Dr. Meghan Lane-Fall - Improving Handoffs for Better Perioperative Care

to our Department of Aniscusia Grand Rounds. Hopefully everyone can hear me okay. It is my great pleasure to introduce our guest lecture this morning, Dr. Megan Lane-Fall, who will be speaking to us today on the topic of perioperative handoff. Dr. Lane-Fall is an anesthesiologist and an intensivist at the Hospital of the University of Pennsylvania. She is Associate Professor of Aniscusiology and Critical Care and Associate Professor of Epidemiology at the University of Pennsylvania at the University of Pennsylvania, after which she obtained her master's in health policy research at the University of Pennsylvania. At her hospital, she is the co-medical director of both the trauma surgical ICU and the COVID-19 perioperative ICU. She is a doctor of the University of Pennsylvania and a doctor of the University of Pennsylvania. She is a doctor of the University of Pennsylvania and a doctor of the University of Pennsylvania. She is also the co-director of the Penn Center of Perioperative Outcomes Research and Transformation, Assistant Director and Director of Research for the Penn Center for Healthcare Improvement and Patient Safety, and Director of Research and Scholarship at the Penn Center for Healthcare Improvement and Patient Safety. She is also a Director of Acute Care Implementation Research. In addition to all of these directorship positions, she holds a number of other leadership roles in her institution, as well as nationally and internationally. Many of them tied to quality and patient safety. She is on the board of directors of the Anathesia Patient Safety Foundation, and she's done a substantial amount of research on teamwork, communication, and implementation science in acute care, with a particular focus on perioperative handoffs. So thank you again, so much, Dr. Lanevall, for taking the time to speak to us today, and I will turn the virtual podium over to you. Thanks so much, Laura, and thank you for the kind introduction. Can you confirm that you can hear me? Yes, we can hear you well. Perfect. Welcome morning, everybody, and thank you for inviting me to your grandrounds. I'm delighted to speak with you about perioperative handoffs, which I think is a timely topic, and certainly one that's near and dear to my heart. So I don't have any conflicts of interest to disclose. I'm funded right now by the NIH, RWA Foundation, and then a small grant from Penn Medicine. And then Dr. Rihall already told you about the relationships there. I have a few objectives for the time that we have together. I want to describe a few different kinds of handoffs that we have in anesthesia and perioperative care, because we work in lots of different places in the hospital, and so our handoffs differ as a function of where we work. I want to show you that perioperative handoffs are associated with harm, but they don't have to be. I want to show you how to incorporate some best practices for safe communication into your daily care, and talk about some strategies to standardize perioperative handoffs. First, I want to introduce you to my lab, which is a social science lab. It's a dry lab, so there's no fume hoods or micro-pipedders or anything like that. Our goal is to advance patient-centered, high-quality acute care. So we take direction from the Institute of Medicine, which is now the National Academy of Medicine, and there's six quality aims, and our focus is really on safety and patient-centeredness. This is my north star. So these are the leading causes of death in the United States. Healthcare is not on the CDC's list, but if you put it on there, it would be number three, and there are some quibbles about exactly what this number is, but certainly the order of magnitude would place it at number three, and I think that we shouldn't be harming people as they come to us for help. So this is what guides me as I do my work. So I'm going to talk a little bit about handoffs, understanding that I'm speaking to a Boston Children's Audience. So hopefully I don't have to convince you too much, but I want to make sure we're all on the same page about what handoffs are and why they're important. So I'm going to give you a little bit of a whirlwind background, just so we're all on the same page. Basically, the basis of the work I'm going to talk about is that healthcare errors cause preventable patient harm. And I'll give some representatives citations here, but clearly this is a body of literature that goes back decades. We know that healthcare errors are mostly attributable to communication defects, and the Joint Commission and its work has suggested that 80% of healthcare errors are related to communication. Handoffs are a particular type of communication activity that are particularly prone to defects, as I think many of us who are practicing clinicians have experienced. We know that standardizing handoffs improves handoff quality, and following from that, we know that handoff quality improvement translates into improved patient outcomes. Although I'll be honest with you, this is the weakest part of the evidence, but the best evidence has come out of some folks that are based at your institution. I want to suggest that handoffs are more complicated than you might think, which is why I have based my work on them. I'm showing you the handoff rainbow. This is, think of it as an amonic. This is visual representation of how complicated handoffs are, and I'm going to show you exactly what I mean by this. There's a human factors engineer at Ohio State University named Emily Patterson, who worked with an emergency medicine who since passed away, Bob Wears, to put forth seven functions or framings of handoffs. So the idea is that a handoff does a lot of different things in clinical practice, only one of which is information processing, which is what we tend to think about. There are actually seven different functions that they've elaborated that a handoff is supposed to accomplish. I'm going to share just two of them with you. The first is information processing, which I think is what most of us think about. When we think about handoffs, we think about the transmission of information from one clinician or one team of clinicians to another. And this is why when we're trying to improve handoffs, we tend to think about using templates and checklists. So the goal of this particular function of a handoff is to transfer data through a noisy communication channel, meaning from person to person or from team to team. And the way that we support this function is through standardizing handoffs and using closed-loop communication. There's another one that I really like. The sixth one, distributed cognition, which is the idea that handoffs enable us to share the thinking about patient care. So I'm an intensiveist. It's been lots of time in the ICU. I've had the experience many times of when I'm handing off to a colleague saying, oh, this is Mrs. Smith. She's had a really fast heart rate. We think it's Texas. Oh, but wait, it could be hypolimia or it could be a PE. And then we have a conversation with each other. And we come to insights about care that we never would have had if we hadn't had that conversation. So the idea here is that handoffs create an opportunity for us to think about patients together, to have insights to build and nurture professional relationships. And this is a really important function that isn't captured or wouldn't be supported by, say, a checklist or a template. So the solution here for supporting handoffs, this particular function of handoffs, is to engage other care providers. The iPass work that's been done that I'll touch on very briefly talks about this, when the second S in iPass is synthesis by receiver. So if we are actively active participants in the handoff, we can support this function of distributed cognition. We also need to explicitly acknowledge uncertainty in our handoffs when we handoff accountability and responsibility for a patient to another provider. We should be clear about what we know and clear about what we don't know, so that they have a sense of what to look out for as they assume care and responsibility for that patient. There are some really important challenges to conducting and studying anesthesia handoffs. Elizabeth Lizarra, who's a human factors engineer, based at Embry Riddle Aeronautical University, spearheaded this paper that I was happy to be a part of, that tried to pull all of this together, really using and engineering lens to look at why it's difficult to conduct handoffs and why it's difficult to research and really build a body of knowledge about handoffs. So I'm going to highlight just a few of these to inform the discussion moving forward. First, we have a lot of different kinds of handoffs. So I worked with some collaborators at University of Alabama, Birmingham, to create this taxonomy or sort of language around handoffs in anesthesia practice. Basically, we handoff in the operating room, in the pack you. We're handing off to people who do the same thing that we do. When we do those types of handoffs, we think of those as shift-to-shift handoffs, where they're interchangeable actors. So if I'm taking a lunch break or I'm leaving for the day and I'm handing off to another anesthesiologist, I assume that we're all speaking the same language, that they understand exactly what I'm doing, and that it's really conceptually sort of like that baton pass that people sort of, that they show sometimes when they talk about handoffs. A different type of handoff is a duty relief or a break. It's when I'm going to come right back. So I'm leaving, but you only need to know enough to take care of the patient for five minutes or 30 minutes. I'm not leaving permanently, so maybe I don't tell you that they have got it. Maybe I don't tell you that their dog's name is fluffy. I'm only telling you enough so that you can keep the patient safe in that moment. The third type of handoff, the one that I study the most, is a transition of care, where you have a patient moving from site to site, from team to team. And so you can think of an OR to pack you handoff, an OR to ICU handoff, where the functions of the care, the objectives of the care are very different on either side of that handoff, and the people that are doing the handoffs are different. And so the challenge of standardizing and supporting bolstering that type of handoff is very different than for the others. So having different types of handoffs makes it complicated to both conduct them, standardize them, and to talk about them. I also alluded to the different environments in which we care for patients. So we take care of patients in lots of different places. We take care of them pre-admission in the holding area. We see them in the ER, the OR, the ICU, and the task of standardizing handoffs or accomplishing handoffs across these different sites is different. I've highlighted in blue here the settings that we know the most about. So we know the most about handoffs that happen within the operating room, within the intensive care unit, to some degree within the pack you. And then I've highlighted here in blue the arrows to indicate transitions of care between the site. So we know a lot about OR to ICU handoffs and about OR to pack you handoffs. The other white arrows in the spaces we know very little about. Another problem with studying anesthesia handoffs is that the patients who comprise the sample in any anesthesia handoff study are pretty heterogeneous. Patients who have handoffs, especially intraoperative handoffs, tend to be sicker and their cases are more complex, which makes the comparison of outcomes for patients who have handoffs versus no handoffs statistically very complicated. Here I'm highlighting a study that came out of Canada just to show you the difference between patients who had a handoff versus not in cardiac surgery. And so I'll step you through a couple of really important differences here. If you look at the proportion of patients that had what they're calling a simple cardiac surgery, you'll see there's a higher proportion of patients in the no handoff group that had a simple surgery as opposed to complex. You'll see the cross clamp time is significantly different. The use of aynetropes is significantly different. So if you do sort of a conventional multi variable regression to look at the differences between these two patients, there's going to be a lot of unmeasured confounding that you have an accounted for, which makes it very difficult to compare these two groups. Oftentimes the way that we deal with that is some sort of matching. And so what Chris Hudson and his group did here and what a lot of other investigators have done is what's called propensity score matching, or some sort of matching so that you can equalize the cohorts. The idea here is that you are modeling the likelihood that a patient gets a handoff, and that that estimation of the likelihood of getting a handoff then becomes a variable in your new model so that you can equalize the cohorts somewhat and hopefully deal with unmeasured confounding. And you can see in this particular study that when you do propensity score matching, you do equalize the cohorts and so hopefully you're doing more of an apples to apples comparison. But when you're looking at the relationship between handoffs and outcomes, it's important to make sure that you are actually doing apples to apples comparison so that you are able to have confidence in the findings. So I'm going to move on to talking about interoperative handoffs, which is a big part of the handoff landscape in anesthesia practice. The evidence about interoperative handoffs is pretty weak. So the first studies came out of MGA to Jeff Cooper, and he started, there was actually paper in 1978 where he started to hint at this, but his first paper about handoffs really came out in 1982. And then there was really nothing until he published again in 1989 a paper about short breaks. And Jeff's a very good friend of mine now, we collaborate together. And he said, I should have kept going because no one did anything with handoffs until the 2000s. So you see there's a big break, and then we start to see a pickup in papers about interoperative handoffs. And we're seeing more now, but the idea that I can put all of the substantive interoperative handoff papers on a slide, and you can actually read the names, gives you a sense of just how little we know and how much work there is yet to do. So for those of you who are budding researchers, go for it. So if you ask yourself the question, are anesthesia handoff stirring surgery associated with harm? The answer is mostly yes, but we have to be careful here for some of the inferential threat reasons that I alluded to before, the association is not the same as causation. So I'm going to highlight for you a couple of studies that have looked at interoperative handoffs in anesthesia that give you a sense of what the preponderance of the evidence shows. So this study came from life's saga, who did this work at Cleveland Clinic. He since moved to Michigan and now back to Germany, but he used a very large institutional data set to look at the association between anesthesia handoffs and harm. And basically what he found is that every new anesthesia provider added risk of a composite outcome that I'll show you. So what he did was look at Cleveland Clinic non-cardiac surgeries, tremendously large data set, 135,000 patients. Their practice setting included CRNAs, it included residents and attendings, and the handoff rate was pretty high. So almost 40% of patients in this data set had some sort of handoff. This you'll see across the literature, there's wide variability in the handoff rate, which I think speaks to variability and how we care for patients and how we structure our days. But it's a really important variable as we think about the relationship between handoffs and outcomes. But basically what they looked at was a composite outcome of death and major morbidity. Now those of you who do large database analyses understand that composite outcomes are sort of a double edged sword that they're useful because they increase your statistical power to be able to detect an effect. Some people think that it's sort of p-hacking or manipulating the data to try to power your study to find an effect. Really the key here I think with a composite outcome is to make sure that all of the different composites that make up the outcome are causally linked to the exposure. So here they're looking at death, they're looking at acute kidney injury, they're looking at major cardiac events. And if you can conceptually link the handoff to those different adverse events, then it's logically reasonable to combine them all together into a composite outcome. Where we get into trouble with composites is when people just sort of throw a lot of different outcomes together without any clear sense of how the exposure is associated with that individual outcome. So what life in this group found is in odds ratio of 1.08, so an 8% increased odds of the outcome, that composite outcome per handoff. And this created a big splash back in 2014. It was a well done study within the limitations of what you can do with a data set. And so I think this sort of reinvigorated the rigorous examination of the association between handoffs at harm. On the other hand we have a study here from Vanderbilt. This is Maxim Terrakoff, who found no increased risk with handoffs. So what they looked at at Vanderbilt was 140,000 patients, so a data set that was similarly sized to the Cleveland Clinic. Their handoff rate was much lower, so it's 8%. So this gives you a sense there's something very different about what they're doing at Vanderbilt. The exposure is less common there. Their composite outcome was also death in major morbidity. What they found was an odds ratio where the confidence interval crossed 1, so no association between handoffs and death or major morbidity. Knowing Vanderbilt, I know that Matt Wanger, who runs their center for research and system science, he's been working on handoffs and safety and communication and teamwork for about 20 years at Vanderbilt. So I'm pretty convinced there's something different about that institution, but it is an important sort of caveat in the literature that associates anesthesia handoffs with harm. Interestingly, I'll go back to Maxim Terrakoff. They also looked at short breaks and they found a trend toward an improved outcome in patients who had short breaks where they anesthetized through the anesthesiologist left for a short period of time. Along with Jeff Cooper study in 1989, those are the only two studies I'm aware of that have looked at breaks and they both say that breaks are good. So keep breaking, keep doing coffee. It's good for maintaining vigilance. That's a happy message. The biggest most high profile study was published in JAMA in 2018 and this came out of Canada. So what Jones and his colleagues found was an increased risk with what they call complete handover. So that is what I conceptualize as a shift to shift handoff. So it's not breaks. It's when the anesthetist who's taking care of the patient is leaving permanently and handing over responsibility to another anesthetist. So this interestingly is a population wide study. So it's one of the only studies that has looked at multiple institutions across the province in Canada. Their data set is quite large. It's more than 300,000 patients. The institutional settings though are very different. These are empty only cases. So there are no residents here. There are no nurse anesthetists here. And their handoff rate is exceedingly low. So it's 2.9%. It peaked at 2.9%. And they did the study over a number of years. I want to say it's about five years that they did this study. And so I have questions in my mind about how it's an uncommon exposure. And so if people are not used to handing off, you would expect that there might be an increased risk of harm to the patients with this rare event, this relatively rare event. But what they looked at was a composite outcome of death and major morbidity again. And they found an adjusted risk difference of 6.8%. So increased risk of 6.8% for every handover. This paper was followed by about six different letters to the editor in JAMA, either applauding them or saying here are all the problems of it, but it generated quite a bit of chatter in the anesthesia world to get a paper in JAMA and to have this sort of critical look at how we deliver care. So I've given you just a look at three of the big studies in anesthesia and cooperative handoffs. The takeaways are that a preponderance of the evidence, so of all the studies I showed you that looked at an association between anesthesia handoffs and outcomes, show that there is an association between those handoffs and poor patient outcomes that most of the evidence suggests that the patients who experience an anesthesia handoff are at an increased risk of a poor outcome. This seems as not true for breaks, but there's some huge caveats here in that we know nothing about how handoffs are actually executed at the sites where these studies were based. We don't know anything about handoff quality. And so as someone who does a lot of qualitative and mixed methods research, I think of this as hypothesis generating. It's provocative, it's important, it's noteworthy, but I don't think that we can change our practice based on these studies because we wouldn't necessarily know what to do and what direction to change our practice. So I think it's a place to start, but it's not definitive. So you might ask yourself, are checklists the answer? I talked to you about information processing and how checklists can support handoffs, at least that function of handoffs. And we've seen some notable checklists successes in medicine. So Peter Pronevos, who was based at Hopkins, did this work in Michigan with the Central Line Bundle and multiple hospitals in Michigan, the Tulka Wanda and his group had the Safe Surgery Saves Lives checklists and did an international study showing an improvement in patient outcomes when you introduced a checklist at the beginning of surgery. As a mixed methodologist and sort of an amateur sociologist, I find it really interesting that those very same checklists have been shown to fail. So when you look at the Central Line Bundle, there's a group in the UK that did a study that they called matching Michigan, where they tried to do exactly what was done, multi center study to introduce Central Line Bundles to decrease clapsies, Central Line Associated Bloodstream Infection, and they found no improvement and a trend toward harm in the group that actually had the Central Line. So there was also a study in Canada thinking about the Safe Surgery Saves Lives checklists that did the same thing and found no improvement in patient outcomes with the institution of a presurgical checklist. So I think checklists can be helpful. I don't think there are enough by themselves. I think there's something else that comes along with the checklist, whether that's culture change, whether that's workflow change that we need to understand more deeply before we can say that the answer by itself is a checklist. So if checklists alone are not the answer, how do we improve perioperative handoffs if we think that they're associated with patient harm? Clearly you all know iPass. It was born at Boston Children's U-Harm, sure, very familiar with it. I'm not going to touch on it very much, except to say that the work that I'm going to talk about certainly dovetails with iPass in that the principles are that you want to standardize handoffs, you want to create expectations for what people are supposed to do. And that that translates into more reliable, consistent, high quality handoffs, which should then translate into improved patient outcomes. And I'd be happy to talk more in the discussion session about how all of that sort of fits together. But what I want to talk about is some of the work that we've done to standardize ORD ICU handoffs that I think gives some insights into how you can use mixed methods and implementation science informed approaches to improve handoffs. So for those of you who have participated in an ORD ICU handoff, you understand that they can be problematic. Basically you have a patient, often an unconscious patient who's being moved from one site to another site. They're either critically ill or at high risk of deterioration. You're moving them, you're moving the technology, you're moving an ALIN, maybe it's a lumbar train, a pulmonary artery catheter, chest tubes, a bunch of stuff. You've got different teams that are interacting with each other that are from different disciplines, they have different priorities, and all of this is supposed to happen very quickly. The consequences of this happening poorly are that you can injure yourself, you can injure the patients as you're manipulating all of the technology and stuff through the hallway and losing attention on the patient. Patients can deteriorate. If you are focused more or we are focused more on the logistics of moving them, it's very easy to turn your attention away from the patient and their hemodynamics for instance. We can introduce medication errors, we can misinformation and all the chaos and our teams can function poorly. As my team was getting started thinking about ORDICU handoffs, we did a review of the literature on all of the papers that had standardized ORDICU handoffs, which at that point were almost all in cardiac surgery. And actually mostly in pediatric cardiac surgery, the pediatric community was way out in front of us. We had in front of the adult community in terms of thinking about how to make the ORDICU handoff process more reliable. So what my team did was synthesize all of that literature into what looked like a common handoff protocol or process. So what all of the published studies showed us is that when they standardized handoffs, they used a choreographed sequence where there were some sort of pre-handoff preparation. Where before the patient ever left the operating room, there was either a telephone call or a physical visit or some sort of information transmission from the operating room to the ICU so that the ICU could get ready. Then the patient was brought to the ICU, the patient was stabilized, all of that equipment and technology and spaghetti was transferred over. Information was exchanged at that point and when information is exchanged, usually there was a use of a tool or a template or a checklist. So in my mind, this is where the checklist comes in. It's a part of the handoff process. It's a part of the protocol, but it's not the protocol by itself. After you've exchanged information, there's an opportunity to discuss the patient, to ask questions and to specify a plan. When you look at the ORD ICU handoff research that has been published since then, so from 2007 to 2019, we see some variability in what the different studies are looking at. Different ICU types that are involved in these studies. Some of them have looked at sustainability over a number of years, many of them have not. There's some variability in terms of whether they use electronic tools, whether they use a physical template or a checklist and then who's expected to be there in the handoffs. But in general, what we're expecting to see with an ORD ICU handoff is this. So it's a team-based process. Ideally, you have some sort of meaningful communication at the patient bedside. So here, I've got a patient that's unconscious, has no idea what's happening. He's really hoping that we're doing this well. We've got a team here in a nestletist, an ICU ordering provider, who could be a physician, they could be a nurse practitioner, a physician assistant. You've got a surgeon who's looking surly, he's probably hungry because he doesn't eat very much, and then an ICU nurse. So you've got all of these people together having a meaningful conversation about the patient. You've got nurses here that we're calling secondary ICU nurses, meaning they have other patient assignments, but they've come into the room to help get the patient settled so that the nurse who's receiving the patient can participate in the handoff. And then you have the respiratory therapist and any other staff that need to be there, attending to the patient. So I'm going to tell you about a study that we've been doing at Penn for about five years, six years now, called HATRIK, which stands for Handoffs and Transitions and Critical Care. This is a screen grab from our website, which is still live. So if you go to HATRIK.com, you'll see this website. You can sort of go through the study site. The only trick is that you have to remember how to spell HATRIK. So this study is based at Penn Medicine. It was based in two mixed surgical ICUs and two different hospitals. The top one here is Penn Presbyterian Medical Center. That's my IC, where I'm the co-medical director, one of our trauma ICUs. And then the one on the bottom here is Rhodes-5, the surgical intensive care unit at the Hospital University of Pennsylvania. We were interested in patients who were admitted from the OR to the ICU, but really our focus was on the clinicians. We wanted to look at the surgeons, the anesthesia staff, the nurses, the ordering providers, and really engage clinical leaders as well to understand how to improve this process and make it more reliable. So in terms of a timeline, we did contextual inquiry, which I'll explain early in the study. So June to October of 2014, we developed an intervention. We went live with it in 2015. Our pre-specified data collection period was July 2015 to January 2016. And then now we're in the sustainment phase where up until COVID, we continue to collect data to make sure that people were continuing to adhere to the process. Our measures are pretty complicated. So I'm going to go into the structure of the study, but it's a mixed methods, hybrid effectiveness implementation study. So we're interested in whether standardizing handoffs in mixed surgical ICUs works, and I'll say that in quotes. And so the measures of effectiveness here are handoff quality, teamwork quality, and then a lot of quantitative outcomes related to whether the handoff is doing what we expected to do. The primary outcome in bold here is information omissions, which is out of a fixed list of topics that should be discussed in handoff, how many were omitted. And the reason we chose that as a primary outcome is that all of the studies that preceded ours had looked at information omissions, and we wanted our work to be interpretable in the context of the pre-existing literature. We were also interested in implementation outcomes, which precede sort of effectiveness. So it's one thing to create a standardized handoff, but people have to actually do it. So the implementation outcomes help us understand whether people do what we ask them to do, why they do, why they don't, what some of the challenges to implementation might be, that then undermine the effectiveness of the intervention. So we looked at concepts like acceptability, appropriateness, fidelity, and sustainability. So I've mentioned implementation science, but I haven't defined it, so I think it's useful to do that here. Implementation science is a relatively new field. Some people call it old wine and new bottles. It's based on social science, public health, human factors, engineering, management theory. But basically it's a field that's focused on promoting the systematic uptake of research findings into practice to improve the quality and effectiveness of health services and care. And the idea behind implementation science is that there is what we call an evidence to practice gap, meaning we have a lot of evidence, scientific evidence, and many different domains that never makes it to the bedside for some reason. Whether it's difficult to implement, whether it interferes with or clothe, there are lots of reasons that something that works that is efficacious might not actually reach a patient. And implementation science is concerned with narrowing the evidence to practice gap. So this is one of the fields that underpins the study that I'm telling you about. Implementation science fits in the translational research spectrum as conceptualized by the National Institutes of Health. So here I've got a translational research spectrum populated with people from Penn. Some of you may or may not know. The point of this schematic is to show you that all of our biomedical research relates to each other. So I'm over here on the right doing implementation research and population and outcomes level research. But my favorite example of a person on the other end is Vicki Badell, who's ever in the upper left corner. I don't know if you can see my mouse. But she does zebrafish research. And I don't know what a zebrafish is. I don't know what it looks like. I couldn't pick one out of a lineup. But what she does and the things that she discovers will never reach a patient unless I do what I do. So as I'm explaining what I do to a basic science colleagues, it sort of helps to really all of our research to each other. So implementation science is in there all the way on the other end. So back to Patrick. I want to show you how the study unfolded and I'm going to use an implementation framework called Epis to do it. One of the reasons I'm doing this is that implementation science relies very heavily on theories, frameworks, and models to structure the work that we do. And so this is an example of how that might play out. Epis is actually a pretty complicated framework. So I'm just using part of it, but it helps show how you might step through an implementation study. So in the exploration phase, which I called contextual inquiry, our research question really is how are we doing? Oh, are ICU handoffs and our clinicians actually ready for change? And the way that we approach that is by doing observations. We do interviews, focus groups, surveys, and we review documents that people create to support their work in handoffs. In the preparation phase, the question is what does a standardized handoff look like for mixed surgical ICUs? So we developed our process. We synthesized the literature. We created prototypes and we did insight to simulation to test through the prototypes with our providers to refine them. In the implementation phase, the question is how do you get clinicians to change their behavior and do handoffs in a new way? And so the mainstay of our implementation approach was with training and evaluation, coaching, walking people through the process and helping them troubleshoot the process. And then in the sustainment phase, the question is do clinicians exhibit fidelity over time, meaning do they adhere to the process without active maintenance? So in the sustainment phase, what we've done is step back. We don't do any coaching. We don't do any training aside from the mandatory online sort of digital training that all of the interns have to go through. But we just want to see if the process will sustain itself over time without research assistance sort of standing there and getting people to do it. So in the contextual inquiry phase, we started off with some observations. We observed 94 handoffs. We did some dual observer handoffs because we wanted to calculate interrater reliability. Our final sample after deduplication was 68 handoffs. 64 of which actually included a handoff and four of which involved a patient being moved from the OR to the ICU with no substantive communication at the bedside for any providers. It was a little distressing actually. There were a few patients where a nurse would show up a few minutes after the patient and say, oh, look, I have a patient, which is ideally not what happens for an ICU patient. So this was four observers that did this work over about 500 hours. The modal handoff looked like this. So you would have a patient who went from the OR to the ICU. You would have one nurse in the room trying to get the patient settled. You would have an anesthetist. This is me, by the way, with all the spikey here. You have an anesthetist in the room talking to no one in particular. And then in the hallway, you would have a surgeon talking to an ICU ordering provider. So there was some communication happening. It was pretty dysfunctional. And the nurse oftentimes would leave the handoff not knowing anything about what happened to the patient other than what they could find in the medical record. We used a structured template to observe the handoffs in the room. And in the question and answer session, we can talk a little bit about how we got around the health and effect using video observations. But after each handoff, we also had our observers reflect on some open-ended questions. And this gave us some qualitative data to understand what was happening in the room. So the types of questions that we asked our observers to opine about was to describe the interactions between handoff participants, to describe handoff engagement, and to describe the actors in the handoff scene. And the types of information we would get looked like this. So this says, when patient entered the room, there was a lot of commotion and noise. Nurses began a physical exam while surgeon gave handoff to attending. Nurses did not pay attention to handoff. The room was extremely busy. Anesthesiologist and surgeon seemed on good terms. They joked and spoke back and forth during time in the room. Nurses appeared distracted during the surgeon handoff. So it's insights like these that help us understand above and beyond the quantitative data what's actually happening and give us insight into how we might need to approach restructuring that handoff. We surveyed our clinicians and asked them what would make them consider changing how they did handoffs. Most folks said that they wanted some evidence that a different process would improve communication or patient outcomes. But importantly, almost no one said that they were satisfied with their current practice, which we thought gave us a bit of a burning platform to move forward. We asked people to free list words that they thought about when they thought about OR to ICU handoffs. We did get some positive words, but mostly negative. Mostly words like choppy and disjointed and unstructured. So people were telling us that they thought the handoff process was pretty dysfunctional. We did some interviews and focus groups and the themes we encountered had to do with people prioritizing the handoff below, below many of the other things that they needed to do and being worried about time constraints. People said that handoff quality was affected by professionalism and teamwork and were able to give us examples of both exemplary and poor professionalism and teamwork. And the frustrating aspects of handoffs included that providers were inconsistently present. So those ordering providers that needed to write admission orders or that residents, specifically anesthesia and surgery residents didn't know enough about the patient or the case to participate meaningfully in the handoff. I'm going to share with you an example of what we got from the interviews just because I think it shows the power of qualitative data to give you insights into what is happening with a phenomenon of interest. So we asked one of our advanced practice providers why don't post up handoffs work. She said so there are some people that just bring the patient in and basically run away like one of my colleagues. He does not give you a good sign out. He does not tell you anything unless you specifically ask a question of him. He will not tell you if the order fell apart and you had to reconstruct. And we asked her why might someone run away. She said I think that they're not comfortable in the ICU and if something happened they don't want to be responsible for it. So they leave the patient and they go because they don't want to be caught up in this. I think that sometimes it's because unfortunately you have another case or you're getting called into another room. I think sometimes some people are hungry and they just kind of toss and run. And also there's one person that doesn't like one of the physicians takes care of the heart, Dr. X. So if he sees him he doesn't stay long in the unit. They don't get along. So we created our new process. I mentioned that we synthesized the literature. We came up with a prototype. We used inside to simulation to help us build a new handoff process. And this is what we came up with. This is just one piece of it. We also created a template but I'm just showing you the what we call the Candyland poster or the monopoly poster that relates our expectations for what should happen during the handoff process. Which I swear is exactly the same as the clock but just with a graphic designer paying attention to it. In terms of implementation, we did a lot of planning and education talking with people coaching them. We went to departmental conferences and educational conferences. We wore obnoxiously large buttons. Asked me about handoff buttons that were about four inches wide and we carried around Candy and Grinola to tell people about the handoffs to capture their attention. Well, we found after so we had a wash in period first before we started observing and putting data into our final study data set. We found that information emissions decreased which is what we expected about 21%. We found improvements in teamwork and professionalism scores and I'll tell you more about that on the next slide. We found more consistent presence of providers at bedside. We did not see unit level changes in patient outcomes like length of stay or duration of mechanical ventilation. Mostly because we weren't powered to do it but I'll talk a little bit later about some of the challenges in linking up handoffs to patient outcomes. We did look at the duration of handoff which we were expecting to stay the same or decrease based on the pre-existing literature. We found that the handoff duration actually increased. So we went from about three and a half minutes up to a peak of just over 10 with a learning effect and then in the sustainability phase we've settled out at about six. Normally when you tell perioperative providers that you're going to spend more time doing something it's a non-starter. What our providers told us is that they were spending less time outside of the handoff getting information that they needed and so they thought that this time was well spent. When we looked at fidelity to the HATRIC protocol we found that people mostly adhered. So the HATRIC protocol has 10 steps in it so this is a histogram showing the level of adherence. Certainly it wasn't perfect and some of the mixed methods data some of the interviews give us insight into why people might not always adhere but in general we saw good adherence to the protocol. And more importantly in the quantitative data I'll show you momentarily the improvement in information emissions was only seen in handoffs that adhered to seven or more steps of the handoff protocol. And so even though this is a quasi experimental design we're confident that the findings are related to our intervention because of that association between fidelity and outcomes. So we saw as I mentioned decrease in information emissions which is great but it was only in certain patients and it was only in certain categories of information. So when you look at all patients you see a decrease in information emissions which is what we wanted to see but that was completely confined to new patients to the ICU. So we stratified our analysis by whether patients were newly admitted or whether they were in re admission to a re admission might be someone who went to the OR for a tricky estimate for instance or for a peg tube or for a washout versus a patient who was new. And so we only saw improvement in the new patients and we only saw improvement in these categories that you see on the screen. We could not get anyone to talk about allergies they didn't want to talk about it before they didn't want to talk about it after I'm not sure why nobody likes to talk about allergies but we'll have to dive into that later. We also looked at information transmission as a function of teamwork and professionalism and what you're looking at here is a heat map where it's essentially red yellow green. It may not be colorblind friendly which I apologize for but you can see the different gradations and color so we expected all of these to improve together but what we see is that here in the first column this is information emissions we see an improvement as I showed you before but you don't necessarily have strong teamwork and teamwork and professionalism in those handoffs with low information emissions. So you could have a handoff for instance with sort of a moderate level of information emissions and great teamwork and professionalism or with great information transmission and horrible teamwork and so we're still trying to unpack the relationship between those those characteristics. One of the ways that we're doing that is with something called qualitative comparative analysis which I'll explain shortly. So ongoing work we're still working on this we have a number of offshoots I told you about sustainability we're doing a qualitative study of champions to try to understand why some people in our institutions have decided that they want to champion Patrick so you know they'll walk into a room and they'll go everybody we have to have trick now and we don't know why we didn't ask them to do it so we're trying to understand why it is that they've taken this under their wing so that we can moving forward create the conditions for champions to. To help support a project and then for disco this this is disciplinary differences in communication so what our physician assistance and nurses told us is that they learn about interprofessional communication differently than physicians as we're trying to understand how to support different clinicians in their information needs and and really support the team and its ability to do to communicate effectively across disciplinary boundaries. I mentioned qualitative comparative analysis which it's a little bit of a qualitative person and so turning qualitative data into quantitative data is sort of like here see but this is a way of turning qualitative data into data that you can then use bully and logic to analyze. It's logic essentially so if you like logic puzzles you would like QCA but we're trying to understand what the necessary conditions are for fidelity to the handoff which will help us understand how to support implementation moving forward. We have also moved out of pen into the community so we're doing this work at Christianic Air Health System in Delaware which is an independent academic medical center community based hospital trying to understand how to standardize handoffs there where the work flows are very different. And then we just have even our one that was just funded by NHLBI to spread this to five different health systems so we'll be doing Patrick at Temple University in Philadelphia Cooper in New Jersey Johns Hopkins U.T. Southwestern and then some of the outlying pen hospitals trying to understand how to tailor and standardize handoff and get it implemented reliably. I mentioned briefly that there's a problem with handoffs and outcomes the holy grail of handoff research is to be able to demonstrate a relationship between handoffs and the outcomes that we care about like mortality and major injury. The challenge is that those outcomes are very uncommon and so what's easier to detect are these more common less impactful events non routine events is something that has been advanced by Matt Wenger Vanderbilt. But I think the idea here is that as we're doing handoff research and we're trying to associate handoffs and outcomes we need to look at a lot of different types of outcomes to reliably demonstrate an effect and you're only probably going to see an effect with mortality when you see those very, very large studies like the 130,000, 140,000 patient studies that I showed you earlier. If you're doing smaller scale studies that have different research questions then it's very unlikely that you're going to see changes in these outcomes, these sort of big ticket outcomes. So that's it. I want to acknowledge my team, my very large team, they're not all with me at the same time but lots of these folks have rotated through my lab. My family has been very generous with their time and I'll say thank you and please feel free to email me with any questions. And then I think we have some time for discussion. Thank you again Megan that was a really fantastic presentation. Are there any questions or comments feel free to chime in or chat them in? Megan I want to thank you, Joe Carrero from the NSTG, I'll be department. I just want to ask a couple quick things one is just have you been able to leverage technological advances with this specifically around EMR and monitors within the OR. We've been obviously looking at trying to improve our own handoff and part of that has been looking at whether or not a forced reminder of the handoff that comes up as you have been. So if you want to document the change in provider within the MR should come up and how we might do things like that. And I'm interested in any information you have around how you may have leveraged that part of your information system. Yeah, so we haven't done it yet. So you know what I've been impressed by is that workflow seems to be one of the most important factors in influencing what people are going to do with respect to handoffs. And EMR based solution has been successful in many places for the interoperative handoffs and we've just gone live at our institution with an interoperative handoff tool as well. I think that works well with the interoperative workflow if you have a setup where you've got people anesthesiologists that are essentially stationed at the computer where that's where most of their work is. And so it makes sense. Let's pull up a tool. I'm looking at the computer anyway and you can sort of integrate that into the workflow. With some of the ORD ICU handoff work that we did when we did our contextual inquiry and we were watching people no one touched a computer for 20 minutes. You know, a patient would roll in and there would be this chaos and these conversations with no one was touching a computer. And so we realized that if we wanted to standardize that process if we introduced something computerized it would totally blow up the workflow. And it may be that we eventually get there as people become more facile with using their mobile devices or they do integrate computers. But I think there's an important workflow component to this that needs to be incorporated. So I think the EMR makes a lot of sense. It makes sense in offloading some of the information that you don't necessarily need to say it. You can say look, here's the record. The allergies are right here, the meds are right here and then you can focus on the things that make the patient unique. So I think there is a really important role for the EMR, but it kind of depends on how everything else is structured with work. I see something in the chat. Okay, go ahead. Yeah, I'm sorry. This is Bill Sparks. I was interested in how you know when we introduced checklist when we introduced a new process, there's a lot of attention and energy to it. Has your team developed any abilities to keep the consistency of the product there over the course of duration of time, meaning that at the six week or 12 week period. Is the same process being done that it was when it first started. So I don't have a good answer for you and it's that's one of the holy grails I talked about is the association between handoffs and outcomes and the other is sustainment or sustainability of the interventions that we put into place. What we found with Patrick is that it was a little bit of a happy accident. This became integrated into the culture in a way that we we have lost control over in a good way. And I can't tell you how it happened or why it happened, but it's become a part of speech at 10. So if you go to these particular hospitals, you'll hear people talking about being ready for the hat trick or you have to come to the hat trick. They don't talk about handoffs. It's become an expectation. People are holding each other accountable. If someone doesn't show up for the hat trick, then they're supervisor here is about it. So I think part of what you want to look for in terms of sustainment or sustainability is figuring out how to promote ownership, how to create champions, how to get people to take responsibility for whatever it is that you want them to do. And that helps sustain the process. In terms of the quality of the product, we've seen degradation over time and what people are doing with Patrick. So we've seen, you know, they adhere a lot upfront and then you start to see it trickle off a little bit. Sometimes our surgeons don't stay for the entire handoff. They'll give their report and then they'll leave. And we're honestly still trying to figure out how you sort of do booster shots and maintain the quality of the product. So I think there's a couple of questions built in there, which is how do you keep the product and then how do you maintain it over time. And it does take some active intervention to do that. And just to follow up with respect to the perioperative, the pre-op to interrupt the interop, the simple pack you do you guys apply the same system there or is it a different system? It's a different system. So, you know, I think that sort of thinking about the learning health system concept, this is a process that evolves over time. I think it's difficult to take one particular solution and to kind of fit it into every setting. I think that we could eventually get there. And I think this is sort of where I think about iPass coming in. It's been phenomenally successful in some areas and it hasn't captured quite a foothold, the same foothold and say the perioperative space because the workflow is so different. And so I think going back to that idea about workflow, you want to support the clinicians, you want to make their job easier. And if there's a way to have a common structure that makes sense, I think you should do that, but I don't know that we have to force people to adhere to a common structure just because we want to say that it's the same across the house because people will find ways to subvert that. So, it's a work in progress, but I think the principles, I go back to those sort of seven functions or framings of handoffs, we need to support all of those principles in some way or another. Thank you so much. It was a great presentation. Thank you. There's a question in the chat that you from an iPass advocate, that's great. You're working on Interoperative in ORDA, Pack You Handoffs, and you found the second S in iPass is hardest to implement. Yeah, it's hard to get people to speak up. I think this, it's culture and culture is not easy to change. I think that people are afraid of being judged. I think that people are, it's, it's, it's multifactorial, let's say. I think that if we create expectations for people talking and we say something if they don't, so if you're the receiver and the person, if you're the sender and the person who's receiving doesn't actually synthesize, do you have a culture where the, where the sender can say, so what did you take away from that? I would have trouble saying that's one of my colleagues right now. Honestly, I'll be, I'll be honest with you. So trying to create a system where it's safe to do that is, is difficult, but I think that's what's needed is for us to hold each other accountable. There's a question about how you reconcile medication infusions that are started in the ORDA and continued in the ICU is their standardization of medication concentrations. So we struggled with medication concentrations. We, we finally did standardize them with the exception of phenyl effron. So every, so norbeneffron, epinephrine, vasopressin, dopamine, all of those are standardized. Phenyl effron, what we did was we put it in a different container. And so we don't put phenyl effron infusions in an OR concentration in a bag anymore. We put it in a syringe. So there's a very clear visual sign to the nurse in the ICU that this is a different med. The connector for the laryse pump is actually different. But the way we reconcile is that the anesthesia record is in epic, our version of epic and the, I see record is in there as well. I think, and I don't know 100%, but I think there's a way to pause it and it just kind of carries over, but it, I don't have the details on that. So I'm not going to make it up, but, but we haven't had issues with their reconciling on the transfer. What types of things do you do to encourage champions and support your support their efforts? We're still trying to figure that out because our champion self designated what they told us was the reason that they wanted the reason that they loved Patrick was that it made their life easier. So for the nurses in particular, but the surgeons as well, they ended up doing less work outside the handoff with this intervention. And so the takeaway for me from that is if you design an intervention that complements clinicians workflow that somehow makes their life easier, then it's easier for them to get behind it. So that was really part of the push for Patrick was not here's another thing for you to do. It's, hey, you're doing this already. It's mostly good, but if you just tweak it this way, it'll be so much better and your life will be easier and you'll have fewer phone calls. And that was a really appealing message. And so that's that I think informed people getting behind it. The only people that I'll say suffered, I'll put it in quotes were the anesthesia team because they weren't saving any time. They were spending a little bit more time and we don't have a lot of anesthesia champions, but they're now they understand that this is an expectation. And so the incentive on their end is that they won't hear back from someone who said they didn't participate in the handoff. So it's interesting that the different members of the team have different incentives and different drivers in terms of their behavior. So there's a question about verbal handoffs. One issue about verbal handoffs is that people are then documenting it secondarily after the fact, which is extra work and it could have errors. And the ICU numerous patients are dropped off around three to five. And then an overnight attending comes in in the middle of the chaos 4 p.m. and gets a secondary verbal handoff right after handoff finished. So when I'd have an integrated written summary from surgical and anesthesia team from OR with acknowledgement from ICU, which would prevent which would decrease workload and prevent errors. I agree with you that it would prevent errors. I don't know if it would decrease workload workload. I think you'd have to figure out a way to make it so that it decreased workload. And one of the things that our nurses told us is we were so in the sustainability phase, we wanted to add more of a documentation piece to Patrick to have the nurses do the documentation. And so we went to our nurses and nurse managers and we said, hey, we just want to add this like two little questions because the nurses are already documenting anyway. And they said, no, no, we're not documenting anything else. No, you can't have it hard stop. And we said, oh, okay. So if you can figure out a way to do this written summary, I agree it would be really useful. But if you can figure out a way to do that in a way that doesn't increase anyone's workload, I think you'll have a better chance of it catching on. Yeah, because we all have competing priorities that everybody it's like death by a thousand cuts one more thing and you'll have a revolt is my guess. Thank you again, Megan for a really superb presentation. I think we unfortunately ran up time. Thank you for having me so much fun. Well, we'd love to have you back sometime in person. That would be great. So I think we'll close things out now. Everyone have a great day. Bye everybody. Thank you. Bye.