You asked for information on Massachusetts' Medicare Advantage-Special Needs Plans (SNPs), which offer coordinated care to frail elderly individuals who are eligible for both Medicare and Medicaid (dually eligible) through managed care organizations.

SUMMARY

The Medicare Prescription Drug, Improvement and Modernization Act of 2003 authorized states to create special needs plans (SNPs) for their frail elderly and younger disabled residents using the managed care model begun under Medicare Advantage (MA)(previously called Medicare +Choice). In addition to saving money, Congress hoped that the model will better coordinate acute and chronic care services, particularly for the dually eligible, whose needs and care costs tend to be the highest.

In 2006, over 80% of the 276 approved SNPs were planning to serve dually eligible beneficiaries, and the Center for Medicare Advocacy reports that 321 SNPs are serving dually eligible people in 2007. A big incentive for the plans has been Medicare's willingness to pay risk adjusted rates, taking into account the higher risk the plans face when serving the dually eligible. Many plans under the original MA had dropped that business largely because of the inadequacy of Medicare reimbursements.

Massachusetts was poised to start SNP as it had several years of experience offering managed Medicare and Medicaid for the frail, dually eligible, first with the federally authorized Program of All-Inclusive Care for the Elderly (PACE), followed by a Medicare demonstration program, Senior Care Options (SCO), which integrated PACE into its service delivery model.

Massachusetts SNPs have been operational a relatively short time. But reviews of the SCO experience suggest that care costs in the dually eligible SNPs could be high. It appears to be too early to assess whether these plans will ultimately keep more people in the community and save Medicaid and Medicare funds in the long run.

FEDERAL LAW AND CREATION OF SNPS

As part of the Medicare Modernization Act of 2003, Congress created the opportunity for MA health plans to develop SNPs as a way to better serve individuals eligible for Medicare, many of whom had chronic health problems but whose care management was not being monitored. Where the traditional MA plans offered managed Medicare to relatively healthy seniors, SNPs would offer coordinated and managed care to individuals with chronic health conditions, possibly preventing more costly health interventions (e.g., emergency rooms). Unlike other MA programs, Congress allowed the SNPs to limit enrollment to people who (1) are living in institutions, (2) are dually eligible for Medicaid and Medicare, or (3) have severe or chronic disabling conditions.

For the dually eligible, this offered access to two separate revenue streams, supplemental funds for services not normally available, and care coordination. Providers receive capitation-based, instead of fee-for-service, payments.

In July 2006, the Centers for Medicare and Medicaid Services (CMS) reported it have approved 276 SNPs (that figure grew to over 400 in 2007, according to the Connecticut-based Center for Medicare Advocacy), most of which were serving dually eligible populations.

MASSACHUSETTS' EXPERIENCE WITH COORDINATED CARE FOR DUALLY ELIGIBLE

Massachusetts' experience coordinating care for the dually eligible dates back many years. It is one of several states operating a Program of All-Inclusive Care for the Elderly (PACE), a federal demonstration program begun in the 1970s, which provides integrated, managed Medicare and Medicaid services in the community to people age 55 and over who need a nursing home-level of care.

Building upon its PACE experience, the state, in conjunction with CMS, began developing several years ago a demonstration program to help the state's entire dually eligible population receive services in a more systematic and coordinated fashion. The program, SCO, which is voluntary, provides seniors with the full range of Medicare and Medicaid benefits. The program serves “community-well,” “community-frail,” and institutionalized people aged 65 and over. The community frail component of the SCO is, in fact, the PACE program.

The SCO program began accepting applications in 2004. It is financed by pooling Medicaid and Medicare revenues at the health plan level. The Medicare payment is based on the Medicare Advantage (MA) methodology, except for people requiring the nursing home level of care who generate a higher capitation rate. The state received a Medicare waiver to ensure that the payments were adjusted based on the diagnosis. No Medicaid waiver was required to generate higher Medicaid payments.

The SCO plans must contract with State Aging Services Access Point providers as part of the care management team. These entities are affiliated with the state's area agencies on aging (AAA), which administer the state's home- and community-based service care programs, including a Medicaid home- and community-based care waiver.

With the passage of the 2003 federal act, SCO was revised to incorporate the SNPs.

BRANDEIS SNP STUDY

In March 2007, researchers from Brandeis University reported on an early study it conducted of 11 demonstration health plans in three states contracting with CMS, viewing them as a prototype of the kinds of integrated care the SNP would offer. (All 11 plans achieved Medicare Advantage SNP approval.) Massachusetts was one of the states, and it was one of only two states to implement the fully integrated Medicare- Medicaid model in which those programs' funds were pooled to provide more integrated health care to the elderly.

Shared Features of 11 Demonstration Projects

The researchers listed several features common to most or all 11 plans they reviewed that they believe provide a good foundation for any dual-eligible SNP. These include:

1. voluntary enrollment by both dual-eligible and Medicaid-only beneficiaries,

2. financing through risk-adjusted capitations from Medicare (for those eligible for it) and Medicaid,

3. including in the Medicaid capitation waiver and personal care assistance (PCA) funds to cover community services,

4. including in the capitation responsibility for nursing home care,

5. full prescription drug coverage through Medicaid (now covered by Medicare but with Medicaid wrap-around), and

6. special efforts at care coordination of medical and social services using capitation to cover care management staff and benefits not covered in standard Medicare or Medicaid.

Massachusetts — SCO

Plans. For its demonstration, Massachusetts contracted with three SCO plans, all of which were small, new managed care entities: Senior Whole Health, a free-standing, for-profit; Commonwealth Care Alliance, a freestanding nonprofit; and Evercare SCO, a subsidiary of the for-profit United Health Care. The state wanted plans that would serve all Medicaid-only and dual-eligible elders, become statewide, and be willing to work with the state's existing aging network.

At the time of the Brandeis researchers' visits, SCO sites had been operating only a year.

Teams. The success of all three states' SNP programs was contingent on the plans' use of teams of professionals who could coordinate and manage enrollees' care. In Massachusetts, the contract requires the plans to have a primary care team consisting of a primary care provider (PCP) working in conjunction with a geriatric social services coordinator (GSSC) and a nurse practitioner, registered nurse, or physician's assistant, all of whom must have experience in geriatrics. The researchers found that the nurses tended to run the teams with the GSSC responsible for the community care. The three SCO plans varied in terms of how the teams worked and where they were situated.

The plans used one of three models for connecting community care with acute care: single coordinator; nurse/social worker team; or the multidisciplinary team, which included a nurse practitioner.These teams coordinate with physicians through face-to-face contact, phone, and information systems.

Marketing. Referrals to the SNPs came from various places, including primary care providers. The Massachusetts plans welcomed referrals but relied more on signing up medical groups serving large numbers of dual eligibles in which physicians were willing to work with the plan's care managers. Few plans reported that state waiver programs were good referral sources.

CMS reported that the state hired a marketing and promotions firm to promote enrollment in the SCO demonstration plans, with the Governor's Office authorizing an aggressive outreach plan. And the state's Medicaid program, MassHealth, sends out SCO postcards and birthday cards to notify potential enrollees of the program.

The most successful outreach appeared to go to community health centers and practices serving ethnic minorities and immigrants. The Massachusetts plans believed that these groups were underserved by the home care system and saw the opportunity to include them.

Enrollment and Assessment. The researchers found that initial and follow-up assessments were time-consuming, creating a potential “bottleneck” for enrollment growth. (CMS has noted that Medicaid churning (people go on and off) can also make enrollment challenging.)

Enrollee Demographics. Most of the SCO-SNP enrollees were either nursing home-certifiable (NHC), meaning they reside in the community but meet the pre-admission screening requirements for nursing home admission, or non-NHC. (Almost no nursing home residents were enrolled.) The average age was 75-76, with the highest percentage of enrollees (51%-77%) non-NHC.

Not surprisingly, the researchers found that SCO-SNP enrollees were frailer than the average Medicare Advantage (MA) enrollee, with higher utilization rates for acute care hospitals and prescription drugs.

Funding. Because most SNP enrollees tend to be frailer than the average MA enrollee, both Medicare and Medicaid paid the plans higher rates for those enrollees considered to be NHC. (This frailty adjuster is being replaced by something called the “CMS-Hierarchical Condition Category(HCC)),” which pays a higher Medicare rate depending on the degree to which enrollees need help with activities of daily living (ADL).

The researchers found that all 11 plans had “disproportionately” high HCC scores and frailty adjusters, when compared to the average Medicare community population. In some cases Medicare spent double at the demonstration sites what it spent for the average Medicare recipient in the area. (The authors note that the these frailty adjustments have not been available to other new SNPs.)

In Massachusetts, Medicaid paid a capitation that included the costs of Medicare co-payments and deductibles, prescription drugs (now wrap-around payments for what Medicare does not cover), ancillary services, community care waiver benefits, personal care attendant benefits, and some risk for nursing homes. The state shared the risk with the SCO-SNP contractors during the program's start-up, but the plans assumed full risk after that. According to the report, the combination of risk-adjusted Medicare and Medicaid payments was covering the plans' high costs, and all reported that they were financially viable.

All three states' Medicaid rates for NHC enrollees incorporated (1) the state's average spending for community waiver services for NHC beneficiaries, (2) estimated costs of supplementary services, and (3) a component to cover the plans' risk for nursing home care. The authors suggest that the last component provides a strong incentive to the plans to keep enrollees in the community.

In Massachusetts, the plans were at risk for the first 90 days of nursing home costs, after which the state paid them at a case-mix adjusted rate to cover nursing home costs directly. Likewise, if an SCO plan moved a nursing home resident into the community the nursing home rate continued for three months. (The authors note that the state's risk-sharing arrangement with sites limited overall profits and losses on Medicaid reimbursement and services within pre-defined risk corridors.)

All three states based rates for community care waiver services on beneficiaries who were NHC and actually participated in waiver programs, rather than also including those NHC beneficiaries who did not participate, which had the effect of inflating reimbursement further.

Care Coordination. Although all three states' contracts with SNPs required the plans to identify and serve NHC enrollees, the plans also had a financial incentive to do so since the higher NHC rates from both programs were contingent on the plans assessing the enrollees. In Massachusetts, even the non-NHC enrollees received assessments, care plans, and care coordination, albeit at lower intensity levels.

The plans were required to contract with outside organizations for the PCA and Medicaid waiver services. In Massachusetts, once a plan did an initial assessment, the contracts in two of the three sites require it to use the Area Agency on Aging's (AAA) geriatric social services coordinator (GSSC) to develop and implement community care plans. The third site was basing its team in community health centers, many of which already had home visiting nurses who assumed that role on the SCO team, with the GSSC functioning as an expert but less involved. Aside from their being the established care coordinators for existing home- and community-based services, the AAAs have other revenues for community services that are available for SCO members.

Services. The services the plans offered were comprehensive. They offered not only traditional Medicare services but also those that Medicaid covered, in particular community-based services under the state's existing home- and community-based services waiver and PCA programs, two models of delivering community care.

The plans managed waiver services either through outside contracts or in-house. Managing PCAs was a bit more challenging for the plans. They had to (1) identify staff who were qualified to conduct eligibility assessments and (2) contract with PCA management agencies and fiscal intermediaries for training and paying the PCAs.

To connect community care to acute care services, Massachusetts' SCO adopted the nurse and social worker team approach, in which the team social worker managed community care and the team nurse coordinated with the medical care. Physicians headed the teams but participated “remotely” through the nurse. (The researchers found that only Wisconsin and Minnesota had team models that “consistently and closely” integrated the two care levels, but both required very low caseloads for coordinating staff.)

Massachusetts used a standardized risk screening in two plans and clinical leadership from the medical director in the third plan to aid in the coordination efforts. In all three states, Evercare (one of the contractor SNPs) coordinated care in nursing facilities to prevent hospital admissions.

Service Utilization. Not surprisingly, utilization and cost of key services, such as hospitals and physicians, were high for the SCO plans. Attachment 1 provides a breakdown of utilization and costs for the 11 demonstrations. (The researchers did not report the names of health plans because they did not believe the data supported cross-plan comparison.)

Medicare Special Needs Plans: Lessons from Dual-Eligible Demonstrations for CMS, States, Health Plans, and Providers, Brandeis University, March 2007

MATHEMATICA SITE VISITS

The Medicare Payment Advisory Commission (an independent federal body established by Congress in 1997 to advise it on issues affecting the Medicare program) was interested in an early picture of how SNPs were developing and contracted with Mathematica Policy Research to do site visits in three marketplaces with different characteristics: Boston, Phoenix, and Miami. In Boston, Mathematica staff interviewed five of the six plans offering SNPs in the Boston area as of February 16, 2006. According to a report summarizing the visits, the Boston SNP marketplace is dominated by SCO, and in fact all but one of the SNPs participates in that program. (The one non-SCO-SNP also operates a PACE program and is a Medicaid contractor.) Staff also interviewed staff from relevant state agencies and the regional CMS office.

The site visits revealed a number of promising arrangements as well as significant differences between the SCO-based SNPs and non-SCO SNPs. The Boston SCO-SNPs had well-developed care coordination and management programs, while the other SNPs were in the early stages of developing these programs. The visits also showed that the SCO SNPs had implemented “substantial” quality monitoring and reporting systems, presumably since the demonstration had offered the same. The other SNPs did not appear to have any monitoring in place other than the “minimum that CMS required.” Finally, the SCO-SNPs included Medicaid benefits in their capitated payments while the non-SCO SNPs did not.

With respect to financing and payment, Mathematica found that all SNPs expected their capitated payments would be adequate to cover their costs with the full phase-in of the risk-adjusted payment capitated system that CMS had promised. However, the SCO-SNPs and state officials noted that requiring SNPs to keep separate track of Medicare and Medicaid funding streams was somewhat burdensome for plans.

Medicare Advantage Special Needs Plans Site Visits, Mathematica Policy Research, June 2006

RC:ro

Attachment 1: Service Utilization in Dual-Eligible Demonstrations — Hospital and Utilization Costs

na: data not available

(1) Plan J figures for NHC and non-NHC enrollees pooled

Source: Medicare Special Needs Plans: Lessons from Dual-Eligible Demonstrations for CMS, States, Health Plans, and Providers, William D. Clark, Brandeis University (2007)

Attachment 2: Service Utilization in Dual-Eligible Demonstrations — Physicians and Prescriptions

na: data not available

(1) Plan J figures for NHC and non-NHC enrollees pooled

Source: Medicare Special Needs Plans: Lessons from Dual-Eligible Demonstrations for CMS, States, Health Plans, and Providers, William D. Clark, Brandeis University (2007)

Attachment 3: Service Utilization in Dual-Eligible Demonstrations — Nursing Facilities

(1) Experience tracked for SNF/ECF – not separated between Medicare SNF and Other nursing facilities

(2) Medicare SNF utilization and costs based on one enrollee.

Source: Medicare Special Needs Plans: Lessons from Dual-Eligible Demonstrations for CMS, States, Health Plans, and Providers, William D. Clark, Brandeis University (2007)

RC:ro