From 3M Health Information Systems
Podcast Episode Transcript: Clinical documentation: Easing the burden on physicians
Gordon Moore: Hello, and welcome to the 3M Inside Angle Podcast. This is your host, Dr. Gordon Moore, and with me today is Dr. Beth Wolf. Welcome, Dr. Wolf.
Beth Wolf: Thank you. Happy to be here.
Gordon: Dr. Wolf, you and I have spoken in the past about some of the work that you’ve done, and one of the things that you’ve done professionally is to move from direct patient care into a more administrative non-clinical role. Would you give me just a little thumbnail of your backgrounds and a little bit about that transition?
Beth: Sure. Sure. So my clinical practice started in 1999 as an internist and a hospital-based internal medicine physician. And I did that for about seven years, and then got interested in palliative care—that portion of my hospitalist practice that focused on goals of care and shared decision making. My involvement with the hospital system led to being on a lot of committees, and one of those committees was the medical records back in the day when we actually had paper records.
And so sitting on that committee, it became very apparent to me that although my documentation was about communication to my colleagues, there were a lot more people looking at the chart and trying to glean information from it. And with that, in the hospital that I was working in, got embedded in that process and really started to learn about translation and about coding.
And, as we have moved into value-based care and looking at how our outcomes are impacted by documentation, I’ve really emerged in that role. And, I would say, probably two years ago, about 50 percent of my practice was direct patient care, and there has been such a need to provide this translation and this education to physicians that I’ve moved primarily into that role in the last year.
Gordon: When I think about documentation, and coding, and stuff like that, it sounds incredibly dull and all just about the money. Tell me more about that. How is it not? Or how is it you can stand doing that, if it is?
Beth: That’s a great question. It is very tedious at times. For me, I guess, the background in palliative care and communication has always been important to me, and recognizing that when my words are not interpreted the way that I intend them to be interpreted, it can result in significant patient distress, certainly in the direct-patient-care world, but it can also result in a complete uselessness of the data.
And I think we’re living in a time when we really can actually start to look at this data and make real, meaningful, clinical decisions and interventions, and not just ones that result in Medicare payment or penalty but ones that really help us take care of our local population better.
I think we don’t know if we need a diabetic education clinic, or free insulin, or outpatient palliative care. We don’t know what we need unless we actually are able to study our patient population. I think that in an ideal world, we would enroll every patient that we see in randomized control trials, but that is just not possible. There are too many variables. But now we have an opportunity to capture those variables as directed by physicians using diagnoses into codes that can adequately track not only our current performance but really plan for the future.
So, when people say, “Oh, it’s just about the money; it’s just about the billing,” I don’t negate that. I think that we need to keep the doors of the hospital and the clinic open. We certainly don’t want to pay below average wages. You get below average staff, right? That’s not who we want taking care of our patients. So I think acknowledging that that’s a piece of the puzzle, but really recognizing that we, as physicians, are in a prime position to really make this data better and serve our population.
Gordon: So, I was talking to a physician about the requirements of coding and stuff like that, and he was really frustrated by it. And he was saying, “If I have a conversation with another physician, and I say the patient has congestive heart failure, we all know what we’re talking about. Why do I get beat up for that kind of thing?” And so what do you say to that doc?
Beth: I would argue that we don’t all know what we’re talking about. I think that my visual impression of the words acute respiratory failure, for instance, really demonstrate a picture of a patient who is breathing 30 times a minute, who looks in distress, who has significant hypoxia, and who’s requiring a decent amount of oxygen. They’re not on a ventilator, but they’re pretty sick. That’s my picture.
And I can tell you, I’ve spoken to pulmonologists, and they’re at both ends of the spectrum. I have a pulmonologist who won’t use that language unless the patient’s intubated, and then I have a pulmonologist who says, “Well, if there stats are 89 percent [inaudible 05:14], that’s respiratory failure.” And so, again, I think that when we use adjectives, when we use words like severe, I think we can kind of get that picture, but those words aren’t captured in the codified data.
And so, I guess, I would argue that we don’t all see the same thing. We weren’t all taught from the same dictionary in medical school. And for better or worse, the code book is not a dictionary. It’s like a phone book. It’s a catalog of numbers and words, and the coders aren’t allowed to read between the lines. They literally have to code from our black-and-white language.
Gordon: Why does it matter so much? It seems like we’re splitting hairs with a lot of this stuff.
Beth: Well, again, I think that we have the ability to transfer data across encounters, and, if we’re not painting a picture of the patient that can be abstracted in a way that can be analyzed, we’re really not going to benefit from the technology that we have because we’re not going to go into every chart and read the narrative, and to look for those descriptors, or to go into the vital signs, or read the nursing notes. We really count on that translation of the physician’s diagnoses into code so that we can classify patients correctly and provide them with not only clinical decision support in the hospital system, if you’re doing concurrent coding, but get them to the right specialist or discipline once they’re discharged from the hospital.
Gordon: Oh, that’s interesting. A light bulb just went off when you were describing this. I was thinking, of course, it’s in the chart. We’ve been talking about this patient. We have all this documentation, but then, when I think about diving into a medical record, I am swimming in this ocean of words, and all this toggling between screens, and looking at different parts, and it gets exhausting.
And so what I’m hearing you describe is this shorthand that tells me exactly what’s going on with this person with a degree of specificity where I’m certain that this is left-sided heart failure that’s acute on chronic, or something like that, as opposed to just congestive heart failure, and I have to dig, and look, and find the echo, and do things like that. So is that what you’re talking about?
Beth: Yeah. I think there’s just an infinite amount of data, and certainly, if you’ve experienced the paper world and the electronic world, you realize that when we wrote things on paper, we were very clear what was important to us because I wasn’t going to take 30 minutes to write a note. I was going to write down the key points of the impression on the CT scan, and then my assessment and plan was really the biggest part of my note.
And what we’re seeing now is that we have really this data dump into the note, and so the note is now four pages long before we ever get to the assessment plan, and the assessment plan is really very small and insignificant. Not only is it not robust, and in of itself, but, when we go up to look at the data that was pulled into the note, because it was just so encompassing, I don’t really get a sense of what the clinician was worried about. What were they using to make their decisions? So, again, I feel like we have so much data, but we don’t get the information that we need.
Gordon: Yeah. As I scroll through notes on patients, especially in academic medical centers, I’m reading through cut and paste templated notes that are just seemingly endless from med students, and residents, and senior residents, and fellows, and it’s like, “Oh, my gosh.” Some of the stuff I see in there is—it looks kind of like garbage that’s just forwarded because of this cut and paste and note recreation thing. How do we deal with all that?
Beth: Well, I think my hospital system, we’re sort of in the recovery phase from our electronic health record, which it’s been three or four years now, but we were totally on paper as physicians before that. And we’re starting to look at appropriate use of cloning. And there is appropriate use. I think that we need to count on our colleagues that are not physicians or extenders to be entering information into the chart that we can pull from, but, again, I heard this from—and I can’t remember where exactly—but they talked about being an author not an editor.
And I think I try to remember the most important part of this note is what I bring to it. And, certainly, I want to pull in relevant information, but I still have the power to tell this patient’s story. But I think note bloat, those types of things, are very real. When I go around and talk to physicians, it’s commonsense. “What would you want to read? Make your note in that vein.”
Gordon: So, as you do this work, as you’re teaching other physicians how to be more accurate in their documentation, what are the kind of things they talk to you about pushback on and how do you then deal with that?
Beth: Sure. As a I go throughout the country in a different stage of their evolution and reaction to what we might consider the bureaucracy of medicine, and I think that, as we first discussed, I get a lot of “Well, this is just about the hospital and their billing.” And I would say that, universally, we’re moving to a point of being more aligned in our care of patients, and I think that certainly—and, if we have continued hospital payment separate from physician payments, whether it’s medical necessity on the front end or clinical validation after they’re discharged, there will be that disconnect.
But I do think that once physicians buy into the fact that this is important, right, they want their patients to look as sick on paper as they are in person, and so we get an acceptance of that piece of it. What I hear is that—I’m focused on this one problem. So, if I’m a vascular surgeon, and I’ve got an ischemic foot, I need to get him in there. I’m not going to be up with the latest criteria for sepsis, and so I need some help with this. And, whether it’s an extender, or a physician assistant nurse practitioner role, or whether it’s CDI at the elbow, I think they’re acknowledging that it’s important, but then saying “Listen, this is not my area of expertise. How do I get someone in here to help me with this?”
And, as you know, I cannot document for them. I’m not a hands-on provider. Even though I can look at the criteria, and I can say, “Well, gosh, you’re certainly treating sepsis, and this patient sailed through the hospital stay because you reacted quickly, and you did a great job, but you didn’t document that they had sepsis and severe sepsis, and so you’re not showing how sick all your patients are.”
So, when that one person with sepsis does die, it looks like they’re an outlier. And, of course, the patients who do poorly, we document very well on. We have a laundry list of problems, and we talk about how sick we are. It’s the patients that do well that had those underlying issues that we don’t always capture the picture of.
Gordon: That’s an interesting aspect that you just raised. It’s the mismatch of the perceived illness burden of a person against the reality that could maybe flag somebody as an outlier, to say this patient should not have died given the weight of what you presented. And the presentation of that information is to the coders ‘cause we’re still using encoded data as a way of judging performance in the value-based world—or not. Even if you service Medicare, I guess, the star ratings are going to be based on “Are more people dying on your watch in your service than would be expected?” And the expected, I would guess, is going to be based on the illness burden. Is that true?
Beth: That’s exactly right. I think that when we talk about documentation and thoroughly representing your patient, we’re talking about managing the denominator, okay, the expected. The numerator, that’s a whole other issue, right? If you have a problem with a numerator, then you have to look at your protocols and practices in your alerting systems, in your actual clinical care, right?
But what we’re talking about documentation, we’re talking about managing the denominator. Like I said, when someone does poorly, there’s no paucity of documentation. There’s a lot of explanation about what’s happening. It’s the patients who are sick that do well that oftentimes we don’t reflect their level of severity of illness or risk of mortality.
Gordon: And I guess it also comes back to some of the payment where there’s more work in gauging a patient with a high-illness burden than there is one with a low-illness burden, and that if you think about value-based payment, the concept is you’re working harder, so the hospital should be paid commensurate to the amount of work—what’s the work burden? And, if we’re not representing that accurately, that’s going to have an impact on payment as well.
Beth: Yeah. It’s going to most certainly have an impact on outcomes. If you don’t have the right social workers, and discharge planners, and nutritional therapists, those folks—you don’t get paid more for having those with most of the prospective payment systems, right? So you’re tasked with putting the right people in that patient room that’s going to give them the best quality of care. Because, at the end of the day, quality care cost less. You don’t have the complications or the negative outcomes that are untoward, not unexpected.
Gordon: So you’ve made the shift— because you were starting to work in palliative care and looking at the signals in the documentation, and is this something that you do you full time now? Is this your whole job?
Beth: It is a large portion of my job. I still am on the active medical staff and on the roster to fill in and to pinch it, which I’ve done several times this year. But I think that for me, if I do something, I want to be really good at it, and I think that when you start to move away from direct patient care—I can’t say that I provide care that’s not up to standard. I just think it takes me longer. I’m not as efficient because I have to look up the latest chemotherapy drugs to talk about the burden versus benefit.
And I think that, for me, it was a very difficult decision. I think I consider myself an operational person because I’m “boots on the ground,” even though I’m doing non-clinical care. I, of course, recertified for everything in 2018, so I’m good for 10 years. But, if I go back into it full time, there’s going to be a learning curve for me. And, I think for any of us that decide to focus our energies on the bigger picture versus that individual patient care, you have to figure out what’s going to give you satisfaction over time. And, for me, again, obviously I want the doors of my hospital to stay open. I want to be able to be taken care of there when I get sick.
But again, for me, I’m most interested in how we really impact the overall care of the patients in our community, and I think that we have just limitless data that we can use to do this. But physicians, again, we need to be in control of it because we actually know what the patient has. So, for me, that’s my passion for it, but it’s not for everyone. I’m a firm believer that you always have the right to change your mind. You do what you’re drawn to and what you love, and, if you don’t love it after a period of time, then you can do something else.
Gordon: How long have you been doing this?
Beth: Since 2013 is when I got my certified professional coding certificate, which was the hardest part of this, by the way. They have a very difficult job, and it’s very tedious. And I think, from that point on, I became more involved in my local hospital system, and I work about 900 hours a year for our medical records department in many capacities. One of which is documentation.
But I’ve also started to do some more education in hospitals outside of my area. Which, I think, part of doing that is seeing how other people do things and not feeling necessarily alone in your struggles, and then obviously being able to share some of the best practices that you see across the country.
Gordon: When I think about the dissatisfaction that I hear about from colleagues in practice, a lot of it is focused on the electronic medical record, and a bunch of that is focused on documentation burden and coding. What is the future—I get the argument for we need to accurately represent what’s going on with the person, and we need to do it with a degree of specificity that truly represents this individual so that we know compared to the next person. And, therefore, we can do all sorts of things around measuring quality, etc. So that’s great. But it seems to be such a burden when I talk to colleagues. What do we say to them? What do we do for them? Is it just more of the same in the future? Is there any ray of hope?
Beth: Well, there’s always a ray of hope. I can’t negate many of those comments. I think that we’ve been asked to do things that really are not necessarily within the original scope of our practice. I think a colleague of mine equated the electronic medical record as the self-checkout in the grocery store, right? It seems like it should be easy; it seems like it would be more efficient, but I’m not a checker. I don’t know how to put my groceries in the bag, and I never get through the line without having to ask someone to come and help me.
And so I feel like physicians feel that way about the electronic record, and obviously there are so many important things about it. I really do want to be efficient and accurate in placing orders, and I think handwriting was a huge issue, a safety issue, at making sure patients got the right orders. I think we’ve traded some of the technical errors, and we’re seeing that—in the electronic record, if you can write a horrible note on paper, you can write a horrible note in an electronic record. You can just actually read it.
I have to believe—again, coming from a generation that did not have cellphones—that we’ve come so far, and I just I have to believe that we’re going to be better with the electronic health record, both longitudinally, with artificial intelligence, with translation, but it’s a team effort.
We need to start having codes that pick up our language, like, okay, heart failure, you mentioned; stage one, two, three, and four heart failure, if you tell a physician that, they know what you’re talking about. Those have functional definitions, but we can’t capture those in the codified data. So I think it’s a constant effort to make sure that the language that I use is the language that can be picked up, not the other way around, right?
So I sometimes I have to go tell physicians, “Well, if you want to get credit for how sick the patient is, then this is the language that you need to use.” Well, that’s frustrating, right? Because maybe it wasn’t what I was taught in medical school, or maybe it’s just behind the times, and the code book hasn’t caught up with that.
I think it’s going to be possible. I think that technology has to help us do that better, and then, again, looking at—we have scribes. We have physician assistants and nurse practitioners that have helped with that. We have more robust consultative services so that the surgeon isn’t the one having to document sepsis. I think that it will be a combination of solutions, but it’s not going away.
And I think, for me, one of the things that will probably change our note writing the most, in terms of the content and readability, is going to be access to information. When patients are on their portal immediately after discharge, and they’re able to see my notes and the consultant’s notes, we’re going to get better at writing them in a way that communicates and that’s understandable.
Gordon: So I still have hopes in technology. I’m frustrated by the gap between the promise of technology and what it currently delivers, but I see advancements in—or the ability of machines to capture language in a way that is less error written than when dictation software originally came out and had to be laboriously trained. Now, my smartphone can pick up stuff and correct pretty darn quickly. So I have hope for how that can work in health care and ease some of this burden.
And I also have hope for the ability to clearly capture a lot of information. Early in our conversation, you were mentioning the ability to study what’s happening to people. I think that’s one of the promises of codifying data around people. What we say turns into codes, turns into the ability to aggregate information across many people and do those kind of studies, and so that holds a lot of promise, so I think that will be good stuff as well. So what’s your future? Do you see your role evolving over time?
Beth: I do. I think that I see it evolving into really working with software, and all these data points that get entered by non-physicians, those have value. They can point to particular outcomes and particular risk factors. And so I think that—this was very difficult for me to take a spot in medical school, right, old school thinking, and say “Well, I’m not going to use that to take care of patients,” right? I took someone else’s spot.
I think it took me a long time to realize that if we’re going to make this environment better, you have to have physicians who are willing to play a part in the development. And it’s not for everyone. It can be tedious. But I think that we play an important role in straddling that world. And I think, for me, maintaining contact in my own hospital system, where I meet with physicians that I’ve practiced with, who are certainly not shy about telling me what they think, is helpful. Because you lose a little bit of that as you move away from direct patient care.
And again, you can’t be great at everything, but I think asking open-ended questions, being a good listener, those are all things that those of us who have moved into non-clinical roles—and we’re still trying to represent the clinical roles—we need to be very attuned to. And I think that you never have fresh eyes again. You never learn the electronic medical record for the first time again.
And so when you move away from that, in time, you forget what that was like, and you forget how you felt, and how you saw things. And part of being administrative and operational is being able to effectively communicate what the patient-facing clinicians are feeling into the world of technology or regulatory decision making.
So I hope that I can continue to do that. I hope that I can continue to be a good listener. Because as I move out of the clinical world, I’m not going to have as direct a perspective, but I do think there’s a role for it, and hopefully I can make things better, not just for one patient but for a population of patients.
Gordon: Tell me, you mentioned team before—and it’s something that struck me as well—is looking at the cadres of individuals who, like the CDI nurses and coders, how is it that we need that team and how do you—what are your recommendations on how teams like that function?
Beth: It’s funny that you mention that. But my first experience of being on a team was being on the palliative care team, right? So, in medical school, it’s a pyramid, right? The physician’s at the top, and so, when something goes wrong, the buck stops here, and you, many times, felt alone in that. And I think when I joined the palliative care service at my hospital, after I got certified, it was fascinating to me because when we sat down to meet with the family and a patient, I had a social worker with me; I had a Chaplin with me; I had the nurse practitioner with me; sometimes the nutritional therapist; sometimes the pharmacist, and I was not necessarily the most important person in the room besides the patient. The patient was maybe for more interested in the spiritual professional.
And so that was one piece of it to realize “You know what? I don’t have to own everything that’s happening to this patient,” right? “I can utilize the highly skilled people around me.” The second thing about being on a team was, after we left the room, we literally did a 360 debrief, and I sort of envisioned the circle, right? People gave me feedback just like they gave everybody else feedback. And at first it was a little uncomfortable, right, because I wasn’t quite sure how to take it, but, at the end, it actually made you so much better.
That aspect of being on a team, I think, has been incredibly transformational for me. And I think there’s a reason that professional teams have coaches, right? I think that you need that feedback, and I think being on a team lends itself to making everybody better. And so I took that aspect of my professional patient care practice and really have tried to apply that in other aspects of what I do now.
Gordon: I’m inferring from our conversation that you would say that that kind of team is important in how we document, and how we code stuff, and how that information flows.
Beth: Absolutely. To reference back to the meeting I had just this morning where the vascular surgeon said, “I can’t do this alone. I need help. I understand that it’s important, but I need help.” And so, again, yes, how do we hardwire the people to be in place so that he can do his job efficiently and effectively, and we can still capture all these other things. So 100 percent.
Gordon: I think about my background as a family physician. I remember as a resident rotating through pediatric oncology and thinking “Thank goodness, there are people who feel like they have the calling to do that work because it is way beyond me.”
And I think, as I look at the coding world, it’s like, “Oh, my gosh.” The degree of detail and the stuff you have to know to do that well. It’s like, “Just, thank goodness, there are people who are good at that and who want to do that, and then I can collaborate with them, and they can do their thing well, and I can focus more on the clinical stuff that would be more interesting to me.”
Beth: Yeah. Absolutely. And, when I go and talk to other physicians, I don’t pretend to have a handle on all of this, and that’s why I really do talk up the other people involved in this process because they are experts in their field, and we need to treat them as experts, and we need to surround ourself with them.
Gordon: Then hopefully technology, then, helps in terms of easing the communication and making it seamless. So that’s the promise. Well, Dr. Wolf, any last comments or thoughts?
Beth: No. I’ve enjoyed talking with you.
Gordon: Thank you so much.
Beth: You’re very welcome.