AHIP Comment

September 3, 2015
Andy Slavitt
Acting Administrator
Centers for Medicare and Medicaid Services
U.S. Department of Health and Human Services
200 Independence Avenue, SW
Washington, DC 20201
Submitted electronically via http://www.regulations.gov
Re: CMS Emergency Clearance Information Collection Request to Support Data
Validation Under the Risk Corridors and MLR programs (CMS-10401/OMB Control
Number 0938-1155)
Dear Administrator Slavitt,
We are writing on behalf of America’s Health Insurance Plans (AHIP) and the Blue Cross and
Blue Shield Association (BCBSA) to offer comments in response to the Department of Health
and Human Services, Centers for Medicare and Medicaid Services (CMS) information collection
requirements related to the risk corridors and medical loss ratio (MLR) programs.
The information collection requirements—outlined in the notice and supporting materials and
issued under emergency review procedures—are intended to assist CMS in conducting program
integrity reviews of data previously submitted and resolve any potential material differences
between the data collected during the EDGE sever process for risk adjustment and reinsurance
and the separate data submission process and requirement for risk corridors and MLR.
We recognize and support the importance of assuring the integrity of the risk corridors and
implementing the program in a timely manner to ensure the ongoing stability of the exchange
risk pool. Our member plans have worked diligently to complete timely submissions of data
required to administer risk adjustment and reinsurance programs (under the distributed data
collection process) and through the risk corridors and MLR forms. We are committed to working
1

collaboratively with CMS to resolve any questions about the risk corridors and MLR data
submissions.
While the Preamble to the 2015 Notice of Benefit and Payment Parameters Final Rule previewed
the approach of validating risk corridors data against other data sources, CMS also recognized
that it would only do this if "other data source is sufficiently reliable and can be appropriately
compared, including with respect to any data submitted through the dedicated distributed data
environment for 2014.”1 Because the data submitted in the risk corridors and MLR reporting
form and the data submitted through the distributed data environment uses separate and distinct
data fields, instructions, requirements and timeframes and is collected for separate programs, the
data in those two submissions are not comparable and differences – potentially of a significant
magnitude – should be expected (See Appendix A). Thus, we believe CMS should avoid using
the term “discrepancies,” which implies that the data from the two sources should be the same,
and should simply refer to material or significant differences in the reported data.
Given the level of detail required to complete the mandatory worksheets, we believe that CMS’
estimate of the burden on issuers is substantially understated. For example, one company said
that their staff had spent 20 hours in meetings trying to understand the process before even
attempting to complete the forms. Another estimated approximately 120 person hours for the
completion of one FEIN submission. For a company with multiple FEINs, the burden for
completing this submission could be hundreds of hours.
We have attached detailed comments and recommendations on the checklist and instructions to
address the comparability of data, illustrate technical issues, and provide clarifications. These
recommendations are based on input derived during this shortened emergency comment period
from our members' health plan operational leaders and technical experts. These
recommendations balance the need for assuring program integrity while promoting smooth and
uniform completion of the forms and completing CMS’s goal of reconciling differences between
the two data sources, as appropriate, in a timely and expeditious manner.
Our comments are aimed at assuring the integrity and validity of the data submission process
while facilitating the timely implementation of the risk corridors program—which plays a critical
role in promoting market stability and affordability for consumers in the health insurance
marketplaces. We believe it is critical for this data validation process be concluded in a timely
manner in order to avoid negative downstream effects on consumer rebates under the MLR
program or on the timetable for completing risk corridors payment transfers. Of note, the postaudit process—issued under final regulations2—provides for a post-payment audit methodology
1
2

2015 Notice of Benefit and Payment Parameters Final Rule. 79 FR 13785
2015 Notice of Benefit and Payment Parameters Final Rule 79 FR 13836

2

to address audit findings specific to the risk corridors programs while a separate but similar
process exists to address data validation under the distributed data collection process.
We look forward to working in partnership with you to resolve any issues arising from this data
validation process as expeditiously as possible in order to promote the shared goal of assuring
effective and timely implementation of these critical programs.
Sincerely,

Sincerely,

Matt Eyles
AHIP

Kris Haltmeyer
BCBSA

3

AHIP-BCBSA Recommendations and Technical Comments on the PRA Notice and
Supporting Materials



The instructions included in the supporting materials should clarify that plans can
document and describe other differences between EDGE server and MLR/RC data
submissions that do not currently appear in the checklist. CMS appears to have
identified three reasons for claims differences between MLR/RC and EDGE data
(capitation, rejected claims, and hospital claims crossing benefit years) and four reasons
for premium differences (billed/earned differences, impact of grace periods, retroactive
adjustments, and partial month proration). As currently formatted, the worksheet could be
read to imply that these are the only appropriate reasons for differences between the two
data sources, which we hope was not CMS’s intent. We believe issuers should be
afforded the opportunity to describe other legitimate differences between MLR/RC and
EDGE data that do not fall into these seven categories. At a minimum, we recommend
changing the titles of Column S on the claims table (p. 16 of the instructions) and
Column Q in the premium table (p. 25 of the instructions) to “Remaining Differences”
and clarifying that these fields should be used to quantify and explain other acceptable
material differences in the data not captured in other fields (e.g. differences driven by
paid date between EDGE data - which could have been as late as early May given the
5/15 grace period - and the MLR/RC paid-through date of March 31.



We recommend that CMS improve the functionality of the process by allowing
plans to save work on the web form and/or provide plans with an alternative
method (e.g., Excel spreadsheet) to address any contingencies, such as a system
failure. Many plans are concerned that CMS has not provided a template for completion
of the claims and premium worksheets, and that the web form must be completed in one
sitting. Given the complexity of the data involved and the importance of later data audits,
we recommend that issuers have the ability to save progress when completing the
checklist and worksheets. This functionality will enhance the accuracy of the data
validation. Alternatively, CMS could provide issuers with an Excel template (including
the macros that will be used to auto calculate certain fields) to assist issuers in making
timely and accurate submissions.



We recommend plans have an additional response option when completing the Risk
Corridors Submission Checklist. The instructions for completing the checklist provide
issuers with three response options: “Y” if the element is accurate, “R” if the element is
not accurate but will be upon resubmission, and “N” if the element is not accurate and
will not be corrected in resubmission because there is no impact on risk corridors
payments or charges or MLR rebates. There are unique situations where issuers’
responses may not fit into these three categories. We recommend either providing issuers
the opportunity to include an explanation for a “Y” response or creating a new response
(e.g. “E”) that would allow issuers to fully explain how the specific element was
completed.
4



Given the extremely compressed timeframe issuers have to complete these
submissions, we recommend CMS identify support staff and have dedicated
mailboxes specifically for questions regarding the submission process. It is critical
that issuers have a contact at CMS that can quickly respond to any problems issuers may
encounter. Again, this is particularly important given that the submission must currently
be completed in one sitting.



While we strongly support the goal of completing the data validation process as
quickly as possible, we recommend CMS also consider a process for limited
extensions to assure the accuracy and completeness of data necessary to administer
these programs. Some plans with large numbers of HIOS IDs are concerned about the
ability to complete this submission within the required timeframe—either September 8 or
September 14. While we recognize the importance of completing this validation as soon
as possible, we recommend CMS consider a process for extensions under extraordinary
circumstances given the volume and complexity of information that issuers must review
in the data validation process.



We are concerned that quantities on the worksheet must be reconciled to 0.25% of
the applicable amount. For many issuers these tolerances could likely only be achieved
on the largest line items. We recommend that reconciliation by tied in the aggregate to a
certain percentage of total premium (e.g., 0.5% of total premium).

We also offer additional technical questions and request clarifications to the instructions and
worksheets (see details below)-o We recommend CMS confirm a typo in Column K1 (page 22 of the instructions).
This field should also exclude Column I1, which is currently not listed.
o Premium report: Column H1 should be able to accept negative values because
adjustments can go either direction.
o List HIOS ID for which issuer has submitted discrepancy report: Do issuers need to
report on all discrepancy reports, even those that have been resolved or only
outstanding discrepancy reports?

5

Appendix A: Differences between EDGE Data and MLR/RC Reporting
The requirements for EDGE server data submissions differ substantially from the requirements
for Medical Loss Ratio/Risk Corridor (MLR/RC) reporting. As a result, it is important to
consider and account for these differences when comparing EDGE data to MLR/RC filings. The
chart below outlines the differences in the datasets for both claims and premium.
Claims

EDGE

MLR/RC

IBNR

Not included

Included3

Service Dates on Claims

Discharge date or service
through date in 2014

Admit date or service from
date in 2014

Runout Period

Claims paid through 4/30,
although EDGE submissions
generally will not include all
claims paid through 4/30 due
to time needed for preparing
and loading data to EDGE

Claims paid through 3/31

Other claims

Only claims processed on
claims system included

Includes all types of claims
payments, including manual
checks, deductions for drug
rebates and other items

Capitation

Rules require that encounters
be re-priced

Actual capitation amounts
paid to providers

EDGE claim rejections

Issuers prioritized claims
impacting reinsurance and risk
adjustment

Included

Orphan claims

Not included in reinsurance
summary report

Included

3

IBNR can be significant and vary based on such factors as HMO vs PPO, benefit design (e.g., deductible levels), use
of capitation, knowledge of large claims, and use of TPAs.

6

Premium

EDGE

MLR/RC

Basis

Intended charged rate

Actual earned premium

Proration of premium and
calculation of member months

Based on actual
beginning/ending enrollment
dates with 30 day months.
CMS calculates member
months by taking the actual
number of days in a period
and dividing it by 30 (even if
the month had 31 days or 28).
Thus contracts with full year
membership will have
premium overstated by the
ratio of 365/360, or 1.4%.

Based on issuer conventions
for proration of partial month
premium. Member months
represent actual months rather
than calculated months.

EDGE issue on proration

CMS system assumes the full
month premium is entered and
calculates the proration, but
some issuers entered prorated
premiums for partial months.

Based on issuer conventions
for proration of partial month
premium

Retroactive premium
terminations (90-day grace
period)

Could include as active
members who subsequently
did not pay premium, e.g. end
of year no-pays may not have
terminations processed before
the EDGE submission
deadline

Retrospective view of
premium; does not include
premium for lapsed members

Retrospective enrollment
changes

Enrollment reconciliation
continued for 2014 after the
EDGE submission and may
not be reflected in EDGE

Retrospective view of
premium; includes
adjustments for retrospective
changes

APTC premium vs full value
of premium for 1st month of
grace period

Includes full value of premium Issuer may only report the
APTC portion of the premium

7