David Schwartz
Head of Global Policy
Cigna Federal Affairs

,···~

3-1r~ Cigna
December 21, 2018

VIA ELECTRONIC SUBMISSION TO www.regulations.gov

Centers for Medicare & Medicaid Services
Departmen t of Health and Human Services
ATIN: CMS-4185-P
PO Box 8013
Baltimore, MD 21244-8013

701 Pennsylvania Avenue, NW
Suite 720
Washington, DC 20004
(202) 719-6499
David.Schwartz@Cigna.com

Re: CMS-4185-P: Medicare and Medicaid Programs: Medicare Advantage, Medicare Prescription Drug Benefit,
Program of All-inclusive Care for the Elderly, Medicaid Fee-For-Service, and Medicaid Managed Care Programs
for Years 2020 and 2021; Policy and Technical Changes
To Whom It May Concern:
Cigna welcomes the opportunity to respond to the proposed rule that would make changes t o the Medicare
Advantage (MA) and Part D programs beginning in 2020 (CMS-4185- P). We appreciate the Centers for M edicare
& Medicaid Services' (CMS or the Agency) efforts to improve the MA and Part D programs for beneficiaries by
advancing access to care through greater use of telehealth services, increasing integration and coordinatio n
between Medicare and Medicaid for dual-eligible enrollees, and making improveme nts to the Star Ratings
System to better measure performance and provide more meaningful quality information to beneficiaries.
We support severa l of the proposals, and believe they will strengthen the MA program by increasing access to
vital services forMA enrollees and support plan efforts to im prove quality for our members. Specifically, we
welcome the proposed flexibility t o offer additional telehealth services to enrollees, support the proposed path
forward for increased integratio n of dual-eligible special needs plans (D-SNPs), and agree with the proposal to
make Medicare Parts A and B claims data available to stand-a lone prescription drug plans (POPs). We have
conce rns about the provisions around provider and prescriber preclusion lists because they do not resolve
ongoi ng questions about how the preclusion requirements will take effect in 2019.
This letter does not include comment on the Risk Adjustment Data Validation (RADV) audit provisions included
in the proposed rule; w e will address those provisions und er separate cover in advance of the announced
deadl ine of April 30, 2019 . We appreciate CMS' recognition that the RADV issues discussed in the proposed rule
are important to the MA program and welcome the Agency's decision to extend the comment period on those
topics through April 30, 2019. We also welcome CMS' decision to disclose additional information and data
regarding the empirical study described in the proposed rule. We hope the additional disclosures will respond to
the issues raised by our prior correspondence,1 and will be made in a sufficiently timely manner to enable
stakeholders to evaluate and provide meaningful comments on the Agency's study.
1

For reference, our prior correspondence is collected and resubmitted as Appendix A to these comments.

"Cigna" is a registered service mark, and, "the 'Tree of Life"' logo is a service mark, of Cigna Intellectual Property, Inc., licensed for
use by Cigna Corporation and its operating subsidiaries. All products and services are provided exclusively by such operating
subsidiaries and not by Cigna Corporation. Such operating subsidiaries include Connecticut General Life Insurance Company
(CGLIC), Cigna Health and Life Insurance Company (CHLIC), and HMO or service company subsidiaries of Cigna Health
Corporation and Cigna Dental Health, Inc.

 CMS-4185-P
December 21 , 2018
Page2

Cigna Corporation, together with its subsidiaries (either individually or collect ively referred to as Cigna), is a
global health services organization dedicated to helping people improve their health, well-being and sense of
security. Our subsidiaries are major providers of medical, dental, disability, life and accident insurance, and
related products and services. Worldwide, we offer peace of mind and a sense of security to our customers
seeking protection for themselves and the ir families at critical points in their lives.
Cigna serves approximately 1.7 million people through ou r MA, Medicare Prescription Drug Program and
Medicare Supplemental products. Our focus on this market has allowed us to develop a unique approach to
health care coverage. We have a deep understanding of the needs and challenges facing both patients and
physicians, and thus have developed an evolving collaborative model that provides greater access to highquality preventive care for our customers while offering physicians what they need to deliver that care.
With that perspective in mind, we offer the following comments on the proposed rule:
Requirements for Medicare Advantage Plans offering Additional Te lehealth Benefits
CMS proposes to implement provisions passed by Congress in early 2018 that allow MA plans to provide
additional telehealth benefits to members, beyond those covered under Medicare Part B, and to consider the
additional benefits as basic benefits when submitting bids. Under the CMS proposal, plans would have flexibility
to determine which services would be appropriately furnished via telehealth, but could not require that
beneficiaries access Medicare covered services via telehealth.
Cigna supports the additional flexibility to offer telehealth benefits provided by the Bipartisan Budget Act of
2018 (BBA) and CMS' proposed implementation as outlined in the rule. We have identified a number of areas
where additional clarification is needed to ensure that plans are operationa lizing the new flexibility
appropriately:
Contracted Providers: We ask CMS to cla rify the requirement that additiona l telehealth benefits be offered by
contracted providers only. Would this allow a plan to contract with an entity such as Teledoc or MDLive, which
offers telehealth consultations with physicians and providers, to be our exclusive telehealth provider for a
selected set of services? How would a plan list such providers in the provider directory? Would t he plan need to
list every provider associated with a contracted service ent ity?

Also, we ask CMS to clarify how plans should address a case where a provider refers a member to a noncontracted telehealth provider. Would the plan be required to pay for such a claim?
Evidence of Coverage: CMS proposes that telehealth services must be identified in the Evidence of Coverage
(EOC) and noted in provider directories made available to members. We ask that CMS be more specific about
what information would need to be listed. We also ask CMS to provide mode l language for use in the EOC listing
of telehealth benefits. For example, should cost sharing be specified by service, or in a sepa rate section
dedicated to telehealth services?
Network Adequacy: Cigna recommends that CMS allow plans t o include telehealth providers to demonstrate
network adequacy in certain circumstances. Specifically, we suggest that CMS allow the following to count
toward network adequacy requirements: 1) mental and behavioral health providers offering services via

 CMS-4185-P
December 21, 2018
Page 3

t elehealth; 2) the availability of telehealth providers to members in rural or underserved areas; or 3) limited
specialties that are shown to lack a sufficient number of providers in selected areas. Allowing telehealth
providers to be counted for network adequacy purposes in these limited circumstances would increase access to
MAin rural and underserved areas generally, and would improve access to critical mental and behavioral health
services in particular. We do not believe it is necessary to count telehealth providers for network adequacy in
other circumstances at this time.
Medicare Diabetes Prevention Program (MDPP) : We ask for clarification on how the MDPP is to be treated for
purposes of the proposed telehealth flexibility. Would plans be able to offer the MDPP benefit via telehealth
beginning in 2020? We understand that CMS has previously rejected telehealth-only provisions of the MDPP.
Cigna has had success using a digital diabetes prevention program in ou r commercially-insur ed population, and
believe that the flexibility to offer a similar program to our MA members would increase access and improve
completion and success rates for this population. Therefore, we urge CMS to specify that MA plans may offer
the MDPP via telehealth in the future.
Diagnoses from Telehealth Encounters: Finally, we ask that CMS clarify how diagnoses identified through a
telehealth service or encounter should be treated with respect to risk adjustment coding submissions to CMS.
Should plans treat telehealth visits and services as in-person visits with respect to the identification of diagnosis
codes? We ask CMS to instruct plans clearly on how to treat such diagnoses.
Dual Eligible Special Needs Plans Provisions
CMS proposes regulations to implement provisions in the BBA that require integration of Medicare and
Medicaid benefits beginning in 2021 for D-SNPs. These regulations would create new expectations of D-SNPs to
coordinate access to Medicaid services for dual-eligible enrollees and to report hospital and skilled nursing
facility (SNF} discharges to state Medicaid agencies for at least a subset of enrollees, as defined by individual
states. CMS also proposes requirements to create an integrated appeals and grievances systems for certain
types of D-SNPs.
Cigna supports the requirements proposed for D-SNPs beginning in 2021. We agree with CMS' stated
understanding of the BBA provisions, which allow for three types of D-SNPs to operate in 2021 and beyond:
Fully-Integrated D-SNPs (FI DE-SNPs), D-SNPs that are part of a parent orga nization that also operates a Medicaid
managed care plan in a state (HIDE-SNPs), and sta nd-alone D-SN Ps that are not FIDE-SNPs or HIDE-SNPs, but
actively assist enrollees in accessing and coordinating Medicaid services. We also support CMS' proposa l to
implement an integrated appeals and grievance process for beneficiaries enrolled in FIDE-SNPs and HIDE-SNPs.
A core principle of the MA program is that competition among private plans translates into higher quality and
greater value for beneficiaries. Allowing the continued operation of all types of D-SNPs ensures that enrollees
and taxpayers will continue to benefit from vigorous competition through access to greater supplementa l
benefits and services, along with lower cost sharing. Restricting that com petition would reduce the value MA
offers to these vulnerable beneficiaries, which we believe is preserved by allowing beneficiaries to proactively
elect the benefit plan most appropriate for their health care needs.
As CMS moves forward in finalizing these provisions and considering any additional D-SNP integration
requirements, we recommend the Agency prioritize the following goals:

 CMS-4185-P
December 21 , 2018
Page4

•

•
•

Provide f lexibility f or D-SNPs to accommodate reasonable state design differences (such as stateimposed variations in managed long-term care services and supports and behavioral health coverage,
carve-outs and capitations);
Reduce complexity for be neficiaries and minimize plan and state administrative burden, where possible,
by building, refining or aligning existing processes; and
Encourage and enable partnerships between states and D-SNPs, perhaps by requiring that st ates and
plans submit a joint plan for achieving a mutually-agreed-upon level of integration, enrollment
alignment, and timelines for meeting FIDE-SNP status or higher levels of integration where possible.

In addition, as CMS moves forward to finalize these rules, we ask CMS to provide clarification on several issues
t hat will be important for successful implementation:
•
Coordination and Assistance : CMS should provide more specific information about expectations that DSNPs coord inate and assist enrollees in accessing Medicaid services, including accessing benefits or help
in submitting grievances or appea ls for Medicaid-covered services. We ask that CMS specify the range of
activities D-SNPs would be expected to undertake as part of the required coordination and assistance.
•

Hospitai/SNF Discharge Data Exchange : While Cigna recognizes that states will implement the reporting
requi rements individually, we ask that CMS provide technical assistance and guideli nes for how such
reporting should be done. Standa rdization of data exchange and health informa tion technology will
contribute to the value of the data for benchmarking and quality improvement activities.

•

Defining " High-risk" Populations: CMS proposes that states would define the population of dual-eligible
enrollees for which D-SNPs wou ld be required to report on hospital and SNF discharges. We ask that
CMS provide guidance to plans and states on how high-risk populations should be defined.

We also ask CMS to undertake provider education about these new requirements as they move forward. The
information that D-SNPs will report to states will only be as accurate or timely as the information they receive
from providers. Prompt and complete claims submissions from hospitals and SNFs will enable states and
M edicaid managed care plans to use the informatio n received to integrate needed Medicaid benef its and
services.
Proposa l for Prescription Drug Plan Sponsors' Access to Medicare Parts A and B Claims Data Extracts
CMS proposes to establish a process for POPs to request and receive Medicare Parts A and B claims data about
plan enrollees, as required under the BBA. CMS outlines proposed permitted and prohibited uses of the data,
along with a timeline f or how and when POPs would receive the data.
Cigna supports the proposed framework for providing access to the Parts A and B data, including the allowed
and prohibited uses. We do, however, ask that CMS clarify how POPs may use the data for fraud and abuse
detection activities. Specifically, we believe that PDP sponsors should be allowed, on a retrospective ba sis, to
use diagnosis information provided in the Parts A and B data to consider whe re use of selected Part D drugs is
for medically accep ted, Pa rt D-eligible indications. Patterns of prescribing for non-medically accepted indications
can be reviewed against information submitted to the plan as part of fraud and abuse detection. CMS should
clarify that while POPs may not use the Parts A and B data to change individual coverage determination
decisions alone, they may review this data as part of an effective fraud and abuse detection program.

 CMS-4185-P
December 21,2018
Page 5

Medicare Advantage and Part D Prescription Drug Plan Quality Rating System
CMS proposes several changes to the way that measure cut-points are determined, with the goal of reducing the
likel ihood of large jumps in those cut-points, and corresponding rating thresholds, from year to year. Cigna
appreciates CMS' efforts to provide greater stability to individual measure cut-points over time. We are
generally supportive of the proposals outlined below, which would smooth the trajectory of change in cut-points
over time and better account for outliers. However, we also believe that CMS should do more to prevent
outliers from skewing measure cut-points and we offer some suggestions for additional steps the Agency should
take.
Mean Resampling : In an effort to reduce the impact of outliers on measure cut-points determined via the

clustering methodology, CMS proposes to randomly separate measure scores into 10 equal-sized groups.
Clustering would be done 10 times, each time leaving one of the 10 groups out. The mean of the 10 runs would
become the measure cut-point.
Cigna appreciates CMS' efforts to reduce the volatility of cut-points, which results from small plans experiencing
large changes in scoring from year to year due to small numbers. The impact of such plans is especially
important in the ratings of POPs, because the population size is limited (49 POPs received a Star rating for 2019).
We believe that mean resampling is a positive step forward as a way to reduce the impact of outliers on
measure cut-points, but we believe CMS should do more. Specifically, we urge CMS to weight individual plan
scores by enrollment to better reflect national beneficiary experience and further reduce the impact of small
plans that may experience large changes in scoring from one year to the next due to small numbers. Using
enrollment-we ighted scores to determine cut-points would reduce the impact of small plan outliers on measure
ratings, and more appropriately reflect the true experience and performance of the majority of MA enrollees.
We also ask that CMS provide additional information on the impact the proposed mean resa mpling would have
on cut-points. For example, CMS should use data for 2019 ratings to conduct a simulation of measure cut-points
using the proposed methodology, and make the simulation available to plans for evaluation .
Cut-point Guardrails: CMS proposes adding a bi-directional cap that restricts cut-point movement both above
and below the prior yea r's cut-point by five percent for measures that have been included in Stars for at least
three years. We are generally very supportive of CMS' efforts to better control and manage year-over-year
variations, especially when driven by outliers.

As with the mean resa mpling, Cigna supports CMS' efforts to better account for large changes in cut-points,
especially for measures that see large shifts from year to year, driven by outliers (e.g., 2019 Comprehensive
Medication Review rates). We support limiting increases in cut-points, but do not believe the limit should apply
when cut-points decline. Plan consolidations, unanticipated changes in the marketplace, or clin ical factors may
result in declines in industry performance that should be reflected in rat ings thresholds. Further, excluding
downward guardrails will allow measures that currently have artificially high cut-points driven by performance
outliers to return to parity more quickly. If CMS does implement bi-directional guardrails, we ask that the
Agency consider how baseline cut-points are set to account for outlier influence on existing cut-points.

 CMS-4185-P
December 21,2018
Page6

With rega rd to the proposed limit on movement of five percent, Cigna asks t hat CMS explain why five percent
was chosen as the th reshold. We also ask that CMS share information on unadjusted cut-points so plans can
understand how performance standards will change in subsequent years.
Finally, we ask CMS to specify how measures that move to display due to significant specificat ion changes would
be treated when they return to active status. Will they be subject t o the three-year prohibition when they return
as active Star measures?
Controlling High Blood Pressure : CMS proposes moving the measure to display for 2020 and 2021 Star ratings
due t o specification changes introd uced by the Nat ional Committee for Quality Assurance (NCQA). Cigna
supports t his proposa l.
Medicare Plan Finder (MPF) Price Accuracy: CMS proposes maintaining the existing MPF Price Accuracy
measure on display for 2020 and 202 1 Star ratings while a new measure specif ication is developed. Cigna
supports the proposal to keep the existing measure on display. However, we do not support the proposed
changes to the measure going forward . We believe that measuring instances where a drug price paid at the
point of sa le exceeds the price reflected on the MPF by a penny is not a meaningful metric for beneficiaries or a
meaningful measure of plan quality. Instead, we believe CMS should measure accuracy based on when a sale
price exceeds the MPF price by more than five percent and the variance is at least $0.50.
Plan All-Cause Readmissions: CMS proposes moving the measure to display for 2021 and 2022 Star ratings due
t o NCQA specification changes. Cigna supports this proposal.
Timeliness Monitoring Project (TMP) Audit: CMS proposes adding an additional regulatory provision t hat would
assign a 1-Star rating to the app licable appeals measure(s) if a contract fails to submit data requested as part of
a TMP aud it. Cigna supports the proposal because it would bring more consistency to Star ratings. We would like
more information on the impact to Star ratings and Star cut-points for applicable measures if plans fail to submit
TMP audit results prior to a regulatory provision being added for 2020. CMS should further consider add ing a
regulatory provisio n requiring that plans be provided with TMP audit results in advance of the First Plan Preview
Period .
Review of Sponsors Data: CMS proposes to require that plan requests to the Independent Review Entity (IRE) or
CMS to review appeals or complaints submitted to the Complaint Tracking Module (CTM) be received by June
30th each year. Re-openings would not be included in the proposed deadline.
While Cigna does agree that plans should make best efforts to be proactive in identifying and addressing data
issues in advance of the plan preview period, we do not support a deadline of June 30th for plan requests to the
IRE for review of appeals or complaint decisions. We recommend that the cut -off date be the end of the First
Plan Preview Period. Removing a plans' ability t o examine IRE data on decisions during the plan preview period
negates the point of that exercise. Despite ongoing review of IRE decisions, a particula r decision may only
emerge as problematic during the plan preview period. For example, new information may emerge that changes
a pla n's interpretation of cases, such as TMP audit results or auditor feedback. Plans should be able to review
the IRE decisions and challenge decisions that were incorrectly decided and result in inaccurate scores for
affected measures.

 CMS-4185-P
December 21 , 2018
Page 7

Extreme and Uncontrollable Circumstances: CMS proposes codifying Extreme and Uncontrollable
Circumstances that generally mirrors the disaster relief policy CMS adopted for the 2019 ratings. Cigna
appreciates CMS' recognition that events such as natural disasters and other emergencies can affect
beneficiaries' health care experience in ways that are beyond the control of the MA organization (MAO), and
should be accounted for in Star ratings. We also appreciate CMS' effort to develop a consistent and stable
methodology for adjusting ratings affected by such disasters. At the same time, we are concerned that codifying
the methodology, particularly when it is still new, would prevent CMS from making adjustments as new
circumstances arise. We recommend that CMS not finalize the methodology for Extreme and Uncontrollable
Circumstances at this time. Instead, CMS should continue to assess the impact of the methodology over several
years and different contingent circumstances to ensure that it is able to accommodate a range of emergency
circumstances.
In addition, we urge CMS to consider the way ratings would be adjusted in cases where a local geographic area
experiences multiple natural disasters or emergencies across several years. For example, in a case where a
measure score for a given year is deemed to be unusable due to an emergency, CMS proposes to use the higher
of the current year and the prior year score to measure performance. What if the prior year's score was itself
affected by a natural disaster? We recommend that CMS cla rify how measure scores would be determined
under such circumstances.
Additional Suggestions for lncentivizing Improvement: CMS has expressed concern that plans have limited
incentives to improve performance on measures that are put on the display page, since these measures are not
included in the calculation of Star ratings. Cigna believes that CMS could create appropriate incentives for
performance by allowing improvement in performance on display measures to be counted as part of the
Improvement Measure for Star ratings.
We propose the addition of an upside-only incentive whereby points would be added to a contract's
performance on the Health Plan Quality Improvement and Drug Plan Quality Improvement measures when
significant year-over-year improvement occurs on qualifying display measures. This adjustment shou ld only be
calculated based on measures that have been on display for at least two consecutive years without material
changes to the measure specifications during that time. We further recommend that the incremental upsid e
adjustment to the Health Plan and Drug Plan Quality Improvement measures not be factored into cut-point
development in order to maintain continued separation of Stars and display measures. We believe this extra
focus on display measures would result in both improved beneficiary outcomes, and a better performance
baseline when display measures move to active status, red ucing the amount of year-over-year change in
measure performance during the first few years.
Preclusion List Requirements for Prescribers in Part D and Individuals and Entities in MA. Cost Plans. and PACE
CMS proposes changes to upcoming rules that require MA and Part D plans to cease paying providers or
pharmacy claims from prescribers who are included on a Provider/Prescriber Preclusion List beginning in 2019.
CMS states that the Provider/Prescriber Preclusion list ("the list") will become available on Ja nuary 1, 2019.
Under cu rrent rules, MAOs and Part D plans will be prohibited from making payments to precluded providers
and prescribers as of April1, 2019. In the rule, CMS proposes to postpone MA and Part D plan requirements
regarding the list until January 1, 2020.

 CMS-4185-P
December 21 , 2018
Page 8

Cigna strongly supports implementation of the list as a means of safeguarding the Medicare program, reducing
fraud, and protecting beneficiaries. We also support many of the changes to the Preclusion list proposed in the
rule, but are concerned that multiple and varying effective dates will lead to confusion for beneficiaries,
providers, prescribers, and plans, and could result in harm to beneficiaries who are unable to access medications
or needed services. Further, we are concerned that without additional guidance from CMS that clearly specifies
how plans are to implement the list, the intended positive outcomes of the effort will be overwhelmed by
beneficiary confusion, provider and prescriber frustration, dissatisfaction with the MA and Part D program, and
significant potential for beneficiary harm.
In light of the numerous changes to the Preclusion list proposed in this rule, along with the add itional changes
and uncertainty emanating from the recent HPMS guidance on precluded prescribers, 2 we urge CMS to delay
implementation of the Preclusion list until it can fully address these concerns, and the many other issues raised
by stakeholders. Assuming the issues and inconsistencies can be addressed in a meaningful way, we believe a
January 1, 2020 effective date across the board for all requirements would be appropriate.
In the proposed ru le, CMS expands upon and clarifies the "precluded provider" provisions finalized in the CY
2019 MA/Pa rt D Policy & Technical Rule. In many cases, Cigna is supportive of these clarifications. Specifically,
Cigna supports efforts by CMS to: (1) shorten the Preclusion list appeal timeframe from nine months to five
months; (2) place providers and prescribers on the Preclusion list after their first level of appeal; and (3) prohibit
beneficiary appeals of payment denials due to Preclusion List placement. These steps would protect
beneficiaries and ensure precluded providers and prescribers do not continue to receive Medicare
reimbursement, while reducing the administrative burden on plans.
However, rather than making corrections to the Preclusion List policies and delaying the effective date of all of
the policies until January 1, 2020, CMS is proposing to take the unusua l step of implementing the Preclusion List
in a three-step, phased-in process. Such a disjointed roll-out could be extremely problematic for plans,
providers, and beneficiaries. As we understand the framework being proposed, CMS would phase in the
Preclusion List in the following timeframe: (1) beginning on Janua ry 1, 2019, the Preclusion List will go into effect
without any of the proposed reforms introduced in this rule; (2) 60 days following publication of the final CY
2020 rule, Medicare plans will be required to implement the new consolidated appeals provisions; and (3)
finally, any other changes made in the 2020 NPRM will not go into effect until January 1, 2020.
These multiple phases of implementation, with shifting policies and effective dates, will be harmful and
confusing. Beneficiaries will face confusion as the rules under which their provider and/or prescriber may or may
not be placed on the Preclusion List change throughout the benefit year. likewise, they may be particularly
confused as to whether they have appeal rights, with the issue not addressed for 2019 but with appeals then
apparently not allowed as of 2020.

2

CMS, HPMS memo, "Preclusion List Requirements," November 2, 2018. Accessible at
https://www.cms.gov/Research-Statistics-Data-and-Systems/Computer-Data-and-Systems/HPMS/HPMSMemos-Archive-Weekly-ltems/SysHPMS-Memo-2018-Weekl-Nov-12.htmi?DLPage=2&DLEntries=10&DLSort=l&DLSortDir=descending

 CMS-4185-P
December 21 , 2018
Page 9

Based on past experience since provider enrollment was first proposed in 2015, Cigna does not believe CMS'
proposal to implement any of the Preclusion List requirements on January 1, 2019 is realistic. Implementin g
these requirements in the proposed manner will challenge Part D plan operations and cause potential harm to
beneficiary access. CMS must provide clear guidance on how MA and Part D plans are expected to implement
the requirements before moving to hold them accountable or expect beneficiaries, providers, or prescribers to
understand the new rules. These operational issues include:
Prescriber Notice: We ask that CMS clarify whether Medicare plans are required to notify providers and

prescribers of inclusion on the Preclusion List. The proposed rule states that a plan "must ensu re that reasonable
efforts are made to notify the prescriber," yet does not identify what these "reasonable efforts" are. What
standard should plans follow?
Beneficiary Appeal Rights: In the proposed rule, CMS notes that it is updating the regulatory text to confirm
that

if payment is denied because the prescriber or provider is on the Preclusion List, the affected beneficiary would
not have the right to appeal that denial. Cigna supports this clarification, but it raises several issues, including:
•

•
•

•

Will enrollees be permitted to appeal the denial of a claim (due to a provider's placement on the
Preclusion List) during CY 2019 given that beneficiary appeals were not add ressed in the 2018 rule
applicable to CY 2019? If so, how does CMS intend to notify beneficiaries of the change in policy for CY
2020?
While we understand the beneficiary cannot appeal the provider's status of preclusion, we also assume
that they cannot appeal the Medicare plan's denial of a claim. We would ask CMS to confirm this.
We understand a beneficiary who received notice the provider was excluded but continued to use that
provider could not appeal the denial of that claim, however, what about the beneficiary who for some
reason did not receive the notice?
Does CMS intend to add language to the notification letter advising the beneficiary that appeal rights
will not apply when a claim is denied due to the precluded prescriber programs? We urge CMS to do so.

In light of the unanswered questions regarding implementa tion of the Preclusion list, and the fact that this
proposed rule would make changes to the rules for implementa tion after the new requirements are scheduled
to go into effect on January 1, 2019, Cigna strongly recommends that CMS delay implementa tion of the effective
date until Janua ry 1, 2020. At a minimum, CMS should make clear that enforcemen t of the new requiremen ts
will not go into effect before 2020 and before CMS ha s released guidance that clarifies the operational issues
discussed above.
We note that CMS has delayed implementa tion of the precluded provider and prescriber provisions several
times in the past, and Cigna does not recommend further delay lightly. As stated earlier, we strongly support
efforts to reduce fraud, waste and abuse, and believe the precluded provider and prescriber provisions would be
a step forward. However, the existing guidance is insufficient to make the rules work for beneficiaries, providers,
prescribers, or the Medicare program and the changes proposed for 2020 would only add to the confusio n.
Beyond setting a new effective date, CMS must commit to working with plans to identify and overcome
operational challenges, and to do so on a timeline that allows for careful implementa tion of t he new rules in
advance of the effective date.

 CMS-4185-P
December 21 , 2018
Page 10

In summary, Cigna supports CMS' efforts to improve the MA and Part D programs through increased
use of
telehealth , improvem ents to the Stars Rating System, and greater integratio n of Medicare and
Medicaid services
for dual-eligible enrollees. We have serious concerns with the implemen tation of the provider
and prescribe r
preclusion lists, and urge CMS to work with MAOs and Part D sponsors to ensure the new requireme
nts improve
the integrity of the programs without ca using ha rm to enrollees.
Thank you for your considera tion of these comments .
Respectfully,

'

a
'

 CMS-4185-P
December2 1 , 2018
Page 11

Appendix A: Prior Correspondence on RADV Audit Provisions

November 5, 2018
Jonathan Smith, Senior Technical Advisor
Joanne Davis, Senior Analyst
Centers for Medicare & Medicaid Services
Department of Health and Human Services
Attention: CMS-4 185- P
• Mail Stop C4-26-05, 7500 Security Boulevard
Baltimore, MD 2 1244-1850
Re:

CMS~ 185- P, Proposed Rule: Medicare and Medicaid Programs; Pol icy and Technical
Changes to the Med icare Advantage, Medicare Prescription Drug Benefit, Program of
All-Inclusive Care for the Elderly (PACE), Medicaid Fee-for-Ser vice, and Medicaid
Managed Care Programs for Years 2020 and 202 1

Dear Mr. Smi th and Ms. Davis:
You are listed as the contacts for further information in the proposed rule that the Centers for
Medicare and Medicaid Services (CMS) released on October 26, 2018 and published in the Federal
Register on Novem ber 1, 2018. See 83 Fed. Reg. 54,982. I write on behalf of Cigna Corporatio n, whi ch
provides health insurance coverage to hundreds of thousands of Americans as part of the Medicare
Advantage (MA) program. As a pat1icipant in the MA program, Cigna has a direct interest in the proposed
rule.
Cigna values its partnership with CMS. Both CMS and Cigna have a shared commitment to the MA
program. We believe that our shared commitme nt to the program has contributed in important ways to its
success.
Because of that shared commitment, Cigna has repeatedly called on CMS to work with it and other
stakeholders in addressing the issue of RADV extrapolation and a fee for service adjuster. Unfortunat ely,
the proposed rule was not the product of that kind of cooperative, collaborative process. Cigna hopes that
the process, going forward, can be improved to reflect a more cooperative, collaborati ve approach .
In that regard, it wi ll be important that Cigna and other stakeholders have a meaningfu l opportunity
to submit comments on the proposed rule. At present, however, Cigna and other stakeholders are unable
to do so because CM S has failed to release the data and analysis upon which it relied in issuing the
proposed rule.
The need for a meaningfu l opportunity to comment could not be more important here. In a sharp
reversal of its prior po li cy CMS proposes "[t]o establish that CMS would use extrapolatio n in RADV
contract-lev el audits," retroactively to payment year 20 I I, and without a fee- for-service (FFS) Adjuster.
83 Fed. Reg. at 54,984.

 CMS-4185-P
December 21, 2018
Page 12

The proposed rule purp011s to rely, among other things, on various data and analyses that CMS
asse11s supports its position. Much of this information, however, is either undisclosed or only
incompletely disclosed. For example, CMS has not adequately disclosed or explained:
•

The "FFS data" from "the 2004-2005 file used for model calibration" (see Tech. App'x 5);

•

The "MA sample" consisting of the "two million records that were sampled from the 20 II
overpayment run (see Tech. App'x 5);

•

The "sample of FFS claims and associated medical records" that were collected from the (CERT)
audit" (see Tech. App'x 6);

•

The "subset of 8,630 unique claims" from the CERT data used "for a RADV -like medical record
review" (see Tech. App'x 6);

•

The " HCCs mapped from diagnoses on the claims" from the CERT data used for the "RADV -like
medical record review" (see Tech. App'x 6);

•

The calculation of the "average number of underlying claims" per beneficiary in the 2004-2005
fee-for-service dataset (see Tech. App'x 9, table 2b);

•

The bases for CMS's conclusion that each claim has an "independent" and equal probability of
being unsupported (see Tech. App' x 9);

•

The "estimate [ofJ a proxy for the CMS-HCC model on the original uncon·ected dataset and
estimate [of] the original factors" (see Tech. App'x 13);

•

The application of "these original factors to the MA sample" and the resulting estimate of the
"original risk scores for the uncorrected calibrating dataset" (see Tech. App'x 13);

•

The "simulation[s] calculating 50 cases of the impact analysis" (see Tech. App'x 15);

•

The results from each such simulation, including the "50 independent corrected FFS data files,"
the 50 "re-calibrat[ions of] the CMS-HCC model," the 50 "re-normaliz[ations of] a new set of risk
scores," and the 50 "average[s] of the percentage difference in risk score" from the original score
(see Tech. App 'x 15- 16);

•

Other data and analysis CMS considered in undertaking its review; and

•

Communications, reports, or other information received or exchanged with third parties, including
consultants, that CMS considered or relied upon in conducting its analysis.

These are just examples and are not an exhaustive listing of all of the undisclosed data and analysis.
We urge CMS to immediately make publicly available all of the data, studies, and technical analyses
underlying the proposal, as required by the Administrative Procedure Act (APA). "Under AP A notice and
comment requirements, '[a]mong the information that must be revealed for public evaluation are the
' technical studies and data' upon which the agency relies [in its rulemaking]. '" Am. Radio Relay League,

 CMS-4185-P
December 21, 201 8
Page 13

Inc. v. FCC, 524 F.3d 227, 236 (D.C. Cir. 2008) (alterations in original). "[l]n order to allow for useful
criticism, it is especially important for the agency to identify and make available technical studies and
data that it has employed in reaching the decisions to propose particular rules." !d. (quoting Conn. Light
& Power Co. v. NRC, 673 F.2d 525, 530 (D.C. Cir. 1982)). Indeed, unless commenters are able to review
and digest this information and assess the agency's use thereof, " useful criticism" is impossible. See id.
That is true especially where, as here, the agency's proposal incorporates complex technical and statistical
analyses: "Public notice and comment regarding relied-upon technical analysis ... are ' [t]he safety valves
in the use of .. . sophisticated methodology. "' !d. (quoting Sierra Club v. Castle, 657 F.2d 298, 334, 39798 & n.484 (D.C. Cir. 1981 )); see also Engine Mfrs. Ass'n v. EPA, 20 F.3d 1177, 1182 (D.C. Cir. 1994)
(an agency must make public the "basis for ... key assertions'' in its analysis).
CMS's failure to release the required data and analysis has already prejudiced Cigna's and other
stakeholders' ability to comment on the proposed rule, given the complexity of CMS's analysis and the
short time frame provided to respond to the proposed rule. Release of this information is necessary to
"ensure that [CMS's) regulations are tested through exposure to public comment, []afford affected parties
an opportunity to present comment and evidence to support their positions, and thereby [ ] enhance the
quality of judicial review." Am. Radio Relay League, 524 F.3d at 236. It is also required by law. See id.
(agency " fai led to comply with the APA by not disclosing in full certain studies by its staff upon which
[it] relied in promul gating the rule"). We ask that CMS immediately re lease to the public all of the data,
studies, and analyses on which the proposal relies.
Please do not hesitate to contact me if you have any questions regarding this request.

Sincerely,

cw~

William A. Sarraille
cc:

Janice Hoffman, Office of the General Counsel,
Department of Health & Human Services

 CMS-4185-P
December 21 , 2018
Page 14

November 27, 2018
VIA EMAIL AND OVERNIGH T MAIL
Jonathan Smith
Center for Program Integrity
Centers for Medicare & Medicaid Services
7500 Security Boulevard
Baltimore, MD 2 I 244

Joanne Davis
Medicare Plan Payment Group
Division ofPayment Validation
Centers for Medicare & Medicaid Services
7500 Security Boulevard
Baltimore, MD 2 1244
Re:

CMS-4 I 85-P, Proposed Rule: Medicare and Medicaid Programs; Poli cy and Technical
Changes to the Medicare Advantage, Medicare Prescription Drug Benefit, Program of
All-Inclusive Care for the Elderly (PACE), Medicaid Fee-for-Servi ce, and Medicai d
Managed Care Programs for Years 2020 and 2021

Dear Mr. Smith and Ms. Davis:
1 am writing on behalf of Cigna Corporation (Cigna) to follow up on my letter of November 5,
2018, concerning the Medicare Advantage (MA) and Part 0 proposed rule that the Centers for Medicare
& Medicaid Services (CMS) published on November 1, 2018. See 83 Fed. Reg. 54,982. As expla ined in
my prior letter, the Administrative Procedure Act (APA) requires that Cigna and other stakeholders have
a meaningful opportunity to submit comments on the proposed rule. That can only occur if Cigna and
others have access to the data, methodology, assumptions, and analysis upon which CMS relied in issuing
the proposed rule.

My colleague and I spoke with your colleagues in the Office of General Counsel, copied below,
on November 19, 2018. We are grateful for their willingness to discuss our client's serious concerns about
the access issue. We discussed in that call the possibil ity of a joint policy and legal meeting to potentially
continue the discussion. We believe that a fUJther, substantive discussion would be beneficial, and we are
available to meet at your convenience.
In the telephone call, we also discussed this supplementary letter, which your colleagues thought
might be helpful. We discussed the following topics: ( I) additional concerns that have arisen as Cigna
has attempted to analyze the proposed rule; (2) the mechanisms by which the requested information, to
the extent that it includes Protected Health Information (PHI) within the meaning of the Health
Information Portability and Accountability Act of 1996 (H fPAA), can be shared with Cigna and others;
and (3) whether C igna could prioritize its requests for access. I address each of these in turn.

 CMS-4185- P
December21 , 2018
Page 15

First, Cigna's concerns about the level of access provided and the implication s for the ability of
our client and others to comment meaningfu lly have only increased since I first wrote to you. Specificall y,
Cigna has identified the followin g additional questions and issues that were not adequately answered or
disclosed in the proposed rule:

3

~

•

Did CMS develop a protocol or statistical analysis plan prior to conducting the study
described in the proposed rule? It is well recognized that the risk of bias increases
substantially if a study does not follow an advance protocol and/or statistical analysis plan.
If investigators fa il to specify how a study or analysis wi ll proceed in advance, they can
alter their methodologies in ways that affect the ultimate results. CMS therefore must
specify whether a protocol or similar plan was developed in advance of the study discussed
in the proposed rule and make any such protocol or plan publicly available.

•

Did CM S combine data from multiple Comprehen sive Error Rate Testing {CERT) audits?
The Technical Appendix (TA) states that CMS analyzed CERT data "for CY 2008 dates
of service." However, CERT audits do not appear to be organized by calendar year.
CMS 's website states that the 2009 CERT audit reviewed claims submitted between Apri l
I, 2008 and March 31 , 2009. 3 Claims from the first quarter of 2008 were, presumably ,
reviewed during a prior audit. CMS must disclose whether and how it consolidated data
from multiple CERT audits, along with any analyses conducted to assess the impact of
such consolidation. We note that the CERT methodologies were substantially strengthened
between the 2008 and 2009 audits, which led to a sharp increase in the number of claims
deemed to be improper payments based on incomplete or mi ssing medical record
documentation. 4

•

How did CMS create its subset ofCERT data? The TA also indicates that CMS created "a
subset of 8,630 unique claims." Although the TA describes " inclusion" criteria, it does not
specify whether additional exclusion criteria were used to create the subset. For instance,
CMS likely excluded claims classified by the CERT Contractor as "No Documentati on
Errors." CMS also may have excluded claims classified as "Insufficient Documentation
Errors." CMS must disclose all exclusion criteria used to create its CERT data subset, and
any analyses conducted to assess the impact of those criteria. We note that excluding
claims where the provider fail ed to submit complete documentation would obviously bias
the subset towards supported diagnoses.

•

Which specific diagnosis codes were found to be discrepant in CMS 's review of its CERT
subset? The T A discloses HCC-level discrepancy rates at Table 2A, but not the specific
diagnosis codes deemed to be discrepancies. Because most HCCs include multiple ICD-9

https : //www . ems . gov /Resenrc h-Statisti cs Data - and - Systems / Monitm ing
091ams/ Medicare fFS-Comp l1ance - Pro grams / CERT/ CERT - Reports

Itemn / Nov~mbci:2009CERTReport.html?DI..Page=2&DLEntries = l0&DLSort = O&DLSortD
il = dencendi

ng
4

See Improper Medicare Fee-For- Service Payments Report - November 2009, at 11-12.

 CMS-4185-P
December 21 , 2018
Page 16

codes, Cigna cannot review and comment meaningfully on the Agency's observed
discrepancy rates without knowing which specific conditions were at issue.
•

What analysis supports the Agency's decision to estimate "beneficiary-level" discrepanc y
rates? The Executive Summary (ES) describes, at footnote 8, a prior analysis of claimlevel discrepancy rates in the FFS program. Documents obtained through the Freedom of
Informatio n Act (FOlA) indicate that the prior anal ysis was also conducted with a subset
of CERT data, potentially the same subset described in connection with the proposed rule.
However, the ES criticizes the prior analysis on the ground that it failed to estimate
"beneficiary-level" discrepancy rates. CMS must disclose any analyses supporting that
critique and/or the Agency' s decision to exponentially reduce the observed, claim-level
discrepancy rates by simply multiplying probabilit ies. We note that this is appears to be a
results-driven change in methodology.

•

CMS must disclose all data, results, and methodologies from its prior review ofCERT data.
The Agency' s prior review of CERT data is centrally important to Cigna' s ability (and that
of other stakeholde rs) to assess and comment on the study described in the proposed rule.
See, e.g. , Ramapough MI. indians v. Babbill, No. 98-2136,2 000 U.S. Dist. LEX lS 14479,
* 15 (D.D.C. 2000) ("An agency acts arbitrari ly and capriciously when it ignores studies
and data in its own fi les."). Yet none of the data, methodologies, or results from that prior
review have been disclosed. As a result, Cigna and other stakeholde rs cannot evaluate
CMS's assertions, including the highly suspect claim that the prior review "did not analyze
the effect of FFS diagnosis error on the CMS-HCC model." ES at 4 n.8. We note that
FOIA documents describe the prior review as an e ffort to "estimate[] how much higher [a]
plan 's payment would be if the model had been built using perfect data."

•

CMS must fully disclose the methodologies for the simulation s described in the proposed
rule. The proposed rule describes "fifty simulation s" based on estimates of beneficiarylevel error rates. Neither the ES nor theTA, clearly describes how the simulations were
run. This information must be provided before Cigna and other stakeholders can comment
meaningfu lly on the proposed rule.

•

CMS must disclose the recalculated coefficients resulting from the simulations described
in the proposed rule. The proposed rule indicates that CMS recalculated the I-ICC risk
score coefficients 50 times based on random deletions of HCCs at the estimated
beneficiary-level discrepanc y rates. None of the resulting coefficients have been disclosed.
Disclosure of the simulated coefficients is essential to enable Cigna and other stakeholde rs
to evaluate and respond to the Proposed Rule.

•

CMS must disclose its normalization methodology: TheTA describes a process by which
CMS developed new coefficients generated from the "simulated corrected data." TA at
13. TheTA then states: " In the next step, we take the new coeffici ents and apply them on
the origi na l FFS data set, normalizing a new set of relative factors to one." ld. Were the
new coefficients normalized to 1.0 using the "corrected data" from which they were
generated or the original FFS data that still contains the FFS diagnoses later removed to

 CMS-4185-P
December21 , 2018
Page 17

generate the "simulated corrected data" ? Please describe in detail how CMS conducted the
normalization. This is an important step of CMS's methodology and must be addressed
before Cigna and other stakeholders can comment meaningfully on the proposed rul e.
•

CMS must disclose how the " MA Sample" was created. The TA states that each time
corrected HCC coefficients were simulated, the coefficients were "applied to the
beneficiary profiles in the MA sample." T heTA further states that the MA sample was
created from " the 20 11 overpayment run." CMS must disclose whether the MA sample
reflects beneficiary profiles before or after retractions were applied. CMS also must
disclose any and all inclusion or exclusion criteria used to create the MA sample.

Second, there are, in our view, multiple avenues under HIPAA to release the information at issue
to Cigna and other affected issuer stakeholders. As an initial matter, we do not agree that HIPAA is
relevant to many of our requests. For instance, a full and fair disclosure of the Agency' s assumptions and
methodologies would not reveal any PHI.
As to disclosure of the underlying data and medical records, the most fundamental point is that,
for the reasons outlined in my November 5 letter, CMS is required under the AP A to release the
information underlying its proposal to pem1i t meaningful cri tique of the proposed rule. HIPAA
regulations specifically state that covered entities may disclose PHI to the extent that such disclosures are
required by law. 45 C.F.R. § I 64.5 12(a).
CMS also may disclose PHI for payment purposes. 45 C.F.R. § 164.502(a)(l )(ii); id. §
164.506(c)(3). T he proposed rule is, in relevant part, a payment rule, as the presence or absence of a feefor-service adjuster has a direct effect on payment. See, e.g. , 83 Fed. Reg. at 55038 (" We stated that the
FFS Adjuster would calculate a permissible level of payment error (for example, a percentage of the total
payments made on an MA contract in a given year) and limit RADV audit recovery to payment errors
above that level. "). Because Cigna is itself a covered entity and would use the information for purposes
of helping to determine appropriate payments (and to help identify potential overpayments), the disclosure
of the information by CMS to Cigna is permissible under HIPAA.
HIPAA a lso pem1its the disclosure of PHI for research purposes. See, e.g., 45 C.F.R. §
l64.512(i)( 1). Cigna and others wish to conduct research on the study reported in the proposed rule. That
research is perm itted under the applicable regulations. See id.
Finally, whi le fully de-identified data would not provide Cigna and others the data elements
required to review and comment meaningfully on CMS ' s proposal, CMS could disclose the necessary
information pursuant to HIPAA ' s expert determination de-identification methodology, provided that
sufficient information is retained to permit meaningful evaluation of the analyses. 45 C.F.R. §
164.514(b)( I).

Third , with respect to the question of prioritization, we stated during the telephone call on
November 19 that this misses the point. To meaningfully comment on the proposed rule, Cigna needs to
recreate the model that CMS advances. Without all of the data, analysis, assumptions, and methodology
elements used by CMS, Cigna cannot evaluate the study described in the proposed rule. Comment will

 CMS-4185-P
December 21 , 2018
Page 18

necessarily and inappropriately be restricted if the study described in the proposed rule is not fu
ll y
disclosed. See, e.g., Am. Radio Relay League, Inc. v. FCC, 524 F.3d 227, 236 (D.C. Ci r. 2008); Engine
Mfrs. Ass 'n v. E.P.A., 20 F.3d 11 77, 1182 (D.C. Cir. 1994); see also Nat 'I Parks Conservation Ass 'n
v.
EPA , 803 F.Jd 15 1, 167 (3d Cir. 20 15) (" [W]e, as a reviewing court, need an agency to show its work
before we can accept its conclusions.")
Since we understand that CMS requires additiona l time to consider disclosure of the requested
information, Cigna asks that CMS extend the comment period on the proposed rule. An extension
is
necessary so that Cigna can obtain access to the information with sufficient time to permit it to review
and
comment on the proposed rule. As indicated in my Novembe r 5 letter, we believe that the failure
to
provide access to the missing data, analysis, and other information has already prejudiced Cigna and other
stakeholders.
We would be happy to discuss our concerns further, and we hope the Agency would meet with us,
as is traditionally the case during pending rules. It would be unfortunate if CMS would break from
its
settled practice and, in the case of this extremely important rule, refuse to meet with interested stakehold
ers
like Cigna.
We look forward to your response.
Sincerely ,

William A. Sarrai lle

cc:

Janice Hoffman, Office of the General Counsel, Department of Health & Human Services
Susan Lyons, Office of the General Counsel, Department of Health & Human Services
Sean Griffin, Sidley Austin LLP