← Becker’s Healthcare Podcast
Becker’s Healthcare Podcast
Tackling Health System Capacity and Throughput with Mohan Giridharadas, Founder & CEO of LeanTaaS
00:24:22
Loading summary
Transcript29 lines
- [00:00]
A
Welcome to a special episode of the Becker's Healthcare podcast, A rapid fire Q and A with operational leaders and Mohan Girardadas, CEO of Lean Toss. I'm Molly Gamble and today's conversation is part of transform. That's a virtual event co hosted by Lean Toss and Beckers Healthcare September 16th and 17th. To join Transform, find the link in today's episode description or visit conferences.beckershospelvew.com September-2025 Transform Virtual Event now speaking of dashing, let's jump right in and hear Mohan's insights on the challenges and opportunities shaping health system capacity. These were questions straight from attendees at Transform, and Mohan answered them on the fly with me. Hospital capacity, operations, throughput and access. These aren't just challenges, they're urgent pressure points growing more acute by the day. At Beckers Healthcare, we report on both the reality straining health systems now and the forecast that signal an even tougher road ahead. Projected shifts in insurance coverage, worsening physician shortages, rising patient acuity, hospital closures and shifting care volumes. But these stakes aren't off in the distance right now, emergency departments are overflowing, beds are backed up and wait times are climbing. For hospital leaders, these issues demand action. The only variable is speed. I'm Molly Gamble with Becker's Healthcare, joined by Mohan Giraudadas, Founder and CEO of leantas. Hello Mohan. Thank you so much for joining me today.
- [01:29]
B
Hey Molly, good to be with you.
- [01:31]
A
Likewise. We are setting out today to tackle these capacity realities head on with real answers to unfiltered questions submitted by executives attending the Transform Hospital Operations Virtual Summit. That's a collaboration between Lintas and Beckers Healthcare. So Mohan, thanks again for being here. I'm looking forward to your responses to these questions that are again are unfiltered. Are you feeling ready to dive in and tackle some of the health system leaders most pressing issues and questions?
- [01:59]
B
Let's give it a shot. Depends on what the questions are.
- [02:02]
A
Well, let's start with the question from a CFO of a 650 bed flagship academic medical center. Their question? How to leverage AI to improve the bottom line.
- [02:17]
B
AI and algorithms in general can drive operational excellence. At the end of the day, improving the bottom line is improving the operational excellence with which hospitals and health systems execute their day to day work. If I think about healthcare in the US over the last 30 years, the clinical advancements have been magical. Robotics, genomics, precision medicine. But operationally, many things are still way behind two People still chat and make appointments. Surgeons are still faxing cases into the depot Et cetera. To drive operational excellence, three things need to happen. First, the asset needs to be optimized. Optimize the asset in terms of utilization, how well it's used. Second is optimize the staffing on the asset, because a fully configured operating room without the right staff to run it is not fully there. So optimizing the staffing on the asset. And third is optimizing the flow of the patients through the asset. This is a classic network optimization problem. When we think about what airlines do, they optimize the use of the plane. The plane's only making money when it's full and it's in the air. So they turn it around quickly, they fill their seats quickly, they optimize the flow through the assets. Meaning large hub airports move many, many more people than smaller regional airports. So this is a known problem that can be solved with math and algorithms and AI gives us a unique opportunity to go after it.
- [03:47]
A
This next question, Mohan, is from the CIO of a regional academic health system. Not so much a question as a sentiment that I think we can probably all relate to. There must be a better way of doing things. What do you say in response to this thought?
- [04:01]
B
There is, and. And we're doing it. The key thing here is healthcare is a service. In a service, supply and demand have to continuously match. In a manufacturing context, they don't have to. If you get an order for more stuff than you can produce, you just pull it from the shelf and ship it. But in a service, you can't put either the patient or the cardiologist on the shelf. Therefore, the supply of the assets and the staff have to match the pace at which the patient demand is coming in. There are mathematical algorithms for predicting the demand, the volume, the mix, the timing, as well as on the supply side, the staff, the equipment, the rooms, the availability, the constraints, and trying to make it balance. Second is automating the routine work. A third of what the frontline does is stuff that a seventh grader could do. Pull this number from the ehr, write it on a yellow, post it note and go to a huddle meeting. Cut and paste this spreadsheet into another document. Look at this dashboard. Go chart diving in the ehr. These are things that could be automated and therefore you're right, there is a better way of doing things. It's automating the routine work. And the third thing is to predict and prescribe scoreboards. Don't win football games. Better plays win football games. And in order to run the better play, you have to be able to predict in advance what's going to happen, prepare to be able to execute that, and then prescribe the steps. And so that's kind of how we think about the better way of doing things.
- [05:36]
A
I love that scoreboards quote. It's the perfect response to anyone who thinks another dashboard is the solution for things to run differently. Let's go to the question from a CEO of a rural community health system. This is again, not so much a question, but a need. This leader is facing ways to lower cost and continue to give high quality care and keep a positive bottom line.
- [05:57]
B
We started the ORs. The ORs are the financial engine of the hospital. They contribute half the revenue, more than half of the income. So optimizing or utilization drives both top and bottom line growth. Managing inpatient flow unlocks economic performance as well. Just because getting higher flows means more inpatient admissions, means fewer transfer declines, it means fewer excess nights, etc. The important reason to think about improving the bottom line through Superior Operations is every $2 unlocked in operating margin improvements is equivalent to a hundred dollar increase in net patient revenues. So if you multiply that by a million to improve operating performance by 2 million is quite doable. Improving net patient revenue by $100 million is a multi year effort and so the leverage is very much on improving operational performance of the existing assets. In the past, health care got very comfortable with building their way out of trouble whenever capacity got tight. It was build more hospitals, build more units, build more beds, build more, build more ORs, hire more staff. With the financial environment the way it is, building the way out of trouble is not as feasible as it used to be. Therefore, improving the utilization and throughput of the existing assets is where the leverage is.
- [07:26]
A
Especially now. Mohan, I know that backers in the newsroom were just reporting on organizations that are hitting pause on different expansion efforts and capital planning projects. So it was true then. It's especially true after this summer. Thanks for your answer to that leader's question. We've got another question here from a senior Regional director of a major cancer center. How are major EHR companies like EPIC and Cerner incorporating AI into OR and same day surgery workflows?
- [07:53]
B
I might be a bit biased on this answer, but EHRs have been around for 45 years. Both EPIC and Cerner were founded in 1979. For 43 of the 45 years, the words capacity optimization and EHR were never used in the same sentence. So this is what we've got. EHRs are excellent, excellent sources of truth. They're excellent repositories of patient data. They are not the source of intelligent optimization. They are not the source of predictive and prescriptive analytics because they don't have the view of all of the data. They've just got the EHR data. All of the AI announcements from EPIC and Cerner in the recent weeks seem to be focused on ambient listening, inbox management, et cetera, which are very valuable and helpful things to do. They aimed at physician burnout from the data entry tasks. When I think of optimization and AI in the context of OR and same day surgery flows, they're related but not interchangeable. Optimization is the act of figuring out the supply, demand, balance, figuring out the flows, et cetera. We simply have not seen evidence of sophisticated prediction, yield management like optimization, thinking, et cetera, from either of the EHRs. In the way that ORs are managed, their utilization is optimized, blocks are allocated, staffing is optimized, or the end to end flow of patients from the surgical clinic, often run by community surgeons who are not even on the EHR in the first place, all the way through the surgery actually being executed. So we think there's a ways to go there.
- [09:35]
A
I understand what you're saying, Moha. It sounds like that the specialty there is retroactive or retrospective data, whereas you're talking about predictive where things are going and how you can meet those demands. Let's move on to a vice president of case management at a large academic health system. Their question here is what are the current strategies to remove bottlenecks?
- [09:56]
B
Let me take that with a metaphor. Think about before we had GPS in our cars and on our phones. How did we know about a traffic jam? It was after we got stuck in it. That was the first we knew that there's going to be a traffic jam. Up until then we were blissfully ignorant as we drove down our chosen route. And how did we get out of the traffic jam? We went back to first principles and if we happened to know the neighborhood, we knew how to get off the freeway and navigate. If we didn't, we pulled off the road, opened the glove box, took out a map and figured it out. So we didn't know about a problem until we were neck deep in it. And we then, after being neck deep in it, started to figure our way out. That's what healthcare does every day. So healthcare lives in a pre GPS world. Every day most hospitals wake up to the same message. The EDs are boarded, the PACUs are boarded. It's taking four hours to get a patient into a bed and Discharges are delayed. So it's the same traffic jam, but they wake up to it and have to deal with it. From first principles back to the metaphorical world with Google Maps, we predict the drive time and the traffic pockets days in advance. If it's an important drive for an important meeting, we figure out the route. And even if we didn't do that while we're driving, we know when traffic is starting to form, what the alternative routes are automatically, and we still decide whether to take the recommendation or not this can be applied. Think of the discharge process. We normally get delayed in a discharge because something happened. Oh, this test wasn't run or this happened. Well, we could have predicted that three days ago and we could have done something about it, but we waited till we got stuck in the traffic jam. So what? The way to solve this is think of it as predicting what's going to happen, figuring out the right pathways through it, anticipating the barriers, and then nudging the work list, nudging the actions, sending the right messages, trying to be like the invisible hand that makes it work. And when that happens, automatically, discharges just move quicker and that's how you end up removing the bottlenecks.
- [12:12]
A
So appreciate the gridlock and traffic example. And it seems like in some of these questions there's some familiarity with the bottlenecks, like they're always going to be there, Mohan. So I think you're really challenging us to think differently about some of these things that seem so permanent. We have another question from a CMO at a large regional health system. They'd like to know how to improve throughput with limited SNF beds and limited resources. This is something a lot of health systems are facing.
- [12:41]
B
This is a big problem because the SNF availability results in a bottleneck. And then a patient can get discharged because they're getting discharged to a snf. Well, if the SNF can't take them, that patient isn't going anywhere. And so essentially they're blocked. We've looked at this multiple different ways. SNF capacity is actually quite opaque because the SNFs have, in the interest of their own economics, very tight views on who they take. And so they don't want to advertise. They've got plenty of extra beds. They want to make sure they're getting the right kind of patient for the right care with the right payer model, et cetera. So SNF capacity is always a challenging problem. But what tends to happen is hospitals focus on the hardest to discharge cases and they get stuck because moving to a sniff is a hard to discharge case and they get stuck on that. Think of Russia on the freeway. Again, back to what you were just talking about, Molly. Why is it that the commute time is half on a bank holiday? Half the people in the world aren't bankers. 5% of the people in the world are bankers. But it's the right 5%, the kind of work 8:30 to 5 if you get them off the freeway, life becomes great. For the other 95%, their commute times are half. So the trick is find the bankers. Now, in banking it's easy. The banker yesterday is a banker tomorrow. In healthcare, the patient yesterday is not even in the hospital tomorrow. So it's a dynamic find the bankers. So when you focus on accelerating the discharges, despite the limited sniff, there are other people where the discharges can happen faster. And imagine if you found them and moved it through. You unlock the capacity and the velocity goes up. So rather than focusing on what's not doable, think of it as focusing on the bankers. Focus on accelerating the discharges where you can actually move the needle, anticipate, predict, prescribe, make the invisible hand thing happen and you can get the throughput that you're looking for.
- [14:43]
A
Right. So Mohan, instead of focusing on the 25%, you're saying find the bankers, focus on first 5%, move those through quicker and then the system in some ways will take care of itself and you'll be on to your next patient population to move.
- [14:57]
B
Exactly.
- [14:59]
A
Fantastic. Let's go on to. We have a system director in patient logistics at another regional health system. This is a two prong question they'd like your thoughts on. Mohan. Number one, expediting discharge process, getting discharged patients out of their beds and then number two, ER and boarding workflows. We could probably have an hour long webinar on each of those. But what would you first and foremost want them to keep in mind about those pain points?
- [15:24]
B
Let's start at the expediting the discharge process and getting discharged patients out of the bed. It relates a little bit to what we were just talking about. But the mental model for me is think about what a hotel does. What hotels do is they don't let you check in before 4:00 clock in the afternoon and they force you to check out at noon. So they've bought themselves four hours of downtime to get the room ready for the next guest. Hospitals unfortunately end up doing it the other way around. The arrivals start in the morning from overnight boarding, from surgeries etcetera but the departures don't happen until late afternoon. And so arrivals before departures creates the problem. Up until now we got away with it because we had 30 or 40 extra beds. So that was a bit like a credit card float where you can spend money you don't have because you're counting on getting the money before the bill shows up. And so when those extra beds are gone, every day starts with the same gridlock problem. And so what needs to happen is the discharge process has to be linked to a very tight census prediction. Because every hospital every day says, let's get more discharges done by 11. Well, a 500 bed hospital will do 100 discharges a day. Trying to get all of them done by 11 is not practical. And so the chasing tends to be not surgical and precise. The right way to do it is to predict the census at a fine level of granularity, unit by unit. So imagine if we could predict unit two is going to be two beds short at 2:00pm this afternoon. I mean, it's predicting that at six in the morning and updating that prediction every five minutes or every 15 minutes, knowing that unit two is going to be two Beds short at 2:00pm you look at the 100 discharges slightly differently. You say, are any of those Unit 2 discharges? Because by the way, if I got them out of the way, by the time 2 o' clock in the afternoon rolls around, unit 2 won't have a problem anymore. And so by tying it to a unit level census prediction, you can now use the invisible hand to accelerate unit two discharges. Bump them up a little bit on the imaging queue, bump them up on the durable medical equipment being delivered to their room, bump them up on the multidisciplinary round, whatever else you need to do. That starts to happen and it creates the invisible hand that then gets the right discharges out. It's like fixing a pothole before the car went over it. So the unit 2 didn't even know it, dodged a bullet that it was going to be bottlenecked. That's kind of one way of thinking about it. The second thing is on the second part of the question, the ers and the boarding workflows, part of the reason we get ED boarding is because there's no place to put them in. A very high percent of the ED admits end up in a bed. And so if you started to predict, based on the census of who's coming into the ed, what kind of a unit they're likely to be transferred into, and bump those up on the discharge priority. The way I was just describing what's going to happen is then there'll be a safe landing spot. The ED will then get unblocked which then reduces your left without being treated metrics at the ED level. So these things are all interconnected and solving one without the other is a little difficult. So you've got to think of it as a network, a fully connected network and solve it in that manner.
- [18:49]
A
Right, Right. It sounds like the specificity is a big, big piece of what you're underlining is most important those unit level specifics and insights. We've got two more questions, Mohan. This next one. An AVP of surgical services at a large pediatric health system. I would like to understand the potential ROI and increase in or utilization using this product. By this product meaning Lean Toss services.
- [19:16]
B
Collectively our product is in 700 hospitals. About 6,000 ORs are being optimized on it. And of course the. The typical impact varies. It depends on a lot of things, depends on the underlying demand, the starting point, et cetera. But we typically find improving the staff room utilization by four to six points is a very doable thing. And the vast majority of our deployments do better than that. In economic terms, we mentally count on delivering 80 to $100,000 per or per year, which is quite significant. And we are so confident of it, we guarantee it, meaning we say at the end of the contract period, if we haven't delivered at least twice the cost of the product, we will refund the gap. That means there's no doubt whatsoever that it's going to be a 2x return on investment. Typically we get 3 to 4x and so the 2x in some sense is just guaranteeing and taking into account that there'll be some ramp up for the implementation. Things may vary, et cetera, but we felt that that's the right way to strike the balance to demonstrate our commitment and confidence.
- [20:38]
A
You're really sticking your necks out with that ROI pledge. That's really distinct and special. Our last question from a manager in business operations. This is at a heart institute with an academic medical center. With such little information available on the evaluation of AI. What are some of the best ways to evaluate the technology before bringing it into your institution? Any gut checks you would recommend, Mohan?
- [21:00]
B
Yeah, I'm going to focus on operational AI. We don't do anything in clinical AI, so those have a whole bunch of other challenges with it. Because you certainly don't want to be bringing in AI that makes incorrect clinical diagnoses and puts patient lives at risk, et cetera. So I'm not going to touch anything on clinical, since we don't do that on the operational, which is more like our products on optimization and AI, we think you can understand what the product will do at a detailed level and the impact that it could potentially have. Long before you deploy anything, you could watch the demo, build ROI prediction models and so on, and get a fairly good handle on what do we think the impact's going to be, etc. Many health systems we see deployed at one facility. So if it's a hospital with multiple, if it's a health system with multiple hospitals, it'll start at one hospital, demonstrate the results, make sure it's working, make sure the staff like it, make sure the change management and the technology and the automation are all there as advertised, and then scale it across the rest of the hospitals in the system. But I would say there is a minimum unit. It's impractical to think of partial deployment. For example, if you had a hospital with 20 ORs, you can't really say, let's deploy this at two of the ORs and see what happens. It's got to be a bit of an all or none at a hospital, otherwise you'll confuse the staff where they say, oh, if I'm scheduling into or number one or two, I use this method. Otherwise I use that method that becomes not very practical. And so you've got to find the right minimum scope that is contained, is safe and delivers the impact before you worry about scaling it.
- [22:49]
A
Fantastic. That brings us to the end of our questions. I want to thank all of our participants and our attendees for transform for their questions. It's always so helpful to understand the pain points and pressure points that you're up against. And Mohan, I want to thank you too for your unfiltered guidance and answers. Is there anywhere that you would direct attendees to transform for more information if they want to learn more about lean tossed services and even your thinking about these capacity issues that health systems are facing?
- [23:17]
B
Well, we wrote a book. It's an excellent sleep aid for anyone who wants to learn more about how we think about it. It's called Better Healthcare through Math. It's available on Amazon, but we are happy to send people ebooks or physical books if they want it. And then of course our teams are always available to engage and do demos. And we work at 200 health systems across the country and I'm sure many of them would be happy to talk with you and speak on a firsthand basis. As to what they have seen from working with us.
- [23:49]
A
Fantastic Mohan. Always a pleasure catching up with you and learning from you. Thank you so much.
- [23:53]
B
Thank you Molly. It was great to be with you.
- [23:56]
A
That was just a preview of the discussions you'll find at Transform, our two day virtual event on September 16th and 17th. For the full lineup and to reserve your your spot, check the link in the episode description or head to again conferences. Beckershospitalview.com September 2025transform virtual event. Hope you'll join us and we hope to see you there. Thanks so much.