June 21, 2016

Andrew M. Slavitt, Acting Administrator
Centers for Medicare & Medicaid Services

Thank you for the opportunity to comment on the Medicare Access & CHIP Reauthorization Act’s (MACRA) Merit-Based Incentive Payment System (MIPS) and Alternative Payment Models (APMs) rule you proposed. The National Committee for Quality Assurance (NCQA) strongly supports this transition to paying clinicians for the value, rather than volume, of care they provide, and is eager to help.

  • Requiring PCMH and PCSP recognition from a national-in-scope, widely used third party program to receive automatic full Clinical Practice Improvement Activity (CPIA) credit.
  • Working toward more and better outcome measurement and measures derived from data entered into electronic systems as a natural part of clinical workflow.
  • Offering bonus points for electronic reporting and reporting in high priority areas.
  • Promoting joint accountability by averaging MIPS scores among each APM’s participants and applying that average score to each participant.
  • Setting an appropriately high bar for “Advanced” APMs that earn automatic 5% bonuses.

Establish insufficient criteria for identifying and including quality measures for reporting in MIPS.

  •  Undermine AAPM viability by basing “nominal risk” on total care cost rather than Medicare revenue.
  • Detailed comments on these and other issues in the proposed rule are below.

    MERIT-BASED INCENTIVE PAYMENT SYSTEM (MIPS) Quality, which assesses clinical care.
  • Clinical Practice Improvement Activities, which should improve outcomes if effectively implemented.
  • However, given MIPS’ complexity and wide variation in clinician readiness, you may want to simplify the requirements.

    We also encourage you to closely monitor for specialties where significant numbers of clinicians report less than the minimum number to help focus measure development on filling any related measurement gaps. And we support your proposal to develop a validation process, and then give scores of zero, if clinicians can, but fail to, report on the minimum number of measures. To provide real value and meet MACRA’s goals, measures must also be clinically important, transparent, feasible, actionable and rigorously audited to ensure accuracy and fairness. There is no point to measures that are not clinically important to health care consumers and the system overall. There is great potential for gaming if measures are not publicly transparent on what is being measured and how. There is little ability to report measures for which it is not feasible to collect and validate the data. There is no point to measures that are not actionable in promptly showing where clinicians must improve. And rigorous auditing is essential to ensuring real quality improvement and honest reporting. All-Payer Data: One of MACRA’s most important themes is its move toward assessing not just Medicare, but all payers’ patients in APMs. To best support this and ensure data completeness when clinicians in APMs report on non-Medicare patients for MIPS, we encourage you to move away from the traditional visit-based measurement. Instead, you should instead focus more on assessing outcomes for a whole population. For example, many measures look for documentation of a follow-up plan. Population-based assessment would look for whether follow-up actually occurred and whether the outcome or symptom was re-assessed after follow-up. This approach would encourage and reward clinicians for providing and coordinating care across the continuum of a condition rather than a single encounter. However, we note that population-based assessment is difficult without prospective enrollment that informs clinicians in advance about the patients for which they are accountable. The ability to improve on population health measures therefore is limited in MIPS. Providing bonus points for electronic reporting is especially important, as it will help move to a future state where measurement is based on electronic data derived as a part of, not in addition to, normal clinical workflows. This electronic reporting should be through systems certified to produce accurate and reliable electronic quality measure results. We encourage you to also consider providing bonus points to incentivize clinicians to aggregate into virtual groups. We further support capping the number of bonus points so they do not mask low quality in other areas. Consumer Assessment of Healthcare Providers & Systems: Patient experience reporting for groups of >100 should eventually be mandatory. However, we support voluntary reporting until the survey and its reporting mechanism improve. The current CAHPS survey is too long and generates low response rates. It also does not provide the prompt, targeted feedback clinicians need. Our own psychometric testing suggests it could be cut by at least one third, as answers to many questions consistently predict answers to others. We also believe that electronic alternatives to the paper-and-mail-based survey administration process could increase both response and feedback rates. We therefore urge you to work with clinicians, Agency for Healthcare Research & Quality CAHPS stewards and other stakeholders to develop a better means of obtaining patient experience data and require its use as soon as feasible. Ideally, this improved tool will have better data systems and standardized collection methods that support inclusion of other payers’ patients in CAHPS samples. For incorporating other payers’ patient experience data, CMS should establish standardized methods for drawing samples, validating that logic and auditing to ensure the logic is applied to the full eligible population. There has been minimal progress on existing behavioral measures, and NCQA is aggressively working to address that and fill gaps in behavioral measurement. Appendix A lists nearly 30 behavioral measures we have or are now developing. We encourage you to consider including them in MIPS when they can be applied appropriately and fairly to hold clinicians accountable in order to drive much-needed improvement in this critical area. We further encourage you to provide bonus points to incentivize reporting on them. Avoidance of Antibiotic Treatment in Adults with Acute Bronchitis;
  • Use of High-Risk Medications in the Elderly; and

  • However, appropriate use measures face key challenges that require broader understanding.

    • Overuse measures must account for symptoms and history that make an otherwise questionable service appropriate for individuals. However, information on whether individual patients have such factors often is not in claims and requires extensive medical record review, at least until measurement derives from electronic health records. Even then, once all appropriate exclusions are identified, the number of cases that can be confidently categorized as overuse can be quite small.
    • Assigning accountability for appropriate use measures is challenging when more than one provider may be responsible for inappropriate care. As discussed above, measurement of high-risk medication use should include all prescribed medications, not just those from a single practitioner, and it is not reasonable to hold an individual practitioner accountable for medications prescribed by another practitioner.
    Hospital-Employed Clinicians: Allowing hospital-employed clinicians to choose hospital measures is appropriate for some, such as hospitalists whose performance is closely linked with their facility’s. It would not be appropriate for others, such as clinicians in outpatient clinics or surgeons, who should be required to report other MIPS measures more specific to their own individual practice.
    However, there are important challenges to address in getting to better outcomes measures. Measurement today is largely based on claims that lack data outcomes. Measurement that derives from data entered into electronic systems as a natural part of clinical workflow will enable better outcomes measurement and reduce the burden required for measurement reporting. CMS should encourage electronic health record (EHR) vendors to incorporate PROMs, as well as development and use of PROMs, while carefully monitoring their use for potential gaming, such as copy-and-paste documentation. Also, it is difficult for clinicians in small practices or other minimally organized systems to have substantial impact on outcomes, which are influenced by many factors beyond clinical care. MACRA addresses this by encouraging clinicians to move toward APMs for which population-based outcomes assessment is more fair. One example is the Global Cardiovascular Risk (GCVR), a predictive model that creates cardiovascular disease risk scores from patient data in electronic systems. Rather than focusing on population-based targets (e.g., blood pressure of 140/90, A1c of 9.0) GCVR acknowledges that the greatest benefit for one individual may be smoking cessation and for another may be bringing systolic blood pressure of 200 down to 150. The foundation of GCVR scores is that they are sensitive to changes in individual patients’ risk profile, rather than achievement of static thresholds. The system uses electronic clinical data to gauge which care elements would register the most significant impact on outcomes for the individual patient. Measures for Duals & Other Complex Populations: We support adding the Medicaid Adult Core Set to MIPS, which is particularly important for people dually enrolled in Medicare and Medicaid who have greater need and higher costs. MACRA Measure Development Plan. Socio-Economic Status: We appreciate your interest in this area and also eagerly await the HHS Assistant Secretary for Planning and Evaluation’s study on this topic. Our own research shows that while there is a correlation between SES and quality, it is small and many positive outliers achieve high quality despite serving low SES populations. The correlation between SES and resource use is much clearer, as meeting lower SES patients’ greater needs and challenges costs more, for example in care coordination supported by care management fees. This is why Medicare is now adjusting payments to Medicare Advantage plans to account for the higher costs of serving low-SES enrollees who are dually eligible for Medicare and Medicaid. We therefore support risk adjusting resource use measures for SES because there is a clear and plausible association between SES and care costs. We oppose risk adjusting quality measures where SES associations are at best murky. NCQA’s PCMH program is by far the nation’s largest, with over 17% of all primary care physicians and more than 56,600 clinicians overall at over 11,420 sites across the country. Our PCSP program is the only such program that currently exists for specialists. Our PCMH and PCSP programs together help clinicians build well-coordinated “medical neighborhoods” that are ideal AAPM foundations. To earn NCQA recognition practices must meet high standards for expanded access, care coordination, population management, and helping patients to better engage in their own health and health care. The programs strongly support Congressional intent to promote high quality, efficient patient-centered care. We encourage you to establish specific thresholds for assessing which programs meet those criteria and to require up-to-date market penetration data to verify whether other programs meet those thresholds. Regarding how to assign PCMH credit when only some practices using the same tax identification number (TIN) have recognition, we note that some practices have more than one TIN and some clinicians practice in more than one site under different TINs. We are asking our PCMH and PCSP practices how you might address this challenge and will share any insights from them with you. We also support your proposal to study simpler ways to collect and verify CPIAs and provide more rapid feedback, and are eager to help in any way we can. We further support Achieving Health Equity, Integrated Behavioral and Mental Health, Promoting Health Equity and Continuity and Social and Community Involvement CPIA subcategories. In refining individual CPIAs over time, we encourage you to focus on those that are most effective in helping clinicians advance towards APMs. Advancing Care Information: We support the proposal to assess clinicians progress in using EHRs to improve care and move away from blunt thresholds. Over time, we urge you to gradually reduce the “base score” for simply reporting on EHR use and to increase the “performance score” in order to encourage further progress. We also strongly favor your proposal to define success in this category as achieving a 75% score, and reduce the category’s weight when 75% of clinicians reach that threshold. Reducing the weight when just 50% reach the threshold would prematurely thwart efforts to increase data sharing for care coordination and quality improvement. However, there are many sparsely populated regions that support only small numbers of small practices which may not readily aggregate into virtual groups. Beneficiaries in these regions could lose needed access to care if MIPS makes these practices financially unsustainable. In other areas, pushing too hard and too fast for aggregation could have the unintended consequence of incentivizing practices to join large hospital-based systems that drive higher costs system-wide. ALTERNATIVE PAYMENT MODELS (APMs) Given this significant incentive, we generally agree with the proposed high bar for earning AAPMs status but believe additional models to those listed in the proposed rule also qualify. However, the proposed definition of “more than nominal risk” required for AAPMs may limit the ability of some clinicians, particularly those in smaller or solo practices, to participate in AAPMs by basing “nominal risk” requirements on total care cost rather than Medicare revenue. We appreciate your concern about calculating the amount of risk based on Medicare revenue alone for each individual practice. However, the work should be minimal since CMS must do this for both the Medical Home Model Standard and to calculate the percent of revenue threshold for determining qualified AAPM Entities. Given the benefit and incremental work required, we strongly urge you to base the standard nominal risk model on Medicare revenue, as you propose for the Medical Home Model. Our comments and suggestions on specific APM proposals and comment requests are below. Given that proposed AAPMs are time-limited demonstrations, we urge you to clarify what becomes of clinicians in these demonstrations once the demonstrations end. More than Nominal Risk: You should not base “more than nominal risk” AAPMs benchmarks on total cost of care as that much risk threatens financial viability. You should instead base benchmarks on APM Entities’ Part A and B Medicare revenue to limit the risk and maintain financial viability. detailed comments to CMS on the issue. Measures Comparable to MIPS: As to whether to let APMs use all measures comparable to the criteria set for MIPS measures – evidence-based, reliable and valid – we reiterate our belief that MIPS measure criteria are insufficient and must be much more stringent to achieve MACRA’s goals. To provide actual value, measures must also be clinically important, transparent, feasible, actionable and rigorously audited to ensure accuracy. There is no point to measures that are not clinically important to health care consumers and the system overall. There is great potential for gaming if measures are not publicly transparent on what is being measured and how. There is little ability to report measures for which it is not feasible to collect and validate the data. There also is no point to measures that are not actionable in showing where clinicians must improve. And rigorous auditing is essential to ensuring real quality improvement and honest reporting. However, we do not support the proposal to limit Medical Home Model AAPM’s to only practices with 50 or fewer eligible clinicians. This arbitrary limit would prohibit practices with 51 or more clinicians from becoming Qualified Providers for no apparent reason. Other Payer APMs: The ability to include not just Medicare but all payers’ patients in APMs is among MACRA’s most important provisions. The current cacophony of different measures and payment incentives across payers needlessly increases clinician burden and frustration. It also makes it exceedingly difficult for other stakeholders to make apples-to-apples comparisons across innovative payment models. The standardized attribution model described above will be particularly important for ensuring fairness when including other payers. So will establishment of parallel requirements, such as to have at least one outcome measure for other payer APMs. HELPING CLINICIANS ADVANCE TOWARDS APMS Virtual Groups: Virtual groups let clinicians join together to have enough patients for valid measurement, and can be a critical first step towards development of more organized systems and ultimately AAPMs. They also may be a significant departure from previous experience for many clinicians. We appreciate CMS’ need for additional time to develop a virtual group registration. However, Medicare should also use this time to encourage and support development of much-needed guidance and assistance on how clinicians can take this first step toward building organized systems that are more capable of improving care and move up the continuum toward AAPMs. We are eager to collaborate with you and other stakeholders to help clinicians aggregate into virtual groups and begin the journey toward APMs. Identifying virtual group partners, such as recognized PCMH and/or PCSP practices that MACRA actively promotes. Recognized PCMHs and PCSPs have demonstrated commitments to well-coordinated, high-quality, patient-centered care and thus greater potential to improve MIPS scores. These could be:
  • Other PCMH and PCSP practices in the same community or geographic region; or
  • Groups of similar PCSPs likely to report the same specialty measures.
  • Drafting written agreements to establish virtual groups and share accountability and financial risk;
  • Developing skills and tools for group reporting that will be new to virtual groups;
  • Developing skills and expertise in analyzing data and addressing any quality gaps in order to improve MIPS scores and succeed as virtual groups; and
  • Developing further skills and expertise to maximize use of CEHRT, base pay on performance and take two-sided risk in order to become APMs.
  • Margaret O’Kane,