This is the accessible text file for GAO report number GAO-10-403 
entitled 'Medicare Advantage: Relationship between Benefit Package 
Designs and Plans' Average Beneficiary Health Status' which was 
released on June 1, 2010. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as 
part of a longer term project to improve GAO products' accessibility. 
Every attempt has been made to maintain the structural and data 
integrity of the original printed product. Accessibility features, 
such as text descriptions of tables, consecutively numbered footnotes 
placed at the end of the file, and the text of agency comment letters, 
are provided but may not exactly duplicate the presentation or format 
of the printed version. The portable document format (PDF) file is an 
exact electronic replica of the printed version. We welcome your 
feedback. Please E-mail your comments regarding the contents or 
accessibility features of this document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

Report to Congressional Requesters: 

United States Government Accountability Office: 
GAO: 

April 2010: 

Medicare Advantage: 

Relationship between Benefit Package Designs and Plans' Average 
Beneficiary Health Status: 

GAO-10-403: 

GAO Highlights: 

Highlights of GAO-10-403, a report to congressional requesters. 

Why GAO Did This Study: 

Nearly 11 million Medicare beneficiaries are enrolled in Medicare 
Advantage (MA), Medicare’s private health insurance option. Benefits 
vary by MA plan and may include coverage for services not available in 
traditional Medicare. To ensure   do not 
discriminate against beneficiaries in poor health with high expected 
health care costs, the Centers for Medicare & Medicaid Services (CMS) 
reviews and approves all benefit packages yearly. 

GAO examined (1) MA plan benefit packages by average health status of 
plans’ enrolled beneficiaries, (2) distribution and characteristics of 
MA plans by average beneficiary health status, and (3) CMS’s process 
for ensuring that benefit packages do not discriminate with respect to 
health status. Using 2008 data on beneficiaries’ expected health care 
costs, the most recent data available, GAO sorted 2,899 plans 
enrolling 7.5 million beneficiaries into three groups: good health 
(below-average expected costs), average health, and poor health (above-
average expected costs). GAO then analyzed MA plan benefit packages by 
health group and reviewed CMS documentation and interviewed agency 
officials on CMS’s benefit package review process. GAO did not 
determine whether plans structured benefit packages in response to 
enrolled beneficiaries’ health status or beneficiaries in particular 
health groups chose plans because of the benefits. 

What GAO Found: 

In 2008, plans in the good health group generally had lower premiums, 
higher cost sharing for certain services, and fewer additional 
benefits than plans in the poor health group. Almost half of the plans 
in the good health group did not have an MA premium for medical or 
drug coverage, while about one-fifth of plans in the poor health group 
had no MA premium. Plans in the good health group had higher cost 
sharing, weighted by enrollment, for inpatient hospital care, skilled 
nursing facility stays, and renal dialysis than plans in the poor 
health group. Plans in the good health group were more likely to have 
an out-of-pocket (OOP) maximum, but the average OOP maximum for plans 
in that group, weighted by enrollment, was 55 percent higher than that 
for plans in the poor health group. Comprehensive dental and hearing 
aid benefits were more likely to be included in the benefit packages 
for beneficiaries in the poor health group of plans whereas fitness 
benefits were more likely to be included in the benefit packages for 
beneficiaries in the good health group of plans. 

Forty-three percent of plans were in the good health group, 37 percent 
in the average health group, and 20 percent in the poor health group. 
Twenty-nine percent of MA beneficiaries were in plans in the good 
health group, 55 percent in plans in the average health group, and 16 
percent in plans in the poor health group. Among the five largest 
companies sponsoring MA plans, beneficiary health varied: one sponsor 
had 17 percent of its beneficiaries in plans in the good health group 
and 17 percent in plans in the poor health group; another sponsor had 
49 percent of beneficiaries in plans in the good health group and less 
than 1 percent in plans in the poor health group. Average beneficiary 
health status also varied by other factors, such as plan type and plan 
size. 

CMS has revised its process for reviewing MA plans for the likelihood 
of discrimination. It developed a new methodology for setting cost-
sharing thresholds—criteria used to identify benefit packages likely 
to discriminate against certain beneficiaries. For contract year 2010, 
CMS contacted all MA plans with benefit packages identified as likely 
to discriminate, and all plans subsequently met cost-sharing 
thresholds. The new methodology for setting cost-sharing thresholds 
allowed higher cost sharing for some services relative to 2009. For 
example, among plans without an OOP maximum or one above $3,400 for 
2010, allowed cost sharing for a typical inpatient mental health stay 
doubled, from $61 per day to $130 per day, and allowed cost sharing 
for a typical skilled nursing facility stay increased from $53 to $70 
per day, compared to 2009. 

In comments on a draft of this report, CMS noted that GAO’s findings 
are consistent with the agency’s experience. CMS also stated that, 
prior to contract year 2010, it targeted for cost-sharing reductions 
plans with the most egregious cost sharing and often reduced cost-
sharing amounts, but to amounts that were still above the thresholds. 

View [hyperlink, http://www.gao.gov/products/GAO-10-403] or key 
components. For more information, contact James C. Cosgrove at (202) 
512-7114 or cosgrovej@gao.gov. 

[End of section] 

Contents: 

Letter: 

Background: 

The Good Health Group of Plans Generally Charged Lower Premiums and 
Higher Cost Sharing for Certain Benefits, and Offered Fewer Additional 
Benefits: 

More MA Plans and Beneficiaries Were in the Good Health Group than in 
the Poor Health Group of Plans: 

CMS Recently Revised Its Process for Ensuring That MA Plan Benefit 
Packages Do Not Discriminate against Beneficiaries in Poor Health: 

Agency and Other External Comments and Our Evaluation: 

Appendix I: Medicare Advantage Plans and Beneficiaries by Average 
Indexed Risk Score and Health Group, 2008: 

Appendix II: Comments from the Centers for Medicare & Medicaid 
Services: 

Appendix III: GAO Contact and Staff Acknowledgments: 

Table: 

Table 1: Percentage of MA Plans that Exceeded CMS Cost-Sharing 
Thresholds, Contract Years 2008 and 2009: 

Figures: 

Figure 1: Average Combined, Part C, and Part D MA Plan Premiums by 
Health Group, 2008: 

Figure 2: Average MA Plan Cost Sharing for Selected Services by Health 
Group, 2008: 

Figure 3: Percentage of Beneficiaries in MA Plans with Selected 
Additional Benefits by Health Group, 2008: 

Figure 4: MA Plans According to Health Group, 2008: 

Figure 5: MA Beneficiaries According to Plan Health Group, 2008: 

Figure 6: Percentage of Beneficiaries in MA Plans, by Plan Health 
Group, for the Five Largest MA Organizations, 2008: 

Figure 7: Percentage of Beneficiaries in MA Plans, by Plan Health 
Group and Plan Type, 2008: 

Figure 8: Percentage of MA Plans, by Plan Health Group and Plan Size 
(excluding SNPs), 2008: 

Figure 9: Percentage of SNPs, by Plan Health Group and Plan Size, 2008: 

Figure 10: CMS's Process for Identifying Plans with Cost Sharing 
Likely to Be Discriminatory and Contacting Plans for Cost-Sharing 
Reductions, Contract Year 2010: 

Figure 11: CMS's Process for Identifying Plans with Cost Sharing 
Likely to be Discriminatory and Selecting Plans for Cost-Sharing 
Reductions, Contract Years 2008 and 2009: 

Figure 12: Number of Medicare Advantage Plans by Average Indexed Risk 
Score and Health Group, 2008: 

Figure 13: Percentage of Medicare Advantage Beneficiaries by MA Plans' 
Average Indexed Risk Score and Health Group, 2008: 

Abbreviations: 

AHIP: America's Health Insurance Plans: 

CMS: Centers for Medicare & Medicaid Services: 

DME: durable medical equipment: 

FFS: fee-for-service: 

HMO: health maintenance organization: 

MA: Medicare Advantage: 

MAO: Medicare Advantage organization: 

MedPAC: Medicare Payment Advisory Commission: 

PBP: plan benefit package: 

PFFS: private fee-for-service: 

PPO: preferred provider organization: 

OOP: out-of-pocket: 

SNF: skilled nursing facility: 

SNP: special needs plan: 

[End of section] 

United States Government Accountability Office:
Washington, DC 20548: 

April 30, 2010: 

Congressional Requesters: 

Nearly one out of every four Medicare beneficiaries is enrolled in a 
Medicare Advantage (MA) plan--an alternative to original Medicare fee- 
for-service (FFS)--in which private insurance plans offer health care 
coverage to Medicare beneficiaries. As of December 2009, nearly 11 
million Medicare beneficiaries were enrolled in approximately 4,700 
plans offered by 188 MA organizations (MAO).[Footnote 1] Medicare 
payments to MA plans were approximately $109.7 billion in fiscal year 
2009.[Footnote 2] In addition to covering services paid for under 
Medicare FFS, many MA plans offer additional benefits, such as vision, 
hearing, or dental care and MA plans typically have premiums lower 
than those of Medicare supplemental policies purchased by FFS 
beneficiaries (known as Medigap).[Footnote 3] MA beneficiaries 
generally have an array of plans to choose from, each with different 
coverage, premiums, and cost sharing--the portion of medical expenses 
that the beneficiary is responsible for paying out-of-pocket (OOP). 

The Centers for Medicare & Medicaid Services (CMS)--the agency that 
administers the Medicare program--pays MA plans a fixed amount to 
cover each beneficiary and oversees plan benefit designs. CMS adjusts 
payments to plans using risk scores, which estimate the expected 
health care costs of the beneficiaries enrolled in a plan.[Footnote 4] 
Risk scores are developed from data on individuals' demographics and 
diagnoses. As a result, MA plans receive relatively lower payments for 
beneficiaries who are healthier than average, and relatively higher 
payments for beneficiaries who are sicker than average. 

The Medicare Payment Advisory Commission (MedPAC) reported that 
Medicare's risk-adjustment methodology may tend to set payments too 
low for beneficiaries in poor health, who tend to have high expected 
health care costs, and set payments too high for the healthiest 
beneficiaries who tend to have low expected health care costs. 
[Footnote 5] As a result, MAOs may have a financial incentive to 
discourage new or continued enrollment of beneficiaries in poor 
health. It is possible that some MAOs could do this through the design 
of their plan benefit packages. For example, a plan that charges 
relatively high cost-sharing amounts for certain services commonly 
used by beneficiaries in poor health may prove particularly 
undesirable to them. MA plans have flexibility in designing their 
benefit packages, but (1) they must provide all Medicare-covered 
services except hospice care, (2) their overall cost-sharing 
requirements must be actuarially equivalent or lower than those under 
Medicare FFS,[Footnote 6] and (3) they cannot discriminate on the 
basis of health status. By law, CMS may not approve MA plan benefit 
packages if their designs are likely to substantially discourage 
enrollment of certain beneficiaries.[Footnote 7] To determine whether 
an MA plan benefit package is likely to substantially discourage 
enrollment of certain beneficiaries, CMS reviews cost sharing for 
certain services for which excessively high cost sharing could be 
considered discriminatory. 

We reported in 2008 that 2007 cost sharing for certain services, such 
as home health and inpatient hospital stays, was higher in some MA 
plans than in Medicare FFS.[Footnote 8] However, we have not 
previously examined how plan benefit design varied among MA plans by 
average beneficiary health status. Therefore, you asked that we review 
MA plan benefit designs, enrollment patterns, and related CMS 
oversight. This report examines (1) benefit packages of MA plans by 
average health status of plans' enrolled beneficiaries, (2) the 
distribution and characteristics of MA plans by average beneficiary 
health status, and (3) CMS's process for ensuring that MA plan benefit 
packages are not discriminatory with respect to health status. 

To address these issues, we focused on four types of MA plans that 
together accounted for nearly all of MA enrollment in 2009--health 
maintenance organizations (HMO), private fee-for-service (PFFS) plans, 
local preferred provider organizations (PPO), and regional PPOs. 
[Footnote 9] We included special needs plans (SNP), which are allowed 
to limit enrollment to a targeted beneficiary population; SNPs can be 
HMOs, local PPOs, or regional PPOs.[Footnote 10] After certain 
exclusions, we analyzed data for 2,899 plans (including 621 SNPs), 
offered by 192 MAOs that enrolled 74 percent of all MA beneficiaries-- 
7.5 million--as of July 2008.[Footnote 11] Of the 973,923 
beneficiaries enrolled in SNPs, over two-thirds were in plans that 
targeted enrollment to dual-eligible beneficiaries--those entitled to 
both Medicare and Medicaid. In some of our analyses, we reported our 
results for SNPs separately in order to highlight the distinct 
characteristics of such plans. 

Using CMS's 2008 plan-level risk score data by county, the most recent 
data available, we categorized the average health status of MA plans 
as good, average, or poor in the areas that they served. Risk scores 
are based on beneficiaries' projected health care costs, which CMS 
develops using demographic information and diagnosis codes.[Footnote 
12] At the plan level, risk scores indicate the average health status 
of the plans' beneficiaries. Because the overall Medicare population 
may be healthier in some geographic areas than in others, we 
calculated an indexed (relative) health risk score for each plan as 
the ratio of an MA plan's risk score to the overall Medicare average 
risk score in each county. We then averaged each plan's indexed risk 
score across the counties that comprised the plan's service area, 
weighted by July 2008 enrollment. With 1.00 as the average risk score 
for all Medicare beneficiaries--in FFS and MA--we placed the MA plans 
in one of three groups. 

* Good health group: MA plans with an average indexed risk score less 
than 0.90, meaning the projected health care costs for the average 
beneficiary in the plan were at least 10 percent lower than those for 
an average Medicare beneficiary living in the plan's service area. 

* Average health group: MA plans with an average indexed risk score 
between 0.90 and 1.10, meaning the projected health care costs for the 
average beneficiary in the plan were within 10 percent of those for an 
average Medicare beneficiary living in the plan's service area. 

* Poor health group: MA plans with an average indexed risk score 
greater than 1.10, meaning the projected health care costs for the 
average beneficiary in the plan were at least 10 percent higher than 
those for an average Medicare beneficiary living in the plan's service 
area. 

To compare the benefit packages of MA plans by the average health 
status of plans' enrolled beneficiaries, we analyzed CMS plan benefit 
package (PBP) data for contract year 2008 (as the risk score data used 
in our analysis is based on MA plan enrollment as of July 2008). CMS's 
PBP data contain each MA plan's premiums, cost-sharing requirements, 
and additional benefits. We analyzed cost sharing for eight services 
for which CMS considers high cost sharing to be potentially 
discriminatory because these services typically are used by Medicare 
beneficiaries in poor health and are usually associated with acute and 
chronic conditions, high utilization, and high cost.[Footnote 13] For 
several services, we simulated beneficiary OOP costs using average 
beneficiary utilization profiles developed by CMS and accounted for 
plans that charged service-specific deductibles or had service-
specific OOP maximums--dollar limits on the amount a beneficiary must 
pay in cost sharing in a period of coverage (typically 1 year). 
[Footnote 14] We also determined whether MA plans provided additional 
coverage for dental, vision, or hearing services; and fitness 
benefits.[Footnote 15] We did not determine whether MAOs structured 
their plan benefit packages in response to enrolled beneficiaries' 
health status or whether beneficiaries of a given health status chose 
certain MA plans specifically because of their benefit package designs. 

To examine the distribution and characteristics of MA plans by 
beneficiary health status, we analyzed 2008 MA plan-level risk score 
data, the most recent data available, and calculated the number of MA 
plans and the percentage of plan beneficiaries in each health group. 
We further examined the differences across the health groups (1) among 
the five largest MAOs, (2) by plan type, (3) by plan size, (4) and by 
the percentage difference between the MA payment benchmarks and 
estimated FFS spending across a plan's service area. 

To describe CMS's process for ensuring that MA plan benefit packages 
are not discriminatory with respect to health status, we interviewed 
CMS officials, including staff responsible for reviewing and approving 
plan benefit packages; and reviewed relevant laws and regulations, CMS 
standard operating procedures, and other agency documentation on the 
review process. We determined the outcome of CMS's review process by 
analyzing data for contract years 2008 and 2009 on MA plans that CMS 
initially identified as having potentially discriminatory cost sharing 
and data on MA plans' final cost-sharing requirements for contract 
years 2008 through 2010. 

We interviewed CMS officials about reliability of the CMS data used in 
our analysis. We also reviewed data documentation and performed 
certain data checks to ensure the data were reasonable and consistent. 
For example, we compared the results of our cost sharing analysis 
using PBP data with information from the Medicare Options Compare Web 
site.[Footnote 16] We determined that the data were sufficiently 
reliable for our purposes. We conducted our work from April 2009 
through April 2010 in accordance with generally accepted government 
auditing standards. Those standards require that we plan and perform 
the audit to obtain sufficient, appropriate evidence to provide a 
reasonable basis for our findings based on our audit objectives. We 
believe that the evidence obtained provides a reasonable basis for our 
findings. 

Background: 

Most Medicare beneficiaries can choose to receive covered services 
through Medicare FFS or through an MA plan--which operates under 
Medicare Part C--if one is offered where they live.[Footnote 17] MAOs 
are allowed flexibility in designing their plan benefit packages and 
cost sharing for certain services can vary widely by MA plan. 

Medicare Advantage Coverage and Payment: 

MA plans operate under annual contracts between MAOs and CMS and must 
offer benefits that are covered under Medicare FFS.[Footnote 18] These 
benefits consist of Part A hospital insurance, which covers inpatient 
stays, care in skilled nursing facilities, and some home health care; 
and Part B medical insurance, which covers certain physician, 
outpatient hospital, and laboratory services, among other services. 
All beneficiaries enrolled in Part B are charged a Part B premium. In 
general, in order to enroll in a MA plan, beneficiaries must be 
entitled to benefits under Part A and enrolled in Part B. 

Regardless of their source of coverage, all Medicare beneficiaries 
have the option of receiving prescription drug coverage through 
Medicare Part D. Medicare FFS beneficiaries can enroll in stand-alone 
prescription drug plans, which are operated by private plan sponsors, 
and they generally must pay an additional premium to receive Part D 
coverage. MA beneficiaries who also want prescription drug coverage 
generally receive that coverage through their MA plans, which may or 
may not charge an additional premium for Part D coverage. 

In addition to monthly premiums, beneficiaries in Medicare FFS or in 
MA plans typically are responsible for cost sharing, which can be in 
the form of a deductible, coinsurance, or a copayment.[Footnote 19] To 
help provide financial protection to beneficiaries who might otherwise 
have high cost-sharing expenses for Part A and Part B services, MAOs 
may voluntarily establish OOP maximums, or dollar limits on the amount 
a beneficiary must pay in cost sharing in a period of coverage 
(typically 1 year).[Footnote 20] For contract year 2010, CMS sought to 
allow MA plans with an OOP maximum at or below $3,400 greater 
flexibility in establishing cost-sharing amounts.[Footnote 21] 

For each MA beneficiary, CMS pays MA plans a monthly amount determined 
by the plan bid--the plan's estimated cost of providing Medicare Part 
A and Part B benefits--in relation to a benchmark, which is the 
maximum amount the Medicare program will pay MA plans in a given 
locality. If a plan's bid is less than the benchmark, the difference 
is partially rebated to the MA plan and must be used to reduce 
premiums, reduce cost sharing, or provide additional benefits for plan 
beneficiaries. If a plan's bid exceeds the benchmark, the plan will 
charge each of its beneficiaries an additional premium to make up the 
difference. MA plans offering prescription drug coverage have a 
separate payment benchmark for Part D prescription drug benefits. CMS 
risk-adjusts the monthly payments to MA plans to take into account the 
health status of the plan's beneficiaries. 

 : 

As previously reported, cost sharing can vary widely among MA plans 
for particular categories of services as a result of the flexibility 
given MAOs in designing their plan benefit packages.[Footnote 22] For 
example, in 2007, 9 percent of beneficiaries were enrolled in MA plans 
that had no cost sharing for inpatient services, whereas 16 percent of 
beneficiaries were enrolled in MA plans with cost sharing for 
inpatient services that was higher than that of Medicare FFS. 
Similarly, an AARP study reported that in 2008 the average MA 
beneficiary with a 10-day inpatient hospital stay would incur $823 in 
cost sharing, less than the $1,068 incurred for beneficiaries in 
Medicare FFS, but 12 percent of beneficiaries would incur cost sharing 
of $2,000 or more.[Footnote 23] 

The Secretary of Health and Human Services is obliged by statute to 
not contract with a MAO if its plan benefit design is likely to 
substantially discourage enrollment in the MA plan by certain 
individuals.[Footnote 24] To implement this provision of the statute, 
CMS identified in the Medicare Managed Care Manual certain services 
for which high cost sharing could be considered potentially 
discriminatory and provides further guidance in its annual MA Call 
Letter to MA plans on how it will review benefit packages for the 
likelihood of discrimination. CMS has never barred a plan from 
participation in the MA program because of cost sharing that was 
likely to substantially discourage enrollment. 

Certain aspects of MA cost-sharing requirements and payment will 
change as a result of the Patient Protection and Affordable Care Act. 
Beginning with bids submitted for contract year 2011, the Secretary of 
Health and Human Services has the authority to not contract with an 
MAO if it proposes significant increases in cost sharing or decreases 
benefits offered by a plan.[Footnote 25] In addition, for plan years 
beginning January 1, 2011, the cost sharing required for chemotherapy 
services, renal dialysis, skilled nursing care, and any other service 
that the Secretary determines appropriate can be no more than the cost 
sharing required in Medicare FFS.[Footnote 26] 

The Good Health Group of Plans Generally Charged Lower Premiums and 
Higher Cost Sharing for Certain Benefits, and Offered Fewer Additional 
Benefits: 

The good health group of plans--MA plans in which the average 
beneficiary had projected health care costs at least 10 percent below 
those for an average Medicare beneficiary within the plan's service 
area--generally charged lower premiums and had higher cost sharing for 
certain services compared with the poor health group of plans. Plans 
in the good health group also were less likely to include additional 
benefits, such as vision and dental care coverage. 

The Good Health Group of Plans Generally Charged Lower Premiums or 
None at All: 

The good health group of plans generally charged lower premiums 
relative to the poor health group of plans. (See figure 1.) For 
example, in 2008, 

* almost half of the good health group of plans--46 percent--did not 
have a premium for Part C (medical) or Part D (prescription drug) 
coverage, while about one-fifth of the poor health group of plans had 
no premium. 

* for MA plans that included prescription drug coverage as part of 
their benefit package, the combined (Part C and Part D) monthly 
premium for the good health group of plans--$24--was lower than that 
for the poor health group, which was $31. 

* taken separately, Part C and Part D premiums for the good health 
group of plans also were lower than the corresponding premiums for the 
poor health group. 

Figure 1: Average Combined, Part C, and Part D MA Plan Premiums by 
Health Group, 2008: 

[Refer to PDF for image: illustrated table] 

Premium: Part C and Part D[A]; 
Good health group: $24; 
Average health group: $37; 
Poor health group: $31. 

Premium: Part C; 
Good health group: $12; 
Average health group: $24; 
Poor health group: $19. 

Premium: Part D; 
Good health group: $10; 
Average health group: $11; 
Poor health group: $14. 

Source: GAO analysis of 2008 CMS Plan Benefit Package data. 

Notes: Dollar amounts are weighted by July 2008 enrollment and include 
zero-premium MA plans. Using CMS's 2008 plan-level risk score data by 
county, we categorized the health status of MA plans as good, average, 
or poor based on the average health status of their enrolled 
beneficiaries in the areas that they served. This analysis included 
2,899 MA plans that enrolled a total of 7,553,600 beneficiaries. 

[A] The combined Part C and D premium reflects the premium for plans 
that included the Part D prescription drug coverage as part of their 
MA plan benefit package. 

[End of figure] 

In 2008, 12 percent of plans in the good health group reduced 
beneficiaries' Part B premium, while 4 percent of plans in the poor 
health group did so. Among plans that reduced the Part B premium, 
plans in the good health group reduced it by a larger amount than 
plans in the poor health group--averaging $48 and $36, respectively. 
[Footnote 27] 

The Good Health Group of Plans Generally Had Higher Cost Sharing for 
Certain Services: 

The good health group of plans tended to have higher cost sharing for 
the services we reviewed--which included inpatient hospital acute 
stays, inpatient mental health stays, skilled nursing facility (SNF) 
stays, and renal dialysis--than the poor health group of plans. 
[Footnote 28] (See figure 2.) For example, in 2008, 

* the average plan in the good health group charged about $100 more in 
cost sharing for a typical inpatient hospital stay (6 days)[Footnote 
29] and about $150 more for a typical inpatient mental health stay (21 
days)[Footnote 30] than the average plan in the poor health group. 

* the average plan in the good health group charged about $500 more in 
cost sharing for a typical SNF stay (35 days) than the average plan in 
the poor health group.[Footnote 31] 

* the average plan in the good health group charged over $300 more in 
cost sharing for a year of renal dialysis (156 sessions)[Footnote 32] 
than the average plan in the poor health group. 

Figure 2: Average MA Plan Cost Sharing for Selected Services by Health 
Group, 2008: 

[Refer to PDF for image: illustrated table] 

Inpatient hospital stay (6 days): 
Good health group: $701; 
Average health group: $604; 
Poor health group: $601. 

Skilled nursing facility stay (35 days): 
Good health group: $1,545; 
Average health group: $1,485; 
Poor health group: $1,026. 

Inpatient mental health stay[A]: 
Good health group: $895; 
Average health group: $881; 
Poor health group: $746. 

Renal dialysis (156 sessions): 
Good health group: $2,118; 
Average health group: $1,798; 
Poor health group: $1,802. 

Source: GAO analysis of 2008 CMS Plan Benefit Package data. 

Notes: Dollar amounts are weighted by July 2008 enrollment and include 
MA plans that did not charge cost sharing for the indicated services. 
Using CMS's 2008 plan-level risk score data by county, we categorized 
the health status of MA plans as good, average, or poor based on the 
average health status of their enrolled beneficiaries in the areas 
that they served. This analysis included 2,899 MA plans that enrolled 
7,553,600 beneficiaries. 

[A] CMS does not calculate an average cost per day for an inpatient 
mental health stay. As a result, we were only able to estimate the 
average OOP cost for a 21-day inpatient mental health stay for plans 
that charged a copayment. Of the 2,899 plans in our analysis, 
approximately 90 percent indicated that they charged a copayment for 
an inpatient mental health stay. 

[End of figure] 

The good health group of plans had similar cost sharing for Part B 
chemotherapy drugs, other Part B drugs, DME, and home health compared 
with the poor health group of plans.[Footnote 33] For example, in 2008, 

* an average plan in both the good health group and the poor health 
group charged 20 percent coinsurance for Part B chemotherapy drugs and 
charged 20 and 19 percent coinsurance, respectively, for other Part B 
drugs. 

* an average plan in the good health group charged 23 percent 
coinsurance for DME, compared with 21 percent for the average plan in 
the poor health group. 

* for the relatively small share of plans that charged cost sharing 
for home health care, an average plan in the good health group charged 
$568 for 27 home health visits compared to $597 for the average plan 
in the poor health group. 

In 2008, a greater share of plans in the good health group had an OOP 
maximum that limited their beneficiaries' overall financial risk 
compared to plans in the poor health group--63 percent compared with 
40 percent, respectively.[Footnote 34] However, the OOP maximum for 
the good health group of plans averaged 55 percent higher than the OOP 
maximum for the poor health group ($3,515 compared with $2,262). 
[Footnote 35] In designing their benefit packages, some plans 
excluded cost sharing incurred for certain services from expenses that 
counted toward the OOP maximum.[Footnote 36] Approximately 14 percent 
of plans in the good health group had an OOP maximum and excluded one 
or more services for which they required cost sharing, compared with 
13 percent of plans in the average health and poor health groups. 
[Footnote 37] Plans were most likely to exclude Part B drugs (15 
percent of plans), renal dialysis (6 percent of plans), and DME (6 
percent of plans) from their OOP maximum.[Footnote 38] 

The Good Health Group of Plans Generally Had Fewer Additional Benefits: 

Plans in the good health group were less likely to include certain 
additional benefits in their benefit packages, such as vision and 
dental care coverage, but were more likely to include a fitness 
benefit than plans in the poor health group.[Footnote 39] (See figure 
3.) For example, in 2008, 

* seventy-three percent of plans in the good health group included 
coverage for eye exams and 60 percent included coverage for eyewear 
compared with 78 percent and 74 percent, respectively, of plans that 
covered these benefits in the poor health group. 

* twenty-three percent of plans in the good health group included 
coverage for comprehensive dental benefits compared with 33 percent of 
plans in the poor health group.[Footnote 40] 

* while plans in the good health group were more likely to cover 
hearing tests, they were less likely than plans in the poor health 
group to cover hearing aids--36 percent versus 48 percent. 

* the percentage of plans in the good health group that included 
fitness benefits was approximately twice that of plans in the poor 
health group--55 percent compared with 28 percent. 

Figure 3: Percentage of Beneficiaries in MA Plans with Selected 
Additional Benefits by Health Group, 2008: 

[Refer to PDF for image: illustrated table] 

Vision eye exams: 
Good health group: 73%; 
Average health group: 76%; 
Poor health group: 78%. 

Vision eye wear: 
Good health group: 60%; 
Average health group: 64%; 
Poor health group: 74%. 

Dental prevention: 
Good health group: 46%; 
Average health group: 48%; 
Poor health group: 51%. 

Dental comprehensive: 
Good health group: 23%; 
Average health group: 28%; 
Poor health group: 33%. 

Hearing exams: 
Good health group: 78%; 
Average health group: 76%; 
Poor health group: 67%. 

Hearing aids: 
Good health group: 36%; 
Average health group: 40%; 
Poor health group: 48%. 

Fitness benefits: 
Good health group: 55%; 
Average health group: 49%; 
Poor health group: 28%. 

Source: GAO analysis of 2008 CMS Plan Benefit Package data. 

Notes: Percentages are weighed by July 2008 enrollment. Results 
include MA plans that provided dental, vision, hearing, and fitness 
benefits as mandatory supplemental benefits. Using CMS's 2008 plan-
level risk score data by county, we categorized the health status of 
MA plans as good, average, or poor based on the average health status 
of their enrolled beneficiaries in the areas that they served. This 
analysis included 2,899 MA plans that enrolled 7,553,600 beneficiaries. 

[End of figure] 

More MA Plans and Beneficiaries Were in the Good Health Group than in 
the Poor Health Group of Plans: 

In 2008, the plans in the good health group differed with those in the 
poor health group by plan type, plan size, and, for HMOs, by the 
difference between MA payment benchmarks and estimated FFS spending. 
Our analysis of MA plans' average indexed risk scores, including SNPs, 
by health group showed that 1,254 plans (43 percent) were in the good 
health group, 1,068 plans (37 percent) were in the average health 
group, and 577 plans (20 percent) were in the poor health 
group.[Footnote 41] (See figure 4.) Across all 2,899 MA plans in our 
review, the average indexed risk scores--adjusted for geographic 
variations in the average health of all Medicare beneficiaries--ranged 
from 0.38 to 2.65. 

Figure 4: MA Plans According to Health Group, 2008: 

[Refer to PDF for image: pie-chart] 

Good health (non-SNP): 1,218; 
Good health (SNP): 36; 
Good health total: 43%; 
Average health (non-SNP): 891; 
Average health (SNP): 177; 
Average health total: 37%;
Poor health (non-SNP): 169;
Poor health (SNP): 408; 
Poor health total: 20%. 

Source: GAO analysis of 2008 CMS risk score data. 

Notes: Using CMS's 2008 plan-level risk score data by county, we 
categorized the health status of MA plans as good, average, or poor 
based on the average health status of their enrolled beneficiaries in 
the areas that they served. This analysis included 2,899 MA plans, of 
which 621 were SNPs that are allowed to limit enrollment to a targeted 
beneficiary population. 

[End of figure] 

Our analysis of the percentage of beneficiaries, including those 
enrolled in SNPs, by health group showed that 29 percent were in the 
good health group of plans, 55 percent in the average health group of 
plans, and 16 percent in the poor health group of plans. (See figure 
5.) 

Figure 5: MA Beneficiaries According to Plan Health Group, 2008: 

[Refer to PDF for image: pie-chart] 

Good health (non-SNP): 29%; 
Good health (SNP): less than 1%; 
Good health total: 29%; 
Average health (non-SNP): 51%; 
Average health (SNP): 4%; 
Average health total: 55%;
Poor health (non-SNP): 7%;
Poor health (SNP): 9%; 
Poor health total: 16%. 

Source: GAO analysis of 2008 CMS risk score data. 

Notes: Using CMS's 2008 plan-level risk score data by county, we 
categorized the health status of MA plans as good, average, or poor 
based on the average health status of their enrolled beneficiaries in 
the areas that they served. This analysis included 7,553,600 MA 
beneficiaries, of which 973,923 were enrolled in SNPs as of July 2008. 
SNPs are allowed to limit enrollment to a targeted beneficiary 
population. 

[End of figure] 

Average indexed risk scores ranged from 0.90 to 1.03 for the five 
largest organizations in the MA program, which together accounted for 
nearly 50 percent of MA enrollment in 2008. The five MAOs varied in 
the extent to which their plans fell into the good, average, and poor 
beneficiary health status groups. (See figure 6.) For example, one MAO 
had 17 percent of its beneficiaries in plans in the good health group 
and 17 percent in plans in the poor health group, while another had 49 
percent of its beneficiaries in plans in the good health group and 
less than 1 percent in plans in the poor health group. 

Figure 6: Percentage of Beneficiaries in MA Plans, by Plan Health 
Group, for the Five Largest MA Organizations, 2008: 

[Refer to PDF for image: stacked vertical bar graph] 

Percentage of Beneficiaries: 

Organization A: 
Good health: 17.4%; 
Average health: 66.1%; 
Poor health: 16.5%. 

Organization B: 
Good health: 21.2%; 
Average health: 64.2%; 
Poor health: 14.6%. 

Organization C: 
Good health: 24.0%; 
Average health: 55.6%; 
Poor health: 20.4%. 

Organization D: 
Good health: 38.0%; 
Average health: 49.0%; 
Poor health: 13.0%. 

Organization E: 
Good health: 48.5%; 
Average health: 51.4%; 
Poor health: 0.1%. 

Source: GAO analysis of 2008 CMS risk score data. 

Notes: Using CMS's 2008 plan-level risk score data by county, we 
categorized the health status of MA plans as good, average, or poor 
based on the average health status of their enrolled beneficiaries in 
the areas that they served. These five MA organizations accounted for 
48 percent (3,656,396 beneficiaries) of MA enrollment, including 
enrollment in SNPs, as of July 2008. SNPs are allowed to limit 
enrollment to a targeted beneficiary population. 

[End of figure] 

The four types of plans we reviewed differed as to the average health 
status of their beneficiaries. Excluding SNPs, regional PPOs and local 
PPOs had the largest percentage of beneficiaries in plans in the good 
health group--95 percent and 66 percent, respectively. Regional PPOs 
did not have any beneficiaries in the poor health group of plans and 
local PPOs had 1 percent of their beneficiaries in the poor health 
group of plans. HMOs and PFFS plans had the largest share of 
beneficiaries in the average health group of plans--68 percent and 48 
percent of beneficiaries, respectively--and each had less than 10 
percent of their beneficiaries in the poor health group of plans. 
[Footnote 42] (See figure 7.) SNPs, by definition, had the largest 
percentage of beneficiaries (71 percent) in the poor health group of 
plans and had nearly all of their remaining beneficiaries in the 
average health group of plans. 

Figure 7: Percentage of Beneficiaries in MA Plans, by Plan Health 
Group and Plan Type, 2008: 

[Refer to PDF for image: stacked vertical bar graph] 

HMO: 
Beneficiaries: 4,254,320; 
Good health: 22.4%; 
Average health: 68.1%; 
Poor health: 9.4%. 

PFFS: 
Beneficiaries: 1,651,221; 
Good health: 45.7%; 
Average health: 47.8%; 
Poor health: 6.5%. 

Local PPO: 
Beneficiaries: 469,891; 
Good health: 65.7%; 
Average health: 33.0%; 
Poor health: 1.3%. 

Regional PPO: 
Beneficiaries: 204,245; 
Good health: 94.8%; 
Average health: 5.2%; 
Poor health: 0%. 

SNP: 
Beneficiaries: 973,923
Good health: 1.3%; 
Average health: 27.6%; 
Poor health: 71.1%. 

Source: GAO analysis of 2008 CMS risk score data. 

Notes: Results are based on MA enrollment as of July 2008. Using CMS's 
2008 plan-level risk score data by county, we categorized the health 
status of MA plans as good, average, or poor based on the average 
health status of their enrolled beneficiaries in the areas that they 
served. SNPs are allowed to limit enrollment to a targeted beneficiary 
population. 

[End of figure] 

Excluding SNPs, MA plans with lower enrollment were more likely to be 
in the good health group than plans with higher enrollment. (See 
figure 8.) About two-thirds of MA plans with 10 to 75 beneficiaries 
were in the good health group of plans, while roughly one-third of MA 
plans with the largest enrollment (2,863 to 92,950 beneficiaries) were 
in the good health group. 

Figure 8: Percentage of MA Plans, by Plan Health Group and Plan Size 
(excluding SNPs), 2008: 

[Refer to PDF for image: vertical bar graph] 

Quintile 1 (10-75 beneficiaries): 
Good health: 64.5%; 
Average health: 26.8%; 
Poor health: 8.8%. 

Quintile 2 (76-277 beneficiaries): 
Good health: 60.0%; 
Average health: 31.9%; 
Poor health: 8.1%. 

Quintile 3 (278-879 beneficiaries): 
Good health: 55.6%; 
Average health: 39.3%; 
Poor health: 5.1%. 

Quintile 4 (880-2,862 beneficiaries): 
Good health: 50.4%; 
Average health: 43.2%; 
Poor health: 6.4%. 

Quintile 5 (2,863-92,950 beneficiaries): 
Good health: 36.7%; 
Average health: 54.5%; 
Poor health: 8.8%. 

Source: GAO analysis of 2008 CMS risk score data. 

Notes: Results are based on MA enrollment as of July 2008. Using CMS's 
2008 plan-level risk score data by county, we categorized the health 
status of MA plans as good, average, or poor based on the average 
health status of their enrolled beneficiaries in the areas that they 
served. SNPs are allowed to limit enrollment to a targeted beneficiary 
population. 

[End of figure] 

While SNPs of any plan size were most likely to have beneficiaries in 
the poor health group, 58 percent of the smallest SNPs were in the 
poor health group and 73 percent of the largest SNPs were in the poor 
health group. (See figure 9) 

Figure 9: Percentage of SNPs, by Plan Health Group and Plan Size, 2008: 

[Refer to PDF for image: vertical bar graph] 

Quintile 1 (10-75 beneficiaries): 
Good health: 13.5%; 
Average health: 28.6%; 
Poor health: 57.9%. 

Quintile 2 (76-277 beneficiaries): 
Good health: 5.7%; 
Average health: 35.0%; 
Poor health: 59.3%. 

Quintile 3 (278-879 beneficiaries): 
Good health: 4.0%; 
Average health: 29.0%; 
Poor health: 66.9%. 

Quintile 4 (880-2,862 beneficiaries): 
Good health: 4.0%; 
Average health: 24.2%; 
Poor health: 71.8%. 

Quintile 5 (2,863-92,950 beneficiaries): 
Good health: 1.6%; 
Average health: 25.8%; 
Poor health: 72.6%. 

Source: GAO analysis of 2008 CMS risk score data. 

Notes: Results are based on MA enrollment as of July 2008. Using CMS's 
2008 plan-level risk score data by county, we categorized the health 
status of MA plans as good, average, or poor based on the average 
health status of their enrolled beneficiaries in the areas that they 
served. SNPs are allowed to limit enrollment to a targeted beneficiary 
population. 

[End of figure] 

HMOs, including SNPs, in the good health group of plans tended to be 
located in areas where the percentage difference between the 2008 MA 
payment benchmarks and estimated 2008 FFS spending was smaller. 
[Footnote 43] MA payment benchmarks averaged 13 percent higher than 
estimated FFS spending in areas where HMOs in the good health group of 
plans were located, but benchmarks were 16 percent higher where HMOs 
in the poor health group were located.[Footnote 44] 

CMS Recently Revised Its Process for Ensuring That MA Plan Benefit 
Packages Do Not Discriminate against Beneficiaries in Poor Health: 

For contract year 2010, CMS modified the process it used to ensure MA 
plan benefit packages do not discriminate against beneficiaries in 
poor health. The agency revised the way it determines cost-sharing 
thresholds used to identify benefit packages that are likely 
discriminatory, and as a result, some of the thresholds used in 
contract year 2010 reviews allowed higher cost-sharing amounts than in 
previous years. CMS also revised its process for contacting plans for 
benefit package modifications by contacting all, instead of a 
selection of, MA plans found to have cost sharing that exceeded CMS 
thresholds. For contract year 2010, all plans met cost-sharing 
thresholds after discussions with CMS, but if contract year 2009 
thresholds were applied to the approved contract year 2010 plans, 
approximately 38 percent of the plans we examined would have exceeded 
cost-sharing thresholds. 

CMS Revised the Way It Identifies Benefit Packages Likely to be 
Discriminatory, and Some New Review Thresholds Allowed Higher Cost 
Sharing: 

For contract year 2010, CMS modified the benefit package review 
process used in previous years. CMS developed a new methodology to 
determine cost-sharing thresholds and included actuarial equivalence 
tests as part of the review process. Under the new process for 
reviewing MA plans for the likelihood of discrimination, CMS examined 
plans' 2010 OOP maximums and identified MA plans with comparatively 
high cost-sharing amounts. Plans' OOP maximums determined the level of 
scrutiny the benefit packages received, with greater scrutiny given to 
plans with no OOP maximum or with an OOP maximum above CMS's OOP 
maximum threshold.[Footnote 45] For selected services--those typically 
used by sicker beneficiaries--CMS compared plans' cost-sharing amounts 
with threshold amounts set relative to cost sharing for all MA plans. 
Benefit packages were considered likely to be discriminatory if the 
cost sharing for one or more of the selected services exceeded the 
threshold and was higher than cost sharing for Medicare FFS.[Footnote 
46] For contract year 2010, the process was as follows. 

* If a plan had an OOP maximum at or below the amount specified in the 
annual Call Letter ($3,400 for 2010),[Footnote 47] CMS limited its 
benefit package review to cost sharing for five selected services: 
renal dialysis (156 sessions), Part B drugs, home health (37 visits), 
inpatient mental health stays (15 days), and SNF stays (42 days). MA 
plans were considered likely to discriminate if their cost sharing for 
any of these services was at or above the 95th percentile relative to 
the other MA plans and was greater than the cost sharing under 
Medicare FFS.[Footnote 48] 

* If a plan did not have an OOP maximum or if the OOP maximum exceeded 
the amount specified in the Call Letter, CMS reviewed cost sharing for 
14 selected services: the 5 listed above plus inpatient hospital stays 
(10 days), inpatient hospital catastrophic stays (90 days), physician 
mental health visits, Part B chemotherapy drugs, Part B radiology, and 
4 DME services--equipment, prosthetics, supplies, and diabetes tests. 
MA plans were considered likely to discriminate if cost sharing for 
any of the 14 services was at or above the 75th percentile relative to 
other MA plans and was greater than the cost sharing under Medicare 
FFS.[Footnote 49] 

As part of the revised review process, in addition to determining 
whether MA plans' benefit packages were actuarially equivalent to 
Medicare FFS, CMS began reviewing MA plans' pricing data for the 
likelihood of discriminatory cost sharing. CMS reviewed MA plans' cost 
sharing for five selected services. For inpatient hospital care 
(including mental health), SNF stays, home health visits, DME, and 
Part B drugs, CMS calculated the difference between the plans' cost 
sharing and an amount that was actuarially equivalent to cost sharing 
under Medicare FFS. If the difference was higher than a tolerance 
amount--the greater of 50 cents or 5 percent of FFS cost sharing--the 
benefit package was identified as likely to be discriminatory and 
plans were instructed to modify the cost sharing in the bid pricing 
tool submitted to CMS. 

The 2010 contract year process differs from the process for previous 
years in key ways. For contract years 2008 and 2009, a plan with cost 
sharing that exceeded CMS thresholds for one or more selected services 
but with an OOP maximum at or below the amount specified in CMS's Call 
Letter was not considered likely to be discriminatory because the OOP 
maximum would limit beneficiaries' OOP costs. If a plan had cost 
sharing for one or more selected services that exceeded the CMS 
thresholds and did not have an OOP maximum or if the OOP maximum was 
above the amount specified in the Call Letter, CMS considered the plan 
likely to be discriminatory.[Footnote 50] In contract years 2008 and 
2009, CMS generally set thresholds based on cost sharing under 
Medicare FFS, not at amounts that were relative to all MA plans' cost-
sharing amounts. For example, for contract year 2008 CMS set the MA 
cost-sharing threshold for Part B drugs at 20 percent coinsurance, 
equivalent to that under Medicare FFS. 

The methodology CMS used in contract year 2010 resulted in thresholds 
that allowed higher cost-sharing amounts for some services than those 
applicable in contract year 2009. For example, among plans without an 
OOP maximum or one above the amount specified in the 2010 Call Letter, 
the copayment allowed for a typical inpatient mental health stay 
doubled from $61 to $130 per day and the copayment allowed for a 
typical SNF stay increased from $53 to $70 per day. For other services 
that CMS reviewed for discriminatory cost sharing, new thresholds 
allowed cost-sharing amounts in contract year 2010 that were 
comparable or lower than those in contract year 2009.[Footnote 51] For 
example, the copayment allowed for a typical inpatient hospital stay 
was reduced from $213 to $175 per day and the cost sharing allowed for 
home health visits was reduced to zero. 

CMS Revised How It Selects Plans to Contact for Benefit Package 
Modifications; All Plans Met New Cost-Sharing Thresholds: 

Under the revised process, for contract year 2010 CMS contacted all MA 
plans identified as having benefit packages likely to be 
discriminatory--those that exceeded agency thresholds. In addition, 
all plans contacted subsequently reduced cost-sharing amounts to at or 
below agency thresholds. In contrast, in previous years, CMS did not 
contact all plans exceeding the thresholds and not all plans contacted 
reduced their cost-sharing amounts to equal to or below agency 
thresholds. In previous years, CMS's policy did not require 
disapproval of a plan benefit package if cost sharing was above CMS's 
cost-sharing thresholds. 

CMS's new selection process is intended to ensure that all MA plans 
identified as likely to be discriminatory are contacted regarding 
lowering their cost sharing to meet the new thresholds. According to 
agency officials, of the 2,930 MA plan benefit packages submitted for 
contract year 2010, about 40 percent were identified as likely to be 
discriminatory. CMS staff contacted all of those plans about complying 
with cost-sharing thresholds, and all plans subsequently reduced cost- 
sharing amounts to at or below the new thresholds.[Footnote 52] 
However, if CMS had applied the contract year 2009 review methodology--
with its lower thresholds for certain services--to the plans submitted 
for contract year 2010, approximately 38 percent of the plans in our 
analysis would have exceeded cost-sharing thresholds.[Footnote 53] 

In previous years, CMS approved some MA plans that did not meet its 
cost-sharing thresholds. For contract years 2008 and 2009, over one in 
four and nearly one in three plans, respectively, had cost sharing 
that exceeded one or more CMS thresholds. (See table 1.) Our analysis 
indicated that the percentage of plans initially exceeding and 
remaining above one or more cost-sharing thresholds in those 2 years 
varied by plans' average beneficiary health status, as follows: 

* Among the good health group of plans reviewed for contract year 
2008, 39 percent were initially identified as having cost sharing that 
exceeded CMS's cost-sharing thresholds and 33 percent were approved 
with cost sharing that exceeded CMS's cost-sharing thresholds, a 
decrease of 6 percentage points. The percentage of plans in the poor 
health group that exceeded CMS's thresholds prior to and after plans 
were approved also decreased by 6 percentage points, from 20 percent 
to 14 percent. 

* Among the good health group of plans reviewed for contract year 
2009, 35 percent were initially identified as having cost sharing that 
exceeded CMS's cost-sharing thresholds and 29 percent were approved 
with cost sharing that exceeded CMS's cost-sharing thresholds, a 
decrease of 6 percentage points. The percentage of plans in the poor 
health group that exceeded CMS's thresholds prior to and after plans 
were approved decreased by 1 percentage point, from 38 percent to 37 
percent. 

Table 1: Percentage of MA Plans that Exceeded CMS Cost-Sharing 
Thresholds, Contract Years 2008 and 2009: 

Contract year 2008: All plans (n=2,899); 
Percentage of plans that initially exceeded one or more CMS 
thresholds: 34; 
Percentage of approved plans that exceeded one or more CMS thresholds: 
28. 

Contract year 2008: Good health plans (n=1,254); 
Percentage of plans that initially exceeded one or more CMS 
thresholds: 39; 
Percentage of approved plans that exceeded one or more CMS thresholds: 
33. 

Contract year 2008: Average health plans (n=1,068); 
Percentage of plans that initially exceeded one or more CMS 
thresholds: 35; 
Percentage of approved plans that exceeded one or more CMS thresholds: 
30. 

Contract year 2008: Poor health plans (n=577); 
Percentage of plans that initially exceeded one or more CMS 
thresholds: 20; 
Percentage of approved plans that exceeded one or more CMS thresholds: 
14. 

Contract year 2009: All plans (n=2,482); 
Percentage of plans that initially exceeded one or more CMS 
thresholds: 37; 
Percentage of approved plans that exceeded one or more CMS thresholds: 
30. 

Contract year 2009: Good health plans (n=1,050); 
Percentage of plans that initially exceeded one or more CMS 
thresholds: 35; 
Percentage of approved plans that exceeded one or more CMS thresholds: 
29. 

Contract year 2009: Average health plans (n=912); 
Percentage of plans that initially exceeded one or more CMS 
thresholds: 38; 
Percentage of approved plans that exceeded one or more CMS thresholds: 
29. 

Contract year 2009: Poor health plans (n=520); 
Percentage of plans that initially exceeded one or more CMS 
thresholds: 38; 
Percentage of approved plans that exceeded one or more CMS thresholds: 
37. 

Source: GAO analysis of CMS data. 

Notes: Using CMS's 2008 plan-level risk score data by county, we 
categorized the health status of MA plans as good, average, or poor 
based on the average health status of their enrolled beneficiaries in 
the areas that they served. The services we analyzed for contract 
years 2008 and 2009 included inpatient hospital stays, inpatient 
mental health stays, SNF stays, home health visits, renal dialysis, 
Part B drugs, Part B chemotherapy drugs, Part C premium, and Part C 
deductible. This analysis included 2,899 MA plans that were offered in 
2008 and 2,482 MA plans that were offered again in 2009 and we 
maintained plans in their 2008 health groups. 

[End of table] 

In contract years 2008 and 2009, CMS did not contact all plans 
initially found to have cost sharing that exceeded one or more 
thresholds. CMS primarily selected plans to contact for cost-sharing 
reductions from among those identified as likely to be discriminatory 
by considering (1) the number of services for which cost sharing 
exceeded the CMS threshold (2) how much the plan exceeded the cost- 
sharing threshold, and (3) how the plan's cost sharing compared with 
that of other MA plans within the same service area. In addition, CMS 
officials told us the process used to select plans for cost-sharing 
discussions in these years had the potential to be subjective; 
decisions were based on individual reviewers' evaluations of cost 
sharing and judgments about how much the cost sharing exceeded CMS's 
thresholds. Among the plans contacted for cost-sharing reductions in 
these years, CMS reported that some reduced their cost-sharing 
amounts, but they remained above the thresholds. For example, for 
contract year 2008, CMS reported that nearly half of the MA plans 
initially identified as likely to be discriminatory reduced cost 
sharing as a result of being contacted by CMS. 

Figures 10 and 11 show the review process for contract year 2010 and 
for contract years 2008 and 2009, respectively. 

Figure 10: CMS's Process for Identifying Plans with Cost Sharing 
Likely to Be Discriminatory and Contacting Plans for Cost-Sharing 
Reductions, Contract Year 2010: 

[Refer to PDF for image: illustration] 

Does MA plan have an OOP maximum at or below $3,400? 
No: 
Review cost sharing for 14 services and determine whether it is below, 
at, or above 75th percentile.[A] 
Above 75th percentile[B]: 
* Plan likely to be potentially discriminatory; 
* Contact plan for cost-sharing reductions. 

At or below 75th percentile[B]: 
* Plan not likely to be potentially discriminatory. 

Yes: 
Review cost sharing for five services and determine whether it is 
below, at, or above 95th percentile.[A] 
Above 95th percentile[B]: 
* Plan likely to be potentially discriminatory; 
* Contact plan for cost-sharing reductions. 

At or below 95th percentile[B]: 
* Plan not likely to be potentially discriminatory. 

Approve or deny plan. (For contract year 2010, all plans were approved 
and all reduced their cost sharing to amounts at or below CMS’s 
thresholds.) 

Does MA plan have an OOP maximum at or below $3,400? And: 
For five services, calculate the difference between the plan’s cost 
sharing and an amount that is actuarially equivalent under Medicare 
FFS and compare the difference to a tolerance amount (greater of 50 
cents or 5 percent of FFS cost sharing).[C] 
More than tolerance amount: 
* Plan likely to be potentially discriminatory; 
* Contact plan for cost-sharing reductions. 
Less than tolerance amount:
* Plan not likely to be potentially discriminatory. 

Approve or deny plan. (For contract year 2010, all plans were approved 
and all reduced their cost sharing to amounts at or below CMS’s 
thresholds.) 

Source: GAO analysis of CMS documents and interviews. 

[A] CMS reviews cost sharing for renal dialysis, Part B drugs, home 
health visits, inpatient mental health stays, and SNF stays if the 
plan has an OOP maximum at or below the amount specified in the annual 
Call Letter ($3,400 for contract year 2010). If the plan does not have 
an OOP maximum or it is above the amount specified in the annual Call 
Letter, CMS reviews cost sharing for the five services listed above 
and for inpatient hospital stays, inpatient hospital catastrophic 
stays, physician mental health visits, Part B chemotherapy drugs, Part 
B radiology, DME equipment, DME prosthetics, DME supplies, and DME 
diabetes tests. 

[B] CMS does not consider an MA plan with cost sharing above the 75th 
or 95th percentile, but below Medicare FFS cost sharing, as likely to 
be discriminatory. 

[C] CMS analyzes MA plan bid pricing data on cost sharing for 
inpatient hospital care (including mental health), skilled nursing 
facility, home health, DME, and Part B drugs. 

[End of figure] 

Figure 11: CMS's Process for Identifying Plans with Cost Sharing 
Likely to be Discriminatory and Selecting Plans for Cost-Sharing 
Reductions, Contract Years 2008 and 2009: 

[Refer to PDF for image: illustration] 

1) Does MA plan have an OOP maximum at or below $3,250 in 2008 or 
$3,350 in 2009? 

Yes: 
Plan Not considered likely to be discriminatory; 
Approve or deny plan. (For contract years 2008 and 2009, all plans 
were approved, but not all reduced cost sharing to amounts at or below 
CMS’s thresholds.) 

No: 
2) Compare cost sharing for high-risk services against CMS’s 
thresholds;[A] 
No services over threshold: 
Plan Not considered likely to be discriminatory; 
Approve or deny plan. (For contract years 2008 and 2009, all plans 
were approved, but not all reduced cost sharing to amounts at or below 
CMS’s thresholds.) 
One or more services over threshold: 
Plan considered likely to be discriminatory. 

3) To select a plan for negotiations, consider: (1) the number of 
services for which cost sharing exceeds the CMS threshold, (2) how 
much the plan exceeds the cost-sharing threshold, and (3) how the 
plan’s cost sharing compares to other MA plans; 
Plan Not chosen for cost-sharing reductions: 
Approve or deny plan. (For contract years 2008 and 2009, all plans 
were approved, but not all reduced cost sharing to amounts at or below 
CMS’s thresholds.) 
Plan chosen for cost-sharing reductions: 
Contact plan for cost-sharing reductions; 
Approve or deny plan. (For contract years 2008 and 2009, all plans 
were approved, but not all reduced cost sharing to amounts at or below 
CMS’s thresholds.) 

Source: GAO analysis of CMS documents and interviews. 

[A] High-risk services where high cost sharing could be considered 
likely to be discriminatory include inpatient hospital stays, 
inpatient hospital catastrophic stays, inpatient mental health stays, 
SNF stays, home health visits, physician mental health visits, Part B 
drugs, Part B chemotherapy, renal dialysis, and DME cost sharing as 
reported in the bid pricing tool. 

[End of figure] 

Agency and Other External Comments and Our Evaluation: 

We obtained comments on a draft of this report from CMS and America's 
Health Insurance Plans (AHIP), a national organization that represents 
private health insurance companies, including those that participate 
in the MA program. CMS provided written comments (see appendix II) and 
technical comments that we incorporated where appropriate. 
Representatives from AHIP provided us with oral comments. 

CMS Comments: 

CMS stated that the agency's general experience with beneficiaries-- 
that beneficiaries select MA plans based on their individual health 
status--is consistent with the report's findings. CMS also noted that 
the report showed that plans with, on average, beneficiaries in poor 
health did not charge comparatively higher cost sharing. The agency 
believes that this is a result, in part, of CMS's efforts in ensuring 
that plan benefit packages are not likely to discriminate on the basis 
of health status. 

CMS commented that the report does not mention that the agency's 
benefit package review process prior to contract year 2010, by design, 
encouraged plans to establish OOP maximums by affording them 
flexibility in establishing cost-sharing amounts for individual 
services. We believe that the third finding of the report adequately 
conveys that information. Figure 11 shows that plans with an OOP 
maximum at or below the amount specified in the Call Letter were not 
considered likely to be discriminatory regardless of their cost-
sharing amounts for particular services. Further, CMS pointed out that 
its policy until 2009 did not require disapproval of a plan benefit 
package if cost sharing was above CMS's cost-sharing thresholds. We 
added this information to the report. In addition to these points, CMS 
described other aspects of its previous and current benefit review 
process for context. 

CMS suggested a number of presentation modifications to the report. 
For example, because the agency's reviews focus on protecting 
vulnerable beneficiaries, CMS suggested that we present more 
information on the benefit packages of plans in the average or poor 
health group. The section of our report on benefit package designs 
contrasted the good health group of plans with the poor health group 
and the figures present information on plans in the average health 
group as well. 

AHIP Comments: 

AHIP representatives stated that the report's analysis demonstrates 
that MAOs are offering beneficiaries the types of plans that best meet 
their health care needs. The findings also show that special needs 
plans, which target enrollment to certain vulnerable groups of 
Medicare beneficiaries that typically have higher health care costs, 
are serving the types of beneficiaries that the program intended. 

As agreed with your offices, unless you publicly announce the contents 
of this report earlier, we plan no further distribution until 30 days 
from the report date. At that time, we will send copies to the CMS 
Administrator and interested congressional committees. The report will 
also be available at no charge on the GAO Web site at [hyperlink, 
http://www.gao.gov]. 

Should you or your staff have any questions on matters discussed in 
this report, please contact me at (202) 512-7114 or cosgrovej@gao.gov. 
Contact points for our Offices of Congressional Relations and Public 
Affairs may be found on the last page of this report. GAO staff 
members who made contributions to this report are listed in appendix 
III. 

Signed by: 

James C. Cosgrove: 
Director, Health Care: 

List of Requesters: 

The Honorable Henry A. Waxman: 
Chairman: 
The Honorable John D. Dingell: 
Chairman Emeritus: 
Committee on Energy and Commerce: 
House of Representatives: 

The Honorable Sander M. Levin: 
Chairman: 
Committee on Ways and Means: 
House of Representatives: 

The Honorable Frank Pallone, Jr. 
Chairman: 
Subcommittee on Health: 
Committee on Energy and Commerce: 
House of Representatives: 

The Honorable Pete Stark: 
Chairman: 
Subcommittee on Health: 
Committee on Ways and Means: 
House of Representatives: 

The Honorable Charles B. Rangel: 
House of Representatives: 

[End of section] 

Appendix I: Medicare Advantage Plans and Beneficiaries by Average 
Indexed Risk Score and Health Group, 2008: 

Figure 12: Number of Medicare Advantage Plans by Average Indexed Risk 
Score and Health Group, 2008: 

[Refer to PDF for image: stacked vertical bar graph] 

Average indexed risk score: Less than 0.65; 
Good health—MA (SNP): 1; 
Good health—MA (non-SNP): 88. 

Average indexed risk score: 0.65-0.70; 
Good health—MA (SNP): 0; 
Good health—MA (non-SNP): 98. 

Average indexed risk score: 0.70-0.75; 
Good health—MA (SNP): 1; 
Good health—MA (non-SNP): 160. 

Average indexed risk score: 0.75-0.80; 
Good health—MA (SNP): 5; 
Good health—MA (non-SNP): 195. 

Average indexed risk score: 0.80-0.85; 
Good health—MA (SNP): 7; 
Good health—MA (non-SNP): 296. 

Average indexed risk score: 0.85-0.90; 
Good health—MA (SNP): 22; 
Good health—MA (non-SNP): 381. 

Average indexed risk score: 0.90-0.95; 
Average health—MA (SNP): 33; 
Average health—MA (non-SNP): 333. 

Average indexed risk score: 0.95-1.00; 
Average health—MA (SNP): 30; 
Average health—MA (non-SNP): 270. 

Average indexed risk score: 1.00-1.05; 
Average health—MA (SNP): 56; 
Average health—MA (non-SNP): 187. 

All Medicare beneficiaries in plan’s service area = 1.00, 

Average indexed risk score: 1.05-1.10; 
Average health—MA (SNP): 58; 
Average health—MA (non-SNP): 101. 

Average indexed risk score: 1.10-1.15; 
Poor health—MA (SNP): 57; 
Poor health—MA (non-SNP): 66. 

Average indexed risk score: 1.15-1.20; 
Poor health—MA (SNP): 55; 
Poor health—MA (non-SNP): 42. 

Average indexed risk score: 1.20-1.25; 
Poor health—MA (SNP): 55; 
Poor health—MA (non-SNP): 23. 

Average indexed risk score: 1.25-1.30; 
Poor health—MA (SNP): 48; 
Poor health—MA (non-SNP): 17. 

Average indexed risk score: 1.30-1.35; 
Poor health—MA (SNP): 33; 
Poor health—MA (non-SNP): 8. 

Average indexed risk score: Greater than 1.35; 
Poor health—MA (SNP): 160; 
Poor health—MA (non-SNP): 13. 

Source: GAO analysis of 2008 CMS risk score data. 

Notes: Using the Centers for Medicare & Medicaid Services (CMS) 2008 
plan-level risk score data by county, we categorized the health status 
of Medicare Advantage plans as good, average, or poor based on the 
average health status of their enrolled beneficiaries in the areas 
that they served. This analysis included 2,899 MA plans, of which 621 
were SNPs that are allowed to limit enrollment to a targeted 
beneficiary population. 

[End of figure] 

Figure 13: Percentage of Medicare Advantage Beneficiaries by MA Plans' 
Average Indexed Risk Score and Health Group, 2008: 

[Refer to PDF for image: stacked vertical bar graph] 

Average indexed risk score: Less than 0.65; 
Good health—MA (SNP): 0; 
Good health—MA (non-SNP): 0.2%. 

Average indexed risk score: 0.65-0.70; 
Good health—MA (SNP): 0; 
Good health—MA (non-SNP): 0.4%. 

Average indexed risk score: 0.70-0.75; 
Good health—MA (SNP): 0; 
Good health—MA (non-SNP): 1.7%. 

Average indexed risk score: 0.75-0.80; 
Good health—MA (SNP): 0; 
Good health—MA (non-SNP): 2.6%. 

Average indexed risk score: 0.80-0.85; 
Good health—MA (SNP): 0; 
Good health—MA (non-SNP): 9%. 

Average indexed risk score: 0.85-0.90; 
Good health—MA (SNP): 0.1%; 
Good health—MA (non-SNP): 15.3%. 

Average indexed risk score: 0.90-0.95; 
Average health—MA (SNP): 0.4%; 
Average health—MA (non-SNP): 18.1%. 

Average indexed risk score: 0.95-1.00; 
Average health—MA (SNP): 0.3%; 
Average health—MA (non-SNP): 17.2%. 

Average indexed risk score: 1.00-1.05; 
Average health—MA (SNP): 1.5%; 
Average health—MA (non-SNP): 11.6%. 

All Medicare beneficiaries in plan’s service area = 1.00, 

Average indexed risk score: 1.05-1.10; 
Average health—MA (SNP): 1.3%; 
Average health—MA (non-SNP): 4.1%. 

Average indexed risk score: 1.10-1.15; 
Poor health—MA (SNP): 1.8%; 
Poor health—MA (non-SNP): 2.8%. 

Average indexed risk score: 1.15-1.20; 
Poor health—MA (SNP): 1.2%; 
Poor health—MA (non-SNP): 2.8%. 

Average indexed risk score: 1.20-1.25; 
Poor health—MA (SNP): 1.3%; 
Poor health—MA (non-SNP): 0.3%. 

Average indexed risk score: 1.25-1.30; 
Poor health—MA (SNP): 1.5%; 
Poor health—MA (non-SNP): 0.6%. 

Average indexed risk score: 1.30-1.35; 
Poor health—MA (SNP): 0.5%; 
Poor health—MA (non-SNP): 0.1%. 

Average indexed risk score: Greater than 1.35; 
Poor health—MA (SNP): 2.7%; 
Poor health—MA (non-SNP): 0.1%. 

Source: GAO analysis of 2008 CMS risk score data. 

Notes: Results are based on Medicare enrollment as of July 2008. Using 
CMS's 2008 plan-level risk score data by county, we categorized the 
health status of MA plans as good, average, or poor based on the 
average health status of their enrolled beneficiaries in the areas 
that they served. This analysis included 7,553,600 MA beneficiaries, 
of which 973,923 were enrolled in SNPs that are allowed to limit 
enrollment to a targeted beneficiary population. 

[End of figure] 

[End of section] 

Appendix II: Comments from the Centers for Medicare & Medicaid 
Services: 

Department Of Health & Human Services: 
Office Of The Secretary: 
Assistant Secretary for Legislation: 
Washington, DC 20201: 

April 7, 2010: 

James C. Cosgrove: 
Director, Health Care: 
U.S. Government Accountability Office: 
441 G Street N.W. 
Washington, DC 20548: 

Dear Mr. Cosgrove: 

Enclosed are comments on the U.S. Government Accountability Office's 
(GAO) report entitled: "Medicare Advantage: Relationship Between 
Benefit Package Designs and Plans' Average Beneficiary Health Status" 
(GA0-10-403). 

The Department appreciates the opportunity to review this report 
before its publication. 

Sincerely, 

Signed by: 

Andrea Palm: 
Acting Assistant Secretary for Legislation: 

Enclosure: 

[End of letter] 

Department Of Health & Human Services: 
Centers for Medicare & Medicaid Services: 
Administrator: 
Washington, DC 20201: 

Date: April 5, 2010: 

To: Andrea Palm: 
Acting Assistant Secretary for Legislation: 
Office of the Secretary: 

From: [Signed by] Charlene Frizzera: 
Acting Administrator: 

Subject: Government Accountability Office (GAO) Report: "Medicare 
Advantage: Relationship Between Benefit Package Designs and Plans' 
Average Beneficiary Health Status" (GAO-10-403): 

Thank you for the opportunity to comment on the GAO report, "Medicare 
Advantage: Relationship between Benefit Package Designs and Plans' 
Average Beneficiary Health Status" (GAO-10-403). The report compares 
the benefit packages of Medicare Advantage (MA) organizations based on 
enrollees' average risk scores: "good," "average," and "poor" and 
found that, generally, those plans with enrollees in "good" health 
based on their average risk scores charged lower premiums and higher 
cost-sharing for certain benefits and offered fewer additional 
benefits. In contrast, those plans with enrollees in "poor" health 
charged higher premiums and lower cost-sharing for certain benefits. 
Notably, however, the report did not determine whether MA 
organizations structured their plan benefit packages in response to 
enrolled beneficiaries' health status or whether beneficiaries of a 
given health status chose certain MA plans specifically because of 
their design. We believe that the report's findings are consistent 
with our general experience with beneficiaries — that beneficiaries 
select health plans based on their individual health status, and 
therefore, beneficiaries with poorer health status tend to choose 
plans with higher premiums and lower cost sharing, and, conversely, 
healthier beneficiaries choose to make lower premium payments in favor 
of higher cost sharing. 

We find promising the conclusion that plans with beneficiaries in poor 
health status are not charging higher cost-sharing; we believe that 
this is a result, in part, of CMS' efforts to prevent MA organizations 
from designing benefit packages that discriminate against sicker 
beneficiaries and discourage them from joining their plans. We have 
made it a priority to protect our most vulnerable beneficiaries from 
being faced with choices of MA plans that discriminate against them. 
We examine benefits and cost sharing designs for situations in which 
the plan offered is designed to discourage enrollment or to charge 
excessively high cost-sharing after enrollment for those beneficiaries 
who are in poor health and more likely to use health care services. To 
that end, we would like to see more information presented in the 
report about the benefit packages of plans with enrollees in "average" 
and "poor" health. 

Since the report describes our previous efforts to review benefits and 
protect beneficiaries against discriminatory plan offerings, we would 
like to provide additional context for CMS' benefit review efforts 
prior to contract year (CY) 2010. While CMS' selection of plans with 
possible discriminatory cost sharing was less comprehensive prior to 
CY 2010, and we did not contact all plans with high cost sharing 
issues, we did identify and negotiate with plans that had the most 
egregious cost sharing issues in an effort to reduce cost sharing to 
our target amount or as close to it as we were able to negotiate. We 
used a variety of factors, in addition to cost sharing values, to 
assess whether to contact plans to request changes in cost-sharing. 
The report fails to mention that, by design, we encouraged plans to 
establish a maximum out-of-pocket (MOOP) limit for all Medicare-
covered benefits by affording them flexibility in their cost-sharing 
amounts for individual services. 

Additionally, we determined that excessive charges for certain 
benefits (e.g., renal dialysis and Part B prescription drugs) were the 
most problematic for beneficiaries, and considered not just a plan's 
cost-sharing requirements, but the amount by which the plan's cost 
sharing exceeded the CMS target amount. Also, part of our decision to 
negotiate cost sharing amounts was based on how a plan's cost sharing 
charges compared with those of other MA plans under review in the same 
service area because CMS wanted to evaluate whether higher cost 
sharing was a market phenomenon or possible discrimination by a single 
plan. In many cases, although CMS negotiated cost sharing 
improvements, the decreases were not great enough to put plans' levels 
below the CMS threshold and thus CMS' impact on reducing high cost 
sharing is not obvious and is subject to understatement. It is also 
important to point out that CMS' policy until 2009 did not require 
disapproval of a plan benefit package if cost sharing was above the 
threshold. 

Finally, the report correctly points out that CMS revised its 
methodology for identifying, reviewing and approving benefits packages 
for Contract Year 2010. The change in methodology included the 
development of standards based on outlier cost sharing analysis — 
higher allowable cost sharing standards for plans that offered a 
qualified MOOP of $3,400 and more restrictive cost sharing for plans 
without a qualified MOOP. The new methodology also required MA plans 
to adhere to CMS' cost sharing standards in order to receive approval 
of their bid. Therefore, given the significant differences between our 
processes in 2009 and 2010, we suggest that the report describe the 
process for these two years separately. 

In addition to these thematic concerns, CMS has several general and 
technical comments which are attached. Again, we thank you for the 
opportunity to review and comment on this draft report. 

[End of section] 

Appendix III: GAO Contact and Staff Acknowledgments: 

Contact: 

James C. Cosgrove, (202) 512-7114 or cosgrovej@gao.gov: 

Acknowledgments: 

Other contributors to this report include Rosamond Katz, Assistant 
Director; Elizabeth Conklin; Ba Lin; Hillary Loeffler; Kevin Milne; 
Elizabeth Morrison; and Kristal Vardaman. 

[End of section] 

Footnotes: 

[1] MAOs may offer multiple MA plans with different combinations of 
authorization requirements, non-Medicare benefits, cost sharing, and 
premiums. MA plans must cover Medicare-covered benefits except hospice 
care. 

[2] Centers for Medicare & Medicaid Services (CMS), CMS Financial 
Report Fiscal Year 2009 (Baltimore, Md.: November 2009) 

[3] Medicare FFS beneficiaries can purchase Medigap insurance 
policies, offered by private insurers, that help cover cost-sharing 
amounts for Medicare-covered services. 

[4] Medicare spending is concentrated among a small proportion of 
beneficiaries. The most expensive 5 percent of Medicare beneficiaries 
account for roughly half of all spending in the Medicare FFS program. 
See MedPAC, Report to the Congress: New Approaches in Medicare 
(Washington, D.C.: June 2004). 

[5] Medicare Payment Advisory Commission (MedPAC), Report to the 
Congress: Issues in a Modernized Medicare Program (Washington, D.C.: 
June 2005). 

[6] Actuarial equivalence is demonstrated by a qualified actuary's 
certification that overall cost sharing in an MA plan is no more than 
the overall cost sharing in Medicare FFS. 

[7] 42 U.S.C. § 1395w-22(b)(1)(A). 

[8] See GAO, Medicare Advantage: Increased Spending Relative to 
Medicare Fee-for-Service May Not Always Reduce Beneficiary Out-of- 
Pocket Costs, [hyperlink, http://www.gao.gov/products/GAO-08-359] 
(Washington, D.C.: Feb. 22, 2008). 

[9] Beneficiaries in HMOs are generally restricted to seeing providers 
within a network, while PFFS beneficiaries can see any provider 
authorized to provide Medicare services that accepts the plan's 
payment terms. Beneficiaries in PPOs can see both in-network and out-
of-network providers but may pay higher cost-sharing amounts if they 
use out-of-network services. A regional PPO serves an entire state or 
multiple states, whereas local PPOs may serve a county, partial 
county, or multiple counties. 

[10] SNPs are permitted to target enrollment to (1) beneficiaries 
entitled to Medicare and Medicaid (dual-eligible beneficiaries), (2) 
beneficiaries with severe or disabling chronic conditions, and (3) 
institutionalized beneficiaries. They may either exclusively or 
disproportionately enroll beneficiaries from one of these three 
categories. 

[11] We excluded from our analysis employer plans, religious fraternal 
benefit plans, Part B only plans, CMS demonstrations, programs of all- 
inclusive care for the elderly, cost plans, provider sponsored 
organizations (as they constituted less than 1 percent of MA 
enrollment), medical savings accounts, beneficiaries with end-stage 
renal disease, beneficiaries located in areas outside the 50 states 
and the District of Columbia, and MA plans with fewer than 10 
beneficiaries. 

[12] Diagnosis codes are assigned by providers and reported for FFS 
beneficiaries in Medicare claims data. MA plans are responsible for 
providing CMS with the appropriate diagnosis codes for their enrolled 
beneficiaries as the Medicare program does not have claims data for 
these beneficiaries. 

[13] Services typically used by sicker beneficiaries, for whom CMS 
considers high cost sharing to be potentially discriminatory, include: 
inpatient hospital acute care, inpatient mental health care, renal 
dialysis, chemotherapy drugs and other part B drugs, skilled nursing 
facility stays, home health visits, and durable medical equipment 
(DME). 

[14] A deductible is an amount (typically annual) that a beneficiary 
is responsible for before an insurer will make payments. 

[15] Dental, vision, hearing, and fitness benefits (which may include 
gym memberships and fitness classes) are not provided under Medicare 
FFS but may be offered by MA plans as mandatory or optional 
supplemental benefits. Mandatory benefits must be provided for every 
person enrolled in the plan, whereas optional supplemental benefits 
are available to those enrollees who elect and pay for them. 

[16] Medicare Options Compare (available at www.medicare.gov) provides 
information, using CMS PBP data, to beneficiaries so they can find and 
compare MA plans available in the area where they live. 

[17] Individuals with end-stage renal disease are not eligible to 
enroll in most MA plans. However, if these individuals develop the 
disease while enrolled in an MA plan, they may remain enrolled in 
their plan or move to a different MA plan if their plan is terminated. 
42 U.S.C. § 1395w-21(a)(3)(B). 

[18] The exception is hospice care, which FFS covers and MA plans do 
not. 

[19] Coinsurance is a percentage payment for a given service that a 
beneficiary must pay, such as 20 percent of the total payment for Part 
B drugs. A copayment is a standard amount that a beneficiary must pay 
for a given service, such as $200 per day for days 1 through 6 of an 
inpatient acute hospital stay. 

[20] Regional PPOs are required to have OOP maximums, and Medicare FFS 
has no OOP maximum. Beginning January 1, 2011, CMS will require local 
MA plans to have an OOP maximum, the amount of which would be set 
annually by CMS. See 75 Fed. Reg. 19678, 19709-19711 (2010). 

[21] Each year CMS publishes the OOP maximum threshold in its annual 
Call Letter to MA plans. This value represents the maximum amount that 
85 percent of Medicare FFS beneficiaries are expected to incur in 
Parts A and B deductibles and coinsurance in a given year. For 
calendar years 2008 and 2009 the OOP maximum thresholds were $3,250 
and $3,350, respectively. 

[22] See GAO, Medicare Advantage: Increased Spending Relative to 
Medicare Fee-for-Service May Not Always Reduce Beneficiary Out-of- 
Pocket Costs, [hyperlink, http://www.gao.gov/products/GAO-08-359] 
(Washington, D.C.: Feb. 22, 2008). 

[23] Gold, M. and Hudson, M.C., Medicare Advantage Benefit Design: 
What Does It Provide, What Doesn't It Provide, and Should Standards 
Apply?, A report for AARP Public Policy Institute (Washington, D.C.: 
March 2009). 

[24] See 42 U.S.C. § 1395w-22(b)(1)(A). Regulations provide that CMS 
shall review and approve MA plan benefit packages to ensure MAOs are 
not designing benefits to discriminate against beneficiaries, promote 
discrimination, discourage enrollment or encourage enrollment, steer 
subsets of particular beneficiaries, or inhibit access to services. 
See 42 CFR § 422.100(f). 

[25] See Pub. L. No. 111-148, § 3209, _Stat._ (2010). 

[26] See Pub. L. No. 111-148, § 3202, _Stat._ (2010). 

[27] Results weighted by July 2008 enrollment. 

[28] Of the 2,899 plans in our analysis, 5 percent required enrollees 
to pay a deductible before plan coverage began. We considered an MA 
plan to have a deductible if the plan had either an in-network 
deductible or a deductible for both in-network and out-of-network 
services. 

[29] For a 6-day inpatient hospital stay, 15 percent of MA plans in 
our analysis did not charge cost sharing. 

[30] For inpatient mental health stays, CMS does not calculate an 
average cost per day. As a result, we were only able to estimate the 
average OOP cost for a 21-day inpatient mental health stay for plans 
that charged a copayment. Approximately 14 percent of MA plans in our 
analysis did not charge cost sharing for a 21-day stay. 

[31] For a 35-day SNF stay, 16 percent of MA plans in our analysis did 
not charge cost sharing. 

[32] For 156 renal dialysis sessions, 37 percent of MA plans in our 
analysis did not charge cost sharing. 

[33] Among the plans in our analysis that charged cost sharing for 
Part B chemotherapy drugs, other Part B drugs, and DME, 85 percent 
charged coinsurance for Part B chemotherapy drugs, 84 percent charged 
coinsurance for other Part B drugs, and 99 percent charged coinsurance 
for DME. For 27 home health visits, 75 percent of plans in our 
analysis did not charge cost sharing. Among the plans that charged 
cost sharing for home health, 62 percent charged a copayment. 

[34] We considered a plan to have an OOP maximum if the plan had 
either an in-network OOP maximum or an OOP maximum for both in-network 
and out-of-network services. If a plan had two OOP maximums--one for 
in-network services and one for combined in-and out-of-network 
services, then we used the in-network value for this analysis. 

[35] Results were weighted by July 2008 enrollment. 

[36] A plan was considered to have excluded a service category from 
the OOP maximum if the OOP maximum did not cover that service category 
and if the plan had no service-specific maximum for that category. 
Plans that excluded a certain service category from the OOP maximum 
did not necessarily exclude all services from that category. 

[37] We examined whether a plan charged cost sharing and excluded the 
following services from their OOP maximum: inpatient hospital acute 
care, inpatient mental health care, skilled nursing facility stays, 
home health visits, renal dialysis, Part B drugs, and DME. 

[38] In determining whether plans charged cost sharing for services 
excluded from the OOP maximum, we examined plan cost sharing for a 6- 
day inpatient hospital acute stay, a 21-day inpatient mental health 
stay, a 35-day skilled nursing facility stay, 27 home health visits, 1 
year of dialysis (156 sessions), Part B drugs, and DME. 

[39] Results include MA plans that provided dental, vision, hearing, 
and fitness benefits as mandatory supplemental benefits. 

[40] Comprehensive dental benefits may include restorative services 
(e.g., fillings), root canals, oral surgery, extractions, and dentures. 

[41] See appendix I for the number of MA plans and percentage of 
beneficiaries by average indexed risk score and health group. 

[42] In 2008, we reported that beneficiaries in PFFS plans tended to 
be healthier than beneficiaries in other MA plans and Medicare FFS. 
Specifically, projected health care expenditures for PFFS 
beneficiaries were 7 percent less than the projected average for 
beneficiaries in other MA plans and 10 percent less than the projected 
average for beneficiaries in Medicare FFS. See GAO, Medicare 
Advantage: Characteristics, Financial Risks, and Disenrollment Rates 
of Beneficiaries in Private Fee-for-Service Plans, GAO-09-25 
(Washington, D.C.: Dec. 15, 2008). 

[43] To calculate the extent to which MA payment benchmarks exceeded 
estimated FFS spending, we first projected 2008 FFS spending by county 
using 2007 estimates in the 2008 MA ratebook updated by the CMS 
estimate of growth in national spending for 2008. We discounted 
spending related to the double payment for indirect medical education 
payments made to teaching hospitals. We used the statutory benchmarks 
in our analysis of regional PPOs and did not examine the competitively 
set benchmark amounts. 

[44] We did not find a relationship between an MA plan's average 
indexed risk score and the percent difference between MA payment 
benchmarks and estimated FFS spending in 2008 for the other types of 
MA plans included in our analysis (local PPOs, regional PPOs, or PFFS 
plans). 

[45] For contract years 2011 and later, CMS will require local MA 
plans to have an OOP maximum, the amount of which would be set 
annually by CMS. If plans establish OOP maximums lower than the 
mandatory amount, they will be allowed more flexibility in 
establishing cost-sharing amounts for individual services. See Fed. 
Reg. 19678, 19709-19711 (2010). 

[46] CMS annually determines the length of stay or number of visits or 
sessions and the cost of certain services in order to determine 
whether the cost-sharing amount charged by an MA plan is potentially 
discriminatory. 

[47] According to CMS, if a plan with a qualifying OOP maximum 
excluded any Medicare service from its OOP maximum, it was evaluated 
based on more stringent standards. 

[48] For 2010, beneficiaries in Medicare FFS are charged 20 percent 
coinsurance for renal dialysis and Part B drugs; nothing per home 
health visit; $1,100 for a 15-day inpatient mental health stay; and 
$3,025 for a 42-day SNF stay. 

[49] For 2010, beneficiaries in Medicare FFS are charged $1,100 for a 
10-day inpatient hospital stay; $9,350 for a 90-day inpatient hospital 
catastrophic stay; 45 percent coinsurance for a physician mental 
health visit; and 20 percent coinsurance for Part B chemotherapy 
drugs, Part B radiology, and DME (including equipment, prosthetics, 
supplies, and diabetes tests). 

[50] In prior years, CMS also considered MA plan stability grades that 
assessed whether plan cost sharing amounts differed substantially from 
one year to the next. According to CMS officials, these grades were 
implemented because beneficiaries were often unaware of plan benefit 
changes from one year to the next and CMS wanted to identify plans 
that were making significant changes to their benefit packages. 
However, CMS did not use the stability grades to determine if a plan 
was likely to be discriminatory and is not using plan stability grades 
in the new review process. 

[51] For plan years beginning January 1, 2011, cost sharing for 
chemotherapy services, renal dialysis, and skilled nursing care may be 
no more than cost sharing in Medicare FFS. See Pub. L. No. 111-148, § 
3202, _Stat._ (2010). The Part B chemotherapy drugs and renal dialysis 
cost-sharing thresholds for contract year 2010 were equal to the cost 
sharing charged in Medicare FFS. For MA plans without an OOP maximum, 
or with an OOP maximum above the amount specified in the 2010 Call 
Letter, the cost-sharing threshold for a 42-day SNF stay was less than 
the cost sharing charged in Medicare FFS for a Part A SNF stay. For MA 
plans with an OOP maximum at or below the amount specified in the 2010 
Call Letter, the cost-sharing threshold for a 42-day SNF stay was $359 
more than the cost sharing charged in Medicare FFS for a Part A SNF 
stay. 

[52] Of the 1,719 MA plans included in our analysis for contract year 
2010, 35 had cost sharing that appeared to exceed CMS's thresholds. 
Agency officials commented that these plans had extenuating 
circumstances or other financial protections for beneficiaries 
enrolled in those plans. 

[53] We analyzed data for 2,899 MA plans that were offered in contract 
year 2008, 2,482 MA plans that were offered again in 2009, and 1,719 
MA plans that were offered again in 2010 and we maintained them in 
their 2008 health groups. Some plans withdrew from the MA program, 
were terminated, consolidated, or split into multiple plans between 
contract years 2008, 2009, and 2010. 

[End of section] 

GAO's Mission: 

The Government Accountability Office, the audit, evaluation and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the performance 
and accountability of the federal government for the American people. 
GAO examines the use of public funds; evaluates federal programs and 
policies; and provides analyses, recommendations, and other assistance 
to help Congress make informed oversight, policy, and funding 
decisions. GAO's commitment to good government is reflected in its core 
values of accountability, integrity, and reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, go to [hyperlink, http://www.gao.gov] 
and select "E-mail Updates." 

Order by Phone: 

The price of each GAO publication reflects GAO’s actual cost of
production and distribution and depends on the number of pages in the
publication and whether the publication is printed in color or black and
white. Pricing and ordering information is posted on GAO’s Web site, 
[hyperlink, http://www.gao.gov/ordering.htm]. 

Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537. 

Orders may be paid for using American Express, Discover Card,
MasterCard, Visa, check, or money order. Call for additional 
information. 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

Congressional Relations: 

Ralph Dawn, Managing Director, dawnr@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, D.C. 20548: 

Public Affairs: 

Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, D.C. 20548: