AI in Action: Improving OR Utilization at Beverly Hills ASC

Ep. 127: Benita Tapia & Conor McGinn – AI in Action: Improving OR Utilization at Beverly Hills ASC

Here’s what to expect on this week’s episode. 🎙️

In this episode, Benita Tapia, Administrator & Director of Nursing at Beverly Hills ASC, joins Alex — along with Conor McGinn, CEO of Akara — to unpack how she’s using AI to improve everyday operations: tightening block utilization and surgeon accountability, speeding turnovers, answering scheduling questions in real time, and boosting registration accuracy to cut preventable denials. We also cover the guardrails — privacy, policy, and a human in the loop — that keep these tools safe and useful. 

In our data segment, we look at Otolaryngology benchmarks from nearly 112,000 cases across 202 centers. OR block utilization sits around 46%; net revenue per case is up 26% ($2,981 to $3,765); deposits slipped, and denials ticked up while days to bill improved. We highlight trends to watch, and simple steps leaders can take now to tighten collections, curb denials, and move closer to that 70% utilization goal. 

Watch the episode on YouTube or listen to the full discussion on your favorite podcast platform.

Episode Transcript

Alex Larralde: Hi everyone. Here’s what you can expect on today’s episode. I’m joined by two guests, Benita Tapia, the administrator and director of nursing at Beverly Hills ASC, and Conor McGinn, the CEO of Akara. Akara has an AI platform that collects and analyzes surgical event data to help optimize scheduling. They use a thermal heat sensor that sits on the wall of the operating room to passively monitor key events in real time.

Recently they partnered with Beverly Hills ASC to pilot their technology. So today we’re going to talk about how that’s going, results that they’re seeing, and hear a little bit about where we think this kind of AI-enabled technology can take ASCs down the line. Benita is also piloting an AI tool that helps ensure patient data accuracy to reduce claim denials downstream, which she’s going to share a little bit about as well.

Then in our data segment, we’re going to cover the key benchmarks from our latest specialty data report on Otolaryngology. Revenue per case climbed year over year, but patient deposit collections dipped and claim denials increased. We’ll talk about why we think that might be happening, and a few practical ways that centers in this specialty can shore those numbers up.

I hope you enjoy today’s episode, and here’s what’s going on this week in surgery centers.

Alex Larralde: Welcome back to this Week in Surgery Centers. So, before we get into the questions, I’d love it if you two could just each briefly introduce yourselves, your role, and the work you’re doing, Benita, tell us a little bit about yourself.

Benita Tapia: I am Benita Tapia. I run four ASCs in Beverly Hills. Three of them are multi-specialty, then the other one is a GI center. I basically have started from the ground upwards. I actually trained in England. I don’t hold that against Conor, that he’s Irish. I did train with a lot of Irish nurses, but basically I’ve done pre-op, I’ve done inter-op, I’ve done PACU. Now as an administrator and a DON.

I always say now I’ve done a little bit of everything. I do nothing well, but I definitely look to make my life simple and, this is why I’ve been interested in the AI that Conor’s team has been working on.

Conor McGinn: I’m Conor McGinn. I’m the co-founder and CEO of Akara. We’re actually originally a spin out from Trinity College in Dublin, where I was a professor of AI and robotics for a number of years. We developed technology to automate routine tasks in operating rooms as we focus on both hospitals and ASCs.

And our objective is to help hospitals and ASCs reclaim time that they’re losing through preventable inefficiencies, ideally so that they can perform more surgeries.

Alex Larralde: Let’s get into it. I’m curious, how did you decide to bring in Akara and the other, AI technologies that you’re piloting right now?

Benita Tapia: So actually, I sat in the background in the beginning when Conor’s team came on and they’re like, okay. They said, robots, I know now it’s not robots that are coming in there. I think of like Dr. Who and Daleks and things moving around. But in the beginning, I didn’t take much notice. Then I started to see some of the information that was coming out of what Conor’s team, Akara, were producing, and frankly, I saw that there was a way for me to do my job more efficiently too, to look at utilization, look at what they call gap times look at data input too for, another system called Orbit., because data entry for the longest time has been painful because it’s human error. Then all of a sudden, I started to wake up and I’m like, oh, this could make my job easier, like on the ground level.

And that is exactly what I’m finding right now.

Alex Larralde: Conor, tell us a little bit about what exactly you’ve got going on with Benita’s team at Beverly Hills ASC.

Conor McGinn: Sure. So, I suppose the insight for us came when we started to meet with, people like Benita, who obviously understand the ASC and the operating room very well.

But the data they have at their disposable to make decisions is oftentimes error prone. because so much of the data comes from manual data entry where nurses who are already very busy are writing things down. And it can be really difficult to make decisions on this data because if you can’t rely on it you know, you might be thinking you’re making a good decision, but you’re not.

And these knowledge gaps we’ve seen endemic in, in operating rooms. So like understanding how long a surgery will take, for example, if you’re just taking the surgeon’s estimate, like that’s oftentimes not very accurate. Using historical data, again, if the data itself, you aren’t certain how accurate it is, that’s really, really hard to do.

Similarly in an ASC context. You’ve got such a big mix potentially of surgeons at your disposal, being able to kind of understand which surgeons should be allocated, how much block time, and when. Again, it’s a complex decision that really requires data science and a really good understanding of the fundamentals and the data.

What we’re starting to see now as we dig into like things like case costing is that there’s so many moving parts and variables to understand like, what are the practical things that can be done to improve the margins on cases. And it’s really in this area that we try and bridge the gap.

And again, we use the word knowledge gap or data gap, it’s just essentially being able to help connect the ground source of the truth of what’s actually happening with the key people who are making decisions. And working with Benita and the team in Beverly Hills, and in the affiliate group has been really helpful because essentially, they’ve been really helpful in sharing with us where their challenges lie.

And we’ve been working with them as a development partner, as a design partner around connecting those dots.

Alex Larralde: And Benita, what were the challenges that you brought to the table when you were kind of designing this project and these initial use cases? Did you have a stack ranked list of priorities or were you just open to whatever it was that you were going be able to start tracking?

Benita Tapia: I know my pain points. This is a pain point I can tell you for any administrator, especially with us being really, really busy, and one of the things is utilization, right?

That we, basically utilizing those operating rooms to the max. And also scheduling. So scheduling, like when a surgeon comes in, inherently he’s going to tell us what it takes 60 minutes? Well, we all know they lie and it takes a lot longer than that. But you do have your good surgeons, you have your surgeons that are efficient that underrun, and then ones that overrun time. So, for a start, a pain point for me, utilization to make sure that surgeon is using that block time efficiently.

The other thing is that, you know, the patients in the room and they’re waiting for the surgeon to come in. So, you’re like, okay, well where is that surgeon? Our job is to turn room over quickly and get that surgeon back into that room to be doing those cases so that the next surgeon starts on time. So that’s one pain point, which, definitely Akara has been helping with.

And then the other one is demographic errors. You miss a number off an insurance card. You miss a name or a date of birth, and all of a sudden you are getting a denial from the insurance company for a very simple thing. So, with Orbit, for instance, basically the intake comes in from the surgeon’s office and then the robot inputs all of that information into HST. And then it cross references to make sure that the insurance is correct, and the address for the insurance is correct, and the date of birth matches what’s on the insurance.

Alex Larralde: That’s great. And have you seen any improvement to date? What does the ROI look like so far?

Benita Tapia: Okay, so yes, return on investment. So first and foremost, for the orbit where they’re inputting the demographics, we are at 20% increase of not having as many denials. So automatically we’re not using staff to go back and look and double check.

I will say though, you always need a human to overlook and you also always need knowledge of an ASC and what to expect. So, it doesn’t take away from that human intervention. And then for Conor’s team, for sure, right now I have actually condensed surgeons’ time. Okay. Because then realizing you don’t need a block every week, you actually can push that into two or three blocks, right?

And then we’re looking at, okay, two rooms. Some surgeons have two rooms. They think it’s great, they think they’re more efficient, but the fact is they’re actually less efficient. You are having a room stay idle for a long time with staff in there. So that is a big waste of time, and we look to see actually, doctor, you can be more efficient with one room than two rooms. So yes, we already are getting some return on investment. We’re still in the early stages, right? So my whole goal right now is to kind of make surgeons accountable. To give them a moment in time to look at how long it really does take them to do these cases.

And hey, you’re not utilizing your block time. You have the next quarter to try and make that difference and make that up, or. Let’s like look at a different way of condensing the cases into you know, one or two blocks instead of four blocks. And now we’re moving to the other side, right? We’re moving now into the contribution margin we already have the revenue of the contracts are put into the system.

And now Conor’s team has already given me some information, it’s like, hey, there’s some cases here that there is no revenue on there. What happened with the contract. So, we’re moving now to looking at cost of supplies and preference cards, and then we’ll look into staffing.

So truly we’re going to have a good contribution margin of, who is actually efficient, who is most productive, who is actually bringing the revenue into the system to look to be more efficient for the future.

Alex Larralde: Super exciting. Conor, I would love to hear from you, are you seeing similar results in other pilots that you’re running right now?

Conor McGinn: Yes, absolutely. And I think it’s, the subtext to how we often work is like, what we like to do is measure, month on year data. So what we’ll do once we work with a partner each month, we’ll compare that month with the same month of the previous year. That’s the best way of benchmarking, efficiency and output at a center just because there can be variability between two consecutive months.

And what we’ve been able to see is a clear trend with Benita and her team that the volume of cases and also the revenue being produced in July and in August of this year has been noticeably improved over the year previously.

I wouldn’t allocate our improvements as being directly, uh, consequent. because ultimately it’s been people on the ground, like Benita and her team to make those decisions. But ultimately, we’re collecting the metrics that are helping inform some of that. And for us, that’s a measure that our kind of data is helping in that journey to increase throughput.

And for us, that’s the key metric to look at, is the center actually moving in the right direction and are we a part of that? I think as we move forward, and Benita alluded to this, is like helping address some of the cost related applications. We see a lot of opportunity here and a lot of the ASCs we’re speaking with, it’s the margin that they see the greatest opportunity to improve.

And whether that’s finding ways in which you can load balance staff a bit more so that you’re able to reallocate staff during the busier periods and maybe, over the quieter periods you’re able to cover some of that cost. We’re starting to see opportunity to be able to add value there as well.

Alex Larralde: I’m guessing Benita, your leadership team is really excited about this. How are you sharing this data back and keeping people updated on the progress that you’re making?

Benita Tapia: Well, I’m very excited because it makes me look good, right? I’m a nurse and my background’s clinical, so it’s not technical. You know, there’s times when I’ve been asking my children, like how I can do this Excel spreadsheet. In minutes, I get this data and actually, Conor’s team gives me a lot of data. So one of the things that I am super excited about now is that he has something else coming up where I ask Clara, give me this ENT doctor, uh, the last two months of his case utilization? Could you give me his contribution margin? Could you graph that out for me? Things that would take me so long to pull out of the system. I’m getting this beautiful data like really quickly.

And then, you know, Conor, I was talking to you about case mixes too. So if they have Blue Shield, if they have Medicare, if they have liens and everything else, and then I can pull very quickly like, am I missing Blue Shields not paying us for some reason. And then it will also let me know, yeah, this surgeon is efficient, he’s got a great case mix. He’s not using his turnover time or what? I love this, the gap time. I always think of gap time. I think about being back in the UK and I think about being at the train station and hearing “mind the gap.” In this case it really is “mind the gap,” because they need to get into that room, they need to turn it over that data, instead of me just having to like, and it takes me a while, you know, I’ve got staff, I’ve got patients that need help. So to get that data quickly is so valuable. Yeah.

Alex Larralde: Wow, that’s great. So tell me more about this, Conor. Okay. So you are querying the data in real time. Is that how it’s working or…

Conor McGinn: Yeah, so this is actually something that we built as an internal tool initially. I think the backstory of this is quite interesting. So what was happening was that we were capturing the data on our devices, and I have one in front me here. This is typically what sits on the wall in the operating room and it uses thermal sensing to kind of essentially see what’s going on. And it’s like an eye in the sky for the operating room. And the data, it was collecting, we were essentially aligning it up with the case schedules that we would’ve gotten from HST.

And we started to work within the data streams package that you’ve provided, and we started to see that there was actually opportunity to get access to a lot of really valuable data. And combining the two things together, we were able to connect with Benita and Andy and the team and like, the conversations we’d have, there would commonly be like lots of questions thrown at us.

Again, the kind of things that would be really difficult to be able to extract from any single part of an EHR, because it involves, comparison and aggregation of lots of different types of data. Our data science team was essentially like, going to each part of the, the database to try and pull this data from.

We realized that we could actually do this really efficiently using what are called AI agents. So these are little software bots that are trained specifically to do tasks involving manipulation of data. And what we were able to do is build a kind of a ChatGPT type of experience where, you know, rather than relying on a dashboard or a spreadsheet, you could actually ask it a question.

And our agents were able to automate some of those tasks. And this started to work really, really well. The consultancy work that we were doing with human teams, which was taking days and in some cases, weeks, we were able to do it in like, you know, 30 seconds. And we’ve gotten really excited by this because the opportunity for us to do this for not just getting data but being able to do predictive things, like, given these CPT codes, how long will this case take? These are the kind of directions we see it moving.

And it’s super interesting because being able to work with administrators, but also, nursing staff and you know, various different stakeholders within the ASC space, we’ve been able to understand all of the different angles that we might want to query this thing. And you can get agents to do so much stuff. It’s not just what I’ve just mentioned, but you know, the idea that you can build these little automations to do lots of tasks that would normally take a ton of time is quite exciting.

Alex Larralde: Yeah, that is exciting. I mean, it’s transformative. You know, you don’t need an analyst, you don’t need to have that expertise and you can just ask it, plain language questions, it sounds like, and get that data back in a matter of seconds. It’s phenomenal.

Benita Tapia: And it could be in real time. So you have the surgeon and the surgeon says, Hey, I want more time.

I need to increase my cases. And then instead of me saying, ‘Hey, I’ll get back to you in a few days,’ I’ll be able to say, oh yeah, I can give you this time because Clara says I can move this patient, this surgeon, and we can condense his time and we’ll put you in this time. So, for me, on the ground level, or any other administrator, who’s extremely busy, right? Is that we can get real time answers quickly.

Alex Larralde: Yeah. Yeah. Yeah. That’s a game-changer for sure. And really exciting.

Conor McGinn: One thing that maybe I can add, Alex. An interesting insight that we’ve seen, like across the board and most of the hospitals we’ve worked with, is that there’s oftentimes, a handful of people who are local domain experts, whereas they capture a huge amount of kind of anecdotal, local knowledge that the moment isn’t codified at all.

And that might be, you know, simple things like, there might not be an inventory system in place and just somebody has it in their head exactly where things are and how much things cost. And you might have a situation where you know, you understand which surgeons like, which operating rooms and their availability and all of these kind of things.

I think having a knowledge-based system that can represent this, has the potential to be a real game changer because it allows for scalability and, to decouple dependence on individual people. Because again, we’ve seen time and time again if someone’s sick or if something happens where you need to move, expand it to different sites or relocate people, then it’s really hard to do it because the infrastructure doesn’t exist. And I don’t think that there’s a really good alternative to this right now. So something like this I think could really help.

Alex Larralde: Yeah, that’s a great point. I mean, staffing and turnover have been challenges within the industry for some time. So being able to, future proof and build that business continuity into your model seems really helpful.

I’m curious about safeguards and what you do when something can go wrong. So, AI is powerful and there’s so much exciting stuff on the horizon, but there is also this question, I think, lingering in the back of everyone’s minds around accuracy and making sure that there’s, you know, some level of human intervention.

How are you thinking about that? I’ll start with you Conor. How is that kind of built into the product and into your systems?

Conor McGinn: Of course, a couple of points. Firstly, like what we’re doing is not like a clinical decision support tool, so we’re not using AI to kind of infer clinical or therapeutic outcomes.

So in that respect it’s much more about managing efficiency. So the, the cost of a mistake for us is usually quite small. In so far as like, you know, get a timestamp a few minutes out it’s not going to have an impact from a patient standpoint. In terms of like, where the biggest risks in our technology lies, I think privacy and the kind of legal side of things is an important consideration.

To address this, we use thermal sensing primarily as our data. And the nice thing about thermal sensing is that we’re measuring the heat signature of people in the room. So we can’t see any visually distinguishable features. It’s essentially a kind of a, a yellow blob that represents a person.

And we process that data so we can qualitatively see what’s happening in a room. We can see that there is a, or is not a patient present. We can see that a surgery is or is not taking place, and we can capture certain timestamps within that process. But like we don’t have any identifiable information that if you were to get that data, you’d be able to derive. You know, you could identify the place, the people, and that’s, that’s not something that can be done easily or at all with the data for that matter. The other key piece, which again, we’re very aware it’s a consideration for a hospital, is that if, if there was a case involving medical malpractice, having data from that room could be potentially, a problem, again, with the data we capture is an extremely low resolution. So it sits on the wall, it’s very grainy. You wouldn’t be able to pick up any information like that. So again, when we show the data we capture back, it really puts people’s minds at ease.

In terms of other aspects of the technology, in terms of the decisions the AI is taking. We maintain a human in the loop at all times. So what happens is that the AI will produce an output of what it thinks is happening, and at the end of each day, once the data is aggregated, we have a human look for anomalies.

So if, for example, we see that a surgery that was supposed to take 60 minutes took, 90 minutes like that would flag to us that something is worth a human looking at, and what’ll happen is the one, someone on our team will look at that and if there is a mistake in the AI, we can correct it, in which case the surgical records integrity stays consistent. And if there’s a case where it is 30 minutes out, our person will review it and we can actually make comments as to what happened. At the moment in, in the EHR data, you’ll just see the time being out. You don’t really know why it happened. We can actually help dig into that detail and we can provide feedback loops to the hospital.

Say like, you know, we’ve noticed these cases running late, they’re hand surgeries or whatever. It’s a specific surgeon and like, this has happened three times now. So all of a sudden we’re capturing data that’s not currently being captured, and we’re gaining context on top of the quantitative data, which could be quite helpful if you’re trying to bring about quality improvements.

Alex Larralde: Yeah, absolutely. And then Benita, from your perspective, I’m sure there were probably a lot of questions internally around security, accuracy, privacy. How did you navigate that?

Benita Tapia: I think really the safeguards, there’s no personal health information that can be shared, right? Policies and procedures have to be in place to do that. The other thing with safeguards is really auditing things that are happening that are on our server and are occurring. We did have a bad situation with ChatGPT and that like kind of alerted us to how things were being used by employees.

But the other thing is that for safeguards, so for instance, like Orbit where we say, okay. This is all data input, right? But data input that goes onto this scheduling intake form that gets read by a robot has to come from the surgery scheduler.

So you’ve got to have a human overlook that because it learns from what we give it. And if it doesn’t know the information, you’re going to get the incorrect information. And that happened like at the moment with Orbit where supplies were incorrect and the robot hadn’t learned that this is this procedure and this procedure needs this.

And then again, with Akara’s information, as I get it, you have to understand what you’re looking at, right? So if he throws a bunch of numbers out of me, like saying there’s no zero revenue, I basically know for certain cases why that is. That way or why it’s less than the contracted amount.

So you have to have a knowledge of what is happening in your ambulatory surgery center, and that only comes with like an administrator that’s been doing this for a while. What he gives me is a lot of information, but you have to have a safeguard of a human. And you have to have policies and procedures in place so that this AI like ChatGPT does not get used in the wrong way.

Alex Larralde: Absolutely. And that’s something we haven’t touched on yet, but I would love to kind of hear how are you using tools like generative AI to help with processes?

Benita Tapia: I mean, it helps with policies and procedures, right?

With templates, with check sheets. You know, I just did one recently and it was for SPD and it was for compromised instruments. Now, what would’ve taken me such a long time to kind of remember everything, or pulling my SPD people to do that, it produced this checklist. And it was just beautiful.

Again, I had to have the team look at it, but the time that it took to do it, it also analyzes for me. So if I have quality improvement studies, I can throw that information in there and up pops this nice and analytics or graph or whichever way I want it to be. Or it could be common things like, oh, you know, joint commission, your accreditation, people are coming in, tell me the most common questions in deficiencies that they’re asking. So we use it in a lot of ways and, it’s been used in the wrong ways. That’s being given me some problems. Uh, you know, employees that basically we’ll use it instead of talking to a person, coming and saying, I have a problem with my coworker, or This isn’t working instead of face-to-face.

And this is becoming a big deal with chat GPT right now, where, you know, teenagers, employees, they’re using it instead of talking to a person. And that becomes this serious problem, especially when ChatGPT tells your employee to resign. Yeah. Cannot even answer an email or a text message without popping it into ChatGPT. There has to be parameters, there has to be oversight, there has to be human interaction. And as long as all those parameters are into place, this can be such an amazing thing with AI to help and it allows me to have more time to do the things that I feel are very important, which is quality patient care.

Alex Larralde: Absolutely. So that’s a great point. It’s a great tool to streamline things, drive efficiency, but it’s not for everything and it can’t replace human interaction.

I’m curious, if somebody’s evaluating and thinking about how they’re going to get started actually incorporating AI into their tech stack, where should they start? What advice do you have for them? Are there specific things they should be looking for in vendors? Questions they should be asking?

I’ll start with you, Conor, since you have a lot of these conversations all the time.

Conor McGinn: So I think before I’d get into the AI question, I pick a step back and I would say like, how reliable is the data you’re capturing? Because any AI system, and there’s like so many applications for AI, they’re only increasing. But if you’re not very confident in the integrity of the data, the AI is going to give you bad results. So I take a step back and I, I’d say, you know, are you actually measuring the thing that matters or are you just aware of it? So things like, for example, we hear this a lot, like we have a long turnaround.

Okay, that’s good to know. But are you tracking that on a daily basis, a case basis? Are you really confident in the metrics that you’re producing around that? Because awareness can be limiting because if you don’t have a way of measuring it and you try something new or you introduce new practices, it’s going to be very hard for you to tell if it’s moving in the right direction.

Because like, you might have, you know, if you’re doing 10 cases in four operating rooms each day, there’s going to be lots of those data points. And if you’re relying on a kind of an anecdotal. Perception of things and that’s not going to be very effective. So being able to measure those things is really, really important.

And the kind of follow on question I’d say is like, are you at a point where you’re capturing data enough, well enough to be able to focus in on the bottlenecks? Have you been able to quantify your bottlenecks? And if you’re at the point where you’re doing that. And then I think AI is a really good opportunity because you can start to dip into tools that can improve things better.

If it’s a case where you can’t do that, then my suggestion for that ASC would be, you need to invest in improved data collection processes. And again, there’s lots of opportunities, ours being one of them, to do that. But there is of course lots of other ways to do it.

And a lot of it’s a mindset. It’s a database mindset where like in, we’ve seen sport become more data-orientated and how, professional athletes train and prepare. I think we’re starting to see that now filter its way into industries like ASCs and in hospitals.

Alex Larralde: And then Benita, what about you from, from your perspective as the buyer and someone who’s doing this evaluating, what advice do you have for your colleagues and peers who are looking to start doing the types of things that you’ve implemented?

Benita Tapia: I think number one, which was great, because of two of the big AI systems that we’re using right now, which is Akara and Orbit, it’s a trial. So it’s not like if this doesn’t work for me or it doesn’t work for my team, at least we have time to kind of trial it and we’re not already committed to something that maybe doesn’t work for our team.

And the other big thing is communication. Okay. So a lot of time I have vendors and they’re like, okay, this works in this center and you need to use it this way. Well, no, it doesn’t like, like I know what works in this center. My team knows what they want. And I must say a wonderful thing with Akara is communication.

So for instance, like I’m, my day’s busy. At the end of the day, I’m emailing, I’m looking at this thing, and I get the response back about, yes, I think we can do this. This is what I have right now, because ultimately this is here to make our lives easier and simpler. And I need you to communicate and tell me, yes, we can do this, or no we can’t, or ask me why I need that. It’s that big communication that I think is extremely important.

Alex Larralde: I love that. Okay, and so one more question, similar vein, but we’ll zoom out, and this is really just in general for ASC leaders and administrators. What is one piece of advice you would each give to an ASC administrator that they can start making improvements in their surgery center this week?

Benita Tapia: I think this is easy for me. It’s a team effort. It’s not just me. It’s not just Conor the vendor. He becomes part of our team too. And the thing is that you have to listen to your team, and if there’s an issue or an problem, we can work it out as a team.

I am not an administrator that’s ever punitive. If you have a problem or you’ve made a mistake. Come tell me, we’ll fix it together. We’ll figure out a way for it not to happen again.

Alex Larralde: I love that. And what about you, Conor?

Conor McGinn: I think kind of two general kind of points. One would be to like, incentivize people on the ground to engage in technology adoption.

Because again, I think that there can be challenge as an outsider or a vendor coming in because like if people are very busy and like what they’re being promoted and incentivized to do as the day to day, it can be kind of difficult. It takes a while to figure out how to make this work well for them.

And as Benita says, like our job is. It’s to make life easier. But if there’s people in the ground not incentivized to engage with us, then it’s really hard to get to a position of doing that. And understanding that there is a little bit of upfront, you know, support needed in the early days to just get to a point where it adds value in the medium to long term.

And I think going beyond that, and we’re extending that, is that viewing vendors less as a kind of transactional thing. And I think this is a really nice thing within the ASC business. It seems like it’s much more collegial than in some other places we’ve experienced where, you’re kind of working together towards a shared objective and a shared goal. It’s not just a case of like, you know, you do this and we’ll do this, and it stays that way. We’ve benefit mutually by being able to. You know, learn what the actual pain points are and similarly get the opportunity to work closer with them to customize and build bespoke systems.

And that’s a mindset. If you expect it to work outta the box immediately, that’s a much harder expectation to live up to.

Alex Larralde: This has been fantastic, I’ve learned a ton from both of you, so I’m really grateful for you to take the time to join us.

Thanks so much.

Benita Tapia: Thank you.

Conor McGinn: Thank you.

Alex Larralde: This week, let’s take a closer look at Otolaryngology or ENT ASCs. Our latest specialty benchmarking report looked at nearly 112,000 cases across 202 centers, and the story it tells is one of steady volume, flat cancellations, a strong lift in revenue per case, and a few pressure points on collections and claim denials.

Let’s start in the OR. OR block utilization is 46%, which signals stable scheduling, but leaves some room for improvement. Broadcast or availability, enable electronic scheduling and automate block time management, so minutes don’t slip away and you can move closer to that industry gold standard of 70% OR utilization.

Preauthorization approvals edged up from 27% to 28%. Verify the authorizations that the physician’s office secures and use integrated payer technology to streamline the process, reduce administrative burden, and stay compliant. This critical step helps prevent denials and reimbursement delays by forcing confirmation of insurance coverage upfront.

Eligibility verification remains very high at 91% slipping just one percentage point year over year. Consistent verification helps avoid last minute delays and cancellations. The recommended cadence is simple and reliable. Reverify twice when the case is accepted, and again on the morning of surgery, and then run monthly checks to catch coverage lapses, using automation to minimize manual work.

The cancellation rate is flat year over year at 18.3%, and patient decisions are cited as the top reason at 41%. While some cancellations are unavoidable, many can be reduced with better processes. Fixing things like missing labs, scheduling errors and inefficiencies that come from the surgery center are totally within your control.

Analyzing the reasons on a regular basis can help you keep your schedule optimized and reduce lost revenue.

Patient deposit collection dipped from 68 to 66% last year over 2023. Provide accurate estimates one to two weeks before the procedure. Make them clear, easy to read, deliver it by text or email or both, and enable upfront payments. Doing so builds trust, reduces cancellations, and helps improve satisfaction all the way around.

Days to bill improved from 10 to nine days, showing some forward movement on that metric. The fundamentals behind this matter. Everything from having accurate coding and charge entry, having really well-trained coders, making sure that you have an integrated EHR and billing system to help reduce errors and that you’re also tracking denials so that those repeat mistakes don’t creep back in. All of these practices help strengthen financial health while keeping operations smooth.

Claim denial rate increased from 15% to 16%. Remember that efficient claim management depends on using EHR, practice management, and electronic claim systems to help streamline submissions and reduce errors, plus regular monitoring so denials are addressed quickly and reimbursement stays timely.

Now here’s the big headline where we really saw some movement.

Net revenue per case rose 26% last year over 2023 from $2,981 to $3,765 per case. To keep that momentum, make monthly financial reviews a fixture, so trends surface and issues get addressed quickly. Pair those reviews with analytics tools and standardized reporting to support data-driven decisions so that you can improve revenue, control expenses, and protect long-term financial stability without guesswork.

And finally, volume. Average monthly case volume at ENT ASCs Rose from 90 to 93 last year. As you know, tracking volume helps with profitability forecasting, resource allocation, expansion planning, and it feeds back into OR utilization so that you can make sure your staffing matches the pattern of demand as it starts to shift, it’s the steady pulse that keeps clinical and financial operations calibrated.

And that’s all for our episode this week. I’m so grateful that you tuned in, and I really hope that you enjoyed my conversation with Conor and Benita. I’ve learned a ton from both of them about operationalizing AI in the ASC, and I hope that you did too. If you enjoyed this episode, make sure to leave us a rating or review on your favorite platform, and we’ll see you again soon.