WEBVTT
00:00:00.160 --> 00:00:08.720
Hello everyone, my name is Matt Harrington, and I am your Holly Beth host for the day for the Inside ABA Cares podcast.
00:00:08.880 --> 00:00:15.039
Your behind-the-scenes space where we lay the foundation and set expectations for who you'll meet this August.
00:00:15.199 --> 00:00:22.640
Now, most practice owners I talk to can tell you that their billing problem feels like the primary challenge in their business.
00:00:22.800 --> 00:00:26.000
It's almost like a death by a thousand paper cuts.
00:00:26.160 --> 00:00:29.679
Plutus Health is one of the companies that solve this problem.
00:00:29.920 --> 00:00:36.719
They're a sponsor of the summit in August, and they're one of the main reasons we're able to put on this conference at the scale that we do.
00:00:36.960 --> 00:00:47.600
Thomas John, my guest, is the founder and CEO of Plutus Health, which processes$3.5 billion in medical claims a year with denial rates under 5%.
00:00:48.320 --> 00:00:50.479
That's roughly half the industry average.
00:00:50.719 --> 00:01:04.799
He's been doing this for 30 years, and in 2023, he spun up Artemis, which we did a podcast with the other day with Cherie, specifically because he kept seeing the same operational failures repeat across ABA practices.
00:01:05.120 --> 00:01:07.200
Thomas, thank you so much for being here.
00:01:07.359 --> 00:01:13.359
For those who haven't met you yet or heard about Plutus, give us a little bit about your background and what Plutus does.
00:01:13.760 --> 00:01:14.480
Thank you, Matt.
00:01:14.560 --> 00:01:15.519
I appreciate that introduction.
00:01:15.760 --> 00:01:16.879
My name is Thomas John.
00:01:17.120 --> 00:01:19.200
I'm the former and CEO of Plutus Health.
00:01:19.359 --> 00:01:22.640
I'm also the founder and CEO of Artemis ABA.
00:01:22.719 --> 00:01:26.959
And my background is I started in technology about 30 years ago.
00:01:27.359 --> 00:01:31.920
And uh, you know, revenue cycle management found me, and ABA found me as well.
00:01:32.079 --> 00:01:41.120
So it's more of a pattern, you know, that I got introduced into some of these in-market opportunities that that then break the company that we are today.
00:01:41.359 --> 00:01:51.920
Across the last 30 years, I've been dealing with uh, you know, technology and claims in the uh surgery center market and the ER market and in behavioral health in general.
00:01:52.079 --> 00:02:05.439
And one of the things that I noticed over the years is uh just a pattern emerging around uh how revenue cycle management is an operating system and not just a bunch of activities that most people like it to be.
00:02:05.680 --> 00:02:11.439
So my background in technology allowed me to look at this from a process, technology, and people perspective.
00:02:11.599 --> 00:02:18.000
And that's how I ended up uh founding Pluto's Health and focusing on the ABA space over these years.
00:02:18.240 --> 00:02:31.520
For example, uh when I look at ABA, what I realized is that care is sold in in pre-authorized package of units by insurance companies, the way they control this care delivery process.
00:02:31.680 --> 00:02:39.919
And the other thing I found out is that, especially in the ABA market, it is not one pair, it is a lot of different pairs in different states.
00:02:40.080 --> 00:02:51.360
Medicaid governs the majority of the pair landscape, and every state in the country has a different nuance to how it pays and how it adjudicates claims and so on and so forth.
00:02:51.520 --> 00:03:03.439
So that's what drove me to look at ABA as an market in which my experience in technology and RCM could be applied, and that's how we built Bluetooth Health over these years.
00:03:03.759 --> 00:03:13.680
Can you uh zoom in on the way that you've looked at it from an operating systems perspective versus just a general, hey, here's a problem, let's hire someone to solve it.
00:03:13.840 --> 00:03:23.919
I'm curious, especially in the way you think about the vastly different ways different insurances handle claims and denials and all of that stuff.
00:03:24.319 --> 00:03:24.960
Great questions.
00:03:25.120 --> 00:03:34.719
When you look at the insurance landscape in the ABA world, the first big difference is the commercial versus the gentleman payers.
00:03:34.879 --> 00:03:40.879
So commercial payers could be the Blue Cross, Blue Shields, the Optons, and or even ITNA and so on.
00:03:41.039 --> 00:03:45.520
And then on the other hand, you have all the Medicaids that are state-specific.
00:03:45.680 --> 00:04:05.360
So, and then when you look at the processes which are governed by these insurance companies, they start with credentialing pay providers, and there are different requirements of documentation, different timelines to get it done, different frameworks of when a panel is open or not in that world.
00:04:05.520 --> 00:04:13.120
Same thing with prior authorization, different documentation requirements, different timelines for approval, uh, and so on.
00:04:13.280 --> 00:04:26.639
And then finally, the most important thing is the claim itself and what governs documentation standards around session notes, treatment plans, all of these have different flavors in different pair uh structures.
00:04:26.800 --> 00:04:50.000
But the most complex differences come inside the Medicaid world where you might think that a particular claim, when it is built in a certain way, in, you know, let's say in Texas, the same Medicaid in Indiana may approach it differently and say, for example, in Indiana, the BCBA and the RBT both need to be credentialed, whereas that's not the case in, let's say, another state.
00:04:50.240 --> 00:05:02.879
So those are some nuances that drive this, but there could be many more issues that drive both the adjudication process and, for example, the denials that come out of it, and then the resolution of those denials.
00:05:03.040 --> 00:05:17.279
Um, for example, in Plutus, we built something called Pi, which is a payer intelligence engine, which has aggregated everything that we know about every pair that we have built for over the last several years, running into billions of dollars of claims.
00:05:17.439 --> 00:05:30.800
And we have synthesized that into a model that we can refer to from time to time as we look at a claim or look at a credentialing application or a prior author application or an appeals package we send out.
00:05:30.959 --> 00:05:40.160
And keep in mind that Pi is pair intelligence engine as its uh acronymous based on today's information that may become outdated tomorrow.
00:05:40.319 --> 00:05:42.399
So we have to constantly refresh that.
00:05:42.639 --> 00:05:47.759
So a company with our experience and resource itself finds it challenging.
00:05:47.839 --> 00:06:08.240
So I can understand why uh a small uh ABA practice with one BCBA providing care and then also billing uh on the weekends may not be able to comprehend all the nuances at every payer and every state level can be I want to dive into AI a little bit later, but before we get there, I'm curious.
00:06:08.480 --> 00:06:16.000
I heard a quote the other day that said something along the lines of every tech and AI company is actually a data company in disguise.
00:06:16.240 --> 00:06:22.560
And that came to mind when you're sharing your billions of claims over seven years.
00:06:22.800 --> 00:06:33.759
So I'm curious as you're training this model, as you're sharing this data with your staff, were there any surprises that came from analyzing the trend of that much data over that many years?
00:06:34.079 --> 00:06:43.839
The surprise that I would allude to is that a lot of the analysis shows a lack of consistency and adjudication across time, right?
00:06:43.920 --> 00:06:51.920
A same pair may adjudicate a claim in a certain way at a certain point, and then you will see that it's not always the same pattern.
00:06:52.079 --> 00:07:03.680
So it's it would have been nice if we could have fallen into that pattern and analyzed that pattern, but the pattern is that there is sometimes no pattern to adjudication or decision making on the payer side.
00:07:03.839 --> 00:07:05.519
So that's one thing I can talk about.
00:07:05.600 --> 00:07:12.720
The other is again the inconsistencies of policies and procedures across payers, across states.
00:07:12.800 --> 00:07:14.800
Uh that's something interesting that we found.
00:07:14.959 --> 00:07:33.519
Also, I think one of the things that we see is that while the claim form itself has a finite number of variables to it that you can optimize towards, the reality also is what happens upstream is where I think all the gotchas on the pain points of ABA is.
00:07:33.759 --> 00:07:39.120
Example, credentialing, how many of your clinicians are not credentialed at the time of care?
00:07:39.279 --> 00:07:58.399
Or, you know, how many times a provider ends up providing care without an underlying authorization, you know, how many times the documentation doesn't meet the payer guidelines that is found out after the fact, and then the payers come back and audit it and then start acclawing back money.
00:07:58.639 --> 00:08:02.959
So those are a lot more challenging than the claim form itself.
00:08:03.040 --> 00:08:13.600
Like a lot of times we think, okay, as long as I can keep my claim form clean, whether the right modifiers and the right codes and all of that, things will work well.
00:08:13.759 --> 00:08:29.040
And what surprised me is that a lot of the problems in ABA, or in fact in medical billing in general, are issues that happen upstream or earlier in the process that come back and surface as denials or as no cash in the bank.
00:08:29.439 --> 00:08:35.200
Which brings us to some of the conversations and the way you look at broken revenue or broken RCM.
00:08:35.440 --> 00:08:41.360
So you've said in the past that broken revenue isn't really a billing problem, it's more of a care problem.
00:08:41.679 --> 00:08:43.200
Can you expand on that, though?
00:08:43.679 --> 00:08:48.639
Revenue and care are two sides of the same coin or different sides of the same coin, right?
00:08:48.879 --> 00:08:53.840
So what I mean by that is, for example, when we look at ABA as a service, right?
00:08:53.919 --> 00:08:59.440
I mean, it is ultimately clinical capacity financed by cash conversion, right?
00:08:59.600 --> 00:09:19.600
So the your ability to deliver care is a function of your ability to convert your claims into dollars, because ultimately you are a services business, and every dollar that you generate is going back to pay for a future care process or a care uh or a service uh you know structure.
00:09:19.840 --> 00:09:39.039
So what I mean by that is, for example, if a clinician, a BCBA that owns a practice is unable to convert their previous months' claims into cash, there is going to be a significant impact on their ability to pay for employees like BCBAs and RBTs, or even just grow their business by hiring more people.
00:09:39.200 --> 00:09:45.759
So to that, so that's the primary driver of uh you know revenue and care going hand in hand.
00:09:45.840 --> 00:09:51.919
Or if one is not getting logically converted to cash, the other is going to be sooner than later impact.
00:09:52.080 --> 00:10:09.120
There might be a lag between the time you look at, okay, I'm not able to convert every one of my claims into dollars, and my future growth plan is contingent upon having that cash in the bank in order to finance my continued operation and also my growth.
00:10:09.279 --> 00:10:11.679
The second mechanism is a trust, right?
00:10:11.759 --> 00:10:13.679
It behaves like retention economics.
00:10:13.840 --> 00:10:17.679
So to a family, the bill is part of that clinical product.
00:10:18.080 --> 00:10:24.399
An error-driven balance in a statement to a family is going to drive churn, right?
00:10:24.480 --> 00:10:27.759
I mean, your parent is like, okay, I mean, why am I getting this wrong bill?
00:10:27.919 --> 00:10:32.799
I'm gonna, you know, be upset about it, I might move away, I might go to a different clinic.
00:10:32.960 --> 00:10:42.720
So your ability to then deliver that care with that family is a function of accurate billing, accurate payment uh collections from that from that family.
00:10:42.879 --> 00:10:45.120
So it's the revenue is a care problem.
00:10:45.360 --> 00:10:55.200
It's not just being philosophical, basically naming the two places, capacity and retention, where the back office sets the ceiling on clinical outcomes every month, right?
00:10:55.279 --> 00:11:07.840
Because if your back office is not doing a good job of converting that that care that you've already provided into cash, the ability for you to do that effectively in the future is hurt significantly.
00:11:08.639 --> 00:11:20.000
So what are some of the warning signs that owners should look out for operators when it comes to their back office not converting efficiently enough and care eventually being impacted?
00:11:20.159 --> 00:11:21.440
What's the canary in the coal mine?
00:11:21.840 --> 00:11:25.200
Well, I would I would say that very easy KPIs out there, right?
00:11:25.279 --> 00:11:28.720
I mean, the most basic one is denial percentage, right?
00:11:28.879 --> 00:11:33.200
So if you're sending out 100 claims, what how many of those claims are getting denied?
00:11:33.279 --> 00:11:34.879
So what is your denial percentage?
00:11:35.039 --> 00:11:50.080
That's the most easiest number to track and use as the uh early warning signal for you to know that if your denial percentages are higher and higher month over month, something ain't right in that in that business model, right?
00:11:50.240 --> 00:12:02.960
We can get into why that's happening in a minute, but since the fact if you're sending out 100 claims a month and you're only getting paid, let's say 85 of them and the other 15 is getting denied, that's an early warning sign for you.
00:12:03.200 --> 00:12:05.440
The other area is days in AR.
00:12:05.519 --> 00:12:10.080
Like what is your turnaround of claims to dollars, right?
00:12:10.240 --> 00:12:12.720
If your days in AR is less than 30 days, great.
00:12:12.879 --> 00:12:17.919
If it is getting to 40, 50 days, 60 days, that means something is wrong.
00:12:18.240 --> 00:12:21.600
And you need to be looking into why those claims are not getting paid.
00:12:21.759 --> 00:12:30.399
Now, why I kind of differentiate between a denial percentage and a days in AR is that sometimes payers don't deny, they just delay.
00:12:30.559 --> 00:12:39.039
Now, whether that delay is because of something that you did or something that the payer is doing, that's another nuance to get into in a minute.
00:12:39.200 --> 00:12:47.039
But the reality is that if you are seeing outstanding claims aging past a certain point, you need to be worried, right?
00:12:47.200 --> 00:12:59.519
Then, of course, the most important one is net collection ratio, which is after all is said and done, let's say three months into this process, what percentage of your claims did you finally collect?
00:12:59.600 --> 00:13:02.639
And what percentage of claims did you finally have to write off?
00:13:02.799 --> 00:13:08.399
If that is like 99% collected, 1% written off, I think you're in a healthy practice.
00:13:08.559 --> 00:13:15.120
If that is 90% collected and 10% written off, you're probably skating on thin ice.
00:13:15.279 --> 00:13:20.799
Because now you're going to be concerned at some point that PBA is not a high margin business.
00:13:20.960 --> 00:13:22.799
It's a pretty tight margin business.
00:13:22.879 --> 00:13:31.519
So if 10% that already tight margin business is being written off, then you're probably losing money and you don't know about it yet.
00:13:31.679 --> 00:13:38.799
Because keep in mind, you know, PL is not the right metric to define the health of your practice.
00:13:38.960 --> 00:13:40.720
It is a net collection ratio.
00:13:40.960 --> 00:13:48.799
Because you could book revenue by seeing a kid, and after you sent out a claim, you're saying, oh, I generated this revenue.
00:13:48.960 --> 00:13:52.799
But that revenue converts to cash based on your net collection ratio.
00:13:52.960 --> 00:14:00.559
Because you can easily go around showing off a PL that shows that you generated a million dollars in top line.
00:14:00.639 --> 00:14:04.720
But how much of it was realized in cash is the real question, right?
00:14:04.879 --> 00:14:22.399
And many a time, as even a seasoned CFO, and I've seen a lot of private equity back seasoned CFOs not realizing that yes, you can book revenue by seeing providing care, seeing a kid, sending out a claim, but did that claim convert to cash?
00:14:22.480 --> 00:14:27.519
And did it convert to cash in a reasonable period of time is what drives this.
00:14:27.679 --> 00:14:42.720
So these are some of the red flags that I always encourage both individual practice owners and even larger companies with seasoned CFOs to look into because sometimes those PL numbers might be misleading.
00:14:43.039 --> 00:14:46.320
So you talked about, and I'll just run through some of the KPIs.
00:14:46.399 --> 00:14:55.279
So correct me if I miss them, but you have a denial percent of about 5%, you have AR of 30 or less, you have net collection 99%.
00:14:55.840 --> 00:15:00.639
Let's say you come to somebody or they join in Plutus, and those numbers are way off.
00:15:00.720 --> 00:15:02.879
They're nowhere near kind of the baseline.
00:15:03.039 --> 00:15:08.159
What does the process look like of getting them back into the correct location?
00:15:08.639 --> 00:15:11.840
We start with what is called as an revenue cycle management assessment.
00:15:12.080 --> 00:15:21.679
And we generally do this free for almost all of our clients, our prospective clients, or even anyone who kind of gets a sense of where they are from a baseline perspective.
00:15:21.840 --> 00:15:28.320
And in that baseline, we are trying to identify the number one denial rate is the first indicator.
00:15:28.480 --> 00:15:29.919
But then that doesn't stop there.
00:15:30.000 --> 00:15:34.080
It goes to what is constituting your denials, right?
00:15:34.240 --> 00:15:37.759
What is the upstream issue that drove that denial?
00:15:37.919 --> 00:15:44.639
And we talked about it in briefly at the beginning of this conversation: credentialing, prior authorization, benefits verification.
00:15:44.879 --> 00:15:55.200
I can bet that 90% of all practices, when they have denials, they are related to almost always one of these three issues or all the three issues.
00:15:55.840 --> 00:16:02.159
So our goal is to identify those root causes because denials don't happen in the back office.
00:16:02.320 --> 00:16:07.200
Denials happen in the front office during the delivery of care.
00:16:07.440 --> 00:16:10.639
And those are sort of visible in the back office.
00:16:10.799 --> 00:16:21.200
So most CEOs, most BCBA owners, or even most CFOs are hyper-focused on saying my billing team is the problem, my billing is a problem.
00:16:21.440 --> 00:16:23.600
Really, it's not the billing that's a problem.
00:16:23.759 --> 00:16:28.639
Billing is only showing you what is the problem somewhere else in the business.
00:16:28.960 --> 00:16:35.519
Example, you may be hiring BCBAs, but you know, lagging the credentialing associated with it.
00:16:35.679 --> 00:16:45.759
You may be scheduling kiddos with the wrong BCBA, as in the BCBA that does not have the right credentialing for the coverage that the kiddo has.
00:16:45.919 --> 00:17:01.840
You may be getting the prior authorization, but then the parents quits their job or switches their job and has a new insurance, and you are still assuming that the previous prior authorization that you got is valid and you're providing care on a wrong assumption.
00:17:02.000 --> 00:17:16.400
You might be providing care, but the authorization expired, and somebody who got the authorization did not inform the clinician, and now the clinician is making a wrong assumption that this kid still has units to provide service.
00:17:16.480 --> 00:17:18.880
And then you may have done it in good faith.
00:17:18.960 --> 00:17:25.039
I mean, saying, okay, you know what, the kid needs that service, I'm going to do it, but underlying choir authorization was never there.
00:17:25.200 --> 00:17:32.640
So these are some of the things that we look at, and then we try to work with the practice to tighten those front-end processes.
00:17:32.720 --> 00:17:39.680
And remember when I started the conversation, I said process comes first, and then comes technology and people in order to fix this.
00:17:39.759 --> 00:17:44.559
So that process is what we are fixated on solving for.
00:17:44.720 --> 00:17:49.680
And each and every process we solve in that front-end is more dollars in the back end.
00:17:50.160 --> 00:17:58.400
What does the general timeline look like from solving these front-end processes to seeing the effect on the back-end KPIs?
00:17:58.640 --> 00:18:01.039
Is it a month, two months, varies?
00:18:01.839 --> 00:18:07.680
To fully run through the process of we will attack this problem from a multiple dimensions, right?
00:18:07.839 --> 00:18:10.640
One is we solve for what has to be prevented.
00:18:10.799 --> 00:18:16.079
One is second is we solve for what we need to mitigate, you know, you know, uh go back and fix.
00:18:16.559 --> 00:18:20.480
For example, there's a bunch of denied claims that needs to go back and get fixed.
00:18:20.559 --> 00:18:27.440
So we may have to try out and see if we can get some retros on the credentialing and prior loss, which is very difficult most of the time.
00:18:27.599 --> 00:18:31.920
Most payers do not even agree to go back and retroauthorize something.
00:18:32.000 --> 00:18:34.799
Uh and if they do, it's limited periods.
00:18:34.960 --> 00:18:39.759
If they are doing credentialing and it's delayed, uh, they may retro for some date.
00:18:39.920 --> 00:18:42.000
So these are things that we can go back and fix some.
00:18:42.319 --> 00:18:48.480
So the cash coming from already denied claims being fixed generally takes 60 to 90 days.
00:18:48.799 --> 00:18:55.359
And then the preventions, they can start solving, giving you uh cash in the bank a little faster, maybe within 30 days.
00:18:55.599 --> 00:18:59.359
So I would say when you're solving for this, you're solving for it at two levels.
00:18:59.519 --> 00:19:00.960
One is what can you prevent?
00:19:01.119 --> 00:19:08.640
And when you prevent something, you can see the impact within 30 days, because 30 days is normally the period of time it takes to adjudicate a claim.
00:19:08.799 --> 00:19:24.720
Uh, but if you're trying to go back and fix something that was already broken, uh, it may take 60, 90 days because payers are known to fix, pay you on fixed claims or appeal claims in a longer period of time because they have to manually review everything.
00:19:24.799 --> 00:19:26.559
So it's probably 90 days.
00:19:26.720 --> 00:19:37.279
So overall, I think you'll start seeing an improvement in cash flow, maybe 90 to 120 days out from the time a PLUDIS health team comes into the equation.
00:19:37.599 --> 00:19:44.400
Now, what does the owner or maybe the team of the operations team experience during those 90 to 120 days?
00:19:44.559 --> 00:19:47.039
Are they working with a representative at Pluto's?
00:19:47.200 --> 00:19:50.000
Do they have somebody helping them with those processes?
00:19:50.240 --> 00:19:55.039
What's the support like as they go through the started a new RCM process?
00:19:55.279 --> 00:19:57.440
Also, they said that we have to fix a bunch of things.
00:19:57.599 --> 00:19:58.799
What does that look like?
00:19:59.119 --> 00:20:04.240
When we come in, we look at what is it that was a problem in the back office?
00:20:04.319 --> 00:20:09.119
Maybe an existing billing team missed a lot of obvious things.
00:20:09.200 --> 00:20:17.599
And those are very low-hanging fruits, like not putting the right modifiers or not using the right, you know, payer information, things like that.
00:20:17.680 --> 00:20:31.039
Like you sent the claim to the wrong pair, or you had the wrong modifiers that are already retired by the insurance company, or maybe not applicable to a particular state, but you nevertheless use something that you would have used in one state in a different state.
00:20:31.200 --> 00:20:33.039
Things like that are easy, low-hanging fruit.
00:20:33.200 --> 00:20:40.319
So that we don't even go to the provider, we just do it ourselves, we fix it, they start seeing the benefits of that in cash in the bank.
00:20:40.480 --> 00:20:54.480
Then comes the more harder and difficult part of this process, which is identifying all those process gaps and then communicating that to the practice owner, to the CEO, or whoever is the key leaders within the organization.
00:20:54.640 --> 00:21:08.720
This is where our leader, someone with stature and gravitas, comes in and has real conversations with the leadership on the practice side and say, listen, these are some challenges we see that we need to come together to fix it.
00:21:08.960 --> 00:21:14.400
But I always tell my team and the client that your problems are my problems.
00:21:14.559 --> 00:21:28.160
My problems are still my problems as a billing company, but I will also own and co-own your problems, which means, for example, if your scheduler is not looking at prior auth units before scheduling a kid.
00:21:28.400 --> 00:21:40.240
If your scheduler is not looking at a credentialing roster to look at the BCBAs who are credentialed for a particular kiddos pair, now these are sometimes also require some homework to be done.
00:21:40.400 --> 00:21:47.119
Do you have a good tracker of all your BCBAs and their associated payer credentialing statuses?
00:21:47.519 --> 00:21:51.359
Like many, many practices don't even keep track of it in one single place.
00:21:51.519 --> 00:21:59.759
So that when a scheduler is scheduling, do they even know that Matt is not credentialed with a blue cross or blue shield in Texas?
00:22:00.160 --> 00:22:06.880
Or MAD is actually credentialed only with uh United or Aitna and not with some of the other pairs.
00:22:07.039 --> 00:22:21.759
Because those are nuances that A need to be tracked, and B, at the time of scheduling, you need to coach the scheduler to look out for those gachas where you end up credentialing or scheduling a kiddo with the wrong BCBA who does not have the credentialing and stuff like that.
00:22:21.920 --> 00:22:31.599
So that is where we will spend time in the front office creating those best practices, but then also guiding people, for example, recruiting.
00:22:31.759 --> 00:22:48.000
Like put a process in place where when you hire a clinician, immediately, as soon as a hiring offer letter goes, you notify credentialing team to start the process of collecting the documents that are required to credential with the payers that you're gonna be seeing.
00:22:48.079 --> 00:22:50.480
Now, do you also know who are your top payers?
00:22:50.640 --> 00:22:55.519
Then you know how long it takes to credential because you need to back up into all of that.
00:22:55.680 --> 00:23:10.240
Like if you are hiring a clinician and you know it takes 90 days to credential that clinician with these pairs, what are you gonna do about that for the next 90 days when you have a wait list, when you have kiddos, how are you gonna match all of these things?
00:23:10.400 --> 00:23:21.599
So these are all detailed operational discussions to be had with both leadership and with department and functional leaders who will then have to come back and tell us how to do this.
00:23:21.920 --> 00:23:22.720
Very interesting.
00:23:22.880 --> 00:23:26.079
And that's where the processes and the operations come in.
00:23:26.240 --> 00:23:35.200
And then once you get through the processes, now we get into what I think is the fun part, which is the tech, the AI, the pie, and all of that stuff.
00:23:35.359 --> 00:23:45.759
So can you explain your general thought process and kind of where you've directed Plutus to leverage AI, whether that be in the process of adjudication, submission?
00:23:46.319 --> 00:23:49.359
Where in general does AI touch your RCM process?
00:23:49.759 --> 00:23:52.960
We have really leaned into all of those elements.
00:23:53.039 --> 00:24:00.559
And I've already talked about the, you know, there are some processes like payment posting, where it's a much more deterministic approach to automation.
00:24:00.640 --> 00:24:07.920
So we can use automation to solve for posting payments, verifying against a bank statement, and so on and so forth.
00:24:08.000 --> 00:24:10.400
So those are easy, uh low-hanging fruits.
00:24:10.559 --> 00:24:14.240
We use APIs to check eligibility on ACE every two weeks.
00:24:14.400 --> 00:24:17.119
We use APIs to do insurance discovery.
00:24:17.279 --> 00:24:22.640
If we find out that a benefit of a parent is lapsed, does that parent have any other eligibility?
00:24:22.720 --> 00:24:24.880
These are all things that AI is not required.
00:24:24.960 --> 00:24:28.240
You can use pretty much automation and APIs to do it.
00:24:28.480 --> 00:24:49.279
Where we have used AI is, for example, today we have the ability to build a credentialing application completely from scratch using documents that VCBAs provide and using AI to validate timelines and document checklists and formats in which applications need to be provided.
00:24:49.440 --> 00:24:53.119
All of those can be done using uh AI today.
00:24:53.200 --> 00:24:55.279
So that's something we have put in place.
00:24:55.519 --> 00:24:57.599
Same approach with prior authorization.
00:24:57.759 --> 00:25:13.359
We have AI today that can take a treatment plan and all the underlying documents that payers ask for and create a prior auth package that can be sent to the payer's sort of portal in order to get prior auth approval done.
00:25:13.599 --> 00:25:16.000
We have denial analysis AI.
00:25:16.079 --> 00:25:40.000
So as soon as a denial comes from a payer, we can run it through our AI, and the AI would analyze the denial correspondence from the payer, the claim documentation, the underlying medical documentation, the underlying benefits documentation, underlying credentialing documentation, and come back and give you a plan of action to overturn that denial with the payer.
00:25:40.400 --> 00:25:49.680
We have been using able to use AI to increase our denial overturn, which was pre-AI at about 50-60% to about 70 to 80%.
00:25:50.000 --> 00:25:55.920
So you've seen an uplift of around 20% just by using AI to analyze the denial.
00:25:56.079 --> 00:26:13.039
We also are currently using AI to create appeals packages where we are able to look at all this underlying documentation that I mentioned to create a comprehensive appeal, again, referring to our Pi and payer policies, historical payer adjudication trends.
00:26:13.200 --> 00:26:25.519
We are able to create this comprehensive appeal that can go back to the payer with all the nuances that a human may have missed or does not have the time to frankly go and build an appeals package.
00:26:25.680 --> 00:26:29.119
Because a typical human appeals package takes several hours.
00:26:29.279 --> 00:26:35.440
I can shorten that time to maybe a few minutes, and then you can have a human review it, and then we can send that.
00:26:35.599 --> 00:26:37.599
So these are some places where we have done it.
00:26:37.759 --> 00:26:43.440
We also have the most important piece, I leave it to the last, which is the documentation review.
00:26:43.519 --> 00:27:05.200
So we are now using AI that can not only review the claim but also review the underlying documentation and scrub it ahead of time so that we are fully compliant with payer policies from a documentation perspective, because nine out of ten times you may have had the right intent and provided the right service to the kiddo, but did you document it well?
00:27:05.440 --> 00:27:10.000
Did you document it as per the payer's guidelines and policies is a million-dollar question.
00:27:10.160 --> 00:27:20.720
And many a time most clinicians trip on that, and we try to stay ahead of that by you know auditing those, just as I mentioned in the past, the payers are going to do the same thing.
00:27:20.799 --> 00:27:54.160
So we are preempting that and saying even before a claim gets created and sent to the payer, we will validate that that underlying documentation is kosher and that your claim and the documentation match so that if and when payer comes back in three months or six months and asks you for that documentation, you're not caught napping, or you know, that employee has left the building and you're like you know figuring out what do I do now because the payer is asking for documentation that that employee did not complete you know properly, but I bid for it and I got paid.
00:27:54.240 --> 00:27:57.359
Now I have to re return that money, and that's a revenue problem.
00:27:57.519 --> 00:27:59.440
So those are some of the use cases matter.
00:27:59.839 --> 00:28:02.240
There's so much of that that I want to dive into.
00:28:02.559 --> 00:28:03.839
Let's start here.
00:28:04.240 --> 00:28:13.440
One of the common things that I hear when I talk to your clinicians, your ABA owners about AI, is they say, Yeah, I used ChatGPT the other day.
00:28:13.680 --> 00:28:17.680
It told me to put ketchup on my pizza instead of tomato sauce.
00:28:17.839 --> 00:28:20.640
Like, why would I trust AI for this type of thing?
00:28:20.799 --> 00:28:26.799
And you know, in marketing, there's some obviously sometimes we use AI and we make it more efficient.
00:28:26.880 --> 00:28:30.319
But whenever you're thinking RCM, you're talking millions of dollars.
00:28:30.559 --> 00:28:40.160
So can you talk a little bit about the quality control, the training process that you use to make sure your AI is responsible with the millions that it's processing?
00:28:40.480 --> 00:28:42.480
Very great, very good question, I would say.
00:28:42.640 --> 00:28:49.680
So let me say that you just throw something into and it magically gives you something in return that is perfect.
00:28:49.920 --> 00:28:56.559
So what we have done essentially is we look at the LLMs as just computing capacity, right?
00:28:56.720 --> 00:29:10.240
It is the ability to process information in large scale, but it's up to us as the people who are running this to control the outcomes with good quality uh learning and guidance.
00:29:10.480 --> 00:29:19.519
So our RAG model, our Pi, these are sort of guardrails that clearly define the way AI is being utilized.
00:29:19.599 --> 00:29:26.000
So think of AI like a young kid out of college that you are hiring as an intern.
00:29:26.319 --> 00:29:34.480
And then now you need to train that intern on everything that you know in your business, and then you need to guardrail what that intern is doing.
00:29:34.640 --> 00:29:42.559
You know, he's got he's very smart, he's got you know perfect logical and analytical thinking skills, but he doesn't know anything about your business.
00:29:42.720 --> 00:29:59.519
Now you are going to use him, but you're going to use him in a manner where you have those guardrails of training in the SOPs that you built over the last 20, 30 years, your ability to comprehend the nuances of pair differentiation.
00:29:59.680 --> 00:30:04.160
Like these are things that Plutis has built over a 30-year period, right?
00:30:04.319 --> 00:30:30.319
The data that we have, the guidelines that we have, the SOPs that we have, these are things that are sitting around this AI computing power, and we are just merely taking that massive computing capability of the AI to look at 10,000 lines of document or pages of documents, hundreds of thousands of claims, which a human, let's be honest, cannot look at the same capacity like an AI can.
00:30:30.480 --> 00:30:31.680
So that's the difference.
00:30:31.839 --> 00:30:32.640
So you're right.
00:30:32.720 --> 00:30:38.319
I mean, you can't just throw a claim into AI and expect it to magically tell you what to do.
00:30:38.640 --> 00:30:45.039
You can write a small prompt, but it's probably going to hallucinate with what you can do with the way we have built AI.
00:30:45.279 --> 00:30:55.519
We are giving it the power of these additional capacity of uh SOPs, you know, guidelines, process documents, reference manuals, APIs.
00:30:55.680 --> 00:31:01.680
That's what makes AI powerful, not just AI's ability to throw something into Chat GPT and get an answer.
00:31:02.000 --> 00:31:15.759
You know, you don't just get highly accurate AI without more data than you think is possible, which is why Pluto's health a player for a long time can perform at a higher rate of accuracy in their AI than someone who just got started.
00:31:16.079 --> 00:31:19.359
Which is actually a question I have next.
00:31:19.680 --> 00:31:32.480
We see a lot of new companies, whether they be RCM companies or AI-powered companies, pop out of the blue now with a website and they're pitching ABA companies.
00:31:32.720 --> 00:31:41.119
In your opinion, what are some of the things that an ABA owner should look for to discern this company is one I can trust?
00:31:41.440 --> 00:31:44.880
Let me preface that by saying everyone has a role to play in this industry.
00:31:45.039 --> 00:31:46.079
This industry is massive.
00:31:46.240 --> 00:31:49.680
There is, you know, you can solve the problem from many different directions, right?
00:31:49.839 --> 00:31:53.920
There are companies coming into this, uh solve problems from a technology perspective.
00:31:54.000 --> 00:31:59.200
There are companies coming in from people uh capacity perspectives.
00:31:59.599 --> 00:32:01.279
And I think there's a place for everyone.
00:32:01.519 --> 00:32:06.559
But the real question is who can solve for this in a more comprehensive manner?
00:32:06.799 --> 00:32:13.839
So PLUTUS essentially characterize Plutus as five pillars of how we look at this business, right?
00:32:14.319 --> 00:32:38.079
The first pillar is that Pluto's almost always comes into a relationship with an ABA provider with an end-to-end offering, which means we like to come in and help our ABA clients from credentialing, prior authorization, benefits verification, claim scrubbing, payment posting, denial, AR follow-up, and uh patient payments.
00:32:38.319 --> 00:32:42.559
Now that kind of an end-to-end process is the first pillar of our business.
00:32:42.880 --> 00:32:46.079
We're not doing only one or two or three processes, we do everything.
00:32:46.240 --> 00:32:54.400
The second element is we almost always get paid on an outcome basis, which means we don't get paid if you don't get paid.
00:32:54.640 --> 00:33:07.119
So we very rarely, if ever, only in in absolutely rare cases do we ever say we will do uh you know something on a different pricing model, like uh, you know, FTE or transaction.
00:33:07.200 --> 00:33:11.200
We are always saying you get paid, we get a small piece of what you get paid.
00:33:11.279 --> 00:33:15.599
And we are accountable for the KPIs that as SL is in that contract.
00:33:15.759 --> 00:33:17.680
The third element is we are tech forward.
00:33:17.759 --> 00:33:27.279
So we have invested in technology and not just uh AI technology, AI automation, analytics are the three A's of our technology.
00:33:27.599 --> 00:33:33.440
The fourth element of it is that we look at this from the lens of speciality expertise.
00:33:33.599 --> 00:33:37.200
We have been in ABA and behavioral health for the last several years.
00:33:37.359 --> 00:33:39.359
That's a majority of our revenue.
00:33:39.599 --> 00:33:44.240
So we really understand the nuances of uh behavioral health and ABA.
00:33:44.400 --> 00:33:47.519
And last is we are also a very people-driven business.
00:33:47.680 --> 00:33:58.079
So we have a fairly large footprint of people, and we always have believed the human in the loop element to it, because the last mile of any process will need a human.
00:33:58.240 --> 00:34:02.000
And a lot of the technology companies may come and say, I'll give you this technology.
00:34:02.240 --> 00:34:06.000
What happens outside of that technology's capacity is up to you.
00:34:06.240 --> 00:34:08.719
We're not going to be able to be held accountable.
00:34:08.880 --> 00:34:11.280
And so that's the last element of our differentiator.
00:34:11.360 --> 00:34:13.679
So those are the five elements of how I look at it.
00:34:13.920 --> 00:34:23.519
So now when you go back to the companies that are only offering services as RCM companies, they're missing some of these pieces, like the ability to use technology.
00:34:23.599 --> 00:34:25.840
Because technology has a play in this business.
00:34:26.000 --> 00:34:27.199
We can't negate it.
00:34:27.360 --> 00:34:31.199
We need to embrace it and be willing to use it in every aspect of it.
00:34:31.440 --> 00:34:36.639
Having said that, we also can't say this is only tech, there is no people, there is no outcome.
00:34:36.800 --> 00:34:49.760
So a lot of the technology companies spinning out of Silicon Valley, I have noticed they're coming back and saying, Oh, I'll charge you 10 bucks for authorization, I'll charge you$3 for benefits, I'll charge you$20 for something else.
00:34:50.000 --> 00:34:53.760
The problem with that approach is those dollars add up pretty quickly.
00:34:53.840 --> 00:34:55.840
And are they correlating to outcomes?
00:34:56.000 --> 00:35:00.400
Most of these technology companies will say, that's not my problem, that's a you problem.
00:35:00.559 --> 00:35:18.559
But when we come into the picture, we are saying, we'll invest and provide you all that technology, but with towards the end of delivering dollars in your bank at the KPIs that I mentioned about, not that, okay, I'm just selling you some fancy, shiny next thing, and then it's up to you how we utilize it.
00:35:18.719 --> 00:35:24.079
Now, your technology may be perfect, but it requires sometimes some stewardship of people.
00:35:24.159 --> 00:35:34.880
And if you're not providing and you expect the mid-sized ABA practice with a BCBA owner to go figure it out all, then you may not be delivering the outcomes that they are expecting.
00:35:35.039 --> 00:35:39.039
So I always point my practices towards outcomes.
00:35:39.280 --> 00:35:45.920
No matter whatever you're doing, your bottom line is outcome because that's what pays the bills, and that outcome is cash in the bank.
00:35:46.079 --> 00:35:53.280
And if you can get 99% of that cash in the bank, every technology, every person, every expertise is worth every penny.
00:35:53.360 --> 00:35:54.880
Otherwise, it's worth nothing.
00:35:55.199 --> 00:36:00.079
The focus on accountability, I think, cuts through a lot of the marketing hype.
00:36:00.239 --> 00:36:02.559
As someone in marketing, I can attest to that.
00:36:02.800 --> 00:36:07.039
Let's talk about the worst letter people receive in the mail, the audit.
00:36:07.280 --> 00:36:15.840
Can you talk a little bit about the audit process and how Pluto can help with that awful time that AVA companies go through?
00:36:16.079 --> 00:36:29.039
Yeah, I think a lot of times people look at audit as some kind of an integrity check and that, but I think a lot of times payers are really trying to narrow what they're looking for, right?
00:36:29.360 --> 00:36:33.840
Payers, you know, in this context of audit are not really concerned about your intent.
00:36:34.159 --> 00:36:40.320
They're really concerned about are you uh you know documenting uh what you did correctly.
00:36:40.480 --> 00:36:49.519
So the point I think we need to uh inform uh owners who receive these audits is that are you substantiating that claim correctly, right?
00:36:49.599 --> 00:36:50.559
That's the bottom line.
00:36:50.719 --> 00:37:02.079
It doesn't matter that you know the kid needed the service, you actually provided the service, this was all uh just uh the payer just taxing you with with a lot of and burdening you with a lot of paperwork.
00:37:02.320 --> 00:37:10.239
I mean, the reality is the payer wants to pay you based on documented evidence, and all they care about is that evidence, right?
00:37:10.320 --> 00:37:13.199
You you know, they paid you, now they want to see an underlying evidence.
00:37:13.440 --> 00:37:22.000
Evidence has to be aligned to the guidelines that they have already provided you as part of their contract or as part of their general policies and procedures.
00:37:22.239 --> 00:37:26.639
So what gets tested is the tie between the note and the that's the bottom line, right?
00:37:26.800 --> 00:37:29.199
Does the session note support the units built?
00:37:29.360 --> 00:37:34.159
Is the provider on the claim who delivered is the one who also delivered the service, right?
00:37:34.480 --> 00:37:40.079
Is the required DCBA supervision RBT documented at the ratio that the payer mandates?
00:37:40.400 --> 00:37:41.679
Little things like that, right?
00:37:41.760 --> 00:37:45.199
I mean, are you doing those things is what the payer is looking for.
00:37:45.440 --> 00:37:59.119
So, you know, my recommendation to most owners is you've got to be living in the reality of doing this every day and not be sort of flippant about it and say, I'll deal with it when I get an audit.
00:37:59.360 --> 00:38:06.239
This has to be ingrained into the DNA of your business that no claim goes out without appropriate documentation.
00:38:06.400 --> 00:38:08.400
Now, that's a culture issue, right?
00:38:08.639 --> 00:38:15.039
A lot of practices that I've seen that have been very successful drive discipline from day one.
00:38:15.199 --> 00:38:29.840
They will be uncompromising with their providers, with their RBTs, with everybody in the organization, that the DNA of the organization is built around proper documentation, and that documentation cannot be compromised with.
00:38:30.079 --> 00:38:37.360
Doesn't matter that you provide great care to the kiddo, doesn't matter that you're a hardworking provider or clinician.
00:38:37.519 --> 00:38:40.159
What matters is did you document what you did?
00:38:40.320 --> 00:38:47.679
And if you didn't, you're risking the practice, you're risking risking revenue, you're risking integrity and compliance and all of that.
00:38:47.840 --> 00:38:51.199
So I tell the providers that the best solution is to prevent it.
00:38:51.360 --> 00:38:58.400
But now, if you still end up in an audit, yes, we need to come down and sit with you and start looking at every document.
00:38:58.639 --> 00:39:06.079
And document by document, you need to improve the quality of your deliverables to the payer.
00:39:06.239 --> 00:39:10.880
And once you reach a certain threshold of quality, the payer is going to take you out of the audit.
00:39:10.960 --> 00:39:14.079
Until then, they're going to be like after you on your case.
00:39:14.559 --> 00:39:33.760
And frankly speaking, fatal for practices that don't have the deep pockets of cash that they can fund while the payer is going to review every single claim, every single document of yours, and come back and pay you after 60, 90 days, where originally you would have been paid in 25 or 20 days.
00:39:33.920 --> 00:39:36.559
Now that becomes a 90-day cash flow problem.
00:39:36.719 --> 00:39:40.320
And it also creates all the stress with your clinicians.
00:39:40.480 --> 00:39:46.079
So the best solution is the day you start, be enforceable processes in place.
00:39:46.239 --> 00:39:50.719
Put enforceable processes in place with your clinicians and get the documentation right.
00:39:50.880 --> 00:39:59.280
But if you're already in an audit and you don't have documentation, you need to grid up and start putting the discipline in correcting everything.
00:39:59.440 --> 00:40:01.039
Eventually it'll come out of the audit.
00:40:01.119 --> 00:40:03.840
It may take 90 days, it may sometimes take six months.
00:40:04.159 --> 00:40:10.159
It all depends on how disciplined you are with uh how you're correcting the documentation and sending it off to repairs.
00:40:10.480 --> 00:40:14.719
We're almost at time, but I have one more question that I want to ask you.
00:40:14.960 --> 00:40:22.639
There's a lot of practice owners who are doing billing internally, billing externally, small teams, big teams with Plutus, not with Plutus.
00:40:22.800 --> 00:40:34.559
I'm curious what are the questions that they should be asking their billing team, internal, external, etc., to cut through everything and just say this is working or this isn't working?
00:40:34.880 --> 00:40:38.000
First thing I would say is start with the denials.
00:40:38.320 --> 00:40:52.079
If you're a practice owner, if you're a CEO or a CFO, I would encourage you to look at your denial report every month, if not every week, depending on how closely you want to monitor and manage this.
00:40:52.239 --> 00:40:59.360
I would, if I were them, I would probably monitor it every week because sometimes a month is too late in order to make corrective action.
00:40:59.440 --> 00:41:03.119
And the reason I say that is because again, I go back to my original point.
00:41:03.280 --> 00:41:10.400
That denial data has a wealth of information that you can use to transform your front-end operations.
00:41:10.639 --> 00:41:18.159
If you see that a particular clinician's claims are getting denied for credentially, that you need to arrest that.
00:41:18.320 --> 00:41:30.639
If you see that a kiddo's claims are getting denied for benefits, you know you need to go back and find out what that parent's job, what happened there, did he stop paying for his insurance?
00:41:30.800 --> 00:41:32.559
We need to be able to do that quickly.
00:41:32.719 --> 00:42:03.599
So my first guideline or encouragement to a practor is look at the denial data every week and not just individual claims or individual denials, try to collate it into trends: payer-wise denials, provider-wise denials, you know, wise denials, location-wise denials, depending upon how your business is, you need to start slicing and dicing that denial data and look for some inferences that then become your front-end plan of action.
00:42:03.760 --> 00:42:09.039
Like what do I solve in my front office based on the denial data in the back office?
00:42:09.199 --> 00:42:10.800
That's the number one thing I would do.
00:42:11.280 --> 00:42:16.400
Because a lot of times leaders think billing is a bill department, but thing I need to worry about it.
00:42:16.480 --> 00:42:22.639
No, what happens in the billing department, the recording that comes out of it is inputs for you to go work on stuff.
00:42:22.800 --> 00:42:29.840
The other thing is what I just mentioned already, so I don't want to go back and repeat it a lot more, but the clinical responsibility of billing, right?
00:42:30.000 --> 00:42:35.840
So what I mean by that is the clinician needs to look at documentation, needs to look at authorization units.
00:42:36.000 --> 00:42:39.920
And if you don't do that, it's going to show up as denials in the back end.
00:42:40.000 --> 00:42:48.800
So the clinician knowing what they are doing in relation to billing is sort of my second most important element of how I would look at it, right?
00:42:48.960 --> 00:42:52.079
And metrics-wise, I would definitely go back to my original point.
00:42:52.239 --> 00:42:53.760
What are our first pass ratio?
00:42:54.000 --> 00:42:59.360
Like how many claims are going out and getting paid, how many are getting rejected, which is different from denial.
00:42:59.519 --> 00:43:08.559
Rejection says what are the claims that the clearing house or your software is scrubbing up and showing up in uh you know, rework.
00:43:08.800 --> 00:43:16.960
That again is a wealth of information out there for you to then go back and say, you know what, I'm there are errors here in my process that I need to fix.
00:43:17.199 --> 00:43:21.519
Outside of that, I would keep a very close watch on the AR above 90 days.
00:43:21.679 --> 00:43:25.920
If claims are not getting paid after 90 days, you have a problem with that claim.
00:43:26.159 --> 00:43:30.000
Within 90 days, I can see some denial, I can see some fixes.
00:43:30.159 --> 00:43:37.519
But if things are going past 90 days, more than likely you have to kiss goodbye to those claims, which means that that is going to be a write-off.
00:43:37.760 --> 00:43:43.840
Most people sit on those claims to 180 days, 200 days, and they think that's all valuable cash.
00:43:44.079 --> 00:43:44.800
It is not.
00:43:44.960 --> 00:43:52.960
I mean, you've got to be honest about it, and you gotta say, you know what, if this claim didn't get paid in 90 days, what are the chances of it getting paid in the next 90 days?
00:43:53.199 --> 00:43:54.079
Probably very low.
00:43:54.239 --> 00:44:08.079
And as long as you're honest about it, and as long as you're starting to plan around it, you know, you are you're gonna be more successful than owners who just fall in love with their car and they just sit on it and they hope that someday somebody will bring some cash out of it.
00:44:08.239 --> 00:44:12.400
And the reality is there is nothing there, most likely after the 90th day.
00:44:12.639 --> 00:44:26.159
So those are some of the things that I ask owners to keep an eye on and watch for and be very inquisitive about and be willing to ask those hard questions and act on it once they know what the answers to those are, kind of.
00:44:26.960 --> 00:44:35.599
So, all those questions, all those KPIs for y'all listening, that's the next conversation that we expect you to have with your team, your billers, etc.
00:44:36.000 --> 00:44:47.599
And depending on the answers, I have great news if you need somebody else, because they'll be at ABA Care's in August in Boston, and of course, available virtually for those as well.
00:44:47.679 --> 00:44:50.400
And that'll all be linked in this podcast as well.
00:44:50.639 --> 00:44:53.199
So, Thomas, thank you so much for being here.
00:44:53.440 --> 00:45:06.320
Thank you so much for representing Plutus, and especially thank you for letting me nerd out on all the data that you've collected and the way you've integrated tech into an operations and systems first approach with Plutus Health.
00:45:06.480 --> 00:45:13.760
For all you listening, you can find the Plutus Health booth at the ABA Care Summit August 4th through the 7th.
00:45:13.920 --> 00:45:23.280
Bring your worst AR report, bring the denial you can't figure out, bring the questions that you're kind of scared to get the answer to, head over to their booth and they'll be able to help you out.
00:45:23.519 --> 00:45:25.519
Thank you again, Thomas, for being here.
00:45:25.679 --> 00:45:29.199
And if you're going live to the conference, we will see you there.
00:45:29.360 --> 00:45:31.679
If not, we'll see you over the live stream.
00:45:31.840 --> 00:45:32.639
Thank you all so much.
00:45:32.800 --> 00:45:33.360
Thank you, Matt.
00:45:33.519 --> 00:45:34.000
Appreciate it.
00:00:00.160 --> 00:00:08.720
Hello everyone, my name is Matt Harrington, and I am your Holly Beth host for the day for the Inside ABA Cares podcast.
00:00:08.880 --> 00:00:15.039
Your behind-the-scenes space where we lay the foundation and set expectations for who you'll meet this August.
00:00:15.199 --> 00:00:22.640
Now, most practice owners I talk to can tell you that their billing problem feels like the primary challenge in their business.
00:00:22.800 --> 00:00:26.000
It's almost like a death by a thousand paper cuts.
00:00:26.160 --> 00:00:29.679
Plutus Health is one of the companies that solve this problem.
00:00:29.920 --> 00:00:36.719
They're a sponsor of the summit in August, and they're one of the main reasons we're able to put on this conference at the scale that we do.
00:00:36.960 --> 00:00:47.600
Thomas John, my guest, is the founder and CEO of Plutus Health, which processes$3.5 billion in medical claims a year with denial rates under 5%.
00:00:48.320 --> 00:00:50.479
That's roughly half the industry average.
00:00:50.719 --> 00:01:04.799
He's been doing this for 30 years, and in 2023, he spun up Artemis, which we did a podcast with the other day with Cherie, specifically because he kept seeing the same operational failures repeat across ABA practices.
00:01:05.120 --> 00:01:07.200
Thomas, thank you so much for being here.
00:01:07.359 --> 00:01:13.359
For those who haven't met you yet or heard about Plutus, give us a little bit about your background and what Plutus does.
00:01:13.760 --> 00:01:14.480
Thank you, Matt.
00:01:14.560 --> 00:01:15.519
I appreciate that introduction.
00:01:15.760 --> 00:01:16.879
My name is Thomas John.
00:01:17.120 --> 00:01:19.200
I'm the former and CEO of Plutus Health.
00:01:19.359 --> 00:01:22.640
I'm also the founder and CEO of Artemis ABA.
00:01:22.719 --> 00:01:26.959
And my background is I started in technology about 30 years ago.
00:01:27.359 --> 00:01:31.920
And uh, you know, revenue cycle management found me, and ABA found me as well.
00:01:32.079 --> 00:01:41.120
So it's more of a pattern, you know, that I got introduced into some of these in-market opportunities that that then break the company that we are today.
00:01:41.359 --> 00:01:51.920
Across the last 30 years, I've been dealing with uh, you know, technology and claims in the uh surgery center market and the ER market and in behavioral health in general.
00:01:52.079 --> 00:02:05.439
And one of the things that I noticed over the years is uh just a pattern emerging around uh how revenue cycle management is an operating system and not just a bunch of activities that most people like it to be.
00:02:05.680 --> 00:02:11.439
So my background in technology allowed me to look at this from a process, technology, and people perspective.
00:02:11.599 --> 00:02:18.000
And that's how I ended up uh founding Pluto's Health and focusing on the ABA space over these years.
00:02:18.240 --> 00:02:31.520
For example, uh when I look at ABA, what I realized is that care is sold in in pre-authorized package of units by insurance companies, the way they control this care delivery process.
00:02:31.680 --> 00:02:39.919
And the other thing I found out is that, especially in the ABA market, it is not one pair, it is a lot of different pairs in different states.
00:02:40.080 --> 00:02:51.360
Medicaid governs the majority of the pair landscape, and every state in the country has a different nuance to how it pays and how it adjudicates claims and so on and so forth.
00:02:51.520 --> 00:03:03.439
So that's what drove me to look at ABA as an market in which my experience in technology and RCM could be applied, and that's how we built Bluetooth Health over these years.
00:03:03.759 --> 00:03:13.680
Can you uh zoom in on the way that you've looked at it from an operating systems perspective versus just a general, hey, here's a problem, let's hire someone to solve it.
00:03:13.840 --> 00:03:23.919
I'm curious, especially in the way you think about the vastly different ways different insurances handle claims and denials and all of that stuff.
00:03:24.319 --> 00:03:24.960
Great questions.
00:03:25.120 --> 00:03:34.719
When you look at the insurance landscape in the ABA world, the first big difference is the commercial versus the gentleman payers.
00:03:34.879 --> 00:03:40.879
So commercial payers could be the Blue Cross, Blue Shields, the Optons, and or even ITNA and so on.
00:03:41.039 --> 00:03:45.520
And then on the other hand, you have all the Medicaids that are state-specific.
00:03:45.680 --> 00:04:05.360
So, and then when you look at the processes which are governed by these insurance companies, they start with credentialing pay providers, and there are different requirements of documentation, different timelines to get it done, different frameworks of when a panel is open or not in that world.
00:04:05.520 --> 00:04:13.120
Same thing with prior authorization, different documentation requirements, different timelines for approval, uh, and so on.
00:04:13.280 --> 00:04:26.639
And then finally, the most important thing is the claim itself and what governs documentation standards around session notes, treatment plans, all of these have different flavors in different pair uh structures.
00:04:26.800 --> 00:04:50.000
But the most complex differences come inside the Medicaid world where you might think that a particular claim, when it is built in a certain way, in, you know, let's say in Texas, the same Medicaid in Indiana may approach it differently and say, for example, in Indiana, the BCBA and the RBT both need to be credentialed, whereas that's not the case in, let's say, another state.
00:04:50.240 --> 00:05:02.879
So those are some nuances that drive this, but there could be many more issues that drive both the adjudication process and, for example, the denials that come out of it, and then the resolution of those denials.
00:05:03.040 --> 00:05:17.279
Um, for example, in Plutus, we built something called Pi, which is a payer intelligence engine, which has aggregated everything that we know about every pair that we have built for over the last several years, running into billions of dollars of claims.
00:05:17.439 --> 00:05:30.800
And we have synthesized that into a model that we can refer to from time to time as we look at a claim or look at a credentialing application or a prior author application or an appeals package we send out.
00:05:30.959 --> 00:05:40.160
And keep in mind that Pi is pair intelligence engine as its uh acronymous based on today's information that may become outdated tomorrow.
00:05:40.319 --> 00:05:42.399
So we have to constantly refresh that.
00:05:42.639 --> 00:05:47.759
So a company with our experience and resource itself finds it challenging.
00:05:47.839 --> 00:06:08.240
So I can understand why uh a small uh ABA practice with one BCBA providing care and then also billing uh on the weekends may not be able to comprehend all the nuances at every payer and every state level can be I want to dive into AI a little bit later, but before we get there, I'm curious.
00:06:08.480 --> 00:06:16.000
I heard a quote the other day that said something along the lines of every tech and AI company is actually a data company in disguise.
00:06:16.240 --> 00:06:22.560
And that came to mind when you're sharing your billions of claims over seven years.
00:06:22.800 --> 00:06:33.759
So I'm curious as you're training this model, as you're sharing this data with your staff, were there any surprises that came from analyzing the trend of that much data over that many years?
00:06:34.079 --> 00:06:43.839
The surprise that I would allude to is that a lot of the analysis shows a lack of consistency and adjudication across time, right?
00:06:43.920 --> 00:06:51.920
A same pair may adjudicate a claim in a certain way at a certain point, and then you will see that it's not always the same pattern.
00:06:52.079 --> 00:07:03.680
So it's it would have been nice if we could have fallen into that pattern and analyzed that pattern, but the pattern is that there is sometimes no pattern to adjudication or decision making on the payer side.
00:07:03.839 --> 00:07:05.519
So that's one thing I can talk about.
00:07:05.600 --> 00:07:12.720
The other is again the inconsistencies of policies and procedures across payers, across states.
00:07:12.800 --> 00:07:14.800
Uh that's something interesting that we found.
00:07:14.959 --> 00:07:33.519
Also, I think one of the things that we see is that while the claim form itself has a finite number of variables to it that you can optimize towards, the reality also is what happens upstream is where I think all the gotchas on the pain points of ABA is.
00:07:33.759 --> 00:07:39.120
Example, credentialing, how many of your clinicians are not credentialed at the time of care?
00:07:39.279 --> 00:07:58.399
Or, you know, how many times a provider ends up providing care without an underlying authorization, you know, how many times the documentation doesn't meet the payer guidelines that is found out after the fact, and then the payers come back and audit it and then start acclawing back money.
00:07:58.639 --> 00:08:02.959
So those are a lot more challenging than the claim form itself.
00:08:03.040 --> 00:08:13.600
Like a lot of times we think, okay, as long as I can keep my claim form clean, whether the right modifiers and the right codes and all of that, things will work well.
00:08:13.759 --> 00:08:29.040
And what surprised me is that a lot of the problems in ABA, or in fact in medical billing in general, are issues that happen upstream or earlier in the process that come back and surface as denials or as no cash in the bank.
00:08:29.439 --> 00:08:35.200
Which brings us to some of the conversations and the way you look at broken revenue or broken RCM.
00:08:35.440 --> 00:08:41.360
So you've said in the past that broken revenue isn't really a billing problem, it's more of a care problem.
00:08:41.679 --> 00:08:43.200
Can you expand on that, though?
00:08:43.679 --> 00:08:48.639
Revenue and care are two sides of the same coin or different sides of the same coin, right?
00:08:48.879 --> 00:08:53.840
So what I mean by that is, for example, when we look at ABA as a service, right?
00:08:53.919 --> 00:08:59.440
I mean, it is ultimately clinical capacity financed by cash conversion, right?
00:08:59.600 --> 00:09:19.600
So the your ability to deliver care is a function of your ability to convert your claims into dollars, because ultimately you are a services business, and every dollar that you generate is going back to pay for a future care process or a care uh or a service uh you know structure.
00:09:19.840 --> 00:09:39.039
So what I mean by that is, for example, if a clinician, a BCBA that owns a practice is unable to convert their previous months' claims into cash, there is going to be a significant impact on their ability to pay for employees like BCBAs and RBTs, or even just grow their business by hiring more people.
00:09:39.200 --> 00:09:45.759
So to that, so that's the primary driver of uh you know revenue and care going hand in hand.
00:09:45.840 --> 00:09:51.919
Or if one is not getting logically converted to cash, the other is going to be sooner than later impact.
00:09:52.080 --> 00:10:09.120
There might be a lag between the time you look at, okay, I'm not able to convert every one of my claims into dollars, and my future growth plan is contingent upon having that cash in the bank in order to finance my continued operation and also my growth.
00:10:09.279 --> 00:10:11.679
The second mechanism is a trust, right?
00:10:11.759 --> 00:10:13.679
It behaves like retention economics.
00:10:13.840 --> 00:10:17.679
So to a family, the bill is part of that clinical product.
00:10:18.080 --> 00:10:24.399
An error-driven balance in a statement to a family is going to drive churn, right?
00:10:24.480 --> 00:10:27.759
I mean, your parent is like, okay, I mean, why am I getting this wrong bill?
00:10:27.919 --> 00:10:32.799
I'm gonna, you know, be upset about it, I might move away, I might go to a different clinic.
00:10:32.960 --> 00:10:42.720
So your ability to then deliver that care with that family is a function of accurate billing, accurate payment uh collections from that from that family.
00:10:42.879 --> 00:10:45.120
So it's the revenue is a care problem.
00:10:45.360 --> 00:10:55.200
It's not just being philosophical, basically naming the two places, capacity and retention, where the back office sets the ceiling on clinical outcomes every month, right?
00:10:55.279 --> 00:11:07.840
Because if your back office is not doing a good job of converting that that care that you've already provided into cash, the ability for you to do that effectively in the future is hurt significantly.
00:11:08.639 --> 00:11:20.000
So what are some of the warning signs that owners should look out for operators when it comes to their back office not converting efficiently enough and care eventually being impacted?
00:11:20.159 --> 00:11:21.440
What's the canary in the coal mine?
00:11:21.840 --> 00:11:25.200
Well, I would I would say that very easy KPIs out there, right?
00:11:25.279 --> 00:11:28.720
I mean, the most basic one is denial percentage, right?
00:11:28.879 --> 00:11:33.200
So if you're sending out 100 claims, what how many of those claims are getting denied?
00:11:33.279 --> 00:11:34.879
So what is your denial percentage?
00:11:35.039 --> 00:11:50.080
That's the most easiest number to track and use as the uh early warning signal for you to know that if your denial percentages are higher and higher month over month, something ain't right in that in that business model, right?
00:11:50.240 --> 00:12:02.960
We can get into why that's happening in a minute, but since the fact if you're sending out 100 claims a month and you're only getting paid, let's say 85 of them and the other 15 is getting denied, that's an early warning sign for you.
00:12:03.200 --> 00:12:05.440
The other area is days in AR.
00:12:05.519 --> 00:12:10.080
Like what is your turnaround of claims to dollars, right?
00:12:10.240 --> 00:12:12.720
If your days in AR is less than 30 days, great.
00:12:12.879 --> 00:12:17.919
If it is getting to 40, 50 days, 60 days, that means something is wrong.
00:12:18.240 --> 00:12:21.600
And you need to be looking into why those claims are not getting paid.
00:12:21.759 --> 00:12:30.399
Now, why I kind of differentiate between a denial percentage and a days in AR is that sometimes payers don't deny, they just delay.
00:12:30.559 --> 00:12:39.039
Now, whether that delay is because of something that you did or something that the payer is doing, that's another nuance to get into in a minute.
00:12:39.200 --> 00:12:47.039
But the reality is that if you are seeing outstanding claims aging past a certain point, you need to be worried, right?
00:12:47.200 --> 00:12:59.519
Then, of course, the most important one is net collection ratio, which is after all is said and done, let's say three months into this process, what percentage of your claims did you finally collect?
00:12:59.600 --> 00:13:02.639
And what percentage of claims did you finally have to write off?
00:13:02.799 --> 00:13:08.399
If that is like 99% collected, 1% written off, I think you're in a healthy practice.
00:13:08.559 --> 00:13:15.120
If that is 90% collected and 10% written off, you're probably skating on thin ice.
00:13:15.279 --> 00:13:20.799
Because now you're going to be concerned at some point that PBA is not a high margin business.
00:13:20.960 --> 00:13:22.799
It's a pretty tight margin business.
00:13:22.879 --> 00:13:31.519
So if 10% that already tight margin business is being written off, then you're probably losing money and you don't know about it yet.
00:13:31.679 --> 00:13:38.799
Because keep in mind, you know, PL is not the right metric to define the health of your practice.
00:13:38.960 --> 00:13:40.720
It is a net collection ratio.
00:13:40.960 --> 00:13:48.799
Because you could book revenue by seeing a kid, and after you sent out a claim, you're saying, oh, I generated this revenue.
00:13:48.960 --> 00:13:52.799
But that revenue converts to cash based on your net collection ratio.
00:13:52.960 --> 00:14:00.559
Because you can easily go around showing off a PL that shows that you generated a million dollars in top line.
00:14:00.639 --> 00:14:04.720
But how much of it was realized in cash is the real question, right?
00:14:04.879 --> 00:14:22.399
And many a time, as even a seasoned CFO, and I've seen a lot of private equity back seasoned CFOs not realizing that yes, you can book revenue by seeing providing care, seeing a kid, sending out a claim, but did that claim convert to cash?
00:14:22.480 --> 00:14:27.519
And did it convert to cash in a reasonable period of time is what drives this.
00:14:27.679 --> 00:14:42.720
So these are some of the red flags that I always encourage both individual practice owners and even larger companies with seasoned CFOs to look into because sometimes those PL numbers might be misleading.
00:14:43.039 --> 00:14:46.320
So you talked about, and I'll just run through some of the KPIs.
00:14:46.399 --> 00:14:55.279
So correct me if I miss them, but you have a denial percent of about 5%, you have AR of 30 or less, you have net collection 99%.
00:14:55.840 --> 00:15:00.639
Let's say you come to somebody or they join in Plutus, and those numbers are way off.
00:15:00.720 --> 00:15:02.879
They're nowhere near kind of the baseline.
00:15:03.039 --> 00:15:08.159
What does the process look like of getting them back into the correct location?
00:15:08.639 --> 00:15:11.840
We start with what is called as an revenue cycle management assessment.
00:15:12.080 --> 00:15:21.679
And we generally do this free for almost all of our clients, our prospective clients, or even anyone who kind of gets a sense of where they are from a baseline perspective.
00:15:21.840 --> 00:15:28.320
And in that baseline, we are trying to identify the number one denial rate is the first indicator.
00:15:28.480 --> 00:15:29.919
But then that doesn't stop there.
00:15:30.000 --> 00:15:34.080
It goes to what is constituting your denials, right?
00:15:34.240 --> 00:15:37.759
What is the upstream issue that drove that denial?
00:15:37.919 --> 00:15:44.639
And we talked about it in briefly at the beginning of this conversation: credentialing, prior authorization, benefits verification.
00:15:44.879 --> 00:15:55.200
I can bet that 90% of all practices, when they have denials, they are related to almost always one of these three issues or all the three issues.
00:15:55.840 --> 00:16:02.159
So our goal is to identify those root causes because denials don't happen in the back office.
00:16:02.320 --> 00:16:07.200
Denials happen in the front office during the delivery of care.
00:16:07.440 --> 00:16:10.639
And those are sort of visible in the back office.
00:16:10.799 --> 00:16:21.200
So most CEOs, most BCBA owners, or even most CFOs are hyper-focused on saying my billing team is the problem, my billing is a problem.
00:16:21.440 --> 00:16:23.600
Really, it's not the billing that's a problem.
00:16:23.759 --> 00:16:28.639
Billing is only showing you what is the problem somewhere else in the business.
00:16:28.960 --> 00:16:35.519
Example, you may be hiring BCBAs, but you know, lagging the credentialing associated with it.
00:16:35.679 --> 00:16:45.759
You may be scheduling kiddos with the wrong BCBA, as in the BCBA that does not have the right credentialing for the coverage that the kiddo has.
00:16:45.919 --> 00:17:01.840
You may be getting the prior authorization, but then the parents quits their job or switches their job and has a new insurance, and you are still assuming that the previous prior authorization that you got is valid and you're providing care on a wrong assumption.
00:17:02.000 --> 00:17:16.400
You might be providing care, but the authorization expired, and somebody who got the authorization did not inform the clinician, and now the clinician is making a wrong assumption that this kid still has units to provide service.
00:17:16.480 --> 00:17:18.880
And then you may have done it in good faith.
00:17:18.960 --> 00:17:25.039
I mean, saying, okay, you know what, the kid needs that service, I'm going to do it, but underlying choir authorization was never there.
00:17:25.200 --> 00:17:32.640
So these are some of the things that we look at, and then we try to work with the practice to tighten those front-end processes.
00:17:32.720 --> 00:17:39.680
And remember when I started the conversation, I said process comes first, and then comes technology and people in order to fix this.
00:17:39.759 --> 00:17:44.559
So that process is what we are fixated on solving for.
00:17:44.720 --> 00:17:49.680
And each and every process we solve in that front-end is more dollars in the back end.
00:17:50.160 --> 00:17:58.400
What does the general timeline look like from solving these front-end processes to seeing the effect on the back-end KPIs?
00:17:58.640 --> 00:18:01.039
Is it a month, two months, varies?
00:18:01.839 --> 00:18:07.680
To fully run through the process of we will attack this problem from a multiple dimensions, right?
00:18:07.839 --> 00:18:10.640
One is we solve for what has to be prevented.
00:18:10.799 --> 00:18:16.079
One is second is we solve for what we need to mitigate, you know, you know, uh go back and fix.
00:18:16.559 --> 00:18:20.480
For example, there's a bunch of denied claims that needs to go back and get fixed.
00:18:20.559 --> 00:18:27.440
So we may have to try out and see if we can get some retros on the credentialing and prior loss, which is very difficult most of the time.
00:18:27.599 --> 00:18:31.920
Most payers do not even agree to go back and retroauthorize something.
00:18:32.000 --> 00:18:34.799
Uh and if they do, it's limited periods.
00:18:34.960 --> 00:18:39.759
If they are doing credentialing and it's delayed, uh, they may retro for some date.
00:18:39.920 --> 00:18:42.000
So these are things that we can go back and fix some.
00:18:42.319 --> 00:18:48.480
So the cash coming from already denied claims being fixed generally takes 60 to 90 days.
00:18:48.799 --> 00:18:55.359
And then the preventions, they can start solving, giving you uh cash in the bank a little faster, maybe within 30 days.
00:18:55.599 --> 00:18:59.359
So I would say when you're solving for this, you're solving for it at two levels.
00:18:59.519 --> 00:19:00.960
One is what can you prevent?
00:19:01.119 --> 00:19:08.640
And when you prevent something, you can see the impact within 30 days, because 30 days is normally the period of time it takes to adjudicate a claim.
00:19:08.799 --> 00:19:24.720
Uh, but if you're trying to go back and fix something that was already broken, uh, it may take 60, 90 days because payers are known to fix, pay you on fixed claims or appeal claims in a longer period of time because they have to manually review everything.
00:19:24.799 --> 00:19:26.559
So it's probably 90 days.
00:19:26.720 --> 00:19:37.279
So overall, I think you'll start seeing an improvement in cash flow, maybe 90 to 120 days out from the time a PLUDIS health team comes into the equation.
00:19:37.599 --> 00:19:44.400
Now, what does the owner or maybe the team of the operations team experience during those 90 to 120 days?
00:19:44.559 --> 00:19:47.039
Are they working with a representative at Pluto's?
00:19:47.200 --> 00:19:50.000
Do they have somebody helping them with those processes?
00:19:50.240 --> 00:19:55.039
What's the support like as they go through the started a new RCM process?
00:19:55.279 --> 00:19:57.440
Also, they said that we have to fix a bunch of things.
00:19:57.599 --> 00:19:58.799
What does that look like?
00:19:59.119 --> 00:20:04.240
When we come in, we look at what is it that was a problem in the back office?
00:20:04.319 --> 00:20:09.119
Maybe an existing billing team missed a lot of obvious things.
00:20:09.200 --> 00:20:17.599
And those are very low-hanging fruits, like not putting the right modifiers or not using the right, you know, payer information, things like that.
00:20:17.680 --> 00:20:31.039
Like you sent the claim to the wrong pair, or you had the wrong modifiers that are already retired by the insurance company, or maybe not applicable to a particular state, but you nevertheless use something that you would have used in one state in a different state.
00:20:31.200 --> 00:20:33.039
Things like that are easy, low-hanging fruit.
00:20:33.200 --> 00:20:40.319
So that we don't even go to the provider, we just do it ourselves, we fix it, they start seeing the benefits of that in cash in the bank.
00:20:40.480 --> 00:20:54.480
Then comes the more harder and difficult part of this process, which is identifying all those process gaps and then communicating that to the practice owner, to the CEO, or whoever is the key leaders within the organization.
00:20:54.640 --> 00:21:08.720
This is where our leader, someone with stature and gravitas, comes in and has real conversations with the leadership on the practice side and say, listen, these are some challenges we see that we need to come together to fix it.
00:21:08.960 --> 00:21:14.400
But I always tell my team and the client that your problems are my problems.
00:21:14.559 --> 00:21:28.160
My problems are still my problems as a billing company, but I will also own and co-own your problems, which means, for example, if your scheduler is not looking at prior auth units before scheduling a kid.
00:21:28.400 --> 00:21:40.240
If your scheduler is not looking at a credentialing roster to look at the BCBAs who are credentialed for a particular kiddos pair, now these are sometimes also require some homework to be done.
00:21:40.400 --> 00:21:47.119
Do you have a good tracker of all your BCBAs and their associated payer credentialing statuses?
00:21:47.519 --> 00:21:51.359
Like many, many practices don't even keep track of it in one single place.
00:21:51.519 --> 00:21:59.759
So that when a scheduler is scheduling, do they even know that Matt is not credentialed with a blue cross or blue shield in Texas?
00:22:00.160 --> 00:22:06.880
Or MAD is actually credentialed only with uh United or Aitna and not with some of the other pairs.
00:22:07.039 --> 00:22:21.759
Because those are nuances that A need to be tracked, and B, at the time of scheduling, you need to coach the scheduler to look out for those gachas where you end up credentialing or scheduling a kiddo with the wrong BCBA who does not have the credentialing and stuff like that.
00:22:21.920 --> 00:22:31.599
So that is where we will spend time in the front office creating those best practices, but then also guiding people, for example, recruiting.
00:22:31.759 --> 00:22:48.000
Like put a process in place where when you hire a clinician, immediately, as soon as a hiring offer letter goes, you notify credentialing team to start the process of collecting the documents that are required to credential with the payers that you're gonna be seeing.
00:22:48.079 --> 00:22:50.480
Now, do you also know who are your top payers?
00:22:50.640 --> 00:22:55.519
Then you know how long it takes to credential because you need to back up into all of that.
00:22:55.680 --> 00:23:10.240
Like if you are hiring a clinician and you know it takes 90 days to credential that clinician with these pairs, what are you gonna do about that for the next 90 days when you have a wait list, when you have kiddos, how are you gonna match all of these things?
00:23:10.400 --> 00:23:21.599
So these are all detailed operational discussions to be had with both leadership and with department and functional leaders who will then have to come back and tell us how to do this.
00:23:21.920 --> 00:23:22.720
Very interesting.
00:23:22.880 --> 00:23:26.079
And that's where the processes and the operations come in.
00:23:26.240 --> 00:23:35.200
And then once you get through the processes, now we get into what I think is the fun part, which is the tech, the AI, the pie, and all of that stuff.
00:23:35.359 --> 00:23:45.759
So can you explain your general thought process and kind of where you've directed Plutus to leverage AI, whether that be in the process of adjudication, submission?
00:23:46.319 --> 00:23:49.359
Where in general does AI touch your RCM process?
00:23:49.759 --> 00:23:52.960
We have really leaned into all of those elements.
00:23:53.039 --> 00:24:00.559
And I've already talked about the, you know, there are some processes like payment posting, where it's a much more deterministic approach to automation.
00:24:00.640 --> 00:24:07.920
So we can use automation to solve for posting payments, verifying against a bank statement, and so on and so forth.
00:24:08.000 --> 00:24:10.400
So those are easy, uh low-hanging fruits.
00:24:10.559 --> 00:24:14.240
We use APIs to check eligibility on ACE every two weeks.
00:24:14.400 --> 00:24:17.119
We use APIs to do insurance discovery.
00:24:17.279 --> 00:24:22.640
If we find out that a benefit of a parent is lapsed, does that parent have any other eligibility?
00:24:22.720 --> 00:24:24.880
These are all things that AI is not required.
00:24:24.960 --> 00:24:28.240
You can use pretty much automation and APIs to do it.
00:24:28.480 --> 00:24:49.279
Where we have used AI is, for example, today we have the ability to build a credentialing application completely from scratch using documents that VCBAs provide and using AI to validate timelines and document checklists and formats in which applications need to be provided.
00:24:49.440 --> 00:24:53.119
All of those can be done using uh AI today.
00:24:53.200 --> 00:24:55.279
So that's something we have put in place.
00:24:55.519 --> 00:24:57.599
Same approach with prior authorization.
00:24:57.759 --> 00:25:13.359
We have AI today that can take a treatment plan and all the underlying documents that payers ask for and create a prior auth package that can be sent to the payer's sort of portal in order to get prior auth approval done.
00:25:13.599 --> 00:25:16.000
We have denial analysis AI.
00:25:16.079 --> 00:25:40.000
So as soon as a denial comes from a payer, we can run it through our AI, and the AI would analyze the denial correspondence from the payer, the claim documentation, the underlying medical documentation, the underlying benefits documentation, underlying credentialing documentation, and come back and give you a plan of action to overturn that denial with the payer.
00:25:40.400 --> 00:25:49.680
We have been using able to use AI to increase our denial overturn, which was pre-AI at about 50-60% to about 70 to 80%.
00:25:50.000 --> 00:25:55.920
So you've seen an uplift of around 20% just by using AI to analyze the denial.
00:25:56.079 --> 00:26:13.039
We also are currently using AI to create appeals packages where we are able to look at all this underlying documentation that I mentioned to create a comprehensive appeal, again, referring to our Pi and payer policies, historical payer adjudication trends.
00:26:13.200 --> 00:26:25.519
We are able to create this comprehensive appeal that can go back to the payer with all the nuances that a human may have missed or does not have the time to frankly go and build an appeals package.
00:26:25.680 --> 00:26:29.119
Because a typical human appeals package takes several hours.
00:26:29.279 --> 00:26:35.440
I can shorten that time to maybe a few minutes, and then you can have a human review it, and then we can send that.
00:26:35.599 --> 00:26:37.599
So these are some places where we have done it.
00:26:37.759 --> 00:26:43.440
We also have the most important piece, I leave it to the last, which is the documentation review.
00:26:43.519 --> 00:27:05.200
So we are now using AI that can not only review the claim but also review the underlying documentation and scrub it ahead of time so that we are fully compliant with payer policies from a documentation perspective, because nine out of ten times you may have had the right intent and provided the right service to the kiddo, but did you document it well?
00:27:05.440 --> 00:27:10.000
Did you document it as per the payer's guidelines and policies is a million-dollar question.
00:27:10.160 --> 00:27:20.720
And many a time most clinicians trip on that, and we try to stay ahead of that by you know auditing those, just as I mentioned in the past, the payers are going to do the same thing.
00:27:20.799 --> 00:27:54.160
So we are preempting that and saying even before a claim gets created and sent to the payer, we will validate that that underlying documentation is kosher and that your claim and the documentation match so that if and when payer comes back in three months or six months and asks you for that documentation, you're not caught napping, or you know, that employee has left the building and you're like you know figuring out what do I do now because the payer is asking for documentation that that employee did not complete you know properly, but I bid for it and I got paid.
00:27:54.240 --> 00:27:57.359
Now I have to re return that money, and that's a revenue problem.
00:27:57.519 --> 00:27:59.440
So those are some of the use cases matter.
00:27:59.839 --> 00:28:02.240
There's so much of that that I want to dive into.
00:28:02.559 --> 00:28:03.839
Let's start here.
00:28:04.240 --> 00:28:13.440
One of the common things that I hear when I talk to your clinicians, your ABA owners about AI, is they say, Yeah, I used ChatGPT the other day.
00:28:13.680 --> 00:28:17.680
It told me to put ketchup on my pizza instead of tomato sauce.
00:28:17.839 --> 00:28:20.640
Like, why would I trust AI for this type of thing?
00:28:20.799 --> 00:28:26.799
And you know, in marketing, there's some obviously sometimes we use AI and we make it more efficient.
00:28:26.880 --> 00:28:30.319
But whenever you're thinking RCM, you're talking millions of dollars.
00:28:30.559 --> 00:28:40.160
So can you talk a little bit about the quality control, the training process that you use to make sure your AI is responsible with the millions that it's processing?
00:28:40.480 --> 00:28:42.480
Very great, very good question, I would say.
00:28:42.640 --> 00:28:49.680
So let me say that you just throw something into and it magically gives you something in return that is perfect.
00:28:49.920 --> 00:28:56.559
So what we have done essentially is we look at the LLMs as just computing capacity, right?
00:28:56.720 --> 00:29:10.240
It is the ability to process information in large scale, but it's up to us as the people who are running this to control the outcomes with good quality uh learning and guidance.
00:29:10.480 --> 00:29:19.519
So our RAG model, our Pi, these are sort of guardrails that clearly define the way AI is being utilized.
00:29:19.599 --> 00:29:26.000
So think of AI like a young kid out of college that you are hiring as an intern.
00:29:26.319 --> 00:29:34.480
And then now you need to train that intern on everything that you know in your business, and then you need to guardrail what that intern is doing.
00:29:34.640 --> 00:29:42.559
You know, he's got he's very smart, he's got you know perfect logical and analytical thinking skills, but he doesn't know anything about your business.
00:29:42.720 --> 00:29:59.519
Now you are going to use him, but you're going to use him in a manner where you have those guardrails of training in the SOPs that you built over the last 20, 30 years, your ability to comprehend the nuances of pair differentiation.
00:29:59.680 --> 00:30:04.160
Like these are things that Plutis has built over a 30-year period, right?
00:30:04.319 --> 00:30:30.319
The data that we have, the guidelines that we have, the SOPs that we have, these are things that are sitting around this AI computing power, and we are just merely taking that massive computing capability of the AI to look at 10,000 lines of document or pages of documents, hundreds of thousands of claims, which a human, let's be honest, cannot look at the same capacity like an AI can.
00:30:30.480 --> 00:30:31.680
So that's the difference.
00:30:31.839 --> 00:30:32.640
So you're right.
00:30:32.720 --> 00:30:38.319
I mean, you can't just throw a claim into AI and expect it to magically tell you what to do.
00:30:38.640 --> 00:30:45.039
You can write a small prompt, but it's probably going to hallucinate with what you can do with the way we have built AI.
00:30:45.279 --> 00:30:55.519
We are giving it the power of these additional capacity of uh SOPs, you know, guidelines, process documents, reference manuals, APIs.
00:30:55.680 --> 00:31:01.680
That's what makes AI powerful, not just AI's ability to throw something into Chat GPT and get an answer.
00:31:02.000 --> 00:31:15.759
You know, you don't just get highly accurate AI without more data than you think is possible, which is why Pluto's health a player for a long time can perform at a higher rate of accuracy in their AI than someone who just got started.
00:31:16.079 --> 00:31:19.359
Which is actually a question I have next.
00:31:19.680 --> 00:31:32.480
We see a lot of new companies, whether they be RCM companies or AI-powered companies, pop out of the blue now with a website and they're pitching ABA companies.
00:31:32.720 --> 00:31:41.119
In your opinion, what are some of the things that an ABA owner should look for to discern this company is one I can trust?
00:31:41.440 --> 00:31:44.880
Let me preface that by saying everyone has a role to play in this industry.
00:31:45.039 --> 00:31:46.079
This industry is massive.
00:31:46.240 --> 00:31:49.680
There is, you know, you can solve the problem from many different directions, right?
00:31:49.839 --> 00:31:53.920
There are companies coming into this, uh solve problems from a technology perspective.
00:31:54.000 --> 00:31:59.200
There are companies coming in from people uh capacity perspectives.
00:31:59.599 --> 00:32:01.279
And I think there's a place for everyone.
00:32:01.519 --> 00:32:06.559
But the real question is who can solve for this in a more comprehensive manner?
00:32:06.799 --> 00:32:13.839
So PLUTUS essentially characterize Plutus as five pillars of how we look at this business, right?
00:32:14.319 --> 00:32:38.079
The first pillar is that Pluto's almost always comes into a relationship with an ABA provider with an end-to-end offering, which means we like to come in and help our ABA clients from credentialing, prior authorization, benefits verification, claim scrubbing, payment posting, denial, AR follow-up, and uh patient payments.
00:32:38.319 --> 00:32:42.559
Now that kind of an end-to-end process is the first pillar of our business.
00:32:42.880 --> 00:32:46.079
We're not doing only one or two or three processes, we do everything.
00:32:46.240 --> 00:32:54.400
The second element is we almost always get paid on an outcome basis, which means we don't get paid if you don't get paid.
00:32:54.640 --> 00:33:07.119
So we very rarely, if ever, only in in absolutely rare cases do we ever say we will do uh you know something on a different pricing model, like uh, you know, FTE or transaction.
00:33:07.200 --> 00:33:11.200
We are always saying you get paid, we get a small piece of what you get paid.
00:33:11.279 --> 00:33:15.599
And we are accountable for the KPIs that as SL is in that contract.
00:33:15.759 --> 00:33:17.680
The third element is we are tech forward.
00:33:17.759 --> 00:33:27.279
So we have invested in technology and not just uh AI technology, AI automation, analytics are the three A's of our technology.
00:33:27.599 --> 00:33:33.440
The fourth element of it is that we look at this from the lens of speciality expertise.
00:33:33.599 --> 00:33:37.200
We have been in ABA and behavioral health for the last several years.
00:33:37.359 --> 00:33:39.359
That's a majority of our revenue.
00:33:39.599 --> 00:33:44.240
So we really understand the nuances of uh behavioral health and ABA.
00:33:44.400 --> 00:33:47.519
And last is we are also a very people-driven business.
00:33:47.680 --> 00:33:58.079
So we have a fairly large footprint of people, and we always have believed the human in the loop element to it, because the last mile of any process will need a human.
00:33:58.240 --> 00:34:02.000
And a lot of the technology companies may come and say, I'll give you this technology.
00:34:02.240 --> 00:34:06.000
What happens outside of that technology's capacity is up to you.
00:34:06.240 --> 00:34:08.719
We're not going to be able to be held accountable.
00:34:08.880 --> 00:34:11.280
And so that's the last element of our differentiator.
00:34:11.360 --> 00:34:13.679
So those are the five elements of how I look at it.
00:34:13.920 --> 00:34:23.519
So now when you go back to the companies that are only offering services as RCM companies, they're missing some of these pieces, like the ability to use technology.
00:34:23.599 --> 00:34:25.840
Because technology has a play in this business.
00:34:26.000 --> 00:34:27.199
We can't negate it.
00:34:27.360 --> 00:34:31.199
We need to embrace it and be willing to use it in every aspect of it.
00:34:31.440 --> 00:34:36.639
Having said that, we also can't say this is only tech, there is no people, there is no outcome.
00:34:36.800 --> 00:34:49.760
So a lot of the technology companies spinning out of Silicon Valley, I have noticed they're coming back and saying, Oh, I'll charge you 10 bucks for authorization, I'll charge you$3 for benefits, I'll charge you$20 for something else.
00:34:50.000 --> 00:34:53.760
The problem with that approach is those dollars add up pretty quickly.
00:34:53.840 --> 00:34:55.840
And are they correlating to outcomes?
00:34:56.000 --> 00:35:00.400
Most of these technology companies will say, that's not my problem, that's a you problem.
00:35:00.559 --> 00:35:18.559
But when we come into the picture, we are saying, we'll invest and provide you all that technology, but with towards the end of delivering dollars in your bank at the KPIs that I mentioned about, not that, okay, I'm just selling you some fancy, shiny next thing, and then it's up to you how we utilize it.
00:35:18.719 --> 00:35:24.079
Now, your technology may be perfect, but it requires sometimes some stewardship of people.
00:35:24.159 --> 00:35:34.880
And if you're not providing and you expect the mid-sized ABA practice with a BCBA owner to go figure it out all, then you may not be delivering the outcomes that they are expecting.
00:35:35.039 --> 00:35:39.039
So I always point my practices towards outcomes.
00:35:39.280 --> 00:35:45.920
No matter whatever you're doing, your bottom line is outcome because that's what pays the bills, and that outcome is cash in the bank.
00:35:46.079 --> 00:35:53.280
And if you can get 99% of that cash in the bank, every technology, every person, every expertise is worth every penny.
00:35:53.360 --> 00:35:54.880
Otherwise, it's worth nothing.
00:35:55.199 --> 00:36:00.079
The focus on accountability, I think, cuts through a lot of the marketing hype.
00:36:00.239 --> 00:36:02.559
As someone in marketing, I can attest to that.
00:36:02.800 --> 00:36:07.039
Let's talk about the worst letter people receive in the mail, the audit.
00:36:07.280 --> 00:36:15.840
Can you talk a little bit about the audit process and how Pluto can help with that awful time that AVA companies go through?
00:36:16.079 --> 00:36:29.039
Yeah, I think a lot of times people look at audit as some kind of an integrity check and that, but I think a lot of times payers are really trying to narrow what they're looking for, right?
00:36:29.360 --> 00:36:33.840
Payers, you know, in this context of audit are not really concerned about your intent.
00:36:34.159 --> 00:36:40.320
They're really concerned about are you uh you know documenting uh what you did correctly.
00:36:40.480 --> 00:36:49.519
So the point I think we need to uh inform uh owners who receive these audits is that are you substantiating that claim correctly, right?
00:36:49.599 --> 00:36:50.559
That's the bottom line.
00:36:50.719 --> 00:37:02.079
It doesn't matter that you know the kid needed the service, you actually provided the service, this was all uh just uh the payer just taxing you with with a lot of and burdening you with a lot of paperwork.
00:37:02.320 --> 00:37:10.239
I mean, the reality is the payer wants to pay you based on documented evidence, and all they care about is that evidence, right?
00:37:10.320 --> 00:37:13.199
You you know, they paid you, now they want to see an underlying evidence.
00:37:13.440 --> 00:37:22.000
Evidence has to be aligned to the guidelines that they have already provided you as part of their contract or as part of their general policies and procedures.
00:37:22.239 --> 00:37:26.639
So what gets tested is the tie between the note and the that's the bottom line, right?
00:37:26.800 --> 00:37:29.199
Does the session note support the units built?
00:37:29.360 --> 00:37:34.159
Is the provider on the claim who delivered is the one who also delivered the service, right?
00:37:34.480 --> 00:37:40.079
Is the required DCBA supervision RBT documented at the ratio that the payer mandates?
00:37:40.400 --> 00:37:41.679
Little things like that, right?
00:37:41.760 --> 00:37:45.199
I mean, are you doing those things is what the payer is looking for.
00:37:45.440 --> 00:37:59.119
So, you know, my recommendation to most owners is you've got to be living in the reality of doing this every day and not be sort of flippant about it and say, I'll deal with it when I get an audit.
00:37:59.360 --> 00:38:06.239
This has to be ingrained into the DNA of your business that no claim goes out without appropriate documentation.
00:38:06.400 --> 00:38:08.400
Now, that's a culture issue, right?
00:38:08.639 --> 00:38:15.039
A lot of practices that I've seen that have been very successful drive discipline from day one.
00:38:15.199 --> 00:38:29.840
They will be uncompromising with their providers, with their RBTs, with everybody in the organization, that the DNA of the organization is built around proper documentation, and that documentation cannot be compromised with.
00:38:30.079 --> 00:38:37.360
Doesn't matter that you provide great care to the kiddo, doesn't matter that you're a hardworking provider or clinician.
00:38:37.519 --> 00:38:40.159
What matters is did you document what you did?
00:38:40.320 --> 00:38:47.679
And if you didn't, you're risking the practice, you're risking risking revenue, you're risking integrity and compliance and all of that.
00:38:47.840 --> 00:38:51.199
So I tell the providers that the best solution is to prevent it.
00:38:51.360 --> 00:38:58.400
But now, if you still end up in an audit, yes, we need to come down and sit with you and start looking at every document.
00:38:58.639 --> 00:39:06.079
And document by document, you need to improve the quality of your deliverables to the payer.
00:39:06.239 --> 00:39:10.880
And once you reach a certain threshold of quality, the payer is going to take you out of the audit.
00:39:10.960 --> 00:39:14.079
Until then, they're going to be like after you on your case.
00:39:14.559 --> 00:39:33.760
And frankly speaking, fatal for practices that don't have the deep pockets of cash that they can fund while the payer is going to review every single claim, every single document of yours, and come back and pay you after 60, 90 days, where originally you would have been paid in 25 or 20 days.
00:39:33.920 --> 00:39:36.559
Now that becomes a 90-day cash flow problem.
00:39:36.719 --> 00:39:40.320
And it also creates all the stress with your clinicians.
00:39:40.480 --> 00:39:46.079
So the best solution is the day you start, be enforceable processes in place.
00:39:46.239 --> 00:39:50.719
Put enforceable processes in place with your clinicians and get the documentation right.
00:39:50.880 --> 00:39:59.280
But if you're already in an audit and you don't have documentation, you need to grid up and start putting the discipline in correcting everything.
00:39:59.440 --> 00:40:01.039
Eventually it'll come out of the audit.
00:40:01.119 --> 00:40:03.840
It may take 90 days, it may sometimes take six months.
00:40:04.159 --> 00:40:10.159
It all depends on how disciplined you are with uh how you're correcting the documentation and sending it off to repairs.
00:40:10.480 --> 00:40:14.719
We're almost at time, but I have one more question that I want to ask you.
00:40:14.960 --> 00:40:22.639
There's a lot of practice owners who are doing billing internally, billing externally, small teams, big teams with Plutus, not with Plutus.
00:40:22.800 --> 00:40:34.559
I'm curious what are the questions that they should be asking their billing team, internal, external, etc., to cut through everything and just say this is working or this isn't working?
00:40:34.880 --> 00:40:38.000
First thing I would say is start with the denials.
00:40:38.320 --> 00:40:52.079
If you're a practice owner, if you're a CEO or a CFO, I would encourage you to look at your denial report every month, if not every week, depending on how closely you want to monitor and manage this.
00:40:52.239 --> 00:40:59.360
I would, if I were them, I would probably monitor it every week because sometimes a month is too late in order to make corrective action.
00:40:59.440 --> 00:41:03.119
And the reason I say that is because again, I go back to my original point.
00:41:03.280 --> 00:41:10.400
That denial data has a wealth of information that you can use to transform your front-end operations.
00:41:10.639 --> 00:41:18.159
If you see that a particular clinician's claims are getting denied for credentially, that you need to arrest that.
00:41:18.320 --> 00:41:30.639
If you see that a kiddo's claims are getting denied for benefits, you know you need to go back and find out what that parent's job, what happened there, did he stop paying for his insurance?
00:41:30.800 --> 00:41:32.559
We need to be able to do that quickly.
00:41:32.719 --> 00:42:03.599
So my first guideline or encouragement to a practor is look at the denial data every week and not just individual claims or individual denials, try to collate it into trends: payer-wise denials, provider-wise denials, you know, wise denials, location-wise denials, depending upon how your business is, you need to start slicing and dicing that denial data and look for some inferences that then become your front-end plan of action.
00:42:03.760 --> 00:42:09.039
Like what do I solve in my front office based on the denial data in the back office?
00:42:09.199 --> 00:42:10.800
That's the number one thing I would do.
00:42:11.280 --> 00:42:16.400
Because a lot of times leaders think billing is a bill department, but thing I need to worry about it.
00:42:16.480 --> 00:42:22.639
No, what happens in the billing department, the recording that comes out of it is inputs for you to go work on stuff.
00:42:22.800 --> 00:42:29.840
The other thing is what I just mentioned already, so I don't want to go back and repeat it a lot more, but the clinical responsibility of billing, right?
00:42:30.000 --> 00:42:35.840
So what I mean by that is the clinician needs to look at documentation, needs to look at authorization units.
00:42:36.000 --> 00:42:39.920
And if you don't do that, it's going to show up as denials in the back end.
00:42:40.000 --> 00:42:48.800
So the clinician knowing what they are doing in relation to billing is sort of my second most important element of how I would look at it, right?
00:42:48.960 --> 00:42:52.079
And metrics-wise, I would definitely go back to my original point.
00:42:52.239 --> 00:42:53.760
What are our first pass ratio?
00:42:54.000 --> 00:42:59.360
Like how many claims are going out and getting paid, how many are getting rejected, which is different from denial.
00:42:59.519 --> 00:43:08.559
Rejection says what are the claims that the clearing house or your software is scrubbing up and showing up in uh you know, rework.
00:43:08.800 --> 00:43:16.960
That again is a wealth of information out there for you to then go back and say, you know what, I'm there are errors here in my process that I need to fix.
00:43:17.199 --> 00:43:21.519
Outside of that, I would keep a very close watch on the AR above 90 days.
00:43:21.679 --> 00:43:25.920
If claims are not getting paid after 90 days, you have a problem with that claim.
00:43:26.159 --> 00:43:30.000
Within 90 days, I can see some denial, I can see some fixes.
00:43:30.159 --> 00:43:37.519
But if things are going past 90 days, more than likely you have to kiss goodbye to those claims, which means that that is going to be a write-off.
00:43:37.760 --> 00:43:43.840
Most people sit on those claims to 180 days, 200 days, and they think that's all valuable cash.
00:43:44.079 --> 00:43:44.800
It is not.
00:43:44.960 --> 00:43:52.960
I mean, you've got to be honest about it, and you gotta say, you know what, if this claim didn't get paid in 90 days, what are the chances of it getting paid in the next 90 days?
00:43:53.199 --> 00:43:54.079
Probably very low.
00:43:54.239 --> 00:44:08.079
And as long as you're honest about it, and as long as you're starting to plan around it, you know, you are you're gonna be more successful than owners who just fall in love with their car and they just sit on it and they hope that someday somebody will bring some cash out of it.
00:44:08.239 --> 00:44:12.400
And the reality is there is nothing there, most likely after the 90th day.
00:44:12.639 --> 00:44:26.159
So those are some of the things that I ask owners to keep an eye on and watch for and be very inquisitive about and be willing to ask those hard questions and act on it once they know what the answers to those are, kind of.
00:44:26.960 --> 00:44:35.599
So, all those questions, all those KPIs for y'all listening, that's the next conversation that we expect you to have with your team, your billers, etc.
00:44:36.000 --> 00:44:47.599
And depending on the answers, I have great news if you need somebody else, because they'll be at ABA Care's in August in Boston, and of course, available virtually for those as well.
00:44:47.679 --> 00:44:50.400
And that'll all be linked in this podcast as well.
00:44:50.639 --> 00:44:53.199
So, Thomas, thank you so much for being here.
00:44:53.440 --> 00:45:06.320
Thank you so much for representing Plutus, and especially thank you for letting me nerd out on all the data that you've collected and the way you've integrated tech into an operations and systems first approach with Plutus Health.
00:45:06.480 --> 00:45:13.760
For all you listening, you can find the Plutus Health booth at the ABA Care Summit August 4th through the 7th.
00:45:13.920 --> 00:45:23.280
Bring your worst AR report, bring the denial you can't figure out, bring the questions that you're kind of scared to get the answer to, head over to their booth and they'll be able to help you out.
00:45:23.519 --> 00:45:25.519
Thank you again, Thomas, for being here.
00:45:25.679 --> 00:45:29.199
And if you're going live to the conference, we will see you there.
00:45:29.360 --> 00:45:31.679
If not, we'll see you over the live stream.
00:45:31.840 --> 00:45:32.639
Thank you all so much.
00:45:32.800 --> 00:45:33.360
Thank you, Matt.
00:45:33.519 --> 00:45:34.000
Appreciate it.
