Integration & Coordination for Dual Eligibles
Integration + Coordination: Increasing the Number of Dual Eligibles in an Aligned Medicare/Medicaid Model
A fully-integrated model offers the best outcomes for dual eligibles. For example, when a single health plan covers her Medicare and Medicaid benefits, a dual eligible who needs Medicaid-covered long-term services and supports (LTSS) and Medicare-covered physician services and prescription drugs would have a single plan of care that integrates the two programs’ benefits. This would reduce Medicare-covered avoidable hospitalizations (due to Medicaid LTSS supports that prevent falls at home), and it would reduce Medicaid-covered avoidable long-term nursing facility stays (due to Medicare’s medical services to manage chronic illnesses).
Beyond better health outcomes, integrated care also aligns financial incentives, and actively encourages expenditures in one program that produce equivalent or greater savings in the other program, such as those Medicaid-financed LTSS investments that result in fewer Medicare-paid hospitalizations.
Despite the promise of integrated care, though, two recent reports are a reminder that the scale of integrated care programs remains quite small, at only about 10% of all dual eligibles who receive full Medicaid and Medicare benefits. Why? In considering the challenge of scaling integrated programs, it’s important to recognize one often-overlooked factor: given choice in the Medicare program, a rational dual eligible might choose not to be in an integrated model, because other available Medicare options offer more supplemental benefits, a higher STAR rating, or a broader provider network.
While still favoring an integrated approach with tools like default and passive enrollment, a state could consider adding a coordinated model alongside an integrated approach to move more full benefit dual eligibles into an aligned model designed by a State Medicaid agency, and reduce the number of dual eligibles in a completely unaligned model those individuals otherwise might choose. This note will explore that option, and mention other policy changes too.
GAO Report: Medicare and Medicaid, Alignment of Managed Care Plans for Dual Eligible Beneficiaries, released in March 2020
In March, GAO released a report on Medicare-Medicaid integration for dual eligibles. After noting that there are about 12 million dual eligibles – roughly 8.5 million who qualify for full benefits in both programs, and roughly 3.5 million who are “partial duals” and don’t receive full Medicaid benefits – GAO set the stage for the integration challenge:
In January 2019, of those roughly 8.5 million full-benefit dual eligibles, less than ten percent -- only 818,000 -- were in an integrated/aligned model:
386,000 were in the same health plan for Medicaid and Medicare in a model that utilized a Medicaid managed care organization (MCO) contract, coupled with a Medicare dual eligible special needs plan (DSNP);
388,000 were in a Medicare-Medicaid Plan (MMP) under the Financial Alignment Initiative (FAI) model; and
44,000 were in a Program of All-Inclusive Care for the Elderly (PACE) program.
Also in January 2019, roughly 2.2 million dual eligibles were enrolled in a DSNP. This includes the 386,000 individuals in the aligned model above. It means:
The 386,000 DSNP members in an aligned Medicaid MCO constitute about 18% of all DSNP members, and the remaining 82% of dual eligible in a DSNP are not in an aligned Medicaid plan; and
The vast majority of the 12 million dual eligibles (and the vast majority of 8.5 million full benefit dual eligibles) are not in any DSNP, MMP, or PACE program. In other words, the vast majority of dual eligibles are either in Medicare fee-for-service (FFS), or in a non-DSNP Medicare Advantage plan.
The GAO report focused on integration and alignment challenges in seven states: Arizona, Florida, Kansas, New Jersey, Pennsylvania, Tennessee, and Virginia. GAO selected these states, partly because all of them enroll dual eligibles in Medicaid MCO contracts for managed LTSS. In conducting its work, the GAO interviewed state Medicaid agency officials, other key stakeholders in these states, and three national managed care organizations that participate in Medicaid and Medicare in alignment models in these states.
Medicaid officials reported several challenges in integration:
Six states reported that the Medicaid agency could not determine the state-level quality results of the DSNPs, because the DSNPs were often on multistate Medicare Advantage contracts with CMS, and available quality scores were only reported at the aggregate multistate level. This made it difficult for states to evaluate outcomes, or strategically promote the high-quality local DSNPs.
Six states reported challenges receiving timely data from CMS on soon-to-become Medicare beneficiaries (for reasons other than age), to support default enrollment into an aligned plan with a goal of stickiness to that integrated plan.
Six states noted that they did not have timely or good information on which dual eligibles disenroll from aligned models in favor of other options.
Five states reported challenges in “controlling” how enrollment into Medicare Advantage products (including but not limited to DSNPs) was managed, partly because of sales activities that are permitted by Medicare even when disallowed by Medicaid.
Four states reported challenges competing for enrollment against so-called “lookalike” plans: Medicare Advantage plans that were not DSNPs, but targeted their benefits and sales activities to enroll dual eligibles.
Two states reported that even when the Medicaid MCO and the DSNP were offered by the same parent company, the two sides might have very different provider networks for Medicaid and Medicare, so true alignment in networks did not exist even in a so-called integrated model.
Health Management Associates: Medicare-Medicaid Integration: Integrated Model Enrollment Rates Show Majority of Medicare-Medicaid Dual Eligible Population Not Enrolled, released in April 2020
Under a grant from Arnold Ventures, Health Management Associates (HMA) released a report in April. Consistent with the findings in the GAO report, HMA noted that, while enrollment in integrated models grew at a rapid rate between 2014 and 2019, overall enrollment in some form of integrated model remained at about ten percent (10%) of all full benefit dual eligibles. HMA also found:
In 2019, 15 states (including the District of Columbia) did not offer any integrated option for full benefit dual eligibles.
The states that have participated in the Financial Alignment Initiative (FAI) have seen high opt-out rates, meaning dual eligibles disenroll from the FAI model in favor of other available options.
Policy Options
Several organizations have weighed in on policy options to improve dual eligible take-up in integrated models. Some of these options focus on states, with various carrots and sticks (e.g. enhanced Medicaid financing match rates; greater opportunity to share in Medicare savings; wrapping Medicaid benefits around the DSNP a dual eligible already is enrolled in; a mandate to offer an integrated option; a requirement that Medicaid procurements for dual eligibles sync with the Medicare Advantage calendar and bid schedule; etc.). Some options focus on CMS and the Medicare program (prohibition on lookalikes that are an end-run on state involvement and DSNPs; allow DSNPs to mirror Medicaid’s network adequacy standards to permit statewide DSNPs, assuming a transportation benefit to access providers; etc.)
All of the common policy options take, as a given, that Medicare freedom of choice rules remain in effect – i.e., that a dual eligible will not be required to join an integrated model, but rather will always retain her right to choose from among all the available options in Medicare, including Medicare FFS or a non-DSNP Medicare Advantage plan. This note similarly will assume the presence of Medicare freedom of choice.
And I will make another assumption: that a dual eligible who elects not to be in an integrated model might be making a rational choice, often based on discussions with agents and brokers at the time of enrollment and typically driven by supplemental benefit comparisons, STAR ratings, and whether the provider network is large and open. For example, a rational dual eligible might prefer:
Enrolling in a 4+ STAR plan that uses its extra quality-based revenue from CMS to offer $60-80 extra, each month, in supplemental benefits like dentures, eyeglasses, and over-the-counter drugs, or
To remain in Medicare FFS or join in a Medicare Advantage Preferred Provider Organization with a broader network of providers than the available DSNPs or MMPs.
To join a Medicaid MCO that offers a broad network for Medicaid services like LTSS or specialty behavioral health and a separate organization’s DSNP or non-DSNP Medicare Advantage plan that includes the person’s physicians in its network.
Given this, one reasonable approach a state might take is to focus on a broader approach to participation by DSNPs – for example, welcoming participation by DSNPs that don’t necessarily have a Medicaid contract if they have a 4+ STAR rating, or an open network. This would increase overall DSNP membership in the state.
In this model, the state would use its two contract vehicles -- the Medicaid managed care contract, and the state contract a DSNP is required to file with CMS, to create a framework in which the state requires the Medicaid MCOs and the DSNPs to coordinate the dual eligible’s care and provider experiences.[1] This is the mechanism to drive an organized system of care between Medicaid MCOs and DSNPs that don’t participate in Medicaid program, while broadening the overall membership that is subject to the state’s vision for dual eligibles. Using the two agreements in this way would create a framework in which a dual eligible’s separate health plans work together to achieve positive outcomes. The framework built through these two contracts could include requirements around:
Data sharing on transitions of care;
Sharing plans of care and health risk assessments with each other and/or the collaborative development of a single plan of care;
Joint case conferences on members with complex needs, as identified by either Medicaid criteria (e.g. LTSS, mental illness) or Medicare (e.g., chronic health conditions; dialysis; cancer);
Joint work on discharge planning for all discharges from hospital or post-acute facility;
Warm transfers of member and provider calls between the organizations;
Joint reconciliation of appeals, grievances, fair hearings;
Coordination on utilization management, to minimize disputes on which payer is accountable for coverage of benefits like home health and DME
This approach leverages the final regulation that CMS promulgated in April 2019, which imposed new coordination requirements on DSNPs. While this isn’t as ideal as a fully-integrated model, there is something to be said about making available, as a state partnered-DSNP, all attractive DSNPs in the market. A coordinated approach could achieve the state’s desired outcomes without expecting a dual eligible to make an irrational choice to remain in a DSNP or MMP that offers fewer benefits or a narrower network than other available Medicare options.
In this approach, a state could realize Medicare-related savings based on how they approach the negotiations around the mandatory state-DSNP contract that must be filed with CMS. A state could require inclusion of certain supplemental benefits that offset Medicaid costs. For example, a state could require DSNP plans to cover the 20% Medicaid coinsurance for physician services, or the state could require DSNP coverage of a supplemental benefit like dental benefits. When the DSNP files a benefit like this, Medicare becomes primary, and Medicaid can see savings. The state might overplay its hand, of course, because too many requirements might lead insurers to abandon their DSNP product and seek to move dual eligibles into non-DSNP Medicare Advantage product, so playing this card needs to be done strategically.
While this approach certainly wouldn’t avoid the fundamental issue that a dual eligible might choose something else (Medicare FFS or another Medicare Advantage option), it would broaden the tent of available DSNP options in the state, and likely increase overall DSNP enrollment – by including all DSNPs with higher STARs ratings/better benefits, or broader networks – and it would do so in a way in which the state drives its vision of organizing a system of care for dual eligibles to drive positive outcomes.
For more information:
Medicare and Medicaid: Alignment of Managed Care Plans for Dual-Eligible Beneficiaries. GAO-20-319, issued in March 2020. https://www.gao.gov/assets/710/705345.pdf. In footnote 6, GAO summarizes peer-reviewed evaluations that note better health outcomes in integrated models.
Medicare-Medicaid Integration: Integrated Model Enrollment Rates Show Majority of Medicare-Medicaid Dual Eligible Population Not Enrolled. Health Management Associates, Issue Brief # 1, April 2020. https://www.healthmanagement.com/wp-content/uploads/04-20-2020-Issue-Brief-1-final.pdf
CMS’s integration requirements for DSNPs were finalized in April 2019 and become effective January 1, 2021. See especially 42 CFR Section 422.107 (requiring DSNPs to provide transition of care and admissions data for members who are not enrolled in a Medicaid MCO related to the DSNP) and 42 CFR Section 422.562 (requiring DSNPs to assist a member in accessing Medicaid benefits and resolving Medicaid-related grievances).
MACPAC’s forthcoming Chapters 1 and 2 in the June 2020 report, on integrating care for dual eligibles, were previewed in this presentation in its April 2020 meeting: https://www.macpac.gov/wp-content/uploads/2020/04/Integrating-Care-for-Dually-Eligible-Beneficiaries.pdf
Promoting Integration in Dual-Eligible Special Needs Plans, June 2019 Report, Chapter 12, MEDPAC. http://www.medpac.gov/docs/default-source/reports/jun19_ch12_medpac_reporttocongress_sec.pdf?sfvrsn=0
[1] The contract is sometimes called a MIPPA agreement (for the Medicare Improvements for Patients and Providers Act of 2008 that created the requirement), and it’s sometimes called a SMAC (for State Medicaid Agency Contract, which is the formal name for the document required by CMS to support the annual bid and benefit process for DSNPs). Use the acronym of your choice.