Bill Gates, Mathematical Averages, and Making Harmful Cuts to Medicare Advantage

By Susan Dentzer, MS, President and Chief Executive Officer, America’s Physician Groups

The punchline of an old joke is that whenever Bill Gates walks into a bar, everyone present becomes a millionaire — on average. That’s the problem with averages: Although they may be accurate mathematically, they don’t illuminate the many variations that lie underneath.

That’s also the problem with the statement released by the U.S. Department of Health and Human Services this week about proposed changes in the Medicare Advantage (MA) program that, on average, would increase payments to the plans that deliver MA benefits by 1 percent in 2004.

Mathematically, it may well be true that the proposed regulation that the Centers for Medicare & Medicaid Services (CMS) unveiled recently would on average increase those payments to plans by 1 percent. In the spirit of high-school math teachers who always urge students to “show their work,” it would have been helpful if CMS had released far more information about its calculations to demonstrate that this final result — 1 percent — is indeed accurate, as well as to show the variable impact on different health plans and providers.

This 1 percent average masks a wide range of potential effects that could make the picture look very different across the range of MA plans — with equally variable impact on providers and patients. One preliminary analysis by Wakely Consulting’s Value-Based Payment practice estimates a range of as much as a 20 percent decline in payments from the government to some MA plans to as much as a 10 percent increase for others. So alongside any Bill Gates-level characters who may fare well under these changes, there are others who will be walloped financially. These unfortunates include many disadvantaged older adults who rely on MA not just for their health care, but for a range of benefits that provide them with nutritious food, transportation, and other items essential to their health and wellbeing.

Here’s why.

What CMS attempts in its proposed rule is to address longstanding concerns about a feature of MA that, quite simply, has had both positive and negative effects. Unlike in traditional fee-for-service (FFS) Medicare, in which providers simply treat patients and then bill Medicare for each service provided, in MA the government sends lump-sum, “capitated” or per-person monthly payments to health plans for everyone who signs up for their MA programs. As with all financial incentives, capitation has potentially positive and negative effects. One danger is that insurers could profit by withholding needed services from people. Another potential negative effect is that insurers could be discouraged from signing up MA beneficiaries who are very sick, and whose care would be costly — possibly causing any health plans enrolling them to lose their shirts.

As a result, MA is equipped with various features designed to blunt the impact of these perverse incentives. One is an elaborate “Star Rating” quality measurement system that essentially forces out of business any MA plans that earn fewer than four out of five possible stars. Another is risk adjustment, which ensures that payments from Medicare to MA health plans are sufficient to support the health needs of sicker beneficiaries.

To account for such risk, the lump-sum payments to health plans are adjusted based on beneficiaries’ demographic characteristics, such as age, and their level of health or sickness. Their health status is determined through careful examinations by clinicians, who diagnose the enrollees for any ailments and submit diagnostic codes to Medicare. Medicare then “risk adjusts” the payments to reflect the level of care that will be needed. No comparable level of coding goes on in the traditional Medicare program, which many experts have characterized as “under-coded.” Translation: Although the government has a pretty good handle on diagnoses of the MA population, by contrast, it isn’t clear just how sick or not the roughly 30 million people in the traditional Medicare program actually are.

Risk adjustment in MA also has both positive and negative aspects, and has unarguably spurred some providers and health plans to “upcode” or “overcode” diagnoses — in effect, exaggerating beneficiaries’ conditions — to get more money from Medicare. The program has guardrails to protect against excessive coding, but these can operate with a lag and don’t always work perfectly. What’s more, there’s often a fine line between thoroughly documenting MA enrollees’ diagnoses — particularly if they are older, sicker, and suffer from multiple chronic conditions — and overcoding. And in any case, risk adjustment offers an advantage — to patients and arguably, to society — that is nonexistent in the traditional Medicare program, in that clinicians actually need to conduct comprehensive exams of patients to determine their true health status and code them appropriately.

These exams are far more in-depth than what typically occurs in the average “Welcome to Medicare” visits in the traditional Medicare program. The detailed diagnostic codes that clinicians participating in MA send along to Medicare far better reflect actual patients’ care needs than the billing codes generated from a typical office visit for patients in the traditional Medicare program. Payments to MA plans — and in turn, the payments that plans make to providers — are then tailored to meet these more clearly delineated care needs. Meeting their needs can include not just providing clinical care such as office visits or hospitalization, but also addressing enrollees’ health-related social needs. For example, many MA plans and providers draw on their Medicare payments to deliver and pay for healthful food to diabetes patients, or supply regular transportation to disabled older adults who would otherwise be confined to home.

The net effect of all these program features is that, according to the Medicare Payment Assessment Commission, MA spending was about 4 percent higher than expected FFS expenditures for similar beneficiaries as of 2021. The differential has led to broad recommendations that MA program spending be cut. CMS is taking these recommendations to heart, and has now chosen to overhaul the risk adjustment system as the means of curbing differences between MA and FFS. CMS argues that despite this significant overhaul, payments to MA plans under its latest proposals would still rise 1 percent on average. But returning to the “Bill Gates walks into a bar” analogy, let’s look at what would actually happen to the other “barhoppers” in the MA picture.

My organization, America’s Physician Groups, represents more than 360 large physician groups around the country that care for nearly 30 percent of all MA patients. Much of the time, they’re providing routine primary care to older adults with multiple chronic conditions. Often, these are relatively low-income people who don’t always look much like the vibrant and active senior citizens featured in health plan advertising. Without their regular check-ins with our doctors and care teams, their conditions would probably worsen, and they would wind up even sicker, in hospitals or nursing homes, and probably die sooner as a result.

When these MA enrollees first see our doctors, they are diagnosed and “coded” so that health plans and clinicians will be paid to provide the appropriate levels of care. In a dizzyingly complex set of changes, CMS has now decided to downgrade many of these codes, so that both health plans and clinicians will in effect be paid less to care for the same patients. CMS says that the effect of all these code changes — both upgrades and downgrades — will be about negative 3 percent on average. But once again, the average masks variable effects that would be higher or lower for some MA plans and providers. Although some of our physician groups will not be hard hit, others estimate that they could face payment cuts of up to 10 percent to 20 percent. Many of these groups are in so-called value-based arrangements in which they are accountable for costs and quality, and caring for the same sick and disadvantaged people that CMS purports it wants to help.

As our groups stare at this cold math, they wonder how they will maintain the level of care they provide to these patients, especially at a time when their own costs for staff and supplies are soaring, and they are scrambling to find enough medical assistants and others who can help them provide care. Some of our members will bear the force of these cuts directly because they are in “delegated” arrangements with health plans, in which they directly assume the risks of caring for MA patients. Others anticipate that the MA plans whose enrollees they care for will respond by slashing the payments that they make to our providers.

What else will have to go on the chopping block when these payments are cut? Will our doctors have to curb office hours or cut telehealth visits, or close their inner-city or rural clinics? Slash their own incomes or those of their staffs, or undertake layoffs that will worsen access to care? Take away some of those extra benefits that supply food or transportation to patients in need, or eliminate valuable programs to help people with pre-diabetes avoid progressing to full-blown diabetes? Charge lower-income beneficiaries more in out-of-pocket payments so that our medical groups can make the numbers add up and stay in business?

Our medical groups are the first to acknowledge that the MA program is far from perfect, and that the government should make changes over time. We all have a stake in paying plans and providers appropriately, toward the ultimate goal of getting Medicare beneficiaries the care that they truly need. A good starting point would be recognizing that the traditional Medicare program is also far from perfect, and that constantly comparing the two programs as if they are accomplishing the same goals for Medicare beneficiaries is nonsensical, because they aren’t. We’ve asked CMS to put the changes it is now proposing on hold for a year to evaluate their likely effects — not on average, but across the distribution of plans, providers, and patients, including ours. It’s also essential to the integrity of the program to continue to pursue and punish bad actors who do take advantage of the risk adjustment system to fraudulently inflate diagnostic codes and reap too much payment.

But please, CMS — don’t treat everybody involved in the MA program as similarly “average.” Whether you want to admit it or not, you’re threatening cuts for many — and it’s primarily older adults and people with disabilities who are enrolled in Medicare Advantage who will suffer as a result.