Through ADEPT and MeasureDx, a program of case review and improvement has been created, aimed at enhancing diagnostic safety within Hospital Medicine. This presentation discusses the program’s initial intervention, the diagnostic cross-check, and how this is only the beginning of a journey to diagnostic excellence.
Um, welcome to session one A. Ignite to innovate fueling the fire of diagnostic excellence. Um, I'm Kim Krause. I'm with Department of Medicine Patient Safety and Quality with Washington, uh, University School of Medicine. Um, love my job, love patient safety, very passionate about it. Um, I also have worked very closely with the presenters on this project, so, um, happy to be seeing them here presenting today. Um, as a reminder, just submit your questions via the poll everywhere or if you wanna ask a question verbally, um, we recommend, uh, use of the microphones, uh, located, um, where are they? Oh, up there, OK, on the stands at the top. Thank you. Sorry, I can't see with my glasses and I can't see without my glasses, so it's a challenging situation, um, so you can access the poll via the QR code. Um, I'm gonna introduce both of the, um, presenters here. So the first presenter we're gonna have is Doctor Rada Devi. They're actually gonna take turns. So, um, Doctor Debbie is an associate professor of medicine and hospital medicine at Washington University, um, School of Medicine. She serves as a patient safety and quality committee. She co-leads the Center for Diagnostic Excellence and is the director. Of the recruitment committee within hospital medicine, she also serves as a medical director for the North Campus medicine, um, floors of the 7900 and 8900. I know she's got some peeps here today, um, cares for patients on the direct hospitalist service, firm medicine service, um, the, uh, BMT service, and is involved with the teaching residents and medical students. Um, Rada's areas of interest include patient safety, quality improvement, interprofessional teamwork, diagnostic excellence, and diversity initiatives. One, interesting fact about Rada is she loves sustainable native gardening. She has her silver certification. She's working towards platinum, so, um, definitely very passionate about that. Uh, Doctor Tim Evans is an assistant professor in medicine and pediatrics within the respective divisions of hospital medicine. Uh, he attends as a hospitalist for oncology and teaching services at Barnes Shoes Hospital, as a hospitalist and a nocturnist at Saint Louis Children's Hospital, so he does both adults and pediatrics, um, with Children's Hospital on the teaching and non-teaching teams. He provides direct care to the pediatric patients in the emergency department, um, of several community hospitals as well. He co-leads the Center for diagnostic excellence within hospital medicine. Um, his interests include patient safety, quality improvement, improving the diagnostic safety and inpatient care of patients with diabetes as they transition back home. I've actually gotten to work with him on that project too. Um, he also has interest in, uh, graduate and undergraduate medical education, mentorship, and curriculum development and patient safety and quality improvement, and Doctor Evans' spare time, he likes to spend time with his kids. Um, he likes to hike and he's very passionate about bird watching, was just in Panama, um, and added 99 birds to his life list. So with that, I will let, uh, Doctor, uh, Debbie and Doctor Evans take over. Thank you, Kim. Can you hear me OK? I'm Rhoda and um as you heard I'm Tim and today we're gonna talk to you about diagnostic excellence and what that means for us today is um studying and reducing diagnostic errors and improving what we are calling diagnostic safety. So, um, the standard disclosure, neither Rada nor myself have any financial disclosures. The ADET study which we are a part of, um, is funded by this AHRQ grant listed there. So today we're going to define and identify examples of diagnostic errors. We're gonna hopefully teach you about recognizing the importance of avoiding diagnostic errors and teach you why uh the field of diagnostic excellence hasn't come as far as we would have liked it to up to this point in time. We'll talk about some recent efforts to study diagnostic errors and an intervention that we have underway at Barnes currently. And then we'll end with identifying opportunities to improve diagnostic excellence further. To start, what are diagnostic errors? So you are all probably familiar with this Venn diagram typically um represented by this um usually this represents medical errors but in our case this is gonna be diagnostic errors and adverse outcomes um CNA will represent kind of preventable harm from diagnostic errors. And diagnostic process failures is represented by this big circle in black that is probably new to you all. So a diagnostic error is a failure to establish a timely diagnosis. An accurate diagnosis. Any diagnosis or to communicate with a patient regarding their diagnosis, OK? And as I said, an adverse event associated with diagnosis is called a diagnostic adverse event and is represented by ANC here. A diagnostic process failure. Is something that happens in the process of getting to a diagnosis. Um, what I will say is that you can have misdiagnoses without diagnostic process failures, and you can have diagnostic process failures that lead to misdiagnoses. As an example, um, I will give a very straightforward example of an MI as a misdiagnosis. Let's say for example, a patient presented um in cardiac arrest due to an MI. They had presented to the health care system and so prior to arriving in cardiac arrest they were seen in the ER and had subtle symptoms of abdominal pain and some shortness of breath. Unfortunately they were treated for GERD and discharge. And in this case they did not undergo an EKG or a troponin and so the diagnostic process failure in this example is um not performing an EKG and a troponin that may have caught an ACS or an acute cardiac syndrome event that ultimately led to a heart attack or an MI. And what I'd like to emphasize is that similar to other types of errors. Diagnostic errors have a large contribution from systemic and process failures but also cognitive failures, and there's a larger contribution of cognitive failures in diagnostic errors because cognition is so intertwined in the process. So why do we care about diagnostic errors? If you all would take out your phones, scan this and take our poll, we'd like you to tell us why you went into the field of healthcare, and we'll come back to this, um, in a few slides. And if you go ahead and pull up the. There's, it's gonna have a word bubble here in a second. We'll give this about maybe 20 to 30 more seconds. And the QR code's still there on the side. All right, I'm seeing lots of helping people, helping others. Stop suffering, making a difference. Service for humanity, patient wellness. Front and center is helping people. Which is what I was expecting. OK, I'll give you a few more seconds. All right, thank you all for. Taking our survey. So the first reason why we should care about diagnostic excellence and reducing diagnostic errors is because we're all human and we hate to see others in pain and we hate to see harm come to them. So these are two cases that were published in The New York Times. The first was a case in 202012 and the second was a case in 2023. The first was a 12 year old boy who presented to the ER. He came in with abdominal pain and vomiting. It was noted that he had had a cut to his arm a couple of days prior and he was also complaining of leg pain. He was discharged from the ED with supportive care and told that he had a viral gastroenteritis. Unfortunately he passed away several days later from strep pyogenes sepsis. Strep pyogenes are a common skin organism that can cause cellulitis, but in this case it caused sepsis and led to his demise. This case was in 2023 during the COVID pandemic. So This was a college student. He was 20 years old at the time. He presented to the ED with chills and a headache similar to our other case, he was diagnosed with. Um, a viral syndrome. He had a 4 in 1 swab which was negative and didn't reveal any viral illness at the time. He came back the next day, which is always a red flag for anyone that works in the ER. He had similar symptoms, but this time he had shortness of breath. In addition, he had pretty significant vital sign abnormalities because he triggered a sepsis alert, which was ultimately overridden. Again, he was given fluids and treated with supportive measures, discharged home with what was called a viral illness. He went back to his dorm room and based on text messages with his girlfriend it was felt he became delirious and unfortunately passed away from what was ultimately found to be multi-system inflammatory syndrome from COVID that he had had several weeks prior. Not only do we have cases across the United States but we have cases locally too. So we've changed some of the details of this case to protect patient privacy. Our patient is a 45 year old male who came to the hospital with complaints of back pain. He has a history of diabetes, hypertension, and hyperlipidemia. He had no history of trauma, there were no red flag signs when he presented. His vitals were stable, he was afebrile. He was alert and oriented. On exam, he had a normal lung, heart exam, and abdomen exam. On neuro exam, he had no focal deficits, but he had some palpable tenderness in his lumbar area. He was admitted to the hospital. And started on medications for pain management. Next day, he spiked a fever. An infectious workup was done that included a chest X-ray, a urine analysis, and a viral swab. All this was negative. On day 3, he was seen by physical therapy, who noted left lower extremity weakness. They brought this to the attention of the primary team. who then confirmed the exam findings, and an MRI spine was ordered. This revealed a spinal epidural abscess. He was started on IV antibiotics, underwent drainage and made a complete recovery. Perfect. Thank you Rada. One more poll this is the last one I promise if you wouldn't mind filling out this um what we are asking here is based on our definition that we have given to you has anyone. Ever experienced a diagnostic error or a misdiagnosis as a patient, as a family member, a friend, or a healthcare provider? Wow. We'll give this a few more minutes, a few more seconds rather. OK, we can go ahead and note about 92 to 93% you guys are definitely above average, so great work, uh. We'll come back to that in a second as well. So the next reason why we should all care about diagnostic safety and diagnostic errors is because of our dedication to the IHI's quadruple aims for those who aren't familiar, which I imagine is no one in this room this is improved healthcare provider experience this is better outcomes for patients. This is lower costs and improve patient experience. So Um, I will point out this graph from the IHI here. Um, of those who encountered the healthcare system, it is estimated that about 41% experienced a medical error. Um, now, not specifically a diagnostic error, but apparently, um, because we're all in healthcare, we have all encountered a diagnostic error, but you can see from this graph here that misdiagnosis represents about 59% of those who have experienced a medical error, so. By improving diagnostic safety and diagnostic errors, we can improve the patient experience. Further, a precursor study to adapt, which we will talk about was a study called Upside, and in this particular study we looked at trigger events called, um, trigger events that were ISCU transfers, deaths, or both, and in these trigger events we found that about 23% of these patients experienced a diagnostic error. And in those cases of ICU transfers and deaths, approximately 18% of them. The the diagnostic error contributed to harm. And in the patients that died, approximately 6-7% of them had a diagnostic error that helped contribute to their death and so you can see that by improving. Diagnostic errors and reducing the harm associated with them, we can improve patient outcomes. Further, we should care about diagnostic safety um for reasons of. Cost and reducing the cost of health care overall. So from this graph here you can see that of those malpractice cases that result in adverse legal judgments, it's estimated that 40% of them represent misdiagnosis. From some preliminary numbers from Upside and Adept um with some other studies, it's estimated that about 8% of medical inpatients experience a diagnostic error each year and this number. Represents um an average adept hospital which is about 2,028,000 inpatient admissions. Our hospital we understand has about 54,000 inpatient admissions a year and so for our hospital this would be 4300 inpatient medical diagnostic errors. Further, it makes sense that adept events which include rapid responses, ICU transfers, and deaths via some number crunching are associated with increased costs and length of stay which ultimately increases costs because of decreased bed availability and turnover. And finally, the healthcare worker experience. From the poll that we took the vast majority of you. Told us that you're interested in health care and you went into the field because you want to help others. So, it's our commitment to patient safety and high-value care. That will, um, by reducing diagnostic errors will improve the healthcare worker experience. Further medical errors, specifically diagnostic errors are associated with feelings of guilt, shame, and anxiety, and by reducing these errors we hope to improve those feelings as well. So From the last few slides, I hope it's evident that It is important to study diagnostic errors and reduce. Um, reduce diagnostic errors. But what have been the barriers to study them? The first papers kind of commenting on diagnostic excellence and diagnostic safety were published in the early 2000s, so why have we not made more progress since then? There's a lot on the slide, um, but the big takeaway from this is that the diagnostic process is very complex. So Diagnosis is both a process and an outcome as you can see in that blue box and because of this it's hard to see when the process ends and when the outcome begins. Further, there's a fine interplay between cognitive and system related factors as I'll show you on the next slide. In the process of our thoughts and our diagnosis, our diagnostic process. Is often unconscious not only to the person who's trying to make the diagnosis but it's not shown in the in the EMR to those who are trying to determine if a diagnostic error occurred so I'm gonna skip to the next slide and then come back to this slide. This is a figure that represents a diagnostic timeline, OK? So on the left here is when a patient experiences a health problem and on the right is when we have a diagnosis and we're treating the patient. This is The diagnostic process and you can see that there's a fine interplay between. Objective and subjective data from the patient. And our own thought processes so we have to take a clinical history and perform a physical exam and come up with a differential diagnosis we perform diagnostic tests and we get referrals and we take that information and. Modify our differential diagnosis and we come up with with what we think is a working diagnosis and treat the patient and we gather more information like how the patient is responding to treatment and all of this is complex and iterative and cyclical and because of that. It makes the study of diagnostic errors very difficult. As alluded to. In the yellow box, the diagnostic process is invisible to those while it's occurring at times and also there are documentation gaps and so when we, as Rata will talk about in a few slides, are doing case reviews to determine if diagnostic errors are present, um, we have to rely on, um, assuming things were known and, um, this makes it very difficult to. Discern the thought process at the time that it was being carried out further we have no we do have a gold standard in diagnostic errors and those are autopsies we don't perform autopsies very often and so we really have no gold standard. And finally, the diagnostic determination process when we come back and do these case reviews, there's a lot of hindsight bias and we assume more was known at the time than was actually known. And finally, organization and cultural barriers we are fortunate that uh most of these things do not exist at barns, but other systems do encounter these barriers. Reporting culture we are fortunate that at Barnes we have what I would say is a just culture that is mostly non-punitive and we take errors and adverse events and study them in order to improve our systems. We Previously had a lack of feedback for our providers we had no system of reporting diagnostic errors and because we can't report diagnostic errors we don't know when they're occurring and we can't study them. Prioritization is an issue as well. The money is in things that are not diagnostic safety at this time. Hospitals care about the the. The processes for which they get paid and diagnostic safety at this point in time is not one of them. And grant funding is often for specific disease processes, um, not nebulous concepts like diagnostic safety. And then the last point I want to make is that the diagnostic process is often occurring amongst different hospital systems, so SSM, SLU, Mercy, Saint Luke's, our own system, and what was known when is not always clear and so it's hard to go back in hindsight and figure out what was known and if a diagnostic error occurred. So I'm gonna hand it over to Rada to talk about our study. So what is AdepT? Achieving diagnostic Excellence Through Prevention and Teamwork. So, 14 hospitals from around the US got together to lead the charge to reduce the burden of diagnostic errors. As Tim previously mentioned, Started with Upside, which was a study done to To make out the, the risk of diagnostic errors. It was a utility of predictive systems in diagnostic errors. So this study was done across 29 hospitals. They reviewed cases of around 2500 patients who either had a rapid response, death, or an ICU transfer, and the mean diagnostic error rate was around 23%. From that stemmed ADET, which is achieving diagnostic excellence through prevention and teamwork. What were the aims of this study? We wanted to implement an enhanced case review structure that can accurately identify diagnostic errors and processes at participating hospitals. We wanted to develop site and group-level benchmarking reports of diagnostic error rates, processes, and performance, and incorporated into our site's safety and quality structure. We wanted to introduce some interventions, pilot safety 1 and safety 2 interventions. And then based on all this, we wanted to go back and analyze these rates of diagnostic errors before and after implementation. So what does this involve? It involves case review. And what principles are we looking at in case reviews? We want to create a shared mental model of diagnostic error. We want to build something that's reliable. Ensure what's an ideal systematic approach to review these cases? Embed principles of equity. And also integrate some assessment of what goes right. So patients arrive at the hospital through the ED through inter-hospital transfers, or through direct admissions. Cases are selected if patients experience a rapid response, ICU escalation or death. We start with the primary diagnosis. What is the condition that led to the hospitalization? We focus on their initial encounter with the primary team on the day of admission. We also look at the decompensation diagnosis. This may or may not be the same as the primary diagnosis. And then we focus, when did these signs and symptoms first manifest? How quickly were they recognized and diagnosed? Exclusion criteria. Patients less than 18 years, those who were transferred to the ICU for reasons other than clinical changes, ie hospital policy, patients who were admitted for end of life or comfort care. RRC for another reason. So we excluded stroke codes from our study. Patients who were transferred within the 1st 48 hours of admission. So this time period was thought to be too narrow to determine whether a diagnostic error had been made. And patients that were identified in error since our study was on the medicine floors. We did leave it up to the sites to determine if there was another reason for which the case should not be included, and this was discussed with the larger collaborative. So what is the case review process look like? So we identify patients. Through RRT ICU escalation, and death. Cases are selected randomly 10 every month from Red Cap. Cases undergo review using safer DX principles by a primary reviewer and a secondary reviewer independently. All our site reviewers have been trained by the larger collaborative. If there is a disagreement, There is a 3rd reviewer who adjudicates the case. And then finalizing the case review and submitting it to Red Cap. We also present cases at our quarterly patient safety meetings. We conduct case reviews both retrospectively and concurrently. So retrospectively, we look at cases in the past year. Concurrently, we look at cases happening in real time. We send a survey to their current primary team to incorporate what's going well. Is there something you want to tell the, tell your colleagues about what went well with the case? So what is Safer DX? It's a socio-technical system in which diagnosis exists. So it has various domains. Did the history alert you to a differential diagnosis? Was there something else in testing? Was there, uh, something that affected your clinical reasoning? Were there any alarm symptoms that were missed? So going through these different domains, we can rate. From strongly agree to strongly disagree if a diagnostic error has been made. If we are certain, we come to dear taxonomy. I did not know that moose was a type of deer. It's actually the largest type of deer. So what's dear taxonomy? This framework tells us where exactly in the diagnostic process did the error occur. Was it an access issue? Was it the way history was taken, physical exam, testing, formulation of your working hypotheses, consults, or follow-up. So, if you think back at our patient that we presented, Was it an access issue? Probably not. Was it a history issue? Did they go back and ask the patient on day 3 about any new weaknesses? Did the pain worsen after the fever? Could be. Was it a physical exam issue? The physical therapist noted left lower extremity weakness. Did the primary team. And then that may have led to a downstream failure of delay in ordering a test. So if we ascertain that an error has been made, then we can assign a degree of harm. This is adapted from the Medication error Prevention by NCC and the blue area shows. Circumstances existing. That have the capacity to cause an error. The orange area shows that an error occurred, but no harm reached the patient. The yellow resulted in some degree of harm. And the red resulted in death. So in summary, We identify cases based on inclusion exclusion criteria. We review them based on safer DX. We determine if a diagnostic error was made. If yes, we can ascertain a degree of harm. And then we also incorporate what went well in the diagnostic process. That leads to intervention and next steps. We also collect some demographic data for further analysis down the line. So this brings us to, how can we improve? That led to diagnostic cross-check. So various interventions were proposed in an effort to reduce diagnostic errors. Can we give physicians feedback about diagnostic errors? Can we use EHR best practices? Can we embedded into our current quality and safety structure? Can we engage patients? How about second opinion or diagnostic timeout? What came out of this was the diagnostic cross-check. So what was the background for diagnostic cross-check? There is evidence from prior studies that more than one clinician is more likely to arrive at an accurate diagnosis than a solo clinician. Unstructured peer consultation is a common practice, both in self-reported surveys and some observational studies. Peer consult is valued, but there's a lot of structural barriers to its use. And a study done in the ED called the ED CHARM trial showed that cross-checking reduced medical errors and misdiagnosis in the emergency department. So how do we perform diagnostic cross-check? When an RRT is called on a patient, we reach out to the primary team to perform a verbal, structured, peer to peer diagnostic huddle. We reach out via Epic secure chat. The goal is to keep it less than 10 minutes. We try to avoid Monday morning quarterback and hindsight bias. Nothing is recorded in Epic. And at the end of this, we send a survey to clinicians and to cross-checkers. We are doing this on the hospitalists direct care service. which includes our nurse practitioners. Our students, residents and farm teams. And our diagnostic cross-checkers are members who have undergone training in the larger collaborative through simulated cases. Since this is structured, we also have a framework. So what is our cross-check framework? It's a series of questions that we use to guide this huddle. What was your main working diagnosis for the rapid response? Was there something about the patient's case that didn't really fit? Were you considering any other diagnoses? And based on all the clinical things that you have observed, would you go back, do a chart review, or elicit any additional history? How about any additional physical exam? Any extra tests you would order, consults or referrals you think would be helpful. Also, what do you think went well? What lessons do you want to share with others? And based on this, what are your next steps for this patient? So the diagnostic cross-check is followed by a primary clinician and cross-checker survey. If found helpful, they can spread the word. Preliminary survey results have shown that both primary clinicians and cross-checker surveys are highly favored among our clinician teams, and they have led to change in diagnoses for certain patients. We hope that this will help foster a culture of collaboration, diagnostic reflection, and patient safety. Back to you, Tim. Um Thank you, Rana. So, From Rada's slides you can see that we have established um a way to get feedback about our diagnostic error rate. And We have started an intervention which we hope will improve diagnostic safety. But We envision that there will need to be a lot more. To move the needle and diagnostic safety. So as she. Told us We've Been a part of the ADEPT trial and from the ADEPT trial we have created what we are calling the Center for Diagnostic Excellence. This is the group of 14 of our hospitalists that we are now calling a subcommittee of our quality improvement and patient safety Committee. That participates in our case reviews and our diagnostic cross checks. We have embedded and uh provided an option for diagnostic errors to be listed and reported in SEMS and ERS and for those of you not in our system, those are our event reporting systems which will allow us to get additional feedback about diagnostic issues that are arising at Barnes and provide feedback for our providers. Further, we have instituted a regular review of diagnostic error cases at our quality improvement and patient safety meeting. When I made this slide. I was a little disappointed because we've put in a lot of time and a lot of effort and. This is a summary of all the things we've done. But what I think is not emphasized strongly enough on this slide is that the amount of work and person power that's gone into this is tremendous. We have these 14 individuals who, for the love of diagnostic safety dedicate their time to case reviews and participate in diagnostic cross checks. And so while initially disappointed, I became proud that we have accomplished a lot up to this point, but there's a lot more that we need to do. So I'm gonna get a cheesy for a moment. Uh, I would like to analogize and, uh, represent this group of individuals as this fire right here. The fire represents the size of our group of excited individuals, but it also represents the excitement and the fervor that they possess. But we need to improve the system and we need to improve our healthcare teams to actually move the needle on diagnostic excellence. So This is there's a lot on the slide so don't don't please don't get overwhelmed but um this is in summary, um, a lot of things that can be done to improve the system on diagnostic safety from an extensive literature review and fortunately we actually have a lot of these things in place but we have room for improvement. So things that we have and things that we want to improve accountability from the top is a must, and we have been fortunate that leaders within our division and leaders within our department and even leaders within the hospital have um started to pay attention to diagnostic safety. And we hope that the fact that it was highlighted in this symposium means that we are moving in the right direction. And we need to own diagnostic safety not only from the top but from the bottom and we this 14 individual hospitalists own diagnostic safety within our division. And we need adjust culture which I alluded to earlier we do possess. We need to improve communication. Communication is key between health care providers, one primary team and the team that transferred the patient. The primary team and the consulting team. But more importantly, because I know not everyone here is a front line provider, everyone has a role to play in communication bedside nurse, case management, social work, PT, OT, RT literally anyone that comes into contact with a patient has a role in communicating with us things that are amiss with the patient. Further, the infrastructure of both Epic, which we both love and hate, and our system can be improved, so utilizing clinical decision support tools, making sure our system can talk to other systems, and making sure every physician every provider has access to the health care system and in addition that patients have access to their notes and what is going on with themselves. The system infrastructure I think we've made some progress. We have feedback on our diagnostic accuracy via our case review process and our reporting system. We have implemented the second review process with the diagnostic crosscheck. And we are able to survey these local issues and talk about them via case reviews. And at our at our local patient safety meeting. But we also need to improve individuals that are part of the healthcare team. Some of these concepts are specific to those who make the final diagnosis, but I want to emphasize that there are things within each of these boxes that every one of you in this room can actually achieve and improve upon. So simply slowing down, considering alternatives and understanding that we have biases and um we can de-bias ourselves is very important. We need to follow up on our patients. We need to ensure that we're all participating in M&M's and case reviews to work on local issues. We can all use cognitive aids that help alleviate some of the mental load that we carry every day. I mentioned communication so I won't harp on it further but embedding technology in our systems and enhancing our skills, making sure that we are all practicing at the top of our license and staying up to date on the most recent medical information. So through the diagnostic cross check as Rada alluded we want to make an existing practice of collaboration we wanna lower the barrier to. Letting others collaborate and communicating for those who experience rapid responses and we hope that this will allow us to be the most collaborative diagnosticians in the country and allow us to reflect on our patients more. Explicitly. But by doing and implementing some of these systems and health care team related interventions we hope to become one of the leading academic centers for the education of the awareness of and the improvement of diagnostic safety. So getting cheesy again. I wanna take and we wanna take this fire which represents where we are now and get people excited about diagnostic safety and diagnostic excellence and join us in our efforts. To become this Thank you, Tim. So in conclusion, The study and improvement of diagnostic errors is important, and gaining momentum. Achieving diagnostic excellence is challenging and requires a multi-modal approach. It takes teamwork. Everyone has a role to play in diagnostic errors, and it takes teamwork to provide the healthcare we want, need, and deserve. Adept and cross-check are our first steps. There's much more to come. We hope to ignite a spark about diagnostic excellence and welcome those who are interested in joining our efforts. We do want to give acknowledgement to our team here. And to all our affiliated authors and investigators. Which is a lot. We have collaborators at 13 other institutions, the Institute for Healthcare Quality, Safety and Efficiency at University of Colorado, and the Home Run Network. I'll take any questions or comments. Thank you. And for those who don't know, this is actually the two-headed monster, because 2 heads are better than one. Do you want to start? So we have a question. Have you identified any specific barriers to accurate diagnosis? Done. So, we have identified a couple of barriers. One is just recognizing that it was a diagnostic error. A lot of this happens in hindsight. So it's hard to go back in time and see what the primary team was thinking at that moment. Do they have all the pieces of data available to make the accurate diagnosis? I think that's been one of the biggest issues. And sometimes as a team, we tend to give credit to our own team. Oh, we would have done the same. But if we go about it in a structured manner, it has helped us to kind of identify, yes, was it really missed or was the data available. Um, the next question I think is, are the multidisciplinary, I think it's supposed to be huddles in person or vir virtual, and did you notice any changes or benefits from either? Um, so I, I will comment that I think um. Any type of Multidisciplinary communication and collaboration is helpful and, and that's actually some of where we wanna put our efforts is um getting better communication between those who are spending more time with the patients and the, the families with the healthcare providers because the frontline diagnostics diagnosticians are not always at bedside unfortunately as we um spend 3 to 4 hours a day charting so we would like um those huddles and the communication. Um, to actually kind of. Enhance our diagnostic skills and so whether it's virtual or in person I don't think really matters as long as the communication occurs. What was your strategy for recruiting hospitalists? Was word of mouth. But I think we are blessed to have a team who have a real passion to reduce these rates of diagnostic errors. As we saw from the poll, most of us went into healthcare to help people, and I think from that. Ignites that spark to make us better. Um, the next question is how do we move from a reactive approach to a diagnostic error prevention before the RRG. This is a, a wonderful question because while we love the diagnostic cross check, we do feel like we need to move further upstream and prevent things before they occur. So, um, we've been scratching our heads about that, but kind of thinking about, um, you know, reaching out to leaders on units about patients that are of concern, uh, before they decompensate using some, um, surveillance tools embedded in Epic would be an idea, um. I think the other big piece of this is working on faculty development which is another piece of the puzzle so um the kind of health care provider, uh, piece of the puzzle I think needs to be emphasized and so we have some things in the work regarding faculty development and, um, ensuring that we are all aware of these biases and up to date I think will be very important as well. So since things are moving, I will pick up this one. How can bedside nurses or nursing leaders help make an impact? So this is something I always talk with, and this is something very near and dear to me as well. I was going through the slides at the beginning that, um, There was a leadership award given to a nurse on the floor who noticed that before the PET scan, the patient was on dextrose-containing fluids. We always look to our nurses to let us know you're working within your scope of work, you are frontline, so if you think something is wrong, speak up. We want to hear your voices. Um, let's see, so. We're heading towards AI diagnostic support, aren't we, uh, so I, I will say that some of the sites, um, that we are collaborating with have, um, tried to embed this in the cross check and have used tools to help them not obviously as a final say, but more like. Well let's plug in the case um at the time of the rapid response and see if we can have some alternate diagnosis that we hadn't considered yet so yes I I do think that there it is on the horizon and I think it will play a huge role in our efforts. How do you provide patient and family feedback regarding these diagnostic errors? A lot of the retrospective cases, um, happened in the past, but for cases that are concurrent, uh, and we've had to tell the family, it comes from a point of we want to grow, errors occur, we all accept it. And how can we go forward from this moment? How can we make what is wrong, what is right? So that's where, you know, it comes from. So we did have an opportunity recently to talk to a family, and we kind of understood the emotions they were coming from, but how can we together move forward? Uh, with so many categories for scoring degree of harm, do you notice a lot of middle of the road scoring, or is the degree of harm accurately represented? Yeah, this is a great question, and I think some of this is, uh, uh, it's muddied by the fact that. These aren't. I gave an example of a very clear diagnostic error. um, I will say in, in reality these diagnostic errors are not so clear cut and so to actually ascribe the level of harm is, is a little more difficult than, um, it, it would be in that case that I presented so um I think that it's accurate for for. Our current system, but um, yeah, I could use some work. Do you see more error in diagnosing a certain area of the hospital, ER versus labor and delivery. Unfortunately, our study is only on the medicine service. There are some prior studies looking at diagnostic errors in the ED because that's a diagnosis and evolution. Nothing has been diagnosed just yet. Any considerations on developing a more formal debrief after each act rapid response involving all the staff similar to what is done after Code Blues, yeah, so, um, in an ideal setting we would actually be doing these, um, diagnostic cross checks real time. Some of the sites have their hospitalists engaged in rapid responses and so that would be something that we would like to strive for, um, because obviously, um, there's a lag time between when we are able to contact. And if we could get to them at the time of the rapid response, I think we'd have a more, um, robust effect on them. What is the threshold for placing in STEMs? I think you should report anything and everything. There are very hardworking nurses at the back end who, who would like go through all these numerous reports. There's always system changes, as a very innocuous thing that might seem to you. You should go ahead and report it. Uh, is there a distinction between diagnostic and treatment errors? Do we know how many bad outcomes are related to treatment errors rather than diagnostic ones? Uh, yeah, great question. Um, we, we're specifically looking at diagnostic errors. Um, however, you know, if you have the wrong diagnosis, presumably you have the wrong treatment, and so, um, those are often intertwined, but we are not specifically looking for treatment errors in this study. Do you think AI can play a role in identifying diagnostic errors before harm reaches the patient? Actually, a few of our sites are looking into how, how large language models can help with diagnosis. So that may be coming down the pipeline. Yeah, and I'll comment on that because I think that is a very exciting feature. Um, I envision something like. AI or a large language model being being being able to sift through the chart and seeing diagnoses across all systems perhaps an example would be a PE that was diagnosed at an outside hospital that we didn't see or a lower extremity clot, let's say for example and a patient present. To our hospital with tachycardia and elevated heart rate. Now we did not put 2 and 2 together, but the system is able to see all the data and say, oh, this patient's at high risk for a misdiagnosis because of these data. So I, I can definitely envision something along those lines. Do any of the participating institutions use an early warning score for initiating the rapid response? If so, does the data from the early warning tool help inform the cross-check? So, we are not using an early warning score, but I know some sites are using it to identify their rapid responses. Um, it's too early to say whether it's actually making a difference. So we'll know once the analysis is complete. With error reviews, what type of cognitive feedback have you received when discussing the case, uh. You. Our reviews are mostly focused on being a diagnostic huddle, so we're not here to assign blame. It, you know, it is mostly what we found is self-recognition. Once we talk through the steps in a structured manner, clinicians or the team recognize maybe we should have done this, but we don't, we're not here to assign blame on other people. And so it's been more of knock on someone's door and say, hey, I have this case. Do you want to talk it over with me? That's the culture we're trying to promote. Do you find higher rates of diagnostic errors and other opportunities in ED hold patients? Are there difficulties with assigning responsibility of those patients? Yeah, so, um, I'm, I'm not involved in the nitty gritty of our, our hospitals ED operations, but yes, um, we, we have not looked specifically, um, as to whether there are more diagnostic errors in that population, but I do think that that would be an interesting thing to look at because we know that those are a particularly vulnerable population given the multiple transitions and. Um, inability to care for them in the same manner that they would get on the floor. Uh, this is not a question. Yes, we are headed toward AI diagnostic support. I'm glad we agree. Um, with so many categories for scoring degree of harm, oh wait, we already did that one, OK. And then what role does the inter interdisciplinary team play in assisting accurate diagnosis? Well, I'm glad that this is the last question because I love this topic, um, and I think Rada does as well. I can't emphasize how much I think interdisciplinary communication will play a role, and it's something that, uh, I think has a lot of merit and should be looked into further. So, um. Anyone who interacts with the patient, as I mentioned in the presentation. Has the ability to help frontline providers so if a case manager goes to see a patient and something seems amiss like they're more confused than they were the last time I would love to know that information bedside nurses, they spend the most time with the patient and having that direct line of communication and ensuring that they're not afraid to speak up when something happens is so, so, so important and is so helpful to knowing when something. Or when a problem arises with the patient, so yes, it definitely has a role. And I think That is all the questions. This one we did, OK. Perfect. Anything else? All right, well, thank you all. Thank you.
