SCCM Pod-447 eSIMPLER: A Dynamic, Electronic Health Record–Integrated Checklist

The increasing use of electronic health records (EHRs) has inspired the need for a more dynamic checklist. Geva et al conducted a before-after quality improvement study by replacing a static checklist with an updated dynamic checklist. Host Margaret M. Parker, MD, MCCM is joined by Alon Geva, MD, MPH, to discuss how the implementation of eSIMPLER provided clinical decision support prompts with display of relevant data automatically pulled from the electronic health record and improved certain care processes (Geva A, et al. Pedtr Crit Care Med. 2021 October;22:898-905). Dr. Geva is an Associate in Critical Care Medicine, Department of Anesthesiology, Critical Care and Pain Medicine and an Assistant Professor of Anesthesia at Harvard Medical School in Boston, Massachusetts. This podcast is sponsored by Mölnlycke.

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

Category: PCCM Podcast

Transcript:

Babies and children aren’t immune to pressure injuries. Mölnlycke® Z-Flo Fluidized Positioners put your youngest patients in the best position. Z-Flo is designed for confirmational positioning across the continuum of care from developmental growth in NICU patients to pediatric pressure injury management. Z-Flo positioners conform and mold to every body, big or small. For more information visit molnlycke.us.

Dr. Parker: Hello and welcome to the Society of Critical Care Medicine’s iCritical Care podcast. I’m your host, Dr. Margaret Parker. Today, I will be speaking with Alon Geva, MD, MPH, about the article “eSIMPLER: A Dynamic, Electronic Health Record-Integrated Checklist for Clinical Decision Support During PICU Daily Rounds.” To access the full article, visit pccmjournal.org. Dr. Geva is an associate in critical care medicine in the Department of Anesthesiology, Critical Care and Pain Medicine at Boston Children’s Hospital and an assistant professor of anesthesia at Harvard Medical School in Boston, Massachusetts, USA, as well as affiliated faculty in the computational health informatics program at Boston Children’s Hospital. Welcome, Dr. Geva.

Dr. Geva: Thank you very much. Glad to be here.

Dr. Parker: Before we start, do you have any disclosures to report?

Dr. Geva: I do not.

Dr. Parker: Would you start by giving us some background to your study? What kinds of checklists have you been using in the past? What made you decide to work on a dynamic checklist?

Dr. Geva: First of all, Dr. Parker, thank you for inviting me to speak here today and thanks to SCCM for organizing this. I’m excited to share our work. As you and others probably know, there’s a lot of literature about checklists in general and, in particular, in the operating room, starting especially with Atul Gawande’s seminal work on improving patient safety. But the results of checklists in general have been more mixed. In particular, there are lots of challenges with their implementation, particularly for daily rounding workflow like we do in the ICU. Nonetheless, we had been using a more static form of the checklist going back almost a decade.

That checklist started off as a piece of paper. It was a very distinct, hot pink piece of paper that everyone at Boston Children’s recognized well, but that checklist really was a checklist. You literally could just check off: Did you talk about sedation? Did you talk about adjusting medications for renal dysfunction? But that checklist didn’t actually provide you any information. It was just a reminder. Over time, as electronic health records started to be used more and more, and we integrated them better, we digitized the checklist and we started to pull some information in from the electronic health record that would be relevant to the checklist questions.

So we were able to, for instance, display the list of medication orders that the patient had, so that when you got a question in the checklist about medications, you could look at the medications in front of you and perhaps use that to inform your answer to the checklist questions. But most of the information in that checklist was still static. It still asked the same questions for every patient regardless of their clinical context. And it didn’t really give you a lot of information to help you with decision-making the way we think about it. It asked you if you should adjust medications for renal dysfunction, but it didn’t give you any information about their renal function, for instance. So our goal here was to create a checklist that would be 1) easier to use and 2) provide you with more contextual information that you could use in using that checklist so that the answers you gave to the questions on the checklist could be better informed.

Dr. Parker: How did you go about developing this new checklist? That seems like it must’ve been quite a challenge.

Dr. Geva: It was a challenge. We have a wonderful IS department who I have to give a lot of credit to for the technical aspects of the checklist. The information is all there in the electronic health record, it’s just a matter of integrating that information in a way that’s logical for how clinicians think. So the principle that we set out with was to say that the checklist shouldn’t ask you questions about things that it should already know the answer to. For instance, because we document when a patient is receiving mechanical ventilation in the electronic health record, if it knows from the electronic health record that a patient is not mechanically ventilated, it shouldn’t ask you questions about sedation or keeping patients comfortable while orotracheally intubated. We use that kind of logic to clean up the checklist to make it more efficient, to try to reduce the number of clicks, to try to reduce the cognitive load that the checklist required of users.

The other thing we did was to try to integrate clinical decision support into that checklist. One of the questions that the original eSIMPLE always asked was, Does the patient require stress ulcer prophylaxis, GI prophylaxis? As many of the listeners know, there’s been a shift over my relatively short career from giving stress ulcer prophylaxis to every single patient basically in the ICU to recognizing some of the potential harms of stress ulcer prophylaxis and trying to target it a little bit better toward the patient in whom the benefits would outweigh the risks. I don’t want to get into the details of that because that’s a little outside the scope of this discussion. But one of the things we tried to do is not just ask our clinicians, Should this patient be on stressful ulcer prophylaxis: yes or no, but actually give them information to guide their answer to that decision, to say, Here are some indications for stress ulcer prophylaxis that experts agree on, and here’s which of these criteria our patient here, that we’re rounding on right now, actually meets. This patient is getting high-dose steroids. Therefore that is an indication for stress ulcer prophylaxis or, conversely, there is no indication that we recognize for stress ulcer prophylaxis. That doesn’t mean necessarily that the patient shouldn’t be on it. You can still order it, but at least we don’t see, from our electronic health record data, an obvious reason that the patient should be on stress ulcer prophylaxis. Maybe we were hoping, with this nonintrusive clinical decision support, we might gently guide clinicians to using less stress ulcer prophylaxis when it wasn’t indicated.

We tried to keep something called the “Five Rights” of clinical decision support in mind when we designed this clinical decision support. Basically the Five Rights suggest that clinical decision support is most effective when it’s delivered to the right person through the right intervention format and the right channel at the right time in their workflow. That’s a fancy way of saying you want to give information that someone knows what to do with and what he/she should do with that information when they get it. By integrating this clinical decision support directly into the rounding format, it allowed our fellows, who usually are the ones who we say run the simple checklist, who are going through this checklist while we discuss all the other patient care elements during rounds, they could integrate that information in the appropriate point during rounds and guide our residents or NPs or other frontline clinical staff in our multidisciplinary rounds with the nurses to have a discussion about that care element and make adjustments to orders, for instance, to the patient who should be on stress ulcer prophylaxis, to put in an order for ranitidine, if one did not exist and the patient should be on it, right at that point during rounds.

The other important thing we did in this context, I should mention here, is that we put in the relevant orders for that part of the checklist right in that section of the checklist. So rather than just having a medication list in the middle of the checklist as a separate section, like we used to have, what we did with this update was, when we asked, Should this patient be on stress ulcer prophylaxis, for instance, we actually put into the checklist, pulled from the EMR, Here is the stress ulcer prophylaxis that this patient is ordered. This patient is already ordered for ranitidine, or this patient is not ordered for stress ulcer prophylaxis, but they should be. So there’s your prompt to maybe go ahead and order it for them.

Dr. Parker: Are there differences in the issues that get addressed with your new eSIMPLER compared to your original simple checklist?

Dr. Geva: We didn’t change the content too much, which helped us to be able to compare between the original eSIMPLE and eSIMPLER, because we had our focus areas for the ICU, which were our focus areas both in the original era with eSIMPLE and our new era with eSIMPLER. So we focused on sedation, which was both maintaining sedation while a patient was intubated and weaning sedation after they were extubated. We focused on invasive lines and tubes; that’s the “I” in the acronym SIMPLER. We focused on medications and measurements, which included adjustment for renal dysfunction or hepatic dysfunction, if present, as well as making sure we were dosing medications for appropriate weights. We focused on prophylaxis, which included stress ulcer prophylaxis and VTE prophylaxis. We focused on lab plans and we focused on enteral nutrition, measurements of growth. Then we always had an “other” section that really was checklist-type items, like making sure we were doing pregnancy tests for older girls, for instance. So the things we focused on in the checklist were consistent between the two periods, and we weren’t really linking this specifically with explicit QI interventions, other than trying to improve the way we displayed information and the information that we provided to the rounding team.

Dr. Parker: How did you evaluate eSIMPLER?

Dr. Geva: We took two approaches to evaluating it. We wanted to see, first of all, how usable the checklist was going to be. We did this by following the rounding teams and doing real-time assessment by trained observers of how often the checklist led to productive discussions during rounds and to interruptions to rounds by the nurses who were correcting us, for instance, about inaccurate information. We also looked at how long the checklist added to rounds because other studies have found that the time to use a checklist was a significant predictor of its usability and how often it was used and how much it improved outcomes in those units. We didn’t have patient outcomes per se, because we don’t necessarily expect these kinds of quality improvement interventions to make obvious changes to a patient outcome, like length of mechanical ventilation, for instance, because that would be a very complex process. But we did look at how consistently our indications for stress ulcer prophylaxis matched our orders for stress ulcer prophylaxis, how often we prescribed VTE prophylaxis. We also looked at the number of line days for central venous catheters, how often we were measuring weights to keep track of patient weights on a regular basis and length and head circumference as well for appropriately aged patients. So we looked at those kinds of process measures as well.

Dr. Parker: How did you transition from the old static checklist to the dynamic one? That can’t have been too easy either.

Dr. Geva: That part I’m proud to say was not too difficult. I think one of the benefits of us designing something that was custom—this was a web-based application and it used core principles of usability and good user interface design—rather than being locked into the standard out-of-the-box EMR (we happen to use Cerner, but I think there are similar problems with Epic)—is that it allowed us to design around our users. I had a mentor years ago who said that, rather than us being tested on how to use an EMR or trained in how to use it, how well we can use it out of the box should be an evaluation of that EMR.

We tried to keep that in mind with eSIMPLE. The transition ended up being pretty easy because I’m proud to say that our team was able to create something that made sense to users. The information was just there in front of them where they expected it to be. All of the questions were clear, the buttons were where they expected them to be and did what they expected them to do. So very little actual training was required to get our fellows and attendings to use eSIMPLER. We did have a watching period when we did our evaluation just to make sure that if there were any hiccups in transitioning they didn’t affect our usability statistics or anything like that. But fellows, who are the primary users, picked it up very quickly. I think we’re generally pleased with it as well.

Dr. Parker: What did the clinicians think of the new checklist? You said they’re generally pleased. What kind of feedback did you get?

Dr. Geva: I will admit we had a relatively small sample size because we only had about two years of fellows who had used both the old and the new checklist. We only had about 15 fellows who even could potentially answer our survey, but overwhelmingly they said that the new version, eSIMPLER, was easier to use, that the interface was clear and understandable, and overall that it was an improvement from the prior eSIMPLE. So they were happy with it. I think part of that was because, when we looked at the amount of time they were taking to complete eSIMPLER—I have to acknowledge here that there’s a little bit of overlap because some of the questions can be answered as the resident or NP went through the other data on rounds, there was a little bit of simultaneous between the presentations and the filling out of eSIMPLER—the actual running through, confirming the answers and getting us all on the same page with the entire care team, the nurses, the respiratory therapists, etc., took about half as much time to complete eSIMPLER, only about 30 seconds as opposed to about a minute with the old system. So I think it helped move rounds along more efficiently as well.

Dr. Parker: What about the process measures? Did you find a change in the process measures you looked at?

Dr. Geva: We did. We found that most of them improved; they were small improvements as expected, given the complexity of critical care and all the factors that go into it. But we reduced the mismatch, for instance, between stress ulcer prophylaxis orders and between the indications for stress ulcer prophylaxis by about 5%, from 14% to 9%, we increased the number of patient days that VTE prophylaxis was prescribed by about 16%. And when patients had renal dysfunction, as quantified by their estimated GFR, we increased the number of the percent of days that clinicians recognize that medications should be adjusted for renal dysfunction as well. So we improved over a bunch of our process measures. There were a few that did not improve, we should acknowledge, like CV, the central venous catheter utilization did not actually change between the two epochs.

But I think some of that is not necessarily unexpected because all we did for central lines in eSIMPLER, as opposed to eSIMPLE, was improve the clarity of the display, we didn’t really have significant new clinical decision support that could help clinicians get central lines out any sooner than they otherwise might. And we were already, I think, pretty good about knowing which central lines were in. So not everything had to change. This isn’t a panacea. It was not a cure-all for all our process improvement measures but, across a wide array of areas that we did concentrate on with this checklist, we were able to show improvement in those process measures.

Dr. Parker: Is this a tool that could be applied at other institutions?

Dr. Geva: I certainly think that the concept could be applied in other institutions. It’s one of the things that I think is really helpful in designing tools like this—having them customized to the workflows of that particular institution. I suspect even different ICUs or other units like an oncology or BMT unit within an institution might have different needs. So I think it’s important to engage, not just the technical folks from the ISD side, but also the clinicians as well to design a tool that fits with your workflow and with what you want to get out of it, because it’s one of those things that’s going to be perfectly designed to achieve the ends for which it is designed. I think a lot of clinician engagement, iterative process, is really important in designing a tool that would be helpful to clinicians.

The other thing that I think is an important limitation of this work to acknowledge is that we did not build it using standard ontologies. We built it using our Cerner EMR and, more specifically, we even built it to our Cerner EMR, which is very far from the standard Cerner EMR build. So this is not something that could be plugged into another institution and start working out of the box; a lot of customization and a lot of what we call mapping of data elements would need to take place. Partly that has to do with the nonstandard nature of our EMR internally at Children’s Hospital, so we had to do it that way. But part of it, I think, is also a limitation of these standard ontologies. I work with Ken Mandl, who is one of the pioneers of SMART on FHIR applications. This idea of being able to have an app store for EMR applications that would be plug-and-play. FHIR is the acronym for a standard that is meant to be usable across EMRs as a way of exchanging health data and is becoming more and more accepted.

But the standard ontologies unfortunately do still have limitations in terms of the granularity of data or especially some of the things we want to look at in the ICU. And the seminal example of that for me is dosages for continuous infusions. The standard ontologies have concepts for medication administration certainly, but they’re not really designed around minute-by-minute or sometimes second-by-second titration of doses that are micrograms per kilogram per hour, let’s say, for something like a fentanyl infusion. So those kinds of data elements, which I think are important for tracking sedation, let’s say, if you want to know what the last sedation dose your patient was on, are things that are still not well integrated into the standard ontologies. So I think custom mapping from local EMR data models to the standard ontology is something that the field in general still needs to develop a lot more extensively before these applications can be easily transported across settings.

Dr. Parker: Where would you like to go next in your search to improve the use of checklists or improve the flow of rounds or the care of the patient in general? What would you think is your next step?

Dr. Geva: This is not my next step, because this is a dream state further in the future, I think unfortunately, but where I would love for this to go—the next step to truly make this follow the Five Rights of clinical decision support that we talked about and to help clinicians integrate it better into the workflow—is to integrate these kinds of checklists into the EMR. Right now, we have a custom web application that pulls information out of the EMR and we track the data and we track the results, so we record it into a database, but we’re not able, in an integrated fashion, to have the flexibility of our custom design to do that web-based application development and also have the benefits of writing directly into the EMR, not just as a static document, but as a dynamic process that allows the clinicians as they answer checklist questions, for instance, to put in orders, to use those data elements, to help automate or facilitate the attendings’ documentation that we have to do for billing and compliance purposes.

So, better, true integration, deep integration with the EMR is one of my longer-term hopes and wishes for this kind of thing. A more feasible next step, a shorter-term next step, I think, for this is to integrate it into the EMR as a more static document to be able to keep track of this information as a way of informing shift-to-shift handoffs. I think there’s an opportunity here because I suspect our ICU functions similarly to a lot of other ICUs, where a lot of plans are made longitudinally for patients by the daytime team, that is, the quote-unquote primary team for a patient, the longitudinal team for a patient, and a lot of the elements of those plans have to be carried out by overnight teams that are caring for the patient only for one or several nights at a time.

So one of the things that we built into eSIMPLER that we are not using effectively right now, for instance, is something as simple as saying, here is the lab plan, here is a chest x-ray plan, here is a fluid balance goal for this patient. We’re making those plans as part of our morning rounds. And we hopefully record those in our notes, but we aren’t summarizing it in a way that a night team can get quick and easy access to and use it to inform their plans in a way that facilitates night rounding and overnight patient care. So that’s one of the quality improvement things that we’re working on with eSIMPLER.

Dr. Parker: You’ve got some great ideas and you’ve done some great work here. Do you have any additional comments you’d like to make?

Dr. Geva: I said this at the beginning and I really want to say it again—we are very fortunate in my department to have wonderful support from a very wonderful group of programmers who come from various backgrounds. In terms of the future of pediatric critical care, I hope we are able to further develop these kinds of collaborations between clinicians, informaticists, programmers, and data analysts and create clever and original applications that can facilitate and improve our care in ways that traditional EMRs, which in many ways are just digital versions of our paper records, aren’t able to do.

Dr. Parker: Thank you very much for talking with us today, Dr. Geva. This has been very interesting and informative.

Dr. Geva: Thank you very much again for having me, Dr. Parker. I really enjoyed our conversation.

Dr. Parker: We have been talking with Dr. Alon Geva from Boston Children’s Hospital about the article “eSIMPLER: A Dynamic, Electronic Health Record-Integrated Checklist for Clinical Decision Support During PICU Daily Rounds,” published in the October 2021 issue of Pediatric Critical Care Medicine. This concludes another edition of the iCritical Care podcast. For the iCritical Care podcast, I’m Dr. Margaret Parker.

Babies and children aren’t immune to pressure injuries. Mölnlycke® Z-Flo Fluidized Positioners put your youngest patients in the best position. Z-Flo is designed for confirmational positioning across the continuum of care from developmental growth in NICU patients to pediatric pressure injury management. Z-Flo positioners conform and mold to every body, big or small. For more information visit molnlycke.us.

Margaret M. Parker, MD, MCCM, is professor emeritus of pediatrics at Stony Brook University in New York, and is the director of the pediatric intensive care unit (ICU) at Stony Brook University Medical Center. She is a former president of the Society of Critical Care Medicine and currently serves as associate editor of Critical Care Medicine and senior associate editor of Pediatric Critical Care Medicine (PCCM).

In her role as associate editor, Dr. Parker conducts interviews with authors of PCCM articles and other pediatric critical care experts. Dr. Parker received her bachelor of science and medical degrees from Brown University. She trained in internal medicine at Roger Williams General Hospital in Providence, Rhode Island, USA, and in critical care at the National Institutes of Health (NIH) in Bethesda, Maryland, USA. She spent 11 years in the Critical Care Medicine Department at the NIH where she was head of the Critical Care Section. In 1991, she accepted a position in the pediatric ICU at Stony Brook University and became the director of the unit, where she served for 27 years.

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 iCritical Care podcast is the copyrighted material of the Society of Critical Care Medicine.

All rights are reserved. Statements of fact and opinion expressed in this podcast are those of authors and participants and do not imply an opinion or endorsement on the part of the Society of Critical Care Medicine, its officers, volunteers, or members, or that of the podcast commercial supporter.

Some episodes of the iCritical Care 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.