Encompass Health Corporation (EHC) Q3 2022 Earnings Report, Transcript and Summary

Good morning, everyone, and welcome to Encompass Health's Third Quarter 2022 Earnings Conference Call. [Operator Instructions] Today's conference is being recorded. [Operator Instructions] I will now turn the call over to Mark Miller, Encompass Health's Chief Investor Relations Officer. Please go ahead, sir.

Thank you, operator, and good morning, everyone. Thank you for joining Encompass Health's Third Quarter 2022 Earnings Call. Before we begin, if you do not already have a copy, the third quarter earnings release, supplemental information and related Form 8-K filed with the SEC are available on our website at encompasshealth.com. On Page 2 of the supplemental information, you will find the safe harbor statements, which are also set forth in greater detail on the last page of the earnings release. During the call, we will make forward-looking statements, which are subject to risks and uncertainties, many of which are beyond our control. Certain risks and uncertainties like those relating to regulatory developments as well as volume, bad debt and labor cost trends that could cause actual results to differ materially from our projections, estimates and expectations are discussed in the company's SEC filings, including the company's earnings release and related Form 8-K, the Form 10-K for year ended December 31, 2021, and the Form 10-Q for quarters ended March 31, 2022, June 30, 2022, and September 30, 2022, when filed. We encourage you to read them. You are cautioned not to place undue reliance on the estimates, projections, guidance and other forward-looking information presented, which are based on current estimates of future events and speak only as of today. We do not undertake a duty to update these forward-looking statements. Our supplemental information and discussion on this call will include certain non-GAAP financial measures. For such measures, reconciliations to the most directly comparable GAAP measure is available at the end of the supplemental information, at the end of the earnings release and as part of the Form 8-K filed yesterday with the SEC, all of which are available on our website. [Operator Instructions] With that, I'll turn the call over to Encompass Health's President and Chief Executive Officer, Mark Tarr.

Mark, thank you, and good morning, everyone. We continue to see strong underlying demand for our services in Q3, with total discharge growth of 7.5%, including a 4.1% increase in same-store discharges. Our value proposition is resonating across payers, and we are gaining share in what remains an underserved market. Our hiring and retention strategies again proved effective at reducing our reliance on agency staffing, leading to another quarterly sequential decline in contract labor FTEs and expense in Q3. Market conditions for skilled clinicians remain challenging and the benefits of lower contract labor in the quarter were partially offset by higher sign-on and shift bonuses. Our Q3 profitability was negatively impacted by hurricane activity, delayed openings at 2 de novos and inflationary effects on supplies and utility costs. These factors led us to revise our 2022 guidance, as Doug will address in greater detail in just a few minutes. These are not the results we'd hoped for, but we are extremely proud of how our team has continued to respond to this volatile and challenging environment. Our de novo embedded edition strategies are increasing the availability of the highly specialized services we provide to meet this rising demand. Our commitment of substantial capital to these capacity expansions underscores our confidence in the future prospects of our IRF business. We opened 3 de novos in Q3, making that 9 year-to-date. We expect to open 8 de novos in 2023. We added 20 beds to existing hospitals in the quarter, raising the year-to-date total to 87. We expect to add an additional 100 to 125 beds in 2023. In the design and construction of our de novos, we are furthering our utilization of prefabrication to enhance speed to market, ensure consistent quality and contain costs. We are working on a number of new initiatives for the next generation of prefabrication and look forward to sharing more details with you in early 2023. As we have discussed previously, we have responded to the challenging labor market for skilled clinical resources in a number of ways, including adding substantial resources to our talent acquisition team. Our strategy of building a centralized recruiting function is generating successes. Same-store net RN hires were 459 for the first 3 quarters of 2022 as compared to 289 in the same period last year, a 62% increase. Q3 same-store net RN hires of 183 is up from 149 in Q2 and 127 in Q1. We also remain keenly focused on managing productivity. Our EPOB for Q3 was 3.39%, in line with our expectations. We achieved this productivity even while opening 3 de novos in the quarter. We are also continuing to create favorable outcomes for our patients and thereby our payers. Our Q3 discharge to community rate was 81.9% as compared to 81.3% in the year ago period and our discharge to SNF rate was 7% as compared to 7.5% in the year ago period. Moreover, our staffing challenges have not tendered our ability to provide great patient care as evidenced by our favorable Net Promoter scoring trends for patient satisfaction. On October 1, the fiscal year 2023 IRF rule became effective providing us with an approximately 4% Medicare price increase. This is the largest price increase we've had in quite some time. The market basket update leading to this increase is based on trailing information and lags prevailing inflationary environment. We are optimistic that ensuing reimbursement changes will close that gap. The benefit of this Q4 reimbursement increase is being partially offset by sequestration. On the regulatory front, the improving Medicare Post-Acute Transformation Act of 2014, known as the IMPACT Act requires collection and reporting of quality measures in standardized patient assessment data elements across post-acute providers. The 2022 IRF-PAI 4.0 form is now 30 pages of admission and discharge interdisciplinary data elements. Consistent with our prior regulatory update, we began preparations for these changes well in advance to ensure that our associates were well trained and fully ready for transition. We have implemented IRF-PAI 4.0 across all of our hospitals with little to no disruption to our operations. While CMS has not yet announced a start date for IRF review choice demonstration, it is slated to begin in Alabama first. We have been proactively communicating with our Medicare administrative contractor for Alabama regarding the potential implementation process. With regard to our revised 2022 guidance, we now expect net operating revenues of $4.32 billion to $4.35 billion, adjusted EBITDA of $800 million to $820 million and adjusted EPS of $2.71 to $2.86 per share. The key considerations underlying this guidance can be found on Page 13 of the supplemental slides. As I mentioned previously, Doug will provide more details about our revised guidance. With that, I'll turn it over to Doug.

Thanks, Mark, and good morning, everyone. I'll start with a quick recap of Q3. Q3 revenue increased 7.8% over the prior year to $1.09 billion and adjusted EBITDA declined 2% to $195.3 million. Our year-to-date adjusted free cash flow was $294 million. As Mark discussed, we continue to see strong volume growth that combined with a modest 0.7% increase in revenue per discharge produced out 8.2% inpatient revenue growth in the quarter. Once again, staffing challenges did not limit volume growth, but did result in the continuation of elevated costs. Our Q3 contract labor plus sign-on and shift bonuses of $49 million was comprised of $24.8 million in contract labor and $24.2 million in sign-on and shift bonuses. Contract labor expense in Q3 declined approximately $10.3 million or 29% from Q2 and $17.1 million or 41% from the Q1 peak. We experienced sequential declines in contract labor expenses and FTEs for every month in Q3, and in fact, for every month since March of 2022. Contract labor FTEs, which peaked at 749 in March, declined to 597 in June, fell further to 447 in September. Agency rates also declined during the quarter. The Q3 agency rate per FTE was $205,000 as compared to $223,000 in Q2. Agency rates do remain market specific and highly variable. As an example, after steady monthly declines, we experienced an uptick in average agency rates in September. Consequently, while we expect our total contract labor costs to decline again in Q4, our revised guidance assumes the pace of improvement will be slower than previously expected. Sign-on and shift bonuses increased sequentially to $24.2 million from $21.8 million in Q2. Sign-on bonuses increased approximately $1.8 million, and shift bonuses increased approximately $600,000. We had expected less of an increase in sign-on bonuses and we also expected a decline in shift bonuses as we move from Q2 to Q3. The increases stem from the net new RN hires, higher sign-on bonuses at the 3 de novos opened during the quarter and the use of shift bonuses to cover periods of unexpectedly high clinical staff PTO usage during the summer. Our updated guidance assumes sign-on and shift bonuses will decline in Q4, albeit at a slower pace than previously expected. As Mark noted, we remain very pleased with the results of our de novo and bed addition programs. We had previously expected our 2022 de novos to be breakeven to modestly accretive in the second half of the year. However, during Q3, opening delays at 2 of our hospitals, higher initial staff recruiting costs and hurricane-related disruptions at our newly opened Naples facility led to the approximately $5 million of net preopening and ramp-up costs incurred in the quarter. Our updated guidance assumes that 2022 de novos will be approximately breakeven in Q4. There are 3 other items that arose in Q3 that were outside of our expectations, negatively impacting our adjusted EBITDA for the quarter and contributing to our revised guidance. Net provider tax revenues, which can be volatile from quarter-to-quarter and thus are difficult to forecast declined by approximately $2 million as compared to Q3 of last year. Utility costs exceeded our expectations based both on higher utilization and increased rates. The higher utilization occurred primarily in Texas and Florida, our 2 largest markets, which both experienced very hot summers. Utilities expense per patient day increased by approximately 13% over Q3 last year and 21% over Q2. And finally, food costs increased beyond our expectations as inflationary pressures drove the cost per patient day up 15% over the prior year period and accelerating from up almost 8% in Q2. An enumeration of key assumptions underlying our revised guidance may be found on Page 13 of the supplemental slides. And with that, we'll open the line for questions.

[Operator Instructions] With that, I'll turn the call back over to Mr. Mark Tarr. It sounds good. I'm going to pull the first question. First question comes from Kevin Fischbeck with Bank of America.

So I guess I wanted to focus on labor. Obviously, you took up the labor outlook for Q4. I mean how are you thinking about the improvement into next year? How should we think about wages? How should we think about contract labor? And then I think sometimes people think that there may be too much of a lift to the bottom line for contract labor going down because you have to fill that with a full-time person. How should we think about kind of wage growth next year, but then the net-net of contract coming down? And is this normalized next year? Or is this a 2- or 3-year kind of thing that we should be thinking about for labor?

So Kevin, it's Doug. I don't know that we'll be completely normalized because we're not sure what that means anymore in 2023, but we certainly do expect the outline that we'll see sequential improvement from Q3 to Q4, and we would anticipate with regard to 2023 as a whole, that you'll see improvement over 2022 in both contract labor and sign-on and shift bonuses. There's still a lot of volatility out there. You heard me mention during my remarks that we saw -- not overly alarming, but we did see an increase in the average rate in September after seeing sequential declines in that rate from March all the way through. Now contract labor expense in total is much [Technical Difficulty] 250 and 300. Now that number is probably 300 to 350, with the low end of that range being optimistic. And the two other things that I would throw in that add some uncertainty and make it difficult to give you just a very straight and concise answer to your question is you get a lot of PTO in the fourth quarter around the holidays as well. And so we don't know what that will necessitate similar to what we experienced in the summer with associates taking PTOs regard to the need to fill in, that could have an impact on the rate as well. And then finally, although right now, it appears like the coronavirus is not going to be a severe winter as some had earlier feared, I think there are some [traditions] about the confluence group, some kind of winter surge of COVID combined with a maybe a more volatile flu season. Both of those things, as we entered 2023, would cause some potential staffing challenges in the period that is normally our highest volume in the year.

Kevin, this is Mark. I just want to add in there. We mentioned earlier about adding the talent acquisition resources. I mean we've done a number of things to make sure that we're well positioned going forward to do everything that we can to support our hospitals and fill the open positions that we have. And we now have over 60 full-time recruiters that have centralized that recruitment function that take some of the load off of the local hospitals in order to do that. And as evidenced by our net RN hires, we're making some significant progress on that. So we are doing everything we can to strategically position us against this very challenging environment.

So, the last thing I'd throw in there, I'm probably piling on, but it doesn't relate to specific contract labor, but rather to the sign-on bonuses. And we have made some significant progress in kind of getting some standardization around those. And also, we've been successful at now paying less of the amount upfront and tying more of it to retention periods. And those vary a little bit from market to market based on the competition. But whereas earlier in this year, we were having to pay a significant, if not all, of the sign-on bonus upfront and take risk around retention, we're now able to tie that. Now that creates a bit of [chaos] on those payments, for instance, substantial portion of what you'll probably see in Q4, as our reported sign-on bonuses actually relate to the hires that were made in Q3. But overall, that's a positive trend.

I think it's safe to say we're testing, we're challenging the marketplaces in terms of the overall amount. And if the sign-on bonus is even required as well as the level of shift bonuses and the rate we're willing to pay on contract labor.

And I think we've just had a record for the longest answer to a single question.

Yes. And I guess I'm still looking for the wage growth number, but if you could give that next year expectation? And then maybe just the second question being, when you think about -- your comment that labor is improving a bit more slowly than you thought something we've heard from everyone in the last couple of quarters. Is there something in particular that you would point to as to why that is? And if that's a sign that if you just get better, faster next year? Or whether it just means that we should be prepared for this just to be a slow improvement for longer?

So we -- first of all, we I will tell you, Kevin, I just want to answer your question with regard to 2023. But if you want a bit of sense as to the underlying wage growth when you strip out some of this noise. If we look at our SWB increase, ex sign-on and shift bonuses. So if you pull that out, and of course, that also doesn't include contract labor. In the second and third quarter, it was running up about 3.5%. And that's compared to 4.7%. This is on a year-over-year basis. So we're not seeing anything overly alarming in the core wage inflation. And I would expect that to be at a similar level or perhaps even lower maybe something closer to 3% in Q4. With regard to the contract labor usage, I would say the thing that surprised us the most in retrospect, maybe we should have seen it coming is that our clinical staff had a fairly high amount of accumulated PTO just based on the need for them to work additional shifts and the fact that for a big portion of the early stage of the pandemic it couldn't go anywhere and we have volume coming in, and they just started to take time off this summer. And so we had -- because the volume was so good in our facilities, again, adhering to the philosophy we'll take the volume and we'll figure out the labor situation later on. We used more contract labor in the third quarter than we had initially anticipated.

We'll take our next question from Andrew Mok with UBS.

Just wanted to ask a little bit about the framework for 2023. The midpoint of the revised guidance range implies about $223 million of EBITDA in Q4, which annualizes to about $392 million before considering any growth for next year. Are there any items that you would call out that impact that 4Q run rate? And if not, is that a reasonable way to think about 2023?

Well, Andrew, I know everybody wants to hear more about 2023, but -- you're going to hear from us much in worse here from some of the other provider reports really across industries. And that is -- there continues to be considerable volatility and uncertainty out there, which makes forecasting business trends beyond the immediate future very difficult. We remain very encouraged, as Mark alluded to by the validation of the underlying demand for our business, and it was indicated again in our volume growth for Q4. But trying to forecast, particularly with -- we're now -- 2 weeks after midterm election, you've got geopolitical risks that are out there, you got a whole host of things that make it difficult to say how much of Q4 is going to be representative of what we experienced in Q3. What we do expect is we expect that we'll continue to see good positive volume growth in 2023. We expect that we'll see good contributions from the 2022 de novos that we opened up as those continue to ramp up, and we expect that we'll continue to see improvement in labor expense, but forecasting the particular level is difficult. Unfortunately, it's really going to be with the provision of our 2023 guidance that will be in conjunction with our Q4 earnings report expected to take place in January before we give you a lot of concrete answers there.

Got it. Okay. But there's nothing you would call out as materially impacting the 4Q run rate at this distance?

No. I think anything we can there on the side of too much detail. But if you look at the fourth quarter, it's full year 2022, but there is only one quarter left. If you look at the guidance considerations that we enumerated, I think they're pretty thorough.

Got it. Okay. And then just a follow-up on the volume side, same-store growth was up 4% on a very difficult 3Q '21 comp. Mark, you made some comments in your prepared remarks that you're taking market share. Can you elaborate on what you're seeing and where those share gains are coming from? What are the primary drivers of those gains?

Yes. So I also comment on the quality outcomes in my prepared statements, and I attribute our ability to take market share from other post-acute providers. And if you look back, going back to 2020 with COVID, I mean, I think we really have proven ourselves very capable of taking on a higher acuity patient and doing a great job getting over 80% of them back come back to the community. So I think our continued commitment to quality, development of programs, specifically around stroke and other neuro continues to put us in a position to be the favorite post-acute provider in the marketplaces that we exist.

And in the quarter on a year-to-date basis, we continue to see discharge growth across all payer categories. If you were to look at our patient mix for the quarter, we're actually down a little bit in stroke and neurological, which are 2 of our larger categories because they are comprised of very medically complex patients. But that is because of what we had anticipated we would see all along, which is patient flows start to normalize, we're not losing volume in stroke and neurological. It's simply not growing to some of the other faster categories, which had been a little bit more absent during the early stages of the pandemic. So there's a lot of good news in both the payer mix and the patient mix underlying the support for that volume growth. Just to throw a couple of things out there as well. We've talked before just about how large and underpenetrated we believe the total addressable market is. And as we've mentioned in this context before, one of the proxies that we look at is CMS-13 eligible discharges coming out of the acute care facilities across the nation. And as a reminder, only 60% of the patients admitted into any particular IRF during the course of its fiscal year have to be CMS-13. Now we run at a higher compliance level than that, something in the low 70s. But even with that, only 13.5% of the CMS-13 eligible discharges in the nation who come out of a hospital wind up in an IRF. And we think we're making strides on a market-by-market basis at increasing that conversion rate. I do want to remind everybody as well that as you think about Q4, and we remain very optimistic about growth. But we're up against another tough comp because Q4 discharge growth last year was up 9.6%, including 6% on a same-store basis.

We'll take our next question from A.J. Rice with Credit Suisse.

Why don't we put A.J. back in the queue and move to the next one, operator.

We'll move to Pito Chickering with Deutsche Bank.

Just to ask Andrew's question in a different way. Historically, 4Q annualized has been a really good predictor about next year's EBITDA. And I understand that forecasting today is nearly impossible but all the volatility in the current macro environment, but assuming the macro doesn't change at all, is there any structurally wrong taking that number and annualizing again, with the caveat that things changed, and that's okay before you guys give formal guidance?

The only thing if you're trying to use Q4 as a proxy for the following year, and I'm not suggesting whether or not that's a good practice because I've already given you the caveat that there is a lot of volatility. You've got to remember a factor in seasonality, which is more pronounced in Q4 and Q1, and you would also need to factor in that you will have another year where we're opening up 8 de novos, and so there will be probably a fairly similar level, maybe at a lower level this year because some of those factors that I pointed out in Q3 and start-up and ramp-up costs. There's nothing else that comes to mind that I would suggest it was an anomaly. You do have the impact of sequestration when you're looking at a year-over-year basis, but it is fully implemented now beginning with Q3. So it depends on whether you're just trying to look at it sequentially or on a year-over-year basis.

Okay. Perfect. And then looking at building de novos, did the ROIC change at all with these news for building costs that they're embedded in these projects and the additional sign-in bonuses that needs to pay in order to ramp this up and then potentially slower Medicare certification due to staffing?

Yes. So to some extent, does when you increase the preopening and ramp-up costs. You can -- if you're calculating your ROIC, you can take those incremental losses and just fold it into the invested capital. But if you just do the math and add up the total capital that we were spending on the de novos this year, and the estimate beyond our expectations were up by about $5 million. That alone isn't really moving the needle at all on ROIC. And the fact of the matter is, once we do get them open beyond those costs, we've seen the ramp-up be faster than we expected. The larger issues are like everybody else in the country over the last several years, we've seen about a 25% to 30% increase in key components of our design and construction costs, mostly around materials. That is beginning to subside, which is good news. But when you combine the increase in construction cost with higher labor costs running through the P&L, it puts some pressure on the early years of the ROIC. That again has been largely offset by the more rapid increase in ADC we've seen with those facilities. And the other thing that Mark had mentioned, which is an important consideration is the use of prefabrication for us is holding costs steady right now. Ultimately, you think it result in savings, but it's accelerating our time to market, and that has a very favorable impact on the return characteristics of the de novos. So we may remain very excited about the terms we're seeing the returns we're forecasting with regard to this capital investment.

Pito, one of the things we're liking more -- the more we use, the more we like it is on the prefab is just the speed to market, which I alluded to in my comments, but just to give you some sense of how fast that is, the whole hospital, it cuts the time -- construction time from 9 months to 6 months. And then for bed additions, it reduces it by about half. So the speed to market is very important in these competitive marketplaces. We've specifically seen that in Florida and really like the direction we're heading with this prefab construction.

And those periods that Mark just mentioned related to construction specifically. There's also a design benefit to it. So the more you move towards prefabrication, the faster we can get the design for a new facility, retrofit into the land that we acquire, the faster it can get through the permitting process as well. So we're taking incremental time out of the project and on the front end as well. And so in aggregate, we feel like we've knocked 30% to 40% over the time to open up a facility. That's not going to be the case everywhere. Sometimes permitting and zoning is more difficult in other markets. But as a generalization, that's the kind of result we're seeing.

Okay. And then just really just a quick numbers question on the OpEx, you talked about increase in food costs, energy costs? And then if you can quantify this for us in the OpEx line and then what you can impact in 3Q and 4Q?

Versus our previous expectations, each of those categories was up just over $2 million in Q3.

We'll go ahead and take our next question from Pito Chickering with Jefferies.

I guess I'll just piggyback on the questions on the de novos. So as we think about -- you added a lot of beds from 2019, '20 and '21. But obviously, the pandemic was in the background. So maybe the ramp on those bed adds did not look as good as what you would have expected or should have been. So how are you thinking about the opportunity for a faster or greater-than-expected contribution as these de novos compound and ramp as we think about next year?

Well, it's there in the volume. And the ramp-up on those facilities even opened up during the pandemic has been very positive. And so the flow-through is simply being impacted by the elevated costs that we've talked about. But look at the year-over-year revenue increases, it's evident that those -- the new capacity that's added to those markets is resonating with payers and patients and referral sources, and so as we continue to make progress, now we've got a pricing increase as well. We probably didn't put enough emphasis on that. In Q3, we faced those increasing cost pressures against essentially a flat price increase. You're hearing about nice earnings gains in certain sectors like consumer brands, McDonald's was out today, yes, they're raising their prices because they've got the flexibility to do that. We had to meet these challenges against flat pricing. Mark alluded to the increase that we're getting in Q4. Again, that's going to be partially offset by sequestration, but it really begins to flow through as we enter 2023. And unless there are games that are going to get played in the rule-making process just based on the cost of inputs and you're seeing that across the provider universe right now, we should see another nice price increase in 2023 for the 2024 rule-making process. Then you're really going to see the operating leverage and the efficacy of these de novos and really our entire business start to flow through into EBITDA improvement.

Brian, we've brought on now 17 de novos in the last 2 years, and they've all been very successfully ramping up. And I think just as an indication of the talent that we have out in the field. Our operators that are largely responsible for bringing these hospitals online, staffing them and getting them up to speed. And I just want to do a call out to our teams and the regional presidents that are out there and the depth of talent that we have within this company.

It's like a lot of other things, the more you do, the better you get at it. And so all the aspects of the development process, we've shown improvement. Our development team is increasingly adept at evaluating good markets or adding an IRF, where we believe that a JV relationship would be beneficial. Those negotiations have been streamlined and every aspect of opening the facility and then ramping it up has improved. Now some of the -- we mentioned the delays that we experienced in Q3. Some of that -- all of it was beyond our control. I mean we had 1 facility where an air handler, which is absolutely critical to opening up the facility was delayed in the supply chain. Those things are going to resolve themselves. But by and large, we remain very pleased with the results we're getting out of the de novos.

Next question comes from Ben Hendrix with RBC Capital Markets.

Given the headwinds that you've seen in the past related to claim denial activity. Can you remind us what's involved of the Medicare intermediary TPE? And are there any proactive measures you can take from an operational or documentation perspective to at least partially mitigate future increases and denials?

Yes. So we're definitely seeing increased denials and you see that in the information is included in the supplemental slide at the resumption of TPE. The best thing that we ever did to try to move away from the traditional ADR issues we had was the implementation and utilization of our clinical information system, which is very effective at capturing documentation, and we continue to look for ways to improve that. The biggest difficulty that we're facing right now with the resumption of TPE is that there are radically different interpretations by the intermediaries of Medicare guidelines around patient care requirements. And I mean just as an example, the CMS guidelines require or suggest because it's a guideline that the preponderance of therapy must be administered in a nongroup or a nonindividual setting. And so I think the average person off the street would interpret preponderances as meaning something like a majority. As a matter of fact, if you look it up in the dictionary, it's pretty much what it is. But we have certain TPEs in certain markets where they have looked at a claim that we have submitted where the patient received more than 80% of their therapy on an individual basis and did not decline because it didn't meet the preponderance test. And so we're having to take an increasing number of these denials, and they're not even at alarming level right now through various appeal levels, and it's just -- it's frustrating. We're providing top-quality care to these patients. We're generating great outcomes as Mark indicated earlier. And it's resonating with the patients because our patient satisfaction scores are at all-time highs, and yet you've got uninformed arbitrary decisions being made about claims that we filed.

So Ben, we're educating the auditors on the requirements for -- by CMS for IRFs versus them educating us.

Do you get any -- is there any additional risk? Or is this -- are these issues exacerbated in any way by -- from like a documentation perspective, by the use of an increased number of agency nurses or nurses you may not be familiar with our systems or coming in from other settings, does that create any kind of additional documentation headwind?

Well, it certainly adds the [indiscernible] who'll do that. We're careful to make sure that we educate the agency nurses on the use of our electronic medical record system. That's part of the basic orientation even for an agency nurse. So we do everything we can to prevent that from happening, but that is certainly a risk for organizations that don't take the time on the upfront period with the contract staff.

I would also say the key nature of how we approach care in our hospitals means that a contract labor FTE is surrounded by permanent -- more permanent employees, which kind of helps with that. And then also just because our skilled labor force is comprised of nurses and therapists, there's a hierarchy in every hospital to our nursing structure. And so their supervision of everybody who's in the facility, not just contract labor nurses.

[Operator Instructions] We next move to Matt Larew William Blair.

I wanted to ask a bit about the salaries and benefits trend here. So in the first half, salaries and benefits per FTE, the increase was about 8% in the back half, it's going to be about 3%, and you've walked through kind of a number in moving parts as to why that is improving sequentially. Is it fair to assume that kind of the fact that the second half of the year trend is more emblematic of the go-forward environment? And as a part of that, you've now had agency labor down -- the rate down 3 straight quarters sequentially, but still 30% higher than 2020, which I presume was meaningfully higher than pre-COVID. So is that a bucket that you expect to continue trending down as well?

It is, Matt. And I would just say that there is some uncertainty regarding Q1 for some of the factors I identified earlier, which is that's normally our heavy volume year. And it's also the quarter which historically has been most impacted by a flu season. And with over the last several years, I forget how many years we've been in COVID now, 2 or 3, its been impacted by winter surge in COVID. And so those could play a role in Q1, but we do expect significant further improvement in overall labor cost through 2023. As we sit here today, the fourth quarter is going to be very important for all of us with regard to that.

Okay. Got it. And then you alluded to the shifts on stroke and neuro just as some other patient normalization. But maybe just give us an overall update on how patient acuity trended through the quarter? You mentioned, I think, last quarter, the quarter before, you're seeing some ortho patients come back. But obviously, you have a 4% price increase started 10 1, but from a patient acuity standpoint, maybe how do you expect that to evolve in Q4 and beyond?

Overall, our acuity was consistent and what has been sequentially in the previous quarters. I mean we may have some shifting among the diagnostic categories. But overall, our acuity was relatively flat.

Yes, across all pairs, it was almost exactly flat. I commented that there wasn't a big move, but our Q3 pricing across all payers actually came out slightly lower than we had anticipated. And the culprit was less around the patient mix and more around the payer mix. I mentioned earlier that we saw discharge growth across all payers in the quarter. And even though it's a small piece of our business, the most rapid growth that we had in the quarter was in Medicaid. And whereas we've been able to close the gap on Medicare Advantage to less than 5% of the discount to fee-for-service. Medicaid on average remains at about a 35% discount. So that was just another factor. I don't necessarily believe that that's going to continue at that level and typically not enough to call out as we move into Q4 and beyond. But the acuity is staying up there nicely.

Next question goes to Sarah James with Barclays.

I'm trying to piece together the comments you guys made earlier with the 25% to 35% rise in construction costs for key components versus the savings on the prefab. And I'm wondering if the net of those 2 change your preference on de novo versus in-facility bed expansion? And does the increased use of the prefab change your strategy at all for building excess footage into your new facilities?

Yes. So those are good questions. One is bed expansions where they're justified by the occupancy -- existing occupancy rate in an existing facility. And where there aren't limitations either related to CON or the physical structure of the building are always at the top of our prioritization list with regard to CapEx. We've mentioned before, you're building into an existing demand, the referral sources and the payer contracts are established. You've already been marketing out there into the market. So there's familiarity and you're leveraging key components of the stack and the infrastructure, and so the returns are very high. And so that will always be at the top of the list. The second question was extremely insightful and you're on target there, which is one of the things that we have been doing is -- whereas if you roll the clock back, maybe just as recently as 3 or 4 years ago, we are entering a lot of markets with a 40-bed, all private room facility. Now we're pushing up towards 50 or 60. And we're doing so because of 2 things. One is because of the very rapid increase we've been seeing in the ramp-up of our new facilities and how we have this end of the learning curve with regard to our efficiency with operating those. But the second is the point that you've made, which is doing that on the front end allows us to bring down the average cost per bed. And so that is the strategy we will likely use. I should note, this is kind of thinking out down the road, the more we push from a 40-bed chassis at opening to a 50 or 60 bed, the more that might cause some lumpiness on future period bed expansions because some of the bed expansion numbers that we've had here recently, have been driven by the fact that we've seen a lot of de novos opened that within 2 to 3 years, we're already requiring good expansions. As a matter of fact, we have some of the recent vintage, North Campo comes to mind as an example, where that's occurred really fast.

Great. And just to follow up on that, the lower cost per bed that you're able to achieve, is there any way that you can frame that [savings upfront]?

Right now, what's really allowing us to do is kind of hold the line versus further increases. We do think with some of the strategies with regard to speed to market and increased fabrication, the next gen that Mark alluded to in his comments that we're going to be talking more about in Q1 that we're actually going to bend that curve back around. And that will happen -- so it's going to be that strategy combined with improvements that we're already seeing in the supply chain and then what hopefully will be a somewhat of a decrease in construction labor costs as well.

And ladies and gentlemen, that does conclude today's presentation. I will turn the conference back over to Mr. Mark Miller for any closing remarks.

Thank you, operator. If anyone has additional questions, please call me at (205) 970-5860. Thank you again for joining today's call.

Ladies and gentlemen, that concludes today's program. You may disconnect at any time.