In this episode of the AMA STEPS Forward® podcast, Elizabeth Harry, MD, senior medical director of well-being at UCHealth, discusses the connection between cognitive workload and physician burnout and how health care organization leaders can address systemic issues to reduce burnout and promote physician well-being.
- Elizabeth Harry, MD, senior medical director of well-being, UCHealth
- Jennifer Mathews, practice transformation and sustainability, AMA
Intro: Hello and welcome to the AMA STEPS Forward® podcast series. We'll hear from health care leaders nationwide about real-world solutions to the challenges that practices are confronting today. Solutions that help put the joy back into medicine. AMA STEPS Forward® program is open access and free to all at stepsforward.org.
Mathews: Hi everyone and welcome to the AMA STEPS Forward® podcast. Today I am joined by our guest, Dr. Elizabeth Harry, a practicing physician and senior medical director of well-being at UCHealth in Colorado. Our topic will be physician cognitive workload or task load and its relationship to burnout. Dr. Harry, thank you so much for joining me today.
Dr. Harry: Thank you so much for having me. I'm really excited to be here with you.
Mathews: Let's start by telling the listeners a little bit about yourself and your background, and how you got into the world of physician cognitive load and burnout.
Dr. Harry: Absolutely. So, I am an internal medicine physician. I have actually done about 10 years of hospital medicine and then recently made a transition to primary care. And when I came out of residency, I was doing a lot of quality improvement work, and one of those projects had a lot of great data for patients and a lot of great patient-centered outcomes. But our providers had a really hard time with the initiative and I thought that was so interesting. And I started asking them why they were struggling with this initiative that improves all of these patient care metrics that we care about so much. And I kept hearing, "I'm just so tired of change. I'm so tired of one more quality improvement project. It doesn't matter what it is; it’s just that I don't want to handle any more change."
So, I actually started researching change fatigue at that time, and that brought me to this concept of cognitive load, and we can talk about what that is, but from there as I started to get interested in cognitive load, I started wondering what impact cognitive overload had on providers. And that's what led us to this study.
Mathews: Got it. So yes, let's start with that definition that you were talking about. What is cognitive load?
Dr. Harry: So, everybody has a fixed amount of working memory and working memory is what you need to be able to learn anything or to be able to produce anything from the knowledge you have. And it's really important to recognize that it's fixed because we overload ourselves and we don't sort of acknowledge that it's a fixed, limited resource that can be used up. And cognitive load theory is a theory that talks about how much of your working memory is being used at any given time. And it's basically impacted by how hard the task is, how much experience you have doing the task and then how well the task is sort of presented or the different elements of the task are able to be accessed. So, if things are disorganized and you have to go to many different places to get them. Or a clinical example would be if you go to one clinical room and it's organized in one way, and yet another clinical room is organized in another way; part of your mental energy or part of your workload is being used to think about how it's organized and not just think about the task.
So, you can use your working memory and deplete it, and then it can become repleted with attention breaks, which are really important but that's not how many of us practice.
Mathews: Right, exactly. And within the umbrella of cognitive workload, the paper that you referenced, which was published last year, which we'll get to in a minute, discusses three types of cognitive load. Can you tell us more about those three types?
Dr. Harry: Absolutely. So, if you think about this fixed working memory and then cognitive load being how much of that working memory is being used at any given time, there's three things that impact that. So, the first is intrinsic cognitive load, and that's how hard is the task itself. And some tasks are just more complex. They have higher elemental interactivity or in other words, more elements that need to interact to get the task done. And that makes it have a higher intrinsic cognitive load. You can't do much to change that. It just is what it is. The second is germane cognitive load, and that's how much of your mental bandwidth or working memory are you using to learn the material. So, this is really pertinent for our learners, our residents, our medical students but hopefully all of us are lifelong learners to a certain degree. So, as we're interacting with the data, we're also learning about the data, making mental models, trying to understand the science, understand the patient care.
And so, there's a part of your brain that's always working on building those mental models. And if you're a novice, a larger part of your working memory is dedicated to that and that you can modify over time but obviously not in the short term. And then the final one, which is the most modifiable, which is why we focus on it the most, is called extraneous cognitive load. An extraneous cognitive load is the cognitive load due to the organization or lack thereof of how the information is presented to you. So, if there are three main offenders for extraneous cognitive load, one, is lack of standardization. So, if the information is presented to you differently every time or if you do the same process differently every time or like we talked about, clinical rooms are organized differently every time. The second is redundancy. And this is a really interesting one because in health care we often cite redundancy as a patient safety mechanism but it adds many layers of complexity.
And as our systems get more complex, that becomes a problem. And then the third is split attention. And so, I want you to think about when you're doing a task and then you get interrupted from the task, how long it takes you to reorient back on what you were doing. And it's a very inefficient way to work and it actually makes us more prone to errors. And so split attention is a really good target for intervention because it's a place that we have a lot of interruptions in health care and it's a place that we could make a lot of headway.
Mathews: Yes and I definitely want to get to those specific interventions to reduce extraneous cognitive load. But first I want to delve a little bit more into how do we measure physician cognitive load in an objective way and then back to the paper. So, this paper was a paper you co-authored with Dr. Chris Sinsky and it is called “Physician Task Load and the Risk of Burnout Among U.S. Physicians in a National Survey.” And this was published in the Joint Commission Journal on Quality and Patient Safety in 2021. The main finding of the paper, as I understand it, is that physician task load was strongly associated with physician burnout. Could you tell us a little bit more about the study and in particular how you measured cognitive load as well as the outcome of burnout?
Dr. Harry: Absolutely. Thank you so much. We had a great partnership with Chris and the entire AMA team and many other members of that team, including folks from Stanford and Mayo. And we used the AMA Masterfile, which is, as you know, a nearly complete record of all U.S. physicians independent of AMA membership. And we got a 17% response rate, which was about a little over 5,400 respondents. And we measured burnout, the standard measurement of burnout using the Maslach Burnout Inventory, and we categorized burnout as being positive if someone had a high score on depersonalization or emotional exhaustion, which is standard in how you measure burnout. And then we use the NASA task load index to evaluate physician task load. And the NASA task load index has been validated over 30 years in multiple different industries, and we were very curious to see if it would be applicable to physicians.
They use this often to look at an individual task or process and they'll particularly ask about a process. We asked more about “in the last two weeks, in your work environment.” So, we did ask it differently but it's been used that way in health care literature before. It has six domains and we evaluated all of them but we ended up taking two of them out because they actually varied or tracked with burnout. And so, they would've falsely made our data look stronger because effectively they're measuring burnout. So, the domains are effort, mental demand, physical demand, temporal demand, frustration and perception of performance. And the last two were the two that we ended up taking out because of their tracking with burnout. So, frustration and perception of performance. So, we then took four domains effort—mental demand, physical demand and temporal demand and called them physician or provider task load. And there's actually a nursing paper that did the same factor analysis and actually took out the same two domains. And so, it seems that these four domains are very nice, conserved measure of task load in health care workers.
Mathews: Got it. Is this the most commonly used instrument to measure physician task load?
Dr. Harry: Yes. There're other ways that you can do it that are a little bit more burdensome, such as when someone is actually doing a task, tracking their eye movement, et cetera, to try to see how often they're deviating from the task at hand. But that's not something we could do on a large scale like this. So, it's the most accessible for a study like this.
Mathews: Got it. And then tell me more about the findings of the association between the task load and burnout.
Dr. Harry: I think it's really interesting. So, we actually took all of the physician task load scores and we put them into quintiles and we looked at the top quintile of physician task load and their odds of reporting burnout was 68%. And we looked at the bottom quintile of physician task load and their odds of reporting burnout was 22% and that had an odds ratio of 7.4, was statistically significant. So really large difference between the top and bottom quintile, which I think is really interesting. We also tried to find sort of a linear relationship between it. And so, we found that for every 10% change in physician task load, which was a 40-point change, there was a corresponding 33% change in the odds of reporting burnout. So, if the physician task load score went down by 10%, the odds of reporting burnout went down by a third or 33%.
And if the physician task load score went up by 10%, the odds of reporting burnout went up by 33%. Which is a pretty profound relationship and suggests that we may be able to impact both task load and/or burnout pretty directly by small modifications. In other words, a small change in the task load correlated with a large change in burnout. And that might suggest that decreasing task load might have a large impact on burnout. Obviously, this is not a randomized controlled trial for it to suggest this correlation...and those are further studies that would need to be done.
Mathews: But your point is an important one that the effects seem large proportionally. Can you give me a sense of how much this 10-point difference in task load, would practical changes in a physician's day-to-day cause a 10-point reduction or increase?
Dr. Harry: Well, it's a zero to 400 scale, so each of the domains is zero to 100 and it's subjective so it's self-reported. I think that's a really good question because it depends on the way that people report that and how dramatically they experience a reduction. But over the 400 scale, we would only need a 40-point drop. So, you could just see a decrease of 10 points in each of those domains, and that would decrease the physician task load by 10%, which would decrease the odds of burnout if there is a causal relationship by 33%. And again, we don't know if there's a causal relationship but in theory that would be the expectation and that would be the next study to do would be to study a process, see what the task load is associated with the process, redesign the process to have a lower task load and measure people's burnout before and after to see if there was actually a reduction in burnout associated with the decreased task load of the process.
Mathews: That does sound like a great study, if it's not being done yet, I think that would be huge to kind of quantify that. In terms of interventions, we have always traditionally wanted to focus on things that can be done on the organizational level. So, what can practices and organizations do, but also I think in this topic in particular, there are some individual interventions that can be effective. Can you give us a few examples of each?
Dr. Harry: Absolutely. So, I think thinking about what are the three main drivers of extraneous cognitive load is the place to start. And so, standardization is the first thing. And where can we standardize? And we can standardize in lots of places. We can standardize how we deliver care for certain kinds of patients. We can standardize what units look like, we can standardize the workflow of the day, and then personally, we can standardize our lives. So, I want you to think of Steve Jobs wearing the same outfit every day. That's like an example of standardization. And what's nice about standardization is when you build a habit, you move something from short-term working memory to long-term working memory. And as far as we know, there's no limitation in the size of long-term working memory. And so, you're not going to be using that precious resource, that short-term working memory and you're moving it to a place that is unlimited as far as we know.
So as many habits as we can build in our workflow and as much standardization as we can insert is important. Now I want to just acknowledge that there is a tension between standardization and autonomy. And a lot of times physicians don't want a lot of standardization because they feel that it impacts their autonomy. And I think it's a very interesting conversation to have around what is that balance between standardizing things that can be standardized and just create unnecessary noise in our workload versus leaving that creative autonomy for how you respond. That would be intervention number one is really looking at standardization. The second one would be looking at redundancy and I think this one's really fascinating. So how many different ways do people get the same information? So, if I have a patient that has a positive blood culture, do I get a notification in my EHR?
My resident or intern will tell me, the nurse may call me, I may get a page from the lab and I now have four mostly reliable processes telling me the same information. When what I need is one highly reliable process, giving me that information every time. And the problem with having four mostly reliable processes is that if one fails, I can't be sure which one failed because I am always scanning the environment for where I'm going to get the information from. So, it takes a lot of cognitive energy to have this duplication of knowledge coming at us. So, I think that redundancy piece is really important and thinking about trying to simplify as much as we can in terms of having really highly reliable processes. The third one is split attention and I think we have a lot of opportunity here. So, there's a lot of examples of split attention but the most salient one are interruptions.
And we do a lot of interruptions in health care. So, I've seen physicians interrupt nurses, I've seen nurses interrupt physicians, we interrupt each other all the time. And the reason that's a problem is because when you shift from one topic to another and back again, you lose 90 seconds of data every single time. And that's called an attentional blink and it's a very inefficient way to practice. So, they actually did a study in the nursing literature where they had nurses wear bright yellow vests during med pass time, and nobody was allowed to disturb them while they were passing meds, which for them is their very high intrinsic load, high-risk task. And they saw the rate of medical errors decrease significantly when they weren't being interrupted during these tasks. And so, I think thinking about what is our high intrinsic load or high-risk process that we do as physicians as APPs. And what processes do we need to put in place to protect that space from interruptions?
Unless of course, it warrants the level of urgency to have an interruption. That implies that as a team, you've come together and decided what level of urgency warrants an interruption. And that's something that would be novel in and of itself, which is just coming together with your team and saying, “This level of urgency is okay to interrupt one another, but if it's not that urgent, then let's wait. Consolidate all the information and then when I finish this task, let's come together at that point.”
Mathews: Interesting. The point you brought up with tension between the standardization versus autonomy and then like you were saying just now with the not interrupting and preserving attention but then also the whole model of team-based care and the importance of co-location. And sometimes it is much more efficient to just turn around to your MA and say, "Hey, can you help me with this right now? or Hey, I have a question." I guess I'm just wondering how to find the right balance. And like you said, there's probably no easy way.
Dr. Harry: I think you're right though. I think it's a balance and I think it's not either or. I think the point here is being thoughtful. So, there's definitely a tension between availability and protected attention time. And the more protected attention time I have, the less available I am. And that's important to acknowledge for sure. But the more protected attention time I have, the more efficient I am. And so, in the long run, the more available I am. So, I think thinking as a team, can we take away some of the impulsivity in the way we communicate and save non-urgent things, batch them up and then have scheduled times to address those things.
And then if there is something that's urgent, obviously we work in a high-risk industry and we need to address those urgently, but I think having the conversation is a really good place to start. Because right now we don't consider our attention a resource, which it is and it's a very valuable resource. And so, we want to protect it and we want to pay attention to it and pay attention to where we're paying attention. And so, even just starting with that conversation I think is really important.
Mathews: So, it's going back to this idea that a physician's most valuable asset is their undivided attention. And we talk about that with our patients, the ability to give our patients undivided attention and not be distracted by documenting or by EHR tasks and things like that. But what you're saying is, it's kind of an asset in all parts of our work and life. Is that kind of true?
Dr. Harry: Yeah, that's absolutely right. I think what's really important to realize is that we've actually moved to what's called an economy of attention where people are sort of buying and selling attention and that's basically what social media is. And I think it's really important to think about attention as an asset. And if you think about it like an asset, you would budget it, and you would be aware where you were spending it and you would be thoughtful about where you gave it away. And we don't do that so much. So, if I came up to you and asked for half of your monthly budget, you wouldn't just give it to me. But if I come up to you and I ask for your attention, most of the time people give it away. And so, I think it's really important to realize that we are in the business of exchanging our knowledge, sharing our knowledge. We've all trained quite a bit to have that knowledge to share but if we can't give our attention, we can't share that knowledge effectively.
And if we can't give our attention, we can't keep that knowledge up to date. And so, our most valuable asset absolutely depends on where we pay attention to. And so, I think it would be really interesting to start looking at our physicians' and APPs' and nurses' attention as one of our most valuable assets that we really want to protect and be thoughtful about where we spend it.
Mathews: That is such a good point. I think that is an asset that's just as valuable as anything as our compensation or what not. So how can we shift towards thinking about our attention as an asset?
Dr. Harry: I think the first thing is just for individuals to be aware. So, if we think about our attention as an asset, you can even start by just paying attention to what you pay attention to. So, you start to think about it like your money that you would budget or any other asset you have and think, "I'm giving my attention away to this. I'm giving my attention away to this. Is that where I want to spend my attention? If I'm thoughtful about it and I think about my values and I think about what's important to me today, is this where I would decide to pay attention?" And often we don't do it that way. We're very reactive in terms of where we give our attention so that would be an individual intervention. And I think from the system standpoint, it's protecting attention. So, it's recognizing that if we give people protected space to think, they are going to do better and deliver higher-quality content.
Whatever they're delivering, it will be a higher quality if they have focused attention. And so, creating protected times or protected areas to think and to pay attention, such as we do in the OR, there's a lot of space for focused attention in the OR. And we don't recreate that many spaces outside of the OR. Brains are finite in what they can process. And the answer can't be just process more data because we're not going to be able to do that.
Mathews: We have to de-implement. And I think another part of it is the over-interpretation of a lot of the policies, where some new policy will be introduced or some kind of metric will be introduced that everyone thinks they have to follow and everyone will say, "Oh, it's the joint commission that requires it." But it's actually not and things build up and it just kind of piles on and on.
Dr. Harry: I think to your point, some of the AMA resources around that and some of the myth-busting resources that you guys have are incredible in that way because you're right. A lot of those institutional or sort of cultural myths that we run around with increase our task load and are not actually necessary. And so, I cite that webpage a lot cause I think it's very helpful.
Mathews: Yes, exactly. I could not have said that better myself. Well, are there any other pearls of wisdom or final thoughts you'd like to share with our listeners?
Dr. Harry: The only other thing that I think is interesting from the study is that we saw a lot of variability in task load across specialties, and we saw variability not only in the total task load but also in the different phenotype of the task load and the type of task load that each specialty is experiencing. And I think that's really important because it means that we need to address this differently for each specialty. And so, I think looking at emergency medicine and asking where the sources of extraneous cognitive load there of which there are many, is going to be really important. And then going into an individual family medicine or internal medicine clinic and saying, where are the sources of extraneous cognitive load here? And it's going to be different, and the interventions are going to look different. And so, I think this concept of customization of the interventions is going to be really important when trying to address these things.
Mathews: Absolutely. Well, thank you so much Dr. Harry for taking the time to speak with me today on this very important topic that is certainly one of the key things that is contributing to physician burnout. So, thank you for your time and expertise.
Dr. Harry: Thank you so much for having me. This was a lot of fun.
Outro: Thank you for listening to this episode from the AMA STEPS Forward® podcast series. AMA STEPS Forward® program is open access and free to all at stepsforward.org. STEPS Forward can help put the joy back into medicine by offering real-world solutions to the challenges that your practice is confronting today. We look forward to you joining us next time on the AMA STEPS Forward® podcast series, stepsforward.org.
Disclaimer: The viewpoints expressed in this podcast are those of the participants and/or do not necessarily reflect the views and policies of the AMA.