Episode #329 Smarter Perioperative Triage

Opening Clip: [Ken Cummings] “So I’m Ken Cummings. I’m at Cleveland Clinic. I’m the head of our uh Center for Perioperative Medicine. So, this is one of those things where it’s kind of win-win for everybody. You implement a risk prediction model to triage patients to a pre-operative pathway. Patients who are comfortable with telemedicine win because they get to have a convenient pre-op visit. It’s actually lower cost for us per encounter. So, enterprise-wide can save a significant amount of money each year. It’s easier to scale our staffing with telemedicine. You don’t have to add an exam room if you’re seeing a patient virtually. But most importantly to me is that it frees up in-person access for the patients that we want to see and are going to need to spend more time with preoperatively.”

This is the Anesthesia Patient Safety Podcast and I’m your host, Alli Bechtel. We are so excited to be bringing you our newest series covering the 2026 APSF Stoelting Conference. I was there in person and now we have a great series for you including lots of exclusive clips from the speakers. We also have research article and book recommendations, visual patient safety boards, and more. This year’s conference was “SOS!: Signals of Safety… Safely across the Patient’s Perioperative Journey.”

Before we dive further into the episode today, we’d like to recognize Blink Device Company, a major corporate supporter of APSF. Blink Device Company has generously provided unrestricted support to further our vision that “no one shall be harmed by anesthesia care”. Thank you, Blink Device Company, – we wouldn’t be able to do all that we do without you!”

We are jumping right back in to learn more about how risk must be matched with readiness. Last week, we heard about using AI to support preoperative patient evaluation and now we are going to turn our attention to the postoperative period.

The second speaker of this session is Ken Cunnings, a professor of anesthesiology and director for perioperative medicine at the Cleveland Clinic. His talk is “Matching Resources to Risk: A Mid-Journey Perspective.” Ken tells us a little bit about the Center for Perioperative Medicine which works hard to see the right patients, create periop pathways, and increase continuity of care across the surgical journey. If you have been tuning into this series on the Stoelting Conference, you will start to recognize some themes here related to resource utilization and risk stratification. We can’t see every patient, so we need to see the right patients. We need to determine the appropriate resources for different risk levels, and we want to ensure continuity of care to see patients safely through their perioperative journey.

Let’s look a little closer at the triage challenge so that we can see the right patients and don’t see the patients that we do not need to see. This gets at the question of risk. At the Cleveland Clinic, they decided to capture risk in 4 categories by readmission and mortality data. The highest risk patients are hospice patients that include the top 2.5% risk for 90-day hospice enrollment. High risk patients are the top 10% for 90-day mortality or readmission. Intermediate risk is 36-90^th percentile mortality or readmission, and low risk is the lowest 35% for mortality and readmission. Once we define risk, we can put patients into their risk groups and this can help to make sure that we see the highest risk patients. In practice, this looks like the lowest risk patients having a virtual visit or an existing H and P while the high-risk patients may see an anesthesiologist or hospitalist preop and patients in the hospice category need advanced care planning preop. Triage and risk stratification preop is convenient for low-risk patients, lower cost per encounter, easier scaling with telemedicine for the virtual visits, and there is increased access for high-risk patients. So, lot’s of wins all around.

Now, we are going to fast forward through the surgery and look at what happens after surgery. This is an area of opportunity to improve patient safety and bridge patients to outpatient care. Can we identify the patients who would benefit from more intensive medical management postoperatively? Once again at Cleveland clinic, they are seeing the same pattern that if high risk patients are seen within 14 days postop, their mortality and readmission rate decreases. These are the patients that we need to see. Lessons learned from Cleveland clinic include automated triage is feasible and post-discharge is the next step with earlier follow-up for the highest risk patients. Does this mean primary care partnerships? Or is there a role for expanding the preoperative anesthesia clinic into the postoperative space? Perhaps we will be seeing peri-anesthesia clinics in the future!

We are going to round out this session with Josh Biro a research scientist at the MedStar Health National Center for Human Factors in Healthcare and an Assistant professor in the department of family medicine at Georgetown University. His talk is “Navigating the AI Wave with a Safety Mindset.”

We’re starting by talking about how good AI is when it comes to medicine. Well, Open Evidence is getting pretty good. If we compare Open Evidence to General Purpose LLMs (large language models), they all performed very well. When compared to humans or with humans or alone, LLMs performed the best alone. Check out the Science article from 2026, “Performance of a large language model on the reasoning tasks of a physician.” The takeaway from the study is “In all experiments, the LLM outperformed physician baselines and displayed continued improvement from prior generations of AI clinical decision support.” So, this is going to be a good resource for us to use going forward.

But there are some caveats especially when it comes to patient safety. There is more to being a doctor than just text generation. AI tools can be thought of as teammates for doctors to use to provide better care. So, we need to think about the human-AI team going forward and this is an area for research as well.

Josh takes us through a study of primary care physicians using AI-drafted email responses and there were some benefits, but AI also made some mistakes that physicians failed to correct. So, why are mistakes being made when we team up with AI? There is a tradeoff between safety risk and efficiency burden. Here is Josh to explain it now:

[Josh Biro Clip] “And so, why did AI draft mistakes get missed? There are many reasons. I’m going to talk about some of the cognitive biases that could be at play here. Automation complacency, the fact that for the most part, the AI did a pretty good job means that it’s hard to be vigilant all the time. There’s also automation bias, the tendency to prefer automated suggestions over relying on one’s own judgment. When you’re a primary care physician or a clinician generally, and you have a ton of work and there’s an AI giving you an answer, man, is it hard to go, I’m going to put in the cognitive effort to check that. And so AI presents this fundamental trade-off in many circumstances between safety and efficiency. Where if review becomes too burdensome, AI loses the value that it’s supposed to provide. But if review becomes superficial and not done, that all the risks and failures of AI can become a reality.”

[Bechtel] How do we calibrate trust in AI? There is very little guidance right now but Josh provides some very practical advice. So, here we go. Perhaps we should think of it as a junior teammate with strong training and education, but no letters of recommendations, confident and cocky, no track record or real-world experiences, blind spots they may not recognize, and requires onboarding and supervision. You can interact with this new teammate by learning where they are strong and where they need oversight, let them take on work only after you know what they can handle, share lessons learned with colleagues, and remember that the responsibility still falls on you.

That was such a great session with lots of important considerations especially when it comes to risk stratification and using AI. I really like the idea of treating AI-tools like a junior team mate. I think that is something that we can all relate to and it is a practical way to use AI safely.

Now, its time to think about and talk about the patient experience. The next axiom is the Patient’s Experience as a Vital Sign. Vonda Bates is the moderator for this session. She is the CEO of 10^th dot as well as an alliance builder, public speaker, and coach who helps multidisciplinary teams communicate effectively, align around shared goals, and achieve extraordinary results. She’s going to introduce the session now.

[Vonda Bates Clip] “The goal of this session really is to understand the both the opportunities and the barriers to treating patient experiences as vital signs. I come to this field. And I went into the hospital as well as he did, not knowing anything about, not only did we not know about hospital-acquired VTE or VTE in general, we didn’t know about hacks at all, hospital-acquired conditions. Yogiraj was a teacher. He was an organizational development consultant like I am. And I know with certainty that he would have taken the conundrum that faces healthcare today and offered a contribution that would have been meaningful, especially around relationship. And as we begin to introduce ourselves to the advancements with AI in particular and all of the other technologies that are before us, one of the things that I think we have to keep in mind is that relationship that really was quite the origin story, if you will, between clinicians and patients.”

[Bechtel] What do we mean by treating the patient’s experiences as a Vital Sign. There is a presumption of goodwill that clinicians are interested, paying attention, and care. Patients express their experiences very differently, so we can’t always point to just one thing, like pain, if we are trying to assess the patient experience. It can be challenging to interpret and measure the patient’s experiences against a baseline. Plus, we need to think about the time frame…when do we need to assess the patient’s experience? Perhaps, during all phases of direct care. We need time points that are immediately actionable, but for now this remains a lagging vital sign.

Here’s a quick example in the postoperative period: When patient’s say, “I’m terrified. I have a sense of doom.” What may be heard by clinicians is, “Normal post-operative gloom.” But, this is a real risk for patients since a sense of impending doom can be a sign of a major issue. So, if we consider this, the patient’s experience, a vital sign, we can ask targeted questions and add to the differential diagnosis and maybe even escalate care.

And now it’s time for the first speaker for this session, Mike Grocott who is a Professor of Anesthesia and Critical Care Medicine at the University of Southampton. His talk is COMPAC: Core Outcomes in Anaesthetic and Perioperative Care.” COMPAC-StEP stands for Core Outcomes in Perioperative and Anaesthetic Care and Step is Standardized Endpoints in Perioperative Care. Phase 1 was a systematic review looking at the previous 15 years including big, randomized control trials. The principal finding was that the outcome data collection was not standardized. Some of the outcome domains identified included mortality and survival, morbidity, complications, adverse events, health resource use, and more. Phase 2 was stakeholder consultation, and this was successful in reaching a broad range of stakeholders. The outcomes that came out of the consultation phase included operation being successful, being comfortable after the operation, getting out of the hospital quickly, returning to normal activities, avoiding temporary complications, avoiding permanent complications, not dying during or soon after the operation. Phase 3 is the Delphi Process. There are 7 domains and a number of different outcomes. You can check out the article in the British Journal of Anesthesia from 2022. Key points from the study were that selecting outcomes for clinical trials is necessary to inform trial design, study interpretation and impact, reproducibility and data synthesis. These core outcomes determined by this project included clinical outcome measures of postoperative mortality and morbidity, length of hospital stay and unplanned readmission, and patient-centered outcomes of short and long term patient recovery after surgery. Compac provides a broad stakeholder guide for selecting outcomes for future clinical trials. So, for any researchers out there we hope that this can help guidelines your next project.

Our next speaker is Leah Binder. She is the President & CEO of The Leapfrog Group, a national nonprofit representing employers and other purchasers of health care calling for improved safety and quality in hospitals. Her talk is “Patient Experience as a Vital Sign: A View From Purchasers and Consumers.”

[Leah Binder Clip] “Here she is now. So they founded Leapfrog and they had an approach that’s truly a business approach. It’s they said, let’s publicly report how hospitals are doing on safety, and let’s make that publicly available so that the public at large can compare among hospitals before they choose where to go. And our employees can as well, and we can, as purchasers, begin to actually reward excellence in safety when we see it. Otherwise, we have no idea where the best performance is or the worst, nothing. But we want an open market, basically. When they did this in the year 2000, they were considered extremely quixotic. I mean, this was, they were almost laughed at because it was, you can’t measure such things. It shouldn’t be made public. People will freak out, they won’t know what to do with that information. I mean, there’s a lot of reasons that this was considered just a crazy idea. But fast forward 26 years, and it’s it actually has been working. And because of not just Leapfrog, but also the efforts of many, like many of you in this room probably, the efforts to develop science of how you measure performance, how you measure safety, how you publicly report it, advocacy to get it publicly reported. That’s something Leapfrog’s been instrumental in. Because of all of that, over the period of 26 years, we’ve been able to do this in a pretty sophisticated way, and I think get at some very important information and drive change.”

[Bechtel] The Leapfrog mission is giant leaps for patient safety with a main strategy of healthcare transparency to drive an open market for excellence by publicly rating hospitals and ambulatory surgery centers.

[Leah Binder Clip] “Hospital CEOs need to see, and healthcare in general needs to recognize if a patient says it’s a priority to them, that should be your priority, that should drive your change. And that’s an economic principle, but it’s also, I think, a principle for good health care. And it doesn’t mean that patients are always right. It just means that they often have an insight that you cannot get anywhere else. And sometimes that insight is clinically important.”

[Bechtel] At Leapfrog, they have a hospital safety grade using 32 measures of safety with outcomes measures that come from CMS including never events, falls, injuries, infections, etc. There is an expert panel that helps to frame this. There are also 5 domains in the patient experience survey which are related to patient safety outcomes and largely have to do with communication. Here’s some good news. They are seeing improvements in patient experience over time. We are moving the needle and the patient experience part of this change. Patient experience is the heart of healthcare.

We will be back next week to continue our coverage of the Stoelting Conference. And spoiler alert we’ll be hearing from Thomas Lee, the Chief Medical Officer for Press Ganey Associates and we hope that you’ll tune in.

Before we wrap up for today, we have a few more APSF Book Club recommendations for you. These are books that are written by Stoelting Conference speakers or highlighted during the conference. We hope that you will check out Danielle Ofri’s book, “What’ Patients Say and What Doctors Hear.” Vonda mentioned this book while opening up her session. We also have 2 more book recommendations based on meeting the authors at the conference.

1. Perioperative Quality Improvement published in 2022 with editors, Carol J. Peden, Lee A. Fleisher, and Michael Englesb.
2. Social Capital in Healthcare: How Trust and Teamwork Drive Organizational Excellence by Thomas Lee.

Are there any other books that we should add to our book club list?

If you have any questions or comments from today’s show, please email us at podcast@apsf.org. Please keep in mind that the information in this show is provided for informational purposes only and does not constitute medical or legal advice. We hope that you will visit APSF.org for detailed information and check out the show notes for links to all the topics we discussed today.

If you can, we would love for you to subscribe to the podcast and maybe share it with a friend or a colleague or anyone you know who is interested in anesthesia patient safety. This way you can check out the entire Stoelting Conference Series and over 300 episodes in our archives. Thanks so much for listening.

Until next time, stay vigilant and stay informed so that no one shall be harmed by anesthesia care.

© 2026, The Anesthesia Patient Safety Foundation