# Oral argument — Texas Tech Physicians Assoc v. HHS > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/argument%3A61396 ## Record - **Collection:** Oral argument - **Document type:** Oral argument - **Published:** February 5, 2019 ## Text We're here today to ask that this court set aside the HHS decision that Texas Tech Physicians Association is obligated to refund the management fees that were paid in connection with a program, an HHS program, that was designed to test whether one could actually reduce the medical expenses of a group based on care management. And we're here today because the terms and conditions of the demonstration agreement under which Texas Tech agreed to perform were not honored by CMS. We're here today because in reviewing that decision, the DAB, the HHS Decisional Board, abused its discretion. Now, the government council correctly observes that that's a practically 30-page decision. It certainly cites to a lot of cases. It certainly analyzes a lot of facts. So what are we complaining about? This is what we're complaining about. The DAB analyzed these issues as though it could simply look at the whether there was a savings and whether the participant achieved the savings requirement. There was a 5% savings requirement. That's not in dispute. But the DAB looked as though it only needed to look at that financial protocol. So what's the big deal about that? I'm going to use an exaggerated example of what I think the DAB's analysis could have allowed to have happened. This is a test group. There's an intervention group, which Texas Tech is responsible for their care. And we assume the risk that if our management of those services didn't reduce the costs, then we weren't entitled to a management fee. We contractually agreed to that, and there's no dispute. Now, what we didn't agree to, and here's my outrageous example. If CMS and its contractor had put 5,000 25-year-olds who didn't drink and didn't smoke in our control group, and we didn't achieve a 5% savings, we didn't assume that risk. And it wasn't as In science experiments, it's called a control group. Under our demonstration agreement, it's called a comparison group. And the significance of that comparison group is it provides the benchmark. I think I understand all that, and that's helpful. When I read your brief, I jumped to about page 39 to see where is the alleged breach of the grant agreement. Thank you, Your Honor. But at that point, I noticed you cited record 436 for the proposition that HHS had agreed to ensure that the comparison group would match the intervention group. But it is true, as they point out, that the language directly in front of the insurer says they would pay attention to insuring. Yes, yes. So it's hard to see a breach if the only agreement was that they would pay attention to insuring it, just as presumably you could have paid attention to it, and if there were a huge disparity, you would just have chosen not to enter into this arrangement. Your Honor, in construing, I think we construed, that's exactly the language. But in deciding whether that's an undertaking, sufficient to require that the government honor it in the context of this agreement, the operative word is ensure. And the other part of it is we have to look at what's going on in this demonstration agreement. Texas Tech was involved in the methodology for constituting the group, but not for actually, we had no participation. We don't have access to CMS's database, and that's the magic to constituting the control group, looking at all the information that's in the CMS database. And so when they undertake the responsibility to pay attention to ensure, I think ensure is the operative verb. Pay attention is, you know, I mean, as a driver, I have a duty to pay attention, and that's a pretty serious duty. If I don't do it, I'm negligent. And I think, you know, CMS had a duty to pay attention to ensure. I think pay attention modifies ensure, and ensure is the operative verb here, and it is an undertaking in the context of this comparison group. And in fact, once we saw the problems, we asked for information about that group, and we were denied access. So that attention to ensuring, when the other side unilaterally controls that in the context of an agreement. I mean, that does seem concerning, but their answer to that is multilevel. They say the record actually doesn't bear out that you were denied access, and then they claim that any delay has to do with HIPAA and privacy concerns. I guess, furthermore, they're saying the grant agreement itself, you didn't negotiate to have in there the right to all the data that would affect the process of selecting the comparison group? I'm going to break that into a couple of responses. We got access bit by bit to our participants, not to the control group. And in fact, the DAB criticizes our consultant because we had to do its modeling, it had to sample from outside information. We had certain high-level information about the control group, but not the details of it. And not at a time when it could have saved the project. And so, I think we've got to make a distinction between our control group data and the intervention group. Are you going to press that this is a procurement contract, or are you content to accept that this is a grant agreement? We are not waiving our procurement contract argument. However, I don't think, I mean, the defenses available to us change. There's a lot, the jurisprudence is a lot easier to apply in the contract arena than in this grant. But at the end of the day, if it's a grant agreement, there's still an undertaking. So we haven't abandoned the procurement contract argument, but I don't think that's dispositive. Keep going. The government cites to our participation in approving the methodology as though that somehow makes us responsible. And the DAB says that as well, suggesting that we actually populated that intervention group when in fact we didn't. We had no access to the information, and we were left with this group. So the distinction is you did have the opportunity to review and accept the entire arrangement and then the money that began to flow with access to the methodology for establishing the comparison group. But your point is you didn't have the access to the data, correct? So you knew the methodology. You knew they were going to refer, what is it, RTI? But the point is that they had to give you data. Is there something in the grant that says that, or do you think that's just a good faith? Well, you asked where the breach occurred, and the breach occurred in the implementation. The breach occurred in the fact that the control group is not an appropriate comparison group in this project. And it sounds callous, but unfortunately there are certain drivers of health care costs, poverty, disability, whether one lives in a nursing home or not, whether one is on Medicare because one is disabled or one is over 65. And the Solutio report, you know, the DAB criticizes that our project person who provided an affidavit isn't qualified to render opinions, but we had an outside consultant who studies CMS programs and has qualifications who documented these are drivers that escalate costs, and our group was higher on all of those. And here's an example of the breach. You asked me, you know, where's the contract? Where's the breach? We signed up in January of 2006. There are some modifications going on about the methodology that will be used, and that's finally, you know, they finally agree to this in sort of the end of April 2006 time period. So, you know, we're already going and they're making adjustments. But one of the things we emphasize is there's a difference between Amarillo and Lubbock. Amarillo has a higher socioeconomic by either 10 or 7 percent depending on median. I'm drawing a blank on which is which at this point. And we had many more people from Lubbock. And so on all of these critical drivers of costs, our group had the higher percentage of the group that drove up health care costs. And I think nothing speaks more clearly than if we look at three simple statistics. When the program, even before the solicitation was going on back in 2004 and 2005, even before the program started, they were evaluating eligibility. At that time, the difference between the control group and the intervention group was 4 percent. Then when we take the year before the program even starts, the difference between the two groups is 5.8 percent. Texas Tech's intervention group is already almost 6 percent higher at that point. And then finally you look at the first six months. You know, in the first six months is outreach. We're telling people about the program. We're starting, and the health care costs of that group escalate to 14 percent. And that's the breach. We had to have an appropriate comparison group, and we didn't. And the RTI, I mean, this isn't disputed. The government doesn't dispute our complaints about the group. I'm going to read just from the November 11th. You know, the die is cast. We're not trying to solve the problem at this point. We're throwing stones at it. Well, we were never throwing stones. But, you know, we weren't trying to fix the problem at this point. So the government's November 11th report says, we found intervention group beneficiaries did differ from the comparison group in having fewer Medicare-to-AIDS. And he goes through the statistics, and then he writes, it is possible that systematic differences in beneficiary mix, coupled with the side effects on per-beneficiary, per-month cost trends, could explain the higher rate of cost increases in the intervention group. But am I oversimplifying that they'll say, we have a really generous program around the country, we're trying to reduce costs, so we're making these stimulant demonstration agreements available to any hospital, but you bear the risk. And then, of course, that comparison group's got to be as matched to your intervention group. But it may not be perfect. And I don't see anything in this record that suggests, why would they have any interest in not getting to the perfect, closest match? Because they want this to work, too. So would you have to establish bad faith? Or are you accepting that it may just not have been the best comparison group, and then their response is, you took on that risk? How do you unpack that? I think there's a rough order of magnitude. Very simply, if they can't make an adjustment to even that out so it's quantifiable, then we don't bear that risk. But where does it say that in the agreement? That seems intuitive. Yes, Your Honor, because they've got to be able to measure our savings, and they can't measure the savings that we accomplish. There's a failure of proof on the part of the government. Last question. What's the best case out there? Because I'm sure these grant arrangements with sort of the disgorgement thing must occur all the time, and then this would be the natural objection. Hey, we couldn't meet their benchmarks. What's your best circuit authority on in a grant context? Well, you know, you could look at Brown, a fifth circuit decision. You could look at Thermalon. I mean, what they say is, look at the grant agreement. What are the undertakings? Did they abide by them? You know, the Thermalon has one thing that I think. Okay, and then just because your time is running out, where you're telling us to look in the grant agreement is their duty to ensure a match. That's what you zero in on? Well, one has to look at all of the undertakings. And included in the comparison group, they've got to be able to measure what our savings was, and they have to provide claims data to us. So there are a number of undertakings to review. Thank you, Counsel. Thank you. Mr. Stoltz. Thank you, Your Honor. May it please the Court. I'd like to first address this distinction that Counsel is attempting to draw between CMS's work in preparing a methodology for preparing a control group and then what they call the implementation of the control group. And I'd like to highlight a few items in the record that talk about this, because the record is pretty clear that actually there was information provided. The background is after the demonstration agreement is signed in January of 2006, it's not actually going to start for several months, and there's an enormous back and forth between the parties in trying to essentially create and populate these groups. Now, on April 4, 2006, at page 750 in the record, RTI, which is the contractor, has a memo that they send out that they give to Texas Tech that has all kinds of information about their proposed control group or comparison group. And that memo talks about, for example, that the control group, the most recent data was about $1,814 per person per month in Medicare spend, whereas the Texas Tech group was $1,733. So they have some very granular information about the relative costs. The RTI memo on page 750 of the record, at page 753, actually says there's 5,117 comparison Medicare beneficiaries in this control group. And, again, this is before Texas Tech agrees to this methodology. So the control group has already, in a sense, been created, and Texas Tech has been given information, not only about the exact number of people that's in this group, the exact dollar amount to the dollar of how much this group costs compared to Texas Tech's group. If you look at that memo on page 750, there's also a bunch of tables and other information about specific characteristics of this group, including, you know, how many of them are on dialysis, how many of them are in hospice, all kinds of factors. So this is a memo on April 4. Now, what happens next? Texas Tech actually comes back and requests more data. They've reviewed this, and they want more data. So there's a second memo from RTI at page 759 of the record where they actually give more statistics. And they provide pages and pages of these statistical tables about all kinds of information, how many of these people have had hip fractures, all these drivers that may increase the cost. So all of this information was provided to Texas Tech, and they had an opportunity, and they did in fact ask for more information on selected topics that they wanted. They then presumably satisfied themselves that they had received enough information to enter into this agreement knowingly. And only after receiving these multiple memos on April 24 that page 680 of the record is when Texas Tech formally gives their written acceptance of the control group. So at the time they accept it, they've already received all this information, and they've had the opportunity to ask for more information. Now, there were some questions about what did the demonstration agreement require in terms of data that CMS was going to provide. And I would direct the court to around page 424 of the record, one of the protocols to the agreement, which is one of the appendices, essentially. There's a protocol about data. And that protocol at page 424 of the record has a number of very specific, you know, essentially lists or itemizations of what kind of data would be provided and would be available to Texas Tech. And for the most part, it's all data about the intervention group. It's data about the group that Texas Tech is going to be providing these services to. Now, I think at page 425, as part of that protocol, there is a reference to what data Texas Tech would be receiving about the control group. And what the agreement says is they would get these monitoring reports that would come out periodically during the course of the performance, and they would receive monitoring reports about sort of big picture how the control group is doing versus the intervention group. But the contrast is Protocol 3, you know, has many, many, many items of specific data that will be available about the intervention group, and there's no similar commitment to provide that data, you know, about the other groups. And again, this is... Texas Tech knew this going into the agreement at the front. So I think that the, you know, the notion that there was some sort of breach in the quote-unquote methodology, I just think the record does not support that. And I would also point out that... It would seem that they would have the data as to their intervention group, so what's critical as they're beginning to see reports come out that they're not reducing costs, they would naturally want to know more. So just your brief response is they didn't negotiate for that information well, or there is no evidence that it was withheld, or it was withheld because it had to be under HIPAA. There's... Okay, there's a couple... There's a couple things. One is they did not negotiate and include that in the demonstration agreement up front. Now, there was provisions to receive monitoring reports and things like that, but not just a global dump of all the data. Now, when they did ask, at some point they raised the issue of what kind of data they could receive, and the record is very, very sparse on this point. 580 in the record is an e-mail exchange toward the end of 2006 where apparently there was some request for data, and so CMS sends an e-mail and says, okay, we understand you request data. You know, the rule I specified in the agreement is to the extent you want data, you need a data use agreement to cover it, and they set out basically what needs to happen. The response, Texas Tech's response is on page 580 of the record, and they basically say we haven't received any data that would require an additional data use agreement, and that's it. So the record just doesn't reflect that, you know, that there was some request at the time. What if instead of anxiety about data matching, their concern had been there's mismanagement? HHS has got mismanagement as to the comparison group. They have no recourse? Or, in other words, let's say there really was, put aside bad faith, but mismanagement. Under this grant arrangement, how would they be able to correct that? I guess I'm not sure what the court means by mismanagement. Well, some flawed HHS RTI, if that was the abbreviation, is it RTI? Some flawed construction and assessment of this somewhat difficult to penetrate comparison group. So the government just gives reports saying your numbers are much worse than ours, but in fact there's someone inside the government that's just not calculating correctly. How do they verify? See what I mean? I do understand, Your Honor. I think that the Departmental Appeals Board addressed this issue at least somewhat. It wasn't directly raised, I think. But what they said is that Texas Tech, there was a provision in the agreement for a reconciliation after the termination of the agreement, and the reconciliation would go in and look at all the costs of the intervention group and all the costs to the control group. And the reconciliation was performed by a contractor. And when the reconciliation report came out, the reconciliation report tallied up and found essentially that the control group ended up being less expensive than the intervention group. Now, as the Appeals Board mentioned, Texas Tech did not challenge any of the mathematics of that reconciliation. What they did challenge is they said, it turns out that, you know, two years into this agreement, it turns out that our group is actually more expensive than the control group. And what they essentially said is that there must have been something wrong or inappropriate about the control group the way it was constructed at the beginning, because, you know, we thought that our group would be cheaper. And, you know, I hope the Court can see how, you know, there was a number of these demonstration projects across the country, and every contractor who essentially came out on the wrong end of their risk-sharing arrangement could make a similar argument. They could say, well, essentially, it's not our fault, you know, it's just... Why did it take so long to get the second demand letter, right? Wasn't it about four years from the first to the second? I think that's correct, Your Honor. So the reconciliation... The agreement contemplated that the reconciliation would actually take a long time to do after the agreement terminated, because they'd have to essentially wait for the claims to come in, make sure they'd captured all the data. So the agreement contemplated about a year just to do the reconciliation. That came out in 2008. CMS made a demand in 2009. There was some back and forth in 2009 with Texas Tech saying, essentially, we want you to withdraw the demand, we don't think we should have to pay. And then you're correct that it was 2013 when CMS then said, no, we are, you know, we're sticking with essentially what we said earlier. Now, the record does not reflect what was going on in 2011, 2012. I just... I can't say. They cited two cases, our Brown and our Thermalon case. Do you have any thoughts on their reference to those as being supportive of... I do, Your Honor. I think the Brown case, which is Institute for Technology Development versus Brown, I think it's actually a very good case, and it's on point in the sense that that was a review of a grant agreement dispute very similar to this one. And if you read the Court's opinion in Brown, the grant agreement in that case was constructed much in the way this one was. It was an agreement, there was terms and conditions, and it was essentially a document. And what the Court said is, although this document has a contractual aspect to it, and it imposes obligations on both parties, it's not exactly the same as a contract because of this notion of, you know, grants are a little bit different in the sense that the government is bound by its express undertakings. And what the Court did, though, is it analyzed that contract and the terms and conditions of that contract to find out whether, you know, what the agency had done was authorized. And I think what's also interesting about the Brown case is in Brown, the Court relied on this line of cases that the sort of leading case is Kentucky versus, it's Bennett versus the Kentucky Department of Education. And that's a Supreme Court case in which the Court said, you know, essentially grant agreements are contractual, but they're also a little bit different. And so in Brown, the Court actually cited the Kentucky Department of Education case and sort of relied on those notions. And again, that's exactly the same thing what the Department of Appeals Board did here. And in fact, I'd like to highlight in the briefing in the record, on page 37 of Texas Tech's brief, Texas Tech makes its argument that essentially the agency board ignored authority that a grant agreement can create binding obligations for the government. And in fact, Texas Tech cites to the Institute for Technology Development versus Brown case at that section of the brief. They claim that the DAB ignored that authority. In fact, though, if you go to page 107 of the record in the DAB's decision, the DAB cites the Brown case in their decision from this Court. And they cite the language where the Fifth Circuit says that grant agreements are binding on both the grantor and the grantee, but the dispute needs to be resolved by looking at the actual binding agreement, including any obligations. So again, I think the record is clear that the Board did in fact follow this Circuit's precedent and other precedents, and it certainly, the Board did not act in an arbitrary and capricious manner in following those precedents. I also want to, the notion that the Board essentially ignored or didn't agree that CMS had any sort of obligations under this agreement, I think that that is not supported by the record either, for one, because of this case that they cite, the Brown case. Also, if you look at page around 108 of the record, the Board goes through a very detailed analysis of the terms and conditions of the agreement with respect to what sort of guarantee, if any, CMS was making about the control group, and that's in Protocol 5 of the demonstration agreement. And the Board spends several, I think, several paragraphs, if not pages, really discussing in detail and examining the exact language of that protocol, and that's the language about attention being paid to insurer, but then there's language saying that Texas Tech will have to mutually agree on what is done. So the Board did not just disregard it and say, no, CMS does not have any obligations under this agreement. What the Board did is they assumed, and they knew, that CMS in fact did have obligations under the agreement, but then they went in and analyzed what specifically those obligations were with respect to creating the control group, and their decision was, well, they did not have this obligation to sort of unilaterally guarantee that the control group will turn out to your liking as measured by sort of an after-the-fact judgment about whether your group ends up being less expensive. So the issue is the Board essentially applied its analysis, looked to relevant precedents, and ruled against Texas Tech. Now, I understand why they don't like that decision, but it just, it doesn't meet the very high bar to be shown to be an arbitrary and capricious decision. Okay. You know, I can talk about other issues if the Court has any questions. Besides that, we would rest on our briefs and ask for the decision to be affirmed. Thank you, Counsel. Rebuttal argument? I will try to address the issues raised by Counsel in order. Starting first with the April 4, 2006 memo, and this idea that Texas Tech had access to information that somehow, I would direct the Court to Table 4 on that memo, 757 It's simply another, it's explaining the concept of the methodology, and when it finally summarizes in the last page of that document how the comparison group should look and how the intervention group should look, according to RTI, it's very, very similar. But, of course, we already knew statistically that they weren't similar at that point. So I don't think that the memo supports this notion that we were given access. In fact, the table that one would rely on after reviewing this information about methodology suggests that the comparison and intervention group are actually going to be quite comparable. The references to hospice and dialysis, there was some back and forth about the methodology, and here's an example. If you were on hospice or dialysis, you didn't participate. So that's what that reference was there. Texas Tech, there was a loyalty factor, and then there was exclusion and inclusion factors. And one of the things Texas Tech asked about is, which are you going to run first, the loyalty first or the exclusion inclusion? We did not have access to the CMS, that's the point, to the CMS database. And that's how you know whether these things are similar or not. And this memo... This is the CMS database. That is correct. As we say in the maritime law, that was the rules of engagement. You knew that you weren't going to have access to that. Yes, Justice King, but by the same token, our partner in this program was Trailblazer, which is a fiscal intermediary for Medicare. That means it acts as Medicare's agent in processing claims. And so when we submit, in response to the proposal, we actually outlined that we intended to use our own data. Meaning, you know, we were working with Texas Trailblazer and that we could use their claims data that they had in their possession. It's in our proposal, it's in a diagram, and for you know, it's laid out in the proposal. And so at some point in December of 2006, Texas Tech says, you know, I need to get access to the, we asked for access to the intervention group, and CMS says, no. Texas Tech says, well, I want to create my own pseudo-comparison group because there are problems with mine. And so we can see, so I have an appropriate benchmark for whether or not I'm saving costs or not. And CMS says, you cannot create your own pseudo. And they said, in fact, if Trailblazer has given you any information, you better destroy it and certify it. I was going to say, I wouldn't have thought Trailblazers could give you that information. Well, within the context of this program, we proposed using that information. And we believed we would be able to use it under this, you know, we had a data use agreement as a participant in the program. CMS does not agree with that, but that was our understanding. That was what we proposed, and our proposal was accepted. And that was our understanding, and we didn't learn until December of 2006. And I want to make one comment about, um, we heard the record, we heard the characterization of the record is sparse. In a court of law, what I describe this record as uncontroverted. The Texas Tech project manager provided a declaration that explained she didn't get access to the information she requested, that she, you know, in response to that she had to sign a data use agreement, and that it took months before it was signed, so it wasn't returned. And so that's uncontroverted rather than sparse. And the government also didn't take issue with the flaws in the protocol. That's another uncontroverted fact. Your Honor, could I add one more case to the cases that are worth reading on this, and that's community relations, social development, versus United States. And like Brown... Is that in your brief, or do you want to give us a cite? Yes, it's... Is it in your brief? To be safe, can I go ahead and give you the cite? Well, give the cite, but if it's not in the brief, bringing it up at this point, obviously opposing counsel hasn't had a chance, but go ahead, give the cite. It's 8, Court of Claims, 723. And it's... I mention this case just because it's another example of when the agency takes too narrow a scope in determining what the undertakings are, the reviewing court looks at the documents and finds an abuse of discretion if they narrowly focus on one thing and ignore the other terms and conditions that govern the relationship. That's what happened in Brown. Remember the secretary hadn't looked at... We'll look at it. Thank you very much. Thank you. --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/argument%3A61396. Public record. Not legal advice.