The current desire to transition to value-based payment models is more easily understood when placed in context with national expenditures for health care. From 1980 to 1990, Medicare compensated physicians based on a system of usual, customary, and reasonable charges. This system implicitly increased expenditures by rewarding providers who were able to establish high billing profiles. In addition, this system created inequalities based on specialty and geographic location; in that decade, healthcare expenditures increased by a staggering 182.2%, from $255.3 billion to $721.4 billion.1 This prompted the creation of a Physician Payment Review Commission, which, by Congressional mandate, researched alternatives to the then-existing model. The Commission ultimately endorsed the adoption of a standardized fee schedule based on the Resource-Based Relative Value Scale, which was codified in the Omnibus Budget Reconciliation Act of 1990.2,3 Over the next decade, the growth rate in healthcare expenditures was cut by more than half, to an average of 8.9% annually. However, despite this decline, healthcare expenditures continued to be an expanding component of the national Gross Domestic Product (GDP); by the year 2016, healthcare expenditures accounted for 13.3% of GDP, increased from 12.1% and 8.9% in the years 1990 and 1980, respectively (Figure 1).
The growing share of GDP consumed by healthcare costs continued to concern federal lawmakers, and led to the adoption of the Sustainable Growth Rate (SGR) formula in the Balanced Budget Act of 1997.4 In essence, through a series of somewhat complex calculations that accounted for changes in price structure and the number of Medicare beneficiaries, the SGR formula linked physician reimbursement to changes in the GDP. This was not an issue during the 1990s, when GDP grew at an average of nearly 5.6%; however, in 2001, GDP growth fell to just over 50% of 2000 levels.5 This triggered SGR-mandated payment cuts to providers of 4.8% in 2002.
As can be seen in Figure 2, after 2000, annual percent change in GDP did not correlate with rising healthcare expenditures (Pearson’s r = 0.377). Consequently, allowable expenditures under the SGR continued, to languish; from 2012 to 2015, Congress addressed pending physician cuts proscribed by the SGR formula through 17 separate temporary patches costing over $150 billion, which did nothing to solve the fundamental flaws in the formula. Furthermore, because the payment cuts required by SGR were carried forward, this created a widening divergence between allowable and actual expenditures. This resulted in a payment cliff as each patch expired; by March 2015, reconciliation of the SGR formula with actual payment rates would have necessitated a fee schedule reduction of 21.2%. It was clear that legislative action was necessary to resolve this issue, but there was concern about both the utilization incentives inherent in fee-for-service models and the disconnect between payments and outcomes; as such, there was little desire to simply return to an unfettered fee-for-service model.
Congress passed the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA) as a solution to permanently repeal and replace the flawed SGR formula.6 This legislation had broad bicameral, bipartisan support and attempted to curtail increased healthcare spending and better align payment with technologic and clinical best practices. By default, providers would be placed in a modified fee-for-service model known as the Merit-based Incentive Payment System (MIPS), which would redistribute revenues from within the provider community based on a normative-based scoring system. To facilitate the shift from fee-for-service to value-based care, MACRA also encouraged the development of alternative payment models (APMs) as an additional payment structure either in conjunction with, or in lieu of, the more traditional model offered under the MIPS. Physicians who participated in varieties of APMs would either be totally exempted from participation in the MIPS, or be subject to more lenient scoring in the MIPS program. Unfortunately, although APMs offer many financial advantages over MIPS for qualified practices, urologists are largely excluded from eligibility from participation in APMs; based on historic metrics, the Centers for Medicare and Medicaid Services (CMS) estimated that a mere 88 urologists nationwide were qualified APM participants in 2017.7 This discussion explores the necessity for a specialty-specific urology APM, the requirements for a successful APM, the rationale for newly diagnosed prostate cancer as an APM model, and an overview of the LUGPA APM for initial therapy of newly diagnosed patients with organ-confined prostate cancer.
The MIPS provides physicians with a methodology to continue in a fee-for-service payment model while simultaneously creating incentives to enhance quality and control costs. Under the MIPS, four quality and resource use metrics are combined to form a single score; three of these metrics are adaptations of previously existing programs, and a new fourth program was added. The scores for these four programs are combined into a single aggregate score, after which the Secretary of Health and Human Services will determine the threshold percentiles for bonus and penalties. The various components of MIPS received variable weighting over 3 years, with progressively greater emphasis on resource use during this period. An overview of the historic programs and corresponding MIPS components, with the originally proposed relative weights, is presented in Table 1.
Of the four components of the MIPS, only the resource use component is calculated from claims data; the remaining three require submission by practices on either the basis of National Provider Identification (NPI) number or Taxpayer Identification Number (TIN), depending on the number of providers in the group. To measure resource use, CMS attributes patients to a reporting entity on a retrospective basis using a two-step system. If a primary care physician performed an evaluation and management service, the patient is attributed to that physician. If not, the patient is assigned to the specialty physician who has the plurality of evaluation and management codes. This is challenging for specialists, as they have no control or knowledge of which patients are going to be attributed to them for the purpose of resource use calculation. The physician to whom the patient is attributed is assigned all costs for inpatient, outpatient, ancillary services, and Part B drugs for that patient, regardless of who provides the service. The MIPS was designed to function as a deficit-neutral program, with penalties from lower-scoring providers used to fund bonuses for higher-scoring providers; in addition, a separate bonus pool of $500 million was created for “exceptional” performers. However, concerns regarding the stringency of the program led CMS, through the rulemaking process, to make the initial performance year (2017) of the program transitional; eligibility requirements were altered so that 40.8% of providers were exempted from MIPS reporting.8 For providers not exempted from MIPS reporting, three options were provided: (1) report a minimum amount of data and be exempt from both penalties and bonuses; (2) report a partial year of data and be eligible for limited bonuses; or (3) report a full year of data and be eligible for larger bonuses. Providers eligible for MIPS that did not meet the minimum reporting would be subject to a 4% negative payment adjustment in 2019. In addition, CMS recognized concerns regarding the weighting of the various MIPS components and reassigned the resource use component to 0% for the first year; to compensate for this change, they increased the weight of the quality component to 60% for the 2017 performance period.
CMS reports that it “engaged more than 100 stakeholder organizations and over 47,000 people since January 1, 2017, to raise awareness, solicit feedback, and help clinicians prepare to participate in MIPS”9; this resulted in continued modification of the program in the 2018 Proposed Rule with Comment Period. Of particular note is that CMS greatly expanded the number of physicians exempt from MIPS reporting from 40.8% to 63.7%—an increase of 56.1%. Unfortunately, this expansion is not evenly distributed across specialties. Over 97% of providers for which CMS provided MIPS eligibility data in the 2018 Final Rule can be directly or indirectly cross-walked to the 2017 MACRA Final Rule. Table 2 illustrates the 10 specialties with the highest and lowest MIPS reporting obligations. Only an additional 2.8% of urologists are exempted from MIPS reporting in performance period 2018 versus 2017; as such, urology is behind only critical care medicine in the percentage of physicians who are compelled to participate in the MIPS in performance year 2018.10 Beyond the unequal allocation of MIPS exemptions among different specialties, the attempt by CMS to ease the regulatory burden on the medical community has created another unintended consequence. Rather than have risk spread over a large pool of providers, a smaller number of what were likely more compliant physicians will now be competing against each other for what will be limited performance bonuses.
There is broad concern about the burden and relevancy of the MIPS program. A recent survey of 750 group practices by the Medical Group Management Association identified the Medicare Quality Payment Program (QPP) as their single greatest regulatory burden, with 82% of respondents rating the QPP as “very burdensome” or “extremely burdensome.”11 With respect to the MIPS, in particular, 71% of survey respondents said that they are “very concerned” or “extremely concerned” with overall implementation costs, and an even greater percentage—78% of all respondents—said that they are very or extremely concerned with the relevance of MIPS to specialty care. Perhaps most troubling is the fact that nearly three of every four respondents view the MIPS “[a]s a government program that does not support our practice’s clinical quality priorities.” At the October 5, 2017, Medicare Payment Advisory Commission (MedPAC) meeting, concerns regarding the utility of the MIPS as a mechanism to reward quality prompted nearly unanimous consensus among MedPAC commissioners that MIPS should be “repealed” and “replaced.”12
Being subject to MIPS reporting, with its associated potential penalties, clearly creates a challenging management problem for physicians. However, under MACRA, MIPS represents only one of four payment structures; MIPS only, MIPS with APM, APM Partial Qualifiers, and APM Qualified Participants. Fully and partially qualified APM participants can be exempted from MIPS reporting, and MIPS APM participants are subject to a more lenient scoring system: in addition to receiving full credit for their Country Policy and Institutional Assessment score, MIPS APM participants are exempt from the resource use component of MIPS. This means that MIPS APM participants need only report the Quality and Advancing Care Information components, both of which are directly manageable by the practice.
Although there are substantial benefits to APM participation, at present there are a limited number of advanced APMs available—a full list of these is presented in Table 3. To accelerate the development of APMs, particularly in specialty episodes of care models, Congress authorized a unique form of APM known as Physician-focused Payment Models (PFPMs). PFPMs were such a priority for Congress that they required the creation of the Physician-focused Payment Model Technical Advisory Committee (PTAC), an 11-member standing committee whose sole purpose is to evaluate and facilitate development of PFPMs.
PFPMs present a significant opportunity to engage subject matter experts as partners in our healthcare system’s critical evolution to value-based care. However, the development of APMs is costly and time consuming; after submission, the proposals are then subject to a cumbersome, multilayered review process. To date, no such model has been approved for implementation; consequently, most physicians are excluded from participation in value-based care payment structures. The challenges in APM development are reflected in the fact that only 20 proposals have been submitted to PTAC for review. This is particularly problematic, as the requirements to be a qualified participant in an advanced APM become progressively more stringent: at present either 25% of a reporting unit’s (NPI or TIN) Medicare payments or 20% of their Medicare patients must be enrolled in an advanced APM for performance year 2017. By performance year 2021, this increases to 75% of a reporting unit’s Medicare payments or 50% of its Medicare patients.
In addition to the intensive commitment in time and resources, development of an APM is complicated by the fact that not every disease entity is an appropriate vehicle for an APM. Characteristics of a disease/condition that may be an appropriate candidate for APM consideration include (1) meaningful treatment costs for payers; (2) the presence of numerous treatment options and/or an evolving clinical paradigm; (3) treatment variability in clinical practice across providers or sites of service; and (4) defined quality metrics and outcomes measures to ensure that cost savings do not come at the expense of patient care. If these parameters are met, it is then necessary to ensure that net cost savings occur if treatment variability is narrowed and/or treatment sequencing changed, while simultaneously ensuring that outcomes are improved or at least unaffected.
Prostate cancer remains the most commonly diagnosed solid tumor in men in the United States; in 2017, it is estimated to be the third leading cause of cancer death in American men.13 In 2015, there were approximately 79,000 Medicare fee-for-service beneficiaries diagnosed with prostate cancer, 79% of whom (approximately 63,000 cases) had cancer localized to the prostate. Urologists are the principal caregivers for patients with prostate cancer, and management of prostate cancer accounts for a substantial subset of the clinical activity for many urology practices.14
Just as existing episode-based APMs such as the Comprehensive Care for Joint Replacement Model (CJR) reconcile costs on an annual basis,15 the LUGPA APM development team tabulated the Medicare-allowed amount for all Part A and B claims associated with prostate cancer over a 12-month period, commencing with the prostate biopsy used to establish the cancer diagnosis. The historic cohort was developed using the Medicare 5% Limited Data Set (LDS) Claims Files for 2011 to 2015 for Current Procedural Terminology (CPT), Healthcare Common Procedure Coding System (HCPCS), International Classification of Diseases, Ninth Revision (ICD-9), or Medicare Severity-Diagnosis Related Group (MS-DRG) codes referable to International Classification of Disease Codes Version 9 and 10 diagnosis codes referable to prostate cancer.16 Intervention was defined as any patient undergoing hormonal therapy, surgery, or radiotherapy during the initial 12 months after prostate biopsy. Patients who did not receive active intervention (AI) during the first 12 months after prostate biopsy were considered to have been managed with active surveillance (AS). The results of the cost analysis are presented in Table 4. Based on this analysis, Medicare expenditures for prostate cancer therapy exceed $1.7 billion in the first year after diagnosis alone. The substantial nature of this cost is reflected by the fact that, for 2015, Medicare reports that the total outlay for all non–Part B-related urology professional services for all diagnoses was approximately $1.1 billion.17
Evolving scientific knowledge regarding the natural history of prostate cancer suggests that 43% of newly diagnosed prostate cancer cases had a Gleason score ≤6 and would likely be able to defer AI at the time of diagnosis,18 and instead be closely monitored via AS. Analysis of the Medicare claims data suggests that 23% of newly diagnosed patients with prostate cancer do not have some form of intervention in the first year after diagnosis; given the percentage of patients newly diagnosed with relatively lower-risk disease, it appears that only nearly half of appropriate AS candidates are actually receiving immediate AI.
Part of this issue may be that the current fee-for-service payment system has not kept pace with existing consensus guidelines19; consequently, providers are reimbursed nearly 2.7 times more in the first year of diagnosis for AI than for AS. Although AS requires adherence to a rigorous regimen of follow-up, there is no mechanism in place to compensate providers for the resources needed to properly track and counsel newly diagnosed patients with prostate cancer. This counseling is important, as patients with substantial emotional distress are more likely to choose AI over AS at the time of diagnosis.20 Conversely, once on a surveillance protocol, patients on AS may tend to experience more distress than those who initially chose AI.21 The lack of resources currently available to providers may account for recent reports of high attrition rates for patients on AS22; particularly troubling is data suggesting that attrition rates for AS are higher in underserved communities.23
A surprising degree of regional variability exists in the management of prostate cancer24; limiting variability in disease management through integrated delivery systems has been shown to reduce healthcare expenditures.24 Existing APMs such as the Oncology Care Model and CJR measure costs against benchmarks in regions determined by the United States Census Bureau (Figure 3). To determine if regional variability exists in prostate cancer therapy, the LUGPA APM development team stratified utilization data from the Medicare 5% LDS by Metropolitan Statistical Area (MSA) and then aggregated MSAs into their respective US Census divisions. As shown in Table 5, the results of this analysis revealed a remarkable degree of variation in the management of newly diagnosed patients with prostate cancer, particularly with respect to utilization of AS versus AI. Use of AS varied from a low of 19.6% to a high of 29.2% in the West South Central and New England census divisions, respectively—a difference of 49.7%. Variation in choice of AI was much less dramatic, with use of radical prostatectomy alone varying between 12.4% and 14.1% in the New England and both East and West South Central divisions, respectively. This regional variation of 13.7% was similar to that seen for radiotherapy alone, which varied 13.9% between the 20.1% and 22.9% utilization observed in New England and West South Central divisions, respectively. Given the differential in cost between AS and AI, reducing the treatment variability between these options should provide an opportunity to reduce treatment costs.
At present, there is one proposed and three existing quality measures that are specific to prostate cancer: (1) avoidance of overuse of bone scan for staging low-risk prostate cancer patients (National Quality Forum [NQF] 0389); (2) adjuvant hormonal therapy for high-risk or very high-risk prostate cancer (NQF 0390); (3) radical prostatectomy pathology reporting (NQF 1853); and (4) the proposed bone density evaluation for patients with prostate cancer who are receiving androgen deprivation therapy. In addition, practitioners may report on follow-up after biopsy (MIPS/Physician Quality Reporting System), which is not specific to prostate cancer. Not only are there no quality metrics with respect to AS for prostate cancer, but the only method to determine if a patient is on AS is by excluding all other therapeutic interventions. Longitudinal measurements of prostate cancer treatment patterns would benefit greatly if direct measurements of AS existed.
The LUGPA APM is an episode-based model that aligns incentives for physicians to recommend AS in clinically appropriate patients, allowing them to avoid potentially unnecessary services. The LUGPA APM, development of which was supported in part by Myriad Genetics (Salt Lake City, UT) and Integra Connect (West Palm Beach FL), will compensate physicians for the management time required to responsibly continue newly diagnosed patients on AS, while simultaneously creating benchmarks defined based on a practice’s historic clinical decision making, considering prior use of AS versus AI. Practices are eligible for a performance-based payment if they meet certain quality thresholds and if total episode spending is less than that established at the benchmark date.
Specifically, the LUGPA APM consists of two phases: an initial 12-month total cost of care episode that commences at the initial diagnosis of organ-confined prostate cancer diagnosis, with subsequent episodes of care for qualified beneficiaries who remain on AS after the initial 12-month period. The model includes both a $75 monthly care management fee for AS episodes and a retrospective performance-based payment for enhancing the utilization of AS over baseline levels. The performance-based bonus payments would only apply to the initial 12-month period, while the monthly care management fee would continue for as long as the patient remained on AS. The monthly management fee would be used for (1) tracking AS beneficiaries to ensure compliance throughout episodes; (2) tracking laboratory results longitudinally in a consistent format for optimal PSA testing; (3) continually educating beneficiaries about disease progression; (4) social services and care coordination across practitioners; and (5) reviewing/revising the care plan.
The performance payment would consist of a retrospective comparison of actual initial episode spending for newly diagnosed patients against a practice-specific target price. This target price would blend practice-specific and regional prevalence of AS. However, to enhance inclusivity, target prices for low-volume practices would be regional. The target price would utilize 3 years of historic data on utilization of different treatment modalities for prostate cancer (with an update for later performance years) and include trending methodology to account for unit cost changes. The APM would use CMS payment standardization methodology for all calculations with a back-end geographic adjustment to the performance-based payment.
Apart from the novel payment structure that incentivizes use of AS, the LUGPA APM incorporates quality measures in the domains of efficiency and cost reduction, communication and care coordination, clinical outcomes, and process to ensure that appropriate patients are being selected for therapy. In addition to reporting domains on avoidance of overuse of bone scan as well as reporting prostate biopsy results, two additional domains will be utilized. The first is a modification of NQF 2962, which would create a patient-reported outcome measure regarding prostate cancer shared decision making that would to apply to all beneficiaries in the initial episode of the APM. The second is a completely new measure that would measure time on AS. Specifically, there would be three nonpayable G-Codes used to describe reasons beneficiaries left AS: (1) beneficiary choice; (2) lack of compliance with AS protocol; and (3) disease progression. The calculation of the measure would be straightforward, with the denominator being the number of beneficiaries in initial or subsequent AS episodes and the numerator being the sum of number of months beneficiaries in the denominator were on AS. There is not a current mechanism to track the duration of AS, and this measure would fill that void while allowing APM entities to better track beneficiaries on AS. Moreover, nonparticipating urologists can also use the G-Codes, allowing for national tracking of both the utilization and duration of AS.
Although the focus of any APM should be on ensuring that patient outcomes and satisfaction are enhanced, it is also important that these vehicles do so in an economically responsible fashion. In addition to enhancing shared decision making, improving compliance with recommended clinical pathways, and reducing morbidity associated with AI, the LUGPA APM has the potential to substantially reduce Medicare expenditures. The differential between expenditures for AS and AI is approximately $20,000 in the first year after diagnosis; a 10% increase in utilization of AS would mean savings to the Medicare program in excess of $125 million annually.
Over the past several decades, rapid expansion in healthcare expenditures has exposed the utilization incentives inherent in fee-for-service payment models. The passage of MACRA heralded a transition toward value-based care, creating incentives for practitioners to accept bidirectional risk linked to outcome and utilization metrics. At present, the limited availability of these vehicles excludes all but a handful of providers from participation in APMs. At present, there are no urology-specific APMs available; as such, most urologists are compelled to participate in the MIPS. Although CMS has eased the regulatory burden on the medical community as a whole, the distribution of MIPS exemptions is disproportionate, creating additional management issues for urology practices.
The LUGPA APM supports the goals of the triple aim in improving the patient experience, enhancing population health and reducing expenditures. By requiring utilization of certified electronic health record technologies, tying payment to quality metrics, and requiring practices to bear more than nominal risk, the LUGPA APM qualifies as an advanced APM, thereby easing the reporting burden and creating opportunities for participating practices. The LUGPA APM is designed to have broad appeal to urologists, including those in large and small practices, whether independent or hospital-owned, with or without ownership of ancillary services. Endorsed by both the American Urological Association and the American Association of Clinical Urologists, the LUGPA APM creates a substantial value proposition for CMS and APM participants by utilizing AS when appropriate. Given the unmet need filled by this proposal for urology participation in APMs, the opportunity to reduce medical expenditures, and most importantly, the opportunity to enhance outcomes and beneficiary experience, the LUGPA APM could be the foundation for value-based care models in urology.