The CMS Center for Medicare & Medicaid Innovation (CMMI) has developed several value-based care models that address chronic kidney disease and end-stage kidney disease. In a recent interview, nephrologist Benjamin Hippen, M.D., senior vice president, head of transplant medicine and emerging capabilities at Fresenius Medical Care, explained why it is important that future models include the entire spectrum of kidney disease, including transplantation.
Hippen co-authored a recent article in Kidney International Reports and will speak at the American Transplant Congress in June, outlining a value-based care approach that’s inclusive of kidney transplant and post-transplant care.
Healthcare Innovation: We’ve written about some of the alternative payment models that CMMI has developed to target kidney care. Why have those not focused as much on transplants? Is it because it is so complex and they wanted to start smaller?
Hippen: I don’t know exactly why they didn't take on transplant right out of the gate. I think your answer is as good as any, which is that they really wanted to start with the dialysis population and move into chronic kidney disease. I think that's a reasonable and natural progression in part because the primary stakeholders there, from a care coordination standpoint, are dialysis providers in nephrology practices historically. This is one of the reasons I've been interested in this for so long. The care delivery on the general nephrology and transplant nephrology sides have been fairly siloed. The vast majority of transplant centers are based in academic medical centers, and the transplant nephrologist affiliated with those programs generally do little to no general nephrology and vice versa. I've long held the view and the hope that the silos between general and transplant nephrology in particular could be dismantled and that we can develop payment models that really treat the patient's journey as end-to-end CKD, ESKD, transplant and back again.
HCI: Do you think that the way that the nephrologists and the health centers are paid had previously contributed to that silo effect?
Hippen: Well, certainly they are reimbursed through fundamentally different payment models. Those payment models probably arose around the existing institutional divisions between general nephrology and transplant nephrology/surgery writ large. So the payment models naturally grew up around the existing institutions. My particular experience was quite unusual and still is quite unusual to be in a large community private practice while also devoting most of my clinical time to transplant nephrology. There are a few other nephrology practices around the country that have adopted that model, but by and large, it's the exception rather than the rule.
HCI: Are there some lessons learned in CMMI’s earlier models that could be applied to the transplant-inclusive model that you are proposing?
Hippen: I think so. I really view the transplant-inclusive value-based care model that's outlined in the paper as a natural extension of the current CMMI models in place under the rubric of the Kidney Care Choices model. I was disappointed, I guess, but not surprised that transplant played a relatively small role in the current models, but I think this is emblematic of CMMI’s stepwise approach. This was a big step to try to bridge the silos between general and transplant nephrology. And I think probably no one is fully satisfied — including CMMI — with how transplant was introduced here, but I think they intended it as a first step rather than a definitive approach.
HCI: CMMI doesn’t have a track record of basing alternative payment models on proposals from physicians. Do you have a sense that CMMI is looking at something like what's proposed in your article? You said that you see it as something of a natural progression.
Hippen: I have cautious optimism. I would never go out on a limb and try to predict what CMMI is going to do exactly. I do think that you're right that CMMI has developed a jaundiced eye toward many accountable care models, in part because many of them have not proven to bear fruit either from a quality or cost containment standpoint, or both.
I do think that there's reason to think that nephrology and transplantation are different, in part because the care of patients with end stage kidney disease and chronic kidney disease make up such a large fraction of the Medicare budget and the need for ongoing care and the growth of the patient populations is such that it's unlikely that that cost curve is going to bend all on its own. I think virtually no one would say that care coordination across all of these different episodes of a patient's experience from CKD the ESKD to transplant and back again are optimized. And I think there's also ample opportunity to realize efficiencies through care coordination strategies that work and bend the cost curve and achieve the quality metrics to benefit patients. I think that probably distinguishes it from many other ACO models.
HCI: You are proposing this as a an episode-of-care model rather than a total-cost-of-care model. Can you explain why that's important?
Hippen: It is tempting to look at the transplant finance ecosystem and try to tally up all the dollars that are expended across that entire continuum, and then approach it as a total-cost-of-care model where there are opportunities to reduce that total cost and thereby realized shared savings or some approach along those lines. I think that the transplant financial ecosystem is different for several reasons. One, it is extraordinarily complex. There are a lot of different stakeholders, a lot of different entities involved. This is not to say that CKD and ESRD care models are not complex, but there are additional complexities in the transplant financial ecosystem that I think make it less advisable or perhaps less desirable from the standpoint of unintended consequences of treating it as an all-in, total-cost-of-care model. In particular, and we argue for this in the paper, we think that the costs associated with organ procurement, preservation, logistics and transportation are all in a cost bucket that's widely variable geographically.
There are organ preservation techniques that are designed to improve organ acceptance rates and reduce by extension organ discard rates. My concern is that including that particular cost bucket in a total-cost-of-care model may inadvertently stifle innovation where innovation is desperately needed. I'm sure that CMMI would be interested in trying to bend that cost curve because it is it is not an insubstantial sum. I'm not sure that this particular model, which is really focused on care coordination, is the best tool to do that, if it's advisable at all, in part because the primary stakeholders, who would be in this case, nephrology providers, transplant providers and some sort of value-based care organizations around which care coordination could be funded and supported, are not really in a position to bend the cost curve with regard to organ acquisition costs.
‘
Now there is something of a misnomer or confusion around organ acquisition costs because part of the organ acquisition cost bucket is not just the procurement, preservation, transportation of organs. Included in that bucket are some pre-transplant costs attributable to the transplant candidate. So when an individual gets referred for transplant, they go through the transplant evaluation process and undergo testing. They have professional services by a multidisciplinary transplant team. All of those costs get attached to the Medicare cost report for transplant centers. So it's not just procuring and preserving, transporting the organ; it's all these other pieces. It's conceivable that some of those costs should get carved out and focused on in a different cost bucket. The current reimbursement approach for transplant centers doesn’t allow for that right now, but I think it certainly could.
HCI: CMS and CMMI have put a big focus in the last couple of years on health equity. One of the things that your paper mentions as key in terms of patient-focused goals is alleviating health inequities and access to transplants. Can you talk a little bit about those inequities and how this model might work to address them?
Hippen: Yes. I would say that inequities in terms of access to transplantation — both on the waitlisting side as well as actually receiving a transplant — are longstanding and durable. In particular, younger patients with diabetes who are African American or Hispanic have reduced access to waitlisting and transplantation in a manner that far outsizes their overall prevalence in the total advanced CKD and ESKD populations. We really have not moved the needle substantially on that. So I think bringing a systems approach to this, particularly upstream on the pre-transplant side is important not only to get more visibility on a regional and local level into the dynamics of those inequities, but also to be a path forward to identify the reasons for those inequities, and starting to develop system-level solutions that can be applied across geographies and across markets.
I'm not exactly sure what the right approach is to do that. I view this paper as the beginning of a conversation in the debate about how to accomplish that. But just by way of example, the ESRD Treatment Choices model recently had a revision that gives particular credit for achieving quality metrics for patients who are dually eligible for Medicare and Medicaid, as well as patients who qualify for the low-income subsidy. I'm not sure if that's exactly the right approach for a transplant-inclusive value-based care model, but I think that's in the general neighborhood of what I have in mind.
HCI: Also in the patient-focused goals arena, your paper talks about prioritizing transparency and shared decision-making in organ acceptance decision-making. Can you explain that one?
Hippen: One thing my colleague Sumit Mohan, who's a transplant nephrologist at Columbia University, and colleagues have pointed out in the literature is that there are a great many patients who receive organ offers while they're on the waiting list, and those offers are declined on their behalf by the transplant center. More often than not, patients are not informed that they received an organ offer that was declined on their behalf. There is some evidence that some of these patients have received a dozen or more organ offers that are declined on their behalf, and this issue with regard to shared decision-making and transparency really speaks to approaches to bring the patient into that conversation in a more routinized way.
It's not to say that the organs that are turned down for individual patients are turned down inappropriately, but there is a sense in which the risk/benefit assessment of individual organ offers are being made without a full understanding of what the patient's preferences are. The American Association of Kidney Patients has surveyed its membership with regard to their own risk tolerance for what's called higher risk organs that may not have as optimal graft survival or may have more complications. One wouldn't expect unanimity on the point. But it's clear that the risk tolerance for patients seems to be in general higher than it is for transplant physicians and surgeons.
Now transplant centers also have to operate under fairly stringent quality outcomes in the form of fairly high thresholds for one-year patient and graft survival. But I think it's worth pointing out that even if one looks at all of the transplant centers in the United States, and you bracket them by their performance as measured by one-year patient graft survival, even the least-well-performing tertile of transplant centers in the U.S. confer patient survival in excess of twofold compared to patients who remain on the waiting list on dialysis. So while we can't take our eye off of quality on the transplant center side, I do think that, increasingly, people have recognized that regulators and payers have moved too far in the other direction, and have created behaviors, unintentionally or not, that have resulted in risk aversion, which for transplant centers translates into higher rates of organ discard and avoiding higher risk patients for listing and transplant.
HCI: Your paper also touches on the risk arrangements for transplant centers. It mentions allowing transplant centers to partner with larger, better capitalized entities to take on downside risk in exchange for opportunity to realize upside financial benefits. Is that already happening now or do we need a model to encourage that?
Hippen: I think we do. I mean, kidney transplant programs in particular operate on fairly thin margins. So unless the kidney transplant program’s sponsoring health system is prepared to capitalize an endeavor so as to hedge against downside risk — and I don't think that they really are prepared to do that, at least as a general matter — then partnering with larger and better-capitalized value-based care companies that are focused on nephrology is probably a reasonable approach because these companies have developed care coordination strategies that are working for their patients and working for them from a financial standpoint. Generally speaking, they have been designed to have capital reserves to hedge against downside risk. So bringing transplant centers in in a manner where they are protected from downside risk is advisable and will give comfort to the vice president for finance in large health systems who are responsible for the balance sheet for kidney transplant programs.
HCI: Have you already received some feedback since this paper on the proposed model appeared online?
Hippen: Yes, it's generally been fairly positive. I haven't gotten any angry e-mails. I count that as a good sign. I will be presenting on themes in the paper at the American Transplant Congress in Philadelphia on June 5. I'll be accompanied by some colleagues, Dr. David Axelrod, who's a transplant surgeon at the University of Iowa and has significant expertise in transplant finance, as well as Rajiv Poduval, who’s the co-founder of Panoramic Health, which is one of the other renal value-based care entities, as well as Dr. Mona Doshi, who is the medical director for kidney transplantation at the University of Michigan. I anticipate that there will be some concerns and criticisms aired. I certainly hope so. This is really designed to present these ideas more or less for the first time to a transplant-heavy audience and not designed to be an exercise in cheerleading. It is really much more of a critical exercise because I and my co-authors really view this as a first draft.
HCI: Is there also an opportunity to share it with CMMI folks and get their feedback?
Hippen: Yes, I have shared the manuscript with the payment working groups of CMMI, and I believe they will also be attending the session. I'm fully confident that they are aware of the paper.
HCI: Are CMMI’s current nephrology models ending anytime soon, so that there would be a logical time for them to either extend those or to make the kinds of adjustments that you are recommending?
Hippen: The ESRD Treatment Choices model, the mandatory model, and the voluntary Kidney Care Choices models all sunset at the end of 2027. So if CMMI elects to extend these models in some fashion, I suspect it will be to some extent a revised framework that will begin to happen, I should think, in 2025 or so. These things take a long time to put together.
One of the reasons I was eager to get these ideas out in circulation is that a key deficit that will need to be addressed before these models can be stood up is developing datasets that will allow stakeholders who want to take risk to be able to assess what their costs are. If it's going to be tagged to a historic cost benchmark, how are they doing? What are their quality metrics right now? And are there care models that they can accommodate or develop? Is there opportunity to avoid downside financial risk from an uncertain future without the ability to sort through one's risk exposure and upside opportunities in advance? It's very difficult to justify engaging in these models when the downside risk could be substantial.
HCI: Is a lot of that in CMS claims data or are there a lot of different sources of data that you have to pull together?
Hippen: Some of it is in CMS claims data, but I think that will be inadequate to the task. I think there's a lot of data harmonization that needs to transpire. My colleague, David Axelrod a few years ago published a study looking at Premier-sourced data from hospitals, which is a little more robust than CMS data, trying to predict how much of the cost can be predicted just from the Premier variables, and it was actually a very small amount of the modifiable costs. So a lot of work needs to be done there.
