Medicare initiatives have been instrumental in improving care delivery and payment as exemplified by its role in broadly expanding the use of telehealth during the COVID-19 pandemic. Medicare innovations have been adopted or adapted in Medicaid and by private payers, while Medicare Advantage plans successfully compete with traditional Medicare only because their payment rates are tied by regulation to those in the traditional Medicare program. However, Medicare has not succeeded in implementing new, value-based payment approaches that also would serve as models for other payers, nor has Medicare succeeded in improving quality by relying on public reporting of measured performance. It is increasingly clear that burdensome attention to measurement and reporting distracts from what could be successful efforts to actually improve care through quality improvement programs, with Medicare leading in partnership with providers, other payers, and patients. Although Congress is unlikely to adopt President Biden's proposals to decrease the eligibility age for Medicare or to adopt a public option based on Medicare prices and payment methods in the marketplaces, the Biden administration has an opportunity to provide overdue, strategic direction to the pursuit of value-based payments and to replace failed pay-for-performance with provider-managed projects to improve quality and reduce health disparities.
Since the 1980s Medicare has been instrumental in advancing care delivery system and payment methods for the entire health system. At its very inception in 1966, the new Medicare program was directly responsible for desegregating hospitals for all patients throughout the US (Smith 2016). This past year Medicare played a decisive role in facilitating the greatly expanded use of telehealth during the COVID-19 pandemic by loosening telehealth payment rules and substantially increasing telehealth payment levels at the onset of the COVID-19 public health emergency (PHE) (DHHS 2020). These policy initiatives were quickly adapted by private payers, including Medicare Advantage (MA) plans, and in Medicaid. The substantial increases in telehealth payments not only provided access to care for patients seeking safety from COVID-19 exposure but also provided an indispensable infusion for financially strapped practices (Corlette et al. 2021).
Medicare has had notable success implementing new provider payment methods that have become models adopted by other payers. Medicare ties payment levels to the underlying costs that providers bear, rather than the much higher rates that hospitals and health professionals are able to negotiate with commercial health plans. Indeed, Medicare provides consumer choice of either the traditional Medicare program or care from private insurers, called Medicare Advantage plans, because the plans are able to peg their rates to the lower rates in traditional Medicare (Berenson et al. 2015). And Medicare has been a dominant influence on system-wide adoption of quality reporting and pay-for-performance approaches to accountability, a topic this article explores.
During the campaign, Joe Biden endorsed Medicare's central role in the US health care system by proposing both to make Medicare coverage available to individuals 60 to 64 years old and to provide a public insurance option relying on Medicare prices in the Obamacare marketplaces created by the Patient Protection and Affordable Care Act (ACA) (Oberlander, Singer, and Jones 2020). Given likely opposition to these proposals in the evenly divided Senate and the narrowly divided Democratic House, their prospects for congressional adoption are not bright. However, the Biden Administration can advance Medicare's role as steward of care delivery and payment though a combination of administrative actions and statutory provisions that should not raise the same level of ideological opposition as the campaign proposals, yet can move care delivery to higher value, not only for Medicare beneficiaries but also for the health care system overall.
Although the preceding discussion has emphasized Medicare stewardship successes, this article will address two areas where Medicare leadership has come up short. The article briefly summarizes the current challenges traditional Medicare faces in improving provider payment methods, in essence to reboot the stalled efforts to adopt value-based payment methods as alternatives, or perhaps more accurately, as complements to Medicare's current payment methods. The article then explores Medicare's unfortunate role in leading many payers to adopt conceptually and operationally flawed pay-for-performance approaches to improving quality and achieving accountability for spending. It concludes by proposing an alternative strategy for improving quality while also adopting a more feasible accountability approach.
Seeking the Holy Grail of Alternative Payment Models
Medicare's role as lead innovator in provider payment has been well described, initially with implementation in 1983 of the inpatient prospective payment system using diagnosis-related groups (DRGs), by which Medicare pays a fixed price based on the principal diagnosis responsible for the inpatient stay, rather than paying for each service provided (Mayes and Berenson 2008). DRGs, the resource-based Medicare Physician Fee Schedule implemented in 1992, and other prospective payment methods, many mandated in the Balanced Budget Act of 1997, have been adopted by other US insurance payers and adapted for use in many other countries (Busse et al. 2011).
While mostly known for the insurance expansions through the Obamacare marketplaces, the ACA also created the Centers for Medicare and Medicaid Innovation (CMMI) to develop and test alternative payments models (APMs) that seek to improve value—higher quality at lower cost—than achievable through the established payment methods.1 Furthermore, CMMI's mission was to develop APMs for broad adoption in traditional Medicare but, working through multipayer collaborations, to also be adopted throughout US health delivery.
CMMI basically adopted a “let a thousand flowers bloom” approach to identifying prospective APMs. This approach, unfortunately, has led to successful germination of only a few promising APM shoots able to reduce the rate of spending growth while showing improvement on the few important measures of quality in use. One recent commentary incisively suggested that CMMI's “dizzying array of options and rules may distract from the hard work of transforming care” (Chernew, Conway, and Frakt 2020). Furthermore, even the results of the many APM demonstrations are in dispute (Miller 2020; Navathe et al. 2020). CMMI's own evaluations of the payment demonstrations are sometimes incomplete or inconclusive.
In short, there is an emerging consensus that the Centers for Medicare and Medicaid Services (CMS) needs to adopt a more decisive, strategic direction for payment reform, but so far little agreement on what direction to take. The Biden administration will need to initiate a more transparent process for achieving consensus among the diverse payment policy opinions and among the various affected stakeholders, including many providers who have done well financially under the current payment methods and, thus, resist change. Furthermore, while CMMI has been mounting demonstrations of APMs, CMS has continued to improve design features of the established payment methods. The process for selecting a smaller number of the most promising value-enhancing payment methods should explicitly consider that the established payment methods for health professionals, hospitals, and other providers can also contribute to improved value (Berenson and Ginsburg 2019).
The Biden CMS will face an immediate imperative to develop a post-COVID-19 approach to paying for the expanded use of telehealth, which was tested under the duress of the COVID-19 emergency and quickly accepted as part of mainstream care delivery yet threatens explosive spending growth if continued to be paid fee-for-service under relaxed rules and with overly generous payments (Berenson and Shartzer 2020). The urgency of creating a viable telehealth payment approach can be the catalyst for reinvigorated efforts to accomplish what the ACA set in motion as value-based payment reform.
Medicare's Checkered History in Improving Quality
Until the mid-1980s, Medicare spent minimal effort measuring or working to improve quality of care provided to Medicare patients, beyond setting structural requirements intended to serve as minimal standards for quality assurance (Milgate and Hackbarth 2005). William Roper, the administrator of the Health Care Financing Administration, as CMS was then called, initiated public reporting of quality in 1986 with publication of Medicare hospital mortality statistics (Milgate and Hackbarth 2005).
A focus on quality was given a major boost at the beginning of the twenty-first century with a series of papers and reports that addressed the mediocre state of US health care quality. An influential review of quality studies revealed substantial gaps against accepted care standards, challenging the then common assumption about the superiority of US health care (McGlynn et al. 2003). In the same time period, two Institute of Medicine reports documented widespread quality and safety problems, recommending system-based, “continuous improvement approaches to address six dimensions of health care: patient safety, care effectiveness, patient-centeredness, timeliness, care efficiency, and equity” (IOM 2000, 2001).
However, instead of primarily supporting nascent, systems-based approaches to quality improvement, a follow-up IOM report called for government to establish measures of quality, to assess the performance of each health care provider against the measures, and to publish comparative data for use by consumers (IOM 2002). The report also called for linking payment to performance, proposing an ambitious timeline for adoption. Responding to this recommendation, Congress in 2003 set into motion statutory requirements for quality reporting, initially for hospitals but subsequently for physicians and other providers and for pay-for-performance (P4P), based on the reported measures (Epstein 2006).
Policy analysts lauded “merging financial incentives with a tremendous expansion in public profiling against a broad array of quality measures” (Epstein 2006). There seemed to be broad agreement that P4P was an idea whose time had come, although a New England Journal of Medicine editorial also expressed surprise—and caution—that early evidence did not show that P4P actually improved quality, despite its “compelling” logic (Epstein 2007). It should be noted that the health policy community accepted the rationale for P4P as compelling despite the fact that varying payment based on measured performance had had a checkered and controversial history in other sectors of the economy, such as for teachers and for executives of publicly traded companies (Heywood, Adams, and Rothstein 2009). Yet, under Congress's statutory direction, Medicare assumed the lead, providing a model of accountability relying on public measures and P4P, with other payers following Medicare's lead.
Only years later, after numerous studies indicated the lack of quality-enhancing success of public reporting and P4P approaches, did critics publicly raise concerns about both the conceptual underpinnings of P4P and the operational difficulties inherent in accurately using publicly reported measures to assess quality performance. Behavioral economists and various clinicians raised concerns that in response to public reporting and complementary, financial rewards and penalties, health professionals would likely replace their assumed professional ethos with an orientation more to their own pecuniary interests (Pink 2012; Woolhandler and Ariely 2012). The operational challenges to identifying useful measures, measuring accurately, and addressing provider “gaming” behavior, among a long list of problems, have been daunting and over many years have defied satisfactory solutions (Berenson, Pronovost, and Krumholz 2013).
Furthermore, the preoccupation with measurement perversely has led to the widely espoused policy view that “if you can't measure it, you can't manage it” and its corollary—that if a quality problem cannot be quantitatively measured, the problem either does not exist or cannot be fixed (Berenson 2016). To illustrate, serious errors in making correct diagnoses—failure to diagnose, wrong diagnosis, and the like—occur in 5%–15% of clinical interactions with providers across many conditions and provider types (IOM 2016). These estimates come from research; real time identification and measurement of diagnosis errors by their inherent nature are exceedingly difficult. The need to improve diagnostic accuracy, therefore, gets little policy attention, as do many other quality failures that cannot be easily measured. As succinctly stated by a practicing physician, “Ironically, metrics-based programs can undermine quality improvement by shifting resources and attention to measurement and reporting and away from actually improving care” (Goitein 2020).
Quality-of-care experts who initially had positive opinions about public reporting and P4P have been chastened by accumulating experience. For example, quality expert Elizabeth McGlynn, the lead author of the study documenting quality gaps 17 years earlier, recently concluded, “Despite nearly two decades of experimentation with standardized measurement, public reporting, and reward-and-penalty programs, average quality performance remains about the same” (McGlynn 2020).
Nevertheless, despite the lack of success, requirements for collecting and submitting quality and other performance measures continue and have proved costly and burdensome. A 2015 study estimated that US physicians spend more than $15 billion annually to report quality measures for their various payers (Casalino et al. 2016). Projecting this experience from physician practices to the many other health professionals, hospitals and other providers required to publicly report performance information, the payers collecting and analyzing the data, and the numerous data management firms that transmit and publish metrics (collectively referred to facetiously as the “measurement industrial complex”), one can conservatively estimate the health care measurement enterprise's annual cost as at least $50 billion.
In short, public reporting and P4P are not achieving their objectives but are failing at a high cost. Still, a bipartisan Congress doubled down on the approach in the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA, H.R. 2, 114th Cong.). Although the main purpose of the legislation was, finally, to repeal the failed sustainable growth rate policy to restrain Medicare Physician Fee Schedule spending growth, Congress decided it needed a “quid pro quo” from organized medicine. Figuring that its previously enacted Medicare P4P programs had suffered from insufficiently strong financial incentives, Congress, in MACRA, raised the stakes; physicians scoring low could suffer penalties of nearly 10% of their Medicare payments and, if highly successful, win the jackpot by receiving as much as 36% more (MACRA, H.R. 2, 114th Cong., 2015), bringing to mind the Catskills-resort-era joke, “the food here is terrible, and the portions are too small.”
Muting the impact of the statute, CMS initially managed to exempt nearly half of eligible practitioners from the Merit-based Incentive Payment Systems (MIPS), mostly because they saw too few Medicare beneficiaries for statistical validity and determined that only 5% of the remaining participating clinicians would suffer penalties, at up to 4% of their Medicare payments (Johnston, Hockenberry, and Maddox 2020), but their discretion to minimize the MIPS impact may be more limited going forward (Spivack, Laugesen, and Oberlander 2018). The program has produced perverse results, including rewarding clinicians serving fewer socially and medically complex patients (Khullar et al. 2020).
Changing the Focus of Accountability for Quality
To date, Congress has shown little interest in revisiting the MIPS part of the MACRA legislation, which also provided new incentives for clinicians to participate in APM demonstrations. The Biden administration should work with Congress to alter MACRA's flawed reliance on measurement and P4P. First, CMS should reverse the policy adopted in the late 1980s that changed the accountability approach from “cutting off the tail” of substandard quality to “raising the mean,” that is, altering the focus from weeding out “bad apples” to attempting to assure success of the whole crop (Bradley et al. 2005). Measurement of performance will help, but the measures do not have to be published or used as part of P4P.
Quality measurement on a national scale simply cannot reliably distinguish fair, good, and excellent, but all acceptable, care. Performance measurement in the hospice program provides a good example of the misguided use of public reporting of quality in Medicare. Since 2017, Hospice Compare has included seven measures that seek to gauge whether appropriate processes of care occurred at hospice admission. Virtually all hospices have scored above 90% compliance on six of the seven measures, that is, the performance measures have “topped out.” Yet, the data continue to be collected and published—and likely ignored. The uniformly excellent results on these measures suggest there are no hospice quality problems needing attention (MedPAC 2020). However, the Office of Inspector General (OIG) analysis of state survey agencies and accrediting organizations identified more than 300 hospice providers as poor performers in 2016 (OIG 2019); it appears the measures used do not reveal the substandard performance.
One well-chosen and available outcome measure could replace the process measures that occupy hospices' time and might actually be useful to beneficiaries and their families. That measure is the rate of live discharges from hospice. Hospice is designed for beneficiaries when a physician judges that the beneficiary likely will die within six months. Some level of live discharge occurs because beneficiaries may change their mind and disenroll, or their condition may surprisingly improve. The live discharge national average is 17% (MedPAC 2020).
But the hospices at the 90th percentile have live discharge rates exceeding 40%, suggesting either that the hospice is not meeting patients' needs or that the hospices admit patients who should not have been in hospice in the first place; they are discharged when they come up against the per-patient payment cap that cuts off continued Medicare payment for that beneficiary. In either case, high live discharge rates likely reflect unacceptable performance—or even fraud. Public reporting of this single measure might actually provide useful information for beneficiaries in selecting among hospices.
Resuming Quality Improvement Projects
For improving overall quality performance—raising the mean—Medicare should encourage and assist health professionals' and other providers' intrinsic motivation to improve the care they provide rather than rely primarily on financial incentives targeted to limited areas of quality (Berenson and Rice 2015). Medicare should emphasize quality improvement (QI) projects that providers would undertake to replace routine production of data for measurement—in both traditional Medicare and Medicare Advantage. Of course, QI does require internal performance measurement to assess progress in adopting quality-enhancing process improvements, but typically the measures used for internal QI are not the same as those used for public reporting (Berenson and Rice 2015).
Before turning its focus to public measurement responding to Congress's 2003 legislation, CMS had initiated QI projects for Medicare beneficiaries targeting major quality problems. For example, Medicare-funded Quality Improvement Organizations successfully worked with hospitals to improve outcomes for patients after heart attacks (Marciniak et al. 1998). These collaborative QI projects mostly were sidelined when measurement took over the field.
Almost one in two US adults has hypertension, with even higher rates in Blacks (Muntner et al. 2018). When inadequately treated, hypertension leads to congestive heart failure, end-stage renal disease, stokes, and dementia, among a longer list, all conditions that exemplify the substantial health disparities that have been exposed during the COVID-19 pandemic. Yet blood pressure control in the US has actually declined from 54% in 2013–14 to only 44% in 2017–18, which in turn is five percentage points below the level reported for 2007–8 (Muntner et al. 2020).2 Of note, the substantial decline in national blood pressure control has occurred even as measurement of blood pressure has been commonly required in public reporting and P4P programs.
Clinical experts recently have called for a national commitment to improve care of patients with hypertension (Adams and Wright 2020; Curfman, Bauchner, and Greenland 2020). Recent studies demonstrate that blood pressure (BP) measurement is more accurate when patients self-monitor their BP using their own home device; that medication administration at night, rather than morning, substantially improves outcomes; that more intensive treatment to achieve lower BP targets than previously considered normal improves outcomes; and that team-based care improves hypertension care. Yet, CMS quality and accountability programs, including in the MIPS and in a range of APMs being tested, still require a single office-based BP reading and an outdated target for BP treatment. Hypertension management could become one of the first clinical opportunities for the Biden CMS to replace public reporting and P4P with QI projects, actually improving outcomes for the entire population when clinicians adopt improved hypertension management approaches, initially for Medicare beneficiaries, while reducing health disparities at the same time. CMS should actively engage with other payers to make improved management of hypertension a national effort.
The two hallmarks of value-based payment are embedded incentives for more prudent spending and accountability for quality performance. The new administration has an immediate opportunity both to develop a transparent and explicit strategic direction for development and deployment of alternative payment models and to decrease current overreliance on quality measurement as the basis of accountability. Measurement, only sometimes with public reporting, should focus on substandard care, while acceptable but less than optimal quality should be improved by requiring providers to participate in quality improvement projects targeted to major health care problems. The administration also should work with Congress to accept the evidence-based verdict on pay-for-performance—its time has passed.
The Medicare Physician Fee Schedule stands out as a true payment for individual services. The payment models for other providers mostly use the same methods being tested in APMs, such as payment “bundles” that group services for a single payment. They are volume-based but not acutally fee-for-service.
Although the recommended target for blood pressure control was changed from 140/90 to 130/80 near the end of this period, the standard used for these comparisons maintained the higher level.