Hello, and a warm welcome to today's agilon health Third Quarter 2022 Earnings Call. My name is Candice, and I will be your moderator on today's call. [Operator Instructions] I would now like to pass the conference over to our host, Matthew Gillmor, Vice President of Investor Relations. Please go ahead.

Thank you, operator. Good evening, and welcome to the call. With me is our CEO, Steve Sell; and our CFO, Tim Bensley. Following prepared remarks from Steve and Tim, we will conduct a Q&A session. Before we begin, I'd like to remind you that our remarks and responses to questions may include forward-looking statements. Actual results may differ materially from those stated or implied by forward-looking statements due to risks and uncertainties associated with our business. These risks and uncertainties are discussed in our SEC filings. Please note that we assume no obligation to update any forward-looking statements. Additionally, certain financial measures we will discuss on this call are non-GAAP financial measures. We believe that providing these measures helps investors gain a better and more complete understanding of our financial results and is consistent with how management views our financial results. A reconciliation of these non-GAAP financial measures to the most comparable GAAP measure is available in the earnings press release and Form 8-K filed with the SEC. With that, I'll turn the call over to Steve.

Thanks, Matt. Good evening, and thanks for joining us. Momentum across our business remains strong. And the success of our aligned partnership model is enabling agilon to make rapid progress against our vision to transform health care in communities by empowering primary care doctors. agilon's platform, partnership model and network create the local infrastructure that enables physician organizations to access a new and sustainable model for primary care. One that aligns physician outcomes with improvements in the quality and experience of their senior patients. As a result, patients are receiving higher quality, comprehensive care from their trusted physician. And primary care doctors are enjoying greater satisfaction as they share in the value created from a more effective approach to care delivery. In 2023, our national network will include 2,200 primary care physicians and 500,000 senior patients across 12 states, making agilon one of the largest organizations in the country, supporting full risk value-based care for senior patients. Our growth is a function of the value that agilon's platform delivers to physicians, patients and communities and the value we are delivering continues to expand. Year-to-date, we have reinvested over $130 million back into local communities through surplus sharing from reductions in wasteful spending and quality incentives to doctors for spending more time with their most complex patients. This reinvestment is helping to sustain and grow access to primary care and transform health care delivery. Turning to our results for the quarter. Membership, revenue and medical margins were above the high end of our guidance ranges, and adjusted EBITDA was in line even with a modest drag from direct contracting. Our core Medicare Advantage business performed better than forecasted on all key measures. Year-over-year, MA membership grew 45%, medical margin increased 74%, and medical margin PMPM increased 19%. This reflects strong performance…

Thanks, Steve, and good evening, everyone. I'll review highlights from our third quarter results and our guidance for the full year 2022. Starting with our membership growth for the third quarter. Consolidated Medicare Advantage membership increased 45% to approximately 267,000 above the high end of our guidance range. Direct contracting membership increased 71% to approximately 89,000. Total members live on the agilon platform, including both Medicare Advantage and direct contracting increased to 356,000. Our Medicare Advantage membership growth was driven by the addition of 6 new geographies in January and 14% same geography growth in existing markets. Direct contracting growth was also driven by the addition of 4 markets that joined the program in January as well as growth within existing markets. Revenues increased 52% on a year-over-year basis to $695 million during the third quarter. Year-to-date, revenues increased 47% to $2.02 billion. Revenue growth was primarily driven by MA membership gains from our new and existing geographies. Third quarter revenue included a modest benefit associated with member retroactivity. On a per member, per month basis, or PMPM, revenue increased 3% during the quarter, which primarily reflects benchmark updates, market and member mix, along with year-to-date true-ups with health plans. Medical margin increased 74% year-over-year to $76 million during the third quarter. Year-to-date, medical margin increased 62% to $244 million. Even with the dilution from our membership growth, along with a higher proportion of members in year-1 markets compared to last year, medical margins increased as a percentage of revenue and on a PMPM basis. Medical margins were 10.9% of revenue during the third quarter compared to 9.5% last year and medical margin PMPM increased 19% to $93 compared to $79 last year. Medical margin growth was primarily driven by the maturation of our year-2 plus partner markets. For these…

[Operator Instructions] So, our first question comes from the line of Lisa Gill of JP Morgan.

Just really 2 areas I wanted to start with and one would be just around medical costs as we go into the fourth quarter. You talked about the third quarter coming in roughly in line with your expectation in-patient below, outpatient in line. How do we think about how you're thinking about the trend into the fourth quarter? And will flu play any part of this? I mean we're kind of watching the Southern Hemisphere and different parties are saying different things around flu. So that would be my first question. And then I just had a quick follow-up.

Yes, Lisa, I'll start. I think that from a medical cost perspective, we feel like our high-touch model is really helping us manage within kind of the ranges that Tim talked about. We are seeing a real uptake in terms of flu shots in this period, similarly to how we saw high vaccination rates around COVID boosters. And so we think that, that is going to help us. As we've thought about kind of our guidance, I think our -- we're looking at sort of a regular flu season and what we've laid out through Q4 based on those dynamics. So Tim, anything to add?

Yes. Lisa, thanks for the question. The only other thing I would say is we have seen kind of on a year-to-date basis, a pretty nice pickup year-over-year and however you want to look at this, either on medical cost, on an MLR basis. I think year-to-date, we're up like 110 basis points or something year-over-year or on a PMPM basis. We talked about our medical margin PMPM, obviously, being up significantly year-over-year as well. As we move into the fourth quarter, I think everybody understands in our model now that the absolute numbers fall off a little bit as we move through the year as our mix of members gets less profitable, more agents as well as kind of the older patients falling off that are the most profitable. But in terms of improvement versus a year ago, we continue to expect to see the same kind of trends and improvement. And you can see that in the -- if you look at the midpoint of our guidance for the fourth quarter or the full year.

That's really helpful. And then just as a quick follow-up, I just want to go back, Steve, to your comment on the renewal of the payer relationships being on track or in line with expectations. Is there anything that's new or unusual as to what payers are looking for? And when we think about capitated rates, are they roughly in line with what we've seen historically? Just any incremental color would be really helpful.

Yes. Lisa, I mean, I think we're extremely pleased with the renewal season that we've just come through and we're finalizing the new contracts for the groups that will go live on January 1. But all in, it's been a very successful season. I think that what we're finding is payers see extreme value in working with our partners. They are scaled within their regions. They have excellent retention and management with their folks and payers see real value within that. And so I think when we think about the payer renewals, what we've heard from them is they're really valuing the scale that our partners are bringing the quality. I talked about the 4 and 5 stars that we're seeing, and I think that will serve us very well. And then just the retention, a key thing with the senior population is the year-in, year-out management of complex conditions. And so the ability to have strong retention really helps around that. So we're first to value in these communities and move them that way and payers see value in that, but all of the things we provide along the way.

Our next question comes from the line of Justin Lake of Wolfe Research.

I'm going to apologize upfront guys. I'm going to ask you about RADV and how this might impact you. So the -- I know it's a painful topic, but getting a lot of questions on it. So maybe first, you can tell us the 2 questions I get the most often, how do your contracts work in terms of if there is an audit and there is some payment back to the plan or the plan has to pay CMS, I should say, if it's for your patient or your patients or a certain percentage of the panel, would you be on the hook for some percentage of that? And then secondly, we all know that you guys do a good job on coding. Do you have any feedback from the plans in terms of how -- the plans are doing their own audits of their population, right, just to see how they're doing? Do you have any feel for how you're performing in those audits relative to their more generalized population?

Yes. Sure, Justin. So on the first one, for our contracts, if there was sort of a retro adjustment back on that payback, we would be on the hook for that. I think the periods that are being looked at are earlier than -- except for the very earliest market. So right now, the window we're in, there's probably not as much of an issue on that. But obviously, as it would roll forward as they roll these audits forward every year, that's how it would work if there was a claw back, as you said, on that. And then feedback from plans on the audits. I think we work very closely with them on those audits. I think they feel like the information that we're providing to them and sort of the back and forth in terms of what we go through before they actually do that submission is a fairly tight process. So that's how I'd answer both those questions.

Yes. And Justin, the only thing that I would add is, I think we've said this a few times on previous calls, we do risk adjustment, we believe, the right way. We have a very tight process at the end of -- and almost all of our risk adjustment is done in the office through annual wellness visits or through the provider themselves through the PCP. At the end of the day, those risk adjustment items have to be -- the conditions themselves, of course, are validated and agreed to by the primary care physician that sees the patient. And when all -- and we do it the right way in terms then of both when there are new conditions, we obviously capture them, and that's not only important for risk adjustment, but also important for the plan of the patient. But also, of course, as conditions fall off, at the least we handle that very well. And I think all of that shows up in the fact that when you look at our overall average risk adjustment score, our average wrap it's not really meaningful above a 1. I mean we're not really pushing the envelope in terms of having really out of balance risk adjustment. So I think across the board, we feel pretty comfortable also that we just have a very tight process.

And Justin, I'll just add, I think it's really important as part of our partnership that we do, do this the right way. And we have a very tight process. To Tim's point, we delete codes as well as add codes based on a constant review process. So I think integrity around this process is something that we pride ourselves on, our partners pride themselves on working with somebody that does this in the right way. And so we feel good about our approach.

And if I could just go one layer deeper, Steve, on your contract structure. I just want to make sure I understand it from the perspective of the -- we know that they're only auditing a few hundred, a couple of hundred members in a contract, right? So the likelihood that they're actually auditing your member becomes infinitesimal, right, especially like a point, given your size at the moment. So how does it work in terms of -- is it just if there's 100,000 patients and your patients have to be -- happen to be I'm just going to pick a number, 1,000, right? So 1% of the membership are sitting at the contract is in -- if you have 1,000 patients out of 100,000, do you have to pay 1% of whatever they're asked for? Is that the way it works?

Yes. So I think without getting into the specifics of each of the individual contracts or each of the individual ways that this could come out. I mean generally speaking, if there is a RADV audit and there is a direct line that we can say, hey, that RADV audit applies to our members that we have cap data from that payer then, yes, I mean, we're -- ultimately, we're going to be on the hook for our portion of that. Having said that, depending on what the actual content of the audit is and how that comes out and how we talk to the payer about it, we'll see on a case-by-case basis on how it works out. But obviously, if there is demonstrative line of sight to, yes, they should apply to us as part of their population then we'd be on the hook for it.

Our next question comes from the line of Kevin Fischbeck of Bank of America.

This is Adam Ron on for Kevin. I saw an interesting release in the quarter that one of your anchor partners in Ohio started doing an on-site clinic with JPMorgan and some sort of value-based construct arrangement. And given the interest from such a large client to do value-based care outside of Medicare, I'm curious why not just support them yourselves? Or I don't know if there's interest from such a large client, if that changes your view maybe about expanding beyond Medicare?

So thanks, Adam. Yes, the -- very familiar with -- it's Central Ohio Primary Care, our partner in Columbus and JPMorgan, who are doing their work on this together. They have another partner who works with them on that, that we know very well. And they're doing an excellent job around that. I think our focus has really been focusing on that over 65 population doing great work on that and really solving a major challenge for our groups. The optionality, obviously, is there down the road for that commercial business. But today, we feel very good about the focus on the Medicare population. And there's just to what we shared in the call, there's tremendous value and impact that we can have by focusing on that population.

If I could just ask one quick follow-up. The direct contracting, the ongoing refinement, that was related to what happened last quarter kind of on the retro trend. And then is anything -- does any of your conversations with CMS changing how you view, I guess, like the sustainability of direct contracting?

Yes. Well, let me answer the macro and then Tim can kind of walk through the mechanics because we literally get monthly updates on this. And so you're constantly sort of refining that. But one is, I think, we believe direct contracting is extremely strategic for our partners and for us in that there's a single experience for their primary care physicians, and there's one care model that gets applied across the entire senior population, which reduces variability and allows for much better care and less sort of variation. The second would be were profitable year-to-date. As Tim said, we have [$6 million] of EBITDA. So that program is still profitable even given the adjustments that Tim talked about. And then the third that I tried to call out in my prepared remarks is we are beating national benchmarks in terms of utilization trend. If we continue to do that, that should drive increased profitability over time. And that's sort of the construct in which the program is laid out. We're working with them on adjustments that can be made to the program to take out some of the volatility, to improve the predictability and the sustainability, but that's sort of the macro view. Tim, do you want to talk about?

Yes. And Adam, not to belabor it, but I think everybody understands that the way this works is CMMI puts out an expected trend number at the beginning of the year. And after the year is over, they will say, "Hey, based on the actual observed trend for the full Medicare fee-for-service population," they will adjust that up or down depending on what they see. They provide interim updates during the year on how that's going. They don't actually tell you what the retro trend adjustment is going to be, but they do tell you how trend is looking at various points in the year. And what you're referring to last quarter was the first update came out in May and there was a pretty substantial indication that trend was way lower than what they had put out initially like as much as 8 points lower. And so a lot of people in the industry at that point, you may remember saying, "Wow, we just had to take a big adjustment to revenue in our second quarter results because of that." Now you remember at the beginning of the year, we had said, "Hey, we anticipated there would be a pretty sizable retro trend adjustment because our analysis said that the trend that CMMI was guiding to at the beginning of the year was very aggressive or high to begin with." So when that second quarter update came out in May or the first quarter update that came out in the second quarter in May. We had to take a small adjustment last quarter, but it wasn't anywhere near probably what a lot of other people in the industry were taking if they hadn't already pre-assumed a retro trend adjustment. Now another update came out in August. And then a…

Our next question comes from the line of Ryan Daniels of William Blair.

I wanted to talk about the new partner pipeline with a focus on '24. Great to hear you added 2 more to the class. I'm curious, number one, if you can discuss what this looked like a year ago for the forward year basis? So is this an acceleration, I think, from what you've seen? And then number two, you talked a little bit about macro trends driving this, but I'm curious if you think the pending U.S. recession or global recession is actually helping your partner pipeline because providers can move to more recurring revenue models and see more income upside with your model than they otherwise would. So is this type of environment actually beneficial for you relative to kind of a more stable market?

Yes, Ryan. Well, I think the headline is there is a tremendous inflection in demand. We've been seeing it for a while. It is accelerating. I think it's a combination of macro, CMS pushing more towards full risk value and looking by 2030 to have all seniors in a total care relationship with a PCP, individual health plans, pushing on that in a really significant way. But then I think the success that we're having for virtually every medical group in the country, you can find an agilon partner that kind of looks like you, is organized like you. And so that referenceability is just a huge asset for us. Two is, I think, that more and more of these groups have come to the conclusion that they really need to make this move into full risk value-based care. And so the combination of the success and the desire is really shortening the sales cycle. And so to your question about to compare it to a year ago, we are well ahead to have 4 groups signed to have implementations that will be greater than 12 months, that puts us in really great stead for 2024 and how these groups should start. And so that's really encouraging to us. The fact that you have a mix of primary care only, multi-specialty and distributed networks, partners within that is also really encouraging to us. And as I said in my prepared remarks, health systems remain very actively engaged with us in talking about partnerships. So I think what we've done with MaineHealth, which is going really well. I think the ability for people to be able to pick up the phone and talk with them about that experience really helps. And I think a number of health systems are feeling the need to make this move into value and get the benefits from that. And so all of that is sort of leading to the momentum that we're seeing and sort of being ahead of where we've been in prior years in the cycle.

Okay. That's very helpful. And then, Tim, one for you. I should probably know this nuance, but -- if we look back to the end of last quarter, I think you had about 261,000 MA lives a little bit above and ended this quarter at 266.6. But you referenced the average being over 270,000. So what's the nuance there that the average for the quarter is so much higher than the starting and ending period?

Yes, any time that you see that, Ryan, what's going on is we have some retro members that we've got attributed to us or we have members that have been attributed to us that really should have been attributed as our members from the beginning of the year or near the beginning of the year. And in this quarter, it was about 1,000 or so members. So 3,000 members kind of above the ending time period, but really retro back over 3 full quarters gives you about 3,000 average members higher is the way that it works out. I mean it's really an important part of our model that we're able to do that. Essentially, if you think about this, there's 2 different ways that we have members attributed. One is through members that are in an HMO couldn't be more straightforward, right? They have to actually name who their PCP is officially. On the PPO side, which is now more than half of our business, that's often not the case or generally is not the case. And so we have to go through other attribution methods that we have worked out with the payers. And often cases, it's related to how many times has that patient seen the PCP in the last 6 months, 12 months or 18 months. We often have backup plans where we do things like recorded phone calls with the member to make sure that that's their PCP, et cetera. But we have a really good established process to do that. But what happens is, at some point, we tend to pick up members that maybe should have been attributed to us early and now we catch later in the year. That's particularly true of new markets. And in fact, of the, let's say, 1,000 or so retro members that we picked up driving that difference in the quarter. I think the majority of them were actually in a couple of our newer markets, with newer payers.

Thank you so much, very helpful.

And by the way, while we're switching to the next question, I'd just throw in, it's really important that we are able to do that in that half of our business -- being more than half of our business being in PPO membership is a big deal because that means we can essentially go into a market and bring -- and see if you can comment on this better than me, bring full risk to a market regardless of whether it's heavy PPO or HMO, unlike other markets that may be very heavy HMO that are very easy to get that attribution. Our TAM is wide open for PPO, HMO. We can go basically anywhere and bring full risk across that membership.

Yes, I think the feedback from the national payers is that we're fairly unique in our ability to take full risk on a PPO product. And that's the fastest-growing product. It's roughly half. It will be more here as you see further growth. And so we think we're set up well.

Our next question comes from the line of Whit Mayo from SVB Securities.

Yes. First question just on STARS. I mean there's been a lot of noise about certain plans going from 4 to 3.5. Maybe just frame how you guys are thinking about it and do your contracts have any contingency provisions that may protect you in the event that an H contract does slip to a 3.5 star rating.

Yes. I mean let me start with by saying what I said in my prepared remarks, which is our own performance is extremely strong. 4 stars across the board and 5 stars in the areas that we talked about. Two is there's obviously been this step down across the industry as some of the COVID provisions expired and the better of provision that went with that. We're doing better than the industry average in terms of percent of members that are in 4-plus star plans, and that's a result of us really managing this extremely well. We do have one payer in particular that we have a decent chunk of membership with that has stepped down to 3.5 stars for the majority of theirs. We believe that, that's very manageable for us as we look at that in terms of what that could look like in terms of an impact from a PMPM perspective kind of low single digits in terms of -- for 2024. So I think we believe it's really manageable. We're performing better than others across the industry. I think health plans would like to see more senior patients in agilon partners because of the strong quality performance. But specifically to your question, I think it's a very manageable impact for us.

Okay. Steve, you talked in your prepared comments, maybe this is new, maybe it's not, but I feel like you referenced diabetes, renal, maybe more in the context of perhaps a specialty program. And I think you've established a partnership with Monogram, not sure how new that is. But is there anything to elaborate as you kind of think about this diabetes, renal specialty program?

Yes. I mean we have a really strong partnership. I think we're in 6 or 7 markets to date, and I think we're expanding to 2 or 3 more by the end of the year. I think our early results in those programs are really exceptional. And in particular, the enrollment that we're seeing from patients because it is an opt-in is, in the mid-80s. I think that's a function of this tight relationship that we have between the primary care physician and the patient and when the PCP makes that recommendation, the patient is very likely to agree with that. That is demonstrably different than what they've seen in other markets by working with payers. And so I think it just kind of speaks to the secret sauce that we've got on that. And I just think -- I called out sort of our strengths in terms of A1C control and the benefits that we've got around that which really is extremely strong. So that's what I'd call out.

Okay. Can I just ask one quick one for Tim. Just these new territory costs, geography costs. Is this all 2023 go-lives or the implementations for 2024 too?

It's almost very, very high 90-ish percent 2023 still at this point. We are now just getting into starting to implement 2024. We haven't put in place a lot of infrastructure or cost for that. We will have some in the fourth quarter that start to flow in because we are getting up and going with our implementation for 2024. But for right now, it's primarily going to be 2023 implementation costs. And remember, of course, very big class for 2023 that we're -- very big complex class that we're implementing for 2023.

Our next question comes from the line of Sean Dodge of RBC Capital Markets.

On margins for new year-1 classes, Steve, you mentioned the platform is getting smarter and more efficient. You pointed to the current year-1 members trending toward the high end of your targets for medical margins. You're also building out classes now than you have historically. You referenced the 4 already for 2024, which gives you more time to prepare for their launches. I guess when we think about the combination of the 2, I think you alluded to it a couple of times, but can you maybe put some bookends around how much you think these can help elevate the launch trajectories for margins for future classes?

Yes. No, I appreciate the question. So I think to your point, we are experiencing the benefits of learning from our platform that is allowing our newer partners to perform at the high end of our range, which is really fantastic and encouraging and they're able to get the advantage of some of the programs like we just talked about, complex kidney disease as an example, earlier in the life cycle. So I think that's kind of point one. Point two is, every class is a little bit different in terms of where they start, did they have an ACO before. And so the starting points for each one of those is a little bit different and always takes us a little while with payer contracts and others as we understand what that starting point is. But I think in general, we figure like -- we feel like we're seeing an acceleration sooner for our year-1 markets in terms of the benefits of the high-touch model, and it's coming through in terms of better satisfaction, better health outcomes and ultimately lower costs and better margins overall. So those are the things that I would really call out.

And one thing, Steve, maybe I would add, as we said for the last couple of years that we think the best time to look at that, although we can give you an indication, obviously, that as Steve did and I did in our comment that our year-1 markets are actually performing better than we thought they would this year, certainly at the high end of our expectation. The best time for us to talk about that and talk about that trajectory is when we have a full year of results. We did that at our Analyst Day and showed you -- you actually saw it last year with some of our early markets were actually some of the best performing and fastest trajectory and obviously, when we come back, once we have a full year results this year, we'll show you updated cohorts and how that's working as well.

Yes. And then just one last point I would make. The fact that for this class of '24 that we're talking about that we're this early, and we're going to have that long of an implementation period, they should start in a very strong position as a result of that.

Okay. That's helpful. And then as we think about the runway for medical margin in some of the older cohorts, you guys have talked before about there being a significant amount of other impactable spending that you could start to address. I think you sized it at [$98] per member per month. What are -- what's kind of the biggest bucket there? And I guess, have you started to make some inroads in trying to tackle some of those cost opportunities?

Yes. I mean it's stratifying the population and dealing with those most complex patients. And really, it's -- I mean it's not hard to understand, but it's being able to maintain a multi-chronic and have them spend less time crashing into the emergency room and less time in an inpatient setting. It is moving to more time in the home, and I talked about that the home-based teams that we've got, I think, is a tremendous opportunity for us. COVID has really shifted kind of the site of care in terms of what senior patients are comfortable with around that. You've seen it from inpatient to SNF and now much more to home. And so I think those are the things that we can really go after. And then the other would be really on the drug side in terms of medication adherence and just making sure that you're substituting sort of appropriate therapies. So those would be the big buckets that I think we'll go after.

Sean, as a good example, Steve, you might expand on this. These markets are all really early in their life cycle of moving to full risk. If you talk to any of our partners, they'll just say there's a huge amount of opportunity in my market, even in the most successful partners that we have, but there was this Akron Summit the other week. I think it's just a great example of how the market evolves.

Yes. I mean now Akron is one of our more mature markets, and we have really exceptional -- 2 exceptional partners within that market that have a really meaningful share of the adult primary care capacity, which is one of those real keys to success, but they were able to bring in leaders from more than 100 specialty groups on a Saturday morning. Everybody showed up and really talked to them about, hey, we're not making the move in the full risk value-based care, we're there. And we want you to come with us, but we need a few things in order for that to be possible. If you want to stay in kind of a preferred network tier, you need to share quality and efficiency metrics. We need you to ensure access with expedited appointments. We need to make -- have you used technology to ensure that patient visits are actually getting completed. And then we need all care decisions coming back to this primary care physician. So it's really kind of the shared partnership. And there was tremendous amount of embrace around that and excitement around that. And you can just see that market really beginning to change. And you talk about just scratching the surface in terms of what we could impact from a specialty and facility cost perspective. Once you get -- this is a primary care only group, both of them to have the specialists that engaged around that is very exciting. And so I think we can see this happening in more and more markets as we build the scale. We always are building around those right partners and -- it was just a great sort of evidence of what we're trying to do.

Our next question comes from the line of Brian Tanquilut from Jefferies.

This is Taji Phillips on for Brian. So as it relates to your 2022 guidance, just curious for Q4, I noticed in your guidance that you raised the 4 for the full year, but also mentioned that you didn't account for upside from direct contracting. So I just want to understand what's informing, I think, the raised floor for your 2022 guidance and if there's anything that we're missing, particularly for modeling purposes.

Yes. Taji, I don't think you're necessarily missing anything. I mean I think the way this works is the midpoint of the range for adjusted EBITDA guidance is relatively constant rate. We're going -- went from 0 to 10 with a 5 midpoint to 2 to 7, so I guess a $4.5 million positive midpoint. The ins and outs matter, basically, we are absolutely seeing better performance from our MA business, and you see that flowing through to the medical margin numbers, and we kind of upped the midpoint of our medical margin MA guidance by about $6 million, so you expect? That's actually probably -- that's helping us on a full year basis by about $3 million. You can see that flowing through from Q3 and an expectation for a decent Q4 as well. The flip side to that, and the reason we took the top end of the range down is a little bit more caution around direct contracting, also the fact that we just obviously booked a $3 million loss for direct contracting in the third quarter. We don't expect that to be contributing -- and that's a retro adjustment, obviously. We don't expect that to be positive in the fourth quarter, but we don't expect it to be dilutive either. I think the common -- and then we're probably seeing a little bit better performance, a little bit more leverage out of our platform support costs, which is a huge component of our model. When you put all that together, we want to be a little cautious on the retro trend adjustments so we brought down the top end of the range. But certainly, the really, really strong performance in MA. It's giving us more confidence that we can tighten that up and bring up the bottom end of the range as well.

Our next question comes from the line of Stephen Baxter of Wells Fargo.

I just wanted to ask another one about direct contracting. I appreciate you guys have taken a cautious approach to how you forecast and accrue for the business. I think you said you're engaged with CMS to try to create increased visibility and predictability in the program. I guess what exactly would you be looking to see to achieve that outcome? And then what would the forum for any of those changes potentially to be made at some point down the road?

Sure. Thanks, Stephen. I appreciate the question. We have a great partnership with the Innovation Center. These are pilot programs. And so they typically have adjustments that are made each year. And so we, in concert with a coalition through APG, have been talking to them about potential adjustments that could be made within it that would create more of the stability and predictability. I mean it's a fairly technical calculation, but something that would smooth the revenue balance, which really is what's happening with the retro trend adjustment that Tim talked about. There's like 3 or 4 factors that affect what that revenue number turns out to be. And so there's a lot of modeling going on and a lot of work around that. But I mean, I guess I would leave it there, but I think they are actively engaged. I think they too are surprised by the volatility that's occurring and it's really a function of coming out of this, hopefully, once in a generation COVID-type experience. And so that really swings when you start to do year-over-year and baseline year comparisons, you can see pretty significant adjustments around that. And so that's what we're working with them on as part of a larger coalition.

Our next question comes from the line of James (sic) [Jamie] Perse of Goldman Sachs.

First one quick clarification. The upside on medical margin this quarter versus your guidance. Tim, you mentioned there was a positive retro adjustment on MA. Can you quantify what the benefit there was or if all of the upside was underlying performance?

No, there's no retro trend adjustment or anything like that, that applies to the MA business. That's specifically just related to the direct contracting business. The MA performance overall is just improving in the fourth quarter based on the overall improvement in the model. I mean there are a lot of adjustments that you do in the third quarter. It's a quarter where we have the most data flowing in, in the year from our payers. So we made all the adjustments that we have on the -- from the payers to make sure we got the right mix of members and the right bid rates and the right [RAF] assumptions in there. And so all that helps. We actually got some help, obviously, in the quarter in terms of incremental medical margin dollars because of the retro members and being the top end of our membership range as well. And then on the flip side, the same thing, we did all the updates that we got from the payer data, obviously, on cost as well. But the net of that, yes, is definitely demonstrating a stronger medical margin performance in dollars on a PMPM basis, on an MLR basis, however you want to look at it. And I think that just reflects the continuing strength of the model to drive positive outcomes. But there's no formulaic retro trend adjustment or anything like that on the MA side.

Yes. I meant the patient attribution piece, but I think you...

Yes. So on the patient attribution side, we picked up a -- yes, we picked up about 1,000 members, I think, retro. That's not really an adjustment. It's just, hey, these are members that are seeing our PCPs. They should be attributed to us. We've identified them and worked out with the various health plans that they should be. It's typically that starts to wind down as you get further through the year. But later in the year, we still have some of our new plans in our new markets that this whole attribution process is new to. And so most of those 1,000 members are across some of those payers in those markets. But yes, so that was a pickup of essentially the difference between average membership and ending membership as those retro members.

Okay. And then just on MaineHealth, how is the integration going? And do you feel like you're ready for next year? You mentioned that's a complex one and different from your historical partnership. So anything to call out there just in terms of how you're going and how you're feeling in terms of getting that ready for year 1? And any economic considerations you should factor into our models versus your typical year-1 performance for traditional markets?

Yes, well, first off, I'll tell you, it's going incredibly well. The engagement from the MaineHealth medical group who is our partner there, and Andy and the entire team, who's the CEO of the system has been first rate. I think we've been able to integrate very well with their EMR. I think the payer contracting is going extremely well. It takes a little while as you go through these to sort of finalize exactly where you're going to start. And they are a extremely large -- they're a very large group across the entire state. And so I think I would say it's going very well. The engagement from them is quite strong. I think the integration with their electronic medical record is going to help quite a bit. They have extensive care team resources that are available that can really sort of help to drive performance over time. So I think we feel very good about that, and we'll update on kind of the class of '23 and what that looks like in terms of a starting point as we get a little further on our progression. But right now, I think we believe it will be within our historical ranges and should be good.

Our last question comes from Gary Taylor of Cowen.

Just coming in at the finish here. A couple of questions. One, you beat your revenue guidance by $45 million to $50 million, and I just wanted to understand the components of that. I think enrollment was at the high end, but not materially above. I think the retro attribution was maybe $8 million or $9 million. So I guess most of it was per member per month. And just wanted to understand, it seems like a really significant magnitude revenue beat versus your guidance and wanted to understand that.

Yes, absolutely, Gary. I think the components are how you're calling it out. Probably the biggest not the biggest, but the first thing is the retro membership that we're showing up that 1,000 members, but over the first full 3 quarters of the year, plus I think we beat the high end of the guidance by not quite 1,000 other members within the quarter. So both of those are contributing to definitely a double-digit millions of dollars of incremental revenue. And then the rest of it is really just syncing up our member-level data with the health plans, which we get the most up-to-date information on in the third quarter. And it's a combination of factors, including just getting the appropriate final files on things like do we have the right bid rates in for the plans that our members are in and updating all that as well as any interim midyear updates to our expected risk adjustment scores, and that made up the rest of it.

And then you beat your medical margin guide by $6 million to $11 million. It sounds like DCE was negative $3 million, you thought it might be low single digits, so maybe that was $5 million or $6 million swing and EBITDA came in at the low end. So anything on G&A? Or how should we think about that? I guess, given the $11 million medical margin beat even with DCE coming in below would have thought maybe a touch higher on EBITDA?

Yes. I think platform support cost continues to be in the range that we expect. You can see it running at about 5% of revenue kind of quarter in and quarter out right now. I think that will continue to be the case for the full year. I mean it's a big improvement over a year ago, but it's not a quarter-to-quarter huge driver of variance to our EBITDA guidance. As we go into the full year, there's probably $1 million or $2 million upside versus what was in our original expectations, but it's not a big driver. I mean within the quarter itself, the medical margin be flowing through the network contribution was obviously positive, and we ended up kind of in the low end, but within our guidance range, rather than at the top end of the range or above the range because of the, as you said, $5 million or so difference in our expectation on direct contracting and adjusted EBITDA.

And just last one on DCE. I kind of been following this. So nearly 8% retro adjustment in April, you guys were orderly conservatively accrued for that. It looked like another 2% retro in August. You said there was a little bit of a hit there. But my understanding was in September, there was another interim update that swung 3 or 4 full year, 3 or 4 points to the positive to maybe full year only trending down 6% or 7%. So is that just incorrect? Or is there some nuance in your regional benchmarks versus what we might be seeing nationally?

Yes. First of all, the retro trend adjustment is national. So the same retro trend will apply to all players in it. There's no regional retro trend adjustment. It is the full 30 million-plus Medicare fee-for-service members benchmark that applies to everybody. I think you're right, the original May adjustment was down about 8%, definitely came down further in August. They're giving us interim data updates, which we have seen another 2 of since that August update, which is indicating that it may be kind of trending back in the same direction. We're not seeing anything that would say it would be another, whatever you said 3 or 4 points back. The other thing is when the numbers come out, they're basically just giving us raw data that says, hey, here's what the experience is year-to-date. And right now, we have it through September, I think, for the DC population. But we still have to go through and do our own analysis of what does that mean in terms of seasonality in the fourth quarter. And so what might that really mean for the full year. But right now -- and so obviously, when we came down a little bit further in the third quarter, we wouldn't expect that the full year would be -- rebound from what we saw that by 2 or 3 points by any stretch.

As that's all the questions we have time for, I'd now like to hand the conference back over to the management team for closing remarks.

Great. Thanks, everyone. We really appreciate it. Hope everyone has a good evening.

Ladies and gentlemen, this concludes today's conference call. Thank you for joining. You may now disconnect your lines.