As expected, the final 2018 payment notice and Call Letter for Medicare Advantage and Part D plans posted April 3 contained no drastic alterations to the MA program and few deviations from policies proposed 60 days earlier in the Advance Notice. But one significant change that gave MA plans a bit of relief was a partial retreat on the percent of encounter data used in risk scores, although industry experts are quick to remind plan sponsors that encounter data aren’t going away and that CMS will continue its efforts to make sure plans are submitting accurate data. Meanwhile, plans were disappointed with CMS’s conclusion that it does not have the authority to eliminate the Affordable Care Act’s ceiling on payment rates (aka benchmark caps) and the apparent return of the ACA-established health insurer fee that was temporarily set aside for 2017.

CMS in 2012 began collecting data from MA organizations (MAOs) through the encounter data system and in 2016 started phasing in EDS-based payments, beginning with 10% of the payment based on EDS scoring, while the other 90% came from the old risk adjustment payment system (RAPS) that relies on a much less complex set of data elements. CMS’s intent was to advance to a 50/50 mix for 2017, but it revised the blend to 25% EDS/75% RAPS after hearing the concerns of stakeholders about plan readiness and the quality of the data coming from providers (MAN 4/3/16, p. 1). For 2018, CMS in the February draft notice proposed keeping that same blend but in the final notice walked it back even further to 15% EDS/85% RAPS.

CMS Scales Back Use of EDS

Ankur Goel, a partner in the Washington, D.C., office of the law firm McDermott Will & Emery, in an April 6 client advisory suggested that CMS’s latest revision to the EDS/RAPS blend is a clear indicator that CMS recognizes that there are numerous problems with encounter data, but it “leaves open the prospect that risk scores will not consistently and accurately reflect the health status of an MAO’s member population.”

While EDS is theoretically supposed to capture the same diagnoses in RAPS, various studies have demonstrated that there are significant differences in the RAPS and EDS scores that are being reported back to plans, and researchers say the filtering logic used by CMS is leading to diagnoses that would have been included in RAPS not being applied to the risk score. Recent research from Avalere Health LLC, for example, showed that the average EDS risk score was 16% lower in the 2016 payment year than the average RAPS-based score (MAN 3/2/17, p. 1). And Milliman, which used a slightly different plan composition and methodology to conduct its own study of select MAOs’ risk scores, found that those based on encounter data for 2016 were on average 4% lower than RAPS-based scores.

These differences can have major reimbursement implications for plans. “If a patient has [congestive heart failure] and the risk score isn’t captured, you’ve still got the CHF member, and you have to provide the services. And the risk scores are supposed to be connected with the cost of care. So if you don’t get credit for the disease state that that person has, you’re paying for them, and you just don’t get the money. It’s a big deal,” remarked Stephen Wood, cofounder and managing partner with Clear View Solutions, LLC, speaking during a session of the National Medicare Advantage Summit held April 5-7 in Arlington, Va.

CMS in the Advance Notice asked for feedback on the idea of a “uniform industry-wide adjustment” to the EDS data to account for the differences, but it abandoned that idea in the final notice. Tom Hutchinson, a strategic advisor with the EBG Advisors unit of the health care law firm Epstein Becker & Green and a former top CMS official who also spoke at the National Medicare Advantage Summit, said the problem with such an adjuster is that it would have essentially given “everybody a break” when certain plans were struggling with their scores and others had no issues. “So instead of doing an across-the-board adjuster, going back…to 15% in a sense is a plan-specific adjuster. If you’ve been really good at getting your encounter data in, you’re not getting much of a plus on that. If you haven’t done all that good, that’s actually a fairly decent adjuster.”

CMS’s original goal was to transition to full use of EDS by 2020, but with the latest revision, its timeline is unknown. Sources say that after publication of the final Call Letter, CMS stated during a user call that it was not announcing its plan for percentages in future years. CMS also has twice pushed back the deadline for submitting calendar year 2017 encounter data.

CMS Will Continue to Monitor EDS Compliance

While CMS is scaling back use of encounter data in the risk score calculation, it will continue to enhance monitoring and compliance activities to ensure the accuracy and completeness of data being submitted. CMS has been conducting basic monitoring of MAOs’ encounter data submissions and will now be using performance measures tied to data submission to guide oversight and enforcement in this area. As stated in the draft Call Letter, CMS has identified several performance measures — four on operational performance and three on completeness performance — and will communicate acceptable performance thresholds through future memos and other guidance. Taking an “incremental approach,” CMS will use these measures to identify contracts that are failing to meet performance standards and will conduct compliance activity, “including but not limited to notices of non-compliance, warning letters, and corrective action plans as needed to improve performance.”

CMS also is conducting site visits with a sample of MAOs to understand their different approaches to and issues with encounter data processing, and to identify areas where CMS can improve technical assistance and guidance. Speaking at the same conference session as Wood, Inovalon Senior Director Arati Swadi reminded plan sponsors that this effort is not an audit and represents an opportunity for plans to provide feedback on how the system can be improved. She said CMS is in the process of sharing “very detailed” questionnaires with about 20 plans across the U.S. in an effort to “understand how the systems are behaving.”

She continued, “I think it’s going to be very collaborative, and we are going to make sure we are giving as much feedback as possible when we do meet with CMS.” More information on the benchmarking of encounter data and a simplification of the many data requirements involved in the EDS are two critical areas she said she hopes many plans address in their responses.

CMS Maintains EGWP Pay Blend, CAI

Other changes finalized in the recent payment notice and Call Letter include:

  • Plans will on average see a pay boost of 0.45%, compared with the 0.25% average increase that was included in the Advance Notice. Factoring in an anticipated upward coding trend of 2.5%, plans will see a total pay increase of 2.95%, estimated CMS.

  • In determining payment levels for MA Employer Group Waiver Plans, CMS will continue using a blend of individual market plan bids and EGWP bids from the prior payment year. This is a departure from its original plan of moving from a separate bidding process to one based on the individual market that would result in an overall 2.5% payment reduction by 2018.

  • Stabilizing MA plan payments in Puerto Rico, rates will take into account the significant increases in the fee-for-service physician fee schedule there, and a majority of counties will qualify for double quality bonuses.

  • CMS will continue to use the Categorical Adjustment Index (CAI) to address disparities among low-income subsidy/dual eligible plans and non-LIS/duals plans, although it will exclude plan all-cause readmissions and reducing the risk of falling. CMS also will establish separate adequacy evaluations of provider networks specific to MA Special Needs Plans.

Meanwhile, CMS noted that a “large number” of commenters expressed concern that the current system diminishes incentives for high-quality plans that offer services in higher-cost areas since top-rated plans are not receiving their deserved quality bonus payments because of the rate ceiling established by the ACA. However, CMS responded that it has not “identified discretion” under the law to do away with applying the pre-ACA rate cap or exclude the bonus payment from the cap calculation.

New RFI Seeks Ideas on MA Transformation

In addition, CMS for the first time used the notice to make a request for information (RFI) soliciting ideas on ways CMS can bring more transparency, flexibility, program simplification and innovation to the MA and Part D programs that could be incorporated into future regulatory, subregulatory, policy, practice and procedural changes from the agency. Responses are due April 24.

“The RFI is written broadly, and stakeholders can provide feedback on many topics,” observes Mike Adelberg, a former top CMS MA official who is now principal with FaegreBD Consulting. “The agency likely cannot make wholesale program changes based on feedback, but this is a particularly good opportunity to seek small bore changes, perhaps tweaks to star ratings, benefits or compliance rules and processes.”

During the National Medicare Advantage Summit, Humana Inc. Vice President, Public Policy Analysis, Mark Newsom called the RFI an opportunity to take MA “into the future” and streamline the program, whether it’s through simplified beneficiary communication requirements or alignment between performance measurement under the provider payment portion of the Medicare Access and CHIP Reauthorization Act and MA star quality ratings that reduces provider burden. Hutchinson added that the RFI indicates there is the potential to see some star ratings changes between now and when CMS releases its annual request for comments on enhancements to the program. The final notice included a few changes around star ratings, including a decision to retain the current Beneficiary Access and Performance Problems measure and plans to include a revised version as a 2019 display measure.

View the notice at