Center Comments on Draft 2018 Medicare Advantage Call Letter

March 3, 2017

VIA ELECTRONIC SUBMISSION
AdvanceNotice2018@cms.hhs.gov

Cynthia G. Tudor, Acting Administrator
Centers for Medicare & Medicaid Services
Department of Health and Human Services
Baltimore, MD 21244

Re:  Advance Notice of Methodological Changes for Calendar Year 2018 for Medicare Advantage (MA) Capitation Rates, Part C and Part D Payment Policies and 2018 Call Letter

Dear Acting Administrator Tudor:

The Center for Medicare Advocacy (Center) is pleased to provide the Centers for Medicare & Medicaid Services (CMS) comments on the draft 2018 Call Letter.  The Center, founded in 1986, is a national, non-partisan education and advocacy organization that works to ensure fair access to Medicare and to quality healthcare. At the Center, we educate older people and people with disabilities to help secure fair access to necessary health care services. We draw upon our direct experience with thousands of individuals to educate policy makers about how their decisions affect the lives of real people. Additionally, we provide legal representation to ensure that people receive the health care benefits to which they are legally entitled, and to the quality health care they need.

Attachment II: Changes to Part C Payment Methodology for CY 2018

General Comment re: Medicare Advantage Payment Rates

We recognize that the payment methodologies outlined in the Advance 2018 Rate Notice and Call Letter are consistent with applicable law, particularly the Affordable Care Act (ACA) changes to bring Medicare Advantage (MA) plan payments in line with costs under the traditional Medicare program. The Center continues to support this effort to achieve this payment parity, both as a matter of equity between MA and traditional Medicare, and as a means of appropriately safeguarding public funds. 

Despite much attention given to the annual rate notice concerning MA payment, though, we urge CMS to direct more attention to protecting public funds by ensuring that such payment is accurate.  MA “upcoding” – when an MA plan inappropriately reports an enrollee as being more sick than they actually are in order to obtain a higher risk-adjusted payment from the Medicare program – remains an ongoing  problem that policymakers must address. 

Various studies have attempted to document the scale of inappropriate MA coding intensity, or upcoding, and the resultant overcharges by MA plans.  An investigation by the Center for Public Integrity, for example, found that Medicare paid MA plans nearly $70 billion in “improper” payments, mostly from upcoding, from 2008 through 2013 alone.[1]  More recently, a study published in Health Affairs found that coding intensity practices could result in overpayments to MA plans totaling $200 billion over the next decade.[2]

 In April 2016, the General Accounting Office (GAO) issued a report entitled “Medicare Advantage: Fundamental Improvements Needed in CMS’s Effort to Recover Substantial Amounts of Improper Payments.”[3]  The report states that CMS estimates that about 9.5% of its annual payments to Medicare Advantage (MA) organizations were improper – totaling $14.1 billion in 2013 alone – “primarily stemming from unsupported diagnoses submitted by MA organizations.”  The report also highlights the significant flaws in CMS’ current efforts to address and recoup such payments, including execution of the Risk Adjustment Data Validation (RADV) audit process.

The Center is deeply concerned by these ongoing improper payments to MA plans and CMS’ lack of progress in recouping previous payments and deterring future misconduct.  In order to ensure that the traditional Medicare program is not further disadvantaged by inappropriate overpayments to MA plans, CMS must employ more rigorous oversight of MA payment.

Section G. MA Employer Group Waiver Plans (p. 23)

In the 2017 Call Letter, CMS finalized a proposal to waive the bidding requirements for Medicare Advantage Employer Group Waiver Plans (EGWPs) and to pay these plans using an alternative payment mechanism. According to CMS, these changes were intended to reduce administrative burdens on employer plans and to more accurately capture EGWP costs by eliminating existing incentives to submit bids that are higher than actual projected costs. In the draft 2018 Call Letter, though, CMS suggests that the agency may not complete phase-in of the payment changes in 2018.

In order to ensure greater payment equity between EGWP and non-EGWP MA plans, we urge CMS to not delay in fully implementing the alternative payment policy for MA EGWPs.  This change in payment policy is supported by previous findings from the Medicare Payment Advisory Commission (MedPAC).  MedPAC found that MA payments to EGWPs were previously substantially higher than fee-for-service costs for comparable beneficiaries, despite the provisions that have better aligned overall Medicare Advantage payments with fee-for-service costs.  Moreover, CMS has previously noted that while EGWPs tend to have healthier low-cost enrollees than other Medicare Advantage plans — risk scores are more than 9 percent lower — and face lower administrative costs related to enrollment and marketing than other Medicare Advantage plans, their bids are actually higher, on average.  MedPAC has previously recommended in 2014 that CMS take steps to determine payments for EGWPs in a manner more consistent with how other Medicare Advantage plan payments are determined.[4] 

Section H: MA Coding Pattern Adjustment (p. 28)

As discussed above, inappropriate Medicare Advantage coding intensity – upcoding – is a problem that must be addressed.  Therefore, we recommend CMS increase the coding intensity adjustment for 2018 above the proposed level of 5.91 percent — the statutory minimum.  As MedPAC found in its March 2016 report to Congress, the statutory minimum coding adjustment is highly insufficient to fully offset current coding intensity trends.[5]  MedPAC estimated in that report, for example, that in 2017, the coding difference between MA and fee-for-service could be 9 percent or more.  More recent estimates from MedPAC found that in 2015, Medicare Advantage risk scores were about 10 percent higher than in fee-for-service, resulting in a 4 percent coding difference between Medicare Advantage and fee-for-service remaining after applying the 2015 statutory minimum adjustment of 5.16 percent.[6]  While MedPAC finds that MA payments, on average, now equal 100 percent of fee-for-service costs, this estimate does not reflect unresolved coding differences, which if taken into account would result in MA payments continuing to exceed fee-for-service costs.[7]

Note re: In-Home Enrollee Risk Assessments

Notably absent from the draft 2018 Call Letter, similar to the 2017 final Call Letter, is the topic of MA in-home risk assessments.  We are disappointed that CMS has not moved further along in its effort to regulate in-home risk assessments to ensure that the services provided to beneficiaries through these visits are meaningful and effective, not simply a means for collecting risk adjustment diagnoses without ensuring that meaningful follow-up care is delivered.  We note that we supported CMS’ initial proposal in its draft 2015 Call Letter to exclude, for payment purposes, diagnoses from in-home risk assessments that were not confirmed by a subsequent clinical encounter (a policy change also supported by MedPAC), but CMS did not finalize these proposals.

We are concerned that voluntary best practices – a proposal from the 2016 Call Letter – will not achieve CMS’ goal of linking heightened risk scores to care that addresses beneficiaries health needs. Use of such best practices should be only an interim step while CMS works to develop a strong evidence-based in-home assessment tool. We again urge CMS to exclude, for payment purposes, diagnoses identified during a home visit that are not confirmed by a subsequent clinical encounter. Simply enhancing a risk profile without benefiting enrollees serves no useful purpose to the Medicare program or its beneficiaries.

Section K: Encounter Data as a Diagnosis Source for 2018 (p. 35)

The Center supports CMS’ transition to the use of encounter data in calculating MA risk scores so that risk adjustment can be gauged by actual care provided by MA plans.  However, we do not support applying a uniform industry-wide adjustment to the encounter data-based portion of the blended risk score under the Part C and ESRD models.  The purpose of collecting and using encounter data is to improve the accuracy of the MA risk adjustment system, which may have the effect of lowering risk scores and hence reducing payments to MA plans.  Applying an industry-wide adjustment, though, would seriously undermine the benefits of using encounter data.  Moreover, as MedPAC notes, such an adjustment would “reduce the incentive to submit complete encounter data, preserve the notion of RAPS [Risk Adjustment Processing System data] as the basis for payment, and hinder the transition to EDS-based [Encounter Data System] payment.”[8]

Attachment VI: Draft Call Letter

            Section I – Parts C and D

Enhancements to the 2018 Star Ratings and Beyond

Changes in Measures for 2018 (p. 80)

                        Adjusting Star Ratings for Audit and Enforcement Actions (p. 82)

We strongly support CMS’ work to further evaluate how audits, civil money penalties, and sanctions impact plan Star Ratings and to minimize any disconnect between audit scores and the Star Ratings system.  This disconnect can, at a minimum, be a source of confusion for people with Medicare and those who assist them seeking to evaluate and compare health plan quality, and at worst, undermine the public’s confidence in the integrity of the Star Ratings system altogether.

We urge CMS to ensure that the Star Rating system does not mask or otherwise minimize plan conduct that puts Medicare enrollees at risk.  Particularly when CMS finds that a plan’s conduct poses a serious threat to the health and safety of Medicare beneficiaries, that finding must have an impact on overall ratings. 

Perhaps the most glaring disconnect is when a plan is identified as having the same, repeated serious deficiencies in audit scores while its Star Ratings continue to rise. To address this imbalance, it is critically important that Star Ratings incorporate audit measures and reflect audit results in meaningful ways, while CMS continues to impose significant sanctions and penalties when serious deficiencies are identified. For example, CMS sanctioned Cigna last year, for a “longstanding history” of noncompliance, in which they threatened the health and safety of beneficiaries. Yet, despite these major issues, Cigna did not receive a reduction in Star Ratings, and therefore received a quality bonus payment – which is incongruent with their major compliance problems that impacted beneficiaries. When CMS sanctions a contract, the contract has significant issues which should be apparent to Medicare beneficiaries in the Star Ratings when shopping for plans.[9]

As CMS notes in the Call Letter, the integrity of the Star Ratings and the ability of the ratings to aid in the selection of a plan must not be compromised. This means that Star Ratings must reflect the severity of issues that result in a plan being sanctioned.

To solve this issue, CMS proposes to revise how civil money penalties are reflected in the Beneficiary Access and Plan Performance (BAPP) measure.  When violations are so severe that they trigger sanctions (as opposed to civil money penalties), however, it is not enough to merely include them as part of a measure or sub-measure.  Further, we are disappointed that CMS will apparently allow for any MA plan contract with sanctions levied against it to still receive a quality bonus payment, if achieving a 4-star rating or higher.  Any plan contract with sanctions does not deserve a bonus payment. As past examples have shown, some of the actions by sanctioned sponsors have been egregious and do not reflect a high-quality plan worthy of receiving a bonus payment. 

A minimal impact of a plan’s misconduct on its measure does not send the right signal to plans or to beneficiaries.[10] As such, we encourage the agency to revisit this policy. Specifically, we recommend that CMS reconsider more significantly weighting the BAPP measure and/or adjusting the overall and summary Star Ratings by at least one star for sanctioned plans.

2018 Star Ratings Program and CAI (p. 87)

 We continue to be concerned by CMS’ policy adjusting Star Ratings scores based on socio-economic and disability status. We would like to reiterate our concerns from our comments to the 2017 draft call letter[11] that adjusting Star Ratings based on socioeconomic and disability status might run the risk of masking true differences in quality for these populations. We are concerned that this adjustment would not meet CMS’ requirement that “policies implemented must result in high quality of care and improved health outcomes” for all beneficiaries. As such, we continue to urge CMS to develop a plan and timeline for phasing out this adjustment.

2018 CMS Display Measures (p. 97)

Chronic Use of Atypical Antipsychotics by Elderly Beneficiaries in Nursing Homes (Part D) (p. 101)

We oppose the proposed changes to the reporting of antipsychotic drug use in persons with dementia because they call for less information about these drugs, which are so widely misused and harmful to older people who have dementia but no diagnosis of psychosis.

At present and since 2013, Medicare Part D Compare reports atypical antipsychotic drug payments for residents in nursing facilities through their Part D Plans (called Rate of Chronic Use of Atypical Antipsychotics by Elderly Beneficiaries in Nursing Homes).  This measure focuses on a group of great concern – residents with dementia but no diagnosis of psychosis – and presents more reliable data than facilities’ self-reported MDS data.  As the HHS Inspector General reported in 2011, most of the antipsychotic drugs given to nursing home residents are inappropriate, off-label, and, as described by the Food and Drug Administration, subject to a Black Box Warning (meaning that the drug could kill the patient).[12] 

The final CY 2017 Call Letter discussed CMS’s concern about the unnecessary use of antipsychotic drugs in nursing homes and cited the GAO’s 2015 report, Antipsychotic Drug Use: HHS Has Initiatives to Reduce Use among Older Adults in Nursing Homes, but Should Expand Efforts to Other Settings, GAO-15-211 (Jan. 30, 2015)[13], which calls for additional efforts to reduce antipsychotic drugs for individuals who have dementia and live in the community (“The GAO conducted this study due to concerns raised regarding the use of antipsychotic drugs to address the behavioral symptoms associated with dementia, the FDA’s boxed warning that these drugs may cause an increased risk of death, when used by older adults with dementia, and because the drugs are not approved for this use.”).

The final CY 2017 Call Letter identified future changes to Medicare Part D Compare that would improve information about antipsychotic drugs – creating separate breakout measures for community-only individuals, short-stay nursing facility residents (fewer than 100 days, using MDS data), and long-stay nursing facility residents – and an overall measure (called the Antipsychotic Use in Persons with Dementia, APD), which combines community individuals and long-stay nursing facility residents.  The purpose of these planned revisions was “to continue to draw attention to the inappropriate use of antipsychotics in persons with dementia without an appropriate mental health diagnosis in both the community and nursing home settings.” 

The proposal in the Draft CY 2018 Call Letter goes backwards.  It proposes to remove the long-stay nursing home measure (the Rate of Chronic Use of Atypical Antipsychotics by Elderly Beneficiaries in Nursing Homes) and to replace it solely with the Antipsychotic Use in Persons with Dementia, APD. 

Contrary to the discussion in the final CY 2017 Call Letter, the Draft CY 2018 Call Letter no longer proposes displaying the three breakout measures (community-only residents, short-stay and long-stay nursing home residents), to accompany the APD.   It proposes displaying only the overall APD measure.  No information will be reported specifically on Part D payments for antipsychotic drug use in nursing facilities.

For 2019, the Draft CY 2018 Call Letter also retreats from revisions identified in 2017.  It proposes to report only two breakout measures – community individuals and long-stay nursing facility residents – not the three measures described in the Final CY 2017 Call Letter.  Missing is a short-stay nursing facility resident measure.  

Section II – Part C

PMPM Actuarially Equivalent Cost-Sharing Limits (p. 118)

As noted in the final 2017 Call Letter, CMS will not permit MA plans to charge cost-sharing for the first 20 days of a SNF stay in 2018.   We strongly support this policy – it is in compliance with the statutory requirements, and is in line with cost-sharing in traditional Medicare.

Since passage of the Affordable Care Act, the Center and other advocates have expressed concern that CMS has been misinterpreting section 3202 of the Act that limits cost-sharing to the level required in traditional Medicare for SNF care.  We asserted that MA plans should not be permitted to allow cost-sharing for the first 20 days of a SNF stay, as long as the overall cost-sharing is actuarially equivalent to the cost imposed under original Medicare for the complete SNF benefit.  The average stay in a SNF is well under the 100 day benefit.  The previous CMS policy allowing MA plans to front-load their SNF cost-sharing requirements to the first 20 days undermined the protection that these provisions were designed to establish. 

We appreciate CMS addressing this issue, and encourage other cost-sharing limits be imposed on MA plans in line with traditional Medicare, such as prohibiting cost-sharing for home health. 

Part C Cost-Sharing Standards (p. 119)

We are opposed to the proposed increases in cost-sharing limits for Emergency Care/Post Stabilization Care, and are unconvinced by the rationale provided for such increases. While we understand the need to align cost-sharing with actual costs, we strongly discourage implementing any such changes “as an incentive to use primary and specialty care services for routine care and avoid using the emergency room for non-emergent routine services.”[14]

A body of research demonstrates that increasing a beneficiary’s cost-sharing can reduce both unnecessary and necessary care, and this effect is particularly pronounced in low-income populations.[15] This blunt tool can have serious effects on patients seeking necessary care, and other, more tailored, strategies should be tried before relying on cost-sharing as a disincentive to seek out certain types of care.

Medicare beneficiaries, especially those with low incomes, must not be deterred from using appropriate emergency services. Alternatively, we encourage CMS and MA plans to employ positive strategies, such as care management, enhanced education, and lowered or eliminated cost-sharing, to encourage enrollees to appropriately utilize primary and specialty care.

Decreasing Health Disparities in the Quality of Care that Vulnerable Populations Receive (p. 133)

We appreciate that CMS is reinforcing the importance of addressing health disparities in concrete ways.  With respect to limited English proficient plan members, we ask CMS to require plans to track and report data on frequency of use of interpreters and correlate that information to the encounter data of individuals that self-identify as LEP.  This information would provide an important data point to help identify patterns of underservice to specific vulnerable groups.

Section III- Part D

Tiering Exceptions (p. 136)

We appreciate CMS’ clarifications and reminders for plan sponsors about the guidelines for managing Part D tiering exception requests, particularly with respect to mixed tiers that include both brand-name and generic prescription drugs. 

We incorporate here, by reference, the extensive comments on this topic provided by our colleagues at the Medicare Rights Center.   

Issues Unaddressed in Draft 2018 Call Letter

In addition to the issues discussed above, the Center raises the following topics that are not addressed in the draft Call Letter, about which we provide comment, below.

Seamless Conversion Enrollments

The Medicare statute and implementing regulations allow insurance plan sponsors to petition CMS to auto-enroll an individual currently in one of their commercial or Medicaid products into an MA plan when that person becomes Medicare-eligible, a process known as “seamless conversion enrollment.”  In 2016, the Center began to hear from beneficiaries who were seamlessly enrolled into MA plans without their knowledge or consent.[16]

Out of concern that existing consumer protections were not adequate to prevent such enrollments without informed consent, the Center worked with partner organizations and CMS to address these concerns.[17]  In October 2016, CMS issued a temporary moratorium on its acceptance of any new seamless conversion proposals and released previously unavailable data concerning plan sponsors’ use of this process.[18]

As CMS is reviewing the current seamless conversion enrollment policy, we urge CMS to add stronger consumer protections and transparency to the process before lifting the moratorium. We direct CMS to the Joint Advocates letter, linked below, for specific consumer protections, including: write-in confirmation, a Special Enrollment Period (SEP), 1-800-MEDICARE scripts, tailored notifications and outreach for commercial insurance vs. Medicaid managed care plans, and continued and enhanced transparency.  Further, we ask CMS to provide the opportunity for public comment on any proposed or revised seamless conversion policies.

MA Provider Directories

We appreciate and encourage CMS’ ongoing effort to make sure Medicare Advantage provider directories are accurate for Medicare beneficiaries and those that assist them who rely on such directories to make informed decisions about their coverage and provider options.  Recent CMS findings from a review of 54 MA organizations, showing widespread inaccuracies in MA provider directories, underscore the need for further work in this area.[19] In response to these findings, CMS issued additional guidance reiterating the rules MA organizations must follow for provider directories.  

We urge CMS to continue to strengthen required data elements in future regulatory updates.  In addition, as mentioned in previous comments, we continue to believe that provider directory integration in Plan Finder would greatly enhance beneficiaries’ ability to choose among MA plans based upon the criteria most important to them. Incorporating accurate provider directories in a searchable and integrated way would significantly improve the utility of Plan Finder for MA searches and ease plan selection for people with Medicare. As such, we strongly encourage CMS to incorporate this goal in the agency’s planning as it develops future policy on this issue.

            MA Network Adequacy

As CMS is aware, in 2015 the General Accounting Office (GAO) released a report entitled “Medicare Advantage: Actions Needed to Enhance CMS Oversight of Provider Network Adequacy”.[20]  GAO examined several factors relating to CMS’ oversight of MA organization (MAO) network adequacy, and made corresponding findings, including:

• How CMS defines network adequacy and how its criteria compares with other programs: GAO concluded that “MA criteria do not reflect aspects of provider availability, such as how often a provider practices at a given location …[w]ithout taking availability into account, as is done in some other programs, MA provider networks may appear to CMS and beneficiaries as more robust than they actually are.”
• How and when CMS applies its network adequacy criteria: GAO concluded that CMS applies such standards “narrowly” – for example, from 2013 to 2015, CMS’ reviews accounted for less than 1% of all networks; CMS currently facilitates reviews of networks via standardized data collection; GAO concluded that “[u]ntil CMS takes steps to verify MAO provider information … the agency cannot be confident that MAOs meet network adequacy criteria.”
• The extent to which CMS conducts ongoing monitoring of MA organization networks: GAO found that “CMS does not require MAOs to routinely submit updated network information for review” and while CMS may learn of any network adequacy issues through broader oversight and/or complaints, “contrary to internal control standards, CMS does not measure ongoing MAO networks against its current MA criteria.”  GAO concludes: “Because a plan’s providers may change at any time, CMS cannot be assured that networks continue to be adequate and provide sufficient access for enrollees until the agency collects evidence of compliance on a regular basis.”; and
• How CMS ensures that MA organizations inform beneficiaries about terminations: GAO found that while CMS requires MAOs to provide enrollees with advance notice of provider terminations, “the agency has not established information requirements for those notices and does not review sample notices sent to enrollees” and concluded that “[w]ithout a minimum set of required information elements and a check on adherence to them, the agency cannot ensure that MAO communications are clear, accurate, and consistent.”

As a result of this review and these findings, GAO recommended that “[t]he Administrator of CMS should augment oversight of MA networks to address provider availability, verify provider information submitted by MAOs, conduct more periodic reviews of MAO network information, and set minimum information requirements for MAO enrollee notification letters.”  

While, as noted above, CMS has made efforts to address some of the deficiencies highlighted by GAO, so far such efforts appear to be primarily directed at provider directories alone.  CMS noted in the draft 2016 Call Letter that “[t]he data collected through our monitoring activities could drive additional reviews of network adequacy, as well as future monitoring and/or audit-based activities” [emphasis added].  We urge CMS to more broadly expand its oversight and definition of network adequacy, as suggested by GAO.

            Need to Address Provider Network Terminations

We remain disappointed that CMS has taken no further action, either in Call Letters or in rulemaking, to strengthen consumer protections surrounding MAO mid-year provider network terminations. The most effective way to protect consumers from being trapped in their plans after their own doctors are involuntarily terminated is to prohibit MAOs from terminating network providers mid-year without cause.  Not only did CMS retreat from this option in the final 2015 Call Letter, but there has been no attempt to extend the current 30-day advance notice to affected beneficiaries, as also suggested in the 2015’s Draft Call Letter.  Further, CMS has failed to strengthen or otherwise expand the limited special enrollment period (SEP) right available only to beneficiaries affected by “significant” network terminations.  In addition, the availability of this limited SEP right is not adequately expressed in beneficiary-oriented materials, including those issued by plan sponsors (e.g. the Annual Notice of Change) or by CMS (e.g. Medicare & You and the www.medicare.gov website).  More accurate provider directories, while a welcome improvement in consumer information, is not a cure for this disease.

Conclusion

We appreciate the opportunity to submit these comments. For additional information, please contact David Lipschutz, Senior Policy Attorney, dlipschutz@medicareadvocacy.org, and Kata Kertesz, Policy Attorney, kkertesz@medicareadvocacy.org, both at 202-293-5760.

David A. Lipschutz                                                    Kata Kertesz
Senior Policy Attorney                                               Policy Attorney