GAO-04-960, Medicare Demonstration PPOs: Financial and Other Advantages for Plans, Few Advantages for Beneficiaries


This is the accessible text file for GAO report number GAO-04-960 
entitled 'Medicare Demonstration PPOs: Financial and Other Advantages 
for Plans, Few Advantages for Beneficiaries' which was released on 
September 27, 2004.

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 the Ranking Minority Member, Committee on Finance, U.S. 
Senate:

United States Government Accountability Office:

GAO:

September 2004:

Medicare Demonstration PPOs:

Financial and Other Advantages for Plans, Few Advantages for 
Beneficiaries:

GAO-04-960:

GAO Highlights:

Highlights of GAO-04-960, a report to the Ranking Minority Member, 
Committee on Finance, U.S. Senate

Why GAO Did This Study:

Preferred provider organizations (PPO) are more prevalent than other 
types of health plans in the private market, but, in 2003, only six 
PPOs contracted to serve Medicare beneficiaries in Medicare+Choice 
(M+C), Medicare’s private health plan option. In recent years, the 
Centers for Medicare & Medicaid Services (CMS), the agency that 
administers Medicare, initiated two demonstrations that include a 
total of 34 PPOs. GAO (1) described how CMS used its statutory 
authority to conduct the two demonstrations, (2) assessed the extent 
to which demonstration PPOs expanded access to Medicare health plans 
and attracted enrollees in 2003, (3) compared CMS’s estimates of out-
of-pocket costs beneficiaries incurred in demonstration PPOs with those 
of other types of coverage, including fee-for-service (FFS) Medicare, 
M+C plans, and Medigap policies in 2003, and (4) determined the effects 
of demonstration PPOs on Medicare spending. 

What GAO Found:

CMS used its statutory authority to offer health-care organizations 
financial incentives to participate in the two demonstrations. CMS, 
however, exceeded its authority when it allowed 29 of the 33 plans in 
the second demonstration, the Medicare PPO Demonstration, to cover 
certain services, such as skilled nursing, home health, and routine 
physical examinations, only if beneficiaries obtained them from the 
plans’ network providers. In general, beneficiaries in Medicare PPO 
Demonstration plans who received care from non-network providers for 
these services were liable for the full cost of their care. 
 
The demonstration PPOs attracted relatively few enrollees and did 
little to expand Medicare beneficiaries’ access to private health 
plans. About 98,000, or less than 1 percent, of the 10.1 million 
eligible Medicare beneficiaries living in counties where demonstration 
PPOs operated had enrolled in the demonstration PPOs by October 2003. 
Further, although one of the goals of the Medicare PPO Demonstration 
was to attract beneficiaries from traditional FFS Medicare and Medigap 
plans, only 26 percent of enrollees in its plans came from FFS 
Medicare, with all others coming from M+C plans. About 9.9 million, or 
98 percent, of the 10.1 million eligible beneficiaries living in 
counties where demonstration PPOs operated, had M+C plans available in 
their counties. Virtually no enrollment occurred in counties where only 
demonstration PPOs operated. 

According to CMS’s 2003 estimates, on average demonstration PPO 
enrollees could have expected to incur total out-of-pocket costs—
expenses for premiums, cost sharing and noncovered items and services—
that were the same or higher than those they would have incurred with 
nearly all other types of Medicare coverage. However, relative costs 
by type of coverage varied somewhat depending on beneficiary health 
status. For certain services and items, such as prescription drugs and 
inpatient hospitalization, demonstration plans provided better benefits 
relative to some other types of Medicare coverage.

Although it is too early to determine the actual program costs of the 
two demonstrations, CMS originally projected that the first 
demonstration would increase Medicare spending by $750 per enrollee 
per year and the second demonstration would increase Medicare spending 
by $652 per enrollee per year. Based on the agency’s original 
enrollment projections, which exceed 2003 actual enrollment, CMS 
estimated the demonstration PPOs would increase program spending by 
$100 million for 2002 and 2003 combined. 

What GAO Recommends:

GAO recommends that the Administrator of CMS promptly instruct plans 
in the Medicare PPO Demonstration to provide coverage for all plan 
services furnished by any provider authorized to provide Medicare 
services who accepts the plans’ terms and conditions of payment. CMS 
agreed to implement the recommendation, and stated that it believes 
the demonstrations are worthwhile. 

www.gao.gov/cgi-bin/getrpt?GAO-04-960.

To view the full product, including the scope and methodology, click on 
the link above. For more information, contact A. Bruce Steinwald at 
(202) 512-7119.

[End of section]

Contents:

Letter:

Results in Brief:

Background:

CMS Offered All Demonstration PPOs Financial Incentives, but Improperly 
Allowed Plans in the Medicare PPO Demonstration to Limit Coverage:

Demonstration PPOs Did Little to Expand Health Plan Options, and Have 
Enrolled Relatively Few Beneficiaries:

Demonstration PPOs Left Beneficiaries Exposed to Relatively High Total 
Out-of-pocket Costs, but Offered Slightly Better Coverage for Some 
Benefits:

Demonstration PPOs Were Projected to Increase Medicare Spending:

Conclusions:

Recommendation for Executive Action:

Agency Comments and Our Evaluation:

Appendix I: Analysis of PPO Demonstration Participants' Restriction on 
Enrollee Choice of Provider:

Background:

Discussion:

Appendix II: Scope and Methodology:

Enrollment:

Estimated Out-of-pocket Cost Comparisons:

Benefits Comparison:

Costs of Out-of-network Services:

Effect on Medicare Spending:

Data Reliability:

Appendix III: CMS Comments:

Appendix IV: GAO Contact and Staff Acknowledgments:

GAO Contact:

Acknowledgments:

Tables:

Table 1: Characteristics of HMO, PPO, and PFFS plans under M+C:

Table 2: Quality Assurance and Access-to-Services Requirements for M+C 
PPOs and M+C PFFS Plans Compared to M+C HMOs, 2003:

Table 3: M+C Plan Availability in the 214 Counties Where Demonstration 
PPO Plans Were Available, 2003:

Table 4: Demonstration PPO Market Penetration in Counties with and 
Without M+C Plans, October 2003:

Table 5: Prescription Drug Benefits Offered in Hillsborough County, 
Fla., in 2003, by Type of Coverage:

Figures:

Figure 1: Illustration of a Hypothetical Risk-Sharing Agreement (MLR of 
87 Percent, Shared Risk of 50 Percent, Risk Corridor of 2 Percent):

Figure 2: Location of Demonstration PPOs and M+C Plans by County, 2003:

Figure 3: Percentage of Enrollment in Demonstration PPOs by Plan, 2003:

Figure 4: Estimated Average Beneficiary Out-of-pocket Health Care Costs 
for Premiums, Cost Sharing, and Noncovered Items and Services per 
Month, by Type of Coverage, 2003:

Figure 5: Estimated Average Beneficiary Premiums and Other Out-of-
pocket Costs by Type of Coverage per Month, Excluding FFS Medicare Part 
B Premium, 2003:

Figure 6: Estimated Beneficiary Out-of-pocket Health Care Costs per 
Month by Type of Coverage and Beneficiary Health Status, 2003:

Abbreviations:

ACRP: Adjusted Community Rate Proposal: 
BBA: Balanced Budget Act of 1997: 
CMS: Centers for Medicare & Medicaid Services: 
FFS: fee-for-service: 
Fu: Fu Associates, Ltd.: 
GSA: Geographic Service Area: 
HMO: health maintenance organization: 
M+C: Medicare+Choice: 
MHPC: Medicare Health Plan Compare: 
MLR: medical loss ratio: 
MMA: Medicare Prescription Drug, Improvement, and Modernization Act of 
2003:  
MMCC: Medicare Managed Care Contract: 
MPPF: Medicare Personal Plan Finder: 
OACT: Office of the Actuary: 
ORDI: Office of Research, Development, and Information: 
PFFS: private fee-for-service: 
PPO: preferred provider organization:

United States Government Accountability Office:

Washington, DC 20548:

September 27, 2004:

The Honorable Max Baucus: 
Ranking Minority Member: 
Committee on Finance: 
United States Senate:

Dear Senator Baucus:

In recent years, concerns have been raised that Medicare beneficiaries 
lack access to the type of health plan that is most prevalent in the 
private health insurance market, the preferred provider organization 
(PPO). In 2003, only six PPOs--plans that allow enrollees to obtain 
care from any provider, but charge enrollees less if they obtain care 
from the plans' networks of preferred providers--participated in 
Medicare's program for private health plans, known as Medicare+Choice 
(M+C).[Footnote 1] About 3,000 of Medicare's 41 million beneficiaries 
were enrolled in the six M+C PPOs. In contrast, 4.6 million Medicare 
beneficiaries were enrolled in 142 M+C health maintenance organization 
(HMO) plans, which generally require enrollees to obtain all covered 
services from the plans' networks of providers.[Footnote 2] The vast 
majority of Medicare beneficiaries, about 35.6 million, were not 
enrolled in private health plans participating in Medicare, but rather 
were enrolled in the traditional fee-for-service (FFS) program, in 
which they could obtain care from any Medicare provider. The percentage 
of Medicare beneficiaries enrolled in traditional FFS Medicare has 
increased in recent years, rising from 83 percent in 1998 to 87 percent 
in 2003--as the total number of private plans participating in M+C 
declined, and their benefit packages grew less generous.

The Centers for Medicare & Medicaid Services (CMS), the agency that 
administers Medicare, has authority under section 402(a) of the Social 
Security Amendments of 1967 to conduct demonstration programs to test 
methods of payment that have the potential to increase the efficiency 
and economy of Medicare.[Footnote 3] CMS is authorized to waive 
Medicare payment rules under the demonstrations, which may result in 
increased Medicare spending.[Footnote 4] CMS used this statutory 
authority to initiate two demonstration programs that included health 
plans designed to operate under the PPO model.[Footnote 5] In January 
2002, CMS began the M+C Alternative Payment Demonstration, which was 
intended to encourage certain M+C plans to remain in the Medicare 
program. One PPO, Independence Blue Cross, participated in 
2003.[Footnote 6] In January 2003, CMS began a second demonstration, 
known as the Medicare PPO Demonstration, that included 33 plans. The 
goals of this demonstration were to encourage plans to participate in 
the Medicare program under the PPO model, extend beneficiary access to 
private health plans, and provide a health plan option that would 
attract beneficiaries from FFS Medicare and Medigap plans.[Footnote 7] 
Subsequent to the start of these two demonstrations, Congress passed 
the Medicare Prescription Drug, Improvement, and Modernization Act of 
2003 (MMA), which revised Medicare's program for private plans and 
provided for a new PPO component to begin in 2006.

Because the experience gained through the two demonstrations may help 
guide future efforts to incorporate private plans into Medicare, you 
asked us to study the demonstrations' implementation and outcomes as 
they pertain to PPOs. Specifically, we (1) described how CMS used its 
statutory authority to conduct the two demonstrations; (2) assessed the 
extent to which demonstration PPOs expanded access to Medicare health 
plans and attracted enrollees in 2003; (3) compared CMS's estimates of 
the out-of-pocket costs beneficiaries incurred in demonstration PPOs 
with those of other types of coverage, including FFS Medicare, M+C 
plans, and Medigap policies in 2003; and (4) determined the effects of 
demonstration PPOs on Medicare spending prior to the passage of MMA.

To describe CMS's statutory authority to conduct demonstrations, we 
reviewed applicable federal law and regulations. We also solicited the 
agency's views on its interpretation of relevant statutes and 
regulations. Appendix I contains our legal analysis on the Medicare PPO 
Demonstration plans' restriction of enrollees' choice of providers. To 
assess health plan participation and enrollment, we used CMS's monthly 
reports on Medicare's private health plans, as well as historical 
enrollment data for demonstration PPO enrollees provided by CMS's 
Office of Research, Development, and Information (ORDI), which oversees 
the demonstrations. To compare beneficiary out-of-pocket costs between 
demonstration PPOs, M+C plans, Medigap policies, and FFS Medicare in 
2003, we used estimates generated for CMS by Fu Associates, Ltd. (Fu), 
a private firm. CMS includes these estimates on the Medicare Web site, 
www.Medicare.gov, as a tool to help beneficiaries evaluate their 
coverage options. We conducted these comparisons for beneficiaries aged 
65 through 69, the age group most likely to join M+C plans, in the 41 
counties with approximately 90 percent of the enrollment in 
demonstration PPOs. We analyzed the reliability of CMS's enrollment 
data and estimates by conducting interviews with CMS's Office of the 
Actuary (OACT), ORDI, and Fu and determined that the data were 
sufficiently reliable for our purposes. To determine the effects of 
demonstration PPOs on Medicare spending, we used projections developed 
by OACT and conducted interviews with OACT staff. For all four 
objectives, we interviewed ORDI and OACT staff and reviewed relevant 
CMS materials. Appendix II contains a complete description of our 
methodology. We conducted our work from June 2003 through August 2004 
in accordance with generally accepted government auditing standards.

Results in Brief:

Under section 402(b) of the Social Security Amendments of 1967, CMS was 
authorized to waive Medicare payment requirements for health plans 
participating in the two demonstrations, but improperly waived 
requirements unrelated to payment.[Footnote 8] Under this authority, 
CMS offered financial incentives to Independence Blue Cross and the 
plans in the Medicare PPO Demonstration that were not available to M+C 
plans, such as payment rates that could exceed M+C payment rates and 
the opportunity to bear less financial risk by signing risk-sharing 
agreements with CMS. Under these agreements, a portion of the burden 
resulting from unexpectedly high costs, as well as any financial gains 
if costs were unexpectedly low, would be shared by both the plan and 
CMS. In addition, CMS allowed the plans to charge enrollees more in 
cost sharing than would have been permitted in the M+C program. 
However, CMS exceeded its authority with respect to the Medicare PPO 
Demonstration when it tacitly waived plan requirements that were 
unrelated to payment. By law, these plans should have been required to 
cover all services in their benefit packages even if those services 
were obtained from providers outside the plans' provider networks, as 
long as those providers accepted the plans' payment terms and were 
legally authorized to provide the services. However, the agency allowed 
29 of the 33 plans in the Medicare PPO Demonstration to cover some 
services only when they were obtained from providers within their 
networks. In general, beneficiaries who received care from non-network 
providers for these services in these plans were liable for the full 
cost of their care. Examples of such services include skilled nursing 
and home health, which are covered under FFS Medicare, and dental care 
and routine physical examinations, which are not covered under FFS 
Medicare.

Demonstration PPOs attracted relatively few enrollees and did little to 
expand Medicare beneficiaries' access to private health plans in 2003. 
About 98,000, or less than 1 percent, of the 10.1 million eligible 
beneficiaries living in counties where demonstration PPOs operated had 
enrolled in demonstration PPOs by October 2003. Further, although one 
of the goals of the Medicare PPO Demonstration was to attract 
beneficiaries from FFS Medicare and Medigap plans, only 26 percent of 
enrollees in Medicare PPO Demonstration plans came from FFS Medicare, 
with all others coming from M+C plans. About 9.9 million, or 98 
percent, of demonstration PPO enrollees also had M+C plans available in 
their counties. Virtually no enrollment occurred in counties where only 
demonstration PPOs operated. Demonstration PPO enrollment was 
concentrated in two plans: one that existed prior to the launch of the 
demonstrations and another that replaced an HMO previously offered by 
the same organization.[Footnote 9]

According to estimates prepared by CMS, and available on the Medicare 
Web site in 2003, beneficiaries who enrolled in demonstration PPOs 
could have expected to incur total out-of-pocket costs--expenses for 
premiums, cost sharing and noncovered services and items--that were the 
same or higher than those they would have incurred with nearly all 
other types of coverage if they used network providers. On average, the 
expected beneficiary out-of-pocket costs for demonstration PPOs were 
similar to those for Medigap plans F and I --private insurance plans 
that supplement FFS Medicare. Demonstration PPOs were estimated to have 
out-of-pocket costs that were higher than FFS Medicare, M+C HMO, and 
M+C PPO plans, but lower than M+C private fee-for-service (PFFS) plans, 
which in function resemble FFS Medicare but are operated by private 
companies.[Footnote 10] To the extent that beneficiaries in 
demonstration PPOs obtained services from non-network providers, their 
out-of-pocket costs would have been higher than those estimated on the 
Medicare Web site. For example, a six-night stay in a network hospital 
in 2003 was projected to cost a demonstration PPO enrollee an average 
of $421, while the same length of stay in a non-network hospital cost 
an average of $1,223. In addition, demonstration PPOs compared more 
favorably to other types of coverage for beneficiaries in poor health; 
for these beneficiaries, only Medigap plans F and I offered lower 
costs. Further, while demonstration PPOs showed out-of-pocket costs at 
least as high as most other options, CMS estimates suggest that 
demonstration PPOs generally provided better coverage for certain 
benefits, such as prescription drugs and inpatient hospitalization, 
than some other beneficiary options.

While it is too early for CMS to determine the effect of demonstration 
PPOs on Medicare spending, CMS's OACT originally projected that 
demonstration PPOs would increase Medicare spending by $750 per 
enrollee per year for the M+C Alternative Payment Demonstration for 
2002 and 2003 combined, and $652 per enrollee per year in 2003 for the 
Medicare PPO Demonstration. Overall, the demonstration PPOs were 
estimated to increase Medicare spending by about $100 million for 2002 
and 2003 combined. OACT's estimates were based on monthly payments to 
plans for their enrolled beneficiaries and losses CMS might share with 
plans under the risk-sharing agreements. Specifically, OACT projected 
that the PPO plan in the M+C Alternative Payment Demonstration would 
increase spending by about $25.2 million in 2002 and 2003 combined--
$10.1 million due to monthly payments to plans and $15.1 million due to 
CMS's participation in risk-sharing agreements. OACT projected that the 
Medicare PPO Demonstration would increase spending by $75 million in 
2003, due to monthly payments to plans, but that there would be no 
additional spending due to the risk-sharing agreements. Total 
enrollment in demonstration PPOs has been lower than CMS anticipated, 
so actual spending may be less than OACT projected. CMS does not yet 
have data on the actual cost of the demonstrations in 2003.

We recommend that the Administrator of CMS promptly instruct plans in 
the Medicare PPO Demonstration to provide coverage for all plan 
services furnished by any provider authorized to provide Medicare 
services who accepts the plans' terms and conditions of payment.

CMS agreed to implement our recommendation and said it is working with 
the PPO demonstration plans to ensure that they come into compliance 
with the provisions that govern their Medicare participation. CMS 
expressed concern about the tone of the report and believes that the 
demonstrations are worthwhile.

Background:

To reduce out-of-pocket costs that result from cost sharing and the 
utilization of non-Medicare covered services and items, FFS Medicare 
beneficiaries may either purchase a private supplemental insurance 
policy, known as a Medigap plan, or enroll in a private health plan 
that has contracted to serve Medicare beneficiaries. From 1998 through 
2003, M+C, Medicare's private health plan program, allowed 
participation by a variety of plan types, including HMOs, PPOs, and 
PFFS plans, as long as these plans met certain organizational and 
operational requirements. Unlike in the private insurance market, where 
PPO plans were the most prevalent type of health plan, the vast 
majority of M+C plans were HMOs. CMS launched two demonstrations that 
included plans intended to operate under the PPO model.

Beneficiaries Purchase Medigap Plans or Enroll in Private Health Plans:

Beneficiaries in FFS Medicare, which consists of Medicare part A and 
part B, may incur substantial out-of-pocket costs. Part A helps pay for 
inpatient hospital, skilled nursing facility, hospice, and certain home 
health services, although beneficiaries remain liable for a share of 
the cost of most covered services. For example, Medicare requires 
beneficiaries to pay a deductible for each hospital benefit 
period,[Footnote 11] which was $840 in 2003, and covers a maximum of 90 
days per benefit period. [Footnote 12] Medicare part B helps pay for 
selected physician, outpatient hospital, laboratory, and other 
services. Enrollment in part B is voluntary and requires a beneficiary 
to pay a monthly premium and an annual deductible for most types of 
part B services --$58.70 and $100, respectively, in 2003 - and may 
require coinsurance of up to 50 percent for some services. 
Beneficiaries are also liable for items and services not covered by FFS 
Medicare, such as routine physical examinations and most outpatient 
prescription drugs.

Many beneficiaries in FFS obtain more comprehensive coverage through 
supplemental health insurance provided by a former employer or 
purchased from a private insurer (Medigap). Although many employers do 
not offer supplemental health insurance to their retirees, Medigap 
policies are available nationwide. In most states, Medigap policies are 
organized into 10 standardized plans offering varying levels of 
supplemental coverage. Medigap plan F is the plan most widely selected 
by beneficiaries, although it does not offer prescription drug 
coverage; Medigap plan I offers similar coverage but includes some 
coverage for prescription drugs.[Footnote 13] Beneficiaries with 
Medigap policies receive coverage for services from any provider who is 
legally authorized to provide Medicare services.

Most beneficiaries may also obtain more comprehensive coverage by 
choosing to receive Medicare benefits through private health plans that 
participate in Medicare instead of through FFS Medicare. While private 
Medicare health plans are not available nationwide, about 80 percent of 
beneficiaries in 2003 had access to at least one plan within the 
counties where they lived. Beneficiaries who enroll in a private plan 
may pay a monthly premium, in addition to the Medicare part B premium, 
and agree to receive their Medicare-covered benefits, except hospice, 
through the plan. In return, beneficiaries may receive additional non-
Medicare benefits and may be subject to reduced cost sharing for 
Medicare-covered benefits.[Footnote 14] Beneficiaries who enroll in a 
private plan that contracts with Medicare are entitled to coverage for 
all services and items included in the plan's benefit package, 
regardless of whether the service or item is covered under FFS 
Medicare.

M+C Established Definitions and Operational Requirements for 
Participating Plans:

Congress created the M+C program, in part, to expand health plan 
options for beneficiaries. Previously, private plan participation in 
Medicare had been largely limited to HMOs. The M+C program provided the 
opportunity for PPOs and PFFS plans to serve beneficiaries as well. 
Generally, M+C plan types differed by the extent to which they used 
provider networks.[Footnote 15] M+C HMOs were required to maintain 
networks of providers, and they generally covered services furnished 
only by providers in their networks, except in limited circumstances 
such as urgent or emergency situations. (See table 1.) M+C PPOs were 
also required to maintain provider networks. Unlike M+C HMOs, M+C PPOs 
were required to pay for covered services obtained from non-network 
providers, although they could charge beneficiaries additional cost 
sharing for these services. A third type of M+C plan, the PFFS plan, 
was not required to maintain provider networks. Rather, M+C PFFS plans 
were required to pay for all covered services obtained from any 
provider authorized to furnish Medicare-covered services who accepted 
the plan's terms and conditions of payment.

Table 1: Characteristics of HMO, PPO, and PFFS plans under M+C:

Type of M+C Plan: HMO; 
Does plan use provider networks? Yes; 
May plan require beneficiaries to use only network providers? Yes[A].

Type of M+C Plan: PPO; 
Does plan use provider networks? Yes; 
May plan require beneficiaries to use only network providers? No.

Type of M+C Plan: PFFS; 
Does plan use provider networks? No; 
May plan require beneficiaries to use only network providers? N/A.

Source: 42 Code of Federal Regulations Part 422 (2003).

[A] HMOs must cover certain services received from non-network 
providers, such as emergency care or urgent care.

[End of table]

While many M+C requirements were uniform across the different types of 
plans, two categories of requirements varied by the type of plan: those 
that were intended to ensure that enrollees had sufficient and timely 
access to covered services, known as access-to-services requirements, 
and those that were intended to ensure that services furnished were of 
sufficient quality, known as quality assurance requirements. (See table 
2.) In general, plans that restricted enrollees to provider networks 
were subject to more extensive access-to-services and quality assurance 
requirements than those that did not. Accordingly, M+C HMO plans were 
subject to more extensive quality assurance and access-to-services 
requirements than M+C PFFS plans. M+C PPOs were subject to the more 
extensive access-to-services requirements of M+C HMOs, but the less 
extensive quality assurance requirements of PFFS plans.[Footnote 16] 
For example, in order to demonstrate that they provided sufficient 
access to services, M+C HMOs and M+C PPOs were required to monitor and 
document the timeliness of the care their enrollees received from 
providers, while M+C PFFS plans were not required to monitor care in 
this way. With regard to quality assurance, M+C HMOs each year had to 
initiate a multi-year quality improvement project, such as a provider 
or enrollee education program, while M+C PPOs and M+C PFFS plans were 
not subject to this requirement.[Footnote 17]

Table 2: Quality Assurance and Access-to-Services Requirements for M+C 
PPOs and M+C PFFS Plans Compared to M+C HMOs, 2003:

M+C PPOs; 
Quality assurance requirements (compared to M+C HMOs): Less extensive; 
Access-to-services requirements (compared to M+C HMOs): Same.

M+C PFFS plans; 
Quality assurance requirements (compared to M+C HMOs): Less extensive; 
Access-to-services requirements (compared to M+C HMOs): Less extensive.

Source: 42 Code of Federal Regulations Part 422 (2003).

[End of table]

M+C HMOs, M+C PPOs, and M+C PFFS plans all were paid a monthly payment 
per enrollee according to a statutory formula. The M+C payment rate 
varied by county and could be higher or lower than FFS Medicare's per 
capita spending in a county. An M+C plan was at full risk for the costs 
of covered services for its enrollees. If these costs made up a higher 
than anticipated portion of the plan's total revenues--consisting of 
enrollee premiums and monthly payments from CMS--then the plan would 
have less than it anticipated for administration, profit, and other 
contingencies.

Plan Participation in M+C Did Not Reflect the Private Marketplace:

In recent years, PPO plans have become increasingly prevalent in the 
private insurance market and tended to displace other types of plans, 
such as HMOs, that offered less provider choice. From 1996 through 
2002, the percentage of individuals with employer-sponsored coverage 
who were enrolled in HMO plans decreased from 31 percent to 26 percent, 
while the percentage of individuals with employer-sponsored coverage 
enrolled in PPOs increased from 28 percent to 52 percent. In contrast, 
there were approximately 3,000 Medicare beneficiaries enrolled in a 
total of six M+C PPO plans by 2003. From 1998 through 2003, the total 
number of M+C plans, the vast majority of which were HMOs, decreased 
from 346 to 155. The number of beneficiaries covered by M+C plans also 
fell, from 6.1 million in 1998, or about 16 percent of all 
beneficiaries, to 4.6 million in 2003, or about 11 percent of all 
beneficiaries.

CMS Launched Two Demonstrations That Include Plans Designed to Operate 
as PPOs:

Section 402(a) of the Social Security Amendments of 1967 authorizes CMS 
to conduct demonstrations to identify whether changes in methods of 
payment or reimbursement in Medicare and other specified health care 
programs would increase the efficiency and economy of those programs 
without adversely affecting the quality of services. In addition, under 
section 402(b), CMS may waive requirements relating to payment or 
reimbursement for health care services in connection with these 
demonstrations.[Footnote 18] For example, CMS may be able to offer 
demonstration plans alternative methods of payment or other financial 
incentives that are not offered to other providers in the Medicare 
program. However, CMS does not have the authority to waive rules not 
related to payment or reimbursement.

Prior to the passage of MMA, CMS launched both the M+C Alternative 
Payment Demonstration and the Medicare PPO Demonstration. The M+C 
Alternative Payment Demonstration began in 2002 and included one 
organization offering a PPO in 2003. It is set to expire in December 
2004. The Medicare PPO Demonstration, which began in 2003, included 17 
organizations representing 33 plans. This demonstration is set to 
expire in December 2005.

CMS Offered All Demonstration PPOs Financial Incentives, but Improperly 
Allowed Plans in the Medicare PPO Demonstration to Limit Coverage:

Using its authority to waive requirements related to payment and 
reimbursement, CMS offered financial incentives to Independence Blue 
Cross and the plans in the Medicare PPO Demonstration that they did not 
offer to typical M+C plans. These incentives included potentially 
higher payments and the opportunity to reduce their exposure to 
financial risk by entering into risk-sharing agreements. CMS also 
allowed the plans to exceed the limits on the cost sharing that M+C 
plans could charge beneficiaries. Under federal law, plans in the 
Medicare PPO Demonstration should have been required to allow 
beneficiaries to obtain plan services from providers of their choice, 
as long as those providers were legally authorized to furnish them and 
accepted the plans' terms and conditions of payment. CMS did not have 
authority to waive this requirement, as it was unrelated to payment or 
reimbursement. However, CMS improperly allowed 29 of the 33 plans in 
the Medicare PPO Demonstration to require, as a condition of coverage 
for certain services, that beneficiaries obtain those services only 
from network providers.

CMS Used its Statutory Authority to Offer Plans Financial Incentives to 
Participate in Demonstrations:

Under its authority to waive requirements related to payment for 
demonstration participants, CMS offered demonstration PPOs a number of 
financial incentives to participate in the demonstrations. By waiving 
the M+C requirements applicable to plan payment, CMS offered 
Independence Blue Cross and the plans in the Medicare PPO Demonstration 
an opportunity to receive payment rates that could be higher than those 
received by M+C plans. Per enrollee per month,[Footnote 19] 
demonstration PPOs received the higher of the county-based M+C rate or 
a rate based on the average amount Medicare spent in that county for 
each FFS beneficiary.[Footnote 20] A plan's ability to receive the 
higher of the M+C rate or FFS-based rate could substantially increase 
its payment rates, depending on the counties it served. In 44 of the 
214 counties where the plans in the Medicare PPO Demonstration were 
available in 2003, the FFS-based rate ranged from approximately 0.3 
percent to 15.1 percent higher than the M+C payment rate.[Footnote 21] 
For example, in Clark County, Nevada, the FFS-based rate was $635.79, 
or 5.6 percent higher than the M+C payment rate of $599.95.

CMS also used its waiver authority to allow Independence Blue Cross and 
the plans in the Medicare PPO Demonstration to reduce their financial 
risk through risk-sharing agreements. Risk-sharing agreements were not 
available to non-demonstration M+C plans, which were required to accept 
full financial risk for the cost of providing covered services to their 
enrollees. For contract year 2003, CMS signed risk-sharing agreements 
with 13 organizations offering a total of 29 plans. The terms of the 
agreements varied. Each agreement specified an expected "medical loss 
ratio" (MLR), the percentage of a plan's annual revenue (comprised of 
monthly payments from CMS and any enrollee premiums) that would be 
spent on medical expenses. Generally, plans could designate the 
remaining percentage of revenue for administrative expenses, profit, 
and other contingencies. For the 12 organizations in the Medicare PPO 
Demonstration that had risk-sharing agreements with CMS, medical 
expenses represented a median 87 percent of plan revenue. CMS agreed to 
share a designated percentage, negotiated separately with each plan, of 
any difference between the plan's actual MLR and the expected MLR that 
fell outside a range around the expected MLR, known as a risk 
corridor.[Footnote 22] For each plan, the designated percentage with 
which it would share risk with CMS was identical whether the actual MLR 
was greater or lower than the expected MLR.[Footnote 23]

For example, a plan's contract might have specified an MLR of 87 
percent, a percentage of shared risk of 50 percent, and a risk corridor 
of 2 percent above and below the expected MLR (See fig. 1.) If that 
plan's actual medical expenses exceeded 89 percent of its revenue, CMS 
would pay the plan 50 percent of the amount by which the actual MLR 
exceeded 89 percent. If the plan's actual MLR was lower than 85 percent 
of its revenue, the plan would pay CMS 50 percent of the amount that 
the actual MLR fell below 85 percent.

Figure 1: Illustration of a Hypothetical Risk-Sharing Agreement (MLR of 
87 Percent, Shared Risk of 50 Percent, Risk Corridor of 2 Percent):

[See PDF for image]

[End of figure]

In order to allow organizations with HMO licenses to offer PPO-model 
health plans without having to meet the more stringent quality 
assurance requirements of M+C HMOs, CMS had organizations sign PFFS 
contracts and also waived certain M+C payment requirements. Of the 33 
plans in the Medicare PPO Demonstration, 13 were offered by 
organizations with HMO licenses. Under M+C requirements, a PPO offered 
by an organization licensed as an HMO would have to adhere to the more 
stringent quality assurance standards applicable to HMOs.[Footnote 24] 
CMS indicated that it could permit licensed HMOs to establish PPO-type 
networks without being subject to the more stringent quality assurance 
requirements applicable to HMOs by structuring their plans as PFFS 
plans. CMS contracted with all plans participating in the Medicare PPO 
Demonstration as M+C PFFS plans because M+C did not prohibit 
organizations licensed as HMOs from offering PFFS plans.[Footnote 25] 
Although M+C requires PFFS plans to pay each class of provider 
uniformly, CMS waived this payment-related requirement, thereby 
enabling these plans to establish provider networks by paying providers 
differently depending on whether they belonged to their networks.

CMS also waived the M+C limits on beneficiary cost sharing. An M+C plan 
may set beneficiary cost-sharing requirements that differ from those in 
FFS Medicare, but these requirements are subject to statutory limits 
that vary by plan type. For example, under M+C rules for PFFS plans, 
the actuarial value, or estimated dollar value, of the cost-sharing 
requirements for benefits that CMS requires the plans to cover could 
not exceed the actuarial value of cost-sharing requirements in FFS 
Medicare, which was about $1,200 annually per beneficiary in 2003. 
Because CMS waived this provision for demonstration PPOs, these plans 
were subject to no statutory or regulatory cost-sharing limits.

CMS Exceeded Its Authority by Allowing Medicare PPO Demonstration Plans 
to Restrict Coverage for Certain Services to Network Providers:

Because CMS signed PFFS plan contracts with all of the plans in the 
Medicare PPO Demonstration, these plans should have been subject to all 
PFFS plan requirements. In particular, by federal law, M+C PFFS plans 
were required to allow enrollees to receive all covered services from 
any provider who is legally authorized to provide Medicare services and 
accepts the plans' terms and conditions of payment. CMS does not have 
the authority to waive this requirement because it pertains to 
beneficiary access to providers, not payment. However, CMS allowed 29 
of the 33 plans in the Medicare PPO Demonstration to establish provider 
networks and to exclude coverage for some services, both those covered 
and not covered by FFS Medicare, obtained outside the provider 
network.[Footnote 26] Examples of such services include skilled nursing 
and home health, which are covered under FFS Medicare, and dental care 
and routine physical examinations, which are not covered under FFS 
Medicare.

In response to our inquiries, CMS, in a letter dated June 15, 2004, 
agreed with our view that the restriction of Medicare-covered services 
to network providers by plans in the Medicare PPO Demonstration 
violated Medicare requirements. The agency noted, however, that the 
plans did not place such coverage restrictions on most services in 
their benefit packages. In its letter, CMS said that it would instruct 
plans in the Medicare PPO Demonstration to provide out-of-network 
coverage for Medicare-covered services in 2005, if they want to 
continue to operate as PFFS plans and avail themselves of the quality 
assurance requirements available to M+C PFFS plans. However, CMS 
indicated that it would not require plans that cover non-Medicare 
services only in network to provide out-of-network coverage for these 
services.

We maintain that the Medicare PPO Demonstration plans' restriction on 
coverage of services obtained outside their provider networks is 
unlawful. The Social Security Act[Footnote 27] does not distinguish 
between Medicare and non-Medicare-covered services with respect to an 
M+C PFFS plan's obligation to cover plan benefits. According to the 
law, M+C PFFS plans must allow enrollees to obtain all covered plan 
services-both Medicare-covered and non-Medicare-covered--from any 
provider authorized to provide the services who accepts the plans' 
terms of payment.

Furthermore, allowing plans in the Medicare PPO Demonstration to limit 
coverage of certain benefits to network providers is inconsistent with 
statutory and regulatory requirements intended to promote quality of 
care for beneficiaries in M+C plans. Under M+C, PFFS and PPO plans were 
held to less extensive quality assurance requirements than HMOs due, in 
part, to the greater choice these plans' enrollees have in obtaining 
services from providers.[Footnote 28] However, plans in the Medicare 
PPO Demonstration were allowed to restrict beneficiary choice of 
provider for certain services but were not held to the quality 
assurance standards that apply to M+C plans that restrict choice.

Demonstration PPOs Did Little to Expand Health Plan Options, and Have 
Enrolled Relatively Few Beneficiaries:

Demonstration PPOs did little to expand access to private Medicare 
health plans for beneficiaries who lacked such access. In addition, 
they enrolled relatively few beneficiaries, less than 1 percent of 
those living in counties where they operated. Furthermore, 
beneficiaries who enrolled in Medicare PPO Demonstration plans were far 
more likely to have switched from an M+C plan instead of FFS Medicare. 
About 98 percent of the beneficiaries who lived in counties with 
demonstration PPOs had other Medicare private health plans available.

Demonstration PPOs Did Little to Expand Access to Private Health Plans:

Although demonstration PPOs provided beneficiaries with an additional 
plan option in the counties where they operated, they did little to 
attract private health plans to counties where no M+C plans existed. In 
October 2003, demonstration PPOs were available in 214 counties 
nationwide, where approximately 10.1 million beneficiaries 
resided.[Footnote 29] Some form of M+C plan was available in 205 of the 
214 counties. (See table 3.) About 200,000 of the 10.1 million 
beneficiaries, or about 2 percent, lived in the nine counties where 
only demonstration PPOs were available.[Footnote 30] (See fig. 2.)

Table 3: M+C Plan Availability in the 214 Counties Where Demonstration 
PPO Plans Were Available, 2003:

Type of M+C plan: HMO; 
Number of counties: 193; 
Number of eligible Medicare beneficiaries (millions): 9.6.

Type of M+C plan: PPO; 
Number of counties: 23; 
Number of eligible Medicare beneficiaries (millions): 2.4.

Type of M+C plan: PFFS; 
Number of counties: 95; 
Number of eligible Medicare beneficiaries (millions): 3.4.

Type of M+C plan: Any M+C plan; 
Number of counties: 205; 
Number of eligible Medicare beneficiaries (millions): 9.9.

Source: CMS.

Note: The nine counties where only demonstration PPOs were available 
are not represented in this table.

[End of table]

Figure 2: Location of Demonstration PPOs and M+C Plans by County, 2003:

[See PDF for image]

Note: Data are from CMS's Geographic Service Area file (October 2003). 
Although demonstration PPOs and M+C plans were available in the shaded 
counties, these plans may not have been available to all Medicare 
beneficiaries in these counties because plans may serve only part of a 
county. M+C plans include M+C HMOs, M+C PPOs, and M+C PFFS plans.

[End of figure]

Demonstration PPOs Enrolled Relatively Few Beneficiaries:

Enrollment in demonstration PPOs was relatively low. Of the 10.1 
million eligible Medicare beneficiaries living in demonstration PPO 
counties, about 98,000, or less than 1 percent, had enrolled by October 
2003.[Footnote 31] (See table 4.) These 98,000 enrollees represented 
about 5 percent of the total enrollment in Medicare private health 
plans in demonstration PPO counties. Enrollment in demonstration PPOs 
was particularly low in the nine counties with no M+C plans. In these 
counties, only about 100 of the approximately 203,000 beneficiaries 
living there enrolled.

Table 4: Demonstration PPO Market Penetration in Counties With and 
without M+C Plans, October 2003:

Type of county: 205 counties with a demonstration PPO plan and M+C 
plan; 
Number of eligible Medicare benefiaries: 9,900,803; 
Number enrolled in private health plans: M+C plans: 1,870,384; 
Number enrolled in private health plans: Demonstration PPOs: 98,047; 
Percentage of eligible beneficiaries enrolled in private health plans: 
M+C plans: 18.9%; 
Percentage of eligible beneficiaries enrolled in private health plans: 
Demonstration PPOs: 1.0%; 
Percentage of all enrollees in Medicare private health plans in 
demonstration PPOs: 5.0%.

Type of county: 9 counties with only a demonstration PPO plan; 
Number of eligible Medicare benefiaries: 203,535; 
Number enrolled in private health plans: M+C plans: N/A; 
Number enrolled in private health plans: Demonstration PPOs: 117; 
Percentage of eligible beneficiaries enrolled in private health plans: 
M+C plans: N/A; 
Percentage of eligible beneficiaries enrolled in private health plans: 
Demonstration PPOs: 0.1%; 
Percentage of all enrollees in Medicare private health plans in 
demonstration PPOs: 100.0%.

Type of county: 214 (total number of counties with demonstration PPOs); 
Number of eligible Medicare benefiaries: 10,104,338; 
Number enrolled in private health plans: M+C plans: 1,870,384; 
Number enrolled in private health plans: Demonstration PPOs: 98,164; 
Percentage of eligible beneficiaries enrolled in private health plans: 
M+C plans: 18.5%; 
Percentage of eligible beneficiaries enrolled in private health plans: 
Demonstration PPOs: 1.0%; 
Percentage of all enrollees in Medicare private health plans in 
demonstration PPOs: 5.0%. 

Source: CMS.

Note: Data are from CMS's Geographic Services Area file (GSA) for 2003. 
GSA excludes counties with 10 or fewer enrollees in demonstration PPOs, 
resulting in slightly lower enrollment figures. The number of eligible 
Medicare beneficiaries was drawn from the September 2003 GSA, gathered 
quarterly, because it is the nearest estimate of eligible Medicare 
beneficiaries to October 2003.

[End of table]

Two plans, Independence Blue Cross and Horizon Healthcare of New 
Jersey, accounted for more than 70 percent of all demonstration PPO 
enrollment. (See fig. 3.) Of the approximately 98,000 beneficaries 
enrolled in demonstration PPOs, about 23,000, or 23 percent, were 
enrolled in Independence Blue Cross, the one PPO plan in the M+C 
Alternative Payment Demonstration. Approximately 47,000, or about 48 
percent of all demonstration PPO enrollees, were enrolled in Horizon 
Healthcare of New Jersey, a participant in the Medicare PPO 
Demonstration. The approximately 28,000 remaining beneficiaries were 
enrolled in the 32 other plans in the Medicare PPO Demonstration. These 
plans had an average enrollment of 878 beneficiaries.

Figure 3: Percentage of Enrollment in Demonstration PPOs by Plan, 2003:

[See PDF for image]

Note: Data were prepared by GAO from CMS's Geographic Services Area 
file (2003) and the Medicare Health Plan Compare data set (2003).

[End of figure]

The Medicare PPO Demonstration largely did not fulfill CMS's goal of 
attracting beneficiaries from FFS Medicare; most beneficiaries who 
enrolled in demonstration PPOs came from M+C plans. Specifically, in 
the 211 counties where plans participating in the Medicare PPO 
Demonstration were available, 26 percent of beneficiaries who were 
enrolled in Medicare PPO Demonstration plans were formerly enrolled in 
FFS Medicare, while 74 percent of these beneficiaries were formerly 
enrolled in M+C plans.[Footnote 32] In these same counties, 1 percent 
were enrolled in demonstration PPO plans, 81 percent were enrolled in 
FFS Medicare, and approximately 18 percent of Medicare beneficiaries 
were enrolled in M+C plans.

The disproportionately high enrollment in demonstration PPOs by 
previous enrollees in M+C plans is partially attributable to Horizon 
Healthcare of New Jersey, which terminated its M+C HMO plan at the end 
of 2002 and offered a demonstration PPO plan in 2003 in the same 21 
counties where its HMO had operated in 2002. Nearly all 45,000 
beneficiaries who enrolled in the Horizon demonstration plan in the 
beginning of 2003 were previously enrolled in the HMO plan that the 
demonstration plan replaced. However, even when Horizon enrollees are 
excluded from the analysis, 47 percent of enrollees in the other 
Medicare PPO Demonstration plans were previously enrolled in M+C plans.

Demonstration PPOs Left Beneficiaries Exposed to Relatively High Total 
Out-of-pocket Costs, but Offered Slightly Better Coverage for Some 
Benefits:

According to CMS estimates available on the Medicare Web site, an 
average beneficiary aged 65 to 69 enrolled in a demonstration PPO could 
expect to incur $391 per month in health care expenses for premiums, 
cost sharing, and utilization of noncovered items and services. This 
amount was generally similar or higher than the expected out-of-pocket 
costs associated with other types of health care coverage. Excluding 
premiums, or focusing on beneficiaries in poor health, however, 
somewhat changed the pattern of relative cost by type of coverage. To 
the degree that enrollees in demonstration PPO plans obtained services 
from non-network providers, their average out-of-pocket costs would 
have been higher than CMS estimates. Despite the same or higher 
estimated out-of-pocket costs, demonstration PPOs may have offered 
slightly better coverage for certain items and services, such as 
prescription drugs and inpatient hospitalization.

Estimated Out-of-pocket Costs in Demonstration PPOs Were at Least as 
High as Most Other Types of Coverage:

In 41 counties with approximately 90 percent of enrollment in 
demonstration PPOs, beneficiaries in demonstration PPOs who used only 
network providers were estimated to have incurred average monthly out-
of-pocket costs of $391. That amount is similar to what beneficiaries 
with Medigap plans F and I would have incurred, which averaged $405 and 
$397, respectively.[Footnote 33] (See fig. 4.) Enrollees in M+C HMO and 
M+C PPO plans and FFS Medicare were estimated by CMS to have incurred 
lower monthly out-of-pocket costs, averaging $349 and $340, 
respectively. The highest monthly out-of-pocket costs were estimated to 
have been incurred by beneficiaries in M+C PFFS plans, which averaged 
$423 per month.[Footnote 34] Because the reported out-of-pocket costs 
were averages across beneficiaries, the difference among types of plans 
represent the variation in plans' premiums, covered benefits, and cost 
sharing--not the characteristics of enrollees.

Figure 4: Estimated Average Beneficiary Out-of-pocket Health Care Costs 
for Premiums, Cost Sharing, and Noncovered Items and Services per 
Month, by Type of Coverage, 2003:

[See PDF for image]

Note: Data are from CMS's Medicare Personal Plan Finder (2003). 
Analysis is based on data from 41 counties. The FFS Medicare part B 
premium is included in total out-of-pocket costs.

[End of figure]

Excluding Premiums or Focusing on Beneficiaries in Poor Health Changed 
Relative Patterns of Out-of-pocket Costs:

Monthly premiums, which represent a predictable expense, accounted for 
a relatively high percentage (26 percent) of expected out-of-pocket 
costs in demonstration PPOs compared to FFS Medicare and M+C plans. 
Demonstration PPOs had an average monthly premium of $100, which was 
higher than the average premium of M+C plans ($35 for M+C HMOs and PPOs 
and $86 for M+C PFFS plans) and lower than the average premium for the 
two Medigap plans ($139 for plan F and $172 for plan I). [Footnote 35] 
(See fig. 5.) Excluding premiums, out-of-pocket costs in demonstration 
PPOs were somewhat lower than M+C plans, but higher than Medigap plans. 
Specifically, beneficiaries could expect an average of $231 per month 
in demonstration PPOs, $254 in M+C HMOs and M+C PPOs, and $277 in M+C 
PFFS plans. Beneficiaries with Medigap plans F and I could expect 
monthly expenses for cost sharing and noncovered items and services to 
total $205 and $150, respectively.

Figure 5: Estimated Average Beneficiary Premiums and Other Out-of-
pocket Costs by Type of Coverage per Month, Excluding FFS Medicare Part 
B Premium, 2003:

[See PDF for image]

Note: Data are from CMS's Medicare Personal Plan Finder (2003). 
Analysis is based on data from 41 counties. We excluded the FFS 
Medicare part B premium of $58.70 from the measure of other out-of-
pocket costs, and rounding prevents the combination of premiums, other 
out-of-pocket costs, and the part B premium from adding up to the total 
out-of-pocket costs for each plan as displayed in fig. 4.

[End of figure]

Relative out-of-pocket costs for beneficiaries in demonstration PPOs 
also depended on their expected health status.[Footnote 36] For 
beneficiaries expected to be in poor health, demonstration PPOs were 
estimated to be less costly than FFS Medicare, M+C HMOs and M+C PPOs, 
and M+C PFFS plans but more costly than Medigap plans F and I. (See 
fig. 6.) For beneficiaries expected to be in excellent health, 
demonstration PPOs were estimated to be less costly than M+C PFFS plans 
and Medigap plans F and I, but more costly than FFS Medicare and M+C 
HMOs and M+C PPOs.[Footnote 37]

Figure 6: Estimated Beneficiary Out-of-pocket Health Care Costs per 
Month by Type of Coverage and Beneficiary Health Status, 2003:

[See PDF for image]

Note: Data are from CMS's Medicare Personal Plan Finder (2003). 
Analysis is based on data from 41 counties:

[End of figure]

Obtaining Services Outside the Demonstration PPO Network Would Increase 
Beneficiaries' Costs:

To the degree that enrollees in demonstration PPOs obtained services 
from non-network providers, their average out-of-pocket costs would 
have been higher than those reflected on the Medicare Web site. Most 
demonstration PPOs excluded at least one service from coverage if it 
was furnished by non-network providers. When beneficiaries obtained 
services that were covered outside their plans' provider networks, they 
were required to pay more in cost sharing relative to what they would 
have paid for the same services from network providers. Demonstration 
PPOs anticipated that at least some enrollees would obtain covered 
services from non-network providers. According to 2004 estimates 
submitted to CMS by organizations participating in the Medicare PPO 
Demonstration, a median of 11 percent of enrollee medical costs would 
be associated with covered services from non-network providers and thus 
higher cost sharing. For example, a six-night stay in a network 
hospital in 2003 was projected to cost a demonstration PPO enrollee an 
average of $421, while the same length of stay in a non-network 
hospital cost an average of $1,223. Across all services in the Medicare 
benefit package that were covered both within and outside the plans' 
provider networks, the plans projected to CMS that, in 2004, enrollees 
would bear a median of 7 percent of the costs of those services if they 
obtained them from network providers, while they would bear a median of 
15 percent of the costs of those services if they obtained them outside 
the provider networks.

For Certain Benefits, Demonstration PPO Coverage Was Better Than That 
of M+C Plans or FFS Medicare:

Although demonstration PPOs had higher enrollee out-of-pocket costs 
than M+C plans, except M+C PFFS plans, demonstration PPOs tended to 
offer slightly better coverage for some benefits, such as prescription 
drugs and inpatient hospitalization. While all beneficiaries living in 
counties with demonstration PPOs had at least one demonstration PPO 
with a prescription drug benefit operating in their county, only 61 
percent had an M+C HMO or M+C PPO plan with a drug benefit operating in 
their county, and none had an M+C PFFS plan with a drug benefit 
operating in their county.[Footnote 38] In 16 of the 41 counties in our 
sample, at least one demonstration PPO and one M+C HMO or M+C PPO 
offered prescription drug coverage. In these counties, demonstration 
PPOs offered drug coverage that resulted in the same out-of-pocket 
costs for beneficiaries as the drug coverage offered by M+C HMO and M+C 
PPO plans ($167 per month), but higher out-of-pocket costs than the 
drug coverage offered by Medigap plan I ($124 per month).[Footnote 39]

Demonstration PPOs were more likely than M+C HMO and M+C PPO plans to 
cover brand-name drugs in counties where both types of plans offered 
drug coverage. About 47 percent of the demonstration PPOs in our sample 
offered coverage for brand-name drugs, while 37 percent of M+C HMO and 
M+C PPO plans covered brand-name drugs. All demonstration PPOs, M+C 
HMOs, and M+C PPOs offered some coverage for generic drugs in these 
counties. M+C PFFS plans did not offer any drug coverage. Medigap plan 
I did not differentiate between generic and brand-name drug 
coverage.[Footnote 40]

For example, in Hillsborough County, Florida,[Footnote 41] 
beneficiaries could choose between five different plans offering 
prescription drug coverage in 2003; one demonstration PPO, three M+C 
HMO plans, and Medigap plan I. (See table 5.) The demonstration PPO 
provided both generic and brand-name drug coverage and required a $12 
copayment per prescription for generic drugs, a $55 copayment per 
prescription for brand-name drugs, and capped coverage for all drugs at 
$750 annually. None of the M+C HMOs covered brand-name drugs. However, 
two of the M+C HMOs offered unlimited coverage for generic drugs, while 
the third capped coverage at $500 per year. The three M+C HMOs charged 
between $7 and $15 per prescription. Insurers in Hillsborough County 
offered the standard Medigap plan I drug coverage: a $250 annual 
deductible, 50 percent of all costs, and a $1,250 annual limit. Medigap 
plan I does not differentiate between generic and brand-name drugs.

Table 5: Prescription Drug Benefits Offered in Hillsborough County, 
Fla., in 2003, by Type of Coverage:

Type of coverage: Demomstration PPO; 
Annual deductible: None; 
Beneficiary cost sharing for prescription drugs: Generic: $12/
prescription; 
Beneficiary cost sharing for prescription drugs: Brand name: $55/
prescription; 
Cap on prescription drug coverage: All drugs: $750.

Type of coverage: M+C HMO 1; 
Annual deductible: None; 
Beneficiary cost sharing for prescription drugs: Generic: $7-15/
prescription; 
Beneficiary cost sharing for prescription drugs: Brand name: Not 
covered; 
Cap on prescription drug coverage: Generic drugs: no limit; 
Cap on prescription drug coverage: Brand-name drugs: not covered.

Type of coverage: M+C HMO 2; 
Annual deductible: None; 
Beneficiary cost sharing for prescription drugs: Generic: $15/
prescription; 
Beneficiary cost sharing for prescription drugs: Brand name: Not 
covered; 
Cap on prescription drug coverage: Generic drugs: no limit; 
Cap on prescription drug coverage: Brand-name drugs: not covered.

Type of coverage: M+C HMO 3; 
Annual deductible: None; 
Beneficiary cost sharing for prescription drugs: Generic: $15/
prescription; 
Beneficiary cost sharing for prescription drugs: Brand name: Not 
covered; 
Cap on prescription drug coverage: Generic drugs: $500; 
Cap on prescription drug coverage: Brand-name drugs: not covered.

Type of coverage: Medigap plan I; 
Annual deductible: $250; 
Beneficiary cost sharing for prescription drugs: Generic: 50 percent of 
all costs; 
Beneficiary cost sharing for prescription drugs: Brand name: 50 percent 
of all costs; 
Cap on prescription drug coverage: All drugs: $1,250. 

Source: CMS.

Note: Data are from CMS's Medicare Health Plan Compare (2003) and 2003 
guide to "Choosing a Medigap Policy."

[End of table]

Compared to M+C HMOs and M+C PPOs, FFS Medicare, and M+C PFFS plans, 
demonstration PPOs tended to offer lower out-of-pocket costs related to 
inpatient hospitalization. In 2003, a six-night stay in a network 
hospital would have cost enrollees in demonstration PPOs an average of 
$421, while the same six-night stay would have cost enrollees in M+C 
plans and FFS Medicare an average of $636 and $840, 
respectively.[Footnote 42] A six-night hospitalization for an enrollee 
in an M+C PFFS plan would have cost an average of $750. In contrast, 
beneficiaries with either of the two Medigap policies would have paid 
nothing for a six-night hospital stay.

Demonstration PPOs Were Projected to Increase Medicare Spending:

At the time the demonstrations were launched, CMS's OACT projected that 
demonstration PPOs would increase Medicare spending by about $100 
million over 2002 and 2003 combined. Specifically, OACT projected that 
the PPO plan in the M+C Alternative Payment Demonstration would 
increase Medicare spending by a total of $25.2 million over 2002 and 
2003 combined, or $750 per enrollee per year, due to higher plan 
payments and CMS's sharing in the plan's financial risk. The Medicare 
PPO Demonstration was projected to increase Medicare spending by a 
total of $75 million in 2003, or $652 per enrollee per year, due to 
plan payments. The risk-sharing agreements with Medicare Demonstration 
PPO plans were not projected to result in additional Medicare spending. 
CMS does not yet have data on the actual cost of the demonstrations in 
2003.

Medicare Payments to Demonstration PPOs Were Projected by CMS to Result 
in Increased Medicare Spending:

CMS's OACT projected that for 2002 and 2003 additional payments to 
demonstration PPOs would increase Medicare spending.[Footnote 43] 
According to its estimates, an average of 16,800 beneficiaries per 
month would be enrolled in Independence Blue Cross, the PPO in the M+C 
Alternative Payment Demonstration, in 2002 and 2003, and monthly 
payments for these beneficiaries would increase Medicare spending by 
$4.5 million in 2002 and $5.6 million in 2003, or about $300 per 
enrollee per year. OACT projected that plans in the Medicare PPO 
Demonstration would have an average monthly enrollment of 115,000 in 
2003, and that monthly payments to plans for these enrollees would 
increase Medicare spending by $75 million, or about $652 per enrollee 
during the year.

OACT projected that Medicare spending would increase as a result of its 
risk-sharing agreement with Independence Blue Cross. OACT projected 
that the plan's actual MLR would be greater than the MLR the plan 
projected in 2002 and 2003. OACT estimated that Medicare's share of the 
difference between the actual and projected MLR would be $4.8 million 
in 2002 and $10.3 million in 2003, or an average of $450 per enrollee 
per year. In contrast, CMS expected that it would neither save nor 
incur additional expenses from risk-sharing under any of the agreements 
in the Medicare PPO Demonstration, because OACT projected that the 
actual MLR would equal the projected MLR.

Actual Spending Due to Demonstration PPOs Cannot Yet be Determined:

At present, it is too early to determine the actual costs of the 
demonstrations in 2002 and 2003. As of July 2004, risk-sharing 
agreements had not yet been reconciled for any demonstration PPOs. 
During the reconciliation process, plans will report their actual MLRs 
to CMS, and depending on the difference between the expected and actual 
MLR, payment may be made either by the plan to CMS, or by CMS to the 
plan under the terms of the risk-sharing agreement. CMS also has not 
completed a more recent estimate of the cost of the demonstrations, 
which would compare spending for actual enrollment in demonstration 
PPOs with projected spending on enrollment in other M+C plans and FFS 
Medicare if the demonstrations did not exist. Enrollment in 
demonstration PPOs has been different than OACT anticipated, which 
would affect such a comparison. Actual monthly enrollment in 
Independence Blue Cross averaged 21,840 in 2002 and 22,835 in 2003, 
somewhat higher than the estimated average monthly enrollment of 16,800 
in both years. Conversely, enrollment in the Medicare PPO Demonstration 
in 2003 was roughly half of projected enrollment. While OACT estimated 
an average monthly enrollment of 115,000 across all participating plans 
in that demonstration, the actual average monthly enrollment was 
61,738. In addition to differing levels of enrollment, the 
demonstrations also experienced much higher than anticipated enrollment 
by former enrollees of other M+C plans.[Footnote 44]

Conclusions:

CMS initiated two demonstrations to expand the number of Medicare 
health plans operating like PPOs. To encourage participation in the 
demonstrations, CMS used its statutory authority to provide financial 
incentives to plans, such as payment rates that exceeded M+C rates and 
the opportunity to share financial risk with Medicare. CMS also allowed 
plans in the Medicare PPO Demonstration to require, as a condition of 
coverage for certain services, that enrollees obtain care for those 
services only from network providers. However, such a requirement is 
inconsistent with federal law for plans in the demonstration, and CMS 
did not have the authority to allow plans to restrict enrollees' choice 
of providers so long as they were authorized Medicare providers who 
accepted the plans' terms and conditions of payment.

Despite CMS's efforts, demonstration PPOs have not yet proven to be an 
attractive option for beneficiaries or the Medicare program. The plans 
were primarily offered in areas where M+C plans were already available, 
and enrollment has been relatively low, even in the few areas where no 
M+C plans existed. According to the estimates available to 
beneficiaries on the Medicare Web site, enrollees in demonstration PPOs 
could expect out-of-pocket costs that were higher than those they would 
have incurred in FFS Medicare or M+C plans, other than M+C PFFS plans, 
and no less than those they would have incurred with Medigap plans F 
and I. In addition to potentially higher costs for beneficiaries, 
demonstration PPOs may also have resulted in $100 million in higher 
Medicare spending in 2002 and 2003, according to initial CMS estimates.

Recommendation for Executive Action:

We recommend that the Administrator of CMS promptly instruct plans in 
the Medicare PPO Demonstration to provide coverage for all plan 
services furnished by any provider authorized to provide Medicare 
services who accepts the plans' terms and conditions of payment.

Agency Comments and Our Evaluation:

In written comments, CMS agreed to implement our recommendation and 
said it is working to ensure that Medicare PPO Demonstration plans come 
into compliance with the provisions that govern their Medicare 
participation. CMS also expressed general concern about the tone of the 
report and said that beneficiaries benefit from increased access to 
PPOs. The agency stated that lessons learned from the Medicare PPO 
Demonstration will help the agency implement the new Medicare Advantage 
regional PPO plan option in 2006. CMS's specific comments largely fell 
into four areas: the report's focus on initial demonstration outcomes, 
the inclusion of the PPO plan in the M+C Alternative Payment 
Demonstration in the analysis, the methodology and data we used to 
illustrate potential out-of-pocket costs for the options available to 
beneficiaries, and the discussion of our conclusion that CMS exceeded 
its statutory authority with respect to the Medicare PPO Demonstration. 
A summary of CMS's specific comments and our evaluation is provided 
below. The full text of CMS's written comments is reprinted in appendix 
III. The agency also provided technical comments, which we incorporated 
as appropriate.

First, CMS stated that the report, by focusing on the Medicare PPO 
Demonstration's initial outcomes, did not adequately present the 
context and value of the demonstration. CMS said that the demonstration 
is an experiment designed to increase availability of the PPO model in 
the Medicare setting, and that it will provide valuable lessons for 
nationwide implementation of the new Medicare Advantage regional PPO 
component in 2006. Because the demonstration was not intended to be a 
fully developed program, CMS felt that our characterization of 
enrollment as "low" was unwarranted. CMS also stated that the financial 
arrangements developed for this demonstration, such as the risk-sharing 
agreements, were intended to encourage plans to participate, and they 
provide an example of how Medicare can encourage PPOs to enter and 
remain in the new Medicare Advantage program.

We were asked to evaluate the initial experience of demonstration plans 
operating under the PPO model because this experience could help inform 
future efforts to incorporate private plans into Medicare. We state in 
the report that our findings apply only to 2003, the first year of the 
Medicare PPO Demonstration and the second year of the M+C Alternative 
Payment Demonstration. We based our evaluation on enrollment in 
demonstration PPOs, the out-of-pocket costs Medicare beneficiaries 
could expect in demonstration PPOs relative to other types of coverage, 
and the effect of demonstration PPOs on Medicare spending. Overall, we 
found that less than 1 percent of the beneficiaries living in counties 
where demonstration PPOs operated had enrolled in demonstration PPOs, 
that most of the enrollees came from M+C plans, and that demonstration 
PPOs did not offer lower estimated out-of-pocket costs than most other 
types of Medicare coverage, even if beneficiaries obtained services 
only from network providers. PPO plans in the demonstrations could 
receive higher payment rates and be subject to less financial risk, 
relative to M+C plans. We acknowledge that the demonstrations are 
continuing and that CMS has contracted for independent evaluations of 
the demonstrations.

Second, CMS stated that the inclusion of the Independence Blue Cross 
PPO from the M+C Alternative Payment Demonstration, along with the 
plans from the Medicare PPO Demonstration, was potentially confusing 
and did not adequately distinguish the different objectives of the two 
separate demonstrations. According to CMS, the purpose of the M+C 
Alternative Payment Demonstration was simply to prevent health plans 
from leaving the M+C program by offering alternative payment 
arrangements. Furthermore, CMS stated that the demonstration was not 
designed to encourage alternative delivery systems in general or the 
PPO model specifically, and that Independence Blue Cross's status as a 
PPO was irrelevant.

We thought it appropriate to evaluate the Independence Blue Cross plan 
and the Medicare PPO Demonstration plans together because the plan 
types were similar and because the demonstrations were conducted under 
the same statutory authority. Independence Blue Cross and the Medicare 
PPO Demonstration plans all operate under the PPO model, and in that 
sense the plans in the two demonstrations are indistinguishable to 
beneficiaries. While the purposes of the M+C Alternative Payment 
Demonstration and the Medicare PPO Demonstration differed, as our 
report states, CMS used the same statutory authority to conduct both 
demonstrations. This authority permits demonstrations that are designed 
to identify whether changes in methods of payment or reimbursement in 
Medicare would increase the efficiency and economy of the program 
without adversely affecting the quality of services. CMS's 
characterization in its comments of the purpose of the M+C Alternative 
Payment Demonstration appears to be inconsistent with the statutory 
authority.

Third, CMS expressed concerns with the methodology and data we used to 
compare the out-of-pocket costs beneficiaries could expect to incur in 
demonstration PPOs with those they could expect to incur with other 
types of Medicare coverage. In CMS's opinion, our comparison was 
hypothetical because it was based on estimates of enrollees' 
utilization of services, not actual utilization of services, and 
potentially unreliable because it may not account for regional 
variation in health care costs. CMS also stated that our findings for 
Medicare beneficiaries aged 65 to 69 may not be applicable for older 
beneficiaries. Finally, CMS stated that including Horizon Healthcare of 
New Jersey in our analysis may have skewed our calculations because it 
had the largest in-network deductible for inpatient hospital services 
of all demonstration PPOs.

Our out-of-pocket cost comparisons used the same estimates that CMS 
makes available on the Medicare Web site through the Medicare Personal 
Plan Finder (MPPF), which is intended to help beneficiaries compare 
their health coverage options. These estimates, developed by Fu for 
CMS, enabled us to compare out-of-pocket costs among various types of 
coverage for beneficiaries of various ages and health statuses, which 
actual utilization data would not have enabled us to do. Fu developed 
these estimates by applying utilization and spending data from the 
Medicare Current Beneficiary Survey (MCBS), a national sample of 
beneficiaries, to the 2003 benefit packages and premiums offered 
locally by various types of Medicare coverage. Therefore, the estimates 
for all types of coverage were derived consistently. If utilization and 
spending in our sample were higher than the national average, then 
actual out-of-pocket costs would have been higher than those we 
estimated; however, the relative differences between the types of 
coverage--which form the basis for our finding--would be expected to be 
similar. In conducting our comparisons, we sought to capture the 
typical plan options available to all eligible Medicare beneficiaries-
-not only PPO enrollees--residing in areas with demonstration PPOs. To 
capture the typical plan option in these areas, we chose a sample of 41 
counties containing 90 percent of enrollment in demonstration PPOs and 
weighted our calculations by the number of eligible beneficiaries 
residing in each county. Horizon Healthcare of New Jersey remained in 
our analysis because Horizon's demonstration PPO plan was available to 
32 percent of all eligible beneficiaries in these 41 counties in 
December of 2003. We presented results for beneficiaries aged 65 to 69, 
the largest of the six Medicare age groups for which Fu calculated out-
of-pocket cost estimates. We also conducted our comparison on a 
substantially older age group--beneficiaries aged 80 to 84--and found 
similar results.

Fourth, CMS stated that our legal finding--that the agency exceeded its 
authority by allowing plans in the Medicare PPO Demonstration to cover 
certain services only if beneficiaries obtained them from the plans' 
network providers--should be discussed in the context of the 
demonstration's objectives. The agency agreed with our recommendation 
that Medicare PPO Demonstration plan participants be instructed to 
remove impermissible restrictions on enrollees' access to providers for 
all covered plan benefits, and not just those covered under parts A and 
B, but did not provide a date by which the recommendation would be 
fully effectuated. CMS stated, however, that the legal finding needed 
to be viewed in the context of the policies the agency intended to 
advance through the Medicare PPO Demonstration. CMS reiterated many of 
the factors that it believes discouraged the offering of PPO plans in 
the M+C program, and said that the agency wanted to provide flexibility 
in the demonstration in order to facilitate participation by plans. CMS 
indicated that it had taken sufficient measures during the Medicare PPO 
Demonstration qualification process to ensure that all demonstration 
plans provided enrollees with adequate access to network providers for 
all covered services, and all plans were required to offer some out-of-
network coverage. In addition, the agency indicated that all PPO plans 
were required to provide full disclosure to enrollees concerning the 
costs for in-network and out-of-network services.

CMS had already identified for us many of the reasons that led it to 
implement the Medicare PPO Demonstration in the manner in which it did, 
and we included them in this report. The context within which CMS 
believes the legal finding must be placed is not relevant to the issue 
of whether CMS exceeded its authority. The waiver authority at issue is 
limited, and its use must conform to those limits. CMS's reiteration of 
the policy objectives the demonstration was intended to achieve, its 
explanations for why some plans did not cover all plan services out of 
network, and its discussion of the measures that it took to ensure 
adequate access to services and enrollee education are not relevant 
considerations and do not make CMS's actions any less unlawful.

We are sending copies of this report to the Administrator of CMS and 
appropriate congressional committees. The report is available at no 
charge on GAO's Web site at http://www.gao.gov. We will also make 
copies available to others upon request.

If you or your staffs have any questions, please call me at (202) 512-
7119. Another contact and staff acknowledgments are listed in Appendix 
III.

Signed by: 

A. Bruce Steinwald: 
Director, Health Care, Economic and Payment Issues:

[End of section]

Appendix I: Analysis of PPO Demonstration Participants' Restriction on 
Enrollee Choice of Provider:

In 2003, the Centers for Medicare & Medicaid Services (CMS) initiated 
the Medicare Preferred Provider Organization (PPO) Demonstration. To 
facilitate participation in the demonstration, CMS permitted 
organizations participating in the demonstration (demonstration 
participants) to require their enrollees to obtain specified services, 
including services covered by parts A and B of the Medicare program, 
only from "network" providers in order to be covered. As discussed 
below, we believe that CMS's decision to permit demonstration 
participants to restrict enrollees' choice of providers exceeded its 
authority and was, therefore, unlawful.

Background:

The Balanced Budget Act of 1997 (BBA) established a new part C of the 
Medicare program, known as the Medicare+Choice (M+C) program,[Footnote 
45] adding sections 1851 through 1859 to the Social Security Act 
(act).[Footnote 46] Under section 1851(a)(1) of the act, every 
individual entitled to Medicare part A and enrolled under part B may 
elect to receive benefits through either the Medicare fee-for-service 
program or a part C M+C plan, if one is offered where he or she 
lives.[Footnote 47] In general, M+C organizations[Footnote 48] must 
provide coverage for all services that are covered under parts A and B 
of Medicare.[Footnote 49] M+C organizations also may include coverage 
for other health care services that are not covered under parts A and B 
of Medicare.[Footnote 50] They may satisfy their coverage obligations 
by furnishing services themselves, arranging for enrollees to receive 
services through contracts with providers, or by reimbursing providers 
who furnish services to enrollees.[Footnote 51]

Section 1851(a)(2) of the act authorizes several types of M+C plans, 
two of which are relevant to the Medicare PPO Demonstration: 
"coordinated care plans" and "private fee-for-service plans." M+C 
coordinated care plans include health maintenance organization (HMO) 
plans, with or without point of service options, and PPO plans. As 
defined by CMS, coordinated care plans have a CMS-approved network of 
providers under contract or arrangement with the M+C organization to 
deliver health care to enrollees.[Footnote 52] M+C organizations 
offering coordinated care plans may specify the network of providers 
from whom enrollees may receive services if they demonstrate that all 
covered services are available and accessible under the plan.[Footnote 
53] Unlike most other coordinated care plans, PPO plans must provide 
coverage for all covered benefits out of network.[Footnote 54] 
Generally, PPO plans require enrollees to pay additional costs for 
services furnished by providers outside the network.

Section 1859(b)(2) of the Social Security Act defines the term "private 
fee-for-service plan" for purposes of the M+C program. As defined, 
private fee-for-service plans are required to reimburse hospitals and 
other providers on a fee-for-service basis without placing the 
providers at financial risk.[Footnote 55] These plans may not vary the 
amounts paid based on the number or volume of services they 
provide.[Footnote 56] Moreover, in contrast to coordinated care plans, 
private fee-for-service plans are not required to have networks of 
providers; instead they must allow enrollees to obtain covered services 
from any provider who is lawfully authorized to provide them and who 
agrees to the terms and conditions of payment, regardless of whether 
the provider has a written contract with the plan to furnish services 
to enrollees.[Footnote 57]

While many of the statutory and regulatory requirements governing M+C 
plans are similar, others vary by plan type. M+C organizations 
generally must be licensed as "risk-bearing entities" by the states 
where they offer M+C plans.[Footnote 58] HMO plans and most other 
coordinated care plans, however, are subject to more stringent quality 
assurance requirements than PPO and private fee-for-service 
plans.[Footnote 59] For example, HMO plans are required by statute to 
implement programs to improve quality and assess the effectiveness of 
such programs through systemic follow-up and to make information on 
quality and outcomes measures available to beneficiaries to facilitate 
comparisons among health care options. These requirements do not apply 
to private fee-for-service and PPO plans. HMO plans, as well as other 
coordinated care plans, are also held to more extensive access 
requirements than private fee-for-service plans to ensure timely access 
to care.[Footnote 60] Finally, although M+C PPO plans generally are 
held to less stringent quality assurance standards than other 
coordinated care plans, M+C organizations licensed as HMOs that offer 
M+C PPO plans may not avail themselves of the less stringent quality 
assurance standards applicable to M+C PPOs. Instead, a licensed HMO 
that offers an M+C PPO plan must comply with the quality assurance 
standards applicable to HMOs.[Footnote 61]

CMS is authorized by section 402(a)(1)(A) of the Social Security 
Amendments of 1967 to conduct demonstrations designed to test whether 
changes in methods of payment or reimbursement in Medicare and other 
specified health care programs would increase the efficiency and 
economy of those programs without adversely affecting the quality of 
services.[Footnote 62] Section 402(b) authorizes CMS to waive 
requirements related to payment or reimbursement for providers, 
services, and other items for purposes of demonstration projects, but 
does not authorize the agency to waive requirements unrelated to 
payment or reimbursement.[Footnote 63] Section 402(b) also authorizes 
CMS to pay costs in excess of those that would ordinarily be payable or 
reimbursable, to the extent that the waiver applies to these excess 
costs.[Footnote 64]

According to CMS, the agency initiated the 3-year Medicare PPO 
Demonstration in January 2003 to make the PPO health care option, which 
had been found to be successful in non-Medicare markets, more widely 
available to Medicare beneficiaries.[Footnote 65] Its objective was to 
introduce more variety into the M+C program so that Medicare 
beneficiaries would have more options available to them.[Footnote 66] 
In addition, CMS believed that the PPO demonstration plans would 
introduce incentives that would result in more efficient and cost-
effective use of medical services.[Footnote 67]

CMS entered into contracts with all demonstration participants. To 
facilitate HMO participation in the Medicare PPO Demonstration, CMS 
permitted licensed HMOs, as well as all other demonstration 
participants, to offer private fee-for-service plans.[Footnote 68] 
Exercising its authority under section 402(b), CMS waived statutory and 
regulatory payment requirements applicable to private fee-for-service 
plans, allowing the participating organizations to vary the amount of 
payments among providers, among other things, so that the plans offered 
would more closely resemble PPO plans. As a result, M+C organizations 
with HMO licenses were able to establish PPO-type plans and were not 
subject to the more stringent quality assurance standards applicable to 
HMOs and most other coordinated care plans. The private fee-for-service 
plan model contract provided that requirements that were not expressly 
waived by CMS would remain in effect during the term of the contract. 
Nevertheless, CMS approved plan provisions that required enrollees to 
obtain various items and services, including those covered under parts 
A and B of Medicare, from "network" providers. CMS officials told us 
that prospective demonstration participants had expressed concerns 
about their ability to determine appropriate payment rates for 
providers who were not under contract with the demonstration 
participant, and that the agency had decided to afford demonstration 
participants flexibility in this area in order to get the demonstration 
project underway. CMS officials also indicated that they had encouraged 
the demonstration participants to cover all benefits "out of network" 
before the end of the demonstration period. Notably, guidance issued by 
CMS to assist M+C organizations, including demonstration participants, 
in developing plan brochures for 2004 contained specific instructions 
for demonstration participants to indicate in their brochures if they 
do not cover all Medicare benefits "out of network."

Discussion:

The Social Security Act places restrictions on private fee-for-service 
plans' authority to limit enrollees' selection of providers. 
Specifically, section 1852(d)(4) requires an organization offering an 
M+C private fee-for-service plan to demonstrate that the plan affords 
sufficient access to health care providers by showing that it has 
established payment rates that are no lower than the corresponding 
rates under the Medicare fee-for-service program or that it has 
contracts with a sufficient number of providers to provide covered 
services, or both.[Footnote 69] That section also provides that the 
access standards may not be used to restrict the persons from whom 
enrollees may obtain covered services, thus suggesting that private 
fee-for-service plans are not authorized to limit their enrollees' 
selection of providers, for example, to those within an established 
"network." The definition of the term "private fee-for-service plan" at 
section 1859(b)(2) echoes this provision, stating that such plans do 
not restrict the selection of providers from among those who may 
lawfully provide covered services and agree to accept the terms and 
conditions of payment.

CMS has recognized that private fee-for-service plans may not restrict 
enrollees to specified providers for covered services. In promulgating 
regulations to implement section 1852(d)(4), CMS interpreted that 
provision as requiring private fee-for-service plans to permit 
enrollees to receive covered services from any provider who is 
authorized to provide services under the Medicare fee-for-service 
program. [Footnote 70] Explaining the new regulatory provision--headed 
by the caption "freedom of choice"--CMS stated:

In 42 C.F.R. § 422.114(b), we specify that the plan must permit the 
enrollees to receive services from any provider that is authorized to 
provide the service under original Medicare. This implements that part 
of section 1852(d)(4) that says that the access requirements cannot be 
construed as restricting the persons from whom enrollees of the M+C 
private fee-for-service plan may obtain covered services.[Footnote 71]

In light of the statutory language and CMS's interpretation, we 
conclude that Medicare PPO Demonstration plan provisions limiting 
enrollees to "network" providers are inconsistent with sections 
1852(d)(4) and 1859(b)(2) of the act. Because these sections are 
unrelated to payment, CMS was not authorized to waive them in 
connection with the Medicare PPO Demonstration.

Further, the plans' exclusions of coverage for services furnished by 
"non-network" providers are incompatible with statutory requirements 
designed to ensure quality of care to enrollees in M+C plans. As 
discussed earlier, private fee-for-service and PPO plans participating 
in the M+C program are held to less stringent quality assurance 
standards than HMOs and certain other coordinated care plans. The 
applicability of less stringent quality assurance standards is due, in 
part, to the increased choices enrollees in private fee-for-service and 
PPO plans have in comparison to enrollees in most other types of plans. 
CMS has expressly recognized this rationale for the distinction among 
various types of plans. In connection with an M+C rulemaking on the 
matter, CMS responded to a concern that private fee-for-service plan 
quality assurance requirements were inadequate to protect enrollees by 
explaining that quality assurance standards may not be as important in 
the case of private fee-for-service plans "in which the enrollee has 
complete freedom of choice to use any provider in the country, and is 
not limited to a defined network of providers."[Footnote 72] CMS's 
approval of restrictions on enrollee choice and simultaneous failure to 
apply the more stringent quality standards applicable to HMO and most 
other coordinated care plans were inconsistent with the statutory 
framework under which M+C plans are required to operate.

Moreover, while CMS stated that the demonstration was intended to offer 
beneficiaries greater choice by encouraging the availability of PPO-
type plans, regulatory provisions applicable to M+C PPO plans would 
have precluded demonstration participants from requiring enrollees to 
obtain services only from "network" providers as a condition of 
coverage. CMS has defined a PPO plan, in part, as a plan that "provides 
for reimbursement for all covered benefits regardless of whether the 
benefits are provided within the network of providers."[Footnote 73] 
(Emphasis added). Since this regulatory provision is not related to 
payment or reimbursement, section 402(b) of the Social Security 
Amendments of 1967 would not have authorized CMS to waive it in 
connection with the Medicare PPO Demonstration.

In its written response to our inquiry about the demonstration, CMS 
indicated that the demonstration plans' conditioning coverage of 
"Medicare-covered services" (those services covered under parts A & B 
of Medicare) on their being furnished by "network" providers violates 
statutory access requirements applicable to private fee-for-service 
plans. CMS explained, however, that while it had reviewed all plans to 
ensure that services covered by parts A and B of the Medicare program 
were covered "in network," some organizations had indicated that they 
were unable to cover certain services "out of network" because of the 
complexities associated with determining payment for "out-of-network" 
providers. CMS, nevertheless, believed that "the basic principle of 
out-of-network access was satisfied" because "the demonstration 
products offer access to most Medicare-covered services." CMS also 
denied that it had waived applicable access requirements, stating that 
it did not have the authority to do so.

CMS indicated that it will instruct demonstration participants that 
they must provide out-of-network coverage for all "Medicare-covered 
services" in 2005, the third year of the Medicare PPO Demonstration, if 
they wish to continue to avail themselves of the quality assurance 
standards applicable to private fee-for-service plans. CMS also 
indicated, however, that it will not require plans to provide out-of-
network coverage for other covered benefits for which the demonstration 
plans provide only in-network coverage. CMS did not provide a legal 
basis for distinguishing between Medicare-covered services and other 
plan services with respect to a demonstration plan's obligation to 
provide "out-of-network" coverage.

We disagree with CMS's assertion that it did not waive the statutory 
requirements at issue. CMS knowingly permitted organizations 
participating in the demonstration to operate in a manner that was 
inconsistent with sections 1852(d)(4) and 1859(b)(2) of the Social 
Security Act. The agency's decision to do so achieved a result for 
demonstration participants that CMS acknowledges it did not have the 
authority to provide. Therefore, we view CMS's action as tantamount to 
a waiver.

We also conclude that all benefits covered under a PPO demonstration 
plan, not just services covered under parts A and B, must be covered 
"out of network" by demonstration plans. The Social Security Act 
defines a private-fee-for service plan, in part, as a "Medicare+Choice 
plan" that "does not restrict the selection of providers among those 
who are lawfully authorized to provide the covered services."[Footnote 
74] A "Medicare+Choice plan," for purposes of the definition of a 
private fee-for-service plan, is defined, in part, as "health benefits 
coverage offered under a policy, contract, or plan by a Medicare+Choice 
organization."[Footnote 75] Furthermore, CMS guidance also provides 
that enrollees in M+C private fee-for-service plans can obtain "plan 
covered health care services from any entity that is authorized to 
provide services under parts A and B and who is willing to accept the 
plan's terms and conditions of payment."[Footnote 76] The act, 
therefore, does not distinguish between Medicare covered services and 
other covered services in specifying the private fee-for-service plan's 
obligations to cover plan benefits.

Section 402(b) of the Social Security Amendments of 1967 provides CMS 
with waiver authority, but also limits that authority by providing that 
the agency may only waive requirements related to payment or 
reimbursement. In connection with the Medicare PPO Demonstration, CMS 
overrode the limitation contained in section 402(b), tacitly waiving 
statutory provisions unrelated to payment. As a general matter, 
agencies may not override statutory limitations on their activities by 
administrative action.[Footnote 77] Therefore, we conclude that CMS's 
decision to allow demonstration participants to restrict enrollees' 
access to providers for any services covered by the plans exceeds its 
authority and is, therefore, unlawful.

[End of section]

Appendix II: Scope and Methodology:

This appendix provides additional information on the key aspects of our 
analysis. First, it describes the Centers for Medicare & Medicaid 
Services' (CMS) administrative data sources we used to assess 
demonstration preferred provider organization (PPO) enrollment and plan 
participation. Second, it describes the CMS data sources we used to 
compare estimated beneficiary out-of-pocket costs between six types of 
coverage. Third, it describes CMS data sources used to compare 2003 
benefits between the six types of coverage. Fourth, it describes CMS 
data we used to estimate the proportion of expected 2004 annual out-of-
pocket costs and cost sharing when demonstration PPO enrollees utilize 
services outside of plan provider networks. Fifth, it describes how CMS 
estimated the effect of demonstration PPOs on Medicare spending. 
Finally, it addresses data reliability issues and limitations.

Enrollment:

We used the following CMS administrative data sets to identify the 
number of eligible Medicare beneficiaries and enrollment by health plan 
in each county where demonstration PPOs operated: the Geographic 
Service Area (GSA) file for October 2003, the Medicare Managed Care 
Plan Monthly Report for October 2003, and the Medicare Managed Care 
Contract (MMCC) report of 2003.[Footnote 78] Because the focus of our 
analysis was on plans available to Medicare beneficiaries at large, we 
used plan enrollment data from GSA to exclude demonstration PPO and 
Medicare+Choice (M+C) plans that were employer-only plans; cost plans; 
and demonstration plans only available to specific beneficiaries such 
as Medicare dual-eligibles.[Footnote 79] Demonstration PPO and M+C plan 
county data from GSA were also used to construct our county-level U.S. 
map.

Estimated Out-of-pocket Cost Comparisons:

To compare out-of-pocket costs for beneficiaries, we used 
administrative data from GSA and CMS's 2003 Medicare Health Plan 
Compare (MHPC) data set[Footnote 80] to identify private health plans. 
For each plan in each county, we then used CMS's 2003 Medicare Personal 
Plan Finder (MPPF)[Footnote 81] to obtain estimated monthly out-of-
pocket costs. We then averaged these costs across counties for 
enrollees in demonstration PPOs, M+C health maintenance organizations 
(HMO) and M+C PPOs, M+C private fee-for-service (PFFS) plans, Medigap 
plans F and I, and fee-for-service (FFS) Medicare.[Footnote 82]

First, we used data from MHPC to identify one plan offered by each 
organization in each county where demonstration PPOs were available. 
Because organizations may offer numerous options for each plan, each 
with its own benefit package and premium, we selected the one option 
that was most favorable for beneficiaries in each service 
area.[Footnote 83] Selecting one option for each plan may have resulted 
in underestimated actual beneficiary out-of-pocket costs for 
beneficiaries in some health plans. In addition, we established a 
sample group of 41 counties containing approximately 90 percent of all 
demonstration PPO enrollment. This sample group includes the 21 
counties where Horizon Healthcare of New Jersey's demonstration PPO 
plan was available,[Footnote 84] and the 23 counties that made up 80 
percent of enrollment in demonstration PPOs other than 
Horizon.[Footnote 85]

Next, we used estimated beneficiary out-of-pocket cost data from CMS's 
MPPF to calculate the 2003 average monthly out-of-pocket costs for 
enrollees in demonstration PPOs and the other types of coverage. CMS, 
and its contractor Fu Associates, Ltd. (Fu), estimated all costs 
related to covered and noncovered benefits when an enrollee utilizes 
services within the plan's network of providers.[Footnote 86] We 
calculated average monthly out-of-pocket costs for beneficiaries aged 
65 to 69 for each type of coverage, in each county, and across all 
health statuses.[Footnote 87] We weighted the estimates of 
demonstration PPOs, M+C HMO and M+C PPO plans, M+C PFFS plans, and 
Medigap plans F and I by the distribution of health statuses of the 
beneficiary cohorts used to create Fu's estimates, and the number of 
eligible Medicare beneficiaries in each county. We separated M+C PFFS 
plans from M+C HMOs and PPOs, because the out-of-pocket costs of 
enrollees in M+C PFFS plans tended to be substantially higher than the 
other two types of M+C plans.

Benefits Comparison:

We used CMS's 2003 MHPC administrative data set in conjunction with 
CMS's 2003 guide to "Choosing a Medigap Policy" to compare the benefit 
packages for enrollees in demonstration PPOs, M+C HMOs and M+C PPOs, 
M+C PFFS plans, Medigap plans F and I, and traditional FFS 
Medicare.[Footnote 88] We compared prescription drug coverage and 
inpatient hospital services for each type of coverage using our sample 
of plans in 41 counties. We selected the one plan option for each plan 
that appeared most favorable to beneficiaries. We also compared 
prescription drug coverage between these types of plans in a sample of 
16 counties where at least one demonstration PPO and one M+C plan 
offered prescription drug coverage as a part of their benefit package. 
In addition, data from CMS's Health Plan Management System 
(HPMS)[Footnote 89] were used to compare the non-network benefits 
offered by each demonstration PPO to the 2003 network benefits offered 
by demonstration PPOs.

Costs of Out-of-network Services:

CMS's Office of the Actuary (OACT), which projects trends in Medicare 
spending, provided the data we used to compare the proportion of 
expected 2004 gross annual out-of-pocket costs and cost sharing when 
demonstration PPO enrollees utilize services inside and outside of plan 
provider networks. The data we obtained were submitted by plans to OACT 
as part of their annual revenue and medical expense projections and 
contained estimates of per member per month gross medical costs and 
target medical loss ratio (MLR) for 2004.[Footnote 90] We contacted 
OACT to verify that we possessed a submission for each of the 20 
demonstration PPOs in our sample of 41 counties.

Effect on Medicare Spending:

To determine the effects of demonstration PPOs on Medicare spending, we 
used projections developed by OACT and conducted interviews with OACT 
staff. To arrive at these projections, OACT compared how much Medicare 
would pay demonstration PPOs per enrollee with the amount Medicare 
would spend on those beneficiaries if the demonstration did not exist 
and those beneficiaries were instead enrolled in M+C plans or FFS 
Medicare. OACT also estimated the effect that risk-sharing agreements 
signed between CMS and demonstration PPOs had on Medicare spending.

Data Reliability:

We used a variety of CMS data sources in our analysis; October 2003 GSA 
file, October 2003 Monthly Report, October 2003 MMCC, 2003 MPPF, 2003 
HPMS, 2003 MHPC, and the estimated 2004 Medicare PPO Demonstration plan 
medical cost files. In each case, we determined that the data were 
sufficiently reliable for our purposes in addressing the report's 
objectives.

We verified the reliability of the administrative data we used to 
determine enrollment figures--CMS's GSA, M+C Monthly Report, and MMCC-
-by comparing the list of unique demonstration PPO contract 
identification numbers and organization names to CMS's list of 
participating demonstration PPO plans and organizations. We did not 
find any discrepancies between the two lists. We worked closely with 
CMS staff and Fu to verify the validity of out-of-pocket cost estimates 
from the 2003 MPPF. We verified that the results of our out-of-pocket 
cost analysis were consistent with CMS's initial tests of its own data, 
and that our methodology, in conjunction with its methodology, did not 
introduce bias. In addition, we worked with CMS to verify the validity 
of the 2004 Medicare PPO Demonstration plan medical cost files 
submitted by the health care organizations by assuring that the 
information they provided to us corresponded with our data for the 
sample of 41 counties.

We identified three potential limitations of our analysis; however we 
have addressed these limitations through conversations with CMS and Fu, 
and by using the best available data. Our report focuses on the results 
of our analysis of estimated enrollee out-or-pocket costs for 
beneficiaries aged 65 to 69. We also obtained similar results when we 
analyzed estimated enrollee out-of-pocket costs for beneficiaries aged 
80 to 84. In addition, we verified with CMS and Fu that the trends 
associated with the 2003 out-of-pocket costs of the 65 to 69 age group 
were similar to the out-of-pocket costs of Medicare beneficiaries aged 
70 to 74. Second, for our out-of-pocket cost analysis, we used national 
FFS Medicare estimates, rather than county-level estimates, because 
county-level estimates were not available. Based on our conversations 
with CMS and Fu, we believe that CMS's national figures were more 
accurate than adjusting the national estimates to the county level 
using national FFS spending in each county. Third, while county-level 
Medigap out-of-pocket costs and benefit package information were not 
available to us, we used CMS estimates of national Medigap out-of-
pocket costs and standardized national Medigap benefits descriptions 
for our benefits comparison.

[End of section]

Appendix III: CMS Comments:

DEPARTMENT OF HEALTH & HUMAN SERVICES: 
Centers for Medicare & Medicaid Services:
Administrator: 
Washington, DC 20201:

DATE: SEP 8 2004:

TO: A. Bruce Steinwald:
Director, Health Care - Economic and Payment Issues: 
Government Accountability Office:

FROM: Mark B. McClellan, M.D., Ph.D.: 
Administrator:

Signed by: Mark B. McClellan: 

SUBJECT: Government Accountability Office (GAO) Draft Report: MEDICARE 
DEMONSTRATION PPOs: Financial and Other Advantages for Plans, Few 
Advantages for Beneficiaries (GAO-04-960):

Thank you for the opportunity to review and comment on the above-
referenced draft report.

We are extremely concerned about the tone of this report, however, 
beginning with the title, which is misleading, unfair, and not 
supported by the information presented in the report itself. We believe 
beneficiaries benefit substantially from increased access to the 
Preferred Provider Organization (PPO) model, and will, in the long run, 
be better off for the availability of these types of plans in the new 
Medicare Advantage program.

It is important to note that the PPO Demonstration is just that, a 
demonstration. The financial arrangements developed for the project 
were intended to encourage plans to participate in this experiment, and 
they should he viewed as a first step in the development of better ways 
to establish partnerships between the Medicare program and private 
plans for the benefit of Medicare beneficiaries.

We have learned a great deal from conducting this demonstration, and we 
look forward to applying these valuable lessons in implementing the new 
Medicare Advantage regional PPO plan option in 2006 as a choice for 
Medicare beneficiaries throughout the country. In addition to a risk-
sharing model similar to the one used for some plans in the PPO 
Demonstration, the Medicare Prescription Drug, Improvement, and 
Modernization Act of 2003 (MMA) created a stabilization fund that will 
encourage plans to enter and remain in the Medicare Advantage regions. 
The MMA also allocated funding to help plans contract with hospitals 
necessary to provide beneficiaries with a full range of care in their 
area. In addition, there are certain beneficiary protections available 
to regional PPOs enrollees that are not available to local plan 
enrollees.

The Centers for Medicare & Medicaid Services (CMS) is working with the 
PPO Demonstration plans to ensure that they come into compliance with 
the provisions that govern their participation in the Medicare program. 
Because these plans are under a private fee-for-service contract for 
purposes of this demonstration, they are required to provide coverage 
for all plan services furnished by any Medicare authorized provider who 
accepts the plan's terms and conditions of payment. It should be noted 
that we have received no indication that the current benefits offered 
by these plans has disadvantaged their enrollees. We will comply with 
the recommendation made in the report while working to ensure a smooth 
transition for enrollees and plans.

The report, overall, fails to adequately reflect the context in which 
the PPO Demonstration was developed and the reasons for its 
implementation. The CMS began to consider a PPO demonstration 
initiative in 2001, three years before passage of the MMA, and before 
many of the current policies and procedures pertaining to PPOs had been 
formulated.

At the time the demonstration was developed, beneficiaries had a very 
limited choice of managed care delivery models. The PPO model, popular 
in the private sector, was available only on a limited basis to 
Medicare beneficiaries under the Medicare + Choice program.

Three years later, the demonstrations have clearly provided valuable 
information to both CMS and health care providers as we work to make 
the PPO model a central feature of the new Medicare Advantage program.

We wanted the PPO Demonstration model to include three fundamental 
components that we thought would appeal to beneficiaries. The first was 
a robust network of preferred providers that would be available to 
provide traditional Medicare benefits as well as some additional 
benefits, such as prescription drugs, with modest cost-sharing. The 
second was the ability of enrollees to use services from non-preferred 
providers, with the understanding that there may be significant cost 
sharing. The third was limits on out-of-pocket cost-sharing when 
possible. All of these features are common in private sector PPOs.

In order to provide these features within the structural constraints of 
the existing Medicare + Choice program, we implemented the model using 
a Medicare + Choice private fee-for-service contract vehicle, which 
would allow both the program and the participating plans as much 
flexibility as possible. We also considered the constraints and 
limitations faced by the organizations that volunteered to test the new 
Medicare PPO model. The entire developmental process occurred over a 
four-month period in order for new PPO options to be available for 
Medicare's 2003 open enrollment period.

In this expedited process, we considered what organizational 
requirements would be the most appropriate for the program, health 
plans, and beneficiaries. One factor that had to be confronted was that 
many of the organizations interested in testing the PPO model had 
existing HMO licenses in their States, which would have prohibited them 
from offering formal PPOs. We also wanted the demonstration 
organizations to be able to accurately pay claim costs incurred by 
enrollees who accessed services through non-preferred providers and 
protect them from incurring any costs other than the stated out-of-
pocket cost associated with the particular service. We created a PPO 
demonstration that was not exactly like any of the existing Medicare + 
Choice coordinated care products, yet provided positive incentives to 
beneficiaries, assured quality of care, and served their interests 
well.

There are multiple references throughout the report to the low 
enrollment levels in the PPO Demonstration products. We believe that 
this misses the point of demonstration projects. The primary objective 
of this initiative was to test ways to develop and implement additional 
coverage options for Medicare beneficiaries. It was not-nor was it 
intended to be-a fully developed program.

That said, about 105,000 Medicare beneficiaries are enrolled in the 
demonstration sites as of August 2004, and the number of enrollees is 
growing steadily. We expect that as more PPOs enter the Medicare 
program because of the new Medicare Advantage provisions of the MMA, 
more beneficiaries will enroll in these plans.

A primary reason for conducting demonstrations is to learn about how 
best to design, develop, and implement potential improvements in the 
Medicare program. As this demonstration has progressed, we have learned 
a great deal that can be applied in the future. In particular, as the 
new Medicare Advantage program has been fleshed out. CMS has been able 
to use information from the demonstration experience to address issues 
related to payment mechanisms, the dissemination of information to 
beneficiaries, and quality assurance. We anticipate that all of the PPO 
Demonstration sites will continue to participate in the Medicare 
Advantage program in 2006, and we believe the experience gained from 
their participation in this demonstration will ensure that they will 
continue to offer beneficiaries attractive options under the Medicare 
Advantage program.

We also must point out that your report creates potential confusion by 
combining analysis of the Independence Blue Cross PPO product with the 
Medicare PPO Demonstration. The Independence plan was one of nine 
products and the only PPO product offered as part of the M+C 
Alternative Payment Demonstration. In discussions with GAO, CMS staff 
repeatedly suggested that the Independence PPO plan be considered 
separately from the PPO products offered under the Medicare PPO 
Demonstration. The report combines the Independence discussion with 
that of the PPO Demonstration products in discussions of cost estimates 
and enrollment, but we do not believe the report adequately 
distinguishes the unique objectives of each demonstration initiative.

The intent of the Medicare + Choice Alternative Payment Demonstration 
was to prevent health plans from leaving the Medicare + Choice program. 
Unlike the Medicare PPO Demonstration, there was never any intent to 
explore or encourage alternative delivery systems in general or the PPO 
model specifically as part of the M+C Alternative Payment 
Demonstration. Rather, this initiative was simply an attempt to try 
alternative payment options to prevent plan withdrawal. The fact that 
Independence was a PPO plan is completely irrelevant. Therefore, we 
would again suggest that the Independence PPO plan be discussed 
separately in the report. (We would note that, unless otherwise stated, 
comments in this response are specific to the Medicare PPO 
Demonstration only, not the independence Medicare + Choice Alternative 
Payment Demonstration.)

The key objective in all Medicare demonstration initiatives is to learn 
from different approaches employed in the demonstration model. The 
report does not acknowledge that an independent evaluation of the 
demonstration is being conducted by Research Triangle Institute, Inc. 
This evaluation includes a survey component to assess why beneficiaries 
are or are not attracted to the products offered under the 
demonstration. The survey will provide valuable information with regard 
to: beneficiary awareness and understanding of the PPO option; specific 
reasons for enrollment; experience and overall satisfaction with the 
plan; and reasons for disenrollment.

Thank you again for the opportunity to review and comment on the draft 
report. Attached are more specific comments on the report and its 
contents.

Attachment:

Centers for Medicare & Medicaid Services' Comments to the Draft GAO 
Report: MEDICARE DEMONSTRATION PPOs: Financial and Other Advantages for 
Plans, Few Advantages for Beneficiaries (GAO-04-960):

GAO Recommendation:

That the Administrator of CMS promptly instruct plans in the Medicare 
PPO demonstration to provide coverage for all plan services furnished 
by any provider authorized to provide Medicare services who accepts the 
plan's terms and conditions of payment.

CMS Response:

The CMS is working with the PPO Demonstration plans to ensure that they 
come into compliance with the provisions that govern their 
participation in the Medicare program. It should be noted that we have 
received no indication that the current availability of covered 
services offered by these plans has disadvantaged their enrollees. We 
will comply with the recommendation made in the report while working 
with the plans to ensure a smooth transition.

In responding to GAO's recommendation, we believe it is important to 
provide the policy context for structuring the PPO Demonstration as we 
did. As we had indicated in communications with GAO during the review of 
the PPO Demonstration, the solicitation for the demonstration specified 
that all plans under the initiative must offer access to out-of-network 
benefits. Some of the PPO Demonstration plans were concerned that they 
would not be able to pay the correct Medicare allowable amounts for 
certain services provided out-of-network and believed that there was 
more than adequate ability to provide these services in-network. In 
developing the demonstration, we chose to waive certain payment 
provisions of the private-fee-for-service (PFFS) contract under which 
the demonstration plans operate in order to maintain maximum 
flexibility to create products that resemble commercial PPO models.

We wanted to provide some flexibility in these demonstrations in order 
to mimic private sector PPO models and to facilitate rapid 
implementation for the 2003 open enrollment season. However, we made 
every effort to ensure that all Medicare-covered Part A and Part B 
services were covered in-network and, in fact, a significant focus of 
the qualification reviews conducted by CMS was to confirm that there 
was adequate access to network providers.

In addition, in order to protect beneficiaries, CMS ensured the 
following:

PPO plans had to provide access to all plan-covered services through 
network providers.

PPO plans provided full disclosure to members with regard to coverage 
and costs for all services, both those provided in-network and out-of-
network. CMS has put a significant amount of effort into the 
development of specific guidance for the PPO Demonstration plans to 
ensure that in-network and out-of-network benefits are clearly 
described in plan marketing material and the plans' Evidence of 
Coverage, a document that details the benefits offered by the 
demonstration plans.

PPO Demonstration enrollees had the ability to disenroll on a month-
to-month basis.

These protections will continue to apply in contract year 2005, the 
last year of the demonstration. Moreover, we have received no 
indication that enrollees in these plans have been disadvantaged in any 
way.

General Comments:

1) There is concern about the methodology used to estimate out-of-
pocket costs incurred by demonstration enrollees, as compared to out-of-
pocket costs of beneficiaries with other types of coverage. There is 
insufficient detail regarding methodology for us to fully comprehend 
how the average out-of-pocket costs were derived. GAO's analysis is 
based on national utilization assumptions, not PPO Demonstration 
enrollees' actual utilization of services. Without knowledge of actual 
utilization patterns of PPO Demonstration enrollees, any conclusions 
with regard to out-of-pocket costs are purely hypothetical.

It is also worth noting that GAO used national average spending 
estimates for Medicare FFS and Medigap to compare with the county-
estimated out-of-pocket costs for Medicare + Choice plans and 
demonstration PPO plans. We would note that most of the demonstration 
PPO plans are in high-cost areas; therefore, it is reasonable to 
believe that GAO's out-of-pocket cost comparison is potentially 
unreliable.

We are also concerned with GAO's focus on the 65 to 69 age group to 
estimate and compare out-of-pocket costs for the different types of 
coverage available to Medicare beneficiaries. The analysis completely 
disregards the important fact that most Medigap policies are age rated, 
and premiums increase dramatically as beneficiaries' age. Additionally, 
unlike Medicare + Choice enrollees, Medigap applicants are often 
subject to medical underwriting or screening. In contrast, PPO premiums 
are the same for all beneficiaries, regardless of age or health status. 
As a result, Medicare + Choice PPO plans may experience adverse 
selection. Clearly, PPO out-of-pocket costs are being compared to the 
most favorable Medigap out-of-pocket cost scenario. We believe that 
limiting the analysis to the 65 to 69 age group is a serious limitation 
that should be addressed.

GAO was informed of this concern at the exit conference, but the report 
fails to address the issue.

In addition, GAO used a sample group of 41 counties containing about 90 
percent of all demonstration PPO enrollment. The sample group includes 
Horizon's 21-county service area. Horizon had the largest in-network 
deductible for inpatient hospital services of all the demonstration PPO 
plans. This, too, could have an impact on the analysis, for the sample 
may be skewed towards the high out-of-pocket costs for the 
demonstration PPO plans.

2) The correct reference for "Fu Associates" is "Fu Associates, Ltd."

[End of section]

Appendix IV: GAO Contact and Staff Acknowledgments:

GAO Contact:

James C. Cosgrove at (202) 512-7029.

Acknowledgments:

In addition to the person named above, key contributors to this report 
were: Yorick F. Uzes, Zachary R. Gaumer, Jennifer R. Podulka, Jennie F. 
Apter, Helen T. Desaulniers, and Kevin C. Milne.

FOOTNOTES

[1] The Medicare Prescription Drug, Improvement, and Modernization Act 
of 2003 (MMA) changed the name "Medicare+Choice" to "Medicare 
Advantage." See Pub. L. No. 108-173, § 201, 117 Stat. 2066, 2176.

[2] Approximately 46,000 beneficiaries were enrolled in M+C plans that 
were not HMOs or PPOs. 

[3] See Social Security Amendments of 1967, Pub. L. No. 90-248, § 
402(a), 81 Stat. 821, 930-31 (1968) (codified at 42 U.S.C. § 1395b-
1(a)(1)(A) (2000)).

[4] Pub. L. No. 90-248, § 402(b), 81 Stat. 821, 930-31 (1968) (codified 
at 42 U.S.C. § 1395b-1(b) (2000)).

[5] The term "PPO" appears in a variety of contexts in this report. For 
the remainder of the report, we use the term "PPO" alone to refer to 
the preferred provider model, in which enrollees can obtain services 
from providers outside the network if they agree to bear a greater 
share of the costs for those services. We use the term "M+C PPO" to 
refer to non-demonstration PPO plans participating in Medicare's M+C 
program. We use the term "demonstration PPO" to refer to demonstration 
plans that are intended to operate as PPOs. The distinction between M+C 
PPO and demonstration PPO is important because CMS established 
different sets of requirements for the two types of plans. In this 
report, we do not refer to the health plans participating in the two 
demonstration programs as "M+C plans."

[6] In addition to Independence Blue Cross, seven non-PPO plans 
participated in the M+C Alternative Payment Demonstration in 2003.

[7] A Medigap plan is a private insurance plan designed to supplement 
FFS Medicare by covering some Medicare cost-sharing amounts and 
possibly additional benefits, depending on the type of plan selected.

[8] Pub. L. No. 90-248, § 402(b), 81 Stat. 821, 930-31(1968)(codified 
at 42 U.S.C. 1395b-1(b) (2000)).

[9] The term "organization" refers to a corporation or other business 
entity that may offer one or more health plans within a geographic 
region.

[10] PFFS plans allow enrollees to obtain services from any provider 
who is legally authorized to provide those services and accepts the 
plan's terms and conditions of payment.

[11] A benefit period begins the first day a beneficiary receives care 
from the hospital and ends when the beneficiary has not been 
hospitalized or received skilled nursing care for 60 consecutive days. 

[12] After the first 90 days of inpatient care, Medicare may help pay 
for an additional 60 days of inpatient care (days 91-150). Each 
beneficiary is entitled to a lifetime reserve of 60 days of inpatient 
coverage. Each reserve day may be used only once in a beneficiary's 
lifetime. 

[13] Medigap plans F and I include coverage for additional lifetime 
hospital days, part A and part B coinsurance and deductibles, and 
foreign travel emergencies. Medigap plan I also covers 50 percent of 
prescription drug costs subject to a $250 deductible and a $1,250 cap. 

[14] In some instances, a plan may elect to pay all or part of the 
beneficiary's part B premium.

[15] Though the M+C program has been changed to Medicare Advantage, 
statutory provisions under the M+C program concerning plan designs and 
operational requirements are the same in 2004 as they were in 2003.

[16] Under the Balanced Budget Act of 1997, M+C PPOs had been subject 
to the more extensive quality assurance requirements of HMOs. Section 
520 of the Medicare, Medicaid and SCHIP Balanced Budget Refinement Act 
of 1999 made M+C PPOs subject to the less extensive standards effective 
in January 2000. Pub. L. No. 106-113, App. F, § 520 (b), 113 Stat. 1501 
A-321, 1501A-385-86. 

[17] If an M+C PPO were offered by an organization licensed as an HMO, 
it would have to abide by the quality assurance requirements for M+C 
HMOs, not M+C PPOs. A similar requirement does not exist for 
organizations licensed as HMOs that offer PFFS plans. Pub. L. No. 106-
113, App. F, § 520, 113 Stat. 1501A-321, 1501A-385-86.

[18] Demonstrations may increase Medicare spending as long as the 
additional spending is related to the waiver of the payment or 
reimbursement rule.

[19] As with M+C plans, CMS adjusts a portion of the monthly payments 
to demonstration PPOs to account for the health and demographic 
characteristics of enrollees.

[20] The FFS-based rate for the PPO in the M+C Alternative Payment 
Demonstration was 98.5 percent of average county-level FFS spending per 
beneficiary, as estimated by CMS. The FFS-based rate for the plans in 
the Medicare PPO Demonstration was 99 percent of average county-level 
FFS spending per beneficiary, as estimated by CMS. 

[21] Of the 44 counties where 99 percent of average county-level FFS 
spending per beneficiary is higher than the M+C payment rate, 17 are 
included in the sample of counties we used to compare expected out-of-
pocket costs between types of coverage. This sample consists of 41 
counties with approximately 90 percent of all demonstration PPO 
enrollment.

[22] The risk-sharing agreement between CMS and Independence Blue Cross 
differed somewhat from the risk-sharing agreements between CMS and the 
plans in the Medicare PPO Demonstration. Rather than negotiate directly 
with the plan, CMS derived the expected MLR for Independence Blue Cross 
from the plan's Adjusted Community Rate Proposal (ACRP) for 2003. The 
ACRP is submitted by M+C plans and provides detailed estimates of a 
plan's expected costs and revenues associated with providing covered 
benefits, and a description of the plan benefit package. CMS subtracted 
the plan's expected administrative costs in the ACRP, which CMS 
subjects to certain limits, from the plan's total expected revenue to 
arrive at the expected MLR. In addition, unlike the risk-sharing 
agreements between CMS and the plans in the Medicare PPO Demonstration, 
there was no corridor in which Independence Blue Cross was at full 
risk. 

[23] All 28 plans with risk-sharing agreements in the Medicare PPO 
Demonstration had risk corridors of at least 2 percent above and below 
the expected MLR. 

[24] In commenting on a draft of this report, CMS stated that it had 
interpreted its regulations as precluding licensed HMOs from offering 
PPO plans. We note that more recent guidance issued by CMS to 
prospective Medicare Advantage contractors now states that licensed 
HMOs may offer PPO plans as long as they follow the more stringent 
quality assurance requirements. CMS has not modified the regulations on 
which it based its earlier interpretation.

[25] According to CMS officials, CMS signed PFFS contracts with all 
organizations in the Medicare PPO Demonstration, even those without HMO 
licenses, so that all plans participating in the demonstration would be 
subject to the same set of requirements. 

[26] Nearly all beneficiaries in Medicare PPO Demonstration plans were 
enrolled in plans that excluded coverage for some services obtained 
outside the provider network.

[27] See Social Security Act §1859(b)(2)(C).

[28] See 65 Fed. Reg. 40170, 40294. In response to a comment submitted 
during M+C rule-making that expressed concern that M+C PFFS plan 
quality assurance requirements were inadequate to protect enrollees, 
CMS acknowledged that PFFS plan quality assurance standards were less 
stringent than HMO standards. CMS nevertheless explained that quality 
assurance standards for PFFS plans may not be as important in the case 
of PFFS plans, "in which the enrollee has complete freedom of choice to 
use any provider in the country, and is not limited to a defined 
network of providers." Id. at 40220.

[29] We excluded employer-only M+C and demonstration PPO plans from our 
analysis because these plans, by design, are only available to retirees 
through their former employers, and therefore are not available to all 
beneficiaries. 

[30] The nine counties include Calvert and Charles counties, Maryland, 
and Boone, Hamilton, Hendricks, Johnson, Marion, Morgan, and Shelby 
counties of Indiana in 2003. 

[31] By July 2004, enrollment in demonstration PPOs had increased to 
127,336. 

[32] Between January and July 2004 the proportion of demonstration PPO 
enrollees coming from traditional FFS Medicare increased each month. 
While 30 percent of new demonstration PPO enrollees were previously 
enrolled in FFS Medicare in January 2004, 73 percent were from FFS 
Medicare by July 2004. 

[33] Out-of-pocket spending estimates are based on national averages. 
The 41 counties used in this analysis are high Medicare spending 
counties, so MPPF estimates may be lower than actual out-of-pocket 
costs for beneficiaries living in the 41 counties.

[34] Across our sample of 41 counties, 49 M+C HMO plans and M+C PPO 
plans (46 M+C HMO plans and 3 M+C PPO plans) and 2 M+C PFFS plans 
participated in M+C, although each of these plans were not available in 
all the counties. Because the average out-of-pocket costs of M+C HMOs 
and M+C PPOs were significantly different from those of M+C PFFS plans, 
we reported one figure for both M+C HMOs and M+C PPOs and a separate 
figure for M+C PFFS plans. 

[35] The FFS Medicare part B premium of $58.70 was excluded from our 
analysis of premiums and other out-of-pocket costs, in order to reflect 
the additional monthly costs beneficiaries can expect to incur that are 
unique to each type of coverage. The FFS Medicare part B premium is a 
required cost for enrollees in each of the six types of coverage. In 
some instances, however, a plan may pay all or part of a beneficiary's 
part B premium.

[36] The Medicare Web site provides separate out-of-pocket cost 
estimates for each of five self-reported health statuses: poor, fair, 
good, very good, and excellent.

[37] We also conducted a comparison of estimated beneficiary out-of-
pocket costs by expected health status for beneficiaries aged 80 to 84, 
and found results similar to our analysis of beneficiaries aged 65 to 
69. 

[38] While 2.4 million eligible Medicare beneficiaries live in counties 
where an M+C plan offers drug coverage, 4.0 million eligible Medicare 
beneficiaries live in counties where a demonstration PPO offers drug 
coverage. 

[39] PFFS plans are available in 8 of these 16 counties, but they did 
not include prescription drug coverage in their benefit packages. 

[40] Medigap plan F does not offer prescription drug coverage.

[41] We used Hillsborough County, Florida, for this example because the 
demonstration PPO and M+C HMO plans available to Medicare beneficiaries 
in this county in 2003 each included prescription drug coverage.

[42] We generated the six-night hospital stay estimates based on plan 
benefit descriptions and CMS's estimate for the average inpatient 
hospital length of stay for Medicare beneficiaries.

[43] To arrive at this projection, OACT compared how much Medicare 
would pay demonstration PPO plans per enrollee with the amount Medicare 
would spend on those beneficiaries if the demonstration did not exist 
and beneficiaries were instead enrolled in M+C plans or FFS Medicare. 
For Independence Blue Cross, which previously participated in M+C, OACT 
assumed that the plan's enrollment as a demonstration participant would 
resemble its previous enrollment. For plans in the Medicare PPO 
Demonstration, OACT calculated the market penetration of M+C plans in 
each county and then assumed that, because demonstration PPOs were 
specifically targeting FFS beneficiaries for enrollment, FFS 
beneficiaries would be three times more likely than beneficiaries 
enrolled in M+C plans to enroll in demonstration PPO plans. Increased 
spending would result if demonstration PPO enrollees were drawn from 
FFS Medicare in counties where the M+C rate was higher than the FFS-
based payment rate, or they were beneficiaries drawn from M+C plans in 
counties where the FFS-based payment rate was higher than the M+C rate. 


[44] OACT assumed that between 54 percent and 99 percent of 
demonstration PPO enrollees in each county would have switched over 
from FFS Medicare, depending on the counties where the plans operate.

[45] The Medicare Prescription Drug, Improvement, and Modernization Act 
of 2003 established the "Medicare Advantage" program and replaced the 
term "Medicare +Choice" with "Medicare Advantage." See Pub. L. No. 108-
173, § 201, 117 Stat. 2066, 2176. To avoid confusion, we refer to the 
program by the name it had at the time CMS initiated the Medicare PPO 
Demonstration.

[46] See Pub. L. No. 105-33, § 4001, 111 Stat. 251, 275-327 (adding new 
sections 1851 through 1859 to the Social Security Act)(codified at 42 
U.S.C. §§ 1395w-21 - 1395w-28) (2000 and 2001 Supp.)).

[47] Certain exceptions exist for enrollment in M+C plans by persons 
diagnosed with end-stage renal disease. See Social Security Act § 
1851(a)(3)(B)(codified at 42 U.S.C. § 1395w-21(a)(3)(B) (2000)).

[48] An M+C organization is a public or private entity that meets the 
requirements of part C for offering an M+C plan. See Social Security 
Act § 1859(a)(1)(codified at 42 U.S.C. § 1395w-28(a)(1) (2000)).

[49] Social Security Act § 1852(a)(1) (codified at 42 U.S.C. § 1395w-
22(a)(1) (2000)).

[50] See Social Security Act §§ 1852(a)(1)(B), (a)(3) (codified at 42 
U.S.C. §§ 1395w-22(a)(1)(B), (a)(3) (2000), respectively).

[51] 42 C.F.R. § 422.101(a) (2003).

[52] See 42 C.F.R. § 422.4(a)(1) (2003). 

[53] See 42 C.F.R. § 422.112(a) (2003).

[54] See 42 C.F.R. § 422.4(a)(1)(iii), (iv)(2003).

[55] See Social Security Act § 1859(b)(2)(A)(codified at 42 U.S.C. § 
1395w-28(b)(2)(A) (2000)).

[56] See Social Security Act § 1859(b)(2)(B)(codified at 42 U.S.C. § 
1395w-28(b)(2)(B) (2000)).

[57] See Social Security Act § 1859(b)(2)(C)(codified at 42 U.S.C. § 
1395w-28(b)(2)(C) (2000)).

[58] See Social Security Act § 1855(a)(1)(codified at 42 U.S.C. § 
1395w-25(a)(1) (2000)); see also 42 C.F.R. § 422.501(b)(1)(2003). An 
organization is considered to be licensed by the state as a "risk-
bearing entity" if it is licensed or otherwise authorized by the state 
to assume risk for offering health insurance or health benefits 
coverage, so that the entity is authorized to accept prepaid capitated 
payments for providing, arranging, or paying for comprehensive health 
services. See 42 C.F.R. § 422.2 (2003).

[59] Compare Social Security Act § 1852(e)(2)(A)(codified at 42 U.S.C. 
§ 1395w-22(e)(2)(A) (2000))(applicable to plans that are not private 
fee-for-service or PPO plans) with Social Security Act § 
1852(e)(2)(B)(codified at 42 U.S.C. § 1395w-22(e)(2)(B) 
(2000))(applicable to plans that are private fee-for-service plans and 
PPO plans); see also CMS Pub. 100-16, Medicare Managed Care Manual, 
Chapter 5, Quality Assessment (Rev. 39, 11-07-03). 

[60] See 42 C.F.R. § 422.112 (2003).

[61] The Medicare, Medicaid, and SCHIP Balanced Budget Refinement Act 
of 1999 (BBRA) revised the quality assurance standards for PPO plans 
under the M+C program and defined a PPO plan in that context as one 
that is "offered by an organization that is not licensed or organized 
under State law as a health maintenance organization." Pub. L. No. 106-
113, App. F, § 520, 113 Stat. 1501A-321, 1501A-385-86 (amending section 
1852(e)(2) of the Social Security Act)(codified at 42 U.S.C. § 1395w-
22(e)(2) (2000)). Based on this provision of the BBRA, CMS amended its 
regulations to define a PPO plan, in part, as one that is offered by an 
organization not licensed or organized under state law as an HMO. See 
42 C.F.R. §§ 422.4(a)(1)(iii), (iv), (b)(2003); 65 Fed. Reg. 40170, 
40175 (2000). 

[62] See Pub. L. No. 90-248, § 402(a)(1)(A), 81 Stat. 821, 930-31 
(1968), as amended by Pub. L. No. 92-603, § 222(b), 86 Stat. 1329, 
1391-93 (1972) (codified at 42 U.S.C. § 1395b-1(a)(1)(A) (2000)).

[63] Pub. L. No. 90-248, § 402(b), 81 Stat. at 931, as amended by Pub. 
L. No. 92-603, § 222(b)(2), 86 Stat. at 1393 (codified at 42 U.S.C. § 
1395b-1(b) (2000)). Specifically, section 402(b) provides the 
following: "In the case of any experiment or demonstration project 
under section 402(a)(1), the Secretary may waive compliance with the 
requirements of title XVIII and title XIX of the Social Security Act 
insofar as such requirements relate to reimbursement or payment on the 
basis of reasonable cost, or (in the case of physicians) on the basis 
of reasonable charge, or to reimbursement or payment only for such 
services or items as may be specified in the experiment." 

[64] Id.

[65] See 67 Fed. Reg. 18209 (2002)(soliciting applications for 
demonstration participants).

[66] Id.

[67] Id. at 18211.

[68] Thirteen of the 33 plans participating in the PPO Demonstration 
are offered by organizations with HMO licenses. All plans were 
presented to potential enrollees as being either HMO plans with a point 
of service option or PPO plans, both of which are coordinated care 
plans.

[69] The statute provides as follows: "In the case of an M+C private 
fee-for-service plan, the organization offering the plan must 
demonstrate to the Secretary that the organization has a sufficient 
number and range of health care professionals and providers willing to 
provide services under the terms of the plan. The Secretary shall find 
that an organization has met such requirement with respect to any 
category of health care professional or provider if, with respect to 
that category of provider - the plan has established payment rates for 
covered services furnished by that category of provider that are not 
less than the payment rates provided for under part A, part B, or both 
for such services, or the plan has contracts or agreements with a 
sufficient number and range of providers within such category to 
provide covered services under the terms of the plan, or a combination 
of both. The previous sentence shall not be construed as restricting 
the persons from whom enrollees under such a plan may obtain covered 
benefits." Social Security Act § 1852(d)(4)(codified at 42 U.S.C. § 
1395w-22(d)(4) (2000))(emphasis added).

[70] See 63 Fed. Reg. 34968, 35039 (1998). See also 65 Fed. Reg. 40170, 
40296 (2000)(under § 1852(d)(4), M+C organizations offering private 
fee-for-service plans cannot restrict providers from whom beneficiary 
can receive care).

[71] 63 Fed. Reg. at 35039.

[72] See 65 Fed. Reg. 40170, 40294 (2000). CMS also explained that, in 
enacting the quality assurance standards of the BBA, "Congress 
recognized that not all of the quality assessment and performance 
improvement activities that are appropriate for a plan with a defined 
provider network would be appropriate for . . . an M+C private fee-for-
service plan." Id. at 40220.

[73] See 42 C.F.R. § 422.4(a)(1)(iv)(2003).

[74] See Social Security Act § 1859(b)(2)(C)(codified at 42 U.S.C. § 
1395w-28(b)(2)(C) (2000)).

[75] See Social Security Act § 1859(b)(1)(codified at 42 U.S.C. § 
1395w-28(b)(1) (2000)).

[76] See CMS Pub. 100-16, Medicare Managed Care Manual, Chapter 4, 
Benefits and Beneficiary Protections,150 (Rev. 23, 06-06-03). The 
Medicare Managed Care Manual provides the following: "To be eligible to 
furnish care to a private fee-for-service enrollee: (1) Physicians must 
be state licensed, and either have a Medicare billing number or be 
eligible to obtain one; and (2) Institutional providers, such as 
hospitals and skilled nursing facilities, must be certified to treat 
Medicare beneficiaries." Id. See also 42 C.F.R. § 
422.114(b)(2003)(private fee-for-service plan must allow enrollees "to 
obtain services from any entity that is authorized to provide services 
under Medicare parts A and B").

[77] See Morton v. Ruiz, 415 U.S. 199, 231-32 (1974); Natural Resources 
Defense Council v. EPA, 907 F.2d 1146, 1165 (D.C. Cir. 1990); and 
United States v. Detroit, 720 F.2d 443, 451 (6th Cir. 
1983)(administrative agencies may not override an express statutory 
requirement). 

[78] GSA, which provides a monthly list of service areas for all risk 
and cost Medicare managed care contracts, allowed us to tie plan option 
contract identification numbers to specific service areas. However, 
CMS's Medicare Managed Care Monthly Report, which provides regular 
monthly updates of active and terminated Medicare contracts, and MMCC, 
which provides statistics regarding all Medicare contract managed care 
plans, we also used because they can be more accurate than the GSA at 
calculating overall demonstration PPO enrollment. This is because GSA 
excludes enrollment in service areas that have ten or fewer enrollees.

[79] Employer-only plans, by design are only available to retirees 
through their former employers. Employer-only plans exist in 
demonstration PPO form and in M+C plan form; therefore we specifically 
eliminated employer-only plans from both types of plans. Dual-eligibles 
are Medicare beneficiaries who are also eligible for Medicaid services. 

[80] CMS's Medicare Health Plan Compare is used as an administrative 
data set to track all participating private health plan benefit 
information.

[81] CMS's MPPF is a Web tool intended to assist Medicare beneficiaries 
with selecting a Medicare health plan. MPPF contains information on M+C 
plan availability by geographic location, estimated beneficiary out-of-
pocket costs, and certain benefits provided by each available type of 
Medicare coverage. MPPF can be found at the following Web address 
www.medicare.gov/MPPF/home.asp.

[82] Throughout our analysis we compare demonstration PPO plans to 
Medigap plans F and I. As of 1999, Plan F enrolled the most Medicare 
beneficiaries of the 10 available Medigap policies, and Plan I was the 
most widely offered Medigap policy that included a prescription drug 
benefit. GAO, Medigap Insurance: Plans are Widely Available but Have
Limited Benefits and May Have High Costs, GAO-01-941 (Washington D.C.: 
July 2001).

[83] In cases where a demonstration PPO or M+C plan offered a plan with 
multiple options, or different sets of benefit packages and premiums, 
in a county, we used four criteria to identify the plan option that was 
most favorable for beneficiaries. We first selected plan options for 
each organization that included prescription drug coverage. If there 
was more than one, we selected the option with the lowest premium. If 
there were two drug plans with identical premiums, we selected the 
option that offered brand-name drug coverage and lowest deductible. If 
there was more than one plan that met the first three criteria, we 
selected the option with the lowest out-of-pocket cost spending cap.

[84] Because Horizon's 2003 demonstration PPO is the only demonstration 
PPO that was an M+C HMO in 2002 and Horizon has the largest proportion 
of demonstration PPO enrollees, we included all service areas where 
Horizon's demonstration PPO was available.

[85] Overlap exists between these two groups of counties. Bergen, 
Monmouth and Ocean counties of New Jersey contain Horizon demonstration 
PPOs and have a significant amount of enrollment in Aetna's 
demonstration PPO. 

[86] Fu defined a cohort of FFS individuals based on the 1998 and 1999 
Medicare Current Beneficiary Surveys (MCBS). This cohort provides the 
basis from which to identify the utilization measures and out-of pocket 
costs for the Medicare Personal Plan Finder (MPPF) database. Actual 
2003 premiums, deductibles, and selected fee-for-service copayments 
from this cohort formed the basis for the fee-for-service component of 
the MPPF database. The contract year 2003 plan benefit packages were 
used to define the out-of-pocket costs associated with contract year 
2003 M+C plans and demonstration PPOs. Finally, Medigap premium data 
were used to define the out-of-pocket costs for contract year 2003 
Medigap plans.

[87] Beneficiaries aged 65 to 69 are generally the newest to the 
Medicare program and, in 1999, were the largest age group in the 
Medicare program. We conducted a similar analysis for beneficiaries 
aged 80 to 84.

[88] Data were drawn from Fiscal Year 2003 Medicare Health Plans 
Compare data set on September 11, 2003, the final version for that 
year.

[89] CMS's HPMS captures Medicare private health plans non-network 
benefits.

[90] Prior to the Medicare Prescription Drug, Improvement, and 
Modernization Act of 2003 (MMA) CMS required that organizations 
participating in the PPO demonstration submit estimates of enrollee 
costs for 2004, because they were not required to submit the standard 
M+C adjusted community rate proposal. Plans' estimates of enrollee 
costs, broken down between services obtained within and outside their 
provider networks, do not reflect revenues included in MMA, which was 
enacted in December 2003.

GAO's Mission:

The Government Accountability Office, the 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 the Internet. GAO's Web site ( www.gao.gov ) contains 
abstracts and full-text files of current reports and testimony and an 
expanding archive of older products. The Web site features a search 
engine to help you locate documents using key words and phrases. You 
can print these documents in their entirety, including charts and other 
graphics.

Each day, GAO issues a list of newly released reports, testimony, and 
correspondence. GAO posts this list, known as "Today's Reports," on its 
Web site daily. The list contains links to the full-text document 
files. To have GAO e-mail this list to you every afternoon, go to 
www.gao.gov and select "Subscribe to e-mail alerts" under the "Order 
GAO Products" heading.

Order by Mail or Phone:

The first copy of each printed report is free. Additional copies are $2 
each. A check or money order should be made out to the Superintendent 
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 
more copies mailed to a single address are discounted 25 percent. 
Orders should be sent to:

U.S. Government Accountability Office

441 G Street NW, Room LM

Washington, D.C. 20548:

To order by Phone:

	

Voice: (202) 512-6000:

TDD: (202) 512-2537:

Fax: (202) 512-6061:

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

Contact:

Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: fraudnet@gao.gov

Automated answering system: (800) 424-5454 or (202) 512-7470:

Public Affairs:

Jeff Nelligan, managing director,

NelliganJ@gao.gov

(202) 512-4800

U.S. Government Accountability Office,

441 G Street NW, Room 7149

Washington, D.C. 20548: