ľֱ

Obama CMS Chief: Medicare Advantage Plans Game the System

— Don Berwick says MA growth "should be slowed or stopped"

MedpageToday
A photo of Donald Berwick, MD

Donald Berwick, MD, MPP, a pediatrician, was the administrator of the Centers for Medicare & Medicaid Services (CMS) during the Obama administration and is president emeritus and senior fellow at the Institute for Healthcare Improvement.

ľֱ interviewed him by phone about his concerns that too many Medicare beneficiaries are being misled into enrolling into private Medicare Advantage plans, which he said should be slowed or stopped because the plans have gamed the system to receive billions more than what is spent for traditional Medicare beneficiaries' care.

The interview comes several weeks after ľֱ's interview with Michael Chernew, PhD, chairman of the Medicare Payment Advisory Commission (MedPAC), who, unlike Berwick, believes Medicare Advantage plans deliver better care.

The interview has been edited for clarity and brevity.

Cheryl Clark: Let's start with the big question. Medicare Advantage, or MA, plan enrollment has been growing so fast; 52% of beneficiaries are now enrolled, with 60%, 70% projected in a few years. Michael Chernew, PhD, chair of MedPAC, said some parts of the country are at 90%. I hear a traditional Medicare patient can't find a primary care provider in some parts of the country because all the doctors are locked in MA.

Will every eligible beneficiary have no choice but Medicare Advantage? Was that the goal of Congress to privatize Medicare entirely? And is that a good idea, given your concern about how MA plans have gamed the system to get more money?

Don Berwick, MD, MPP: I think MA growth should be slowed or stopped, at least until we end the extraordinarily high subsidies for MA plans, which are unfair to traditional Medicare and burdensome to the public treasury and many beneficiaries. Many beneficiaries can get better care for themselves and greater choice through traditional Medicare, and that option should remain robust and available.

I don't have a crystal ball, but Medicare Advantage is growing very fast. Some of the systems that apply to Medicare, like the benchmarks that are set for Medicare Advantage or the quality scores, are predicated on traditional Medicare foundations. As traditional Medicare shrinks, some of the original ideas behind comparing traditional Medicare to MA that are the basis for payment become unsustainable. The traditional Medicare population will become less desirable for insurers, meaning more expensive to treat. If everyone eventually ends up in Medicare Advantage – I can say with certainty that was not the intent of Congress when MA was set up.

The original idea behind MA, in its predecessor forms, was not a bad one. It was to give beneficiaries the choice to have what I call good, responsible managed care if they wanted it, but not require them to have it or trap them into it. It didn't take many years for that to be distorted.

Clark: Let's talk about the benchmarks and quality metrics, and how MA plans are overpaid. They'll receive $88 billion more this year than what is spent for the same patients in traditional Medicare. That is a huge concern. But I don't think it's clear to readers how the plans are paid, that it's not necessarily for better care.

Berwick: It's a really complicated process and, I must say, possibly intentionally so, because I think it serves the interests of the private plans to have payment rules that are so hard to understand.

What the benchmark begins with is a comparison to the traditional Medicare population: people like me, who are in fee-for-service, traditional Medicare. There's a calculation of the cost of my expected care in a given year, based on a whole bunch of demographic and other adjustments, such as county expenditure, because care is more expensive in some counties than others. And there's a quality adjustment, because plans are now scored on quality. There's adjustments for age and gender.

Most importantly, the amount that an MA plan gets is adjusted for the number of codes for diagnoses that a beneficiary has, like atrial fibrillation or diabetes, and for each code the plan gets more money, supposedly reflecting the additional care the patient needs. And that's where the gaming nonsense occurs. Many codes have no real implications at all for evidence-based clinical care, but they carry with them extra payment nonetheless. The coding system has created opportunities for these plans to upcode. They comb through patients' histories and try to stuff in as many diagnoses as they possibly can, even if they have nothing to do with the care of the patient.

Clark: During a January 12 MedPAC meeting, one of the commissioners mentioned that the plans were paying doctors to go through each record and look for anything that could add to capitated payment.

Berwick: That's right. That's worth a bit of a dive. There are three ways health plans manipulate the coding processes. What they're after is what you said, which is, comb through the patient's record, send a nurse into the patient's home to find additional diagnoses. Sometimes, in some of the MA systems, they actually pay the doctor to code using a software package the insurance plan gives to the doctor. The package is designed to excavate all the possibilities for coding, and they'll pay the doctor a certain amount per visit or per beneficiary.

Another way is to pay the doctors a percent of the insurance premium that the plan collects from CMS. This approach has embedded incentives to find more codes. And a third and now much more popular, certainly to insurers, is to simply employ the doctors. Once the doctor is an employee, then you can set up all sorts of ways to accelerate upcoding. You can train the doctors or give them incentives to upcode. And now, the largest employer of doctors is a health plan, an MA plan. In this third tactic, the MA plan gets the benefit of all the upcoding, and that's part of the game they are playing.

Now rather than lower costs, MA has much higher costs -- something to the tune of $80 billion a year. Other estimates are as high as $120 billion. For the most part, that money doesn't represent the needs of the patient. In fact, we know that beneficiaries in MA are, on the whole, healthier than those in traditional Medicare, and ought to cost less, not more.

It's just a transfer of money to the private sector. Most of that goes to profit for the plan, or for stock buybacks, high compensation for plan executives, and activities that don't benefit beneficiaries, including, by the way, administrative costs that are much higher than on the traditional Medicare side. As you know, that's now regulated to be no higher than 15% of the total. In traditional Medicare, administrative costs are much lower, more like 3% or 4% or 5% of the total. So the extra administrative expenses in MA represent waste. I doubt it was ever the intent to allow so much money to leach into private hands.

And further, because of all the gaming of diagnoses, it gets really, really hard to compare quality of care and outcomes between Medicare Advantage and traditional Medicare patients.

Clark: You've said you think MA plans should be slowed or stopped. How do you think that should happen?

Berwick: Medicare Advantage is an unstable platform. If things are allowed to continue, our opportunity to restrain its profiteering becomes more and more narrow. How do we put more restrictions on MA, stop the coding games, and assure MA plans serve people better? We need to stop these abuses as quickly as possible. And we need simultaneously to improve the attractiveness of traditional Medicare, which is what tens of millions of Americans still have. It would slow the expansion of MA if the public were more aware of what they lose, not what they gain, when they go into MA. The public would become less susceptible to misleading statements by MA plans.

Clark: Chernew told me that because MA plans receive that capitated rate, they have an incentive to reduce wasteful and potentially harmful care. I recall you saying about 10 years ago that there were so many unnecessary cardiac catheterization procedures, half of the interventionalists could quit working. Do you think that overall, MA plans are fulfilling their mission to deliver better care?

Berwick: No. At least I doubt it. We don't have evidence to support that assertion. It's true that there is no perfect payment system. In fee-for-service, hospitals try to stay full and there are incentives to do more things. It's a real mistake to think that doctors plot and plan to do things patients don't need. But human psychology is complex, so in an atmosphere in which you get paid for doing more, when there's a tough call, you will do more.

There is, indeed, an incentive to do less under MA payment, but it may be to do less to the disadvantage of the patient, to withhold care, to skimp. Under MA systems, you have metrics and monitors that reveal if care was withheld or skimped. And there is evidence that there is some skimping on needed care. When we look at the limited data we can get on pre-authorization barriers, coverage denials, and network inadequacy, what we see is that now these plans have an incentive to withhold things the patient needs, or to make it harder. The for-profit MA plans are out to make a profit, which means they will try to get paid more and spend less. That's the game for them and it's what they're doing right now.

Clark: Chernew said that for chronic conditions, there is evidence that MA plans deliver better care.

Berwick: First, Mike Chernew is a brilliant academician, a good friend and someone who really tries hard to help us all understand what's going on in healthcare. His chairmanship at MedPAC reflects his dedication and expertise. But I don't agree with him on everything.

There is some information that suggests that for some conditions, some MA plans do offer better chronic care but that's a really hard statement to prove, and it certainly doesn't apply across the board. And remember: MA plans are upcoding patients, so they make the patients look sicker, so when you try to assess outcomes, adjusted for severity, you've already fallen into the hole created by this game in which you're no longer comparing apples to apples. The best plans tend to be not-for-profit plans, although not all non-profits are better. But they're the ones that really do have governance and chartering foundations to try to do the best they possibly can for the patients. Overall, saying that MA patients get better chronic care is an overgeneralization beyond what the data support.

Clark: How can providers and beneficiaries know which MA plans are better, are not-for-profit? It's not always obvious without looking up tax forms.

Berwick: It's hard to tell. I think most people have no idea whether their insurer is for-profit or not-for-profit. It's just hard to find out who's paying who, who owns what, what the contracts look like. This is all defended in the MA world as proprietary information. But it is the information that consumers, regulators, and payers -- like employers -- should have to make decisions about what they're paying for. One of the things CMS seems to be undertaking now is to try to make the Medicare Advantage world much more transparent, with more information on ownership, the flow of money, financial performance. That's been an uphill battle, but one worth fighting.

Clark: The MA plans heavily market gym memberships, free trips to the doctor, and cash cards. Do those extra benefits improve health?

Berwick: And Super Bowl ads. I still have to pinch myself to believe that in 2023, the Medicare Advantage industry bought an ad during Super Bowl. Who knows how many millions that cost? You tell me how that reflects investment in the well-being of Americans. Everyone like me is getting these mailings practically every week trying to sell me an MA plan. No, it's not correct that whether it's $88 billion or $100 billion, those funds are transformed into better health and lower costs for Americans.

Clark: We've talked about ways MA plans fortify their bottom lines, and these AI algorithms that prolong approval and deny care. Some people suggest the delay tactic is designed to play the float. Is that the case?

Berwick: Let me rewind that question just one notch to talk about the original concept of a managed care system -- good HMOs. These plans try to overcome a defect in the fee-for-service system, which is a failure to support excellence in care by comparing the care plans for patients to what the science says should be done, like, "Is this the way to diagnosis and treat migraine headache" or "Is this the way to treat breast cancer?" That constructive review is distorted in most Medicare Advantage plans by games to delay or deny payment if they can. Some of that is, as you say, the float. The real answer to your question is: we don't know.

Because we're back to the problem of opacity. Health plans don't have to report coverage denials or authorization barriers or network adequacy at the level they should.

Some of the stories of how they're using computer algorithms instead of real people are scary. They've given artificial intelligence the job of making clinical decisions.

Clark: CMS has struggled to rein in even the most egregious cases of deceptive MA marketing. How can it manage these plans on much more complicated levels?

Berwick: CMS operates with a very, very thin administrative budget. That's one of the reasons administrative costs are so low. My opinion is that, if Congress is serious, and they should be, about getting some of these MA behaviors back under control, they do need to give the agency resources to scrutinize and enforce the public reporting we've been talking about. Right now, it's very hard. The agency has far less administrative support than it would need. MA plans are very wealthy and can and do pour money into lobbying, advertising, and coding software.

Clark: The federally-funded counseling programs, or SHIP, are poorly funded to do the job of advising so many beneficiaries. They also can't write policies. So beneficiaries largely get advice from brokers, whose commissions we know favor selling MA plans.

Berwick: That's another piece of opacity, the broker system. We don't have information that we need on how brokers are behaving. As you said, brokers are given enormous incentives by MA plans -- I've heard about trips to the Caribbean as rewards -- in order to route people toward MA, even when it is not to the clients' advantage.

Clark: You spoke of the quality bonus program, which is a factor in MA capitated payments. How does it work?

Berwick: It's an additional payment to plans that have higher quality scores, for example reflecting better blood pressure control among people with hypertension. But, in my opinion, the quality bonus system needs a big overhaul. It now has been pretty thoroughly gamed by MA plans. They focus on the scored variables, not overall better care – "teaching to the test," as it were. Something like 80% of plans now are four-star or five-star on a five-star scale. I call that the "Lake Wobegon" effect, where everybody is above average. It's a tricky system with unintended consequences. For example, when you're treating a very distressed population with limited resources and there are barriers to treatment, you don't want to take money away from the very organizations that need more money because their populations are harder to treat. But it has just become too easy to get a high score. When I was the Administrator and had to put into effect that quality bonus system, there was enormous political pressure to make sure the vast majority of plans would get money. But the original intent was to have truly higher performance recognized economically, not just to spread the money along average plans.

Clark: Whenever I talk with a doctor over age 65 who isn't employed, I ask if she has an MA plan or traditional Medicare. I have yet to hear one say MA. Have you ever been an MA enrollee?

Berwick: No, I'm in traditional Medicare and I'm happy to be there. It means that I can go anywhere I want for my care. I have not delegated to unknown administrative structures decisions as to where I can go and what care I can get. For-profit Medicare Advantage is mostly accountable to its investors and shareholders, not to patients.

Clark: One argument made for MA plans is that they can weed out bad clinicians, and perhaps know before the public when a doctor should be taken offline. Is that something MA plans have going for them?

Berwick: We don't have evidence on how MA plans choose who will be in their networks. My own suspicion is the choice is usually made not on quality criteria, but on cost criteria. They're just looking for the lowest costs in town. We also have data showing MA plans often don't include academic cancer centers in network. Well, if I have cancer I want to go to a cancer center. If I have a traumatic brain injury, or need complex coronary cardiac surgery, I want to go to a center that specializes in that. And I don't want any insurance company telling me I can't.

I think there should be a list of specialty services or centers of excellence, whether they're academic or not. And MA plans should have to come up with that. I want to defend the interests of patients, not the plans.

Clark: I want to shift to how friends tell me how mind-boggling the system is for seniors and the disabled. We're told we should study all our options every year. But seniors don't always know what they'll need. The average beneficiary has 43 MA plan choices and no easy way to compare them. Some seniors have waning capabilities to juggle so many variables. Have we made the system too confusing for people to make good decisions?

Berwick: That's what it looks like. And I don't think it's an age problem. You could take healthy 26-year-olds and show them an array of 30 MA plans and all the variations, I think they would pretty-well zone out. It's incredibly confusing, I chose traditional Medicare partly because it's simple compared to that telephone book list of MA plans. For me, it's quite simple: I get government insurance, end of story. But we do have to work on Medicare's out-of-pocket costs, which are higher than I believe they should be. And vision, dental, and hearing coverage should be part of traditional Medicare. And by the way, all that money that's going to MA plan advertising is not adding value. That's not a market. That's chaos.

The question to ask ourselves is, how did it get so complicated? I have a strong belief that the complexity serves the interests of the people that make a lot of money from the complexity, basically, the for-profit system.

Clark: What about all these extra benefits, like trips to the doctor, home-delivered meals? Do you think they really add value?

Berwick: Some do, some don't. Zero premiums add value to people that don't have a lot of money. That's why a lot of poor people go into Medicare Advantage. But downstream, they may discover that through pre-authorization and denials and restrictions, they don't get what they thought they would.

Clark: What should CMS do now to rein in upcoding, make the benchmarking and quality bonus programs reflect the care patients are actually getting? Do we just let MA plans continue gobbling up the eligible beneficiary market?

Berwick: I wish I had all the answers for you. At a minimum, first, the steps that CMS is now taking to begin to rein in this extraordinary overpayment, like the rate notices for last year and this year, which are leaning toward taking away some of the coding opportunities over a 3-year period, is now being attacked by the industry. CMS needs to be supported, encouraged. The plans are framing this as a takeaway – a decrease in payments to MA, and that is not true. MA plans are getting paid steadily more; not less.

The second part is harder, which is we need to improve traditional Medicare. We need to make changes in the co-payment structure so patients have less of a burden for hospitalization and outpatient costs.

You might ask, where will that money come from? Well, I've got an answer. How about clawing back and preventing some substantial proportion of those $88 billion in overpayments to MA plans? That would require Congressional action, so Congress has an opportunity to use those overpayments for higher purposes than lining the pockets of the MA plans. Maybe there are other things Congress would want to do with those resources – food and housing security, fossil fuel reduction, lower costs to beneficiaries, and more. If those agendas take money, well, there's the money: in current MA overpayments.

  • author['full_name']

    Cheryl Clark has been a medical & science journalist for more than three decades.