 We have an exciting announcement. We've partnered with MedStar Health to be able to assign CME credits. So, if you are a physician, you can claim one AMA, PRA category, one credit by participating in this webinar today. And you can go to our website to find out information on how to claim the CME as well as you'll be getting an email from me later today with information on. How to claim it there as well. So, with that, I'm going to move on to the next slide and walk you through a real quick agenda. So, for the 1st, 10 minutes, I'm going to be spending some time introducing the patient safety movement foundation and our actionable patient safety solutions that'll be followed by 40 minutes of our expert presentation led by Brandon. And then the last 10 minutes will be a Q and a remember, because there are so many people on this webinar, please do keep yourself muted. There is a chat feature on the webinar, which will allow you to ask questions. So please feel free to post questions throughout the webinar. We will be monitoring those and will be circling back and and pulling those questions during the last 10 minutes. So, with that, about the patient safety movement. So, for those of you who may be joining the patient safety movement foundations webinar for the 1st time, the patient safety movement has this very bold goal of 0 preventable deaths in hospitals. And that's because we believe that we can't settle for anything less than 0 because 1 preventable patient deaths is 1 too many could be you could be me could deal with 1. So, we really have to focus on 0. And how we do that is we foster new efforts by building on the existing patient safety programs through commitments to 0. So we are not a membership based organization. We're commitment space. So we're looking for people and organizations to take action. So I'm going to just real quickly here walk you through the 5 main groups that the patient safety movement foundation works with. So we work with hospitals and health care organizations. We ask hospitals to share publicly what they're doing to improve patient safety. We call that making a commitment. So they fill out a web form sharing the processes that they're putting into place and the action plans that they're putting into place in order to reduce in hospital harm and mortality. The 2nd group that we work with our committed partners. These are essentially the way I think about it is anyone who would be willing to wave a patient safety flag if it existed. So, it can be advocacy groups, professional societies and associations, other nonprofits. Anyone who's interested in patient safety. We believe that there's a way to work with them and instead of being duplicative be multiplicative. So those partners signed commitment to action letters detailing exactly what they're going to be doing to improve patient safety. The 3rd is healthcare technology companies and this is really core to the work of the patient safety movement because of our founder Joe Piani who is an entrepreneur and engineer, but we ask healthcare technology companies to sign an open data pledge. Which says that they will not knowingly interfere lock or charge on top of the product that they're selling in order to improve patient safety. The 4th, our patients and family advocates. So, again, can't say patient safety without including patients in that we encourage stories to be shared so we can learn how, how, what. What good catches have been identified and how we can improve. So, we also utilize resources that have been shared through patients and family members and we also have a mobile application that we encourage patients and family members to use when they go into hospital to improve patient safety. So, the actionable patient safety solutions. These are kind of the evidence based best practice documents that we've put together over the last 8 years and we have 18 overarching subjects. Normally, our quarterly webinars touch on 1 of these, but we were so excited about the eb boarding topic that Brandon brought to us that we wanted to have a 1 off special topic on eb boarding. So these for those of you who are in the hospital setting who have an opportunity to talk to your leadership. Please make sure that you have a process in place for each 1 of these topics. If you don't, then it's going to be really hard to get to 0. So, I'm just going to spend a few minutes talking about the patient safety movements impact today. This graph represents the hospitals that have committed to 0 that have made commitments through our website. And so, as you can see early on, we had a small amount of hospitals that were very passionate about improving safety. And most recently at our summit in January this year, we announced that 4,710 hospitals have now made commitments to the patient safety movement to improve safety and those spread across 46 countries. The 2nd thing that I wanted to focus on is the live saved annually by hospitals. So remember the little asterisk is denoting that the numbers that hospitals report to us are self reported data. We don't go in and do audits. However, you can see that the number of live saved has increased almost exponentially, not quite since 2013. We were able to announce earlier this year that those 4,710 hospitals have saved 90,146 lives in 2018 alone. So it's about 273,000 lives. If you add up all of those cumulatively, since we began. So, with that, I'm going to pass it over to Brandon. I'm going to do a quick introduction. So, Brandon Lau is assistant professor of radiology, radiology and radiological science and health science informatics at Johns Hopkins School of medicine. He's also an associate faculty of the Armstrong Institute for patient safety and quality. His primary research interest is in the use of electronic health record data to improve care quality, patient safety and clinical education. Clinically, the majority of his effort has focused on the prevention diagnosis and treatment of VTE or venous thromboembolism. He's authored and co-authored more than 50 peer reviewed publications and has received more than $5 million in extramural research funding to study and improve the use of health. IT for the improvement of quality and hospitals. And most recently, just a few days ago, I saw that he also received the 2019 Johns Hopkins catalyst award, which is really exciting. So with that, I'm going to pass it over to Brandon. Brandon, are you there? Can you hear me? Yes, I can hear you. Go ahead. Thank you very much for that wonderful introduction. I'm really excited to talk about some work that we've been doing here on using real time data to try to improve clinical decision making and specifically around an issue that seems to exist in emergency departments and hospitals all across the country. And that is prolonged boarding time. So just getting into this, what it is that we're talking about when we say boarding time, boarding begins when a decision is made that a patient will be admitted to the hospital, but that a bed is not currently available for them to occupy. And as a result, they end up spending time, prolonged amounts of time in the emergency department. And this has taken some interest within hospitals and also within accrediting organizations where the Joint Commission has considered the prolonged boarding time in emergency departments as having a negative effect on the health and potentially causing safety issues and concerns for patients. And when you think about this, the decision to admit a patient to the hospital means that they are in need of specialty care. They are in need of more care than what can be provided at home or in the emergency department where there's a lot of focus on highly acute issues. So every hour that a patient is spending in the emergency department is time that's delaying them from receiving the care that somebody has decided that is appropriate for their current health level. So there's already a potential issue here. There's also an economic issue. It's been estimated that for every hour that you can reduce a boarding time in the emergency department, hospitals can actually increase their revenue by about $13,000. So just in case you happen to be a hospital administrator, there is a financial implication to this as well. But perhaps most concerning is that there have been studies that have suggested that longer boarding times are associated with an increase in mortality, which is really the point of what we're going to be talking about today. Looking at the studies that have been published associating boarding hours with mortality, there was a study out of one hospital that looked at 40,000 patients who were admitted. And what they found is that the longer the boarding hours, the higher the level of mortality. So they found that about 2 to 3% of patients with boarding hours less than two hours experienced in hospital mortality, whereas that number jumped to 4.5% for those with boarding hours longer than 12 hours. And that was looking at all patients who were admitted to that particular hospital. There was a separate study looking at about 40,000 patients in a different hospital and specifically looking at non-ICU patients. These are the general medically ill patients, the patients who are coming in with shortness of breath, with chest pain, not an active heart attack, not people who are getting admitted to the ICU or to the cath lab. These are general medical patients, non-ICU patients. And what they found in this group is that longer boarding hours were associated with a greater risk of inpatient mortality, a significantly higher risk. So we think that there's something that's associated with this delay in care, delay in getting into beds that suggests that there could be a positive effect on inpatient mortality if we can potentially reduce boarding hours. That number of national goals that are set for boarding hours, the Joint Commission, which has prioritized reducing ED boarding hours, is expecting that the average length of boarding in the emergency department is four hours or less. Arguably based on more recent data, might suggest that two hours would be a more laudable goal, but certainly need to start with something. If you're on a national level, average boarding hours in the emergency department range from two to over 24 hours, shockingly enough. So there's an incredible opportunity to improve this practice, to hopefully improve other outcomes that patients experience. And I'd like to talk a little bit about what it is that we've done at the Johns Hopkins Hospital to try to address this issue. So to give you a bit of a sense, we're going to focus specifically in our Department of Medicine, our internal medicine department, where we have almost 600 full-time faculty attending physicians and over 3,000 employees including nurses, tax fellows, residents all trying to work together to provide coordinated care for patients during their entire course of hospitalization. And these cover multiple different subspecialties that we admit to the Department of Medicine for, including cardiology, the renal service, our residency program floors. We have four different floors staffed by residents and a supervising attending. Over the course of the hospital, we have about 270 to 280 beds that are staffed that we see over 13,000 inpatient admissions. And as the Department of Medicine, we have over 96,000 outpatient clinic visits per year. So we have a fairly large population that we're attempting to serve within the department. But also in 2012, our hospital underwent a massive renovation, including to our emergency department. Our emergency department greatly expanded in size, including more than 60 private exam rooms. We have a 17-bed acute care facility within the emergency department, six trauma bays, eight emergency psychiatry beds, and also an expanded radiology suite. So we now have the capacity in our emergency department to see a larger number of patients and ideally assess and determine current condition to triage as appropriate. But in the process of building that, it created a number of operational challenges. We didn't have additional capacity within the Department of Medicine to receive additional patients who were being admitted. We were also focused heavily on 30-day readmissions and reducing readmissions, which naturally has the effect of prolonging length of hospitalization to ensure that patients are at a peak readiness for discharge. We also have pressure, of course, from the Accreditation Council for Graduate Medical Education to confine ourselves within the required duty hours of our residents. And perhaps most concerning, and I think that this is an issue in hospitals across the country, is a culture of no information. We collect more information in healthcare than has ever been collected in the history of medicine. And the ability to get that information, the data back to frontline clinicians and decision makers who are trying to do their best for patients every day is somewhat of a challenge. And when you don't have that information, it feeds into this lack of effective communication between care teams, particularly between departments. And as you can imagine, as you're trying to get patients from the emergency department into any other department within the hospital, effective communication between care teams is absolutely critical for timely throughput. Now, as an indicator of the volume that comes through the emergency department and really why we're focusing on our internal medicine department, 23% of the visits to the ED result in inpatient admissions, and 65% of those come to the Department of Medicine. The workflow that goes on between the emergency department and the internal medicine department is absolutely critical to ensure that those 65% of admissions actually find their way to a bed in a very timely manner. So how these two departments interact with each other, the information that's shared with them is absolutely critical to ensuring the patients get into the appropriate bed. And when you look at the process of taking patients from the emergency department and getting into a bed, there's a process that one must go through. So first, a decision needs to be made that a patient should be admitted to the hospital. And when that decision is made, a request is put into the electronic health record system to request a bed. At some point after the request is made, a bed becomes available. There are multiple different systems that determine bed availability that I'll go over in future slides. But after that bed is available, that then allows a clinician to assign the patient to that bed in that unit. And once the patient is assigned to that bed, they can then depart the emergency department. This is the general flow that most emergency departments would follow, the classic demand and then supply and then availability. And when you look at how this is defined, 56% of the time is really focused on waiting for the bed to become available. And 44% of the time is the process behind it of which bed is available, which bed is appropriate for the patient, who else is waiting for a bed to become available, what is the level of priority, and how can we get the patient into that bed. So we see this bed wait time as a huge potential for understanding what the current supply and demand for resources are. And when you think about the decision process that has to go into this, it makes for a lovely figure, a wonderful figure. But anyone who knows how clinical practice works in hospitals, this decision flow looks great but is probably not the way that it actually works in general clinical practice. But basically the idea is that patient is in need of a hospital bed and the patient needs either an ICU bed, a telemetry monitoring bed, or a general floor bed. And within that strata they either need to go to the general internal medicine floor, the cardiology floor, the renal service, the infectious diseases floor, different groups within those general strata. So the decision needs to be made of who is the patient, what type of bed do they need, what service do they need, and what level of care. Once you have that information, you can then start looking through what is the current availability, what beds have a patient who is currently flagged for discharge, what beds are currently empty but are dirty and need to be cleaned, and what beds are currently clean and available to be occupied. And within each of those different levels triggers a different bit of information. If there's a bed that needs to be cleaned, can we reach out to our custodial staff to ask that bed to be cleaned? If the patient is pending discharge but they haven't quite finished the paperwork, can we reach out to the care team to try to expedite the discharge so that we can optimize throughput? And the challenge here is that all of this information doesn't exist in one place. As a matter of fact, the nurse coordinator in our emergency department, when trying to determine who the patients are currently waiting for beds and what beds are currently available, would frequently have to download multiple different reports and would often print them out three different pieces of paper. They would go through with a highlighter to identify the current patients who are waiting for beds, the current priority order of those patients, and also what different beds are currently available or pending availability to try to assign the patients in their general priority to those beds that either are available or are becoming available. And as you can imagine, printing that out, highlighting, looking at the different reports, it could easily be a 30-minute process of trying to assign patients to beds, at which point in a very busy emergency department, the demand may have completely shifted and that data might actually be incredibly old, whereby creating another barrier for getting patients into the appropriate bed. And just as an example, I don't expect anyone to read through each of these reports. These are the kinds of reports that would be generated to try to help somebody in the process of assigning patients to different beds. A very cumbersome process, certainly not one that's consistent with a hospital in the EHR era. But going through that, even in the EHR era where we theoretically have one electronic health record system, we still have multiple silos of data that come in and feed that electronic health record system that require us to go back and tap into different data sources. So the idea of automating the process, generating a report that would facilitate throughput is definitely possible, but it still requires pulling data from multiple different clinical information systems from our admission discharge transfer system, from our electronic health record system, from our admission data sources so that we can, and from our environmental health services, our custodial staff, so that we can understand the current demand and the current supply of resources. It's possible, but it requires pulling from these multiple different sources. And I don't want to go into the complex data architecture, but I'm certainly happy to answer questions about that. But the point is that we need to pull relevant bits of information regarding the current demand and the current supply and the anticipated supply of resources. And what that looks like, we call an ETL process an extraction, transformation, and loading process. So being able to extract information from multiple different data sources, applying an algorithm that analyzes the data that are pulled from those multiple resources, and loading it into a single user-friendly platform that can be accessed by multiple individuals to determine what the current needs are to support decision-making processes. If that wasn't enough, that we wanted data from all of these different sources, we wanted it analyzed and we wanted it put into a platform. We wanted to collect this information in real-time or as close to real-time as possible. As I mentioned in a previous slide, as a nurse was trying to pull these data from multiple different sources, by the point that you've extracted the data that you need, the data could be 30 minutes old, and your entire situation could have changed within the emergency department in that time. So what we want are really all of these data to be refreshed in real-time to give us accurate situational awareness at the point that we need to make these decisions. And we want it to be in a format that clinical care teams can look at if they're at their desktop in the unit, if they are looking at it on their phone or their tablet, or if they're looking at it from home so that they can help facilitate coordination through the throughput process. And what we decided, what we specced out of this is that we wanted a dashboard that would provide real-time information regarding the demand for beds and what the current supply of beds are. We wanted to be able to predict what the current availability is, what the anticipated availability is based on those different measures. If a patient was flagged for discharge, that the bed would become available in the near future. If the bed was dirty and just needed to be cleaned. And if the bed was clean and available for occupancy. And we also wanted to allow the different stakeholders in the process, people within the Department of Medicine and people within the Department of Emergency Medicine, to be able to access this information and to be notified if there are patients who have been waiting for 24 hours or more so that it would prompt appropriate care teams to go back and assess the patient and make sure that those patients are getting the best care that we can provide while they're waiting for a bed to become available. We wanted to pull in all of the relevant stakeholders to ensure that we're closing the quality gap in the care that we're providing. So rather than describing this, let me give you an example of what this dashboard actually looks like. So this is an example of what we initially built pooling data in from multiple different sources. This gave us an indicator of what beds were currently available within the Department of Medicine. And that was if they were a general floor bed, if they were an ICU bed, or if they were a telemetry monitoring bed. It gave us information about the beds where patients were pending discharge, where beds were dirty and waiting to be cleaned, or beds that were empty, clean and ready for occupancy. It also gave us information about which patients in the emergency department were waiting for a bed to become available, which type of bed those patients were waiting for, and an indicator of severity to help us prioritize which patients needed to get into beds fastest. We also had ongoing monitoring of what the current average boarding hours were so that we could benchmark performance. The number of patients who were waiting more than 24 hours so that we would have an opportunity to go and engage those patients in further care. And also giving us an estimate of prediction of what beds would become available in the Department of Medicine over the course of the next 12 hours, where patients who might not have been pending discharge at that point in time, but other indicators suggested that they might be discharged in the near future. And we built this so that it would be available for nurse coordinators in the Department of Medicine and also emergency medicine. And we built it with real-time data. And when I say real-time, I have to apologize. It refreshes every 15 seconds, but that's about as close to real-time in healthcare as we get on most things. And just to show you how this affected our boarding hours, we implemented this in a staged rollout fashion. We looked at this for a period of three months before we implemented this. We implemented this in five of the floors for the Department of Medicine, and then we implemented it on all floors in the Department of Medicine, allowing us to assess any barriers and facilitators for implementation and try to improve the quality of the information that we were providing. And what we found is that the mean boarding hours before we implemented this were about 11, which is certainly not good. After a short period of partial implementation, we significantly reduced the mean boarding hours to about eight, and after we implemented this across the entire Department, we shrunk boarding hours to under five, under five hours. What we did see after implementing this was a slight increase in occupancy, so the demand for inpatient care services increased from about 80% to 87%, and then remained fairly constant for some period of time. And we saw a slight increase in the mean boarding hours to about 5.8 hours, but still significantly below where we were at baseline, simply by making the appropriate information, the best information available to the frontline decision makers as possible. As I mentioned, we saw a significant association between occupancy and boarding hours, which absolutely makes sense. As more beds are occupied, naturally boarding hours will increase because there isn't really the supply for the beds necessary. And the only way, unfortunately, to overcome that is to expand a hospital and add more beds, more capacity to the hospital, but it's a great predictor of what bed availability is going to be, and certainly something that we're using in our predictive analytics to determine the likelihood that beds are will become available. Just to look at this over time, before we implemented this dashboard, the median ED boarding hours were number one significantly higher, but also with a large range of hours. You can see here in the first month that the range of the interquartile range of boarding hours range from about 4.5 hours all the way up to 16 hours. Over the course of time, as we implemented this dashboard, that range got substantially narrower. And you can see in the last month of data that I'm presenting here, our range of boarding hours went from about 2.5 to 4.9. So not only were the median boarding hours reduced, but the range of boarding hours were also significantly narrower, owning to the value of predicting bed availability. So speaking a bit about outcomes, what we saw is that the median boarding hours for internal medicine significantly decreased by 51%, simply by putting the appropriate information into the decision maker's hands, real-time information that facilitated rapid throughput. We saw an added benefit that the median length of hospitalization significantly decreased by 25%. Four days versus three days after full implementation of this dashboard. But the most interesting thing that we saw out of this, which is consistent with what has been published in other studies, is that inpatient mortality significantly decreased by 57%, looking at the pre-implementation compared with the full implementation period. We went from 3.5% inpatient mortality, stepwise down, we saw that stepwise reduction in boarding hours. We saw a stepwise reduction from 3.5% mortality to 2.1% inpatient mortality to 1.5% inpatient mortality. And I absolutely want to give credit or not overstate that there may have been other interventions within the department that have helped reduce inpatient mortality. However, this finding is consistent with what has been published in the literature associating long boarding hours with inpatient mortality. And this is the first study that I've come across where a prospective intervention has decreased boarding hours and has been associated with a decrease in inpatient mortality. So while there may be other things that were happening and helped this, and I certainly want to give credit for it, it is a very, very interesting finding that is consistent with the literature that as you reduce boarding hours, you also reduce inpatient mortality. We also had the opportunity to see a number of unintended benefits of the reduction in boarding hours. By reducing boarding hours, we were able to focus on other potential drivers that weren't simply informational barriers to timely throughput. When we did some qualitative interviews and specifically observations on the floors to look at lingering issues for why beds were delayed, so patients who were no longer occupying beds, the beds were flagged as dirty, we would see that these beds would remain listed as dirty for two, three, four hours. And we started to do some follow-up investigations to look at what the issues regarding this were. And what we found is that there were a number of cases where occupancy on floors was relatively high and nurses would ask the custodial staff, the environmental health services staff to hold off on cleaning the beds. That they know that as soon as the bed is marked as clean and ready for occupancy, they're going to send another patient up, and they're still catching up on patient care issues, documentation, things that needed to be addressed for ongoing patient care, which gave us an opportunity to address certain staffing issues to make sure that we had more nurses and staff on the floors when we had particularly high occupancy to help improve throughput. Things that we never would have looked into if we hadn't started driving down boarding hours and looking at other secondary drivers of why there might be delays in timely throughput. The other thing that we were able to do as a result of this, which I touched on a bit, is that by making this data available and by analyzing larger amounts of these data to look at the full processes of throughput, looking at patients who were pending discharge, looking at how long beds were flagged as dirty, how long it took to get patients into beds that were marked as clean and ready for occupancy, we were able to begin predicting bed availability before patients were even flagged for discharge so that we could begin looking at potential pre-assignment to beds, reducing the amount of time that it took to get patients assigned to beds within the hospital such that we could actually have a patient who was waiting for a bed in the emergency department and they would be assigned to a bed that is already occupied but that we can accurately predict will become available within the next 12 or fewer hours. So a few summaries, we were very excited to see the associated reduction in inpatient mortality. We saw certainly the reduction in ED boarding time but secondarily we saw a reduction in the overall length of stay and we also saw an enhanced level of communication between different care teams, attendings, nurses, social workers, case managers, administrators. The biggest challenge as a clinical informatician that I see in healthcare environments is that we generate data, we generate information and it so rarely makes it back to the frontline providers in a meaningful way. We oftentimes throw out reports and shake our fingers and pound desks and are angry about unsatisfying numbers or bad outcomes but we very rarely empower on a department or hospital level people with that information to change practice and really make a difference and then enable them to see the trending difference over time to see what happens before and after we provide this information and we see an enormous level of satisfaction among staff across the board in providing meaningful information back to frontline providers. And there are many people I need to thank in this. This is such a team effort. I absolutely have to thank Haydel Rupani who over the course of this project was based in both internal medicine and emergency medicine and is really the guru behind the database administration to pull all of the information from different data sources together to make this report happen. Sanjay Desai who is the residency program director in internal medicine at Hopkins and was the executive champion for this project as well as multiple postdoctoral fellows who helped with the analysis of the data doing observations on the floor to try to help improve the quality of the data and try to identify other areas where we could potentially improve throughput practice. And with that I've seen a number of questions that have come in throughout the course of this time. I'm hoping that I've given enough time to address the questions that have come in. And thank you all very much for joining today. It's a very important topic and an interesting topic to me. Brandon that was phenomenal. Thank you so much for leading us through the journey that you all have been on at Johns Hopkins to summarize. I mean the median boarding time, length of hospitalization and mortality decreases that you saw were remarkable. And so I hope everyone who is on the call today and listening to this later on our recording can take some or all of this back to their hospitals to see what they can do. And certainly I know that you highlighted some of the confounding factors and other initiatives that your team was also working on that could have led to the inpatient mortality, but certainly is significant. So with that we would love to open it up to questions and answers. We also identified that there is a feature that you can raise your hand. So if you hover over your name under the participant list if you're on the web. Again if you're just dialing by phone you won't be able to participate in this and if you're on a mobile device you won't be able to but there's a little hand and if you quit on that then I should be able to scroll through and unmute just you so that we don't get a crazy amount of people who are trying to talk over one another. So let me pull up the chat box at least to start and we'll take the first question. So the first question is from Christine and I apologize if I pronounced your last name. So the question is where outcomes strategized across different patient severity in ED levels. We actually didn't. So I will say that the number of instances of mortality inpatient mortality in this case were relatively low even at baseline that we wouldn't have had sufficient power to look at different levels of acuity. But it's a great question and I think certainly one that if we could expand this kind of intervention across multiple different hospitals would be amazing to look at because I think then you can start to also get to the specific drivers of inpatient mortality. This is really such an association and the exact causes of mortality in the population we didn't quite get into but I think that if we can start to get a large enough population that we can drill down on acuity and also the causes of inpatient mortality specifically that would really help to inform future interventions as well. Great. Thanks Brandon. The second and third question are together from Danielle Geyer. First is what system was used to generate the dashboard and second is can you please restate who the key people were who used the dashboard? We generated the dashboard using Tableau which is a wonderful visualization tool. The example that I showed was an early version of it. Tableau has certainly become a lot fancier over the course of time but we actually have found that it works very well to receive data from multiple different sources. The primary users of the dashboard were the nurse coordinators in emergency medicine and in the Department of Medicine for this particular study because those were the individuals who were really focused on the throughput of patients from the emergency department to the Department of Medicine. However, in the spirit of transparency we wanted to make the data available to nurses on the floor in the ICU on the telemetry monitoring floors or intermediate care units so that even nurses who weren't necessarily engaged in the planning and coordination of throughput could still see what might be coming down the pike as far as patients waiting in the emergency department. Really a transparent process. Great. Thank you. The next question comes from Susan Pfeiffer. She says, I see how boarding hours and time to admission were decreased using the dashboard. How was length of stay reduced or did you mean ED length of stay, not inpatient length of stay? We looked at inpatient length of stay in that case and what we were really looking at was any association between inpatient length of stay associated with boarding hours. I think that it's such an important question and one that I wouldn't say that we dealt too far into in this particular case. There could be multiple interventions that are attempting to reduce inpatient length of stay. However, this was purely associated with as ED boarding hours decreased, we also saw a decrease in inpatient length of stay. One of the factors that might be associated with it is if we can get patients into the beds faster, into specialty care beds, appropriate care beds faster, the sooner we get appropriate treatment started, it may have an effect on how quickly patients are then eligible to be discharged. Very much as we saw with potentially that also reduces inpatient mortality. Purely inpatient. Yeah, great. Next question is from Anka Sarbu. She asks, how do you measure the enhancement in communication? We actually surveyed nurses to find out what their level of comfort was, particularly with nurse coordinators, to assess their level of comfort with the process before and after implementation. I think that as we transition, the biggest difference that we noticed was really transitioning from I have to print out three different reports on paper and find a quiet space and highlight and prioritize and assign verses. I can log into this dashboard and all of the information is presented to me in real time. And I can just pick up the phone and call different units and say, hey, I need a sensation to to your floor. It was a survey approach, but largely one that was affected just by making the data available. Thanks, Brandon. Next question is from Greg Swarovski. He says, thank you for your presentation. Did you include your BH patients in your study for the boarding time? BH. I was hoping you knew what that was. Greg, maybe you can type in what BH stands for and we can circle back on your question. Oh, behavioral health. No, we actually didn't. We only included department specifically patients who were admitted to the Department of Medicine. We didn't include any other departments in this particular analysis. But it's a great question. My guess is that the same issues with supply of beds, it will exist there as well. Yeah, great. Next question is from Trish Cruz. She says, it looks like you used Altrix as a data blender and Tableau for displays accurate. So actually, we used Tableau for the data display. The data that we pulled out, we actually ended up using Access from an aggregator standpoint. Oh, excuse me. I'm sorry, SQL Server. SQL Server. Perfect. Next question is from Ken. Did you look at variation in boarding times by hour of day and day of week? That's a great question. I actually haven't looked into that yet, but I think it's definitely an important one. We've certainly seen a lot of variation on other aspects of clinical care based on day of the week and especially Friday, Saturday, Sunday, and evenings. But I haven't looked into that yet. That's a fantastic question, though. Amazing. Well, I think that's the last question. If anyone else has any last minute questions, please do send through now. I'm going to move on to the last slide that I have, which is regarding save the dates for the patient safety movement. Next, save the date is just that we'll be sending out our patient safety newsletter on July 1st. This is our opportunity to update you on different blog articles and spotlights that we do on hospitals that have made commitments and are doing great work. So that will be released on July 1st. If you don't get our newsletter, you can go to our website and click follow our progress to sign up for our emails. The second is our mid-year planning meeting, which is our one of two events that we host every year. It's held on Tuesday, September 17th, which now we can formally announce World Patient Safety Day based on the WHA resolution on patient safety passing at the end of May. It'll be held in Irvine, California, co-convened by UCI Health. You can request your invitation through our website. It is free to attend. And then our next quarterly webinar will be September 4th. The topic is related to our actionable patient safety solution number 10, which is the systematic prevention and resuscitation of in-hospital cardiac arrest. But Dr. Dan Davis will be talking about the advanced resuscitation training or art system of care as a potential scaffolding for reducing preventable deaths. So with that, I haven't seen any other questions come through. We really appreciate you taking time out of your day to joining us on our webinar and we look forward to participating in our other events going forward. Again, thank you so much, Brandon, for the powerful presentation and data that you presented today. Thank you for the honor of joining. Take care. Have a great day, everyone.