At its January 12, 2017 meeting, the Medicare Payment Advisory Commission (MedPAC) made it clear they had reached the conclusion that the Merit-based Incentive Payment System (MIPS) cannot work (see my last post ). MIPS is the larger of the two programs within MACRA; the Alternative Payment Model (APM) program is the other. The commission’s primary rationale for its conclusion about MIPS is that it’s not possible to measure physician “merit” (cost and quality) at the individual physician level.
But rather than recommend that Congress repeal MACRA (the Medicare Access and CHIP Reauthorization Act), MedPAC decided to try to fix it. At the January and March 2 meetings, the commissioners discussed a staff proposal to amend MIPS substantially and to tweak the APM program. Those discussions went nowhere.
I give MedPAC credit for finally stating unequivocally that MIPS cannot work. But MedPAC should never have volunteered to fix MACRA. It can’t be done. By proposing modest amendments to MACRA and thereby implying it’s fixable, they stepped into an intellectual tar pit. I will illuminate this tar pit by describing the commission’s unproductive discussion about the staff’s proposed amendments to MACRA. To give you a sneak preview of what that discussion was like, I give you two excerpts from the transcript of the January meeting:
[Commissioner Kathy Buto] As I look at this whole [MACRA] system, I worry about the complexity of it and whether it’s going to achieve anything.… Medicare has sometimes a way of getting into a system it can’t get out of, and this has that feeling about it, that we could go down a rabbit hole and not be able to get out…. (p. 290, transcript )
[MedPAC director Mark Miller] And the reality of MIPS right now is it’s dead in the water because of those complexities. And that’s what we grind tons of time on, and I just wonder sometimes how we’re going to get out of this…. (p. 298, transcript, January meeting)
Allow me to introduce Scylla and Charybdis
MedPAC’s staff, under the direction of Chairman Francis Crosson, presented its alternative to the MIPS program at the January 12 meeting. The staff proposed reducing MIPS’ reporting burden substantially but also increasing the financial pressure on physicians who refuse to flee MIPS and join a group where, allegedly, physician “merit” would be measurable. According to the staff’s proposal, physicians could either stay in MIPS and sacrifice 1 to 2 percent of their Medicare revenue, or flee to: (1) a MACRA APM (the ACO is the main APM prototype); (2) a “virtual group” (the staff did not explain what that might be); or (3) a group defined solely by geographical criteria that CMS would select (the staff did not explain this ephemeral entity either).
The staff was essentially proposing that the commission abandon one impossible mission (figuring out how to measure “merit” at the level of the individual physician) for another – figuring out how to jam all doctors into functioning groups so that measurement of “merit” can be more accurate. The staff didn’t warn the commissioners they were asking them to swap one impossible mission for another. The staff merely asserted that measuring cost and quality at the individual physician level is impossible because each doctor serves relatively few patients and that herding doctors into large groups would solve the small-numbers problem. As Chairman Crosson put it at the March meeting, “[T]he whole point of [the staff’s proposal] is to achieve a volume of care of Medicare beneficiaries that is actually measureable….” (p. 228, transcript of the March 2 morning session)
As is so often the case when managed care advocates promulgate their proposals, the staff did not identify assumptions that have to be true for their proposal to work. The most fundamental of the unarticulated assumptions is that functioning groups of doctors (be they APMs, “virtual,” or geographically defined) will pop up all over the country so that all doctors who wish to flee MIPS will have a group to run to. In addition to being universally available, these groups must be real (not mere constructs on CMS or Dartmouth Atlas computers), and they must be functional – they must actually improve care and lower costs at the system level (not just for Medicare), something no previous managed care fad (HMO, ACO, “integrated” whatnot, “medical home”) has managed to achieve.
If Dr. Crosson and the MedPAC staff had bothered to articulate these and other assumptions, they might have grasped that measurement at the group level presents obstacles as insurmountable, and as mind-bending to contemplate, as measurement at the individual level. But they didn’t do that. Predictably enough, this failure to lay out the unarticulated assumptions caused the commissioners to descend swiftly down a rabbit hole where productive discussion became impossible.
Adventures in the rabbit hole: The January meeting
As I noted in my last post, when the staff finished presenting their proposal at the commission’s January 12 meeting, Commissioner Craig Sammit asked, “[D]o we have a sense of how accessible and feasible it will be for clinicians to join APMs?” Staff member David Glass replied, “I would say that’s a little hard to say right now.” (pp. 248-249, transcript) That was the commission’s first warning that a rabbit hole was opening under their feet. But no one noticed, and down they went.
The commission’s discussion of Dr. Sammit’s question focused at first on the difficulties small clinics would face in creating an organization that would qualify as a MACRA APM. One of the criteria for an APM is that it must bear “two-sided risk” (the risk that the APM will lose money as well as make money). Can small practices be expected to do that? Commissioner David Nerenz noted that very few of the groups that have signed up for Medicare’s MSSP ACO program signed up for the “track” that requires they take both upside and downside risk, so why should anyone expect small clinics to expose themselves to two-sided risk? “We have no good working examples of people willing … to step into the two-sided risk models of the MSSP,” Nerenz warned. “It’s a tiny, tiny fraction of those that are accepting two-sided risk. You can force people in. You could declare everything else to be a felony punishable by law, but we don’t have examples of how to do it.” (p. 268)
Commissioner Jack Hoadley asked the staff if they could refer the commission to “any history” of successful “physician groups” the commission might learn from “so that we’re not in the business of just creating … organizations that aren’t really doing anything….” (p. 291)
The staff had no answers for Nerenz, Hoadley, Buto and other commissioners who asked how small clinics were supposed to meet the two-sided-risk requirement for a MACRA APM. Commissioner Paul Ginsburg summarized the prevailing evidence accurately when he said, “I think that the reality is that for the fairly large percentage of physicians, they probably don’t have very extensive APM activities now.” (p. 250)
So if smaller clinics don’t want to stay in MIPS where they’ll lose 1 to 2 percent of their Medicare reimbursement, and if they can’t flee to an APM because they can’t handle two-sided risk, maybe they could flee to one of the other two types of groups proposed by the staff – “virtual groups” that doctors might form, and geography-based groups CMS would define. The commission discussed that scenario next. But that discussion also led nowhere. The commission wound up agreeing that groups of any sort (APM, virtual, geographical, whatever) make sense only if the groups are built on a shared culture – a consensus among the participating doctors about how they want to practice. This shared culture cannot be fabricated top-down by CMS or hospital executives seeking to expand their empires. They must arise organically.
At this point MedPAC’s director, Mark Miller, could see that the commission was beginning to realize they had swapped one impossible mission (measuring at the individual level) for another (measuring at the group level). But rather than advise the commission that the problem they were having was the intractability of the obstacles to measurement at both the individual and group level, Miller blamed the commission members for being unable to complete either mission. “[M]easuring in the aggregate, measuring in the individual, you guys are the commissioners. I don’t care,” he began. Then he complained that advocates of measurement mania are damned if they try to measure at the individual level and damned if they try to measure at the group level (as if the measurement advocates were the victims here). “[B]ut at some point in time, somebody has got to decide how this is going to go.” Miller offered no way out of the rabbit hole he had led the commission down. He instead offered only his own frustration at finding himself in the hole. “What are we doing here?” he asked. (p. 274) 
As the January meeting drew to a close, both Crosson and Miller expressed their confusion about how to summarize the commission’s discussion, presumably for the purpose of setting the agenda for the March meeting and to begin writing a chapter in the June 2017 report to Congress. Crosson was reduced to saying “[W]e need to take, you know, a much more effective approach, and I couldn’t describe it right now.” (p. 303) Commissioner Buto gamely tried to help Crosson by asking, “Are we essentially saying that on MIPS that … we don’t think it potentially is going to work the way it’s structured? … I think we still need to fill in the blank as to what we think should happen.” (p. 304) No one could answer Buto’s question; no one filled in any blanks.
Crosson ended the meeting with this plaintive question to the commission as a whole: “So a paper with suggestions. Is that what you’re saying”? (p. 307)
Further adventures in the rabbit hole: The March meeting
Crosson and the staff apparently learned nothing from the January meeting. They returned to the March meeting with the identical proposal they had presented at the January meeting, with one addition: They proposed to change the way in which primary care physicians (PCPs) who participated in ACOs would be paid, but not the amount they would be paid. The staff’s refusal to alter its proposal, and their addition of a partial-capitation method of payment for PCPs, made no sense. Here is my guess at the staff’s rationale for adding the PCP bauble. They left the January meeting aware that the commission was worried that small clinics wouldn’t have any groups to flee to; many of these small clinics have PCPs in them; if the PCPs were given more upfront financial support, smaller clinics might be able to join or start an ACO and survive two-sided risk. I realize this doesn’t make sense, but that’s the only rationale I can imagine. 
Following the staff’s presentation, the commission took up where they left off at the January meeting – posing questions about how the staff’s proposals for measuring “merit” at the group level would work. Commissioner Jack Hoadley led off with a question about why doctors in “virtual groups” and geographically defined groups would have any influence over the cost and quality measures they would be subjected to. “[I]t certainly doesn’t feel like it has any potential for anything I do as a clinician to make those measures go up or down,” he said (p. 207, transcript of March meeting).
Commissioner Brian DeBusk asked whether the staff had “explored” whether groups in lower-income areas would be punished by the proposed “quality” measures because of factors outside physician control. (pp. 209-210) Commissioner Kathy Buto said, “I do not know what we are trying to do with primary care….” (p. 235) Commissioner Alice Coombs (a doctor) said she was struggling with the question of “finding a home for the primary care physician [within] MACRA.” She asked her fellow commissioners to “have a little mercy on the onesie-twosie groups.” She said the “whole notion of attribution [of patients to doctors] within a geographic region” is riddled with problems. (pp. 222-223) 
Again, the staff’s answers to these obvious questions and concerns were grossly inadequate. Staff director Mark Miller was especially unhelpful. He implied the commissioners were spoiled brats asking for the moon. He said they needed to remember how bad MIPS is now and that anything would be better than MIPS. For example, after a brief discussion about whether hospitals could serve as the staging ground for an ACO when few doctors follow their patients into hospitals anymore, Miller said:
In listening to you guys, you are absolutely right. Tons of physicians do not set foot in a hospital, but you could almost have the hospital-based physicians … say, “Okay, let’s reach out and get these sets of physicians so that we can all be measured together.” And then I will just say this again: Keep in your mind, what is the alternative?…. [E]ach time you guys say this I am going to … force you back to the status quo…. [Y]ou are right, this has a bunch of problems; so does the status quo…. And so if there is another idea, that is what we are searching for. (pp. 228-229) 
But no one offered another idea. The closest thing to “other ideas” were bromides like, “I really think that we need to give up on … repelling people from unorganized care and drawing them into organized care” (Ginsburg, p. 237), and, “I think it is time to put our thumb on the scale for ACOs” (DeBusk, p. 246).
At the end of the January meeting, Chairman Crosson made an attempt to summarize the commission’s confused and unproductive conversation. It was embarrassing; Crosson sounded incoherent (see pp. 301-303, transcript, January meeting). His attempt accomplished nothing other than to drive home how much time the commission had wasted discussing the staff’s proposal. But Crosson, whose term expires this month, made no attempt to summarize the commission’s discussion at the end of the March meeting. Instead he ended the meeting abruptly with this cryptic happy talk: “We will be returning to this issue probably early in the next MedPAC term…. We are going in the right direction.” (p. 247)
Crosson himself observed at the January meeting that the commission could not just sit on the sidelines “and watch a dysfunctional thing unravel.” (p. 305) But it appears now that’s exactly what MedPAC intends to do. For the foreseeable future, it appears MedPAC will do what MedPAC always does after they identify defects in a managed care nostrum they previously endorsed – they will dither.
 Here is a lengthier version of the statement MedPAC director Miller made at the January meeting expressing his frustration with the commission’s inability to figure out how to make measurement at both the individual and group level work:
[T]here has been a 15, 20-year discussion about this, where it’s like, okay, you can’t do anything but measure me at my individual level, and … you have to tell me precisely how you are measuring me. [T]he physicians groups argue this. Then they are upset that it is too burdensome, that the comparisons aren’t fair, … and nobody is happy. [This] exchange [about individual versus group measurement] … basically … just stopped the process because it’s too complicated. So then, as an analyst, you’re sort of like, “I don’t know what to do, except go to a more aggregated level and measure.” Then everybody says, “But this isn’t me, and that’s not fair.” So fine, but at some point in time, somebody has got to decide how this is going to go. And I’m going to throw just one last bomb into the middle of this. A few meetings back, [commissioner] Paul [Ginsburg] said, “Why are we measuring and paying for quality?” and I’ll just ask that question. I mean, as long as this is so complicated and nobody can come to any agreement on this – and any model you pick, you’re going to have a bunch of unhappy people and a bunch of logical or analytical failure. Then maybe Paul’s point should come into the conversation and say, “What are we doing here?”…. [W]hat the hell. (pp. 273-274)
 Here is how MedPAC staffer Ariel Winter explained the partial-capitation payment method at the March meeting: “[T]he … approach would allow primary care practitioners in two-sided ACOs to receive an upfront, lump sum payment…. PCPs in ACOs would not receive new money for this upfront payment; instead, they would be shifting some of their own revenue from fee-for-service payments to the upfront payment.” Why, you might ask, should doctors feel happier about being swallowed by an ACO and bearing two-sided risk if they receive no increase in revenues, only a change in the chronological distribution of those revenues (more in January and less over the remaining months of the year)? Winter offered this non-explanation: “The advantage of this upfront payment is that it would give providers more flexibility to invest in infrastructure and staff for care coordination activities.” (pp. 199-200).
This “explanation” illustrates the free lunch syndrome . Winter is saying primary care doctors would now be “free” to spend money on something Medicare won’t currently pay for, for example, nurses stationed at nursing homes to reduce emergency visits, but new revenue to pay for the additional nurses would not be coming from CMS. Let’s see here, where might the extra revenue come from? Tinker Bell?
 Commissioner William Hall also commented on the attribution issue. “It seems to me that with MIPS … we are [telling] a group of physicians who call themselves primary care providers … ‘You are responsible for some kind of a catchment area…. And as we look at your area …, you have some problems – like maybe diabetes is not being cared for … very well, or there seems to be too much alcoholism in your community.’ And we are saying: ‘Why don’t you fix that?’ …. I think we have very muddy expectations what modern primary care should do.” (p. 238)
 The following conversation between Chairman Crosson and Commissioner Alice Coombs about whether hospitals could create “organized staff models,” taken from the transcript of the March 2017 meeting, illustrates one of the many dead ends the commission ran down in their search for the prototype of a physician group that would want to take on two-sided risk:
DR. CROSSON: So, Alice, … let’s just think about the organized medical staff model for a moment [Crosson was for many years an executive within Kaiser Permanente]. As you well know, … it’s kind of a group but not really a group…. Essentially, everybody is … nominally working in the same institution …. Historically, the organized medical staff has had a role … in trying to oversee and manage quality within that institution, right? Now, some are effective, and many have not been, and in fact, the enthusiasm for the value of the organized medical staff model has kind of waned over the last couple of decades. So at least one of the ways I’ve been thinking about this … is that … joining an A-APM would be an option, but then you would have the other option, which is to organize your organized medical staff to be the unit of measurement on these population-based measurements.
DR. COOMBS: So there’s an underlying assumption that you just made. … In our community [Dr. Coombs is from Massachusetts], 80 percent of docs don’t even come to the hospital. The hospital’s program has superseded … every single service. Pediatrics has an inpatient hospitalist. … [B]elieve it or not, the majority of care happens outside the hospital…. In the olden days, you’re right –
DR. CROSSON: Yeah. [pp. 226-228]