A Q&A with Amy L. Hester
WE ALL KNOW that technology can help transform healthcare, but we sometimes forget that no change happens without a large dose of human passion behind it. In observance of national Patient Safety Awareness Week (March 14-21), we sat down with former bedside nurse Amy Hester, who, charged with improving the care of patients in her UAMS hospital unit, worked with a colleague to “invent a better mousetrap”—one that today improves the lot of millions of patients all over the country.
Tell me about yourself—where’d you grow up? What did you study in school? When did you know you wanted to do this kind of work?
I was born and raised here in Arkansas. I grew up in Benton, graduated from Benton High School. So I’m a Benton Panther and proud of it.
I went to college straight out of high school. I did my prerequisites for nursing school at UA Little Rock, then finished nursing school at UAMS, graduating with my Bachelor’s in 1994. I knew I wanted to go back to school at some point, but right then I really wanted to get out and start making money and growing up and living life. So I did that.
What was your job?
I worked as a bedside nurse at UAMS for 12 years, from 1994 until 2006. Working bedside means you’re a frontline staff provider; you’re not an administrator, you’re not sitting in an office somewhere. You’re cleaning people up, starting IV’s, doing everything for the patients. Nursing is a passion, a love. It’s what we do.
And nursing is very physically demanding. But I had a car accident in 2002 and it really monkeyed with my neck, so it became increasingly difficult to lift patients. By 2006, I was continuing to reinjure my neck, and it was just not a good situation. Here I had a pre-existing injury from a car accident, and yet I’m hurting myself on the job. But in my heart, I didn’t feel like that was really a worker’s comp claim. I didn’t know what to do.
So I went to see my doctor—he’d been my primary care physician for as long as I could remember—and he sat me down and said, “You know, Amy, there are lots of things to do in nursing besides just bedside nursing.” And I was really angry with him—really offended. I was a really good bedside nurse, and I was helping people and touching lives. How dare he tell me to leave the bedside?
He actually called me at home after my visit. “Amy,” he said, “I know you’re really upset with the advice I gave you, but let me help you understand what I mean. When you go to work at the bedside, you’re assigned your five patients for those twelve hours and you touch those lives and make a difference in those five lives for that twelve-hour period.
“But there are other positions in nursing that have a bigger reach than that. If you’re a nurse manager or an administrator, you have a whole unit—you’re touching thirty lives every day, all day, twenty-four hours a day, seven days a week. Just look at the impact of care that you can have when you take that leap. So if you’re really the great nurse that you feel like you are, maybe you can do bigger and better things if you’re in a different role. And those roles don’t require lifting and tugging and pulling and doing all the things that you’re doing to hurt yourself today. I want you to just think about that.”
I did think about it, and he was a hundred percent right. So I moved into an administrative position as a unit manager. That was in 2006, which happened to be the same year that the Joint Commission—the body that accredits healthcare facilities in this country—released their first national patient safety goal around patient falls. So as a new manager, I was suddenly faced with meeting a brand new regulatory requirement for preventing falls and injury in my unit.
Before you tell how you did that, please talk about falls in hospitals in general. Why are they such a big deal?
It’s a good question, so let me take you back to 2006 when the Joint Commission made the National Patient Safety Goal to reduce falls with injury. Fast forward only two years to 2008, and that’s when CMS—the Centers for Medicare and Medicaid Services—made falls “a hospital acquired condition,” or a HAC. What that meant was that if you fell when you were in the hospital, and you experienced any kind of treatment related to that fall, your insurance company wouldn’t pay for it. So the hospital couldn’t bill for it, because it was the hospital’s responsibility to not let that fall occur in the first place. Now they were going to have to eat the costs of it.
Incidentally, in hospitals it’s not just older people who fall: The average age of hospital patients who fall is 52, and the average age of falls with injury is 52.5. And during COVID, we’re even seeing really young patients falling because of confusion and delirium related to not having enough oxygen.
The costs due to patient falls are a real incentive for hospitals to do something about falls. Even a fall that doesn’t result in injury has an average cost of anywhere from $3,000 to close to $5,000. The reason for that is that a fall often increases a person’s length of stay in the hospital. Your neuro or your ortho docs still have to come in to consult to make sure that everything’s okay. You’re still doing imaging, like CT scans to make sure there’s not a head bleed if the patient bumps his head. You’re doing plane field imaging to make sure his hip is okay, or his wrists or ankles or wherever he’s having pain is okay. So even in a non-injury fall, there’s cost.
But an injury fall becomes very expensive very fast. A patient who breaks his hip and has to go to surgery and have it fixed, and then stay longer in the hospital—that costs a hospital anywhere from $35,000 to $50,000 per incident. And not only does the hospital have to eat that cost, it’s now required to report such events.
Who do they have to report them to?
They’re required to report them to CMS, as well as to other agencies like Hospital Compare or Leap Frog that make that reporting publicly available so consumers who’re shopping around for where to get their care can be better informed. How a hospital performs in various “quality metrics” affects its overall rate of reimbursement and penalties from CMS. It’s called Value-Based Purchasing. If a hospital’s quality rating is high, CMS reimburses it at a higher rate.
CMS is a hospital’s bread and butter. So if a hospital gets penalized one or two percentage points because its quality performance is low, that’s millions of dollars just right there. Then if you have a staff member who’s injured while trying to prevent somebody from falling, the average cost of that staffer’s Workers Comp injury is $27,000. Then you lay litigation on top of that, because falls are the most commonly litigated medical malpractice event against nurses. Those can amount to millions of dollars per claim. So there are tons of regulatory and financial incentives for hospitals to really get their hands around falls.
In general, an average 200-bed hospital is probably spending somewhere around a million dollars a year on fall-related costs. And that’s assuming that they don’t have a big litigation case, like a wrongful death or something like that. So falls are very costly. They often change a patient’s disposition at discharge, so instead of a patient now going back home, the family has decided they can’t manage him at home anymore, and now we’re looking at a nursing home stint or skilled nursing.
You made your case—I get it now.
And so, back to 2006. My unit was particularly challenging because it was a neurology/neurosurgery unit. Those are the patients with the highest fall risk, and they have the highest percentage of falls with injury because they often can’t protect themselves if they fall. They may be paralyzed or have some other condition that just prevents them from doing that.
So that was how I entered into the world of starting to tackle falls. By 2008, we had pushed the needle a little bit, but mostly we had just wallowed around for two years not making much progress with our population. To my nurse heart, it was bad enough that these patients had had a stroke that was going to change their lives, but now they’ve had a fall with injury on top of it. If I really want to help my patients, I thought, we’ve got to find a better way.
That’s when my colleague, Dees Davis, and I started trying to figure out a better mousetrap for falls prevention in neuro patients. We weren’t trying to cure the ills of the world, we just wanted to fix the issues on our unit. In a neuro population, everybody’s a fall risk, and when everybody’s a fall risk, no one is. What we needed was a way to stratify who’s a low fall risk versus a high fall risk. Dees and I examined lots of tools that were readily available, and what we found was that they didn’t work for the neuro population.
They didn’t tell us the degree of risk, and they didn’t inform the clinician as to the specific reasons individual patients were at risk. Was it because of their medications? Their mobility? Their cognition? Was it behavioral issues? Was it how they were going to the bathroom? There were lots of different factors that contributed, but how do we inform the why so we can intervene and prevent against it?
We knew we needed a better way to do it, so I went to the College of Nursing and said, “We really want to reengineer falls prediction.” And they were like, “Nurses don’t do that.” I was, like, “Well, that’s funny, because all the tools are named after nurses. Morse and Hendrich and Conley, those are all nurses.” The Morse Fall Scale, which came out in 1987, was long considered the gold standard for fall risk training.
What is the Morse Fall Scale?
Developed by Dr. Jan Morse as her Ph.D. dissertation at the University of Alberta in Canada, it’s a scale that nurses complete on every admission, every shift, to determine if a patient is at risk. It looks at some very basic things: age, gender, do they use furniture to get around? Do they have an IV? At the end of that screening, you get a yes/no—this patient is at risk/is not at risk. But the sensitivity of that tool is pretty low, only 73 percent. So if you have a hundred people who go on to fall, that tool’s going to have missed 27 of them. We needed a more precise tool, one that wouldn’t just give us a yes/no; we needed a why and a degree of risk, because in a neuro unit, all of our patients were a yes.
So I decided to reach out to Jan Morse, who I found at the University of Utah. I spent days crafting an email, and when it was done it took me a couple of Crown and Cokes to hit the send button—after all, she’s The Jan Morse. But I finally did it. I basically said, “Dr. Morse, thank you so much for your landmark work. It has made an amazing difference in the world of falls prevention, but I feel like we need something different today. Healthcare’s just evolved and Morse doesn’t seem like it’s keeping up with the demands of patient care today. We really want to find a better way, but I don’t know even where to begin.”
She was so collegial. She sent me all of her handwritten notes from her dissertation, so I had all the work that she’d done up until then. How did she examine fall risk factors? How did she statistically analyze them? How did she model her prediction tool to come up with what was then the Morse Fall Scale? I basically taught myself the necessary statistics at my dining room table, using Jan Morse’s handwritten notes.
At this point, I knew I didn’t have the education to be doing this work. Not to mention that if the work was to be taken seriously, I had to have more than a BSN behind my name. And so, in 2008, I went back to school for my Master’s degree. When I got into the research classes, I was, like, This is where it is for me. It wasn’t clinical judgment, it wasn’t my nurse gut, it wasn’t based on my singular lens of experience of taking care of people. It was so much more than that. It was methodological, a step-by-step process that I could take to evaluate my work, to steer me when I wasn’t getting the answers that I needed.
I bridged from my Master’s program into the Ph.D. program in 2010, and by then we had already developed our prediction model, the Hester Davis Scale©. Because I had Jan Morse’s template and I knew the steps that I needed to take to get everything right, I was really crossing my t’s and dotting my i’s along the way. All of the data analysis and research was done at my dining room table, more than 1,000 hours in validation of our original model. And that was before the days that people had a second monitor, so I was trying to dump everything into that one Excel spreadsheet, toggling back and forth on the same screen. And finally, my son, who’s big into gaming, said, “Mom, let me show you something, I think this’ll help you out a lot.” And he brought his gaming monitor in and plugged it into my laptop, and I was in heaven. “Son,” I said, “you’re never getting this back. I’ll buy you another one, but you can’t have this one back.”
Tell me about the change in collecting data between ’87, when Dr. Morse was doing hers, and 2008 or 2010 when you were working on the Hester Davis Scale©.
Sure. Back in ’87, it was all handwritten documentation—Electronic Health Records (EHR) were in their infancy. But even the original validation that we did of our tool was on paper, because all the nursing documentation was still on flow sheets and writing and paper. Data collection was very tedious—it was me going into our medical records department and collecting the data, chart by chart by chart by chart by chart.
Then in 2011 we moved the tool into what was then our Allscript Sunrise electronic medical record platform. I worked with our Translational Research Institute at UAMS to build a bridge from the data warehouse into the backend of the tool in Sunrise. So now, every time a nurse entered an assessment, it was going into the data warehouse, where I could access it while preserving the privacy of the people whose data I was collecting. I didn’t have to break into people’s medical records anymore, so I didn’t know whose data this was. That’s what you want. The other thing was, it now allowed us to start collecting Big Data.
Even so, after we translated the tool into Allscript Sunrise, I wanted to make sure there wasn’t a difference in the performance of the tool if a nurse used it in an EHR or if they used it on paper. So we went through the trouble of revalidating the tool after we put it in the EHR, and it was the first time any predictive tool had been validated in both paper and electronic medical records. What we showed was that it was a stable instrument, regardless of how you used it. Not only was it stable, it was much more precise than other tools and had a higher sensitivity, and it answered all those questions we needed to answer.
Having that as a foundation allowed us to drive individualized care. I might have two patients who were at high fall risk, but the gentlemen’s reason was medications and mental status and the lady’s reason was volume electrolyte and behavior, so my care plan, as a nurse, was very different for those two individuals. That’s what the regulatory agencies want to see—that you’re individualizing your care, not cookie-cutter stamping the same care for all. Also, we were gaining efficiency because we weren’t wasting interventions on patients who didn’t need it; we were concentrating on those patients who did. The Hester Davis Scale© turned out to be a better mousetrap.
And we developed that before I even knew what prediction modeling really was. It was before machine learning and AI were a thing. We were collecting the right information, even though there was a lot of trial and error. We had access to a limited amount of data, so we had to do our best with what we had.
Today, of course, everything we do is in the EHR. And around 2012, UAMS transitioned from Allscripts Sunrise to the Epic Foundation system. Epic, which is based in Verona, Wisconsin, is the world’s largest electronic records system provider. They are the EHR service vendor for almost every academic medical center in the U.S. So Epic is just that—epic. It’s an EHR that services inpatient, outpatient, lab, long-term care, registration, billing, and research protocols. It’s more than just EHR; it’s really a system that supports healthcare.
When Epic heard about the work we’d done to validate our model in the EHR, they were very interested in that. They came down and said, “We want this model in Epic Foundation. We’ve looked at your validation work, we think it’s fantastic, no one else has ever done this. We want the best evidence-based practice in the Epic Foundation system, and we believe that’s you. We’ll pay you for it, it’ll be just a little bit and it’ll be one time, but….”
So I had to think pretty hard on that, because once it’s out there, it’s out there. But, oh my gosh, there are millions of patients on the Epic EHR platform. So we made the decision to go ahead and license the predictive component of our program, the Hester Davis Scale©, into the Epic EHR platform. And that was really a cathartic moment for me, a flashback to my physician’s phone call that night from years before, saying, “Amy, there are lots of ways for you to touch many more patients’ lives….”
Having our tool in Epic was basically a huge “Good Housekeeping Seal of Approval” for us. I retired from UAMS in 2018 to go full time to HD Nursing, which we had started through UAMS BioVentures in 2012. The University of Arkansas continues to make a royalty off of HD Nursing, so anything that we create and do and cultivate is a benefit to them as well. So it’s a really symbiotic relationship.
Now, 80 percent of our clients are also Epic clients. We don’t have a sales force at all—our website is our sales channel, and our clients come to us through word of mouth or from seeing our clients present at conferences. Epic’s been a good partner with us too, informing people that, “Hey, Epic has just part of the HD Falls Program; it’s great, but if you want the whole thing, you’ll have to go to HD for that.” So it was a really good move strategically for us, and it’s really helped our reputation immensely to have even a part of our work be licensed to and promoted by Epic.
It’s also been a really blessed journey for me as a nurse scientist to be able to be in the lives of people who never even knew I was there. All of the people we’ve prevented from falling, they have no idea of the care that was cultivated to make that happen. And that’s pretty cool. So at HD Nursing, we have an internal motto: “We save lives while we sleep.”