SCCM Pod-475: Data Science and the Future of Critical Care Research

---

Discovery, the Critical Care Research Network, is an SCCM program that aims to expand research and improve outcomes. Discovery has launched the Data Science Campaign to leverage large-scale (big) data for research, seeking to apply these data in a clinical environment through standardized models and shared resources. Kyle B. Enfield, MD, FSHEA, FCCM, was joined by J. Perren Cobb, MD, FACS, FCCM, and Karin Reuter-Rice, PhD, NP, FAAN, FCCM, at SCCM's 2023 Critical Care Congress to discuss the future of data science and critical care research.

*If you are unable to play the podcast please click here to download the file.

Transcript:

Dr. Enfield: Hello and welcome to the 2023 Critical Care Congress edition of the Society of Critical Care Medicine Podcast. I’m your host, Kyle Enfield. Today, I’m joined by J. Perren Cobb, MD, FACS, FCCM, and Karin Reuter-Rice, PhD, NP, FAAN, FCCM, to discuss the Discovery Data Science Campaign and the importance of data science for the future of critical care research. Dr. Cobb is one of the preeminent experts in the field of critical care and is a professor and clinical scholar of surgery and anesthesia at Keck Medical Center of USC in Los Angeles, California. Dr. Reuter-Rice is an acute care pediatric nurse practitioner in the Division of Pediatric Critical Care and an associate professor at the Duke University School of Nursing, School of Medicine, and Duke Institute for Brain Science at Duke University in Durham, North Carolina. Welcome, Dr. Cobb and Dr. Reuter-Rice. Before we start, do either of you have any disclosures you’d like to report?

Dr. Cobb: I’m Perren Cobb. I have three. I have equity interest in three companies, Akido, BauHealth, and GibLib.

Dr. Reuter-Rice: Hi, Kyle. Glad to be here. I have Elsevier book royalties as a book editor, and I also have intramural and extramural federal funding.

Dr. Enfield: Perren, Karin, thanks for joining us today. I’m glad you guys took some time out of Congress to sit down with me. I wanted to start out and ask you what you think the importance of data science is for critical care.

Dr. Cobb: Well, I think it really helps us address the question of how do we get better, whether you’re looking at it from a performance improvement standpoint or from a research standpoint as two sides of the data coin, if you will. Without that, you’re not going to be able to answer the question how we get better. The challenge that we have in critical care is we’ve always had a lot of data. Many people say that it’s the most data-rich environment in medicine. For a long time, we haven’t had the ability to be able to collect all those data and analyze them simultaneously. But now in the era of big data, we do.

Dr. Reuter-Rice: I also think that, with this data richness that we, we have a real variety of talent around how do we translate and how do we understand what the data are? I think we’re in a period of immense growth around data and how we translate that data but also, how do we get the clinician to use data in a way that’s meaningful for the patient, the family, and actually the service itself?

Dr. Enfield: So really, you’re taking that informatics approach of transforming data to wisdom. I wonder, in that thought process, could you give us some examples of where data science has really empowered critical care?

Dr. Cobb: Sure. One of the areas that we’re very interested in is improving compliance with best practice, what we know to be best practice, so that every patient every day gets exactly what they need when they need it. One of the bigger campaigns that the Society’s been promulgating over the last few years is the IC Liberation Campaign, with A to F. We know because we’ve measured it at a national level that compliance with best-practice protocols, if you look across academia in any specialty, is on the order of 25% to 50%. That’s true of the A-F bundle, and we’d like to get it closer to 100% obviously. A straightforward thing with data science is to be able to collect data in real time at the bedside, where providers are providing care and be able to remind that person that they may not have gotten a certain aspect of the bundle yet done, please do that, then at a higher level, be able to show that compliance with the bundle, for example, coupling a spontaneous awakening trial with a spontaneous breathing trial, helps improve outcomes and can measure that patients are getting extubated earlier.

Dr. Reuter-Rice: Yeah. And when I think of a population of children who often don’t have recognizable symptoms to many providers, particularly in areas of care where there aren’t as many pediatric patients who interface with the health system, for example. Thinking about our sepsis campaign and how we recognize children who are presenting with early signs of sepsis, you think about the data that you can actually place within your EHR systems that trigger your warnings, that you’re seeing specific characteristics or phenotypes that would highlight the potential for this child actually being in a septic state, and then allowing you to trigger treatment more immediately, and then also coordinate higher level care as needed.

Dr. Cobb: I think it’s important to make the point, Kyle, that this is important not only for big hospitals that may have a formal infrastructure and processes for data science, but also smaller hospitals. That’s part of the epiphany that data science can provide that data from smaller hospitals or medium-sized hospitals or larger hospitals, those distributed nature by which we provide care, that they can all play a part in contributing data to work toward making sure that we have complete datasets and that we can use those data to make the most informed decisions possible.

Dr. Enfield: You guys have made a pretty good case for the importance of data science, but what do you see right now are the gaps for data science and the impact for the critically ill or injured?

Dr. Reuter-Rice: I think, touching upon the fact of what is the familiarity with data science, when you’re a practicing clinician or provider at the bedside, how much time do you have to actually get that level of either familiarity or even fluency around data science? Are there components to AI or machine learning that might help you change your practice or allow you to start interpreting earlier warning signs to change your practice? I see that as a gap of this kind of fluency for providers to understand how do you best use data science and how do you advocate for the type of support you need within your structures of health? Then, also importantly, how do you present that to patients and families, that this collection of data can really have immense benefit for the individual themselves, but even for larger populations? But I think it’s a gap. We need to help people understand what does big data mean, or what does data science do for them?

Dr. Cobb: And it’s important, I think, to emphasize that it is team science and we have clinical expertise, but we also need the technical expertise. IT departments, information service departments are really challenged these days, right? With all the IT requirements that are going on at the hospital level and trying to find IT groups or individuals with data science expertise who have the bandwidth to be able to support new projects, for example, I think either at the local level or at the national level is going to be really challenging. I think that that’s one of the roles the Society and this new Data Science Campaign can play at a national level is informing people of the importance in helping to motivate funding agencies to help us make some real progress.

Dr. Reuter-Rice: Yeah. I think when you tag on the fact that the analyses are complex and that you really need to have a pretty significant infrastructure, even ways to test models and then replicate those testings to make sure that you’re actually visualizing the data in a way that’s meaningful and accurate, that’s still, I think, a struggle for many institutions, but more importantly, for larger campaigns to have this coordination of analyses as well. I think that there are a number of gaps. But I think the Society is working really strategically with the actual Data Science Campaign to start bringing in those experts and start fleshing in the kind of support and resources people need to become more fluent, to become more engaged and able to interact in ways to bring data forward or to participate in ways of data collection or data translation.

Dr. Cobb: I think that one of the silver linings of the COVID pandemic is that the average person has a much better understanding about why ICUs are important, that the ICU is a stressful place to work, and helping critical care practitioners to be able to work smarter instead of working harder, I think is helpful because, prior to COVID, I don’t think most folks had a good sense of what we do in intensive care units, but I think there’s a much better appreciation now as a result of that and hopefully we can leverage that better understanding and sympathy for critical care practitioners to help promote some of these ideas.

Dr. Reuter-Rice: Yeah. And playing off of that, I think it’s also awakened funders because this is expensive type of work. It’s starting to, I think, become more of a funding-rich environment to really think about why is it important for us to have this data and use it in a way that changes health, not just from a financial bottom line, but really from a quality-of-life perspective, from reducing morbidity and mortality, and how do we use this data-rich environment that Perren mentioned in a way that’s actually going to change the health of the public at large, not just within an individual ICU.

Dr. Enfield: You guys have covered a lot of ground there, and obviously the Society’s got a mission here. Maybe you could walk us through how the Society’s Discovery Research Network is addressing some of those issues.

Dr. Cobb: The Discovery Research Network has existed for several years, providing an umbrella to get investigators who are interested in doing investigations together to recognize how challenging it is to be able to collect data. You can have studies, for example, where you have investigators in different hospitals with different types of EHRs and it’s very difficult to be able to get those EHRs, if you will, to talk to each other. So having this foundation, if you will, of investigators understanding some of the data challenges of being able to do collaborative science, to be able to do team science, I think has helped motivate the Discovery Research Network for its Data Science Campaign.

Dr. Reuter-Rice: Yeah. I think, ultimately, the end point is clinician use and the change of care within the critically ill patient population, right? The impact part is, can we bring more people to the table, can we shore up efficient ways to start thinking about data? Using data in a way that is coopted in a group or across our nation so that we can better inform how we go forward in intervention of care, new druggable targets, really thinking along the lines of extending some of the opportunities we have in data-rich environments to those that don’t have those but would benefit by the same treatment simply because we’ve had this large data collection that informs us and is more generalizable. It’s not unique to just one site. I think that the Society is looking at making a larger impact across all of public health.

Dr. Cobb: To address that, the Society put together, as part of its strategic planning initiative, the Data Science Campaign of the Discovery Research Network, which has three domains. The first domain is establishing a baseline, if you will, guidelines around data sharing and harmonization. It includes a couple of objectives, including the guiding principles panel that’s been created of subject matter experts and, in parallel, an outcomes and definitions work group. The second big domain is a new datahub of SCCM to provide a place for investigators to be able to collect and store their data for ease of storage but also ease of analysis. And then finally, annual datathons that will allow investigators to get together with analytical experts, maybe they’re a part of the critical care enterprise, maybe they’re not, they may be from outside the enterprise, but be able to really bring as much information as possible, as much insight as possible from the data.

Dr. Enfield: You guys have both mentioned investigators, but really that pipeline is from our young investigators. Sometimes when we think about some of the areas, there’s a pathway that people can think of when they want to go in from bench science or clinical science. But where is the entry point for the young fellow or the resident or other person who wants to get involved in this kind of research?

Dr. Reuter-Rice: First I say, “Welcome. Dial us up. We’re happy to include you.” That’s the first thing I’ll say. There are plenty of seats at the table. I would also say that there are a lot of opportunities within the actual Discovery Research Network site that will allow young investigators to start looking at opportunities to participate. But importantly also, if they have novel ideas and they want to tap into databases that are currently existing, they can submit proposals as well. And on top of it, the Society has funding awards for projects that can also leverage data from these existing datasets. So I think, as a new investigator, the experience you have has to be simply desire and eagerness to learn and participate, and then there’s a place for you at the table.

Dr. Cobb: To complement that, the Society has a Research Section, and the mission of the section specifically is to focus on the professional development of new investigators, that they go from just starting off in science and maybe not having much experience to hopefully seasoned funded investigators. That Research Section partners with the Discovery Research Network, whose goal is also, to some degree, professional development of investigators but primarily is focused on being able to foster investigator-initiated hypothesis testing. Whether the projects that are submitted and ultimately accepted are quite small or quite big, even larger than SCCM partnering with other professional societies globally for something like the pandemic.

Dr. Reuter-Rice: On top of that, we also have Discovery Investigators Meetings. Anyone can join them when they’re announced. There is an annual one that we’ve actually just resumed again at the NIH that people are happy to join and come and listen to some of the current projects that are in process, but also get a flavor for what’s happening at the Discovery level. I think even the virtual meetings are really accessible. That’s another way to get your toe in the water and see if it’s something you’d like to participate in.

Dr. Enfield: I also have to acknowledge the fact that, even in this sort of 30,000-foot view look at the data science, we use terms like harmonization and data sharing that aren’t things that any of us went to medical school for. And Tim Buchman this morning at one of our sessions on artificial intelligence raised the question about data literacy. How do you think SCCM, as a Society, can really help promote data literacy within our peer group?

Dr. Cobb: The Society is really good at, as I mentioned, building teams, and the Society also has a number of educational sessions, for example. I could imagine a session, let’s say at next year’s meeting, instead of being dedicated to echocardiography or something, was dedicated to data literacy. Putting together a panel very similar to what Karin and I and others have done for the data science panel and the data science work group of the campaign, get investigators, data science experts, clinicians, performance improvement experts together to present a data literacy or cata science 101 session, for example. My guess is that those would be well attended. I was also at that meeting where Tim Buchman was making that point. I think we are, as I think he mentioned, going to need to get to the point as a Society, if not requiring, having at least the expectation of members that there’s a basic understanding of data literacy and data science. Because so much of the advances that we hope our patients and their families will enjoy in the future are going to be based upon applying it really well.

Dr. Enfield: As people are listening to this, I think one of the questions they’re going to ask themselves is, when can the results of all this work be put in their hands to where they can implement it? What would you say to all those clinicians who are out there wanting to know when the work is going to be completed and when can they use what you’re working on right now?

Dr. Reuter-Rice: I’m going to say we’ve been really fortunate, if you look at several of the projects that sit under Discovery, such as, SARI-PREP and VIRUS, that there are plenty of outcome papers already, so that data is coming out and we’re seeing practice translation happening. I think as they move forward, the recognition that we need a larger data campaign that will bring in all walks of ICU medicine, whether that’s a small rural hospital or a large urban hospital, that we really think about data that we collect in a way that is translatable to all of the population, not just unique to high tertiary care level centers. I think that that’s going to take a little bit of time, but I think what you’re seeing is a significant and very serious investment by SCCM that this be the path forward, that we’re really starting to acknowledge that there’s an importance to this. But in the same breath, a realization that it has to have what I would consider a significant impact if we’re going to do it. I think that’s why you’re seeing this data campaign having so many different spokes to it, because they realize it’s not just one singular domain that’s going to have impact. It’s literally going to take a village of critical care or professionals to help get us to where we need to go.

Dr. Cobb: And maybe I can make an analogy to the genomics revolution. I’m old enough to have been a part of that back in 2000. There was a lot of new lingo, a lot of new terminology, not a lot of familiarity about the omics revolution, what genomics meant to clinicians, and then we learned about it, and what we’ve learned is that it’s a lot harder than you might think. There’s a lot of hype initially, but it’s a lot harder than it might initially appear. I think patience and iteration are going to be a big part of it, and I hope our listeners will recognize that what we’ve tried to provide is a really good basis for that. The Society provides a lot of value added as far as helping to usher in this era of “big data” or informatics and would encourage everybody to be part of it.

Dr. Reuter-Rice: I think what Perren mentioned was there’s going to be an iteration. I can’t imagine data ever being static. I imagine that where we start today will look very different in 5, 10, 15 years from now, which is a good thing. I think that there isn’t really an end in sight for data science. I think it will continue to propagate across all fields of what I would consider health. I think in the same way, it will also make us reflect on what are we missing, what haven’t we even considered yet? But I think where the Society is starting as a really level playing field and a way for us to, what I would say, make it comfortable for everyone to participate if they’re interested in participating.

Dr. Enfield: Thank you guys both for coming down today and speaking with me. I know there are a lot of people out there who are not able to make Congress and hopefully this will inspire them to look at Congress online and watch those sessions on data science and reach out to you to get involved in the Data Science Campaign.

Dr. Enfield: Thank you very much, Dr. Cobb, Dr. Reuter-Rice. This concludes another edition of the Society Critical Care Medicine Podcast. For the Society of Critical Care Medicine Podcast, I’m Dr. Kyle Enfield. Thank you.

Kyle B. Enfield, MD, FSHEA, FCCM, is an associate professor of medicine in the Division of Pulmonary and Critical Care at the University of Virginia. He received his undergraduate degree and joint medical and master’s degree from the University of Oklahoma. He completed his residency and fellowship at the University of Virginia. In July 2013, Dr. Enfield was appointed medical director of the medical ICU at the University of Virginia. From 2009 through 2016, he was the assistant hospital epidemiologist there and remains the medical codirector of the special pathogens unit. Dr. Enfield’s clinical interests are in critical care medicine and transport of critically ill patients. His academic interests are epidemiology and prevention of healthcare-associated conditions, including multidrug-resistant organisms and healthcare-associated infections.

This podcast was recorded during the Society of Critical Care Medicine’s 2023 Critical Care Congress. Access essential education online through Congress Digital. More than 120 sessions are available on an easy-to-use platform. Continuing education credit is also available. Some SCCM members receive complimentary access to Congress Digital. To learn more, visit sccm.org/congressdigital.

Join or renew your membership with SCCM, the only multiprofessional society dedicated exclusively to the advancement of critical care. Contact a customer service representative at +1 847-827-6888 or visit sccm.org/membership for more information.

The SCCM podcast is the copyrighted material of the Society of Critical Care Medicine. All rights are reserved. Find more episodes at sccm.org/podcast.

This podcast is for educational purposes only. The material presented is intended to represent an approach, view, statement, or opinion of the presenter that may be helpful to others. The views and opinions expressed herein are those of the presenters and do not necessarily reflect the opinions or views of SCCM. SCCM does not recommend or endorse any specific test, physician, product, procedure, opinion or other information that may be mentioned.

Some episodes of the SCCM Podcast include a transcript of the episode’s audio. Although the transcription is largely accurate, in some cases it is incomplete or inaccurate due to inaudible passages or transcription errors and should not be treated as an authoritative record.

Disclaimer