By John R. Graham
In March 2015, an overwhelming bipartisan majority in Congress voted for the Medicare Access and CHIP Reauthorization Act (MACRA). The so-called “doc fix,” a component of MACRA, was an attempt to fix the very flawed method Medicare uses to pay doctors and other health professionals. Unfortunately, MACRA is fiscally irresponsible and increases the federal government’s control over how clinicians practice medicine:
It is not paid for. Less than 4 percent of the increased spending authorized by MACRA is offset by other government spending cuts, resulting in an estimated $141 billion increase in the accumulated deficit over 10 years and $500 billion over 20 years, thus abandoning budget neutrality, a commitment previously made by both parties.
It significantly increases federal control of the practice of medicine. In line with the ambitions of Obamacare, clinicians will face increasing requirements to comply with federal regulations in order to get paid. These regulations will likely include greater reliance on government certified Electronic Health Records, which have already proven to frustrate doctors and do nothing to benefit patient care, despite an investment of $30 billion taxpayer dollars.
In May 2016, the Centers for Medicare & Medicaid Services (CMS) issued a 962-page proposed rule initiating the process of developing dramatically new ways to pay clinicians. Most clinicians will be subject to a new Merit-Based Incentive Payment System (MIPS), in which their compensation will vary as much as 18 percent, depending on their Composite Performance Score (CPS). The four categories of measurement comprising the CPS have evolved from previous attempts to influence physicians’ behavior that have had mixed results, at best. The weights given two of the categories will change over time:
• Quality achievement will be 50 percent of the composite score in 2019 and 30 percent in 2021;
• Resource use will be 10 percent of the score in 2019 and 30 percent in 2021;
• “Advancing care information,” the new name for Meaningful Use of Electronic Health Records will be 25 percent; and
• Clinical practice improvement activities will comprise the final 15 percent.
Clinicians’ scores will be distributed along a 100-point range and their incomes adjusted accordingly:
• In 2019, approximately $833 million in Medicare payments will be withheld from clinicians scoring below the yet-to-be-determined performance threshold and awarded to those scoring above it.
• This redistribution will amount to plus or minus 4 percent of reimbursements in 2019, and will increase to plus or minus 9 percent in 2022 and subsequent years.
• On top of this zero-sum game, higher performing clinicians will share up to $500 million in exceptional performance payments each year.
The new system grants the federal government’s imprimatur to quality measurements that are not always unanimously endorsed by the medical profession. Further, the cost of compliance is likely to drive more physicians into very large groups, which can handle the bureaucratic burden better than can small practices. This is only the first of many rules which will follow in the next few years. As the regulations evolve, there will be some limited opportunities for advocates of consumer-driven Medicare to improve them and reduce bureaucratic control. The opportunities which should be identified and advanced are:
• Integrating Medicare Part D (prescription drug) claims into Medicare Part A (hospital) and Part B (physician) claims so the value added by prescription medicines to health care overall is adequately recognized.
• Moving away from the Resource-Based Relative Value Scale to paying for “bundles” of care.
Neither of these changes is certain or even likely, as the rollout of the new physician payment scheme will be bogged down by interest group politics. Nevertheless, they are potentially positive developments.
Introduction: What Is the Problem with MACRA?
For over a decade before April 2015, Congress struggled with a reimbursement formula that paid physicians too little to ensure they would continue to see Medicare beneficiaries. At least once a year, Congress passed a short-term doc fix to prevent Medicare physician fees from dropping by about 20 percent. The last boost expired on March 31, 2015, necessitating rapid congressional response. The solution was MACRA.
There are two major reasons MACRA is a poor doc fix:
• It is not paid for. Less than 4 percent of the increased spending authorized by MACRA is offset by other government spending cuts, resulting in an estimated $141 billion increase in the accumulated deficit over 10 years and $500 billion over 20 years. thus abandoning budget neutrality, a commitment previously made by both parties.
• It significantly increases federal control of the practice of medicine. In line with the ambitions of Obamacare, clinicians will face increasing requirements to comply with federal regulations in order to get paid. These regulations will likely include greater reliance on government-certified Electronic Health Records, which have already proven to frustrate doctors and do nothing to benefit patients care, despite an investment of $30 billion taxpayer dollars.
Paying for Value, Not Volume
The stated purpose of the new rule is to move from paying Medicare physicians and other health professionals for volume to paying them for value through a Merit-Based Incentive Payment System (MIPS). Though “paying for value, not volume” has become a rote incantation in U.S. health policy, the notion that the federal government should determine whether a clinician is achieving this is actually a pretty recent development. Nevertheless, it has broad bipartisan support.
MACRA complements the Affordable Care Act (which established Obamacare) and previous legislation. President Obama confirmed this after he signed MACRA, saying, “I shouldn’t say this with John Boehner here, but that’s one way that this legislation builds on the Affordable Care Act, but let’s put that aside for a second.”2 Actually, this bipartisanship was not surprising, as Steven Findlay notes in Health Affairs:
“Laws passed between 2006 and 2010, including the Affordable Care Act (ACA), were forerunners to Congress’ approach in MACRA. For example, Congress created the Physician Quality Reporting System in 2006 and the Physician Value-Based Payment Modifier in 2010. And in 2009 Congress created the Electronic Health Record (EHR) Incentive Program. The ACA contains numerous provisions that promote transparency, accountability, payment reform, and quality improvement — including the creation of Physician Compare, a website mandated to, over time, contain comparative performance and quality measures on physicians. MACRA’s new physician payments system builds on and is synced up with these efforts.”
The new rule published by CMS on May 9, 2016, prepares the way for the implementation of clinician performance measurements beginning as early as January 1, 2017. These measurements — which are well beyond those currently existing in Medicare — will be used to reward or penalize clinicians stating in 2019. Given the changes proposed, this is a very short timeline.
Even the most sophisticated experts whose livelihoods depend largely on getting this right are struggling with the proposed rule. As John D. Halamka, chief information officer and dean for technology at Harvard Medical School, noted: “The 962 pages of MACRA are so overwhelmingly complex that no mere human will be able to understand them.”
Clinicians will be channeled into one of two payment streams:
• The vast majority of clinicians will be subject to the Merit-Based Incentive Payment System (MIPS). The proposed rule anticipates there will be 687,000 to 746,000 MIPS-eligible clinicians.
• Only 30,658 to 90,000 clinicians will be subject to Alternative Payment Methods (APMs).
MIPS is effectively the default. Although APMs could be more profitable, they require meeting high performance thresholds and will only be attempted by providers with a competitive advantage. Another way to look at it is that MIPS is mandatory and APMs are voluntary. Therefore, this issue brief only addresses the merit system, which will have a far greater impact than alternative methods.
Composite Performance Scoring
Under the merit-based payment system, clinicians will be scored in four categories that will be summed to a Composite Performance Score (CPS) ranging from zero to 100. The weights given two of the categories will change over time:
• Quality achievement will be 50 percent of the composite score in 2019 and 30 percent in 2021;
• Resource use will be 10 percent of the score in 2019 and 30 percent in 2021;
• “Advancing care information,” the new name for Meaningful Use of Electronic Health Records, a bugbear of physicians for a few years now, will be 25 percent; and
• Clinical practice improvement activities will comprise the final 15 percent.
(Note that three of the four measurements will have grown out of smaller, tentative programs CMS has rolled out in recent years.)
The merit system is mostly a zero-sum game. Clinicians’ scores will be distributed along the 100-point range and their incomes adjusted accordingly. In 2019, approximately $833 million in Medicare payments will be withheld from clinicians scoring below the yet-to-be-determined performance threshold and awarded to those scoring above it. This redistribution will amount to plus or minus 4 percent of reimbursements in 2019, and will increase to plus or minus 9 percent in 2022 and subsequent years. On top of this zero-sum game, higher performing clinicians will share up to $500 million in exceptional performance payments each year.
Three of the four performance measurement categories grew from previous attempts to “nudge” the behavior of clinicians. Following is a discussion of each.
The Quality Achievement Component. Although declining in weight significantly within the first three years of pay for performance, quality achievement will continue to be the largest component of the CPS.
It will also likely be difficult for the administration to get physicians to buy into CPS, given their resistance to its predecessor, the Physician Quality Reporting System (PQRS). Established in 2007, the PQRS offers much smaller financial incentives than the MACRA rule. As a result, many clinicians have ignored it, even though they are now being penalized financially for not complying. The problems with PQRS have compounded over the years and are recognized by even the strongest proponents of government-run medicine.
In 2013, Robert Berenson, of the Urban Institute, and a colleague, wrote:
“The meager rate of physician participation in the PQRS also suggests that something is fundamentally wrong — physicians simply do not respect the measures, and for good reason. PQRS measures reflect a vanishingly small part of professional activities in most clinical specialties. A handful of such measures can provide a highly misleading snapshot of any physician’s quality. Research shows that performance on specific aspects of care does not predict performance on other components of care. Primary care physicians manage 400 different conditions in a year, and 70 conditions account for 80% of their patient load. Yet a primary care physician currently reports on as few as three PQRS measures.”
The same year Berenson wrote this, Medicare began to fine clinicians 1.5 percent of their Medicare billing if they did not report quality metrics. (As with MACRA, the actual penalty was levied two years after measurement, in 2015.) The next year, the penalty was bumped to 2 percent. This negative incentive increased participation somewhat to 641,654 eligible professionals in 2013. However, 469,755 (42 percent) failed to report, and were penalized.
In the first year of the new performance score, quality achievement will account for 50 percent of 4 percent of the payment adjustment. After a few years it will account for 30 percent of 9 percent. Thus, its net contribution to the payment adjustment will increase from 2 percent to 2.7 percent. It is not clear whether this will be enough incentive to cause more physicians to invest significantly in complying with the quality reporting requirement.
While a number of old PQRS measures have been jettisoned, the basic design of the program remains the same. A provider must report six measures from either a master list or a specialty-specific list. Most measurements are described as process measures. One measure must be from a subset of 10 cross-cutting measures and one from a list of outcome measures (except for specialists for whom no outcome measures have been approved). Examples of each type of measure are:
• Process measure — the proportion of patients diagnosed with coronary heart disease prescribed aspirin or clopidogrel.
• Outcome measure — the proportion of diabetic patients with hemoglobin A1c under control.
• Cross-cutting measure — the percentage of patients with an advanced care plan or surrogate decision maker documented in their medical record (or who affirmatively declined to have one).
The final list of quality measures to be used for MIPS adjustment in the first year is expected to be published November 1, 2016, but it does not appear the new program is different enough from the previous PQRS to increase clinicians’ enthusiasm about participating.
Further, granting the federal government’s imprimatur to quality measures interferes with the practice of medicine (traditionally regulated by the states) and creates an illusion of unanimity in the medical profession that does not exist. For example, Daniel Musher, M.D., of Baylor College of Medicine, served on the Advisory Committee on Immunization Practice of the Centers for Disease Control and Prevention, which considered guidelines for a dual vaccine approach for pneumococcal vaccination for adults. He disagreed strongly with the published recommendation, but was prevented from publishing his opinion alongside the recommendation.
In 2009, the U.S. Preventive Services Task Force issued guidelines recommending annual mammograms for women starting at age 50, not 40 (as previously recommended). Needless to say, this upset many people. The American Cancer Society maintained its recommendation that preventive screening start at 40, as did the Mayo Clinic. Politicians took note, and made an exception in Obamacare for mammograms, such that the 2009 USPSTF revision was ignored when it came to Obamacare’s “free” preventive care. (In January 2016, USPTF reiterated its recommendation.)
The Resource Use Component. In plain English, resource use means costs, as measured by claims. Broadly, this category descends from the Value-Based Payment Modifier in the Affordable Care Act, which first took effect in 2013 (for payment in 2015). In order to avoid a 1 percent penalty on their claims, clinicians in large groups had to either deliberately report a measure approved by CMS or submit to administrative review of their claims. The measurements address hospital admissions for various acute or chronic conditions, and readmissions within 30 days of discharge. For example, if a practice’s diabetic patients are incurring too much costs in hospitals, this is interpreted as a signal of poor physician care for which the practice is penalized.
The challenge here is that clinicians drive more claims than they actually submit. Physicians admit patients to hospitals, and they prescribe medications. However, these claims are not captured by Medicare Part B. Allocating hospital (Part A) claims to physicians who drive them is (just about) manageable, because claims for Part A and Part B are both processed by administrative claims processors, usually subsidiaries of the private insurers that win contracts with CMS. These contractors do not bear much financial risk, nor do they usually determine fees. They process claims.
Medicare Part D, on the other hand, is run entirely by private insurers that bear financial risk. Further, because Part D plans are chosen by individual beneficiaries during open enrollment, there is no connection between a clinical practice’s patient panel and the Part D plans to which those patients belong. MACRA gives CMS the statutory authority to include Part D claims in the measurement of resource use; however, CMS declined to do so in the proposed rule.
The rule proposes three measures for resource use:
• Total per capita cost of Part A and Part B services used by patients attributed to a primary-care practitioner, whereby the patient is attributed to the primary-care practitioner who provided the most or most recent services.
• Medicare spending per beneficiary (MSPB), which attributes spending to physicians with the highest Part B claims during a period of hospitalization.
• Measurements of total costs for 41 episodes for which patients “trigger” expensive care which can be attributed to the clinician who sent a claim for the “trigger” event or billed at least 30 percent of the costs of an episode.
Beyond the current Value-Based Payment Modifier, the resource use category also descends from the current spate of payment reform initiatives launched by the Centers for Medicare and Medicaid Innovation (CMMI), an agency created by the Affordable Care Act. Energetic and innovative, CMMI is already testing over 20 different payment models within Medicare.
The new resource use category is evolving out of experiments bearing names like Accountable Care Organizations (ACOs) and Bundled Payments for Care Improvement (BPCI). Although still relatively new, these models do not appear to be achieving significant savings. The number of organizations participating in what are called Medicare Shared Savings Program (MSSP) Accountable Care Organizations has increased since they started in 2012. After paying out bonuses to providers, MSSP ACOs saved taxpayers $383 million in 2013 and $465 million in 2014. Whether those savings indicate success depends on how you look at it.
Total Medicare benefit payments amounted to $577 billion in 2013 and $597 billion in 2014. So, MSSP ACO savings are effectively irrelevant to current Medicare spending — less than one-tenth of 1 percent. On the other hand, net savings grew 21 percent while Medicare spending grew only 3 percent. That might count for something, so advocates of free-market reforms should encourage further experimentation by CMMI.
Participation in ACOs and other payment reforms has been voluntary and most providers bear only “one-sided risk.” That is, they can make more money but they cannot lose (other than capital they have put at risk to become an ACO). Only medical groups confident they can earn more by saving Medicare money participate. However, participation in MIPS will effectively be mandatory, so we can anticipate the vast majority which have not participated in experimental payment reforms will struggle to reduce resource use.
Two other conditions will continue to make it challenging to reduce costs via measuring resource use.
First, costs are assigned to clinicians retrospectively based on the patients they have seen. That is, after the end of the year, CMS will examine claims and charge them back to clinicians according to the rules summarized above. A simple example shows how unfair this might be. Suppose a Medicare beneficiary lives in New York for nine months and Florida for three months — a very realistic scenario.
He has a primary-care doctor in each location. After the end of the year, it is determined that slightly fewer than half his claims were incurred in New York and slightly more in Florida. So, Medicare will attribute all his claims to his Florida doctor’s measurement of resource use.
Very forward-thinking planners will respond this might cause the Florida doctor to enter into a private arrangement with the New York doctor whereby he will pay the latter to provide services that will reduce resource use. Markets can solve this challenge. However, it is not clear the financial rewards will be enough to cause the market to respond.
Further, the inability — so far — to integrate claims for prescriptions under Part D in the measurement of resource use is very limiting, because prescription spending generally substitutes for more expensive medical care, including hospitalization. Thus, higher prescription spending is generally associated with lower health spending overall.
The Advancing Care Information Component. One-quarter of the CPS will comprise this category, which is effectively the Meaningful Use of Electronic Health Records (EHRs). This was first imposed in 2009 and has become so painful for clinicians it has to be rebranded. The current program has spent $30 billion to install EHRs and has frustrated doctors and stymied innovation in health information technology.
The new category will measure six criteria:
• Protecting patients’ health information through Security Risk Analysis;
• Electronic prescribing; • Patient electronic access, including educating patients on their EHRs;
• Coordination of care through patient engagement in viewing, downloading, transmitting, secure messaging and incorporating patient-generated health data;
• Health Information Exchange of patients’ records, including requesting and accepting records, and reconciling clinical information; and
• Public health and clinical data registry reporting (including immunization registries).
Complying with previous criteria has proved increasingly challenging for physicians, and these new criteria are similar enough they do not address three challenges that have become apparent as EHR adoption has become institutionalized. First, EHRs are harmful to health professionals’ relationships with patients. This is exemplified by physician Mark Sklar, in a recent essay:
“The push to use electronic medical records has had more than financial costs.
“Yet to avoid future financial penalties from Medicare, I must demonstrate ‘meaningful use’ of the electronic record. This involves documenting that I covered a checklist of items during the office visit, so I spend 90 minutes each day entering mostly meaningless data. This is time better spent calling patients to answer questions or keeping updated with the medical literature.
“My practice quickly adopted the new Medicare requirements for electronically prescribing medications. Yet patients often do not want their prescription sent electronically.
“If I don’t electronically prescribe for a certain number of Medicare patients, I am penalized with a decrease in reimbursement that can rise to a maximum of 5%. Patients should have a choice in how their prescriptions are delivered, and physicians shouldn’t be penalized for how the patients choose.”
Perhaps the most influential report on EHRs was issued in 2014 by an independent scientific group. Referring to the rapid transition from paper to electronic health records, the report concluded: “Furthermore, there are questions about whether that transition will actually improve the quality of life, in either a medical or economic sense.”
This introduces the second problem, which is that federal payments did not compensate clinicians for the cost of installing, maintaining and using EHRs. Finally, one goal which remains out of reach is interoperability, the ability of EHRs from competing institutions to communicate with each other. Indeed, research indicates that Meaningful Use payments encouraged the adoption of EHRs that are deliberately closed to exchange with other parties.
The new rule does not appear to improve these challenges. It is more likely to exacerbate clinicians’ frustration because its payments are a zero-sum game, whereas financial incentives for EHR use to date have only been positive. Forty percent of respondents in a recent survey of physicians working in groups of fewer than 25 doctors anticipate the new requirements will lead to an “exodus” of small practice from Medicare, as they suffer “death by bureaucratic strangulation.”
The Clinical Practice Improvement Activities Component. Of the four categories, this one is the only one that does not descend directly from a previous program. Worth a maximum of 15 points in 100, this might be described as the “A for Effort” category. Clinicians will earn scores if they participate in activities which CMS believes are likely to result in improvement. There are high-weighted and mediumweighted CPIAs (worth half the score of the highweighted ones). Clinicians will achieve a maximum score by participating in three high-value or six threevalue CPIAs, although there are lower bars for certain clinicians in rural areas or with small practices.
An example is Patient-Centered Medical Homes (PCMH) for primary care. The PCMH model preceded the Affordable Care Act, first taking shape around 2007. At its most basic, the PCMH seeks to avoid problems like the one described above, in which a patient divides his time between New York and Florida. More generally, it seeks to create incentives for primary-care doctors to create medical “homes” that coordinate care with specialists such that patients get better care at lower cost.
In a culture where most primary-care doctors have preferred to refer patients to specialists they know, but each doctor bills Medicare independently in the Fee-For-Service system, PCMH models have had mixed success. A recent literature review indicated PCMH models had a moderately positive impact on the delivery of preventive care, moderate improvement in staff experience, some reduction in emergencydepartment visits, but no reduction in hospitalization. There was no evidence of overall cost savings.
Examples of high-weighted CPIAs include providing 24/7 access to MIPS-eligible clinicians, or having at least 60 percent of patients in year one (and 75 percent in year two) who are prescribed the anticoagulant warfarin also participate in an anticoagulation management program beyond taking their pills. Examples of medium-weighted CPIAs include collecting patient experience and satisfaction data, or using telehealth. A clinician can participate in a CPIA for as little as 90 days a year to receive points.
Although this is still the least well-defined category, it is likely that clinicians will quickly find out which CLIAs have the lowest bars to jump over, and focus on achieving them to the detriment of more complex ones.
Opportunities for Improvement
As it stands, the proposed rule appears likely to continue the trend of imposing more burdens on clinicians with mixed clinical results and little or no cost savings. Nevertheless, this is the way of the future, and CMS remains open to changes as the new payment system rolls out. Three opportunities for improvement may arise.
Include Medicare Part D (prescription drug) Claims in Medicare Part A (hospital) and Part B (physician) Claims Measurement of Resource Use. This has proved too complicated to figure out yet. Nevertheless, the statute encompasses this opportunity, and CMS is open to input on meeting the challenge. Including Part D claims in the measurement of resource use will better incentivize clinicians to prescribe appropriately to reduce overall resource use. Achieving this will be a significant technical challenge, but it will be worth it.
Move Away from the Resource-based Relative Value Scale to Paying for “Bundles” of Care. Although MACRA wants to pay for value, not volume, the new payment system is still based on the old fee schedule. This is especially problematic with respect to the resource use category. William Hsiao, the economist who designed the current Resource-Based Relative Value Scale, originally determined the fees as follows:
“He put together a large team that interviewed and surveyed thousands of physicians from almost two dozen specialties. They analyzed what was involved in everything from 45 minutes of psychotherapy for a patient with panic attacks to a hysterectomy for a woman with cervical cancer. They determined that the hysterectomy takes about twice as much time as the session of psychotherapy, 3.8 times as much mental effort, 4.47 times as much technical skill and physical effort, and 4.24 times as much risk. The total calculation: 4.99 times as much work. Eventually, Hsiao and his team arrived at a relative value for every single thing doctors do.”
This would make a Soviet bureaucrat blush. Until the government allows prices to be determined by normal market forces, resources cannot be used effectively. Although the statute does not jettison the fee schedule, the measurements of resource use imply clinicians should be paid for “bundles” of care around diagnoses. In this sense, it approximates a 1983 reform to the hospital fee schedule that paid hospitals by DiagnosisRelated Group (DRG). Advocates of further payment reform should emphasize that paying for “resource use” rather than every single service enables Medicare to move from paying fees for activities to paying fees for diagnosis-related episodes of care, and encourage the current fee schedule to become vestigial.
MACRA is a bipartisan reform to Medicare that is unlikely to be dislodged in the foreseeable future. However, it significantly increases federal intervention in the practice of medicine, and fails to take prescription drug spending into account when measuring value in health care. Nor does it get rid of the Soviet-style price fixing that characterizes the current physician fee schedule.
Future regulatory and legislative reforms must do both more and less than the currently proposed rule does. They must reduce the role of the federal government in setting fees for physicians and determining what quality is, while continuing to move the locus of control to patients and doctors, by continuing to move away from paying for individual procedures toward paying for episodes of care.