh-Walker:But that accountability is really, really interesting. I saw a, a screenshot earlier today from an IBM training manual for computers in the seventies. And the screenshot in all caps was computers can't be accountable, so computers can't make management decisions. The accountability is, I think, such a huge part of what the PACT model instills and and kind of organizes inside of a health system.
Teri Sholder, RN MHA:Yeah, and you're right. Because it goes back to that same question, who owns length of stay? And yes, we have heard people say, well, everyone really does. Well, you know, if everyone does and no one does, right? Someone has got to be accountable to make sure those barriers are addressed and the risks are identified and addressed and that the patient is really ready to go.
Teri Sholder, RN MHA:If a physician in the hospital has to be 36 on their panel that day, is that really the right person to make sure they're getting to the sniff on time? Probably not.
Kai Romero, MD:I wonder if you could wave a magic wand, what is one clinical workflow problem that you'd want, well, it could be AI to solve that nobody has cracked yet? What's something that you see kind of over and over again, aware that you're really kind of rooting a lot of your solution and, in human interaction and transparent communication. What's one workflow issue that you kind of have your eye on where you're like, man, we still haven't figured that one out yet?
Teri Sholder, RN MHA:System alignment, systemness, systemness. You go to a hospital and it's like, Oh, wow, we're a leapfrog grade A, for example. The hospital in the same system is a D. So, it's systemness. And it's not a difficult solution to align your whole system around a common goal.
Teri Sholder, RN MHA:It's setting that goal, making sure it's realistic, achievable. And so, that would be my answer. If I could wave a magic wand and have these systems, all right, systemness. You have one hospital that's just knocking it out of the park, whether it's sepsis, whether it's mortality, whether it's readmissions, what are they doing? And how do you align all your hospitals around that same goal?
Kai Romero, MD:Yeah, it is remarkable how different the culture can be between one part of it and another.
Teri Sholder, RN MHA:And
Kai Romero, MD:some of that rooted in, you know, there might be a trauma center and all of the complications that happen with trauma patients and not knowing who your patients are and whatever that may be. And then next door, an academic medical center versus a community center and you realize they ideally, they do all have the same priorities. And in reality, it's like not just the patients are different, not just the people that choose to work there are different because there is sometimes a choice you're making about identity or mission or something about the place that you work. But aligning yeah. Aligning all of that seems really hard.
Kai Romero, MD:So for health systems just starting a journey into kind of expanding their use of technology and AI, what's one thing that they should invest in before buying anything?
Teri Sholder, RN MHA:I believe the best thing to to invest in before buying for one thing, I think, you know, what what are you trying to fix? What what are you trying to achieve with this technology? You know, what is the end result? So once you figure out what the end result is, how is it going to impact your team? Who are the users?
Teri Sholder, RN MHA:Is it going to change or disrupt workflows? I guess the biggest thing is, and I know I'm not answering this very articulately, I think the biggest thing when a health system is going to get a new technology is really solidifying what it's going to fix. Is it specifically for length of stay? And then you've got to ask how, how this is going to happen. I mentioned being at a conference recently and there was a booth we went up to in the company that said, Oh, we reduced length of stay.
Teri Sholder, RN MHA:And after talking with them, it was about as indirect as you can imagine, it was about building the culture of the organization. That's all well and good. So I guess that's what I would say. If you're a health system and you're looking at a technology, number one, figure out what it's gonna fix. And then make sure you understand how, the how, then, yeah, before you go a step further, because there are a lot of technology companies that I've talked to as a health system executive that made a lot of promises, and they could say the what, but sometimes we're challenged to explain the how.
Andres Krogh-Walker:Teri, thank you so much for spending time today. I feel like I always learn something and need to take a nap after every conversation with you, and this is no different. If folks are trying to find you and would love to talk more about what you do, where's the best place for them to go to find you?
Teri Sholder, RN MHA:Well, we have a website. It's shoulderhealth.org, and there's a link there to to send us an email.
Andres Krogh-Walker:Awesome. We'll drop links in the description too. You're listening to Trust But Verify, the Evidently podcast. You can find us online at evidently.com/podcast, and you can find us anywhere you subscribe to podcasts. See you next time.