Thanks so much Olga. So Olga asked me to talk about common mistakes that are made in quality in performance improvement and they're looking for an expert in someone who has made a lot of mistakes so they came right to me so I'm happy to be that expert for you to share all the mistakes we've made and hopefully a few things that we've learned from them. Paul Batalden, one of the experts in quality improvement in health care has said all improvement requires change but not all change is an improvement. So when you improve something you're going to have to change things at least when you try to improve something but your attempts to make things better don't always go the way that you want them to go. So when we start on an improvement project we often see these blue skies and these clear water and we think things are going to be so much better. We're idealistic, we're enthusiastic, we're energetic, we've got a team, we're ready to go and not uncommonly unfortunately this is what can happen if we're not careful and so this is obviously very discouraging when this does happen but hopefully as this does happen to many of us either we can learn from each other or learn from ourselves and we start to develop an understanding, a better understanding of how to avoid the treacherous waters, how to navigate those difficult waters, what to watch for as we're implementing or we're improving a process. So that's what we're going to talk about today, how do we navigate these waters of change and not end up on the shoals of quality improvement but really get to where we want to go. So first of all we're going to start with a conversation about the difference between implementation and improvement. We'll talk for a minute later about why that's important but first of all an implementation project the way I see it is it's a the process of putting a decision or plan into effect. It's execution whereas performance improvement is a process of modifying a process or procedure in order to better align the actual outcomes with the desired outcomes. In other words when we're saying implementation we're going to put in place a solution and this solution had better be well tested and validated if you're going to put it in place because otherwise you could have real problems if it's not well tested. On the other hand performance improvement says we say we know where we want to go, we know where we are, we don't know exactly how we're going to get there but we've got a process for learning as we go along and we're going to follow that process and we'll probably go on some dead ends in the meantime and we're willing to entertain a couple of different avenues a couple of different possible solutions in order to get there but the focus is somehow we're going to get there we're not necessarily fixed on the solution. So implementation we know the solution works well, improvement there's more uncertainty we're going to develop the solutions as we go along. When I was working at the Agency for Healthcare Research and Quality as an intern a few years ago I was working with Dr. Greg Meyer and I asked him what should what should I read in order to learn about quality improvement and also something about how how to work in the federal government and he said well you should read Sun Tzu's Art of War. I thought well that was an unusual recommendation but I started reading it and you can go online it's not very long and it actually offers some concepts I mean I guess if you want to wage war it helps with that but it also talks about the concept of strategy. So Sun Tzu said the art of war is of vital importance to the state it is a matter of life and death a road either to safety or to ruin hence it is a subject of inquiry which can on no account be neglected. The art of war then is governed by five constant factors to be taken into account in one's deliberations when seeking to determine the conditions in obtaining the field. These are the moral law, the heaven, earth, the commander, method and discipline. These five heads should be familiar to every general he who knows them will be victorious he who knows them not will fail. So what he's talking about on the moral law how do you get people to do what you want them to do the heaven is talking about the weather and the climate and night and day earth he's talking about the terrain the commander is the person in charge and then the method and discipline that you use to move your armies and your strategy. So what this really is about is strategy how do you use strategy strategic in waging war. Well so I'm going to say what is the art of quality improvement and if I were to write that book I would probably start with something like this the art of quality improvement is governed by five constant factors to be taken into account when seeking to determine whether a QI project will be successful. These are leadership, method, resources, culture and execution. These five factors should be familiar to every QI leader he or she who knows them will be successful he or she who knows them not will fail. So this is what we're going to talk about these five factors. So then the other thing one of the other things that Sun Tzu said was now the general who wins a battle makes many calculations in his temple error the battle is fought. The general who loses a battle makes but few calculations beforehand. Thus do many calculations lead to victory and few calculations lead to defeat. How much more no calculation at all is by attention to this point that I can foresee who is likely to win or lose. In other words he can look at the general and look at how they're making calculations going into the war before it even happens that he can foresee who's going to win or lose. I would say it's similar in quality improvement. The successful quality improvement leader makes many calculations and preparations before starting the project. The unsuccessful QI leader makes but few calculations or preparations beforehand. Thus too many calculations lead to success a few calculations to failure how much more no calculation at all. By reviewing their preparations I can foresee whether a QI team is likely to succeed or fail. So if you've made no preparations you're just going to start into it you're probably going to fail. Now that that's okay you can learn from that failure and then move on but if you don't want to fail then it would be helpful to take some of these things into account. The most common mistake I generally see in quality improvement is what I would consider a rookie mistake which is a mistake which is often egregious made due to a lack of experience. So there are tools and there are there's some basic approaches that we can use to overcome or at least prevent some of these rookie mistakes. Can someone tell me what this is? Okay I heard someone say it's a football diagram. Good. Can someone tell me more specifically what this is? All right I wouldn't be able to but I looked it up on the internet and so I'll tell you what what the internet says it means and this is from the Miami Dolphins website apparently the fan site. So this is a configuration for football teams so this is the line of scrimmage. Line of scrimmage on the bottom. Sorry. Line of scrimmage here I gave it away and then the quarterback you can see on the bottom and then the defense is in the X's so the yellow X's is the defense right so this is a configuration so the quarterback can sit down and look at the defense how the defense is lined up and say aha I see how they're lined up and that I'm going to change my strategy accordingly. Now apparently this is a cover one defense so this means that there are no deep safeties and the free safety is moved up in other words you're probably gonna blitz and so they have about three and a half seconds to get this this football off so in other words the point is to us to most people if you're not familiar with this this makes no sense but when you have a little familiarity with it it's something you can read so can you read that's what it means can you read the defense and can you act accordingly so that's what that's our our goal is can we read the signs or another way to look at it can we see the lay of the land as we're going to do a QI project and so the lay of the land again leadership method resources culture and execution so a couple of definitions quality I generally defined as consistently doing something well and you can define what that something is and what that how consistently you want to be domains of clinical performance we have often heard this clinical or technical quality service quality or efficiency and this is essentially the value equation and then finally performance improvement then is either consistently doing something better or more consistently doing something well W Edwards Deming talked about what it takes to understand improvement and in general there are two areas he called this profound knowledge first of all you need to know something about the subject you're trying to improve so in this case radiology and then the appreciation of processes and systems understanding of performance measurement data and variation the theory of developing practical practical knowledge so that's basically running scientific tests in the local environment and the psychology of change all right so you have to understand like how these processes work together understand how to read the data understand how to test and gain knowledge and then understand people and how people think and how they're going to react and if you can understand that then you can do improvement so let's dive into a few cases case number one you implement a change in a workflow that some of the technologists do not like when one of the technologists goes to complain to a manager the manager responds by saying well this is the first time I've heard of this it sounds silly let me go talk to the radiologist and sort this out I don't know if anyone has ever come across this before if this has happened what what did I do wrong in this case well I would say that we didn't understand and appreciate and utilize the role of leadership in improvement so the organizational leader is the person who has accountability and authority over the performance of a unit and so if you're going to change the want to try to change the performance of a unit then you'd better talk to the person who has accountability and authority over that often we don't have a great appreciation when we're physicians of how the accountability and organization structure works on the hospital side but yet that's the thing we're trying to improve on the hospital side there tends to be a more hierarchical organization with at the bottom technologists and then a lead tech and then supervisors and then managers and clinical directors and administrative directors and vice president there so there's a well-respected hierarchy so when we go into it we don't understand it that can create problems so we generally go we go and talk to the the person in charge and their supervisor before we embark on any improvement a radiology practice of course tends to be a little different physicians tend to be more first among equals but still you have a person who is generally the person who is has some accountability over that area so again if you're going to work in that area then you need to make sure that you're on on the same page as the leader in general the organizational leaders role in change improvement specifically they should help set the vision model the behavior that you're looking for empower the right people address skeptics resistors and disruptors so somebody's not going to like this whatever the changes there's someone who's not going to like it and so when they go to to the leader as they inevitably will then that leader needs to be on the same page and ready for that they help mobilize resources and they enable execution by removing barriers and providing guidance now leaders cannot single-handedly affect change but they can single-handedly undermine change efforts so again got to have them on the same page what is not the organizational leaders role in improvement it's to solve it's not to solve the problem so the organizational leader in enables the team the frontline team to help solve the problem but they themselves do not solve the problem okay case number two you discussed with your technologist manner the issue of timely verification of radiographs which has been a problem in your department for years the next day you see your manager and he informs you don't worry it's all taken care of I had a talk with technologists and they now understand the expectations I don't know if this sounds familiar but if you have this big gnarly hairy problem and you go and ask somebody to help solve it and then you come back later that day it's all solved no worries all taken care of so what did I do wrong here well the they didn't understand the method we didn't we weren't on the same page about the method for improvement so there is a method for improvement and we're so the analogy is that we often have it's the scientific method we understand what this method is the purpose of this method is to answer a question and it has a specific approach whereas the the analogy in in improvement is more like the engineering design process and the purpose of that process is to solve a problem that's more what improvement is about is how to solve a problem rather than answer a question the scientific method you ask a question you do your research you construct a hypothesis you test it you make sure it works you analyze your data and you report it in the engineering design process you start with the problem so it's really important to clarify what that problem is then you do your background research you specify your requirements and your constraints then you develop a bunch of prototype solutions and many of them are not going to work and so you test them and you refine them and keep testing and refining so this is the engineering design process and it's a bit different than the scientific method and once you have one that works and gets you to where you want to go then that's something that you can fully implement excuse me so this is the PDSA process that's embedded the scientific method that is embedded into the engineering design process which is akin to the improvement process the improvement method we use and Paul knows you talked this morning about many methods or hundreds of methods but the most common ones are either lean from Toyota or Six Sigma and we also use a combination of that so we use the a3 and a3 is a Toyota developed document it's a single document a project management document and then on that a3 we use basically the Six Sigma process so first of all we make sure we get the right people in the right roles in place we define the problem and state why it's important we measure the performance we analyze the factors causing the problems we improve performance through structured problem solving methods and then we have some type of controls to ensure that performance is sustained this is the a3 that we use at Stanford we have a fairly structured method because we find that when we're all on the same page on how we solve the problem how we accomplish improvement then we can focus more on the actual improvement itself rather than having to fight over the process of process improvement also there are many other methods out there so the most important thing is that you're all on the same page who you're working with that you have a method that you use at your institution if your greater institution has a method then I recommend that you adopt that or at least tap into that and use that as a basis so these are some of the places you can go the improvement guide is kind of the model for model for improvement for the IHI which is the Institute for Healthcare Improvement you can go online to get their their resources all right case number three you gather together a team to work on improvement improving exam completion time meeting attendance is sparse because people are unable to break away from work the team is not able to extract data from the risks the team has no background in improvement and lacks guidance so they're spinning their wheels so what did we do wrong here well I would say we didn't plan to provide adequate resources so there's this book that came out in 1980s called quality is free by Philip Crosby it's a great great little read and not not much has changed at least not in terms of the principles but the point of the book is not that quality is free the point is that it's it is a positive return on investment but not that it doesn't have it it's not a requires no investment so quality does require an investment I think often we think that's the thing that should be happening in the first place so why would we spend money on it well if we want to change the organization that does take time and effort and energy which translates in many cases at least to some dollars so what resources generally does it require to have a successful improvement project well-protected time education and training access to data coaching access to the experts possible investments in equipment software tools and facility remodel perhaps etc so one question to ask of your leadership up front is what resources are you willing to devote to getting this improvement and if it comes down to it we have to add personnel we have to have equipment are you willing to make that commitment and if the answer is absolutely no way well then at least you know your constraints and if they say well only if you there's really no other way then you know that your job is to really try but if you have to you can go back and ask for more resources case number four you and your QI team develop a solution to the problem of delayed start times for CT and that is that ED nurses simply need to call the CT tech ten minutes before the exam when this is shared at a meeting with the ED nurses it's met with anger and hostility so what went wrong I don't know if it is anyone has ever experienced this where we decide you know we've got a great solution we're gonna have somebody else do something all right and then we go and walk in and we encounter this hornet's nest of anger and hostility right well so I would say in this case we didn't understand and appreciate the concept of organizational culture so culture often is referred to its shared norms and values that drive behavior and especially interactions between individuals and units within an organization and this describes what's acceptable versus unacceptable and it can be it generally varies from one organization to another it varies within subunits within the organization sometimes it varies even within sub units of those subunits so it's really important to understand what are those interactions what are the rules of engagement so to speak it often it also reflects feelings that individuals have towards each other and towards the organization and the organizational units so some questions to ask are who's going to be impacted by the change and is everyone who will be impacted included in the discussion and if not you got to get them in the discussion how big is the impact how will people feel about the changes in other words is this going to be a big lift this is going to be a big change you're going to ask of them if so then you're gonna have to it's gonna require a major change effort are people used to change so will they be openly receptive or are they tired of it is there change fatigue so you should appreciate this before walking into it how will people react to exposing their problems sometimes as it turns out people don't appreciate when you point out all their flaws as well meaning as you might be so if you can be strategic and think and anticipate how is that likely to be received and what are the interpersonal dynamics within there so often there are silos between for example the emergency department and CT or other areas of radiology so we need to understand to be able to navigate that case number five you put together a team and start on a project with great enthusiasm after the first meeting a month goes by with nothing happening a flurry of emails indicates is not clear who's supposed to do what some team members try interventions without others knowing frustration builds and the project stalls so what did you do wrong what was our problem here well I kind of gave it away there's so there's my hair whenever I give a talk about errors there's always at least one error in the slide so hopefully that's that's the only one so the problem here is execution and execution is basically the art and science of getting stuff done can now you know what you need to do can you break it down and do it so ways to help with this first of all choose your projects wisely choose your project roles wisely so we we are very specific about what the roles are in a project so the sponsor their job is to commission the project assemble the team follow their progress and remove barriers so they get everyone together and they make sure they're going along if they need some help they're checking in frequently and they provide the help that they need what they don't do is solve the problem the project leader or manager keeps the team on task it clarifies the assignments and they follow up so larger projects huge projects have a project man a dedicated project manager like your PACS implementation, you probably had a project manager. On improvement projects, usually the project leader serves the role as both the leader and the manager. A project coach, so we align a project coach with most every project. They're there to provide guidance on the methods, help with the data, and connect with the sponsors as needed. It's really helpful to have a coach. And it doesn't require a ton of time necessarily, but they're just to kind of check in when they get stuck, they can kind of help them with the methods. And the project participant is, now they're the ones who contribute the ideas, they fulfill assignments, they're the ones who do the heavy lifting as a team, they work together as a team. So there's a whole science to project management. And so this is the PMBOK, as it turns out, the Project Management Body of Knowledge. I recommend you do not read this book. But you can learn just a few simple things about project management. It's not rocket science. It's mainly about recognizing what are the tasks that need to be done, deciding who's going to do it, and then follow up to see if they did it. Okay, so those are the major, what I would say, factors of success. Leadership, method, resources, culture, and execution. So I'd like to stop with just a few final pearls that we often teach in our performance improvement courses. There is always more work to do than there is bandwidth to do it. So if you can't do all the work, all the improvement work that there is to do, that's okay. It's better to do what you're going to do, to do it well, than to try to do everything and not do any of it really well. So we have a parking lot, and we will put improvement projects in the parking lot. And that can be hard for people. They say, what do you mean you're not going to solve my problem yesterday? And we say, well, this is a big, hairy problem that's going to require a lot of collaboration. We want to do it well, and so we'll hold on. And when we have the resources, we'll do it well. It's really important to learn how to fail. So one of the things we reemphasize in our program is we fail fast, fail small, and fail friendly. In other words, you're going to fail. Your first draft is not going to be your final draft. So it's just, you're going to fail. So get used to that, and just accept that right now. But the question is, how quickly can you learn from that? So when you fail, make it happen quickly. Do it in a place where you don't bring the whole house down. And do it among people who are going to be forgiving as you make your mistakes. We solve our own problems. So in other words, we don't solve other people's problems. If there's a problem with the emergency department, then the emergency department team has to be the one at least working with us to solve that problem. Go to Gemba. So Gemba is a Japanese term for the real place, actually is technically what it means, which means the place where the work is done. So the first thing you should do, and do it frequently, is to go observe the actual work. Rather than sitting in a boardroom and talking about what's happening, go actually see and observe it. Do not jump to solutions. There's a reason why the process starts with defining the problem and analyzing it and really looking at it. You learn to appreciate a lot of nuances, and then you come up with a lot better solutions. We have this analogy of the baby, so I'm a pediatric radiologist, so I kind of have to bring a child analogy into it. So the way we think of it is, when there's a problem with the process, we often will take this process, just like I or a physician will, temporarily take the process and care for it for a while, make some changes, and then we hand it back. So we have to be prepared to hand it back as we are making the changes, and the person who's processed it has to be able to let go for a little while while we make some changes. So there's an understanding there, and those handoffs have to occur. If you're having problems getting started, then just stop talking and go try something. But try something small and be prepared to fail, and that's okay because you'll learn from it, and then you'll get back on the right track. And finally, here are some ideas to do when leaders are not supportive. That's what people most commonly ask me, what do you do if my leader is not supportive? Well, here are some ideas. Fly fishing, spending more time with your children or grandchildren, maybe hike the Appalachian Trail. What I do not recommend when leaders are not supportive is to do the improvement project. You've got to have your leaders on board. So if you can see the field, if you can learn to see, and see the lay of the land, and just with a few strategic initiatives before you go into it, a little strategic thinking, you can be successful in your improvement projects. Thank you. Well, good afternoon. My name's Allie Kazerouni, and I'm going to talk about uncommon champions of value, or what I hope you'll understand by the end of my brief discussion is common champions of value, and it really is all of the people, and all of the people you touch who are in this room. The right thing for the right patient, and as we think about a value-centered environment, the right time in the right way for the right price. We talk about right thing for right patient, right time a lot, but when we think about the value proposition, and we think about patient and family-centered care, bringing in the way, the experience, and paying attention to price and price points, putting all that together allows us to become common champions of value. So I'm going to start with a little bit of imagine, I'll talk about the construct of how this applies to the evolution of healthcare, and where we are with value, because I think we're all struggling with what we do, what our role is, where it is in this complex evolving healthcare environment, how we improve quality, drive to value, step back and look at radiology operation, figure out how can we get started, and then fundamentally the commonality of it all. So imagine with me, you're a patient. Pretty much everybody in this room has either been a patient, is a patient, or has a friend and family member who's a patient. And if you try and step out of your healthcare shell, which is where we have extreme bias, because we get biased by the fact that we do this every day for a living, we get sensitized to the fact that there might be bad news. But when you think about it in your own environment, in your own family and patient-centered way, it becomes very different. Imagine you don't know much about healthcare at all. Hard for all of us, because we do. Imagine you don't know anything about healthcare, and you're scared that you have cancer, cancer's getting worse, you've got an aneurysm, Alzheimer's, cardiovascular disease, all those things that fundamentally scare people when they think about what that might mean for them or their family. And imagine you have a family, broadly defined family, whether it's your close family, your friends and relatives, whoever you define as family, depending on you. So you're afraid, you're uncertain, you're concerned, you have people depending on you. What would be your state of mind? And imagine all the people who come through your healthcare delivery door every day who are just like this, who are leaving their workplace or leaving their families or their homes to come and get tested to find out they have bad news or bad news has become worse news and what that means for them and their families. And that's the people who come to seek out our care every day. How would this impact you in all the myriad things that we all think we have so much time or so much little time to do? We all feel busy, overworked, overburdened, fast-paced work paces, life schedules. How would this impact you if you stopped and thought about it? How would you want to interact with your healthcare system, your radiology department, and the providers? Would you want to be 50th in queue in a call center to schedule a radiology test? Would you want to sit in a cold, pasty, white waiting room with hard chairs that you couldn't move or get in a comfortable position because you're feeling sick and have nausea from your chemotherapy and you can't get comfortable when you're waiting for your follow-up CT scan? What would you and your family want in your healthcare? What would you want to know about what's happening to you in radiology? This is, of course, a group of radiologists and those aligned with radiology healthcare. And now imagine how healthcare has changed over the course of the careers of those of us who are here in this room. From students to the most experienced physician in the group or administrator here today, we span a very diverse time of healthcare and it's accelerating in the way this page is changing. And we're yet trying to find the value and drive quality improvement, not quite sure where we're actually going. So we have started from a very chaotic healthcare environment. We've gone through a period of movement towards standardization and now we're drilling towards personalized, whether it's a personalized precision medicine and data and analytics or personalized patient family-centered engaged care in both directions, in both the technical and the patient side of what we're doing. So if you think about the healthcare before I became a physician, let's just say, I won't say how long ago that was, it was somewhat chaotic. It was pretty much the days of individual physicians, individual practitioners, doctor often on a pedestal, doctor as God, doctors influenced by the last case of whatever disease they saw and how the testing did or did not work. Not a lot of data or evidence center practice. It was an era of doctor knows best, of people believing that patients wanted to be told what to do and didn't want to ask questions, just doctor said this, I'm going to do that, that's what the doctor said. Of solo practitioners to small groups of pretty poor communication and documentation, a doctor's shorthand scroll in an office record that nobody else could read or see, no standardization to the process, poor data collection, poor dissemination, and that era, that chaotic era, it spans all the way back to Aristotle and through the era of surgical amphitheaters and teaching, pretty much all the way up through the Marcus Welby era in the U.S. For those of you who've seen Marcus Welby, I caught a few episodes as a child. But it was really the era of the solo practitioner, the solo practice. And we've moved through this period of standardization where we are all part of systems of standards of care, complexity of information and care, patients with multiple physicians, physicians feeling we're losing our individuality and control from how we train to how we're practicing now. It's a fast moving at your fingertips, hard to keep up with, lots of data, patients as we heard earlier this morning, 86% have surfed the internet for information before they see a doctor, more and more informed than ever, more access to content. We have large scale expensive medical trials, data is shared through the internet and available to anybody who wants to grab a personal device. We've evolved many more hundreds if not thousands of practice standards and guidelines that we expect everybody to keep up with. And we have now organizational provider and external accountability from external parties looking at us and demanding excellence and value. And quality improvement is an important tool to help us adjust to the standardized world. But yet we're evolving very quickly again to what is standardized almost one size fits all for signs, symptoms, diseases to this personalized, precision based approach to medicine. In biology this is very real because we're a very data rich specialty, taking biomarkers and data quality metrics out of imaging tests, feeding that together with big data, if you know what this symbol is, it's for IBM Watson Healthcare, the Watson computer, pulling mass quantities of data together that you and I could never comprehend in our heads to try and make the right decision for the right patient at the right time. So beyond standardization and standards and guidelines of care. And then as that has become more personalized, the data itself, the medical care itself, the choices in medical care, you have a rapidly evolving patient and family centered care movement who now wants to be more involved in their care. In the days of Marcus Welby I'm getting told what I'm supposed to do. In the days of standards and guidelines I'm being categorized as a patient with X need and one size fits all. And now it's really about what's right for me and my values and what's important to me at this point in my life when it comes to medical care. And we in radiologists are trying to figure out how to keep up with this evolution. And if we think about things like, whether it's Watson or other big data analytic initiatives, taking these mass quantities and data and delivering it to us, extracting information out of our tests like CT that we don't even regularly report, it feels a bit frightening like we're losing something as healthcare providers. Now take the radiology experience and how that's evolved along this evolution of healthcare and how we as radiologists have intersected with the patient's radiology experience. We have the gateway into radiology departments, the scheduling, the requesting of tests, the access to a schedule, getting on the plots, getting a slot in the schedule. There's some pre-procedure exam preparation, patient's check-in, reception. They have waiting areas, or as I like Dr. Rawson this morning, patient lounges. We don't want to make them preconceived that they're going to wait, they're lounging. We have the test or the intervention itself, the X-ray, the CT, the MRI, the ultrasound. We have post-procedure care in some cases. We have a check-out process back through clerical, reception, changing area. Results happen. Somehow they get out to somebody who's going to act on them. There's a form of communication. And of course, there's infrastructure things like billing to support us doing all those things and continuing to practice. And historically, in kind of 1.0 radiologists, the Marcus Welby kind of example, the radiologist has been involved in this part, the interpretation of the test or doing the procedure, and has been involved in this part to making sure we get paid so we can continue to run our practices and continue to operate as healthcare providers. But in general, especially in hospitals and bigger systems, a lot of these other things were left to other people. And physicians often provided little input, especially outside of private operations in the private sector of radiology, on how all this worked, kind of doctor in a dark reading room. And somebody else is going to do the quality improvement and all those other things because it's not in my space. I just want to sit down and read my tests and deliver what I believe is high-quality care and set those protocols and work with my techs and deliver a high-quality result, a dictation or report. And so there are metrics around these kinds of things, around the test intervention that is us radiologists, things that we can measure and look at. How fast we turn around our reports, peer review or quality overreads of our reports. We look at complication rates of our procedures or how our results impact downstream outcomes. We look at process or technical measures around the test, like how long a test took. We can look at radiation exposure. We can look at things on the billing end, like collection rates, making sure that we're paying attention to rules for billing and what we put in our result for billing compliance. But these are pretty straightforward or simple things to measure and improve on. And so in that era, our customers were primarily our frank physicians. The patients went to where the doctor told them to go, and we were kind of the doctor's doctor. And the patient, you know, kind of was over here. They came in and out, and we were over here in a dark room. So as we kind of move along and we're getting more engaged with our frank physicians directly, what do they want? And our patients and families, what do they want? We start to pay attention to our gateway. How well do we schedule? How well do the people who try and schedule this like our services? Do we survey them and ask them questions? How long are they on hold? How often did the call center drop their calls because they're waiting too long? And we start to think about more about the outgoing function of our results, our communications. And as we heard this morning, people talking about the patient portal and delivering results directly to patients. Do we communicate results to our referring providers? Do we call them, or does everything just go out electronically and use electronic checkboxes for communicating urgent and critical results? And so we start to look at other metrics as we start to get a little more involved on the front end and the back end, the gateway and the communication. We start to look at turnaround times, about scheduling appointments, access to appointments and availability, report turnaround times. Do we actually do critical and urgent results communications? We evolve. And in this era, we've become not just the doctor's doctor, but we actually talk more to our referring providers, and we start to talk more to our patients. So we're no longer just talking to doctors and just talking to patients. We've really started to engage in the patient and family-centered world, and with patient satisfaction scores driving, in many cases, reimbursement and being seen as a source of value, you have to pay attention to your patients and families and what they want. You don't have to do everything they say, but you do have to hear what the fundamental underpinnings of how they would like to have the patient care experience. And those concepts of patients and family-centered care that Kelly Parent, our patient advocate, and family-centered care representative who spoke this morning, talked about were things like respect and dignity for the people, information sharing, transparency of what it is to come into radiology and have a test, what the experience is going to be like, and how they're going to get their information and results, embracing their participation in the process and what choices they might have. As Jim Ross has showed this morning, the color of the lighting in the pediatric fluoroscopy room, improving the efficiency of the pediatric fluoroscopy suite because the patients were more engaged and the exams went faster, and fundamentally, collaboration of everybody in the healthcare team with the patients and families. And why do we want to do this? Our traditional radiology patient relationship has been weak, and we've seen campaigns like the Face of Radiology and Radiology Cares and bringing radiologists back into direct communication in the diagnostic realm. The competition that we're seeing in the marketplace is both within radiology and outside radiology as more and more people look to provide imaging-based services for many specialties. They're their patients. They believe they can add the same value. How do we distinguish ourselves? And then we also see many disruptive technologies coming along that we're trying to integrate. How do they impact what we do? Smaller, less expensive footprint machines, threatening some of our core business, and technologies that we're afraid will put us out of business and become the automated radiologist, computer-aided diagnostics and big data. So if you try and bring this down to why we need to bring patient-family-centered cares into our programs to drive quality and value, part of this is because radiology costs a lot. People demand a lot when things cost a lot. And we need to reassociate with our patient base. We need to direct our care out to patients. People market to patients. Hospitals, pharma companies, healthcare, people market to patients the value and quality of what they do. We've been a little bit behind in that in radiology. What do people want to know about us to bring them into our services? So now, radiology has really evolved into a team, and this really speaks to the fact that every person who touches the patient in a radiology operation from start to finish is part of a team, a continuum of care, to bring it all together. And the efficiency of our product, the tools that we use, the products that we deliver are fundamentally core to embrace all the people along the way. So I can't imagine a quality improvement project in our CT operation that doesn't start with the people doing the scheduling at the front desk knowing that they have the right information to schedule the patient on the right machine in the right location with the right set of symptoms. Without the imaging technologists and the protocols and processes that they operationalize on our behalf as physicians to do the tests. On recording of the radiation exposure, my physicist colleagues to make sure we understand the information we're getting and the reports that we're getting to do process improvement around radiation exposure. From patient satisfaction data, looking at our reception areas, our patient lounges, how they get their test results, what they want. The checkout process, the billing process. As was said earlier today, somehow healthcare systems are always able to get the bill to the patient. But somehow radiologists aren't always able to get the result directly to the patient. Is that something that we can change? So now we start to see externally imposed metrics, superimposed on us, that give us data. Things that we have to respond to in an increasingly complex healthcare reimbursement structure. Patient satisfaction scores being very fundamental to driving how we run our operations and doing quality improvement around the patient processes. H-gap stores, exam appropriateness is another big one coming down the pipeline very quickly where reimbursement may be driven by decision support being used in order to get paid for advanced imaging. Are you doing appropriate imaging? If I ask most of you in the room, do you think you do appropriate imaging? Raise your hand. Pretty much everybody's gonna raise their hand, but can you prove it? Well, I read the requisition, I protocol the right test, I think I'm doing appropriate imaging, but can you actually prove it in any measured manner? Well, that's going to sort of be enforced on us with CMS driving decision support into reimbursement for the future. Somebody else is helping us drive those metrics on appropriateness of the tests we perform. What about the structure of our reports? Can we improve, in a quality improvement project manner, the quality of the information in our report so that the user, whether it's a referring provider or a patient, can understand it? If we're reporting MRIs for multiple sclerosis, can we incorporate the coding scheme and improve the quality of the reports from the perspective of the neurologist user who ordered the test? These are not challenging quality improvement projects, but if you understand your referring physicians and you know the standards of diagnostics and guidelines they expect to be able to use the information, you can incorporate guidelines and structured reporting that gives the information to them that they want and drive up your referring physician satisfaction scores and they will find your resorts more valuable. So as we go along this rapid advancement in healthcare and changes in radiology and radiologists and physicians as leaders and role models in this process, we try and understand what that means for value. We've seen many of these graphics over this last year in particular in many publications in here at the meeting. Value is quality plus service over cost, or quality over cost, where quality is not just the medical outcomes but the patient experience, as we've heard today. And I like the way Jonathan Kressl phrased it this morning. It's what you get for what you're willing to give or pay at a very tangible level, and that's the you, that's the person, that's the person making the judgment. They're the people factor. So the value to me that we're talking about today explains what the mismatch is that we as healthcare providers have had about what our patients see in quality. We've always thought quality was a CT with the lowest radiation exposure for the question being asked with a fastly turnaround report time and access to a test this afternoon. That's us. Those are professionally centered metrics. But in the value world, if I go back, you see that the quality as perceived by the patient includes a lot of the patient experience. And this delta, this disconnect between the value we're driving to and what we've seen as a mismatch between our perception of quality and patients' perceptions of quality is very real. As we define those healthcare metrics as morbidities, mortality, cure rates, complications, hand hygiene scores that have a physician or organization center focus, the patients perceive 90% of their medical care on the non-medical aspects completely. So they're looking for value not just in what we believe are medically important outcomes, they're driving it mostly on the patient care experience to determine their value. And if we don't pay attention to that, we will not be providing the value, we will not be driving the patient satisfaction scores that are so important to us. And quality improvement tools are methods to help us get there. So what do all the value equations and radiology process maps show in common? They all involve the people. Interacting through processes and things like quality improvement tools to interpret data to improve processes. But another question I might ask you is why is healthcare different than most factory-based industries? You make a hammer, it meets specs with 99.9% confidence coming off the line and you sell your hammer. You change your processes and maybe get a little closer to 100%. Well in healthcare, our product is also the people that interact with the system. Our product is the health of the people who come through the doors. It's not that easy to improve the quality of a human being's life because we want to apply standardized processes. They want to impose their own values. So you might be driving down a quality improvement project that says, you know what, in the cardiac surgery clinic, we have patients coming to be evaluated for TAVRs, these new percutaneous aortic valve replacements. And in that assessment, there's about 10 tests they want the patient to have. And they would like to have them all in the same day so the patient doesn't have to come again. Sounds like a great goal. Get it all done quickly and efficiently, save a patient a drive back. And so if you set about doing that as healthcare providers and set it up that way, sounds good. But then what you'll find at the end is that the patients really didn't want what the clinic thought they wanted. If you think about the average age of a patient coming for a TAVR assessment the last couple years, they're somewhere between 75 and 85 years of age. They don't ambulate very well. And can you imagine running around a healthcare system for a day getting 10 different tests? They're exhausted. So we make assumptions as healthcare providers and the efficiency is get it all done back to back one day, but it's not what the customer really wanted. It doesn't drive the patient experience. In fact, they don't want it at all. So you have to ask the right question. If you don't involve patients, you're going to get the wrong solution. The product of our industry is inextricably people-centered. We are treating patients. We are diagnosing patients. So it all gets macked to me. It's not uncommon, core champions of value. It's actually quite common. As I started, it's the people. The people, the people. And as I borrow from our football coach at the University of Michigan, the people in the team. Interacting with processes, looking at data to provide value. It's important that we maintain a customer service. It's important that we maintain a focus on excellence and the way we deliver services. And the very simple things that you can do if you haven't done a big quality improvement project at all is to just ask, how do your patients get into your practice? How does somebody actually get an appointment? What do you ask of them? What's that long set of questions that somebody asks on the phone? Does somebody email and get a test done? Do they send something through some sort of electronic portal to request that? Do they walk up to a counter or to sit on line on a call? How do they access your system? Are they happy with it? Could it be better? If you don't ask, you'll never know. Is it helpful or a roadblock? Do the people who are actually doing your scheduling have good customer service skills? Have they actually ever been trained in customer service? Are they knowledgeable beyond putting a block, a request in a block? Do they actually understand what's going on with a request in a block? Do they actually understand the downstream implications to the technologists and the radiologists about why things are being done? Thanks, Olga. If you don't know what your front door is doing, you should probably ask. People have got to get in your door and want to come in your door. Survey your patients and referring provider offices. This is not a complex quality improvement project. There are many examples. You can Google out there on the literature. Go to the web and you can find surveys that patients have done or physician practices have done of their referring physicians and patients. What do they want in their scheduling process? How is yours working? If you find a defect and something's not working, don't get defensive. Bring the right people together. Try and improve your process. Engage the people from beginning to end in the process, your staff, your patients you're referring, your families. Change something. Don't be afraid. It's your practice. And remeasure and see how you do. A very easy way to get started. How do you deliver your radiology results or communications on the other end? Do you deliver them directly to the patients outside of say mammography? Do the patients really want the results directly? We heard this morning, should we release our results directly to these patient portals with no intervening referring physician modification or translation? We're kind of at the precipice of understanding how to use some of these new tools. In a study from, boy, 20 years ago, 92% of patients wanted to be told if a test was normal. It would give them reassurance. But fewer patients would want to know directly from a radiologist if their test was abnormal. Now in 2016, in a survey that we published this year, 85% of patients expressed an interest in speaking to a radiologist once they understood what a radiologist was and could offer and what their training was. Furthermore, a quarter of patients, and this is a survey both of our large academic medical center and a large private practice hospital in the same city, 25% of patients said they'd pay $40 or more to talk to a radiologist about their test results. So I think more than ever, patients want to be engaged and we really need to consider how we man up, so to speak, how we present ourselves and make ourselves accessible to patients to deliver their information to them. So when you think about how you're delivering results the way out of your department, stop and think about what you're doing. You dictate a result and something magical happens to it and off it goes. Who's it going to? How's it getting to them? Google around the literature. What are people doing? What's new in this space? Survey your patients. Are they happy with how they get the results? Make improvements and survey again. These two very basic quality improvement type things on the way in and out of the way of your department can really change the culture of your operation. And really, look at what's going on in the patient portal space. Patients are increasingly wanting their information electronically. And how we do that is we're at the very beginning of trying to understand. Get input. You can't do quality improvement without patient input. Thank your customer. How often have you bought a car and you get like three or four follow-up calls and letters in the mail to fill out surveys to see how they did? Thank your customers for coming to your services. And if they have any questions, that you're willing to call them and follow up with them. And establish a culture or a philosophy in your practice that makes you pledge to follow and provide the best personal service, to make interactions positive and professional, to be responsible for uncompromising cleanliness, the experience in the patient environment, and taking pride in your personal appearance. And as physician leaders and staff leaders, this has tremendous influence on those who work with us. So at the end of the day, our uncommon champions of value are very common. It's the people. The decisions we make about value are based on our individual values and often financial decisions combined. They're unique to every individual. To truly create value for people requires doing the right things for them at the right time in the right way. And we can improve our processes to do this. So I would encourage you to talk to your patients and families and referring providers and find, through those conversations, opportunities for improving your practice. It's a little bit like Darwin, the survival of the fittest. We have to evolve and change if we want to survive as a discipline. So uncommon champions of value, it's all of you and all the people you touch every day in your radiology practices. Thank you very much. I'm gonna close by, I've got a crystal ball up here, and I'm gonna try and look into it and predict what patients will demand from radiology in 2025, which is really just nine short years from now. I'm gonna start with two fairly broad predictions. It's always more and better. That's what the future promises. And I think healthcare won't be any different. Better will mean better integration between all the different organizations, the different individuals that really are a pretty loosely organized, we call it a system, it's really not. Better will also mean tailoring the diagnostic process and treatment choices to meet the individual patient's unique needs. More will mean greater transparency on topics like pricing and expertise, and we'll have more predictable processes and outcomes, hopefully, in healthcare. And I expect that many of you will have your own ideas about what patients will want, and I invite you to go onto Twitter and use the hashtag radiologyquality2025 and post your ideas, because I think it'd be fun to have a conversation about what do you see the future? And it's not that far away. So another easy prediction. In 2025, we'll be working with more patient data. Some of that data will be coming from our time-honored process of people walking into a room, asking patient questions, walking out of the room, writing it down, or maybe typing it into a medical record. But unfortunately, unless we hire a lot, lot more people in healthcare, and I don't think we can, we won't be really expanding this aspect unless we go to a different source. Other industries, Amazon, Google, they have you type in the stuff. They don't, I mean, you don't, they don't have scribes. I'm a scribe. I walk into a patient's room, I interview them, I take notes, and I walk down and I write it down. I'm a scribe. Google doesn't do that, Amazon doesn't. Other industries have figured out how to get around it. And maybe we should do that too. Maybe we should have patients entering into their medical records to tell us what they expect, what their values are as part of the treatment, because I think that would be very valuable information. Another easy prediction is that we're gonna have more data from sensors. It's gonna be our smartwatches, it's gonna be our smartphones. And many of those sensors will actually be arrays like imaging sensors, and they're gonna be imaging devices, especially bedside devices. And I actually had an interesting experience recently. I had a patient who was acutely short of breath after a splenic artery embolization, and I went the old route, okay? I was asking questions, trying to do a physical exam. The medical resident came down, grabbed our ultrasound, which was sitting right there, took the transducer, put it on the chest, and made the diagnosis of flash pulmonary edema, and was using imaging data to guide treatment choices. So the amount of data, it's actually staggering. You see some of these estimates, you know, exabytes and doubling in 73 days. And it's also gonna be because every site's gonna collect data, and it's pretty easy to share, and you really don't dispose of digital data, so we're gonna have a huge amount of data, and it's gonna be overwhelming. And I've been really, I've been troubled by this. I really am. But I've come up with a simple solution. I'm gonna hire Chloe. Anybody else watch 24? Chloe is amazing, okay? You give her a computer in about 35 seconds, and she can get data from a government, private databases, and then she'll mash it all up together, and she'll solve every one of Jack Bauer's questions. The problem is, is Chloe is fiction, okay? Watson, well, but Chloe is pure fiction. So let's come back to reality and talk about some hard truths about data, information, and communication. And if this was all actually worked out over 70 years ago by Claude Shannon, he saw data as the raw material. And if we start in the upper left-hand corner of this image of his transformational report, I'd say, well, the first pixel is a dark gray. The next pixel is also another dark gray, dark gray. You know, I'd rasterize the image and report it. But you look at this, and you're looking quickly for patterns, and that's what Shannon recognized. It's the patterns within the data that allow you to recognize letters and words, and that data conveys information. And Shannon defined information by its ability to reduce uncertainty. And went on to say that successful communication is reducing uncertainty in the messages recipient. And Shannon went on to prove that error-free communication is possible even when we're in noisy environments, when you're adding noise into images or messages, because it's possible if you use error correction strategies like feedback and redundancy. So this is part of Shannon's model. He recognizes that messages, they're encoded, transmitted, and then decoded at the other end. And it all occurs within a context. In this particular context, it's something that I encounter. I walk into a hospital room. The child has had right lower quadrant pain for several days. His parents are nervous after the CT scan. And my message is that the CT scan shows a ruptured appendix. There's an abscess, and we need to percutaneously drain the abscess. The room's noisy. There might be siblings in there to distract the parents. And all they hear is appendicitis. And when I ask the parents to tell me what they understand, they might respond, well, your plan is to remove my child's appendix. And what that feedback loop tells me is that some of the information has been conveyed, but there's a fair amount of uncertainty that persists. And it's probably going to take a few more cycles in this conversation to adequately address the uncertainty. So the crucial point is that communication, it's not so much about what I say. It's about what the other side, what the recipient understands. It's a dynamic problem and one that clearly has numerous failure modes. So what does all this have to do with managing uncertainty in patients and what they'll want from radiology in 2025? I believe radiology's core strength is that we can gather data faster than anybody, okay, with our imaging sensors, huge amounts of data flowing in. But we also do a great job of transforming mounds of data into small nuggets of information that address diagnostic uncertainty. And to borrow the phrase from Clayton Christensen, medical imaging is disruptive. He also has a theory that he likes to call his jobs to be done, where people don't really want to buy the quarter-inch drill. They want a quarter-inch hole. So that brings us to why do patients hire the healthcare system? And I propose the following model. First of all, patients don't want to hire us. They will do most anything to stay outside our walls. So patients with a symptom, condition, or need, they first assess whether they can take care of it on themselves. And it's only when they can't that they enter the healthcare system where there's some diagnostic uncertainty. That typically leads to some request for information. It might be in a history physical, it might be a lab test, but clearly imaging, again, huge amounts of data flowing quickly, the data's interpreted, and that interpretation transforms data into information, the information goes back, addresses the uncertainty, and once uncertainty decreases below a certain level, you start with treatment. And if treatment's successful, boy, the patients are like a bat out of, and they're leaving your walls. Now, my wife and I got to test this model a few months ago. We were traveling from St. Louis to Ohio, and she developed right lower quadrant pain. We didn't immediately hire the healthcare system. We were far from home, Googled it, said, we're going to hire Advil. It wasn't until the next day that the pain persisted. We entered the healthcare system. The chief source of diagnostic uncertainty was, could this be appendicitis? Go on to get the CT scan. It is appendicitis. And so pretty soon, we're starting on the treatment arm with antibiotics, and then her appendectomy. And I'm happy to report that within a few days, we were discharged, and she's completed a recovery at home. But as I watched all these events unfold, it's pretty easy to see how healthcare has this big data problem. It was uncomplicated appendicitis. I mean, it's run-of-the-mill stuff, but it generated a lot of data. And when I think about the ICU patients, I mean, we're drinking from the fire hose, okay? The lab's coming, the vitals, the imaging, all the notes, all that stuff that's pouring into this medical record, and I worry about the days that Chloe takes off, okay? That it's going to be my responsibility. And even on the days that Chloe's there, she's trying to make sense of this electronic medical record where the data elements are loosely organized text, and you never hear Chloe griping about HIPAA concerns, right? Traditional characters like Chloe can transform mountains of data into nuggets of information in a flash, but my ability to consciously transform data into information is actually very limited. I'm about 100 bits per second on a good day. It's only with pattern recognition that I can take big gobs, you know, maybe approaching a gigabyte per second. And the problem is that patients are constantly producing more data. It's every new heartbeat, breath, lab result, imaging study, note, symptom. And even with pattern recognition, I can't keep up with this. And as every new data point arrives, I'm on this, do I make a decision or not? And there's a voice in the back of my head that's always telling me, well, it could be the next data point that's the crucial key to the puzzle. And so there's this indecision, not wanting to jump in. And so we're going to need help. You know, Watson, other machines, actually, will do more data collection and analysis. So medical imaging, I mean, we already do it. I mean, we collect a gob of information with an array of sensors. We don't use scribes that are writing it down. We present it as images, and we trust the machines to collect and transform raw data into a usable format. And it's presented to us in a format that's actually easy for us to grasp with our visual system. I despise EMRs that present data like this, okay? And this is pretty typical to what it is. I'm hoping the EMR designers, at some point, will pull down Tufte's books and recognize that presenting data as columns of numbers is silly. If we did that with imaging, you'd be looking at arrays, big tables of Hounsfield units, page after page of Hounsfield units that you'd be paging through. No way, no how. I mean, it would make sense to transform this data table into something that's a little bit easier to digest in a flash and recognize patterns, and also integrate it with some of the other important points going on during that hospitalization. And unfortunately, this doesn't happen yet. It needs to. So improved technology will take us through some aspects of healthcare and improvement, but us as humans, I'm not worried about being replaced. My bet is that we'll still be called on to do the advanced data analysis, and we'll also provide empathy. It's the people. It's the emotion. Because even though technology continues to accelerate, the human brain and our ability to analyze data, we haven't changed much over the last 100,000 years. Michio Kaku calls this the caveman principle. It means that during stressful times, humans want to interact with other humans. And I saw this during my wife's appendicitis journey. It was the CT technologist that saw how scared my wife was, talked her through every step of the process, held her hand, calmed her while they were doing the CT. It was the surgery chief resident and floor nurse who saw the paroxysms of pain in the post-operative period, recognized that the pain regimen wasn't enough, made adjustments. It was the patient care technician in the room that recognized that creating small islands of order by cleaning and organizing the room can counter the chaos caused by this simple hospitalization. So what will value metrics be for radiology in 2025? And before going further, I think we need to step back and say, well, the health care system, it's only a minor piece of overall health. It's about 10%, okay? Behaviors, genetics, and environmental factors play a much bigger role. And that shouldn't keep us from trying to improve what happens within the health care system. But maybe it ought to teach us to think maybe how do we expand the reach of the health care system or influence the other factors that govern overall health. So it's my strong belief that value, the indicators of value and quality are really encoded within the patterns of care. A single event in a care process, it's a data point. Patterns of care are informative because they reflect a series of events. And we've all seen patients who get in the cycles where they keep going through the diagnostic loop over and over again because of failed diagnostic cycles. You'll see them also getting failed treatment loops. And it's interesting to think about could we create measures that look for efficient and effective trips through this cycle so that you go through a single diagnostic loop, a single treatment loop, and you're out. And a potential model for this already exists. Low back pain, common condition, CMS put together, and this metric is actually reported on the Hospital Comparer website. Patients first, I mean, they start with Advil, right? Nobody wants to go in to see the physician right off. You'll start. It's only when the symptom persists or reaches a certain threshold that they enter the health care system. They're evaluated with some sort of data collection. If there are no red flags through that collection, they're slotted for conservative therapy and that works. It's only when there are red flags or continued pain in this, I mean, an existing CMS metric that an MRI is considered worthwhile. So plans of care like this one for low back pain not only provide some transparency, but they provide a series of expectations or predictions. They describe a journey through the health care system, and in the future, I expect that patients will want those journeys. Should they be more, they want those journeys to be more like every other journey that you take, okay? Your plane ride home, your itinerary. And if you want to see, this is actually a really entertaining YouTube video that talks about if air travel were like health care, flipping it, and it's worth watching. So these plans would clearly only work if we communicate them. They have to be communicated to the patients, the rest of our care team, the people at the front desk, and the rest of the physicians on the team. Patients I believe will demand itineraries because they allow them to check their observed progress. They want to see the mileposts, you know, fading into the background as they're traveling down the path. They want to know in the coming days, weeks, or hours, you know, who's going to be able to go back to work because their child's ill? When are they going to be able to, you know, get on with their lives? They'll also use this to assess quality. And I define quality as conformance to prediction, that if the health care provider team and the patients have the same set of prediction and expectations, they both can assess quality and the quality metrics would not only reveal the observable checkpoints as they're going down this journey, but they would also provide the timing and the sequencing because sequencing is important. A single large deviation from these plans, well, that's an informative event. They should be investigated to determine what caused their plan to go awry. A series of small but nonrandom deviations also suggest a problem with our predictive model and should be investigated. In each case, we should be looking to improve this predictive model that was really the patient care plan. So taking this back to my wife's appendicitis journey, time, it's easy to observe. We actually had a timestamp in Google Maps that told us when we arrived at the emergency room. We compared it to the timestamp on the CT, and it was an hour and 15 minutes from the time we actually parked the car until the CT scan was done. And I think that's terrific. It's not the 15 minutes that David and his team at Stanford can do for door-to-TPA time, but it made sense for appendicitis, and we thought that was terrific. What was not terrific was the eight hours from that CT scan to the OR, okay? Especially when we were told at 10 o'clock in the morning that we were aiming towards a 1 o'clock surgery slot, we thought that would work. Not 3.30, we're still waiting, and somebody comes out and tells us that, well, the OR, they're turning it over. It'll be another 15 minutes. I have never seen an OR turn over in 15 minutes. I mean, it was. It was, don't lie to me, you know? Please. I know doctor time. Images are also going to demand more transparency, not only about the timing, but also about the cost. We were charged $5,500 for that CT scan. I'm sorry. I mean, that's extortion. It's especially when the acquisition of diagnostic images, it's really pretty close to commoditization. You know, a fair price would be under $1,000. Interpretation, yes, that's something still that's different. But acquisition of images, pretty straightforward nowadays. Patients also want reasonable predictions about their pain, duration, and your expertise. I commonly tell patients now that they should ask three questions before any procedure. They should ask when it's going to start, how long is it going to take, and they should ask if we've ever done this before, okay? And I try my best to sort of answer those questions. If I don't know when we're going to start, I can give them a range, you know, sometime this afternoon, sometime this morning. When I tell them the time, I include it. It's 30 minutes to set up with anesthesia, 30 minutes to finish up. Our procedure's 15 minutes, but your child will be out of your arms for about an hour and 15 minutes. And then when talking about our experience, I tell them that, you know, we've done this procedure many times before, but it's the first time for you and your child, and it's always possible that your child's going to teach us something new, something that we haven't experienced before. And so there's always this element of unknown, but we're prepared to try and deal with the expected unknowns. So we also try to tell patients when we're falling behind a few minutes and let them know why, whether we got started slow, and give them an update, boy, I sure wish we could get to the point where it was like my pizza, that I know where it is in the process, or my FedEx package, or my flight across the country. I mean, they have sensors embedded in their systems that actually can update patient expectations. So patients are clearly frustrated with how we collect lots of data, and then we never seem to use it. We need expert systems that do a better job of using the data to narrow down the list of choices so that they match the needs of an individual patient. We want to share this information with the patient and get their input. It highlights the importance of this choice architecture where default choices are key, and also the value of a short, organized list rather than the exhausted, unorganized list of every possible choice being presented to patients. And we should recognize that if a choice doesn't make sense to you or me, it might be because we don't share the same value system as the patient. I encounter occasionally Jehovah's Witnesses with GI bleeds. I can't make sense of why they refuse a life-saving blood transfusion. It doesn't fit my value system, it fits theirs. And the best I can do is explain the situation and trust that they understand the consequences of what was really their decision. So by this time, I hope you'll recognize that every diagnostic encounter is really just a small addition to this ongoing story. And Claude Shannon, he had this, he was testing his information theory. He was reading detective novels out loud to his wife. And he would stop in mid-sentence and ask her, guess the next word. And he was astounded by the fact that she was often spot on or very close. And it's the same manner that diagnostic imaging, it's just part of the patient's ongoing story. The prior chapters of the past medical history, the past surgical history, you know, a lot of stuff buried in the EMR. And if you had all those, you could actually make a first-order approximation guess of what the next imaging study might be. And your prediction could be updated by information gathered when the exam was scheduled and it could be updated further if there was the history or labs or other more recent information. And this very Bayesian approach would allow you to predict the question being asked and how best to address that uncertainty. And while such integration has a value, it will require clearly some marketed advances in what we have for electronic medical records in those systems to get to it. So as we wrap it up, I need to address some concerns about efforts to measure quality and value. The first is that measurement provides us this ability to assess if we're improving. Too often, I mean, we've all seen it. Improvement efforts that a couple weeks down the road, nobody's measuring and so nobody can really tell us whether we're improving or not. The other concern is it's about wrong or bad data. And the point is data is what the data is. And yes, there might be problems with the third decimal point in some of our measures. But the much bigger problem is when you try and take data and transform it into information and then use that information to make valid decisions. Those are where the more common errors are. So there are also concerns that people out there will try and gain the metrics rather than actually improving the care they provide. And it's completely possible and it actually is expected. But in the short run, it can work. But since quality and value are really encoded in larger patterns of care, gaming efforts at some point create a house of cards that collapses. So it goes back to Clayton Christensen's theory of jobs to be done. And I postulated that patients hire a health care system when they have a problem or a need that they cannot solve on their own. And that imaging's role in this process is addressing the diagnostic uncertainty. And our job isn't done, as much as we'd like it to be, our job isn't done when we broadcast a message. We really have to look downstream to see if the uncertainty's been reduced and actually had an impact on the care path that the patients are going through. So how will we measure quality and value in 2025? We're going to have to look at care as a sequence of events. We're going to look for evidence that our message was actually received, correctly decoded and used to solve the problem in an efficient and effective manner. And this, again, will require looking at sequences of events. So thank you for your attention, and I look forward to seeing your ideas on Twitter.

performance improvement

quality assurance

healthcare sector

change management

strategic insights

implementation

iterative process

strategic planning

organizational culture

leadership

patient-centered care

team execution