Managed Care Reflections: Insights From Richard J. Gilfillan, MD; and Donald M. Berwick, MD, MPP

Am J Manag Care. 2025;31(2):52-53. https://doi.org/10.37765/ajmc.2025.89675

A Wolf in Sheep’s Clothing: How the Medicare Advantage Money Machine Corrupted Value-Based Care

Despite 50 years of public policy initiatives intended to improve health care delivery, America’s health care system persists in delivering poor outcomes at twice the cost of other countries.¹ Advancements in medical care, medical technology, and pharmaceuticals have significantly improved outcomes for patients with serious illnesses. Decreased poverty rates² have improved but not eliminated health inequities. And the effects of the COVID-19 pandemic were mitigated significantly as America’s health care workforce, leaders, and the pharmaceutical industry performed heroically. Americans’ average life expectancy has improved by more than 5 years over the past 55 years.³ However, since 1980 it has steadily fallen behind that of other developed countries. We now live about 4 years less on average than those in comparable countries, with persistent differences across racial and ethnic groups.^3,4 Higher system costs are now passed on to consumers through reduced coverage, resulting in higher family medical debt. Increased financial burdens, barriers to access, and administrative complexity underlie the widespread frustration and anger expressed following the tragic and inexcusable murder of UnitedHealthcare CEO Brian Thompson. This system underperformance is not for lack of public policy efforts. It results from the co-optation of the policy initiatives expected to drive improvement and a failure to capture the mindset of health care delivery system leaders. The co-optation of value-based payment and value-based care is the latest example of this reality; it functions today as the sheep’s clothing hiding the underlying wolf of finance-driven health care, seeking profits and not care improvement.

Managed Care Origins

Managed care, an outgrowth of the Health Maintenance Organization (HMO) Act of 1973, gradually morphed from a clinical strategy to deliver better care for patients into managed payment, a financial strategy to improve insurer and provider finances. Managed care originally referred to prepaying the expected total cost of care for a population to intermediaries, typically primary care–based clinical organizations such as group practice HMOs (eg, Kaiser, Harvard Community Health Plan, and Group Health Cooperative). Many also used their own or a limited number of tightly aligned hospitals. These organizations were expected to manage patients’ care to improve their health and lower total costs. Restructuring care delivery was essential to achieving those outcomes. They were paid a monthly amount equal to the expected total cost of care for their members. If the actual care costs less, they could keep the savings or a portion thereof. Prepaid capitation payments provided the funding for the group practice to invest in care redesign.

In the 1980s, the increasing success of these models led Blue Cross and other traditional insurers to offer HMO products using independent physicians. Employed physicians in group model HMOs were naturally aligned with the plan objectives of decreasing costs and improving care. In some markets, most notably California, insurers sought to similarly align independent physicians by using full and partial risk capitation contracts. These groups developed intensive primary care (IPC) models that provided extended services, care coordination, and quality improvement capabilities all aimed at reducing costs and improving outcomes. (These models grew organically in at-risk medical groups. Although not originally called “intensive primary care,” they were the model for the IPC firms that have become prominent in the past 10 years. We use the term retrospectively to refer to the common shared approach of these varied firms.) In most markets, insurers used fee-for-service (FFS) payments supplemented by incentive payments for primary care providers (PCPs) to improve costs and quality. (Both authors were practicing physicians in this era, Berwick at Harvard Community Health Plan in Boston and Gilfillan with Medigroup Mercer, a Blue Cross Group and independent physician association HMO model in New Jersey. Gilfillan also became a medical director with Independence Blue Cross and participated in the development of many of the managed payment efforts described herein.)

Employers and consumers found these broad-based networks more attractive, and gradually insurance-based organizations replaced provider-based organizations as the dominant intermediary. These managed care organizations (MCOs), driven by their employer and government customers’ singular priority of decreasing their health benefit spend, focused on just reducing costs. Decreasing the unit price and the number of services paid for are fast and predictable ways for insurers to decrease costs. Gradually these administrative levers became managed care and care redesign withered.

Managed Care Provider Pricing: A Market Failure

Initially, insurer HMOs offered higher volume to a limited number of providers in exchange for lower prices. Local Blue Cross plans and HMOs with heavily concentrated market share demanded and achieved significant discounts for their managed care products. But as managed care grew in the 1990s, virtually all hospitals had to participate and accept lower rates for the same volume. Gradually, national insurers such as UnitedHealthcare and Aetna obtained lower rates that allowed them to compete with the Blues.

In the early 2000s, hospitals with strong market power pushed back, demanding—and receiving—markedly higher rates. These “must-have” providers typically derived their market power from their location in well-to-do communities where corporate decision makers lived. Human resources departments negotiating employer-sponsored health insurance (ESI) contracts with insurers required them to include the hospitals used by executives and their families. The result was a bifurcated delivery system—hospitals serving high-income communities received high rates while those in lower-income communities took much lower rates. These hospitals also had a much greater percentage of lower-paying Medicare and Medicaid business, thus accentuating their disadvantage.

The market-based ESI sector has failed to contain hospital pricing. The Congressional Budget Office estimates that commercial payments are more than 200% of Medicare today and that the trend in commercial hospital prices has been 50% higher than Medicare’s.⁵ Ironically, Medicare and Medicaid privatization efforts have only succeeded because both programs require providers to accept government-set rates for patients assigned to MCOs. Price negotiation allowed MCOs to disrupt indemnity insurers in the 1980s and 1990s. But since then, the primary result has been to create winner and loser hospitals, which has led to worsened health care inequities and driven hospital consolidation.

Managed Care Delays and Denials

Managed care’s second major strategy to manage costs was to decrease the number of health care services it paid for. Insurers created administrative processes to determine what services were “medically necessary” and what provider claims should be paid. Managed care leaders and clinical staff pursued preauthorization activities believing that they were improving care for patients. But these activities gradually morphed from preventing unnecessary care to becoming an obstacle to people getting needed care and to providers being paid. Decisions were made by insurer employees operating with a bias towards denying and waiting for appeals. Given that providers and patients appealed only 10% of the time,⁶ the vast majority of denials stood. While patients frequently received unexpected bills when providers went unpaid, in the 1990s MCOs were recalcitrant and unmoved.

In privatized Medicaid, delays and denials were the dominant cost control mechanism from the very beginning in the 1980s. States expected to save money by paying MCOs less than the state’s cost. Medicaid unit prices were already low and there was little opportunity to decrease utilization since most of the cost was for obstetrical and neonatal care. So, Medicaid MCOs were forced to rely on their ability to decrease the number of services they pay for. The Office of Inspector General has demonstrated that Medicaid MCOs deny 1 in every 8 preauthorizations.⁷ Because Medicaid-insured individuals had no ability to pay provider claims, MCOs could simply deny provider claims with no concern about members being billed. Because Medicaid reimbursement rates are very low in most states, providers don’t make them a priority for appealing. Medicare plans at least had the hope of saving money by decreasing the utilization of hospital and postacute inpatient services. The initial focus was on care management, heavy preauthorization, and redirection to less intense sites of care. But as Medicare went through cycles of decreasing payment, plans gradually became more dependent on aggressive claims denials and downgrades.

Managed Care Favorable Selection

Plans also tried to decrease the number of services by attracting healthy people and avoiding sicker people. In employer groups, early HMOs were usually offered alongside traditional indemnity plans. Savvy sales teams, PCP selection, referral requirements, and limited networks (eg, excluding oncologists and comprehensive cancer centers) led sicker people to stay in the indemnity plan while healthier people chose the HMO. Plans also used additional benefits like gym memberships to attract healthy members. Wherever possible, typically in the individual market, plans used preexisting conditions exclusions and medical underwriting to limit their exposure to sicker patients.⁸

Managed Care Backlash

In the mid-1990s, legions of stories about HMO bad judgments made managed care notorious for denying vital care to members. At the same time, employers, realizing that the risk pools in their indemnity programs were deteriorating, started mandating managed care enrollment for all employees. The result was the managed care backlash captured in the 1997 movie As Good as It Gets. Regulators reacted with more patient-friendly rules requiring patients to be held harmless from most claims denials.⁹

Managed Payment

In the face of the consumer backlash and stepped-up regulatory oversight, some insurers relaxed their managed care activities in the early 2000s and tried new disease management approaches.¹⁰ But patients did not enroll in these, and the impact was minimal. With unit costs rising due to provider market power, insurers gradually resumed and extended both their preauthorization and claims denial activities. With managed care growing rapidly, virtually all providers had to participate. MCOs accelerated the development of new claims editing, claims denial, and downgrading activities to simply decrease what they paid. Provider contracts mandated acceptance of any new insurer payment policies and prohibited balance billing patients. Although some must-have providers pushed back, most acquiesced and tried to recoup payment through appeals.

In the early 2000s, rising costs ultimately pushed employers and insurers to consumer-directed health plans that used high deductibles and coinsurance to make patients “more effective consumers” of health care. Studies before¹¹ and since¹² have shown that this approach leads patients of all types to delay both necessary and unnecessary care. Not surprisingly, they also show that there is a strong negative impact, including some increased mortality, for lower-income individuals. They may well be a contributor of unknown magnitude to America’s declining health. These plans basically transfer expenses from payers to individuals. Nevertheless, this trend to underinsure people is now the norm for most employer and Affordable Care Act (ACA) Exchange benefit products.

Today, despite the talk of how digital health, machine learning, and artificial intelligence might transform health care, insurers have used IT primarily to amplify their preauthorization and claims denial activities.¹³ One recent study showed that Medicare Advantage (MA) denial rates increased 56% between 2022 and 2023.¹⁴ UnitedHealthcare explained in a recent analysts call that downgrading claims from inpatient to outpatient status, unrelated to actual care, adds approximately $5 billion a year to their bottom line.

Consumers had seemed to adapt to these realties like the proverbial frog in the pot of boiling water until the recent online outpouring of anger. A well-intended reform policy of managed care had morphed into a nightmare managed payment scheme. It makes care hard to get and impossible for many to afford. It is administered through a vast administrative superstructure built on different proprietary policies and processes that vary by payer and change frequently. It leaves providers, patients, and families confused and frustrated. And it is the reason why America spends at least 25% of every health care dollar on administrative expenses¹⁵—at least 3 times more than comparable countries.¹⁶

The Rise of Value-Based Payment and Care

As insurer-based managed care efforts to address escalating costs fell out of favor, attention turned toward ways to engage providers more directly in improving health care quality and costs. CMS led the way in the early 2000s with pay-for-performance (P4P) programs like the CMS Premier Hospital P4P demonstration (2003) and the Medicare Physician Group Demonstration Project (PGP) in 2005. P4P soon morphed into value-based payment and care as the new approach to improve health care costs and quality.^17,18 In the insurance world, value had been defined as providing better care and improved customer satisfaction at a lower cost. Health care value was codified by the Institute for Healthcare Improvement in 2008 as the Triple Aim: simultaneously delivering better health care to individuals, improved health outcomes for populations, and lower per capita costs.¹⁹ At this time attention turned to primary care providers as the foundation of value-based care.

IPC and the Patient-Centered Medical Home as Value-Based Care Models

Managed care insurers have a long history of paying PCPs for value as a way to align participating physicians with their cost management goals. The California capitated IPC model described above was successful in managing total costs and improving quality metrics such as the Healthcare Effectiveness Data and Information Set (HEDIS). In other locations, managed care insurers paid PCPs using incentive systems that included cost and quality measures as a way of aligning them with the HMO’s goals.20

The patient-centered medical home (PCMH) model had a very different origin. It started as a clinician-driven medical home intended to provide comprehensive care for children with serious illnesses. Combined with the Chronic Care Model described by Wagner and colleagues,²¹ the PCMH focused on improving care for patients with chronic illnesses. Key elements included evidence-based care, patient-informed team-based care, planned visits, activated patients, and some care coordination. In 2006, the Primary Care Collaborative was established to spread the PCMH model as a more effective and efficient health care delivery model.²²

Unlike the IPC model, the PCMH model initially did not target the total cost of care. It did seek alternative payment models from insurers to support the additional operational costs. A 2010 AJMC article,²³ “Value and the Medical Home,” described Geisinger Health System’s ProvenHealth Navigator model, which combined elements of the IPC and the PCMH models. It was a partnership between Geisinger’s health plan and primary care group practice. The medical group provided expanded primary care services focused on delivering the Triple Aim. The health plan supported the PCP group with additional capitated payments to support office-based nurse coordinators, population-based data analytics and incentives payments tied to Triple Aim metrics. Final evaluation of the model showed total cost of care savings of 7.9%, primarily a result of 20% fewer inpatient admissions.²⁴

As the Obama administration and Congress developed the ACA, a central question was how to drive improvement in the delivery system. Existing PCP models like IPC and PCMH and emerging value-based payment and care initiatives like the CMS Premier and PGP demonstrations, Blue Cross of Massachusetts’ Alternative Quality Contract, and various health system value initiatives²⁵ all informed the ACA’s answer.¹⁸

The ACA, Value-Based Payment, and Health System Transformation

The ACA ushered in a new era calling for the creation of a patient-centric high-value health system. Driven by Triple Aim–focused value-based alternative payment models (APMs), providers were expected to move their clinical and business models from a volume to value focus. (Although the Triple Aim label was not used, the population-based metrics of total cost of care, patient experience of care, and improved quality outcomes are replete throughout the legislation.) Most providers had experience with quality and patient experience improvement but little to no experience with or understanding of the total cost of care. Nevertheless, the ACA was replete with proposed value-based payment models for all major provider types. The most significant was the Medicare Shared Savings Program (MSSP) Accountable Care Organization (ACO) program. The Center for Medicare and Medicaid Innovation (CMMI), with an initial 10-year budget of $10 billion, was created and given the authority to test and then scale successful value-based payment and care models nationally. The MSSP ACO, CMMI’s Pioneer ACO, Bundled Payments for Care Improvement (BPCI) episode-based payment, and Comprehensive Primary Care (CPC) models were the first APMs implemented. The stage seemed well set to drive broad system transformation.

PCMH and CMMI Primary Care Models

The CPC primary care model had more in common with the PCMH model than the IPC full risk model. It attempted to include multiple payers, but in most practices, it involved less than 40% of the patients. It basically layered a small value-based operation on top of an FFS practice model with only marginal extra support to deliver the expected extended services. This is distinctly different from the most effective California IPC models that were designed and staffed aggressively to reduce costs and improve quality metrics.

To date, none of the CMMI PCP models have demonstrated significant net savings or improvement in quality metrics.²⁶ There has been some evidence of decreased chronic medical disease admissions, but if there was any impact on total costs it was not enough to cover the incremental population payments. The evidence on quality is about same. These outcomes are consistent with the results reported for other PCMH initiatives, which show very little impact on Triple Aim dimensions.²⁷ One positive finding has been that PCPs operating within ACOs did decrease the total cost of care. PCP-centric Medicare Shared Savings Program (MSSP) ACOs have also produced significant savings and at a higher rate than hospital and specialist-centric ACOs.²⁸

Notwithstanding the CMMI experience, CMS has remained committed to supporting primary care.²⁹ It recently announced a new hybrid payment approach for PCPs focused on improving care for patients with chronic diseases.³⁰ The rationale was based on several US reports of positive results from intensive PCP models3¹ as well as international observational studies that show that countries with higher percentages of PCPs are associated with lower costs and some better outcomes.^27,32

Episode-Based Payments

The BPCI program initially garnered great interest and participation, with about 10% of hospitals participating. Although episode costs did decrease significantly, the programs did not result in savings for CMS primarily because providers were allowed to pick and choose the episodes. The model did significantly alter the way hospitals viewed their interactions with postacute providers. Many hospitals for the first time built networks of aligned post-acute providers. Subsequent iterations of BPCI continue, with CMS recently announcing the mandatory Transforming Episode Accountability Model, which is currently being implemented.

ACOs

The MSSP was broadly accepted and grew rapidly to cover about one-third of traditional Medicare (TM) beneficiaries. It has consistently produced savings from the program perspective, particularly by physician-centric ACOs. However, the overall MSSP savings are well below what was expected. The Pioneer ACO model was successful in reducing costs and certified by CMS but not scaled nationally, despite the extant authority in the ACA. The CMMI Next Gen ACO model had borderline results. Its successor, ACO Realizing Equity, Access, and Community Health (REACH), is reporting significant savings. Independent researchers have questioned the degree to which the reported ACO savings were real.³³ Although there is some evidence that more recently favorable physician selection drives some of this success, there is also evidence of positive impact.^34,35 We believe that the presence of individual ACOs that have repeatedly created meaningful savings over the last 10 years is demonstration of proof of concept. Providers can come together and improve the Triple Aim for populations. But the ACO program to date has underperformed because it offers a very weak business alternative to the incumbent FFS model. CMS has tried to address some of the concerns about shortcomings in the MSSP ACO and ACO REACH models with the introduction of a regional component into the benchmark. These opportunities have created some new interest, but the longer-term prospects remain limited because of the ongoing CMS policy to rebase benchmarks to eliminate prior-year savings.

Value-Based Payment and Care Transformation Stalled

The ACA assumption was that APMs would drive providers to deliver value-based care. In 2013 the Health Care Transformation Task Force, a nonprofit group of providers, payers, patient advocates, employers, established the goal of having 75% of payments channeled through APMs by 2020. CMS established a similar goal shortly thereafter. In 2024 the CMS Learning and Action Network estimated that only about 28% of total system payments flow through 2-sided risk APMs and 45% through upside-only APMs. But both of these estimates are well overstated because as we will describe below, MA, by far the segment cited as having the highest penetration, is counting contracts that are actually the opposite of value based. Without these contracts, the metric would about 20%. The APM percentage for commercial and Medicaid business has barely reached 20%. The TM percentage, primarily reflecting ACO lives, has increased somewhat as TM shrinks due to MA growth. The number of MSSP ACOs has decreased from its peak in 2020 and the total of Medicare ACO beneficiaries been flat in the range of 13 million for the past 5 years.³⁶ Our own conversations with population health leaders across many integrated health systems confirm that in health systems real value-based care is either stalled or being rolled back.

Why Value-Based Care Transformation Stalled

Post ACA, there was real excitement about value transformation. Many large integrated health systems made significant investments in population health, participated in the various APM models, and had some real success in ACOs. But gradually health systems decreased their investment as they realized that real value transformation was not necessary.

There are 5 major reasons for this conclusion. First, APMs left the traditional FFS payment systems in place. APM shared savings are paid through a reconciliation process after the FFS claims are paid. Participants could make limited investments, collect their usual FFS payments, and wait 18 months to see if there were any shared savings. In this wait-and-see context, no one was really forced to change their basic operations. They could just add limited APM operations and pursue their usual business. Second, APM participation was voluntary. Many providers decided not to participate. Participants who were not successful could simply drop out. Third, the shared savings business model was very weak compared with the familiar and successful FFS model. FFS profits today were much preferred to uncertain and small APM profits 18 months later. In fact, many of the theoretical shared savings of the future were their own forgone FFS revenue today. Why give up 100% of the revenue today for a potential return of 50% to 70% of it 18 months later? The argument that they would be avoiding the variable costs of care averted was never really convincing. Worse yet, CMS would ultimately keep those savings by rebasing their ACO benchmark. Revenue in the hand today was always preferred to potential small profits in the bush tomorrow. Fourth, commercial and Medicaid payers did not engage with the models, so at least 50% of most health systems’ revenue would be FFS driven. Health systems, like most organizations, only change if they must. Lacking any forcing function, they did not transform. The dependable FFS model remained dominant.

Most systems had learned this and begun pulling back by 2020 when COVID-19 struck. The pandemic demanded a reordering of priorities. In most institutions, reduced population health teams have persisted post COVID-19. Fortunately, many continue in MSSP to maintain PCP relationships, obtain useful CMS data, and hedge their bets regarding future value-based care initiatives.

The ACO model presented a very different opportunity for PCP-based ACOs, particularly those managed by ACO-enabling firms. Although many of the shortcomings are still relevant, the upside of shared savings is that they are totally incremental to their traditional physician payments. Although difficult for small practices to do alone, ACO enablers have created a business model that works. They bring small investments, typically supporting data analysis, office training, and know-how that allow physicians sometimes with extra staff to persistently deliver reduced costs and improved quality.³⁷ The requirement for downside risk enhances the role of enablers as they insulate small practices from what would otherwise be catastrophic risk. CMS’ unfortunate insistence that small practices accept downside risk has had the effect of driving PCPs into larger corporate firms, hospitals, or ACO-enabling firms.

The good news is that we do have population health management operating across the health system. It just is stalled at the current level. It is stalled because incumbent health systems felt no need to change, physician ACOs have not yet achieved the scale needed to be a transformational driver, and most payers have not embraced real value-based payment for non-Medicare populations. Finally, the fifth and perhaps most significant reason care transformation is stalled is because all of these providers and insurers became focused on a much more lucrative and easy business: the MA risk score game.

The MA Money Machine and the Co-Opting of Value-Based Care

Privatized Medicare has always cost taxpayers more, never less. This has usually been the result of favorable selection. The 2006 introduction of risk adjustment was intended to align an MA plan’s payment with the illness burden of its population. It did just the opposite. In MA, each member has a risk score based on their different medical diagnoses. CMS calculates an expected additional cost for every diagnosis using TM data and total spending. But TM providers get paid per service, not per diagnosis. They don’t submit every diagnosis, only enough to be sure they are paid. When CMS uses the smaller number of diagnoses to calculate the expected additional cost for each diagnosis, it is inflated. Naturally, plans then go to great lengths to find every diagnosis to increase patient’s risk scores. They end up being paid more per diagnosis and for more diagnoses than they should be. The result is huge overpayments. The Wall Street Journal (WSJ) recently reported³⁸ that MA plans on average increased patients’ risk scores about 25% above TM in the first year and a total of 30% after 3 years. That would deliver additional payment of approximately $2000 per member per year above CMS’ average cost of $14,000. The Medicare Payment Advisory Commission estimates that risk score gaming and favorable selection created MA overpayments of $83 billion or 22% in 2024.³⁹ Legislation also requires CMS to pay MA plans another $23 billion in direct subsidies for quality and county bonuses. The total subsidies of $106 billion pay for virtually all the additional benefits that fueled MA’s rapid growth to cover 54% of beneficiaries.

These subsidies and the underlying risk gaming processes have created an MA distortion field that directly impacts most segments of the health care system. An entire industry of risk coding firms was created to aid the effort. One such firm, Signify Health, which facilitates code collection through home visits, was acquired by CVS in 2023 for $8 billion. The Office of the Inspector General found that plans submitted codes worth $5 billion that were only documented via nonphysician home visits and chart reviews.⁷

Most significantly, insurers have created incentives for providers to submit more diagnoses. The percentage of premium full-risk contract is the most powerful tool. It uses a medical cost target based on a percentage (eg, 85%) of the insurer’s premium from Medicare. Providers keep the difference between the target and the actual total medical costs. When the provider submits more diagnoses, the risk scores, premiums, and target go up. Medical costs don’t change since the patients are the same, so the extra premium generates a massive guaranteed provider profit. This MA “Money Machine”⁴⁰ contract can convert a break-even PCP practice into one with a $1 million per physician profit. The WSJ reported that physicians employed by UnitedHealthcare’s Optum subsidiary increased risk scores by 71% for their own customers.³⁸ A risk score that high could provide UnitedHealthcare about $5 billion in additional revenue for every 1 million MA members in Optum practices. They currently report 4 million patients in fully capitated arrangements in those practices, but they do not specify how many are in MA.

MA Plans Acquire the Money Machine

IPC Models Optimize Risk Coding

The Money Machine model that the WSJ describes began in the at-risk California IPC models described above.⁴⁰ Prior to 2006, they were successful because they made long-term investments in redesigning care. Optimizing risk scores was a natural extension of their existing operations and made sense because they could capture up to 85% of the increased payment as pure profit.³⁸ Insurers, getting only 15% of the increased premium, saw that if they owned the practices they would get 100% of the risk coding profits. UnitedHealthcare’s Optum began to acquire these groups starting with AppleCare Medical Group in 2010 followed by Monarch HealthCare and Memorial Health in 2011. Wellpoint (now Elevance) acquired CareMore Health Group in 2011 as well. Non-insurer Davita Inc acquired Healthcare Partners for $4.4 billion in 2012 and ultimately sold it to Optum in 2019. And these physician practice subsidiaries were unregulated with no need to meet the ACA 85% medical loss ratio (MLR) requirement. The excess profits can be consolidated into the parent’s profits as medical expenses so they circumvent the MLR rule.

Optum has been acquiring PCP groups at a rapid clip and promotes itself as having the largest contingent of employed and affiliated physicians with more than 75,000.⁴¹ In a striking and sad irony, the most profitable of these acquisitions result from acquiring nonprofit groups like Atrius and Reliant medical groups in Massachusetts. Both are direct descendants of original nonprofit HMO group practices described earlier. The acquisition of for-profit groups requires a direct negotiation with the owners to determine a fair price. These prices have dramatically increased due to the Money Machine effect. In 2022 Optum reportedly paid $3 billion to acquire the for-profit Kelsey Sebold Medical Group in Texas with about 40,000 MA lives and 400 physicians. Conversely, when acquiring nonprofits, a firm only has to pay what the state decides is the value of the nonprofit assets into a successor non-profit foundation. The result with Atrius was a payment of $236 million in the foundation, about 10% of the valuations of similar for-profit practices like Kelsey-Seybold.^42,43 Massachusetts citizens did not get true market value. Now Atrius, bought on the cheap like the Reliant medical group before it, has become a for-profit firm structured to generate UnitedHealth’s Money Machine profits.

MA Gold Rush

Venture capital (VC) investors took notice of the Money Machine. In 2012, Oak Street Health became the first VC-backed, Medicare-focused primary care company. With attractive and well-staffed PCP offices and an aggressive marketing campaign, it grew rapidly. Percentage of premium contracts with MA insurers was the foundation. Easy and regular access to the PCP office was a real improvement for patients and facilitated the collection of more diagnoses. Although managing medical costs was important, it was not easy, very labor intensive, and results were not guaranteed. An uncertain 2% savings in medical costs seems less significant when compared with a much smaller investment in coding that yield a certain 35% more premium that becomes pure profit. The Money Machine model was the only path to the return on investment their investors expected.

It is hard to overstate the influence of the MA risk coding game on health care system investment and innovation. The acquisition of the California groups and Oak Street’s launch highlighted the easy MA arbitrage profits, leading to a gold rush of MA investment. This included PCP vehicles like Landmark Health, Agilon Health Inc, Iora Health, and Cano Health as well as new MA insurers like Clover Health, Bright Health, and Devoted Health. In total, MA-related startups generated total valuations of over $75 billion and direct investments of more than $10 billion from 2019 to 2023.

MA Risk-Score Gaming Made Real Value-Based Care Look Unattractive

The CMS ACO models did not offer a significant risk-score gaming opportunity in any way comparable to those in MA and required significant investment and ongoing operational costs. Sophisticated Pioneer ACO participants such as Heritage Physician Network recognized that after the first settlements in 2013. Ultimately they exited the model early and focused more on MA. Many health systems continued in the MSSP model to obtain CMS data but minimized their investments and looked to MA. The calculus was and remains simple. MA offers guaranteed easy profits for just coding better. MA risk coding can be optimized through regular care processes with supplemental IT help. Direct profits can be 10% or more and are totally predictable. ACOs require millions in start-up investments and operating costs. Reaching breakeven requires real medical cost savings of at least 5% to 6%, which are exceptional and typically take 2 to 3 years. Success is uncertain at the start and may not be sustainable because of benchmark rebasing—the “ratcheting” down of payments by CMS when the contract is renewed. MA plans get $15 billion annually in quality bonus payments that are a subsidy above FFS spending. Quality scores can provide an additional 5% payment. ACO quality scores are used in exactly the opposite way by CMS: to ratchet down ACOs’ earned shared savings payments. Over the past 10 years, most large health care systems have gradually directed more of their investments toward MA rather than ACOs—either starting their own plan or partnering with established plans.

But not all health systems can obtain advantageous Money Machine–like contracts. Typically, it is must-have providers that either have their own plan or have been able to demand favorable contracts from established MA plans. For many providers, MA is the worst payer because of the denials and particularly the downgrading of inpatient to outpatient services.

Many physician ACOs, whether free-standing or with enabling companies, have sought MA contracts and have benefited from the Money Machine model. For-profit enabling companies can’t help but seek this business as their investors see the much easier profits available from their MA investments.

The result is a perverse policy outcome. ACOs, a successful value-based payment and care model that makes Medicare more sustainable, is not sustainable for participants. The heavily subsidized MA program delivers easy profits for insurers and some providers, but is unsustainable for CMS and taxpayers. In effect, we continue to push providers and beneficiaries into the more costly program that offers more limited choice of providers for patients and more administrative pain for providers.

Gap Closure Rates Become Executive Team Metrics

Star Ratings remained critical to payers for the additional 5% premium and the extended marketing opportunities success brings. Plans flood providers with lists of patients who lack documentation of a particular Star-related service—“quality gaps.” They send similar lists of risk score diagnoses to be documented—"coding gaps”. They drive attention to closing these “gaps” with various incentive plans based on the degree to which they close these quality and coding gaps. The “gap closure rate” has become a key metric in senior system leadership incentive plans. The hard work of actually improving care has been supplanted by the “gap closure chase,” with frontline clinicians spending time documenting preventive services or irrelevant diagnoses.³⁸ Population-based admission rates or emergency department visit rates, key targets for real value-based care, are generally not even reported.

MA Provider Contracts Are Not Value-Based Payment and Inhibit Value Transformation
MA plans and providers are fond of calling their percentage of premium and other incentive plans value-based payment and care. In non-MA business, they can be when used to align PCP groups. But when risk score gaming converts these arrangements into tools to create higher costs for CMS and taxpayers, they are not value-based—they are value-destroying contracts. In the MA population, serious efforts to improve care are abandoned for the much more lucrative risk-score coding opportunity and perform to the test quality gap closures. It takes a good dose of cynicism or willful ignorance to call these activities value based.

In 2025, the MA plan business model is narrowly focused on 4 key operating activities that are the antithesis of value-based care:

1. Risk score optimization to increase revenue
2. Preauthorization to delay and deny clinical services
3. Claims payment delays, denials, and downgrades
4. Pharmacy benefit management that profits from obscuring the price of drugs

The net result of this operating approach is for CMS to pay 22% or $83 billion above what traditional Medicare would spend for the same population.⁴⁰ Total MA subsidies of $106 billion have created a distortion field that forces every health care organization to meet plans’ needs, not patient needs. The value-based payment/care approach has been co-opted and become the latest short-circuit path to profits for insurers and well-positioned providers. The policy goal of health system value transformation has again been lost. Lurking beneath the sheep’s clothing of intensive primary care start-ups, value-based payment, and value-based care is the same financial wolf driving health care costs up and higher CMS spending. The recent consumer rage against insurance companies shows that at least some patients intuitively understand the nature of the beast.

Conclusion

Despite 50 years of public policy efforts to deliver better health care at lower costs, health care in America remains twice as expensive and produces inferior outcomes as those of similar countries.¹ Incumbent payers and providers have been successful in co-opting policy efforts because their legacy business models remained in place:

• Risk code gaming, delays, denials, and downgrades in payment for payers
• FFS volume and market power pricing for must-have providers

We have all adapted to a managed payment system that subjects everyone to a dizzying health care maze structured to make it difficult for patients to get care and providers to get paid, and a world where providers with market power increase prices that become the debts of families whose insurance leaves them exposed. We have created a system where dedicated primary care providers leave practice because they are driven more by financial interests of firms than the needs of patients.⁴⁴ We remain far from the goal of a patient-centric high-value health system.

But we have also made some progress. Government intervention via the ACA has increased access for millions. Government-set prices have dramatically limited the total costs for the Medicare and Medicaid populations, which now represent 56% of health care spending.⁴⁵ In addition to the extraordinarily committed traditional health care workforce, we now have thousands of individuals experienced in and excited about improving the health of populations. The ACA led to the creation of a broad-based population health infrastructure including these people and systems and knowledge that did not exist in 2010. It is simply underpowered because providers treat it as a marginal business and are distracted by MA easy money. While somewhat dormant in many organizations today, it is poised to become the foundation for a real value-based care system, if we can create a compelling business case for organizations to adopt it as their goal.

Reestablishing a Path to a Patient-Centric, High-Value Health System

The failure to create a high-value, accountable health care system results from our failure to capture the mindset of health care leaders. Most are committed health care professionals and executives interested in doing well for their organizations and the people they serve. They face enormous challenges today in just operating their traditional health care businesses. We cannot make light of that and they deserve our gratitude. But we need more from our health care system. So, we need to construct an operating context that supports their current work and compels them toward building patient-centric, high-value care. Charting the way forward will require broad discussion including patients, providers, and policy makers.⁴⁶ We offer the following suggestions to be considered in such a discussion.

1. Eliminate the MA distortion field:

a. CMS should declare its intent to undertake a major initiative to create a new risk-adjusting system that is not susceptible to manipulation by providers or plans and commission a project to that end.

b. Pending that new system, CMS should prohibit all contractual arrangements that reward increased coding by providers including percentage of premium contracts.

c. Congress should level the playing field by creating an improved TM benefit to compete more effectively with MA:

i. Eliminate current sources of MA overpayment using contract-specific coding intensity factors and a “favorable selection” factor to reduce MA benchmarks.

ii. Concurrently, use the resulting MA savings to create Medicare Plus, a modernized TM benefit with integrated Parts A, B and D, a maximum out-of-pocket cap (MOOP), and some supplemental benefits. These improvements will flow through to MA benchmarks, thereby allowing plans to continue to offer most of their current benefits.

d. Congress should order CMS to create a mandatory MA administrative contractor to provide standardized and evidence-based prior authorization and claims adjudication for all MA plans. This would provide an unbiased decision maker, eliminate the proliferation of proprietary processing systems, increase competition for national plans, and reduce complexity and administrative costs for all.

2. Convince provider leaders to build high-value organizations by increasing population-based payments and decreasing FFS payments:

a. Mandate provider participation in APMs, particularly ACOs.

b. Use federal authority to drive commercial and Medicaid payers to use APMs.

c. Mandate systematic use of appropriateness screening tools for high-cost services.

d. Prohibit the use of productivity-based incentive models (eg, relative value units) that drive higher service utilization.

e. Improve ACO models to create a sustainable business model:

i. Include only provider organizations

ii. Provide a full capitation alternative

iii. Create reasonable guardrails for smaller ACOs

iv. Eliminate rebasing, gradually decrease costs using lower trend factors

v. Require meaningful roles for clinicians and patients in governance

f. Gradually decrease FFS reimbursement to make it less attractive.

3. Redefine the quality metric approach to decrease the number of metrics but shift the focus to patient-driven outcome measures and community-based population health measures.

4. Minimize administrative complexity and costs and increase insurer competition by requiring all insurers to use a standard Commercial Administrative Contractor for all preauthorization and claims adjudication activities.

5. Expand current efforts to improve PCP payment to support Integrated Primary Care to include behavioral health, community health workers, and care coordination.⁴⁶

6. Create an all-payer rate system for nongovernmental coverage to assure fair provider reimbursement for all populations and constrain the market power of providers and insurers. (Eliminating provider pricing power may allow for appropriate consolidation that could deliver more integrated care.)

7. Mandate that states increase Medicaid rates to Medicare levels.

8. Create an all-payer database to include all lines of business and use data to create radical transparency for providers and payers.

9. Create an ACA-exchange non-insurer Medicare Plus coverage product for anyone who wants it including those with employer sponsored insurance.

10. Create a Health Services Accountability Board for systemwide oversight similar to the Federal Reserve Board.⁴⁶

11. Decrease the role of for-profit firms in health care delivery.

a. Prohibit the acquisition of established nonprofit providers by for-profit and private equity firms.

b. Regulate the debt-to-equity ratio of health care for-profit firms to avoid Steward Health–like outcomes.⁴⁷

REFERENCES

1. Blumenthal D, Gumas ED, Shah A, Gunja MZ, Williams RD II. Mirror, Mirror 2024: A Portrait of the Failing U.S. Health System. The Commonwealth Fund. September 2024. Accessed January 27, 2025. https://www.commonwealthfund.org/sites/default/files/2024-10/Blumenthal_mirror_mirror_2024_final_v4.pdf

2. Historical poverty tables: people and families – 1959 to 2023. US Census Bureau. Updated September 10, 2024. Accessed January 27, 2025. https://census.gov/data/tables/time-series/demo/income-poverty/historical-poverty-people.html

3. Rakshit S, McGough M, Amin K. How does U.S. life expectancy compare to other countries? Peterson-KFF Health System Tracker. January 30, 2024. Accessed January 27, 2025. https://www.healthsystemtracker.org/chart-collection/u-s-life-expectancy-compare-countries/

4. Dwyer-Lindgren L, Baumann MM, Li Z, et al. Ten Americas: a systematic analysis of life expectancy disparities in the USA. Lancet. 2024;404(10469):2299-2313. doi:10.1016/S0140-6736(24)01495-8

5. The prices that commercial health insurers and Medicare pay for hospitals’ and physicians’ services. Congressional Budget Office. January 2022. Accessed February 19, 2025. https://www.cbo.gov/system/files/2022-01/57422-medical-prices.pdf

6. Levinson DR. Medicare Advantage Appeal Outcomes and Audit Findings Raise Concerns About Service and Payment Denials. HHS Office of Inspector General; September 2018. Accessed January 27, 2025. https://oig.hhs.gov/oei/reports/oei-09-16-00410.pdf

7. Murrin S. Some Medicare Advantage Companies Leveraged Chart Reviews and Health Risk Assessments to Disproportionately Drive Payments. HHS Office of Inspector General. September 2021. Accessed February 20, 2025. https://oig.hhs.gov/oei/reports/OEI-03-17-00474.pdf

8. Pear R. H.M.O.’s warned on recruiting only fit Medicare clients. New York Times. June 11, 1999. Accessed February 19, 2025. https://archive.nytimes.com/www.nytimes.com/library/politics/061199hmos-elderly.html

9. Patients’ Bill of Rights. Office of Personnel Management. Accessed February 19, 2025. https://www.opm.gov/healthcare-insurance/healthcare/reference-materials/bill-of-rights/

10. Felt-Lisk S, Mays GP. Back to the drawing board: new directions in health plans’ care management strategies. Health Aff (Millwood). 2002;21(5):210-217. doi:10.1377/hlthaff.21.5.210

11. Newhouse JP.Consumer-directed health plans and the RAND Health Insurance Experiment. Health Aff (Millwood). 2004;23(6):107-13. doi:10.1377/hlthaff.23.6.107

12. Zheng Z, Jemal A, Banegas MP, Han X, Yabroff KR. High-deductible health plans and cancer survivorship: what is the association with access to care and hospital emergency department use? J Oncol Pract. 2019;15(11):e957-e968. doi:10.1200/JOP.18.00699

13. Shryock T. Cigna faces second class-action suit over automated claims denials. Medical Economics. September 21, 2023. Accessed January 27, 2025. https://www.medicaleconomics.com/view/cigna-faces-second-class-action-suit-over-automated-claims-denials

14. Skyrocketing hospital administrative costs, burdensome commercial insurer policies are impacting patient care. American Hospital Association. September 2024. Accessed January 27, 2025. https://www.aha.org/system/files/media/file/2024/09/Skyrocketing-Hospital-Administrative-Costs-Burdensome-Commercial-Insurer-Policies-Are-Impacting-Patient-Care.pdf

15. Sahni NR, Mishra P, Carrus B, Cutler DM. Administrative simplification: how to save a quarter-trillion dollars in US healthcare. McKinsey & Company. October 20, 2021. Accessed February 19, 2025. https://www.mckinsey.com/industries/healthcare/our-insights/administrative-simplification-how-to-save-a-quarter-trillion-dollars-in-US-healthcare

16. How does the U.S. healthcare system compare to other countries? Peter G. Peterson Foundation. August 16, 2024. Accessed February 19, 2025. https://www.pgpf.org/article/how-does-the-us-healthcare-system-compare-to-other-countries/

17. Robinson JC. Slouching toward value-based health care. Health Aff (Millwood). 2008;27(1):11-2. doi:10.1377/hlthaff.27.1.11

18. Rosenthal MB. Beyond pay for performance—emerging models of provider-payment reform. N Engl J Med. 2008;359(12):1197-200. doi:10.1056/NEJMp0804658

19. Berwick DM, Nolan TW, Whittington J. The triple aim: care, health, and cost. Health Aff (Millwood). 2008;27(3):759-769.

20. Value-based provider incentive payments included in May self-funded invoices. Independence Edge. April 28, 2016. Accessed February 19, 2025. https://media.whatcounts.com/ibc_mktgcomm/2016IndependenceEdge/April282016/042016_Edge_BDTC_Customers.html

21. Wagner EH, Austin BT, Von Korff M. Organizing care for patients with chronic illness. Milbank Q. 1996;74(4):511-544.

22. About us. Primary Care Collaborative. Accessed January 27, 2025. https://thepcc.org/about/

23. Gilfillan RJ, Tomcavage J, Rosenthal MB, et al. Value and the medical home: effects of transformed primary care. Am J Manag Care. 2010;16(8):607-614.

24. Maeng DD, Khan N, Tomcavage J, Graf TR, Davis DE, Steele GD. Reduced acute inpatient care was largest savings component of Geisinger Health System’s patient-centered medical home. Health Aff (Millwood). 2015;34(4):636-644. doi:10.1377/hlthaff.2014.0855

25. Paulus RA, Davis K, Steele GD. Continuous innovation in health care: implications of the Geisinger experience. Health Aff (Millwood). 2008;27(5):1235-1245. doi:10.1377/hlthaff.27.5.1235

26. O’Malley A, Singh P, Fu N, et al. Independent Evaluation of Comprehensive Primary Care Plus (CPC+): Final Report. Mathematica. December 2023. Accessed February 19, 2025. https://www.cms.gov/priorities/innovation/data-and-reports/2023/cpc-plus-fifth-annual-eval-report

27. Shi L. The impact of primary care: a focused review. Scientifica (Cairo). 2012;2012:432892. doi:10.6064/2012/432892

28. Greiner A, Muhlestein D, Kempski A, Fillippi MK, Huffstetler AN. PCC 2024 Evidence Report: An Evaluation of Primary Care in Medicare Accountable Care Organizations. Primary Care Collaborative; November 2024. Accessed January 27, 2025. https://thepcc.org/wp-content/uploads/2024/11/Primary-Care-The-MVP-of-MSSP-2024-Evidence-Report.pdf

29. HHS is taking action to strengthen primary care. HHS. November 7, 2023. Accessed February 19, 2025. https://www.hhs.gov/sites/default/files/primary-care-issue-brief.pdf

30. CMS, HHS. Medicare and Medicaid programs; CY 2025 payment policies under the Physician Fee Schedule and other changes to Part B payment and coverage policies; Medicare Shared Savings Program requirements; Medicare prescription drug inflation rebate program; and Medicare overpayments. Fed Regist. 2024;89(147):61596-62648.

31. Maeng DD, Graham J, Graf TR, et al. Reducing long-term cost by transforming primary care: evidence from Geisinger’s medical home model. Am J Manag Care. 2012;18(3):149-155.

32. Friedberg MW, Hussey PS, Schneider EC. Primary care: a critical review of the evidence on quality and costs of health care. Health Aff (Millwood). 2010;29(5):766-772. doi:10.1377/hlthaff.2010.0025

33. McWilliams JM, Chen A. Understanding the latest ACO “savings”: curb your enthusiasm and sharpen your pencils – part 1. Brookings. November 12, 2020. Accessed February 19, 2025. https://www.brookings.edu/articles/understanding-the-latest-aco-savings-curb-your-enthusiasm-and-sharpen-your-pencils-part-1/

34. Trombley MJ, McWilliams JM, Fout B, Morefield B. ACO investment model produced savings, but the majority of participants exited when faced with downside risk. Health Aff (Millwood). 2022;41(1):138-146. doi:10.1377/hlthaff.2020.01819

35. Lyu PF, Chernew ME, McWilliams JM. Benchmarking changes and selective participation in the Medicare Shared Savings Program. Health Aff (Millwood). 2023;42(5):622-631. doi:10.1377/hlthaff.2022.01061

36. APM measurement effort. Health Care Payment Learning & Action Network. 2022. Accessed February 19, 2025. https://hcp-lan.org/workproducts/apm-infographic-2022.pdf

37. Gonzalez-Smith J, Crook H, Singletary E, Bleser W, Saunders R. How to better support small physician-led accountable care organizations: recent program updates, challenges, and policy implications. Duke Margolis Center for Health Policy. February 2020. Accessed February 19, 2025. https://healthpolicy.duke.edu/sites/default/files/2020-07/How%20to%20Better%20Support%20Small%20Physician-Led%20ACOs.pdf

38. Weaver C, Wilde Mathews A, McGinty T. UnitedHealth’s army of doctors helped it collect billions more from Medicare. Wall Street Journal. December 29, 2024. Accessed January 27, 2025. https://www.wsj.com/health/healthcare/unitedhealth-medicare-payments-doctors-c2a343db

39. Johnson M, Berwick D, Gilfillan RJ. Ending overpayment in Medicare Advantage. Center for American Progress. March 19, 2024. Accessed January 27, 2025. https://www.americanprogress.org/article/ending-overpayment-in-medicare-advantage/

40. Gilfillan R, Berwick DM. Medicare Advantage, direct contracting, and the Medicare ‘money machine,’ part 1: the risk-score game. Health Affairs Forefront. September 29, 2021. Accessed January 27, 2025. https://www.healthaffairs.org/content/forefront/medicare-advantage-direct-contracting-and-medicare-money-machine-part-1-risk-score-game

41. Competition is getting bigger! United Healthcare Optum is now the largest physician group in the country boosting 90,000 physicians. Healthcare Leadership Institute. November 30, 2023. Accessed February 20, 2025. https://www.ctileadership.com/united-healthcare-optum-90000-physicians/

42. Morse S. Massachusetts AG agrees to $236 million Optum and Atrius Health merger. Healthcare Finance. April 25, 2022. Accessed February 20, 2025. https://www.healthcarefinancenews.com/news/massachusetts-ag-agrees-236-million-optum-and-atrius-health-merger

43. Newitt P. Behind Optum’s physician acquisition strategy. Becker’s ASC Review. April 27, 2023. Accessed February 20, 2025. https://www.beckersasc.com/asc-transactions-and-valuation-issues/behind-optums-physician-acquisition-strategy.html

44. The plight of primary care, part 1—ITT episode 12. N Engl J Med. 2023;389(1):e1. doi:10.1056/NEJMp2303852

45. Martin AB, Hartman M, Washington B, Catlin A; National Health Expenditure Accounts Team. National health expenditures in 2023: faster growth as insurance coverage and utilization increased. Health Aff (Millwood). 2025;44(1):12-22. doi:10.1377/hlthaff.2024.01375

46. Berwick DM, Batchlor E, Chokshi DA, et al. From laggard to leader: why health care in the United States is failing, and how to fix it. Health Aff (Millwood). 2025;44(2):179-186. doi:10.1377/hlthaff.2024.01007

47. Krueger H, Bartlett J, McCarthy B. Inside the secret financial dealings of Steward CEO Ralph de la Torre. Boston Globe. September 3, 2024. Accessed January 27, 2025. https://www.bostonglobe.com/2024/09/03/metro/steward-health-ralph-de-la-torre-secret-financial-dealings/