Back to THE HANDOFF
Hospital Operations

Episode 110: Beyond the Surface: The Hidden Potential of Ambient Data

October 18, 2023

Episode 110: Beyond the Surface: The Hidden Potential of Ambient Data

Listen on your favorite app
October 18, 2023

Episode 110: Beyond the Surface: The Hidden Potential of Ambient Data

October 18, 2023

Dani:

Welcome back to the Handoff from Trusted Health. I'm Dr. Dani Bowie, Chief Nursing Officer of Trusted Health. This season we're covering all things healthcare innovation and the importance of fearless leadership for change. Today I have the pleasure of speaking with Dr. Dana Womack, healthcare informatics consultant. We discuss her pivotal career moments before diving into the power of harnessing ambient data or digital dust and turning it into clinical practice insights. We also cover the concepts of resilience, engineering, and adaptive capacity in nursing. Here's my conversation with Dr. Dana Womack. Welcome to the Handoff. I am excited to be speaking with Dr. Dana Womack today.

Dani:

Dana, welcome to the Handoff.

Dana:

Hi Dr. Bowie. It's great to be here with you.

Dani:

Well, let's get started as I've been doing for this season, really centering us back to the patient. Dana, I would love it if you could share a story that's had an impact on you in regards to the patient experience.

Dana:

I'll go back to when I first started nursing, and as you're a excited new nurse, you want to make sure that your patient experiences are really smooth. And I remember a particular Saturday morning when this really didn't happen. I had a patient who I realized was being transferred to a long-term care facility when I saw her on a stretcher in the hallway with an ambulance crew next to her. And this was a surprise to me. And so I was like, oh my gosh, I wonder if her tuberculosis screening is done. I know that's a prereq for them accepting her. So I went to her chart and sure enough, it hadn't been done, so I had to go tell the guy, sorry, we need to put her back in bed. And then I had to fire off a message to pharmacy, can you send me a tuberculosis skin test right away?

Dana:

I had to ask radiology to work her in and then have the crew come pick her up again a couple hours later. And it was uncomfortable for her to do the extra transfers. It was a lot of chaos in the day, which was completely unnecessary, had the system fully recognized and we had done the prep prior to the weekend. Even if somebody had tipped me off and report she was leaving, it might have helped. But I went to my manager then I said, Hey, I said, I'm seeing these things, and it was her response that really has stuck with me. She said, well, Dana, what we need is more conscientious nurses like you. And I was young conflict avoidant. I knew she meant it as a compliment, so I just kind of smiled and left. But I was very unsettled. I was unsettled about my future as a nurse because I realized that what she was saying is our primary strategy is to count on you, the nurses, to overcome broken systems and to create positive patient experiences despite that, and that has stayed with me. I rally against that notion to today, and it really has continued to be a motivator throughout my career.

Dani:

There is a lot of that in healthcare that has implications on the patient, implications on the caregivers and the whole care team. I mean that's no easy feat though. It seems like a small thing of like, oh, you need to get this TB test done. There was multiple players that needed to make that happen, and the coordination, the effort that took was paramount. And I can see how that has shaped some of your career and who you are. And that's what I want to talk about next is you as a clinician and a leader. So can you share with some of our listeners today about any pivotal career moments or highlights that have been impactful and explaining where you are today as well with how you're practicing?

Dana:

So I spent two years on that first med-surg unit in a community hospital. And after that time I decided that I really wanted to understand if the frustrations that I was experiencing as a nurse were unique to my small town facility or if this was more widespread. And so I happened to meet up with a fellow nursing classmate and we were like, Hey, let's go be travelers. So we did for a couple of years. I would say that my experience as a traveling nurse, it was tough, but it was really, really informative and educational because I realized that these situations are pretty universal and that all healthcare facilities kind of struggle with them. And somewhere along the way, I'm not exactly sure where I came across the field of nursing informatics, which back in the early nineties was fairly new. And I'm like, wow, if we could apply better systems to this, that's potential for change.

Dana:

And so I'm like, I've got to switch the focus of my career. So I took a traveling nurse assignment into Salt Lake City and then switched to grad school and completed a master's in nursing informatics there intending to stay in healthcare, but I was recruited to industry afterwards. So I went to Intel Corporation and worked there for a couple of years and along the way, actually I'd taken elective courses in my master's program from the M B A program and in new product development, and I discovered my love of new product concepting and that was kind of what really took me into industry. After working at Intel, I deployed VRA probably in a dozen hospitals. That was impactful because I saw how immediate technological change can impact the work experience of nurses. If it's implemented well, it can be for the better. If not, it can be for the worse.

Dana:

And so it really emphasized the importance of good system configuration and training in deploying systems. Later I had the opportunity to support the office of the national coordinator for health IT during the of electronic health records as a Deloitte consultant during that time, participating in national initiatives was really exciting, but it was a little less hands-on than I preferred. And so when Intel recruited me back, I jumped at the chance and came back to the Pacific Northwest and then later earned a PhD in informatics at O H S U where I focused my dissertation on harnessing ambient workplace data to sense periods of overload at the bedside. Again, kind of coming back full circle to where I had started. So I currently split my time between applied research in the hospital and community settings and then innovation focused independent work in the industry.

Dani:

I love your background and as you're describing more about it, I think it's so critical that the profession of nursing is going into some of the paths that you have paved. There's a lot of controversy around the travel industry, but I've often said that travel nurses bring such a wealth of knowledge because of that exposure to many health systems and they see how one place has a best practice in a certain instance and another place may benefit from that. And so I've always encouraged managers, if you have a traveler, you actually have someone who has this exposure to the health system in a very macro way that can offer some insights. So I love that augmented with all of your experience at Intel, the academia as well. It's just amazing. We're going to talk a little bit more about this ambient data and digital dust and your doctoral work. And so let's shift gears and can you help explain to our listeners more about the use of ambient data, digital dust, what that is and how health systems can use that to improve the experience of the frontline clinicians and the experience of the patient as well?

Dana:

So when we're nurses at the bedside, we touch so many electronic systems and each one of these gives off constant streams of what you can think of as digital dust. And examples would be communication devices, nurse call medication dispensing, electronic doors that you swipe in and out of the electronic health record and many others. These data serve an immediate purpose like dispensing a medication, but then the systems are always logging these events to a behind the scenes backend log system that captures data about this event. So the beauty of

Dana:

The ambient data or this digital dust in the workplace is that nobody has to do anything extra to capture it. It's already produced, they're automatically generated, but currently we do very little to capture and turn these data into information and then knowledge to improve nurses', work experiences and healthcare operations. What can we do with this dust? Dust? We can harness multiple operational data streams to begin to tell the story of a shift, like the context of who is there, which nurses were assigned to which patients, the activity that is taking place. A lot of people would assume that hospitals would be able to produce data about this. Really a lot of our systems are very patient and encounter oriented rather than nurse workplace and work shift oriented. In order to convert these building up from the digital desk, from data to information, you have to derive summary features from the data streams.

Dana:

For example, you can calculate the count or the average duration of a nurse call and then how evenly or unevenly these nurse calls were distributed across the patients. That's just a couple of examples. With the single data stream altogether, you can calculate hundreds of summary metrics from naturally occurring data streams and then you can utilize these summary activity metrics as input into machine learning algorithms or other AI algorithms to begin to identify patterns that are predictive of an outcome of interest. You mentioned my dissertation work. In that exploratory work, I demonstrated that it's feasible to predict unplanned overtime as a proxy indicator of strain with moderate accuracy four to six hours before the end of a 12 hour shift. And what's nice about that is it would afford you time to intervene and level load nurses work before shift end. So the potentials there. Our environment is giving off these digital dust and digital whispers about our working conditions, but we need continued r and d to convert these into insight so that we can intervene before negative nurse and patient consequences occur.

Dani:

What I find interesting about this concept is often you, at least in the profession I've had people say it's so hard to say what nurses do to describe it, what is a nurse doing? And rightfully so because what you just described is like they're touching all these systems, but what you just presented here was an opportunity to create a really objective story around the work of a frontline clinician and what they're doing with how they are using all these different systems to deliver care. And there's so much opportunity there, and even as you mentioned, being able to moderately predict what's going to be indicators of overtime, I think you said was it four to six hours into the shift or was it four to six hours with the shift ending that you could see before the

Dana:

End? Yeah,

Dani:

Before the end and then as you said, readjusting some of the workload or interventions to offset that, and that is just not how we operate and it is usually a reactive space and something that we look at the lagging indicators of over time and premium, things like that. This is something that I think I'm really glad that you're sharing with us today and our listeners because I think that there's such an important need around harnessing the digital dust to create the story even of the work that we're doing. And then as you mentioned, I think the ability to intervene proactively in situations to help the clinicians and the health systems in general. Which kind of brings us to our next topic, which is resiliency. You have some expertise and interest in resilience engineering, and I would greatly appreciate if you could share with our listeners how that differs from individual resilience and how we should be thinking about that in our health systems today.

Dana:

Yes, it's really interesting to me how the term resilience has almost become a negative term in the healthcare space. I think it's because it's almost synonymous with personal coping and psychological resilience and it evokes mental images of self-care, meditation, yoga, and all these things are super important and totally support them, but when you think of resilience engineering, it's looking at the whole system that people operate within and recognizing that that system needs to be resilient as well. I think the reason that clinicians recoil when they hear the word resilience is because hearkening back to that original story and what my nurse manager told me about just needing conscientious nurses, they're kind of hearing the message that they need to be personally resilient to overcome a broken system. So resilience engineering has its roots in safety science. The term was introduced in 2006 to represent a new way of thinking about safety in multiple industries including healthcare.

Dana:

There was growing recognition that rather than seeing the root cause of events as a single point of failure, that these events needed to be explained as combinations of conditions that prepared a path to an error or an event that was unwanted. Historically, safety has been focused on counting incidences and the number of things that go wrong, but resilience engineering defines safety as the ability to succeed under varying conditions. For anybody who wants to learn more, I'd recommend a book by Dr. Robert Weirs called Resilient Healthcare. And in that book, he defines organizational resilience as the ability of the healthcare system to adjust its functioning prior to during or following changes and disturbances so that it can sustain required performance under both expected and unexpected conditions. So if you're an organization that practices resilience, you're putting more effort towards understanding everyday clinical work and how it is that work goes right, and recognizing that this augments your awareness of how to create safe organizations in addition to the continued accident and regular investigations that you're doing.

Dana:

It takes a system-wise perspective. It sees safety, quality, productivity, satisfaction and more as all facets of the same reality. So there's a lot to say. I get super excited about this topic, but I'll highlight one more. Healthcare organizations are complex adaptive systems as opposed to factories. Things are done differently and they have to be, it's not like we can do the same process every time through studying nuclear power plants and many other mission critical systems. The field of resilience engineering has identified three common patterns of failure in complex adaptive systems and one of these, the pattern of decompensation offers a really good explanation of how it is that nurses buffer a system and keeping things going despite challenges. So in decomposition, as pressure enters the system, and in our case we're thinking of nursing work systems, the system sustained its performance by utilizing its adaptive capacity.

Dana:

So nurses are compensating and they start to do this by doing things that have lower consequence like moving their break or skipping their break or spending less time talking with the patient's family. But then if the pressure continues, their adaptive capacity erodes and then they have to start making much tougher trade-off decisions like needing to forego things like oral care or patient mobility or something that's much more impactful or even to the point of asking, getting an after the fact cosign for a medication. I think this is a really useful pattern for us to recognize and for even to have nursing leaders think about talking about, we don't really talk about what we do when there's more work to be done than time available. There's research done on missed nursing care, but there's not a lot of guidance to nurses or it's nurses are hesitant to even talk about the fact that they're missing nursing care. And I feel like there's a lot of opportunity for healthcare to be a lot more open about this and to even collaborate and figure out, well, what are we going to do when we've exhausted our adaptive capacity and recognize that it's not something to hide, but to highlight and to address

Dani:

What you're mentioning here too is resiliency is often the individual perspective that we take as leaders in health systems, but in fact it's something much broader and the decompensation that happens as you're dealing with a broken system and then working to prioritize, do these systems, as you described, studying this from other mission critical areas and your knowledge in this space, did they provide some, here's how to prioritize, do they provide, here's how to solve the system or in a system that you're continuing to solve, you may still deal with some of this compression issue that then causes you to prioritize and here's some ways to think about that because we don't in nursing say, here's how to really prioritize the things that you're dealing with, your breaks, the oral care, whatever it is, it's like it is kind of intuitive and you're just managing it autonomously and then we don't talk about it. Do you have any thoughts about that?

Dana:

It's going to be different for every industry and potentially different settings, but a core central theme is recognizing your current point of operations. Are you in a point where you have adaptive capacity? Are you at a point and it's like it's time, workload, financial constraints, all of those are putting pressure in different ways. Are you drifting towards a safety boundary towards a workload boundary towards what is your dynamic point of operations and realizing that that's moving around right now we don't have a lot of ways to understand when our adaptive capacity is eroding and in this way, I think just maintaining adaptive capacity and having some reserve is a compliment to things like lean, where you're working out all extra waste in the system, but you still have to maintain your ability to adapt when people start to adapt. We theoretically could see some of these adaptive behaviors in our data delaying certain things.

Dana:

They're going to delay a daily thyroid before they'll delay an IV vancomycin, they'll delay oral care before they delay somebody's cardiac arrhythmia. It's like there's kind of like these weak signals that together can start to paint a picture of where the pressure is in the unit. I'd just like to highlight this because I've even encountered it where it's like when you start looking at things like delays and missing this and whatever, nurses get really nervous that you're looking for the bad apple, but we have to create an environment and a culture where we're all in this together and we recognize that you can't do impossible things and we recognize that and when impossible requirements start to pop up that there's the psychological safety of being able to talk about it and bring it forward. And I think that some combination of sensing in the environments combined with humans and they're able to fill in and kind of sense make on top of all of these digital signals that we can bring forward that that could be a potential solution for addressing these situations in a different way in the future.

Dani:

Culture of inquiry, I've been speaking with leaders about that and there's components of culture of inquiry, which is psychological safety and being able to be curious and be safe in that space, these digital ambient data that could be so powerful and harnessing the information and in reframing it that we're not looking for the low performers or the bad apples, but recognizing to air as human and what systems are in place that can stop that or what do we need to do with the systems to make them better. I do think that that's an important component is that psychological safety to be able to address this, which can be tough knowing. I mean, medication error can be deadly and there's so much implication around this, and we were seeing with the Vanderbilt case where the medication error and how scary was as a nurse to watch that with a fellow nurse who made an error, but why was that error made and how can we do better next time? We're talking a lot about the ambient data. Let's talk a little bit more about another concept called digital whispers, and maybe that's the same thing, Dana, I'm new to this context, so least explain to our listeners also about the concept of digital whispers and how digital whispers can be applied to nursing.

Dana:

It relates to the ambient data in healthcare. We really, really don't like uncertainty and ambiguity if you don't like risk. And so we like clear decision support like patient is allergic to penicillin, don't order this, or we consider a new treatment to have low evidence until a randomized clinical trial has been conducted and all of that. But when it comes to operations, we can't always avoid the ambiguity. We work in these messy, complex adaptive systems and we need to pay attention to signals that have perhaps lower signal to noise ratio than we'd like. Sometimes the same actions that could help a situation go are the same actions that could make something go wrong in another situation. And so I think that just advancing the idea of digital whispers, we're not seeing that. We're not even claiming that we can use data to say, you should do this or that in order to get your shift on track because I don't think that that's potentially even the right way to approach it at this point in time.

Dana:

We're seeing we can help harness digital whispers about your environment and then we need intelligent humans charge nurses, frontline nurses, managers to be aware of these as they're coming in real time and use them first to create the observability and then to sense the status of the working conditions and then go ahead and be proactive about it. I think that this idea of human and machine monitoring of workplaces will be a much larger focus in nursing in the future. I have an industrial engineering colleague, Dr. Lindsay Stickie, whose work I admire very much, and she also shares this vision. And together we've identified some digital whispers or weak signals that support insight. For example, the sum of communication minutes and the percent of medications delivered via syringe and the skewness of meds dispensing across nursing was predictive of unplanned over time, and so it was machine learning that helped us identify these patterns.

Dana:

But there's also pragmatic explanation and we've conducted qualitative studies with nurses to confirm this. So in this example, when you need to sort something out or there's a challenge or a logistical problem, you need to start communicating, and so that's going to increase your communication minutes across the shift. The p r n example is like those are often used to manage symptoms, and so if you have a lot of patients with a lot of symptoms that you're managing, your dispensing is going to go up in that category. To the point about skewness in medications across nurses, if you've got a patient that has a really heavy medication load or even if you have differences in assigned patients, those are going to kind of show up in your medication dispensing data. So those are just a few examples of this kind of informal term of digital whispers that I think would be really helpful to incorporate into our everyday monitoring and management in nursing.

Dani:

As you described these examples, I'm like, oh, yeah, yeah, that makes complete sense, right? Increased communication, which is potentially could be a patient situation or just something happening that you have to respond to. We haven't really been trained to think that way and to think about it in terms of these different workload components and the increase or the decrease related to those. Then leading to what's happening in the environment, there's this digital dust everywhere and it requires analysis, curiosity, and understanding compared with the human intervention. So machine and a human coming together for the sake of improvement and tying to the resiliency of an organization to make it that much better. Now, Dana, what are your thoughts on the future of nursing? Is there anything that excites you or worries you as we were talking about present and hopefully then what do you think about the future and the lens that you have, which is so unique in perspective?

Dana:

I think my primary concern is that nursing is not a long-term sustainable career for people, especially at the hospital bedside. We're losing nurses that recognize muddling through unworkable conditions is draining their wellbeing and putting their licenses at risk and creating unsatisfactory patient experiences. This might be a little more controversial, but I'm also worried that the current focus on mandatory ratios might hamper innovation and care model redesign. I totally, totally support the thought and the intention behind the ratios. I just want to see us to be able to continue to innovate in our care models and try different things. I'll offer an example from a hospital that I work with and during Covid there was a lot of quick juggling of roles and responsibilities and they recognized that people who place patients for surgery have particular skill in that area, and they created these proning teams that went around and went through units and helped prone patients, and it was just this incredible relief to the nurses to have this team come in and do that.

Dana:

There's a lot of debate about team nursing versus a hundred percent RN staffing and what should be done if you have a mandatory ratio and then as a result of increasing costs, they take away some of the ancillary services that doesn't really help nurses either. And so I really am concerned about that. We can continue to innovate in this space, but there are some things to be excited about. I am excited about the application of data science to increase the observability in the world. There's dynamic forecasting down to the hour and really small geographies and these technologies are out there and creating big changes in other industries, and I think that we can harness them and use them nursing as well. And I think our notable opportunity right now is just really to expand our thinking beyond staffing is the problem to, yes, staffing is a problem, good staffing is necessary, but on top of that, we can take steps to recognize and respond to strain whenever it rises. Realizing that burnout can happen in punctuated moments across many shifts as well as this kind of chronic stress that people experience in healthcare.

Dani:

Often I think bad staffing outcomes is just a symptom of something that's going down, which you kind of allude to. I would be in the same boat with you around the ratios just because I understand the place and the intent, and we all want great staffing, but it is a bit more complex than a number in the sense that even what you just described, all these different components that lead to a patient experience and the nurse or the clinician being able to deliver care and the systems in which they're using to do that. When you think about innovation and change, I kind of cringe though. I believe in it and I want to be a part of it, but oh, the work that goes into innovation and change, it's so much. So I'm really excited about what you're sharing and hopeful that this continues to evolve over the next 10 to 20 years and excited to see the work that you do as well leading in this space. Dana, where can our listeners find you?

Dana:

My email address is easy to remember, dana@danawomack.com.

Dani:

Perfect. Perfect. And you provided some really great literature for us to review authors and books and things like that, but listeners, if you're hearing it, feel free to send her an email. Dana is amazing and has so much knowledge in this space. Finally, Dana, what would you like to hand off to our listeners today?

Dana:

Ask more of your vendors and stay open to opportunities to continue to capture data that you're producing. So we talked about ambient data that is automatically produced, but nurses produce additional data. For example, when I was in my dissertation, they did huddles every four hours and I would've loved to had the information that came out of the huddles, but that data went on a whiteboard and then was erased. And so it's no longer there. It's not available. So when you're creating data, keep an eye out for that. Also, it would be like if you're interested in helping advance AI in the field of nursing, we really need more shift level outcomes that are nurse oriented. Hospitals ask nurses maybe once a quarter, once a year about their working conditions, but if you're trying to do hour and shift level trying to find out patterns that matter, we really need to say, well, how was the shift?

Dana:

And I used unplanned overtime as a proxy indicator of strain. There's some limitations with that. We could use safety incidences, but there's also limitations with that because they're not always reported. And so even just having nurses do a quick one to 10 on their way as they sign out, would an advancement forward and helping us have more time granular outcomes for nursing to help advance AI in the field. And then when it comes to asking more about from vendors, it's like each system is so kind of like a silo and it's difficult to integrate. The data systems we're created for one purpose in mind, but now we need them to think about nurses and we need to think about the fact that nurses span many patients. And so we need kind of a nurse orientation. We need timeframes or data that we can group it by a shift or something like that in a much easier way than having to go back and reconstruct this out of the data. So those are a couple ways that listeners could participate in the innovation and themselves and generate pressure and interest in these areas.

Dani:

Partnerships are critical and industry partnerships with health systems and clinicians I think is the way of the future. And being able to innovate together with these asks and ways that we can utilize technology, gather new data, and then drive the outcomes that we need to see for change in our clinicians and patient care. Thank you, Dana, for your time and your expertise in this space. I really appreciate it and can't wait to hear more about the work that you're doing over the next couple of years.

Dana:

Thank you, Dani. I appreciate you and the tremendous work that you're doing as well. So the feeling is mutual.

Description

Dr. Dani speaks with Dr. Dana Womack, healthcare informatics consultant. Together, they discuss her pivotal career moments before diving into the power of harnessing ambient data, or ‘digital dust’, and turning it into clinical practice insights. They also cover the concepts of resilience engineering and adaptive capacity in nursing.

Transcript

Dani:

Welcome back to the Handoff from Trusted Health. I'm Dr. Dani Bowie, Chief Nursing Officer of Trusted Health. This season we're covering all things healthcare innovation and the importance of fearless leadership for change. Today I have the pleasure of speaking with Dr. Dana Womack, healthcare informatics consultant. We discuss her pivotal career moments before diving into the power of harnessing ambient data or digital dust and turning it into clinical practice insights. We also cover the concepts of resilience, engineering, and adaptive capacity in nursing. Here's my conversation with Dr. Dana Womack. Welcome to the Handoff. I am excited to be speaking with Dr. Dana Womack today.

Dani:

Dana, welcome to the Handoff.

Dana:

Hi Dr. Bowie. It's great to be here with you.

Dani:

Well, let's get started as I've been doing for this season, really centering us back to the patient. Dana, I would love it if you could share a story that's had an impact on you in regards to the patient experience.

Dana:

I'll go back to when I first started nursing, and as you're a excited new nurse, you want to make sure that your patient experiences are really smooth. And I remember a particular Saturday morning when this really didn't happen. I had a patient who I realized was being transferred to a long-term care facility when I saw her on a stretcher in the hallway with an ambulance crew next to her. And this was a surprise to me. And so I was like, oh my gosh, I wonder if her tuberculosis screening is done. I know that's a prereq for them accepting her. So I went to her chart and sure enough, it hadn't been done, so I had to go tell the guy, sorry, we need to put her back in bed. And then I had to fire off a message to pharmacy, can you send me a tuberculosis skin test right away?

Dana:

I had to ask radiology to work her in and then have the crew come pick her up again a couple hours later. And it was uncomfortable for her to do the extra transfers. It was a lot of chaos in the day, which was completely unnecessary, had the system fully recognized and we had done the prep prior to the weekend. Even if somebody had tipped me off and report she was leaving, it might have helped. But I went to my manager then I said, Hey, I said, I'm seeing these things, and it was her response that really has stuck with me. She said, well, Dana, what we need is more conscientious nurses like you. And I was young conflict avoidant. I knew she meant it as a compliment, so I just kind of smiled and left. But I was very unsettled. I was unsettled about my future as a nurse because I realized that what she was saying is our primary strategy is to count on you, the nurses, to overcome broken systems and to create positive patient experiences despite that, and that has stayed with me. I rally against that notion to today, and it really has continued to be a motivator throughout my career.

Dani:

There is a lot of that in healthcare that has implications on the patient, implications on the caregivers and the whole care team. I mean that's no easy feat though. It seems like a small thing of like, oh, you need to get this TB test done. There was multiple players that needed to make that happen, and the coordination, the effort that took was paramount. And I can see how that has shaped some of your career and who you are. And that's what I want to talk about next is you as a clinician and a leader. So can you share with some of our listeners today about any pivotal career moments or highlights that have been impactful and explaining where you are today as well with how you're practicing?

Dana:

So I spent two years on that first med-surg unit in a community hospital. And after that time I decided that I really wanted to understand if the frustrations that I was experiencing as a nurse were unique to my small town facility or if this was more widespread. And so I happened to meet up with a fellow nursing classmate and we were like, Hey, let's go be travelers. So we did for a couple of years. I would say that my experience as a traveling nurse, it was tough, but it was really, really informative and educational because I realized that these situations are pretty universal and that all healthcare facilities kind of struggle with them. And somewhere along the way, I'm not exactly sure where I came across the field of nursing informatics, which back in the early nineties was fairly new. And I'm like, wow, if we could apply better systems to this, that's potential for change.

Dana:

And so I'm like, I've got to switch the focus of my career. So I took a traveling nurse assignment into Salt Lake City and then switched to grad school and completed a master's in nursing informatics there intending to stay in healthcare, but I was recruited to industry afterwards. So I went to Intel Corporation and worked there for a couple of years and along the way, actually I'd taken elective courses in my master's program from the M B A program and in new product development, and I discovered my love of new product concepting and that was kind of what really took me into industry. After working at Intel, I deployed VRA probably in a dozen hospitals. That was impactful because I saw how immediate technological change can impact the work experience of nurses. If it's implemented well, it can be for the better. If not, it can be for the worse.

Dana:

And so it really emphasized the importance of good system configuration and training in deploying systems. Later I had the opportunity to support the office of the national coordinator for health IT during the of electronic health records as a Deloitte consultant during that time, participating in national initiatives was really exciting, but it was a little less hands-on than I preferred. And so when Intel recruited me back, I jumped at the chance and came back to the Pacific Northwest and then later earned a PhD in informatics at O H S U where I focused my dissertation on harnessing ambient workplace data to sense periods of overload at the bedside. Again, kind of coming back full circle to where I had started. So I currently split my time between applied research in the hospital and community settings and then innovation focused independent work in the industry.

Dani:

I love your background and as you're describing more about it, I think it's so critical that the profession of nursing is going into some of the paths that you have paved. There's a lot of controversy around the travel industry, but I've often said that travel nurses bring such a wealth of knowledge because of that exposure to many health systems and they see how one place has a best practice in a certain instance and another place may benefit from that. And so I've always encouraged managers, if you have a traveler, you actually have someone who has this exposure to the health system in a very macro way that can offer some insights. So I love that augmented with all of your experience at Intel, the academia as well. It's just amazing. We're going to talk a little bit more about this ambient data and digital dust and your doctoral work. And so let's shift gears and can you help explain to our listeners more about the use of ambient data, digital dust, what that is and how health systems can use that to improve the experience of the frontline clinicians and the experience of the patient as well?

Dana:

So when we're nurses at the bedside, we touch so many electronic systems and each one of these gives off constant streams of what you can think of as digital dust. And examples would be communication devices, nurse call medication dispensing, electronic doors that you swipe in and out of the electronic health record and many others. These data serve an immediate purpose like dispensing a medication, but then the systems are always logging these events to a behind the scenes backend log system that captures data about this event. So the beauty of

Dana:

The ambient data or this digital dust in the workplace is that nobody has to do anything extra to capture it. It's already produced, they're automatically generated, but currently we do very little to capture and turn these data into information and then knowledge to improve nurses', work experiences and healthcare operations. What can we do with this dust? Dust? We can harness multiple operational data streams to begin to tell the story of a shift, like the context of who is there, which nurses were assigned to which patients, the activity that is taking place. A lot of people would assume that hospitals would be able to produce data about this. Really a lot of our systems are very patient and encounter oriented rather than nurse workplace and work shift oriented. In order to convert these building up from the digital desk, from data to information, you have to derive summary features from the data streams.

Dana:

For example, you can calculate the count or the average duration of a nurse call and then how evenly or unevenly these nurse calls were distributed across the patients. That's just a couple of examples. With the single data stream altogether, you can calculate hundreds of summary metrics from naturally occurring data streams and then you can utilize these summary activity metrics as input into machine learning algorithms or other AI algorithms to begin to identify patterns that are predictive of an outcome of interest. You mentioned my dissertation work. In that exploratory work, I demonstrated that it's feasible to predict unplanned overtime as a proxy indicator of strain with moderate accuracy four to six hours before the end of a 12 hour shift. And what's nice about that is it would afford you time to intervene and level load nurses work before shift end. So the potentials there. Our environment is giving off these digital dust and digital whispers about our working conditions, but we need continued r and d to convert these into insight so that we can intervene before negative nurse and patient consequences occur.

Dani:

What I find interesting about this concept is often you, at least in the profession I've had people say it's so hard to say what nurses do to describe it, what is a nurse doing? And rightfully so because what you just described is like they're touching all these systems, but what you just presented here was an opportunity to create a really objective story around the work of a frontline clinician and what they're doing with how they are using all these different systems to deliver care. And there's so much opportunity there, and even as you mentioned, being able to moderately predict what's going to be indicators of overtime, I think you said was it four to six hours into the shift or was it four to six hours with the shift ending that you could see before the

Dana:

End? Yeah,

Dani:

Before the end and then as you said, readjusting some of the workload or interventions to offset that, and that is just not how we operate and it is usually a reactive space and something that we look at the lagging indicators of over time and premium, things like that. This is something that I think I'm really glad that you're sharing with us today and our listeners because I think that there's such an important need around harnessing the digital dust to create the story even of the work that we're doing. And then as you mentioned, I think the ability to intervene proactively in situations to help the clinicians and the health systems in general. Which kind of brings us to our next topic, which is resiliency. You have some expertise and interest in resilience engineering, and I would greatly appreciate if you could share with our listeners how that differs from individual resilience and how we should be thinking about that in our health systems today.

Dana:

Yes, it's really interesting to me how the term resilience has almost become a negative term in the healthcare space. I think it's because it's almost synonymous with personal coping and psychological resilience and it evokes mental images of self-care, meditation, yoga, and all these things are super important and totally support them, but when you think of resilience engineering, it's looking at the whole system that people operate within and recognizing that that system needs to be resilient as well. I think the reason that clinicians recoil when they hear the word resilience is because hearkening back to that original story and what my nurse manager told me about just needing conscientious nurses, they're kind of hearing the message that they need to be personally resilient to overcome a broken system. So resilience engineering has its roots in safety science. The term was introduced in 2006 to represent a new way of thinking about safety in multiple industries including healthcare.

Dana:

There was growing recognition that rather than seeing the root cause of events as a single point of failure, that these events needed to be explained as combinations of conditions that prepared a path to an error or an event that was unwanted. Historically, safety has been focused on counting incidences and the number of things that go wrong, but resilience engineering defines safety as the ability to succeed under varying conditions. For anybody who wants to learn more, I'd recommend a book by Dr. Robert Weirs called Resilient Healthcare. And in that book, he defines organizational resilience as the ability of the healthcare system to adjust its functioning prior to during or following changes and disturbances so that it can sustain required performance under both expected and unexpected conditions. So if you're an organization that practices resilience, you're putting more effort towards understanding everyday clinical work and how it is that work goes right, and recognizing that this augments your awareness of how to create safe organizations in addition to the continued accident and regular investigations that you're doing.

Dana:

It takes a system-wise perspective. It sees safety, quality, productivity, satisfaction and more as all facets of the same reality. So there's a lot to say. I get super excited about this topic, but I'll highlight one more. Healthcare organizations are complex adaptive systems as opposed to factories. Things are done differently and they have to be, it's not like we can do the same process every time through studying nuclear power plants and many other mission critical systems. The field of resilience engineering has identified three common patterns of failure in complex adaptive systems and one of these, the pattern of decompensation offers a really good explanation of how it is that nurses buffer a system and keeping things going despite challenges. So in decomposition, as pressure enters the system, and in our case we're thinking of nursing work systems, the system sustained its performance by utilizing its adaptive capacity.

Dana:

So nurses are compensating and they start to do this by doing things that have lower consequence like moving their break or skipping their break or spending less time talking with the patient's family. But then if the pressure continues, their adaptive capacity erodes and then they have to start making much tougher trade-off decisions like needing to forego things like oral care or patient mobility or something that's much more impactful or even to the point of asking, getting an after the fact cosign for a medication. I think this is a really useful pattern for us to recognize and for even to have nursing leaders think about talking about, we don't really talk about what we do when there's more work to be done than time available. There's research done on missed nursing care, but there's not a lot of guidance to nurses or it's nurses are hesitant to even talk about the fact that they're missing nursing care. And I feel like there's a lot of opportunity for healthcare to be a lot more open about this and to even collaborate and figure out, well, what are we going to do when we've exhausted our adaptive capacity and recognize that it's not something to hide, but to highlight and to address

Dani:

What you're mentioning here too is resiliency is often the individual perspective that we take as leaders in health systems, but in fact it's something much broader and the decompensation that happens as you're dealing with a broken system and then working to prioritize, do these systems, as you described, studying this from other mission critical areas and your knowledge in this space, did they provide some, here's how to prioritize, do they provide, here's how to solve the system or in a system that you're continuing to solve, you may still deal with some of this compression issue that then causes you to prioritize and here's some ways to think about that because we don't in nursing say, here's how to really prioritize the things that you're dealing with, your breaks, the oral care, whatever it is, it's like it is kind of intuitive and you're just managing it autonomously and then we don't talk about it. Do you have any thoughts about that?

Dana:

It's going to be different for every industry and potentially different settings, but a core central theme is recognizing your current point of operations. Are you in a point where you have adaptive capacity? Are you at a point and it's like it's time, workload, financial constraints, all of those are putting pressure in different ways. Are you drifting towards a safety boundary towards a workload boundary towards what is your dynamic point of operations and realizing that that's moving around right now we don't have a lot of ways to understand when our adaptive capacity is eroding and in this way, I think just maintaining adaptive capacity and having some reserve is a compliment to things like lean, where you're working out all extra waste in the system, but you still have to maintain your ability to adapt when people start to adapt. We theoretically could see some of these adaptive behaviors in our data delaying certain things.

Dana:

They're going to delay a daily thyroid before they'll delay an IV vancomycin, they'll delay oral care before they delay somebody's cardiac arrhythmia. It's like there's kind of like these weak signals that together can start to paint a picture of where the pressure is in the unit. I'd just like to highlight this because I've even encountered it where it's like when you start looking at things like delays and missing this and whatever, nurses get really nervous that you're looking for the bad apple, but we have to create an environment and a culture where we're all in this together and we recognize that you can't do impossible things and we recognize that and when impossible requirements start to pop up that there's the psychological safety of being able to talk about it and bring it forward. And I think that some combination of sensing in the environments combined with humans and they're able to fill in and kind of sense make on top of all of these digital signals that we can bring forward that that could be a potential solution for addressing these situations in a different way in the future.

Dani:

Culture of inquiry, I've been speaking with leaders about that and there's components of culture of inquiry, which is psychological safety and being able to be curious and be safe in that space, these digital ambient data that could be so powerful and harnessing the information and in reframing it that we're not looking for the low performers or the bad apples, but recognizing to air as human and what systems are in place that can stop that or what do we need to do with the systems to make them better. I do think that that's an important component is that psychological safety to be able to address this, which can be tough knowing. I mean, medication error can be deadly and there's so much implication around this, and we were seeing with the Vanderbilt case where the medication error and how scary was as a nurse to watch that with a fellow nurse who made an error, but why was that error made and how can we do better next time? We're talking a lot about the ambient data. Let's talk a little bit more about another concept called digital whispers, and maybe that's the same thing, Dana, I'm new to this context, so least explain to our listeners also about the concept of digital whispers and how digital whispers can be applied to nursing.

Dana:

It relates to the ambient data in healthcare. We really, really don't like uncertainty and ambiguity if you don't like risk. And so we like clear decision support like patient is allergic to penicillin, don't order this, or we consider a new treatment to have low evidence until a randomized clinical trial has been conducted and all of that. But when it comes to operations, we can't always avoid the ambiguity. We work in these messy, complex adaptive systems and we need to pay attention to signals that have perhaps lower signal to noise ratio than we'd like. Sometimes the same actions that could help a situation go are the same actions that could make something go wrong in another situation. And so I think that just advancing the idea of digital whispers, we're not seeing that. We're not even claiming that we can use data to say, you should do this or that in order to get your shift on track because I don't think that that's potentially even the right way to approach it at this point in time.

Dana:

We're seeing we can help harness digital whispers about your environment and then we need intelligent humans charge nurses, frontline nurses, managers to be aware of these as they're coming in real time and use them first to create the observability and then to sense the status of the working conditions and then go ahead and be proactive about it. I think that this idea of human and machine monitoring of workplaces will be a much larger focus in nursing in the future. I have an industrial engineering colleague, Dr. Lindsay Stickie, whose work I admire very much, and she also shares this vision. And together we've identified some digital whispers or weak signals that support insight. For example, the sum of communication minutes and the percent of medications delivered via syringe and the skewness of meds dispensing across nursing was predictive of unplanned over time, and so it was machine learning that helped us identify these patterns.

Dana:

But there's also pragmatic explanation and we've conducted qualitative studies with nurses to confirm this. So in this example, when you need to sort something out or there's a challenge or a logistical problem, you need to start communicating, and so that's going to increase your communication minutes across the shift. The p r n example is like those are often used to manage symptoms, and so if you have a lot of patients with a lot of symptoms that you're managing, your dispensing is going to go up in that category. To the point about skewness in medications across nurses, if you've got a patient that has a really heavy medication load or even if you have differences in assigned patients, those are going to kind of show up in your medication dispensing data. So those are just a few examples of this kind of informal term of digital whispers that I think would be really helpful to incorporate into our everyday monitoring and management in nursing.

Dani:

As you described these examples, I'm like, oh, yeah, yeah, that makes complete sense, right? Increased communication, which is potentially could be a patient situation or just something happening that you have to respond to. We haven't really been trained to think that way and to think about it in terms of these different workload components and the increase or the decrease related to those. Then leading to what's happening in the environment, there's this digital dust everywhere and it requires analysis, curiosity, and understanding compared with the human intervention. So machine and a human coming together for the sake of improvement and tying to the resiliency of an organization to make it that much better. Now, Dana, what are your thoughts on the future of nursing? Is there anything that excites you or worries you as we were talking about present and hopefully then what do you think about the future and the lens that you have, which is so unique in perspective?

Dana:

I think my primary concern is that nursing is not a long-term sustainable career for people, especially at the hospital bedside. We're losing nurses that recognize muddling through unworkable conditions is draining their wellbeing and putting their licenses at risk and creating unsatisfactory patient experiences. This might be a little more controversial, but I'm also worried that the current focus on mandatory ratios might hamper innovation and care model redesign. I totally, totally support the thought and the intention behind the ratios. I just want to see us to be able to continue to innovate in our care models and try different things. I'll offer an example from a hospital that I work with and during Covid there was a lot of quick juggling of roles and responsibilities and they recognized that people who place patients for surgery have particular skill in that area, and they created these proning teams that went around and went through units and helped prone patients, and it was just this incredible relief to the nurses to have this team come in and do that.

Dana:

There's a lot of debate about team nursing versus a hundred percent RN staffing and what should be done if you have a mandatory ratio and then as a result of increasing costs, they take away some of the ancillary services that doesn't really help nurses either. And so I really am concerned about that. We can continue to innovate in this space, but there are some things to be excited about. I am excited about the application of data science to increase the observability in the world. There's dynamic forecasting down to the hour and really small geographies and these technologies are out there and creating big changes in other industries, and I think that we can harness them and use them nursing as well. And I think our notable opportunity right now is just really to expand our thinking beyond staffing is the problem to, yes, staffing is a problem, good staffing is necessary, but on top of that, we can take steps to recognize and respond to strain whenever it rises. Realizing that burnout can happen in punctuated moments across many shifts as well as this kind of chronic stress that people experience in healthcare.

Dani:

Often I think bad staffing outcomes is just a symptom of something that's going down, which you kind of allude to. I would be in the same boat with you around the ratios just because I understand the place and the intent, and we all want great staffing, but it is a bit more complex than a number in the sense that even what you just described, all these different components that lead to a patient experience and the nurse or the clinician being able to deliver care and the systems in which they're using to do that. When you think about innovation and change, I kind of cringe though. I believe in it and I want to be a part of it, but oh, the work that goes into innovation and change, it's so much. So I'm really excited about what you're sharing and hopeful that this continues to evolve over the next 10 to 20 years and excited to see the work that you do as well leading in this space. Dana, where can our listeners find you?

Dana:

My email address is easy to remember, dana@danawomack.com.

Dani:

Perfect. Perfect. And you provided some really great literature for us to review authors and books and things like that, but listeners, if you're hearing it, feel free to send her an email. Dana is amazing and has so much knowledge in this space. Finally, Dana, what would you like to hand off to our listeners today?

Dana:

Ask more of your vendors and stay open to opportunities to continue to capture data that you're producing. So we talked about ambient data that is automatically produced, but nurses produce additional data. For example, when I was in my dissertation, they did huddles every four hours and I would've loved to had the information that came out of the huddles, but that data went on a whiteboard and then was erased. And so it's no longer there. It's not available. So when you're creating data, keep an eye out for that. Also, it would be like if you're interested in helping advance AI in the field of nursing, we really need more shift level outcomes that are nurse oriented. Hospitals ask nurses maybe once a quarter, once a year about their working conditions, but if you're trying to do hour and shift level trying to find out patterns that matter, we really need to say, well, how was the shift?

Dana:

And I used unplanned overtime as a proxy indicator of strain. There's some limitations with that. We could use safety incidences, but there's also limitations with that because they're not always reported. And so even just having nurses do a quick one to 10 on their way as they sign out, would an advancement forward and helping us have more time granular outcomes for nursing to help advance AI in the field. And then when it comes to asking more about from vendors, it's like each system is so kind of like a silo and it's difficult to integrate. The data systems we're created for one purpose in mind, but now we need them to think about nurses and we need to think about the fact that nurses span many patients. And so we need kind of a nurse orientation. We need timeframes or data that we can group it by a shift or something like that in a much easier way than having to go back and reconstruct this out of the data. So those are a couple ways that listeners could participate in the innovation and themselves and generate pressure and interest in these areas.

Dani:

Partnerships are critical and industry partnerships with health systems and clinicians I think is the way of the future. And being able to innovate together with these asks and ways that we can utilize technology, gather new data, and then drive the outcomes that we need to see for change in our clinicians and patient care. Thank you, Dana, for your time and your expertise in this space. I really appreciate it and can't wait to hear more about the work that you're doing over the next couple of years.

Dana:

Thank you, Dani. I appreciate you and the tremendous work that you're doing as well. So the feeling is mutual.

Back to THE HANDOFF