Subscribe
Share
Share
Embed
The Transform your Teaching podcast is a service of the Center for Teaching and Learning at Cedarville University in Cedarville, Ohio. Join Dr. Rob McDole and Dr. Jared Pyles as they seek to inspire higher education faculty to adopt innovative teaching and learning practices.
Are we implementing assignments that require students to use AI to the level that is needed? I think as faculty, we have to be comfortable first with AI use before we then start making assignments like that.
Narrator:This is the Transform Your Teaching podcast. The Transform Your Teaching podcast is a service of the Center for Teaching and Learning at Cedarville University in Cedarville, Ohio.
Ryan:Hello, and welcome to this episode of Transform Your Teaching. In today's episode, Dr. Rob McDowell and Doctor. Jared Pyle start a new recurring segment on our show called Needle Movers. In this episode, they interview Doctor. Michele Dodds, a professor of nursing at Cedarville University. Thanks for joining us.
Jared:Well, Rob.
Rob:Yes, Jared.
Jared:We are here today interviewing, Dr. Michele Dodds.
Rob:Yes.
Jared:Who is one of our assistant professors at the School of Nursing here at Cedarville University. And I have had the privilege of serving with her, as I'm leading the AI fellows at our school. And she is one of our AI fellows specifically for the School of Nursing here, has been doing tremendous work. I thought we should have her on to talk about what she's been doing and, all that stuff. So welcome.
Michele Dodds:Thank you. Glad to Doctor. Be
Jared:Dodds, thank you so much for joining us. And, we talked about this before we started recording. So I mentioned you're an assistant professor and I mentioned that you're also an AI fellow, but what else is it that you do here at the university?
Michele Dodds:Well, in the school of nursing, I teach in the community health courses, leadership and management. In the graduate program, I do the global public health certificate. And, I'm actually starting to be the nursing health care leadership, track coordinator coming up soon.
Rob:That's great. She's also an NP, I think she said.
Michele Dodds:Yeah. And I work in clinical practice as a nurse practitioner
Jared:Right.
Michele Dodds:Which I've been for twenty six years now.
Jared:Wow.
Michele Dodds:Been a nurse for thirty two. I was a nurse before I was a nurse practitioner.
Jared:Okay. Okay. Interesting.
Rob:Yeah. Well, it's a pleasure to have you here talking with us. Thank you for doing it, first and foremost.
Michele Dodds:Well, thank you for having me.
Rob:But it just some of our interactions that we've had in our our AI fellow meetings, It's been really wonderful to hear things that you've been doing, what kinds of interactions you've been having with your fellow faculty members in terms of generative AI, and and we just wanted to kinda, like, pick your brain. What's going on? What's the impact been for you in terms of, just what you do in the nursing school or or even nursing practice?
Michele Dodds:Well, let's start with nursing practice. So really interestingly, I had an interest in AI about eight to ten years ago. I had just transitioned, went to Wright State Physicians, working there in primary care from a rural practice that I was working in for like eighteen years before I went there. I went there because I was starting my doctoral degree because I knew the Lord had called me to teach. And so I wanted to be in an academic environment.
Michele Dodds:And, I was doing a project there to look at AI in primary care.
Rob:Okay.
Michele Dodds:And really then, like, I kind of thought it was gonna like, this is what it was gonna do. So what you would find is it would do patient reminder visits. It's running through the electronic medical record, looking for those type of things. I really thought that was going be the limit of AI at the time, you know, and we would talk about like, oh, it's going to do reminder visits. We can use it to pull data for stuff like, we would look up ICD-ten codes for like all of our diabetic patients in the practice so we could see what their a one c was and figure out what their glucose control is and what can we do as a population to improve their care.
Michele Dodds:But little did I recognize how AI has progressed and just changed the way we care for patients.
Rob:Okay. Hold on a second. Not just that it progressed. You just said change the way, present tense
Michele Dodds:Mhmm.
Rob:You care for patients.
Michele Dodds:Mhmm.
Rob:I interpret that as
Michele Dodds:now. It's now. Explain. Yes. So so for one, if I go in to see a patient and I'm developing a diagnosis of what's going on with this patient, you know, I can take all of my clinical practice and experience and all the things I've learned and come up with all these potential diagnosis for a complaint.
Michele Dodds:But if I also use an AI tool in addition to that and say to that AI tool, like open evidence is one you can use. And I say, these are the things I'm considering in the patient. Can you give me additional ideas? It's going to give you additional diagnosis. There's no provider out there, no physician, no nurse practitioner or PA who's going to know every potential diagnosis for a complaint.
Michele Dodds:But if you can leverage that tool to improve the care you provide, then that's also gonna prevent misdiagnosis.
Jared:Wow.
Rob:So it's like almost having a team of other NPs or medical professionals around you who have access to these diagnoses or what is what is
Michele Dodds:it what is having an interprofessional team with you to discuss a patient case and make sure you come up with the optimal care plan. You don't rely on it, but you use it to supplement what you already know.
Jared:Okay. So using it to diagnose help with correct diagnoses. Do you see any other application?
Michele Dodds:Absolutely. I was just telling someone earlier about, even the Cleveland Clinic, like just in August released information, like a press release talking about they are in the process of using AI and they're doing a launch, like there's a platform launch, a step launch to it in their neuro ICU for AI to continuously monitor patients' brainwaves. And in that process, you have to think if you can have something continuously monitor something like an EEG that takes neurologists hours to read, we may be able to see before a patient has a stroke or see before they have a seizure. So we're intervening before we have this massive event with the patient who's already critically ill.
Rob:Wow. Reduces cost.
Michele Dodds:Oh, yeah.
Rob:Because you
Michele Dodds:don't have to outcomes go. Ultimately. Improving patient outcomes is our goal for everything.
Rob:Yeah. So, basically, you can sit there and watch for patterns.
Michele Dodds:Mhmm.
Rob:And then alert staff.
Jared:Yep. Oh. That's that's incredible.
Rob:Yeah. But it's also spooky at the same time. I mean, it's taking it to another level because there's probably no neuro doc out there that's gonna sit there and watch an EEG.
Michele Dodds:Right. They're gonna have to read it. Right? Yeah. But, like, now they're gonna have a tool that's gonna help them.
Michele Dodds:They can be spending some of their time optimizing what their treatment plans are for their patients on top of that.
Rob:So instead of having to go through that whole, litany of of proof, you know, that they have the data that they've got, They've got a tool that will show the patterns, and then they can go and validate based on where the tool said they've found the pattern. Absolutely. That speeds things up considerably. Wow.
Michele Dodds:And even now from a nursing perspective, like for our nurses who are graduating or out there working, I can think of things like, there's so many things that come up in the EMR that are AI generated. So let's say I get ready to go, give a medication to a patient. And a lot of times when a patient's in the hospital, they might be on 20 medications. There's no way every nurse is gonna know every drug interaction between all of those drugs in that chart, right? But AI will read that stuff and it pops up notifications in your EMR record so that you are tuned to like, oh, I need to check this before I give this med.
Michele Dodds:So it's giving you extra extra steps, but it really is to protect the patient and improve the care you're providing in lifetime.
Rob:So where is this actually in practice right now? Because you said it's happening now. Where have you seen it yourself?
Michele Dodds:Clinical dis what what we call CDSS, which is clinical decision support systems Mhmm. In the EMR are everywhere in practice. Okay. There's nowhere now that I haven't seen them in use. So I know they're in Epic.
Michele Dodds:My primary care practice, we were in, Athena. I've been in, eClinicalWorks, which are other platforms, but those clinical decision support systems are in all of those.
Rob:Wow. So we walk into our doctor's office just even for a, you know, regular checkup or something like that. They're gonna have those kinds of tools already available.
Michele Dodds:They're in there.
Jared:Wow. That's kind of cool. Do you foresee some resistance, hesitancy, on the part of doctors and such to use AI for this purpose because of the whole, like, you know, we would talk a lot about hallucinations and the fact that it could give a wrong diagnosis. Is there some hesitancy there?
Michele Dodds:I'm sure there is some level of hesitancy, but I feel like it's more with my generation. I'm an older, been out in practice longer, so less exposure to AI. But truly, I think specifically the younger generation, they are already using it. But I don't think it's something we can't unlearn. Like I've unlearned that myself.
Jared:Okay. Okay. So tell me, I'm about the patient side of it because now, you know, my chart is a blessing and a curse because my chart is like, you have your test results from your blood work before you hear from your doctor.
Rob:So you're So you look at it
Jared:and you're like,
Rob:what do you mean this is out of range?
Jared:That doesn't look right. So you can put stuff into a chat, GPT or a clot or Gemini. And let's say that it flies, cause you know, the internet is always like, oh, you have these symptoms, you're near death. Like you have a cough, you're going to die. So these over diagnosis stuff, how does that, how do you think that interacts with like a doctor then following up and maybe they say the test results are different than what a gender of AI would give a patient?
Michele Dodds:All right. I'm gonna tell you as a provider perspective, my chart is good. Federal regulation is the reason why the chart results are released like that before a provider has time to review them. I have a little bit of a negative attitude about that as a provider because I know that patient's full picture. One one lab result doesn't define a patient.
Michele Dodds:Like you need trends of lab results. You need to see the patient's clinical picture, but as providers, we know that the problem with AI on the outside of my chart, doing the example of what you're saying
Rob:Mhmm.
Michele Dodds:Is you're only getting a result based on one thing, whatever you tell it, whatever you feed it is gonna be the result you get. So if that result is I have this abnormal lab value, it's gonna say, oh, this could mean all of these things, and some of them are gonna be very terrible things.
Jared:Yep. Yep.
Michele Dodds:But I don't know that we can combat that. We can't control what people do out in society. There was a federal regulation that required immediate release of labs. And it's it's got to do with some of the transparency things in health care about also the reason now that you have to have more itemized billing for patients to see. Those are positive things, but I can tell you I've had some really negative experiences when things get released, like biopsy results that show cancer before I've ever had a chance to see it because I'm in a room seeing a patient.
Jared:That happened to my mother-in-law.
Michele Dodds:And it's not that you don't care as the provider. It is that you truly don't have the opportunity to look at it. You're in seeing patients, and now someone's gotten this terrible news that that is not how I want my patient to find something like that out.
Jared:Right. Yeah. But I'm always thankful that my, family doctor then responds later is like, you're fine. You know? These test results are fine.
Jared:Just keep eating well. Let's just
Rob:go back real quick. One of the terms you used was EMR record. Mhmm. Would you be able to explain what that is to our listeners? Because
Michele Dodds:Oh, just the electronic medical record.
Rob:Okay.
Michele Dodds:Where we're not so as an old NP, we're no longer on paper charts. That was a real struggle for me to transition from paper to electronic. And now we're doing AI. Like, how times change. We have to learn to change to optimize what our patients need.
Jared:Right. Right.
Michele Dodds:You know, I will tell you though, like when hospitals have cyber attacks and they have to go to paper, I feel like generally, racially, people will tell you the older nurses who were once on paper are tolerating that better than the younger generation. And so I think we also have to understand how to not rely on AI when we have emergencies like that.
Jared:Okay. So you've also been leading, our faculty. Where have you, I want to kind of pivot a bit.
Michele Dodds:Okay.
Jared:You know, how has that experience been leading it? Have you noticed any hesitancies there within faculty? I know you've, we've talked previously that it seems to be like, yeah, bring it on.
Rob:Mhmm.
Jared:But have you noticed anything like that?
Michele Dodds:Interestingly, I was probably the most hesitant person of faculty.
Jared:Really?
Michele Dodds:Because you get concerned about using AI, and will it tempt our students to use it in an inappropriate way? And I think that Yes. Was my area of hesitancy. Right?
Rob:Yeah.
Michele Dodds:But some of that has to do with I I can just I can't speak for all of our faculty. I think everyone there's had a positive attitude, but that was one reason I wanted to serve as an AI fellow because I wanted to learn, and I wanted to learn at a high enough level that it helps me to be confident in how to use it and how to tell my students to use it.
Jared:That's good. So how have the students responded? Are the students using it or how are you, how are you having your faculty have the students use it?
Michele Dodds:And They what has that been are. So it really helped when we had, when you guys provided chat GPT tutor training. I think that's been real instrumental in our nursing program. And so then we took that training and then we did a continuing education event for all faculty. That way we get contact hours for our licensure renewal.
Jared:Oh,
Michele Dodds:yeah. And so we did a whole faculty training during a faculty meeting, and so people got CE for it, which helps our license renewal. And then we just sat and worked with a computer and we all worked on creating a chat GPT tutor. And the more we worked with that, I just think you keep learning different ways to affect its behavior positively. And, I've just met with individual faculty members as they've needed help.
Michele Dodds:I've had students give me feedback in class. So we once you know, you can't just stop with the formation of the tutor. You gotta keep repairing the tutor. Right?
Jared:Oh,
Michele Dodds:yeah. So I just did an evaluation tool. The first week we implemented it, we just did one week's worth of information and had students evaluate it and give me feedback. And each week, I added all of our course content each week, and we just kept trying to improve it, trying to improve it throughout the semester.
Jared:Interesting. So let me ask you this. I've had conversations with, another fellow and he has found that, that he's been attempting to build tutors for his faculty and he's run into barriers where the bots will get the steps right, but the actual computation or the analysis it'll get wrong. Like it'll say it'll nail the formulaic process, but the other side of it, when actually in this case would diagnose would be incorrect. Have you noticed that at all?
Michele Dodds:I haven't had that, but the class I'm using in is also leadership and management. It's not doing tons of, they're learning a lot more concepts and clinical questions I to mean, did budgeting in there, but I didn't see any I haven't seen any wrong answers.
Jared:Okay.
Michele Dodds:One of the things I did learn, which I think might be helpful for faculty across campus that are doing tutors, to know, from student feedback is there I always had to give it, like, a practice quiz for them, know, and do Socratic questioning. But one of the things I learned in that process is the students have a limit to how much AI they can use per day. And so I think it's really important we know as faculty to program in there. Like, if you wanted to give your student five questions based on the content they're studying, you had to add into the instructions. Please give the five questions before they reach their daily limit.
Jared:Right.
Michele Dodds:And that kind of repaired that situation, which I thought was helpful.
Jared:And we also mentioned like giving a heads up to the student when they're almost reaching their limit so they don't end up halfway through something and end up with no responses whatsoever. So, future cast for me, let's think ahead. What skills do you think that our future nursing students will need that maybe previous graduates didn't? Cause you and I have also talked about the work workforce and the expectations of these employers of new hires. What do you see that these future nursing students will need that wasn't the case before?
Michele Dodds:Well, they need to understand how, how clinical decision support systems work. Mhmm. And even employers out in the community have communicated that to us. Mhmm. And they need to not learn to ignore the clinical decision support system.
Michele Dodds:Some of the things in health care that I think are are crucial is, we have a national problem with something called alarm fatigue. So when you're in the hospital setting and you have all these IVs and alarms, things going off, I've been reading studies about how they're using AI to control that. It can really affect patient outcomes. You know, if you hear bells all day alarms all day, it's not that you're not listening, but your brain starts not listening to it. Right.
Michele Dodds:You're focusing on a task. Well, it can really cause bad incidents like, you know, high risk for death incidents.
Rob:Mhmm.
Michele Dodds:AI is being leveraged for that to start trying to reduce alarm fatigue. And even though that's a system's problem and AI can't fix that by itself, it can, however, assess what's a false alarm.
Rob:I have a family member who works in an ICU, and that family member has referenced this with, especially, others where they've had to report something
Michele Dodds:Mhmm.
Rob:Because folks didn't respond as they should have. But at the same time, that this is a real thing.
Michele Dodds:Yes.
Rob:Alarm fatigue. So having AI come in and help.
Michele Dodds:Yes. I can even foresee, and I don't know that it's there yet, and I can't tell you if it isn't, but, like, I've thought about that issue. And what if you hadn't, your AI tool is working alongside you and let's say you go hang your bag of IV antibiotics on someone and you've got the infusion rate set, you've calculated what that infusion rate is, and you tell it, give me an alarm when this is almost out so I can go back and check. So you can continue to take care of other patients.
Jared:Mhmm.
Michele Dodds:But then you don't miss that time frame to come back and check that IV.
Rob:Yeah. I I wonder because we already have this. We've been toying with some things here, in our office, putting AI models on a little machine, like a Raspberry Pi or, you know, something of that nature. And you can put, you know, you can put cameras on those. You could even make them really super portable, carry it around, and it's got a camera.
Rob:So it's seeing what you're doing, especially if you're like maybe you have it on a pair of glasses or something like that. I can foresee that happening at some point. Wearables Mhmm. Where, you know, whatever the protective gear that nurses wear Yeah. You know, they have that on all the time.
Rob:And so when they go to work on a patient, it's aware of where they're at, first and foremost, but it's also can look and see what's going on. It's recording. You know? And you could speak to it and say something like that and say, hey. Set a reminder.
Rob:Check back on this and x number of minutes.
Michele Dodds:In the hospital are busy, and they're seeing so many patients in so many rooms. You can't be in every room. But what if you had an AI tool that was helping you monitor that patient in between? You know, the ICU is one thing when you have like a one to one or one to two patient ratio where you're intensively with that patient, but you're out on a med search floor, and let's say you've got a one to six or one to eight ratio where you've got six to eight patients to care for, and you just don't have the manpower to do it. And we have a national nursing shortage all across the country, so people are working with less and less staff and just AI tools that can monitor like what you're saying could truly just help protect patients and help protect the nurse as well as, you know, she would know like, oh, two rooms down, this just alerted that I need to go check this, finish with this patient.
Michele Dodds:And then even for task management, I think like that in real time could improve patient outcomes.
Jared:Because you've got, I remember I've, my kids have been healthy for a while. Thank you, Jesus. But I remember when they were little, I had RSV and suffering that stuff. And the nurses had the little thing that would hang and they could press it and communicate with somebody. I imagine you could repurpose that in some way to include, something like a small Raspberry Pi, which by the way is a portable small computer for anyone who's curious.
Jared:But you can put that in there and have it alert when, hey, the IV bag and room, whatever is slowing down. We need to replace that or something like that.
Michele Dodds:So that's why our nurses that are graduating now, the tool's already there, and there's gonna be more tools coming out. Just by the way, AI research is just blowing up in health care.
Jared:Yeah.
Michele Dodds:They have to leave school having been exposed to it, having used it, and being comfortable with it so that they can apply it to whatever setting they happen to graduate and go to.
Rob:That makes me think about what are we doing in terms of fluency or being being AI ready for defining that for nursing students? Is that something that that you and your fellow nursing faculty are looking at and starting to lay down and define, or are you getting standards pushed at you from from outside your accrediting agencies, those kinds of things? What are you seeing there?
Michele Dodds:You know, the accrediting agencies just say things like how, like, basic competencies of a entry level nurse, You have to be able to navigate technology and use technology. I think we're doing a good job of, like, we're implementing AI in our courses, but are we implementing assignments that require students to use AI to the level that is needed, I think is the area that may be looking forward that we need to expand to. But I think as faculty, we have to be comfortable first with AI use before we then start making assignments like that. With that being said, however, I have used AI to help us create some simulations, in my leadership and management course, and, it had weaknesses and strengths. We just implemented it, the rough draft, but we're gonna do it here again in, like, eight weeks, and now I know what things I need to change.
Michele Dodds:But it was really beneficial just for efficiency of trying to get the simulation process started Mhmm. And for just grabbing some of the research I needed to put behind it. So I think that's good. I have seen I've used, like, Illicit for literature reviews. Are you familiar with that AI tool?
Michele Dodds:Yes. And it's good, but there's some weaknesses. You know? Like, I think there isn't anything. But, also, if you're in a hurry time wise, it it can help you get some of that initial grab of literature and things like that.
Michele Dodds:But I think those are things we're gonna have to teach our too. We want them to leverage research so they can give good patient care at the bedside. But the open evidence I mentioned earlier that's an AI tool that's used in health care is really good for that because it's using, like, JAMA, like the Journal of American Medical Association, some of those. And that's when you can put I can put, like, a patient case in it, and it'll start just spitting stuff out.
Rob:So Open Evidence for our listeners is an AI tool that assists clinicians Mhmm. In helping to diagnose patients.
Michele Dodds:Yeah. Or or I can be looking for it you can ask it anything. So I might say, I have a 54 year old patient with uncontrolled hypertension. We've treated with the following agents. Is there any new research out there about this specific type of hypertension that will help guide the treatment?
Michele Dodds:You know, there's treatment guidelines for all kinds of things like that, but there's specific, like, diagnosis patients can have where you might need, what's the latest thing out there for them. When you've tried everything and you've not gotten this person to goal Mhmm. That's when I use tools like that.
Rob:Oh, that's cool. It sounds really cool. It sounds scary to some degree too. I have to I have to admit because on the one hand, you have this tool and you said it, like, research is blowing up Mhmm. Which means there's probably gonna be a lot more in terms of clinical trials for all sorts of things Mhmm.
Rob:From cancer to Absolutely. All sorts of of different diseases or clinical presentations that you all see.
Michele Dodds:One of thing is you don't stop with the tool just giving you feedback. If it gives you feedback, I'm gonna go see what is that article. I don't wanna just depend on that tool as the answer.
Jared:Right.
Michele Dodds:If it's giving me an article and that article is a case report with two patients, that's not rely like, that's not the same level of reliability as I go there and I find some type of systematic review that this is gonna be a new recommendation. Like, you're gonna look at that evidence and analyze that evidence.
Rob:It just gets you there faster.
Michele Dodds:Yes. It just helps you find it faster.
Jared:It's that AI literacy that we're talking about where being able to critically evaluate the feedback it gives you.
Michele Dodds:Right. So I also think with nursing students, when I think about where we need to go as AI fellows in the school of nursing, this is just my opinion, but I think we need to once we're more comfortable using the technology, you know, it's our first year doing this. So, I think as we move into the next year that we need to look at how do we teach students to critically analyze that material as well.
Jared:It's good stuff. Love it. I think it's great.
Rob:I think We could just end with that.
Jared:Yeah. I think that's that's really good. I just wanna say thank you for your encouragement to me and to the leadership that you've provided to the School of Nursing. It's made an impact for sure. My daughter is a prospective student, as you're aware, of the School of Nursing, and she fell in love with the program when she came to visit.
Jared:So and that's and she mentioned you specifically as someone who was very encouraging to her.
Rob:So and,
Jared:yeah. So I don't
Rob:see anyone cry on microphone. It's
Michele Dodds:pretty sweet.
Jared:I'm just I'm just saying that it's yeah. So she's really looking forward to it, and I'm very thankful for the school of nursing, what they're could potentially be doing in the Piles family.
Rob:So that you can go couple of things. Honey. I got this thing.
Jared:Yeah. Exactly. Can you check out my wrist? Yeah. Well, thank you, doctor Dodds, for coming in and and talking with us.
Jared:We appreciate it.
Michele Dodds:Thank you.
Ryan:Thanks for listening to this episode of Transform Your Teaching. If you have any questions or comments about our needle mover episode with Doctor. Dodds, feel free to send us an email at ctlpodcast@cedarville.edu. You can also connect with us on LinkedIn and message us there. Finally, don't forget to check out our blog at cedarville.edu/focusblog.
Ryan:Thanks for listening.