Skip front matter and go directly to the report.

MEDICARE ADVANTAGE PLAN AVAILABILITY, PREMIUMS AND BENEFITS, AND
BENEFICIARY ENROLLMENT IN 2006

Authors:
Gregory C. Pope, M.S.,
Leslie Greenwald, Ph.D.,
John Kautter, Ph.D.,
Brian Dulisse, Ph.D.,
Nathan West, M.A.

Project Director: Gregory Pope, M.S.

Associate Project Director: Leslie Greenwald, Ph.D.

Federal Project Officer: Victor G. McVicker, Jr.

RTI International*

CMS Contract No. 500-00-0024, T.O. #17

November 2007

This project was funded by the Centers for Medicare & Medicaid Services under contract no. 500-00-0024, T.O. #17. The statements contained in this report are solely those of the authors and do not necessarily reflect the views or policies of the Centers for Medicare & Medicaid Services. RTI assumes responsibility for the accuracy and completeness of the information contained in this report.

*RTI International is a trade name of Research Triangle Institute

CONTENTS

EXECUTIVE SUMMARY

SECTION 1. PROJECT BACKGROUND AND KEY ISSUES IN THE MEDICARE ADVANTAGE PROGRAM
SECTION 2. METHODOLOGY
SECTION 3. PLAN AVAILABILITY
SECTION 4. PREMIUMS AND BENEFITS
SECTION 5. ENROLLMENT
GLOSSARY OF ACRONYMS

REFERENCES

COMPUTER OUTPUT TABLE

LIST OF TABLES

Table 3-1. Number of Medicare Advantage contracts, by plan type
Table 3-2. Percent of counties with at least one Medicare Advantage contract, by plan type
Table 3-3. Number and percentage of Medicare beneficiaries with access to a Medicare Advantage plan, by plan type
Table 3-4. Percent of counties with at least one Medicare Advantage contract, by plan type and urbanicity
Table 3-5. Percent of counties with at least one Medicare Advantage contract, by plan type and region
Table 3-6. Percentage of beneficiaries and counties with access to Medicare Advantage plan types, 2006
Table 3-7. Percentage of beneficiaries with access to Medicare Advantage plan types, by urbanicity, 2006
Table 3-8. Percentage of beneficiaries with access to Medicare Advantage plan types, by region, 2006
Table 3-9. Percentage of beneficiaries with access to Medicare Advantage contracts, by number of contracts and urbanicity, 2006
Table 3-10. Percentage of beneficiaries with access to Medicare Advantage contracts, by number of contracts and region, 2006
Table 3-11. Number of special needs contracts and plans by plan type and target beneficiaries, 2006

Table 4-1. Mean monthly premiums of Medicare Advantage plans by plan type, 2006
Table 4-2. Percent of Medicare Advantage enrollees in zero premium plans by plan type, 2006
Table 4-3. Percent of enrollees in Medicare Advantage plans, by premium range, 2006
Table 4-4. Mean monthly premiums of Medicare Advantage plans by urbanicity, 2006
Table 4-5. Percent of Medicare Advantage enrollees in zero premium plans by urbanicity, 2006
Table 4-6. Mean monthly premiums of Medicare Advantage plans by region, 2006
Table 4-7. Percent of Medicare Advantage enrollees in zero premium plans by region, 2006
Table 4-8. Part B premium reduction by Medicare Advantage plan type, urbanicity, and region, 2006 Percent of enrollees
Table 4-9. Prescription drug benefits by Medicare Advantage plan type, 2006 Percent of enrollees with Part D benefit type for each MA plan type
Table 4-10. Prescription drug benefits of Medicare Advantage enrollees by urbanicity, 2006 Percent of enrollees with Part D benefit type for each urbanicity category
Table 4-11. Prescription drug benefits of Medicare Advantage enrollees by region, 2006 Percent of enrollees with Part D benefit type for each region
Table 4-12. Cost sharing before the initial coverage limit, by type of Medicare Advantage prescription drug plan, 2006 Percent of enrollees in each Part D benefit type with specified cost sharing
Table 4-13. Drug tiers of Medicare Advantage prescription drug plans, 2006 Percent of enrollees by number of copayment and coinsurance tiers
Table 4-14. Common cost sharing structures in Medicare Advantage prescription drug plans, 2006 Median copayments or coinsurance by drug tier
Table 4-15. Initial coverage limit in Medicare Advantage prescription drug plans, by plan and geographic characteristics, 2006 Percent of enrollees with specified initial coverage limits
Table 4-16. Gap coverage in Medicare Advantage prescription drug plans, by plan and geographic characteristics, 2006 Percent of enrollees with specified gap coverage
Table 4-17. Selected characteristics of Medicare Advantage prescription drug plan formularies, by plan and geographic characteristics, 2006 Enrollment-weighted plan mean of number of drugs
Table 4-18. Selected mandatory supplemental benefits in Medicare Advantage plans, 2006 Percent of enrollees with benefit
Table 4-19. Cost sharing for selected services in Medicare Advantage plans, 2006 Percent of enrollees with cost sharing category
Table 4-20. Out of pocket (OOP) maximums in Medicare Advantage plans, 2006 Percent of enrollees
Table 4-21. Out of pocket (OOP) maximums in Medicare Advantage plans, by plan type, urbanicity, and region, 2006

Table 5-1. Medicare Advantage enrollment, 2000–2006
Table 5-2. Change in Medicare Advantage enrollment, 2001–2006
Table 5-3. Medicare Advantage enrollment by plan, geographic, and beneficiary characteristics, 2006
Table 5-4. Medicare Advantage enrollment, plan type by urbanicity and census region, 2006
Table 5-5. Medicare Advantage enrollment, plan type by urbanicity and census region, row and column percentages, 2006
Table 5-6. Special needs plan enrollment, by plan type, 2006
Table 5-7. Part D enrollment in Medicare Advantage, by plan type, 2006

EXECUTIVE SUMMARY

This report presents findings on Medicare Advantage (MA) plan availability, premiums and benefits, and beneficiary enrollment in 2006. It focuses especially on key new features of MA for 2006, including the Part D prescription drug benefit, the new regional Preferred Provider Organization (PPO) plan type, and the more widely available special needs plans (SNPs). In light of these changes in the program, 2006 is an important year in which to monitor these elements of the MA program. Where feasible, 2006 developments were put in context of 2000 to 2005 trends in MA that were identified in prior work.

Several important changes occurred in the MA program in 2006. The Medicare Modernization Act (MMA) added a major new benefit to the basic Medicare benefit package in 2006, the Part D prescription drug benefit. Many MA plans had offered a drug benefit prior to 2006, but the benefit was usually limited, such as covering generic drugs only and/or having annual drug benefit caps. Beginning in 2006, MA plans were required to offer at least one plan in an area with the standard Part D prescription drug benefit (or an actuarially equivalent benefit). MA plans could also offer enhanced alternative drug coverage.

New types of plans were created by the MMA or earlier Balanced Budget Act (BBA) of 1997 that offer alternative provider access, premiums, and benefits to beneficiaries. These include local PPOs, which allow access to out of network providers at a higher cost sharing level; regional PPOs, which are PPOs that cover an entire region as specified by the Centers for Medicare & Medicaid Services (CMS), and these regional definitions include either an entire State or a mix of entire States; private fee-for-service (PFFS) plans, which permit access to any provider who accepts on a service by service basis the plan's terms and conditions for payment; and special needs plans (SNPs), which are targeted at beneficiaries with special needs. The Conference Report for the BBA of 1997 notes the contrast between innovations in health benefit design and delivery in the private sector and the Medicare program, adding that the creation of the private plan options in the BBA "will allow beneficiaries to have access to a wide array of private health plan choices in addition to traditional fee-for-service Medicare." The Conference Report for the MMA of 2003 cites the decline in plan participation and indicates that the immediate changes to the payment methodology for the MA program were included in the law to "encourage plan entry," adding that "The goal is to increase beneficiary choice, by increasing private plan participation in Medicare." In the discussion of the regional PPO option, the MMA Conference Report also refers to bringing greater health plan choices to areas not previously served by private plans, particularly rural areas.

Also beginning in 2006, payments to MA plans were determined through a new bidding process. Bids below the benchmark (with 25 percent of any difference between bid and benchmark retained in the Medicare trust funds) created rebate funds that are used to enhance benefits, reduce cost sharing, or reduce Part D or Part B premiums; the portion of any bid amount in excess of the "benchmark" rate became the beneficiary premium. For the period March–December 2004, the MMA changed county capitation rates by establishing a fee-for-service (FFS) per capita cost minimum capitation rate, raising floor rates, and establishing a minimum update of the greater of the national Medicare expenditure growth percentage or 2 percent. Another significant payment change is that the phase in of risk adjustment continued; 75 percent of plan payments were risk adjusted in 2006.

We note that data for 2006, as well as some data for 2007, have been made available by CMS that may not match precisely other reports of similar data. As the chapter on our methodology explains, there are multiple sources for data and some data are available for 2006 that had not been available in previous years. In addition, there are varying approaches to data presentation. For example, as we explain in both the text and footnotes, in a number of tables we excluded enrollment in employer plans. Readers should bear in mind that these and other factors mean that findings across sources may vary, but we do not believe the differences represent any material effect.

Key Findings

Plan availability

1) Access to MA plans

• Continuing a trend observed after the passage of the MMA in 2003, the total number of MA contracts increased sharply in 2006 to a total of 387, from 289 in June 2005. All plan types contributed to the increase. Likely drivers were higher MA payment rates and the implementation of the Medicare Part D prescription drug benefit. (Table 3-1)
  
  
• Almost all Medicare beneficiaries had access to at least one MA option in 2006. The policy goal of extending access to MA plans to all areas, including rural areas, had largely succeeded by 2006. PFFS plans and regional PPOs, which typically cover larger service areas than HMOs and local PPOs, were instrumental in achieving the universal access goal. (Table 3-3)
  
  
• Each Medicare Advantage plan type was available in more counties in 2006 than in 2005. In 2006, HMOs were available in just over a third of all counties, 30 percent of counties had access to a local PPO, PFFS plans were available in 96 percent of counties, and regional PPOs were available in just under 90 percent of all counties. SNPs were offered in 23 percent of counties. (Table 3-2)
  
  
• By 2006, all MA plan types were available to a majority of Medicare beneficiaries. HMOs were available to 72 percent of Medicare beneficiaries, 65 percent had access to a local PPO, 89 percent to a regional PPO, and 81 percent to a PFFS plan. (Table 3-3)
  
  
• In 2006, PFFS plans and regional PPOs were widely available throughout urban and rural areas. HMOs, local PPOs, and SNPs were more widely available in large and medium urban areas. (Table 3-4)
  
  
• HMOs, local PPOs, and SNPs were most widely available in the Northeast. PFFS plans and regional PPOs were nearly universally available in the Midwest and South, and had substantial, though lesser, availability in the Northeast and West. (Table 3-5)

2) Access to multiple MA plan types and contracts in 2006

• In 2006, the majority, 54 percent, of Medicare beneficiaries—and 31 percent of counties—had access to all three major plan types of HMOs, PPOs (local or regional), and PFFS. PPOs and PFFS plans were available to another 25 percent of beneficiaries and 60 percent of counties. HMOs and PPOs were available to an additional 18 percent of beneficiaries and 2 percent of counties. Only about .4 percent of beneficiaries had no access to any MA plans as compared to about 15.2 and 3.6 percent in 2004 and 2005 respectively. The MMA's goal of increasing the range of options available to Medicare beneficiaries was largely achieved by 2006. (Table 3-6)
  
  
• Urban beneficiaries were most likely to have access to all three plan types (HMO, PPO, PFFS—62 percent), or to HMOs and PPOs (22 percent). Rural beneficiaries were most likely to have access to PFFS plans and PPOs (64 percent) or to all three plan types (27 percent). (Table 3-7)
  
  
• Northeastern and Western beneficiaries were most likely to have access to all three major plan types, or to HMOs and PPOs. Midwestern and Southern beneficiaries were most likely to have access to all three plan types or PFFS and PPOs. (Table 3-8)
  
  
• About one-quarter of Medicare beneficiaries had access to 10 or more MA contracts in 2006, including 40 percent in large urban areas, 26 percent in medium urban areas, but less than 5 percent in small urban and rural areas. In urban areas, 90 percent of beneficiaries had access to four or more MA contracts, as did a majority, 54 percent, of beneficiaries in rural areas. (Table 3-9)
  
  
• In the Northeast and West, more than 30 percent of beneficiaries had a choice among 10 or more MA contract options, compared to fewer than 20 percent of beneficiaries in the Midwest or South. The South had the highest percentage—22 percent—of beneficiaries with access to three or fewer contract options. (Table 3-10)
  
  
• Three-quarters of SNPs offered in 2006 were HMOs targeted at dual Medicare/Medicaid eligible beneficiaries. (Table 3-11)

Premiums and benefits^1,2

1) Premiums

• Over half (53 percent) of MA enrollees received their Part C and Part D benefits at zero additional premium in 2006. The proportion of enrollees in open access plans (excludes SNP enrollees) paying neither a Part C nor a Part D premium varied from a high of 65 percent for PFFS plans to a low of 16 percent for local PPOs. (Table 4-2)
  
  
• The average monthly MA Part C premium in 2006 was $19.71, the average Part D premium was $11.63, and the average total (Parts C+D) premium was $30.43. (Table 4-1)
  
  
• By plan type, the 2006 average total (Parts C+D) premium was $14.80 (PFFS); $24.73 (SNP); $26.85 (regional PPO); $30.84 (HMO); $68.33 (local PPO). (Table 4-1)
  
  
• Although most MA enrollees paid zero or modest premiums, over one-fifth (21 percent) paid a total monthly premium of $75 or greater and 8 percent paid $100 or more. (Table 4-3)
  
  
• The urban-rural difference in MA premiums was relatively modest: the average MA total monthly premium of urban MA enrollees was $30.16 and of rural MA enrollees was $35.80. MA enrollees in large urban areas paid an average total premium of $27.48, the lowest of any urban-rural category. Enrollees in rural areas not adjacent to urban areas paid an average premium of $29.14, the second lowest of any urban-rural category. Average total premiums of MA enrollees in small urban areas were $46.73, the highest of any urban-rural category. (Table 4-4)
  
  
• Regional premium differences were pronounced. Average total premiums were highest in the Northeast and lowest in the South ($56.05 versus $12.26). Over 7 of 10 Southern MA enrollees paid no total premium, while one in four of Northeast MA enrollees were in plans with a zero total premium. The Northeast had an unusually low percentage of enrollees in MA plans with a zero Part D premium, only 24 percent, compared to at least 60 percent in other regions. (Tables 4-6 and 4-7)
  
  
• In 2006, 2.7 percent of MA enrollees had their Medicare Part B premium reduced by their plan, by an average of $42.07. (Table 4-8)

2) Part D benefits

• About 11 percent of MA enrollees were in plans without a Part D benefit as compared to the 35 percent of PFFS enrollees and 30 percent of rural MA enrollees. PFFS plans are not required to offer a Part D benefit. (Tables 4-9 and 4-10)
  
  
• Sixty-two percent ³ of MA enrollees had an enhanced Part D benefit. Defined standard Part D coverage was uncommon, except in SNPs, where 65 percent of enrollees had such coverage, and only 14 percent of SNP enrollees had enhanced coverage. The Part D low income subsidy presumably exempted most SNP enrollees from the cost sharing in defined standard coverage. (Table 4-9)
  
  
• Northeastern MA enrollees were least likely to have enhanced drug coverage (42 percent), while Southern enrollees were especially likely to have enhanced coverage (77 percent). (Table 4-11)
  
  
• The vast majority (86%) of MA prescription drug plan (MA-PD) enrollees paid no Part D deductible. (Table 4-12)
  
  
• About 93 percent of MA-PD enrollees were in plans with drug copayment tiers before the initial coverage limit. The number of co-payment tiers was almost always two or three. About one-quarter of MA-PD enrollees were in plans without coinsurance tiers. Almost all of the remaining enrollees faced one or two coinsurance tiers. The most common tiering structure (29 percent of MA-PD enrollees) was three copayment tiers and one coinsurance tier. (Tables 4-12 and 4-13)
  
  
• The most common copayment for tier 1 drugs (typically generic drugs) was $5, but it ranged as high as $20 in some plan designs. The copayment for tier 2 drugs (usually preferred brand) was typically about $30, but ranged from $20 to $40. Plans with a third tier (usually nonpreferred drugs) typically charged a $50 to $60 copayment for drugs in this tier. The cost sharing percentage for drugs in a coinsurance tier (usually specialty, injectable, or expensive drugs) was typically 25 or 33 percent, but ranged up to 50 percent in a second coinsurance tier.⁴ (Table 4-14)
  
  
• About three-quarters of MA-PD enrollees were in plans with the standard $2,250 initial coverage limit. About 13 percent had a lower, and about 10 percent a higher, initial coverage limit. (Table 4-15)
  
  
• About 27 percent of MA-PD enrollees were in plans with some form of gap coverage in 2006. Overwhelmingly, gap coverage was for generic drugs only (84% of all enrollees with gap coverage had generics only). (Table 4-16)
  
  
• No PFFS and very few regional PPO MA-PD enrollees had gap coverage. About one-quarter of HMO and local PPO MA-PD enrollees had gap coverage for generic drugs, but less than 5 percent had brand drug gap coverage. (Table 4-16)
  
  
• Urban MA-PD enrollees were much more likely to have gap coverage than rural enrollees (28 percent versus 8 percent). (Table 4-16)
  
  
• Gap coverage varied considerably by region. Only 9 percent of Midwestern MA-PD enrollees had any gap coverage, compared to 35 percent of Southern enrollees. The South also had the highest percentage of brand and generics gap coverage, over 7 percent, compared to none in the Northeast and less than 1 percent in the Midwest. Overall, Southern and Western MA-PD enrollees had more plans with gap coverage than Northeastern or especially Midwestern enrollees. (Table 4-16)
  
  
• HMOs covered fewer of the top 100 drugs on their formularies than other plan types, but had less than $20 cost sharing for a larger number of these drugs. PFFS plans listed the largest number of drugs on their formularies, and required prior authorization for the fewest of the top 100 drugs. Regional PPOs listed the fewest drugs on their formularies, on average. (Table 4-17)
  
  
• Urban MA-PD enrollees ‘plans covered on average fewer of the top 100 drugs than rural enrollees' plans, but the urban plans charged less than $20 cost sharing for a larger number of the top 100 drugs and listed slightly more total drugs on their formularies. (Table 4-17)
  
  
• Western MA-PD plans appear to have more restrictive drug access policies than other regions. Western plans on average listed considerably fewer of the top 100 drugs and fewer total drugs on their formularies, and required prior authorization for a larger number of drugs. (Table 4-17)

3) Other benefits and cost sharing

• Eighty-three percent of MA enrollees had in 2006 vision coverage (eye exams and glasses). About two-thirds of MA enrollees had coverage for hearing exams, one-third dental coverage, about one-quarter coverage for podiatry, and only 6 percent for chiropractic treatment. (Table 4-18)
  
  
• Most MA enrollees faced copayments of $5 to $15 for primary care physician visits. Eleven percent had no primary care copayment while 8 percent paid between $15 and $25. Copayments for specialist physician visits were higher. The most common amounts were in the $15 to $25 range, and one-quarter of MA enrollees paid more than $25 per specialist visit. Emergency department copayments were almost always about $50. More than 85 percent of MA enrollees faced copayments or coinsurance for hospital services, either acute inpatient admissions, or outpatient care. About three-quarters were charged copayments or coinsurance for X-ray and clinical laboratory services. (Table 4-19)
  
  
• About 42 percent of MA enrollees had an out of pocket (OOP) maximum. A typical OOP maximum was $3,000, and most maximums ranged from $1,000 to $5,000. (Table 4-20)
  
  
• OOP maximums were least common in HMOs—only one-third of HMO enrollees had one in 2006. All enrollees in regional PPOs had one, as required by the MMA. Most PFFS enrollees (80 percent) had an OOP maximum, as did about half of local PPO enrollees. Although less common in HMOs, HMO OOP maximums were typically lower ($3,000 median) than PFFS plans' maximums ($5,000 median). Local PPOs had a particularly low median OOP maximum, at $1,500. (Table 4-21)
  
  
• Urban enrollees were less likely to have OOP cost maximums than rural enrollees, but when they existed, urban maximums were typically lower. Regionally, most Midwestern enrollees had OOP cost maximums, but few Northeastern enrollees did. Nearly half of Southern and Western enrollees were in plans with an OOP maximum. (Table 4-21)

Enrollment

1) Medicare Advantage Enrollment Trends, 2000–2006 (Table 5-2)

• Total MA enrollment has varied over the last decade. Enrollment fell by 25 percent between 2000 and 2003, but recently has rebounded. By 2006, MA enrollment was similar to enrollment in 2000.
  
  
• There was a significant increase in MA enrollment between 2005 and 2006, with an overall increase in enrollment of 31 percent. Greater availability of plans and of plan types offering less restrictive access to providers, the addition of the Part D benefit, and more attractive premiums and benefits resulting from higher MA payment rates were likely drivers of higher enrollment.
  
  
• PFFS enrollment rose substantially between 2005 and 2006, by 682,345 beneficiaries. PFFS enrollment grew by nearly as many beneficiaries as HMO enrollment, despite starting from a much smaller base than HMOs, which have historically dominated the MA market.

2) Medicare Advantage Enrollment in 2006

• MA enrollment in July 2006 was 5.5 million, with a penetration rate of 14.2 percent of MA-eligible beneficiaries. Although HMOs are still the dominant players in MA, together PFFS and PPOs (local and regional) comprised about 20 percent of MA enrollment. Regional PPOs had 89,409 enrollees in July 2006. (Table 5-3)
  
  
• Among MA enrollees, 91 percent resided in urban areas, with 9 percent in rural areas. At 17 percent, the percentage of beneficiaries residing in urban areas taking up an MA plan was triple that of the beneficiaries residing in rural areas (6 percent). (Table 5-3)
  
  
• Among the 4.4 million MA HMO enrollees, 4 percent resided in rural areas. This contrasted with the 768,324 PFFS enrollees, of whom 34 percent resided in rural areas. PFFS is increasing MA enrollment in rural areas, consistent with the long-time policy goal of improving MA plan availability and attractiveness in rural areas. (Table 5-5)
  
  
• The regional PPO option was created, in part, to provide more MA options to rural beneficiaries. They drew 18 percent of their total enrollment from rural areas, more than double the percentage of HMOs or local PPOs, but about half the percentage of PFFS. Regional PPOs accounted for only 3 percent of total rural MA enrollment in 2006. Over half of rural MA enrollees were in PFFS plans, with most of the rest in HMOs. In contrast, 84 percent of urban MA enrollees were in HMOs, with 10 percent in PFFS plans, predominantly in medium and small urban areas. (Table 5-5)
  
  
• The South and West had the largest number of MA enrollees among census regions, with 1.7 million each. However, among Medicare beneficiaries eligible for MA, the take-up rate for MA in the West was about twice that of the South (22 versus 12 percent). (Table 5-3)  
  
  
• MA enrollment in the Northeast and West was dominated by HMOs, which had about a 90 percent share of MA enrollment in each of these regions. This differs substantially from the Midwest and South, where PFFS plans were much more popular. Over half of the regional PPO enrollment was in the South. (Table 5-5)
  
  
• Certain populations such as those eligible for Medicare by disability, the very old (age 85 and up), and dual Medicare/Medicaid eligible beneficiaries had lower take-up rates for MA. (Table 5-3)
  
  
• 348,842 (6 percent) of MA enrollees were enrolled in a SNP, with the vast majority (94 percent) enrolled in SNPs serving beneficiaries dually eligible for Medicare and Medicaid. Ninety-two percent of SNP enrollees were in HMOs. (Table 5-6)
  
  
• Among MA enrollees, 93 percent were enrolled in Medicare Part D, with broadly similar Part D take up rates across MA plan types, although PFFS plans had somewhat lower take up rates. Most MA enrollees who were enrolled in Part D were enrolled in an MA-PD (96.4 percent), although 3.6 percent (all PFFS enrollees) were enrolled in a standalone prescription drug plan. About 27 percent of PFFS enrollees with Part D coverage were enrolled in standalone drug plans. (Table 5-7)

SECTION 1
PROJECT BACKGROUND AND KEY ISSUES IN THE MEDICARE ADVANTAGE PROGRAM

1.1 Project Background and Overview of this Report

For more than 20 years, Medicare has offered enrollment in private health plans as an option to beneficiaries in areas where these plans were available. Private health care plans cover all the services of the traditional Medicare FFS program and often offer additional benefits that are attractive to beneficiaries. Plans may charge their enrollees a monthly premium. Many different options are available, including health maintenance organizations (HMOs), which typically provide coverage for services obtained from their "network" hospitals and doctors, and preferred provider organizations (PPOs), which include coverage for services provided "out of network," generally for a higher co-payment. A fast growing option is Private Fee for Service (PFFS) plans, which can and often do operate without formal provider networks.

The Medicare private health plan program is known as the "Medicare Advantage" (MA) program. Medicare pays MA plans a fixed, prospective amount per enrollee per month, independent of the actual medical services used by the enrollee. MA plans historically have participated unevenly around the country, with greater availability in large urban areas and more limited presence in rural areas. Throughout the years, the types of plans and benefit offerings have undergone substantial change. The current phase in the evolution of the Medicare Advantage program will be particularly eventful with the mandated integration of Part D (prescription drug) benefits, the introduction of regional PPO plans, and the expansion of PFFS.

At the broadest level, this report documents the increased enrollment in Medicare Advantage plans in 2006 compared to prior years, as well as the increased availability of MA plan options for beneficiaries. We note that CMS has released data for 2006 and 2007 prior to the release of this report. As Chapter 2 on methodology explains, there are multiple sources for data, and more detailed breakdowns of enrollment data are available for 2006 than had been available in previous years. In addition, approaches to data presentation can vary. For example, as we explain in both the text and footnotes, in a number of tables we excluded enrollment in U.S. Territories, as well as enrollment in employer plans. By contrast, CMS reports on total MA enrollment typically reflect all MA enrollees. Readers should bear these types of data issues in mind if they are using data from other sources. In addition, data from different sources used for this report may result in slight differences in some tables. While we do not believe the differences materially affect the key findings about MA plan enrollment, plan availability, benefits and premiums, we want to advise readers about these data issues.

This project is divided into two phases. The first phase of this project produced a Report to Congress that "described the impact of additional financing provided under this Act (i.e., the Medicare Modernization Act [MMA]) and other Acts (Balanced Budget Refinement Act of 1999 (BBRA) and Benefits Improvement and Protection Act of 2000 [BIPA]) on the availability of Medicare Advantage Plans in different areas and its impact on lowering premiums and increasing benefits under such plans." This report was completed in late 2005 and was transmitted to Congress.⁵ The Report to Congress analyzed trends in the MA program from 2000 through 2005.

The second, and current, phase of this project focuses on monitoring the MA program from 2006 through 2008. This first interim report presents analyses of the program in 2006 in three key areas: plan availability, plan premiums and benefits, and beneficiary enrollment. The next section, 1.2, briefly reviews the key findings from the first phase of this project (2000-2005) as background for this report. Section 1.3 describes the major legislated changes in the MA program taking effect in 2006; they provide an important focus and context for this report. Section 1.4 outlines the goals and objectives of this report. Section 2 describes the methods, including data sources, that were used for this report. The remaining sections present the findings. Section 3 presents findings on plan availability, Section 4 on premiums and benefits, and Section 5 on beneficiary enrollment.

1.2 Review of Key Project Findings 2000–2005

Historically, payments to Medicare health plans were tied to local FFS per capita costs. The Balanced Budget Act of 1997 (BBA) fundamentally changed the method for setting rates used to pay Medicare health plans. BBA established a minimum floor for capitation rates, introduced a blended national/local rate, and limited rate updates in counties with higher rates in an attempt to narrow geographic payment differences. Following BBA, and prompted in part by the limited rate updates in counties with higher rates, large numbers of health plans withdrew from the Medicare program, constricted service areas, raised premiums, and/or reduced benefits. Partly in response to these developments, Congress enacted several laws to refine and modify the payment provisions of the BBA, including the Balanced Budget Refinement Act of 1999 and the Benefits Improvement Protection Act of 2000. However, the next fundamental change in the Medicare health plans program was the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA). The MMA set 100 percent of estimated FFS costs as a minimum payment level in each county, raised payment update amounts, and increased urban and rural floor rates. These changes raised MA payment rates, especially in large, high-cost urban areas. In the previous work of this project (Pope et al., 2006), the following key developments in the MA program from 2000 to 2005 were documented in response to these legislative changes:

Plan Availability

• Medicare plan availability decreased substantially after the implementation of the BBA, and despite interim legislation (BBRA and BIPA) aimed at addressing some of the effects of the BBA, availability of plans did not improve until after the MMA.
  
  
• Managed care availability (HMO and PPO) outside of large and medium urban areas improved under the MMA, but remained relatively weak in these areas. However, access to PFFS plans increased considerably in all areas, especially rural areas.

Plan Premiums, Benefits, and Cost Sharing

• Plan premiums and cost sharing generally increased, and benefits decreased, in response to the BBA. These conditions improved after passage of the MMA, with many plans lowering premiums and cost sharing, and improving benefits, after the March 2004 MMA payment increases.

Enrollment

• Although Medicare Advantage plan enrollment continued to grow through 1999, it declined steadily between 2000 and 2003, and rebounded somewhat in 2005 after the passage and full implementation of the MMA.
  
  
• Enrollment in urban counties continued to dominate the Medicare Advantage program throughout this time period. Enrollment in rural counties improved slightly as of 2005, though overall rural enrollment remained small.

In short, the context for developments in 2006 is that the MA program had declined in the early years of this decade, but had begun to rebound in response to the MMA payment increases taking effect in 2004 and in anticipation of the additional major changes mandated by the MMA for 2006.

1.3 Managed Care Legislative Mandates

A primary focus of this project is the impact of legislated changes on MA plan availability, premiums and benefits, and beneficiary enrollment. This section describes MA legislative provisions taking effect in 2006.

Although the MMA was passed in 2003, many of its most far reaching mandates relevant to the Medicare Advantage program did not become effective until January 1, 2006. The MMA mandates effective in 2006 can be summarized in three categories: bid-based payment methodology, mandate for Part D benefits in MA coordinated care plans, and implementation of new plan type with different operational conditions (regional PPO plans). The details of each of these three categories are discussed below.

Bid-Based Payment Rate Methodology: Beginning in 2006, reimbursement rates to Medicare Advantage plans are no longer based solely on administratively set prices. Instead, payment to plans is based in part on bids submitted by plans, which are compared to county benchmark amounts. The bidding process begins with a county benchmark amount calculated by CMS and released to plans in advance of preparation of their bids. Plan bids include three basic components: projected costs for Part A and B services, projected costs for non-Part D supplemental benefits, and projected costs for the Part D benefit if the plan includes Part D benefits. MA organizations offering coordinated care plans must offer at least one plan that has Part D benefits. How a plan's bid for Parts A and B services compares to the county benchmark determines whether the plan will need to charge an additional beneficiary premium and/or if they will have "rebate" dollars. When the plan bid for Part A and B services falls above the benchmark, the plan receives the benchmark as payment and must charge an additional premium for the amount above the benchmark. If the plan bid falls below the benchmark, 25 percent of the difference is retained in the Medicare trust funds, and the plan must apply the remaining 75 percent, called a rebate, to subsidizing the Part B or Part D premium and/or adding benefits or reducing enrollee cost sharing. Rebate dollars identified through this process (based on Parts A and B services) may also be applied to lower premiums and costs for Part D benefits, and our discussions with organizations through the Medicare Part D payment demonstrations suggests this is a common practice (Greenwald et al., 2007). The county benchmark is based on prior year county rates updated by an actuarially projected national per capita Medicare expenditure growth rate. In addition, at least every 3 years CMS is required to recalculate county FFS estimates. In those years, if the FFS amount is higher than the updated rate for the prior year, the FFS amount is the bench mark for that year. Local plans proposing to serve multiple counties submit bids based on an average of projected costs in each county, weighted by projected enrollment. The comparative benchmarks are similarly averaged and weighted by projected enrollment. Finally, for Part D payments (submitted as a separate bid), a separate Medicare payment is made to MA plans based on bids received. In addition to the base Medicare payment, additional payment variations, such as reinsurance and risk corridors, apply to these separate Part D payments (this is discussed in more detail below).

Bidding for regional PPOs is somewhat different. The bids of these plans, which must serve entire MA regions (there are now 26 specifically defined MA regions), are considered in setting the MA regional benchmark. Regional PPOs also have access to a special Medicare Advantage stabilization fund, available for organizations that offer regional PPOs in underserved areas. Section 221 of the MMA added Section 1858(e) to create this regional PPO stabilization fund. The purpose of the stabilization fund is to attract and retain regional PPO to areas with limited MA penetration. CMS was authorized to make payments as of January 1, 2007 to plans that participate in MA regions where no other plans were offered in the prior year. However, the Tax Relief and Health Care Act of 2006 delayed CMS' implementation of the fund until 2012.

Despite these substantial changes in payment methodology for Medicare Advantage plans, other aspects of the Medicare Advantage payment systems remain. County rural and urban floor rates, now relevant for benchmarking purposes, continue as minimum percentage increase rates, with the 2004 floors "preserved" in subsequent years because they are trended forward by the national Medicare growth percentage, and the majority of counties now fall under one of these rates. Risk adjustment under the CMS-HCC system continues, though the blend increased to 75 percent in 2006 and 100 percent in 2007 and beyond.

Mandatory Part D Benefits: The second major change affecting Medicare Advantage was the mandatory inclusion of Part D benefits for MA plans except PFFS plans (optional) and Medical Savings Account (MSA) plans, beginning in 2006. In this first year, the Part D defined standard prescription drug benefit includes an annual $250 deductible that the beneficiary is responsible for paying. Between $251 and the initial coverage limit of $2,250 in total drug costs, the Part D plan is responsible for 75 percent of costs and the beneficiary pays a 25 percent coinsurance. There is no coverage after the initial plan coverage limitation is met. Beneficiaries are responsible for all costs between the initial coverage limit and when they reach a $3,600 threshold in true out-of-pocket costs (TrOOP).⁶ Catastrophic coverage begins at the attachment point or threshold of $3,600 in TrOOP. Costs in catastrophic coverage are split three ways, with the government providing reinsurance equal to 80 percent, the Part D plan covering 15 percent, and the beneficiary paying the greater of a 5 percent co-insurance, or co-payments of $2 for generic drugs and $5 for non-generic drugs.

Coverage for the prescription drug benefit is provided through MA plans, which offer prescription drug coverage that is integrated with the health care coverage they provide to Medicare beneficiaries under Part C of Medicare. In each service area, MA plans must offer at least one plan with the Part D basic benefit, though they may offer only plan(s) with broader (enhanced) coverage if they are available to beneficiaries at no additional cost. If the minimum required level of coverage is offered, MA plans may also offer additional prescription drug benefits through enhanced alternative coverage for an additional premium. As described above MA plans may use Part A and B rebate amounts to subsidize Part D premiums and cost sharing.

Government payments to Part D plans are made through the following four mechanisms (CMS, 2005a): (1) the direct subsidy equals the standardized bid amount, adjusted for the risk characteristics of the enrollee, minus the monthly beneficiary premium for basic benefits; (2) reinsurance subsidies are equal to 80 percent of the allowable reinsurance costs attributable to prescription drug costs after the Part D enrollee has incurred True Out of Pocket Costs (TrOOP) that exceed the annual out-of-pocket threshold; (3) low-income subsidies are government payments on behalf of certain beneficiaries based on their income and asset levels that cover part or all of the premium subsidy amount and plan cost sharing; and (4) risk sharing arrangements involve symmetrical risk corridors in which the government either pays more of plan costs or recovers payments when a plan has allowable risk corridor costs above or below a target amount by certain percentages.

Implementation and Proliferation of New Plan Types: Effective in 2006, the MMA introduced a new plan type that operates under different implementation rules and modified payment methodologies: regional PPOs. In addition, the MMA consolidated former demonstration plans and created rules for new plans that meet the needs of vulnerable populations. These plans, called special needs plans, became more widely available to beneficiaries in 2006.

Regional PPOs, like local PPOs that were allowed to operate under Medicare since the BBA, offer beneficiaries the option of obtaining services from non-network providers at additional cost. The distinguishing characteristic of regional PPOs is that they are required to offer a uniform plan (or multiple plans) for a single premium for an entire MA Region. The 26 MA regions were defined and announced in December 2004, and are generally comprised of large single states or groups of two or more smaller states. The concept behind creation of the regional PPO option was that the requirement to serve large state-based service areas (rather than selecting service areas county by county) would improve Medicare beneficiary access to coordinated care plans, especially in rural areas. Regional PPO plan payments, though based on a bidding and benchmark process, are calculated on a slightly different basis than those for local coordinated care plans; the regional PPO benchmark takes into account the regional PPO bids, and is not based entirely on MA county capitation rates.

The second plan type authorized under the MMA, and which became more widely available in 2006, is special needs plans (SNPs). SNPs are a new category of plans that offer benefits and services aimed at one (or more) of three types of beneficiaries: dually eligible, institutionalized, or severely chronically ill. A SNP must be first approved as an MA contractor. Many SNPs are former demonstrations (Evercare and dual eligible demonstrations which operated in Minnesota, Wisconsin, and Massachusetts), though new options have also been established particularly beginning in 2006. SNPs are required to offer Part D and other statutory benefits, but are also required to include additional benefits (e.g., disease management) tailored to the targeted special needs population. Because of the targeting of benefits to the special needs beneficiaries, SNPs (unlike any other regular MA plan) are allowed to restrict enrollment to this population. Enrollees of SNPs are also allowed more flexibility in enrollment and disenrollment compared to regular MA plan enrollees. Although SNPs are paid using the same methodology as regular MA plans, in 2006 increased proportions of payments will be based on the CMS-HCC risk adjustment methodology, which, since special needs beneficiaries typically have higher costs, should result in higher payments to SNPs.

A third new plan type, medical savings accounts (MSA), originally authorized in the BBA of 1997, will first be offered to beneficiaries in 2007. The MSA option combines a high deductible health plan with a medical savings account which beneficiaries can access to pay for non-catastrophic expenses. This option will be offered both in the regular MA program, as well as in a demonstration program. Because this option does not begin until 2007, it is not included in this report.

1.4 Goals of this Report

The implementation of the 2006 legislative mandates for the MA program, as well as the continued influence of past legislative mandates, have impacted the MA program. The goal of this report is to document MA plan availability, premiums and benefits, and beneficiary enrollment in 2006 as they evolved in response to these legislative changes and other factors. Where feasible, the 2006 developments were put in the context of trends documented for 2000 to 2005 in prior work of this project. This report focuses especially on key new features of MA for 2006, including Part D prescription drug benefits, the regional PPO plans, and the more widely available SNP plans.

SECTION 2
METHODOLOGY

2.1 Overview

In this section, we provide an overview of our methodological approach for monitoring the Medicare Advantage program in 2006. Additional methodological detail specific to certain analyses is presented in subsequent sections of this report. Our quantitative analyses were performed on CMS administrative data. We describe here the primary methodological definitions, approaches, issues, challenges, samples, and data sources used in our analyses.

2.2 Contracts and Plans

In the report, we conducted analyses at both the MA contract and plan level. The term "contract" refers to a contract between an "MA organization" (typically an insurer) and CMS to enroll Medicare beneficiaries and provide them with medical services in a defined geographic area. The term "plan" refers to a specific benefit package and premium offered by an MA organization in specific counties. Several "plans" may be offered by the same contract (MA organization) in the same county—for example, a plan including the Part D drug benefit and a plan without a drug benefit. In some sections of this report, such as in Section 3 where we are analyzing the availability of MA options to beneficiaries, our unit of analysis is generally the contract. However, since benefits and cost sharing vary by plans within overall contracts, the unit of analysis in Section 4 is the plan, weighted by plan enrollment.

One of our major analytical variables in this report is "plan type," that is, HMO, local PPO, regional PPO, or PFFS. Each MA contract contains only one of these plan types (although a contract may contain multiple plans of the same type). So contracts, as well as plans, may be classified into the plan types and analyzed on that basis. HMO point of service (POS) plans may be offered by HMO contracts and are grouped with them in our analyses. We also group the uncommon "provider sponsored organization" (PSO) plan type with HMOs in our analyses. PSOs are HMO-like plans that are sponsored by a provider organization rather than an insurer.

One important type of MA plan—special needs plans or SNPs—is not also a contract type. SNPs are defined by their targeted population, not by their provider network requirements. A MA contract may offer both SNP and non-SNP plans, or only one or the other. SNPs are allowed to restrict enrollment to their targeted population whereas other non-employer-only MA plans must enroll any beneficiary eligible for MA. We therefore refer to non-employer-only, non-SNP plans as "open access" plans. In our analyses, SNPs are sometimes distinguished as a separate category and sometimes combined with open access plans in other categories such as total MA, HMOs, etc. PFFS plans cannot offer a SNP.

2.3 Types of Plans Analyzed

Our analysis focuses on Medicare Advantage plans. The Medicare law specifies three types of MA plans: (1) coordinated care plans, which include HMOs (with or without a point-of-service option), local and regional PPOs, and provider-sponsored organizations; (2) private fee for-service plans; and (3) Medical Savings Account (MSA) plans. We discuss these options below, except for MSA plans, which were not offered in 2006:

• Health Maintenance Organizations (HMOs)—a traditional form of Medicare coordinated care contract in which enrollees are covered only for services received from a defined network of participating providers. Enrollees usually must choose a primary care provider who authorizes all or most services. A variant of HMOs are HMO/Point of Service (POS) plans, in which out-of-network coverage is available with higher cost sharing on a service-by-service basis.⁷
  
  
• Local Preferred Provider Organizations (local PPOs)—a variant of coordinated care contracts in which non-network health care providers are covered with increased cost sharing. In-network providers can be accessed without referrals from a primary care provider. Local PPOs define their service areas on a county-by-county basis. As of 2006, the Medicare PPO demonstration plans that began prior to 2006 converted to local PPO status. Prior to 2006, we included the PPO demonstration plans in the local PPO category.
  
  
• Regional Preferred Provider Organizations (regional PPOs)—regional PPOs are coordinated care plans and were new to Medicare in 2006. Like local PPOs, regional PPOs offer out-of-network services for additional cost sharing, and do not require in-network referrals. But regional PPOs must offer a uniform product(s), at the same premium(s), in an entire MA region rather than defining their service area on a county-by-county basis.
  
  
• Special Needs Plans (SNPs)—SNPs are coordinated care plans that target beneficiaries with special needs. They can be offered through HMOs or local or regional PPOs. The three types of SNPs are targeted at dual Medicare/Medicaid eligibles, institutionalized beneficiaries, or beneficiaries with a severe chronic or disabling condition. Unlike other MA plans, SNPs are allowed to exclusively enroll or enroll a disproportionate percentage of their target group of beneficiaries. SNPs must provide services tailored to their special population. All SNPs are required to offer Part D drug benefits.
  
  
• Private Fee–For-Service (PFFS)—Most PFFS plans do not have a defined provider network. Enrollees are covered for services from any provider willing to accept the payment terms of the PFFS plan. Enrollee cost sharing for services may differ from traditional Medicare. Providers are paid on a FFS basis, at the traditional Medicare payment rates or higher.
  
  There are a few PFFS plans with a network of providers (providers who have a contract with the plan) for some or all categories of services. Enrollees can still see out-of-network providers willing to accept the payment terms of the PFFS plan, but they may have higher cost sharing. Payment to contracted providers may be less than the traditional Medicare payment rates.
  
  Unique among MA plans, PFFS contracts are not required to offer plans with a Part D benefit (and MSA plans are not permitted to offer Part D; MSA plan enrollees can sign up for Part D). Also unlike other MA plans, PFFS and MSA plans are not considered coordinated care or managed care plans, and federal regulations prevent them from offering SNPs.

Unless otherwise noted, we did not include non-MA plans in our analyses. Non-MA plans include demonstration, cost reimbursement, and other plan types. Non-MA plans often have unique payment arrangements, enrollment limitations, or benefit design features not found in MA plans. We also excluded employer-only plans from our analyses. Employer-only plans are restricted to enrollees sponsored by specific employers, typically retirees of a specific employer, and are tailored to that employer's situation. In 2006, we were able to exclude enrollment from employer-only plans completely because of the availability of plan-level enrollment data. Prior to 2006, only contract-level enrollment was available, and we could not exclude enrollment from MA contracts that offered a mix of employer-only and non-employer plans. Finally, we included only plans that were Part A/B plans.

2.4 Enrollment Weighting of Premiums and Benefits

Unless otherwise noted, our analyses of MA plan premiums and benefits are weighted by plan enrollment. They reveal what premiums MA enrollees paid, and what benefits they received, on average. Enrollment-weighted premiums and benefits reflect both plan offerings and beneficiary choices among available plans. An unweighted analysis, or an analysis weighted by the number of Medicare program enrollees in an area (MA and non-MA), would reflect plan offerings only. An unweighted analysis would count a plan with one enrollee the same as a plan with one million enrollees.

Our previous trend analyses of 2000 to 2005 were limited to basic HMO plans, defined as the lowest-premium plan offered by an HMO contract in a county (Pope et al., 2006). We examined HMOs because we wanted to examine effects of payment changes on trends in the premiums and benefits of a consistent plan type over time. We selected the single basic HMO plan because our analyses were enrollment-weighted and only total contract enrollment, not enrollment for each plan offered by a contract, was available.

For 2006, enrollment weights by contract and plan within contract were newly available. For 2006, we no longer needed to utilize the concept of "basic HMO plan," but rather included all plans in our analysis, weighting each by its enrollment. Our MA totals for premiums and benefits in 2006 include HMOs, PPOs, and PFFS, and include all plans in each contract, not just the lowest-premium plan. The ability to analyze all plans weighted by enrollment gave us a more accurate picture of the premiums paid and benefits received by the average MA enrollee. This is increasingly important as the number of plan types and options proliferates, and provides a basis for examining MA trends from 2006 to 2008 in the remainder of this project.

As a consequence of including all plans in our 2006 premiums and benefits analysis, our 2006 premiums and benefits data are not comparable to premiums and benefits for basic HMO plans from our earlier work (Pope et al., 2006). Hence, our 2006 premiums and benefits analysis is limited, for the most part, to a cross-sectional analysis of 2006. Trend analysis of MA premiums and benefits for 2000 to 2006, or even 2005 to 2006 is not possible.

Even if we had not made the change in enrollment weighting, comparison of 2006 premiums and benefits to earlier premiums and benefits would have been problematic because of the introduction of Part D in 2006. With the advent of Part D, MA plans' prescription drug benefit is separately priced (through the Part D premium); the Part C premium now covers only medical benefits. Previously the drug benefit, if any, was covered by the single Part C premium. Thus, the benefit package covered by the Part C premium has changed, and Part C premium time trends pre- and post-2006 are not comparable. Part D premiums, of course, did not exist before 2006.

2.5 Geographic Areas

In our analysis of plan availability, number and percentage of counties is a key measure of the availability of types of plans. We have data on approximately 3,120 counties throughout our time period (2000–2006). The number of counties may vary slightly for different tables, analyses, or years due to availability of data for several counties. One issue is Broomfield County, Colorado, which was created in 2003, and thus did not exist throughout our study period. Another issue involves counties in Alaska that were not coded consistently across different data sources. To address the latter, we created a single aggregate "county" for "rest of Alaska," which comprises Alaska excluding Anchorage, Juneau, and Fairbanks. Data were not always available for these Alaska "counties" that we created. The Social Security Administration county codes that we used include two county codes for Los Angeles county in California. We combined these into a single Los Angeles county code.

We excluded U.S. territories (Puerto Rico, the Virgin Islands, Northern Marianas, American Samoa and Guam) from all of our analyses.

In addition to national and county-level analyses, we grouped counties by urbanicity and region to examine aggregated impacts by type and location of county. We defined five categories of urbanicity based on the "Beale" codes created by the U.S. Department of Agriculture for the year 2003 based on the 2000 Census. The categories included the following:

• Large urban: counties in metropolitan areas of 1 million or more
  
  
• Medium urban: counties in metropolitan areas of 250,000 to 1 million
  
  
• Small urban: counties in metropolitan areas of less than 250,000
  
  
• Rural, urban-adjacent: non-metropolitan counties adjacent to at least one metropolitan county
  
  
• Rural, non-adjacent: non-metropolitan counties not adjacent to any metropolitan counties

Our regional definition was the four US census regions:

• Northeast
  
  
• Midwest
  
  
• South
  
  
• West

2.6 Beneficiary Sample

Our analysis focuses on options available to Medicare beneficiaries. However, since individuals diagnosed with End Stage Renal Disease (ESRD) are excluded from enrolling in an MA plan (they can however remain in a plan if they are diagnosed after enrollment), we have excluded this population from our analyses that look at penetration and Medicare eligible populations.

2.7 Timing of Data

In our earlier work on 2000 to 2005 we were unable to obtain a consistent month of the year for trend analyses because of data limitations (see Pope et al., 2006 for more details). For 2006, we chose to obtain data for July 2006, the midpoint of the year. July was after the special initial open enrollment period for Part D plans ended in May 2006. Our data represent a point-in-time sample for July 2006, not an "ever enrolled" in 2006 sample.

2.8 Trend Versus 2006-Only Analyses

Our intention was to build on our earlier work for 2000 to 2005 by adding results for 2006, and analyzing trends for 2000 to 2006. For our analysis of plan availability in Section 3, this was largely possible because the necessary data were consistent over time. For the premiums and benefits analysis of Section 4, and the enrollment analysis of Section 5, trend analysis was more problematic.

We discussed above, in Section 2.4, that trend analysis of premiums and benefits proved to be infeasible for two reasons: (1) inclusion of all MA plans in the 2006 analysis versus only basic HMO plans prior to 2006; (2) the introduction of Part D in 2006, which changed the premium and benefit structure of MA plans. We make some comments in Section 4 about how 2006 MA premiums and benefits may compare to earlier MA premiums and benefits, but we do not attempt a systematic trend analysis. Our premiums and benefits analysis is a cross-sectional study for 2006. We will be able to conduct trend analysis of upcoming years as part of this project, comparing 2007 and 2008 to 2006.

Trend analysis of MA enrollment also proved to be difficult. Our 2000 to 2005 enrollment analyses utilized the Medicare Enrollment Database (EDB). In 2006, we began using the Medicare Beneficiary Database (MBD, described in more detail below). Enrollment trends from the two databases were inconsistent. In part, the incomparability between the EDB and MBD enrollments was probably due to our ability to perfectly exclude employer-only plan enrollment in 2006 with the MBD, compared to our imperfect exclusion for 2000 to 2005 with the EDB (see Section 2.3 for further discussion of excluding employer-only plans).

Instead of using the EDB/MBD to measure enrollment trends, we used CMS' Medicare Monthly Contract Reports, available on CMS' Web site. We believe that these data provide a consistent enrollment series. However, they are aggregate data, and do not allow us to apply all the plan sample restrictions that we used in other analyses. For this reason, our enrollment trends analysis is not fully consistent with our cross-sectional enrollment analysis for 2006. For the 2006 analysis, we utilized the MBD, and applied the full set of plan restrictions that we used in the rest of the report. In future reports, we plan to use the MBD for enrollment trend analysis of 2007 and 2008 compared to 2006.

2.9 Data Sources

CMS Health Plan Management System (HPMS). The primary data source used in our analyses was CMS's Health Plan Management System (HPMS), which collects service area, premium, and benefit information for MA and certain other plan types. This information is submitted by plans annually, or more frequently if the data changes. The HPMS Plan Benefit Package (PBP) datasets are available for each month, and contain information describing the benefit package provided by each plan, including information on premiums, copayments, coinsurance and deductible amounts, and drug and other benefit descriptions. The HPMS data was used for the plan availability and plan premiums and benefits analyses. We used July 2006 HPMS/PBP extracts.

HPMS Plan Enrollment Data Extract. Because of delays in obtaining the MBD enrollment data, RTI completed national-level premiums and benefits analyses using an enrollment weight from the Plan Enrollment Data Extract from HPMS. Like the MBD, these data include enrollment at the individual plan level, rather than just the contract level. But they are not available at the contract/plan/county level, and thus the MBD was used to develop an enrollment weight for analyses including a geographic component (e.g., urbanicity, region). For most plans, the HPMS and MBD enrollment data are very similar, but differences are larger for a few plans, perhaps due to differences in the timing of when data feeds from plans are reflected in the two data sources. Thus, premiums and benefits results using an HPMS enrollment weight versus an MBD enrollment weight are very similar and consistent, but are not identical.

Medicare Beneficiary Database. The Medicare Beneficiary Database (MBD) is a beneficiary-level CMS database that contains extensive information about Medicare beneficiaries, including Medicare program enrollment information, Medicare health plan enrollment, Part D enrollment, and beneficiary demographic characteristics. The MBD was used to obtain a contract/plan/county enrollment weight for premium and benefit analyses by urbanicity and region. The MBD was also used for the 2006 cross-sectional enrollment analyses.

Medicare Denominator File. The Medicare Denominator File was used to calculate counts of Medicare beneficiaries eligible to enroll in Medicare Advantage. Eligibility counts were needed for several of our analyses, including descriptive analyses of number of Medicare beneficiaries with access to MA plans, and percentage of Medicare beneficiaries enrolled in MA plans (MA penetration).

Medicare Monthly Contract Reports. These reports, available on CMS' Web site, report a time series of enrollment for MA plans aggregated by organization. These data were used for the trend analysis of MA enrollment.

2.10 Data Consistency and Quality Issues

Developing the analytical data files for this report required merging multiple data sources from the HPMS and MBD. The data from different source files were not always fully consistent (e.g., a small number of plans or counties might not match between data files). We merged files and reconciled data as completely as possible, and merges were usually perfect or nearly so. But because of a small number of non-merges in some instances, the sample (number) of plans, counties, or enrollees may differ slightly among some tables, years, variables, or analyses in this report. These minor inconsistencies should not have any material effect on the results that we report.

In some cases, we found that variables were not reported accurately in the source data. For example, not all MA plans may have responded to certain items on the HPMS/PBP, and certain MBD fields did not contain usable data. If data fields did not appear to be substantially complete and accurate, they were not used in our analyses.

SECTION 3
PLAN AVAILABILITY

3.1 Introduction

One of the primary goals of the legislative initiatives from the BBA through the MMA was to expand the number and type of Medicare health plans available to Medicare beneficiaries, particularly in geographic areas (such as rural counties) that have traditionally been underserved by managed care. Therefore, in this chapter, we describe changes in plan availability between 2000 (after the BBA and BBRA were implemented) and 2006 (2 years after initial MMA provisions were implemented), focusing on recent developments in 2006. We examined changes in total number of contracts participating in Medicare Advantage (MA), contract availability by urban-rural and regional areas, and beneficiary access to different numbers and types of MA contracts.

3.2 Medicare Advantage Contracts by Plan Type: 2000–2006

3.2.1 Number of Contracts

First, we looked at the number of Medicare contracts, in total and by contract type, by year. Findings are presented in Table 3-1. In this analysis, we counted the number of contracts, not individual plans offered under these contracts.⁸

Table 3-1
Number of Medicare Advantage contracts, by plan type

	Nov-00	Jun-01	Apr-02	Apr-03	Feb-04	Jun-05	Jul-06
1 HMO includes HMO POS, 2006 also includes PSO. 2 Includes PPO demonstration contracts from 2003 to 2005. NOTES: SNPs incorporated by plan type. Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System data.
Total MA contracts	264	179	154	178	178	289	387
Total coordinated care contracts	263	178	152	175	175	275	366
HMO1	259	173	147	137	132	176	235
Local PPO2	1	2	3	35	40	93	120
Regional PPO	-	-	-	-	-	-	11
PFFS	1	1	2	3	3	14	21

A marked decrease in the number of Medicare contracts occurred in the post-BBA implementation period. In 2000, 2 years after the implementation of the BBA payment changes, there were a total of 264 MA contracts. By 2002, contracts declined to the lowest point in our analysis period, at 154 MA contracts nationally. In 2003, as a result of the start of the PPO demonstration, the total MA contracts rebounded slowly (increasing from 154 to 178). Total contracts were stable in early 2004 prior to implementation of MMA payment increases in March 2004.

The following year, 2005, was the first full year that MA organizations had the opportunity to take into account the MMA March 2004 and 2005 payment changes. In addition, the MMA created two other special factors in 2005. First, it temporarily prohibited new local PPO contracts in 2006 and 2007, the first two years that the regional PPO option was available. Hence, MA organizations interested in new local PPO contracts had to enter in 2005 or wait until 2008. Second, the MMA established a new Medicare drug benefit beginning in 2006, which may be offered by either MA plans or standalone drug plans. The new drug benefit created an opportunity for MA organizations to gain substantial new enrollment in 2006.

In response to these factors, the total number of contracts rose sharply—by about 62 percent—in 2005. All types of MA contracts increased, especially local PPOs, which more than doubled from 40 to 93. By June 2005, the number of MA contracts, exceeded the number of contracts at the beginning of our analysis period, 2000.

The sharp increase in the number of contracts continued throughout 2005 into 2006, with an increase in the total number of MA contracts to 387. Part of this increase in total 2006 MA contracts was due to the addition of the first regional PPO contracts (11 new in 2006). Other likely drivers were higher MA capitation rates under the MMA and the implementation of the Medicare Part D prescription drug plan. By July 2006, the number of HMO (adding 59) and local PPO (adding 27) contracts increased sharply over June 2005. The increase in PFFS contracts appearing in 2005 also continued into 2006, registering a 50% increase.

From 2000 to 2006, HMOs remained the dominant plan type of MA contract, but alternative types—especially PPOs—grew in importance. In 2000, 259 of the 264 MA contracts, or 98 percent, were HMOs. In 2006, HMOs were 235 of 387 MA contracts, or 61 percent. Local PPOs grew from 1 to 120 contracts from 2000 to 2006, and comprised 31 percent of MA contracts in 2006. Private fee-for-service plans accounted for only a small number of contracts throughout the period, but have also increased, especially in 2005 and 2006. PFFS contracts tend to cover very large service areas relative to other plan types.

3.2.2 Percent of Counties with at Least one Medicare Contract

Because one of the goals of the legislative changes was to improve Medicare beneficiary access to Medicare health care plans, we also analyzed for each year between 2000 and 2006 the percent of counties in which at least one Medicare contract was available. Our findings are shown in Table 3-2. Table 3-2 maps the contracts to counties served, and presents data on the proportion of counties with access by each type of plan.⁹ Continuing the trend from the last several years, each plan type was available in more counties in 2006 than in 2005.

Table 3-2
Percent of counties with at least one Medicare Advantage contract, by plan type

	Nov-00	Jun-01	Apr-02	Apr-03	Feb-04	Jun-05	Jul-06
1 HMO includes HMO POS, 2006 also includes PSO. 2 Includes PPO demonstration contracts from 2003 to 2005. NOTES: SNPs incorporated by plan type. Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System data.
Coordinated care plans
HMO1	25.9	20.3	19.1	17.8	18.5	29.0	33.8
Local PPO2	0.2	0.2	0.4	6.3	7.6	22.7	29.5
Regional PPO	-	-	-	-	-	-	89.9
Non-coordinated care plans
PFFS	52.7	52.7	51.6	54.9	40.6	92.9	96.0

In 2000, almost 26 percent of counties had at least one Medicare HMO contract. By 2003, that figure declined sharply to 17.8 percent. The percent of counties with access to an HMO rose slightly to 18.5 percent in 2004. In 2005, there was a sizeable increase in the percent of counties with at least one HMO contract, to 29.0 percent. In 2006, HMO availability rose moderately, and beneficiaries had access to HMOs in just over a third of all counties compared to just over a quarter at the beginning of the decade in 2000.

The percentage of counties with a local PPO contract remained low until the start of the PPO demonstration in 2003, and increased from that point. In 2003, 6.3 percent of counties had access to a local PPO, increasing from less than one percent the year before. Counties with access to a local PPO increased sharply in 2005. In 2005, 22.7 percent of counties had access to a PPO, the first year in which the number of counties with a PPO approached the number of counties with an HMO. In 2006, 29.5 percent of counties had access to a local PPO, as even more PPOs entered or expanded in the MA program in late 2005 before the PPO moratorium for 2006 and 2007 took effect.

The number of counties with access to a PFFS plan was quite large, particularly considering the relatively small number of PFFS contracts. In 2000, though there was only one PFFS contract, through this contract 52.7 percent of counties had access to a PFFS plan. The structure of the PFFS option appeared to favor large service areas under a single contract umbrella, possibly because of the lack of the need to establish local provider networks under PFFS plans. Though the number of PFFS contracts increased to three by 2004, the number of counties with access to a PFFS plan actually decreased that year to 40.6 percent, suggesting that PFFS plans had reduced the number of counties in their service areas. However, by 2005, both the number of PFFS contracts and the number of counties with access to a PFFS plan increased significantly. In 2005, 92.9 percent of counties had access to a PFFS plan, making PFFS options the most accessible MA option for Medicare beneficiaries. This trend continued in 2006, where with 21 contracts, PFFS plans were available in 96 percent of counties.

Like PFFS plans, regional PPOs represented a small percentage of the number of contracts, but due to large service areas, offered accessibility to a large proportion of the Medicare population. In their first year, regional PPOs accounted for only 11 contracts in 2006, but were available in almost 90 percent of all counties.

3.2.3 Number and Percent of Beneficiaries with Access to a Medicare Contract

In addition to the percent of counties with access to a Medicare plan, we considered the number and percent of Medicare beneficiaries with access to a contract. Just as counting the number of contracts can give an incomplete picture, counting counties does not take into account the number of beneficiaries residing in each county. Table 3-3 addresses this by counting the number of Medicare-eligible individuals in each county, and calculating the proportion of eligibles that have access to each contract type. In looking at the trends in Table 3-3, it is important to note that the data source changes after 2004. The results for 2000–2004 were drawn from data that was formerly posted on the CMS Web site; results for 2005 and 2006 were drawn from the CMS Denominator files.¹⁰

Table 3-3
Number and percentage of Medicare beneficiaries with access to a Medicare Advantage plan, by plan type

I. Number Plan Type	Nov-00	Jun-01	Apr-02	Apr-03	Feb-04	Jun-05	Jul-06
1 HMO includes HMO POS, 2005 and 2006 also includes PSO. 2 Includes PPO demonstration contracts from 2003 to 2005. NOTES: Medicare beneficiaries by county prior to 2005 were obtained from the CMS Web site; beneficiaries from 2005 and 2006 were obtained from the Medicare Denominator file. Beneficiaries include those eligible to enroll in a Medicare Advantage plan. SNPs are incorporated by plan type. Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Web site, Denominator file, and Health Plan Management System data.
MA Plans	33,300,258	32,958,996	32,305,226	32,841,281	31,774,507	37,334,895	38,766,667
HMO1	27,233,843	25,646,057	24,754,752	24,042,140	25,160,074	26,713,737	28,157,310
Local PPO2	598,318	864,952	1,693,642	9,625,333	10,660,896	21,382,705	25,083,176
Regional PPO	-	-	-	-	-	-	34,426,846
PFFS	15,223,535	15,443,348	14,862,682	15,490,096	13,037,695	28,681,100	31,570,787
II. Percent Plan Type	Nov-00	Jun-01	Apr-02	Apr-03	Feb-04	Jun-05	Jul-06
MA Plans	83.3%	80.9%	78.3%	78.5%	74.8%	97.7%	99.6%
HMO1	68.1	62.9	60.0	57.4	59.2	69.9	72.3
Local PPO2	1.5	2.1	4.1	23.0	25.1	56.0	64.5
Regional PPO	-	-	-	-	-	-	88.5
PFFS	38.1	37.9	36.0	37.0	30.7	75.0	81.1

The percent of beneficiaries with access to HMOs and local PPOs was much higher than the percent of counties with access, throughout our analysis period. This is because offerings of HMOs and local PPOs are concentrated in populous urban counties. Conversely, the percent of beneficiaries with access to PFFS is lower than the percent of counties with access, because PFFS service areas were concentrated in less populous rural counties. For regional PPOs, the percent of beneficiaries and counties are almost the same because regional PPOs must be offered throughout entire regions comprising both urban and rural areas.

Considering all types of MA plans together, more than three-quarters of beneficiaries had access to at least one MA plan throughout the 2000-2004 period, though the percentage with access declined from 2000 to 2004. In 2005, the pattern of declining access reversed dramatically, and virtually all beneficiaries (97.7 percent) had access to at least one MA contract. This trend continued in 2006 when 99.6 percent of beneficiaries had access to at least one contract. These high percentages in 2005 and 2006 were driven by the availability of PFFS plans (whose relatively small number of contracts provided access through very large service areas per contract) followed by the addition of regional PPOs in 2006. Regional PPOs, like PFFS, covered wide service areas and large numbers of potential Medicare enrollees through relatively few contracts.

By 2006, all MA plan types were available to a majority of Medicare beneficiaries. Over 72 percent of Medicare beneficiaries had access to an HMO in 2006—above the 2000 level. Nearly 65 percent had access to a PPO, 88.5 percent to a regional PPO and 81.1 percent to a PFFS plan. MA plan types were widely, though not universally, available in 2006.

3.2.4 Plan Availability by Urbanicity

To further analyze how the legislated payment changes impacted access to Medicare plans, we analyzed plan participation by urbanicity. In this analysis, we returned to the percent of counties as the measure of access rather than the percent of beneficiaries. We looked at the percent of counties with at least one HMO, local PPO, regional PPO or PFFS contract by a range of urban/rural categories, from 2000 to 2006. Our results are shown in Table 3-4. Table 3-4 stratifies counties by a measure of urbanicity (Beale Codes) developed by the U.S. Department of Agriculture. The total values (aggregated across all counties) differ in some instances from the results in Table 3-2 because SNPs have been broken out separately; in the categories in which there are no SNPs (i.e., PFFS) the results are the same as in Table 3-2.

Table 3-4
Percent of counties with at least one Medicare Advantage contract, by plan type and urbanicity

Urbanicity	Number of counties	Nov-00	Jun-01	Apr-02	Apr-03	Feb-04	Jun-05	Jul-06
1 HMO includes HMO POS, 2006 also includes PSO. 2 Includes PPO demonstration contracts from 2003 to 2005. 3 SNP are listed as a separate category, and not by plan type, e.g., an SNP HMO would be listed as SNP, and not counted as an HMO. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System data.
TOTAL	3,120
Any open access plan		-	-	-	-	-	-	99.5%
HMO1		25.9%	20.3%	19.1%	17.7%	18.5%	29.0%	30.6
Local PPO2		0.2	0.2	0.4	6.3	7.6	22.7	28.5
Regional PPO		-	-	-	-	-	-	89.9
PFFS		52.7	52.7	51.6	54.9	40.6	92.9	96.0
Special needs plan3		-	-	-	-	-	-	22.8
Urban	1,089
Any open access plan		-	-	-	-	-	-	99.4%
HMO1		51.8%	44.1%	39.5%	36.4%	38.1%	52.0%	55.4
Local PPO2		0.5	0.6	1.0	14.6	17.4	43.5	51.1
Regional PPO		-	-	-	-	-	-	90.0
PFFS		42.9	42.9	41.3	43.4	34.9	88.0	92.0
Special needs plan3		-	-	-	-	-	-	41.0
Large Urban	414
Any open access plan		-	-	-	-	-	-	99.8%
HMO1		75.8%	64.3%	58.5%	52.4%	55.3%	63.3%	66.4
Local PPO2		1.2	1.7	2.4	22.7	27.3	57.0	65.2
Regional PPO		-	-	-	-	-	-	91.5
PFFS		33.6	33.6	31.6	29.7	25.8	81.2	86.7
Special needs plan3		-	-	-	-	-	-	52.4
Medium Urban	324
Any open access plan		-	-	-	-	-	-	99.7%
HMO1		49.1%	44.4%	37.7%	37.0%	39.5%	58.6%	63.6
Local PPO2		0.0	0.0	0.3	13.9	16.4	46.3	57.7
Regional PPO		-	-	-	-	-	-	88.6
PFFS		50.3	50.3	48.5	51.5	41.7	92.0	95.7
Special needs plan3		-	-	-	-	-	-	47.5
Small Urban	351
Any open access plan		-	-	-	-	-	-	98.6%
HMO1		25.9%	19.9%	18.8%	16.8%	16.5%	32.5%	34.8
Local PPO2		0.0	0.0	0.0	5.7	6.6	25.1	28.5
Regional PPO		-	-	-	-	-	-	89.5
PFFS		47.0	47.0	46.2	52.1	39.3	92.3	94.9
Special needs plan3		-	-	-	-	-	-	21.7
Rural	2,031
Any open access plan		-	-	-	-	-	-	99.6%
HMO1		12.0%	7.5%	8.2%	7.7%	8.0%	16.6%	17.3
Local PPO2		0.0	0.0	0.0	1.9	2.3	11.5	16.4
Regional PPO		-	-	-	-	-	-	89.9
PFFS		57.9	57.9	57.1	61.1	43.7	95.5	98.1
Special needs plan3		-	-	-	-	-	-	13.0
Rural—Urban Adjacent	1,061
Any open access plan		-	-	-	-	-	-	99.5%
HMO1		18.9%	12.0%	11.0%	12.1%	12.6%	25.1%	24.9
Local PPO2		0.0	0.0	0.0	3.6	4.3	15.1	22.1
Regional PPO		-	-	-	-	-	-	90.9
PFFS		57.0	57.0	55.8	61.1	44.3	94.9	97.4
Special needs plan3		-	-	-	-	-	-	17.4
Rural—Not Urban Adjacent	970
Any open access plan		-	-	-	-	-	-	99.7%
HMO1		4.3%	2.7%	5.1%	3.0%	2.9%	7.4%	9.0
Local PPO2		0.0	0.0	0.0	0.1	0.1	7.5	10.1
Regional PPO		-	-	-	-	-	-	88.9
PFFS		58.9	58.9	58.6	61.0	43.1	96.1	99.0
Special needs plan3		-	-	-	-	-	-	8.2

From this analysis, a number of interesting trends emerged. Among open-access plans, the primary differences between urban and rural counties were largely confined to HMOs and local PPOs; the proportion of urban counties in which these plan types were available was nearly twice that of rural counties. This contrasts with regional PPOs and PFFS, which had relatively small differences in availability in urban and rural counties. The type of urban area played a large role in the availability of some plan types. For example, in 2006 more than 60 percent of medium and large urban areas were served by at least one HMO, while this value in small urban areas was less than 35 percent. In urban adjacent rural areas this figure was 25 percent, and in non-urban adjacent areas less than 10 percent. Clearly, population density was closely related to the viability of offering an HMO plan. A similar pattern emerged for local PPOs, another type of local provider-network-based, coordinated care plan.

SNPs exhibited a pattern similar to HMOs and local PPOs, albeit at a slightly reduced overall level of availability. SNPs were more common in urban areas (41 percent of counties) than rural areas (only 13 percent of counties). Within urban areas, the availability of SNPs was associated with the size of the urban area, and in rural areas, SNPs were more likely to be offered in urban-adjacent counties. Nationally, in 2006, SNPs were available in 22 percent of counties, and only in a majority of large urban counties.

We also noted interesting patterns among HMOs, still the dominant plan type in 2006 despite the continued growth of PFFS and PPOs. A larger proportion of large urban counties had at least one HMO every year between 2000 and 2006 compared to any other county type. However, between 2000 and 2006, the percentage of large urban counties with at least one Medicare HMO declined from 75.8 percent to 66.4 percent. Some of the decline likely arose from a substitution of local or regional PPO offerings for HMOs in large urban counties. Conversely, the percentage of medium urban counties with an HMO rose from 2000 to 2006, due to a large increase from 2004 to 2005 that continued into 2006. By July 2006, close to the same percentage of medium as large urban counties had access to an HMO. HMO access also continued to rise in 2006 in small urban counties, to a greater level than in 2000. Despite these increases in availability in urban counties, HMO availability in small urban counties remained poor, well below availability in larger urban counties.

3.2.5 Plan Availability by Census Region

To understand plan participation trends in different areas of the country, we analyzed plan availability by census region. Table 3-5 is a complement to Table 3-4 in the sense that counties are stratified by census region rather than urbanicity. Table 3-5 shows the percent of counties with different contract types in the Northeast, Midwest, South, and West.

Table 3-5
Percent of counties with at least one Medicare Advantage contract, by plan type and region

Census Region	Number of Counties	Nov-00	Jun-01	Apr-02	Apr-03	Feb-04	Jun-05	Jul-06
1 HMO includes HMO POS; 2006 also includes PSO. 2 Includes PPO demonstration projects from 2003 to 2005. 3 SNP are listed as a separate category, and not by plan type, e.g., an SNP HMO would be listed as SNP, and not counted as an HMO. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System data.
Northeast	217
Any open access plan		-	-	-	-	-	-	94.5%
HMO1		69.1%	60.4%	58.1%	57.1%	58.1%	63.1%	66.4
Local PPO2		2.3	2.3	2.3	32.7	34.1	56.7	71.9
Regional PPO		-	-	-	-	-	-	69.1
PFFS		30.9	30.9	30.9	30.9	30.9	46.1	74.7
Special needs plan3		-	-	-	-	-	-	60.4
Midwest	1,056
Any open access plan		-	-	-	-	-	-	100.0%
HMO1		17.4	16.0	16.6	14.4	14.9	27.8	30.6
Local PPO2		0.0	0.2	0.4	3.7	5.1	13.1	19.6
Regional PPO		-	-	-	-	-	-	100.0
PFFS		49.3	49.3	49.3	57.8	48.9	100.0	100.0
Special needs plan3		-	-	-	-	-	-	15.5
South	1,425
Any open access plan		-	-	-	-	-	-	100.0%
HMO1		24.8	16.4	13.2	11.6	12.9	23.4	24.1
Local PPO2		0.0	0.0	0.1	4.6	6.0	21.0	25.5
Regional PPO		-	-	-	-	-	-	100.0
PFFS		58.9	58.9	52.7	53.7	33.4	98.2	98.5
Special needs plan3		-	-	-	-	-	-	23.6
West	423
Any open access plan		-	-	-	-	-	-	99.3%
HMO1		28.1	23.4	25.3	26.4	26.2	33.3	33.8
Local PPO2		0.0	0.0	0.0	5.2	5.2	34.9	38.5
Regional PPO		-	-	-	-	-	-	41.4
PFFS		51.3	51.3	64.3	64.2	49.3	80.7	88.7
Special needs plan3		-	-	-	-	-	-	19.1

Of particular note is that all Medicare eligibles in the Midwest and South had access to an open access plan in 2006 due to the existence of regional PPOs. The PFFS option exhibited high penetration rates in these two regions as well: 100 percent in the Midwest and 98.5 percent in the South. These rates stood in contrast to those in the Northeast and West where regional PPOs were offered in only 69 percent of Northeastern counties and in only 41 percent of Western counties. Five of 26 MA regions where regional PPOs were not offered in 2006 were all in the West or Northeast: Alaska, Colorado/New Mexico, Connecticut/Massachusetts/Vermont, and Maine/New Hampshire. Similar to regional PPOs, PFFS plans were less available in Northeastern counties (75 percent) and Western ones (89 percent).

HMOs were most widely available in the Northeast (about two-thirds of counties in 2006 compared to one-third or fewer in other regions). From 2004 to 2005, HMO availability nearly doubled in the Midwest and South, rising from low levels. But HMO access growth was slow in all regions from 2005 to 2006.

In 2006, local PPOs were available in a substantially higher proportion of counties in the Northeast than other regions (72 percent versus 20 to 38 percent elsewhere). Similarly, the Northeast has a much higher percentage of counties served by SNPs than other regions (60 percent versus a maximum of 24 percent elsewhere). These results may be partly a consequence of relatively fewer rural counties in the Northeast.

3.3 Plan Choices Available to Beneficiaries in 2006

Tables 3-1 through 3-5 defined access to MA plans in the most basic way: if a single contract was available in a given county, a Medicare-eligible person was considered to have access to that type of plan. Our analyses focused on changes in this basic definition of access between 2000 and 2006.

In this next set of analyses, (presented in Tables 3-6 through 3-10); we considered the range and combinations of multiple plan choices available to beneficiaries in 2006, defining access beyond just availability of at least one plan type. It generally is believed that the broader the set of choices available to a beneficiary, the more likely that he or she can find a plan closely suited to his or her preferences. One aspect of the availability of choices is the degree to which alternative plan types are available to a beneficiary. For example, the availability of a single HMO plan and a single PFFS plan may comprise a greater degree of plan choice than the availability of two HMO plans without access to a PFFS plan. Tables 3-6 through 3-8 examine the range of choices available to beneficiaries in 2006 by looking at the various combinations of the major MA categories: HMO, PPO, and PFFS. In these tables, local and regional PPOs are combined because, though they have different service area requirements, to beneficiaries they offer a single type of benefit. In Tables 3-9 and 3-10, we considered yet another aspect of access, the numbers of contracts available to beneficiaries in various types of counties.

Table 3-6
Percentage of beneficiaries and counties with access to Medicare Advantage plan types, 2006

Plan types	Beneficiaries	Counties
1 HMO includes HMO POS and PSO plans. 2 Beneficiaries with no access to HMO, PPO or PFFS. 3 PPO includes local and regional PPOs. NOTES: SNPs incorporated by plan type. Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
No MA plans2	0.4%	0.5%
HMO only1	0.2	0.1
PPO only3	0.9	1.1
PFFS only	1.3	4.4
HMO & PPO1,3	17.5	2.3
HMO & PFFS1	0.6	0.6
PPO & PFFS3	25.1	60.2
HMO & PPO & PFFS1,3	54.1	30.8

Table 3-7
Percentage of beneficiaries with access to Medicare Advantage plan types, by urbanicity, 2006

Urbanicity	No plans2	HMO only1	PPO only3	PFFS only	HMO & PPO1,3	HMO & PFFS1	PPO & PFFS3	HMO & PPO & PFFS1,3
1 HMO includes HMO POS and PSO plans. 2 Beneficiaries with no access to HMO, PPO or PFFS. 3 PPO includes local and regional PPOs. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Urban	0.4%	0.2%	0.7%	0.3%	22.1%	0.5%	14.3%	61.7%
Large Urban	0.2	0.1	0.1	0.1	31.8	0.2	6.2	61.2
Medium Urban	0.2	0.5	0.5	0.0	9.4	0.6	15.6	73.2
Small Urban	1.3	0.0	3.3	1.6	4.1	1.4	46.5	41.8
Rural	0.5	0.2	1.6	4.8	0.8	1.0	63.9	27.1
Rural, urban adjacent	0.6	0.3	1.9	3.9	1.0	1.5	57.6	33.1
Rural, not urban adjacent	0.4	0.0	0.8	6.6	0.5	0.1	75.8	15.8

Table 3-8
Percentage of beneficiaries with access to Medicare Advantage plan types, by region, 2006

Census Region	No plans2	HMO only1	PPO only3	PFFS only	HMO & PPO1,3	HMO & PFFS1	PPO & PFFS3	HMO & PPO & PFFS1,3
1 HMO includes HMO POS and PSO plans. 2 Beneficiaries with no access to HMO, PPO or PFFS. 3 PPO includes local and regional PPOs. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Northeast	1.7%	0.9%	0.0%	3.2%	46.9%	1.4%	4.8%	41.1%
Midwest	0.0	0.0	0.0	0.0	0.0	0.0	33.6	66.4
South	0.0	0.0	0.7	0.0	3.5	0.0	40.2	55.7
West	0.4	0.0	3.1	3.5	35.2	1.6	6.7	49.5

Table 3-9
Percentage of beneficiaries with access to Medicare Advantage contracts, by number of contracts and urbanicity, 2006

Urbanicity	Number of contracts					Mean # contracts/county1
	0	1-3	4-6	7-9	10+
1 Weighted by eligibles in county. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Total	0.4%	17.6%	34.8%	22.3%	24.9%	7.48
Urban	0.4	9.9	32.9	25.5	31.4	8.42
Large Urban	0.2	3.4	27.7	28.8	39.9	9.87
Medium Urban	0.2	8.7	40.1	24.9	26.1	7.25
Small Urban	1.3	40.2	41.7	12.6	4.2	4.40
Rural	0.5	45.2	41.9	10.8	1.5	4.12
Rural, urban adjacent	0.6	40.8	43.4	12.9	2.3	4.37
Rural, not urban adjacent	0.4	53.4	39.2	6.9	0.0	3.65

Table 3-10
Percentage of beneficiaries with access to Medicare Advantage contracts, by number of contracts and region, 2006

Census Region	Number of contracts					Mean # contracts/county1
	0	1-3	4-6	7-9	10+
1 Weighted by eligibles in county. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Northeast	1.7%	12.6%	25.4%	28.7%	31.7%	8.31
Midwest	0.0	17.3	34.9	29.7	18.1	6.55
South	0.0	22.0	39.7	18.7	19.6	7.00
West	0.4	14.6	35.0	14.1	35.9	8.66

3.3.1 Choice Among Medicare Advantage Plan Types

Table 3-6 displays the number and percentage of beneficiaries facing each combination of plan choices, as well as the number and percentage of counties in which the particular combinations were offered. In 2006, over half of all Medicare beneficiaries lived in counties where HMOs, PPOs and PFFS were all offered; these counties represented 31 percent of all counties. PPOs in combination with either HMOs or PFFS (i.e., one or the other) were available to another 43 percent of beneficiaries (in 63 percent of counties). Only about 2 percent of beneficiaries had access to only one of these three plan types. As of 2006, very few Medicare beneficiaries (less than 0.4 percent) had no access to any of these three plan types. Put another way, 97 percent of beneficiaries have access to two or more plan types, including at least one coordinated care plan option. If a goal of the MMA was to increase the range of options available to Medicare beneficiaries, this analysis suggests that as of 2006, most Medicare beneficiaries had at least some choice among multiple plan types.

3.3.2 Choice Among Plan Types by County Urbanicity

One focus of the MMA was to increase beneficiary choices of MA plan types in rural and other underserved areas. Table 3-7 examines how access to combinations of plan types varied with county urbanicity in 2006. The percentages in the table are row percentages; that is the proportion of beneficiaries in the specific urbanicity category that have access to a particular combination of plan types.

Very few beneficiaries, particularly in large and medium urban locations, had access to only a single plan type. In urban regions, over 60 percent of beneficiaries had access to all three major plan types. By comparison, only slightly more than a quarter of rural beneficiaries had access to all plan types, due primarily to the paucity of HMO offerings in rural areas (Table 3-4). Beneficiaries in small urban areas were less likely to have access to all three plan types than residents of larger urban areas. This likely resulted from HMOs being less prevalent in lower population urban areas than in higher population ones, and is consistent with findings shown in Table 3-4 that found HMOs to be present in a substantially smaller proportion of small urban areas than medium and large urban areas. While availability of all three (HMO, PPO and PFFS) options was not as commonly found in small urban and rural areas, beneficiaries residing in these county types often had a choice between at least PPO and PFFS options. Growth of PPO options under the MMA through the regional PPO program may explain this finding.

Small and medium urban areas appeared to have more in common with rural areas than with large urban areas. About 88 percent of beneficiaries in small and medium urban areas had access to all three plan types or to the combination of PPO and PFFS, while this value was slightly more than 90 percent in rural areas. In contrast, in large urban areas, less than 70 percent of beneficiaries fell into either of these two categories. Nearly a third of beneficiaries in large urban areas had access to HMO and PPO plans, but not PFFS plans.

3.3.3 Choice Among Plan Types by Census Region

Table 3-8 examines how access to MA plan type varied by census region in 2006. In the Northeast and the West, the most prevalent plan combinations available were all three plan types (HMO, PPO, PFFS) or HMO and PPO only. In the Midwest and the South, on the other hand, the most prevalent plan combinations were all three plan types or PPO and PFFS only. This is consistent with the results in Table 3-5, which showed that PFFS plans are offered in substantially fewer counties in the Northeast and West. In all regions, the two most prevalent plan type combinations are available to 85 percent or more of beneficiaries.

3.3.4 Choice of Multiple Medicare Advantage Contracts

Tables 3-6 to 3-8 present findings on the combinations of different plan types available to a beneficiary, consistent with the idea that an important aspect of "choice" of MA plans is the availability of different plan types that offer different provider access structures. Another aspect of choice, however, may relate to the number of different contracts available in an area (each of which may offer more than one plan). Choice among different contracts in an area may reflect both the sheer number of offerings available as well as the presence of multiple competing organizations (e.g., insurance companies) offering these options. Tables 3-9 and 3-10 use the number of contracts in a county as an alternative way to evaluate "choice" to beneficiaries in that county in 2006.

Table 3-9 stratifies the number of contracts available in a county by urbanicity. Results are weighted by the number of MA-eligible Medicare beneficiaries residing in each county, and therefore show the percentage of beneficiaries with access to the number of contracts. Consistent with our earlier analyses, these results show that the number of contracts in 2006 was related to county urbanicity, with urban areas as a whole having more total contract options than rural areas. Within urban areas, the number of contracts available to beneficiaries was an increasing function of the size of the urban area, with large urban areas having more than twice the number of contracts on average than small urban areas. Similarly, rural urban-adjacent beneficiaries had roughly 20 percent more contracts available on average than non-urban adjacent beneficiaries. Most of the distinction between large and medium urban area counties and all other counties was a result of the discrepancy in counties offering seven or more contracts. More than half of beneficiaries living in large and medium urban areas had access to more than seven contracts, as opposed to roughly 17 percent in small urban counties, 12 percent in urban-adjacent rural counties, and 7 percent in non-urban adjacent rural counties. For counties with 10 or more contracts, the effect was even more pronounced—a quarter of beneficiaries in medium urban counties, and 40 percent in large urban counties, had access to 10 or more contracts, while only 4 percent of beneficiaries in small urban counties had access to this many contacts. The proportion of rural beneficiaries with access to 10 or more contracts was less than 2 percent.

Table 3-10 stratifies the number of contracts per county by census region. The Northeast and West regions had the most mean contracts per county in 2006 (weighted by MA-eligible Medicare beneficiaries residing in each county), and the Midwest and South had the fewest. This finding is an interesting contrast to our findings from Table 3-8, which found that the Midwest and Southern regions offered a larger proportion of Medicare beneficiaries the choice among three (HMO, PPO and PFFS) plan options compared to the Northeast and West. In the Northeast and West, more than 30 percent of beneficiaries had a choice among 10 or more MA contract options, compared to fewer than 20 percent of beneficiaries in the Midwest or South. The South had the highest percentage—22 percent—of beneficiaries with access to 3 or fewer contract options.

3.4 Special Need Plans in 2006

Section 3 presents most information on the availability of MA options to beneficiaries in terms of contracts. However, SNPs are defined by their targeted population and are not defined by a contract type. Table 3-11 disaggregates 2006 SNP options by plan type and target population. The analysis shows that in 2006, SNPs were offered through both HMO and PPO contracts, including three SNPs offered by regional PPOs. About 87 percent of all SNPs were offered under HMO contracts.

Table 3-11
Number of special needs contracts and plans by plan type and target beneficiaries, 2006

Plan Type	Contracts		Plans
	Total1	SNP Only2	Total	Institutional	Dual Eligible	Chronic Condition
1 Offering at least one SNP. 2 Offering only SNPs. 3 HMO includes HMO POS and PSO plans. 4 PFFS plans are not allowed to provide SNFs. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Total	139	48	240	31	199	10
HMO3	117	39	208	18	180	10
Local PPO	19	9	29	13	16	0
Regional PPO	3	0	3	0	3	0

About one-third of the contracts offering at least one SNP specialized in offering SNPs only. A significant number of these contractors were Medicaid only HMOs, that upon passage of the MMA, applied to be SNPs in order to keep their populations served intact. SNPs were heavily focused on dual Medicare/Medicaid eligible beneficiaries, with over 80 percent of total plans aimed at this population. Overall, 75 percent of SNPs were HMOs targeted at dual eligible beneficiaries. 13 SNPs were local PPOs targeted at institutionalized beneficiaries.

SECTION 4
PREMIUMS AND BENEFITS

As discussed in Section 1, several important changes occurred in the MA program in 2006 that affected the premiums and benefits of MA plans. In 2006, the MMA added a major new benefit to the basic Medicare benefit package, the Part D prescription drug benefit. Many MA plans had offered a drug benefit prior to 2006, but the benefit was usually limited, such as covering generic drugs only and/or having annual benefit caps. Beginning in 2006, most MA plans had to offer at least the standard Part D benefit (or an actuarially equivalent benefit). MA plans could also offer enhanced alternative drug coverage. In offering drug benefits, MA plans competed against original Medicare FFS combined with standalone Part D coverage and possibly supplemental Medicare insurance.

Also beginning in 2006, MA plan premiums were determined through a new bidding process. Bids below the benchmark (with 25 percent of any difference between bid and benchmark retained in the Medicare trust funds) created rebate funds that are used to enhance benefits, reduce cost sharing, or reduce Part D or Part B premiums; the portion of any bid amount in excess of the "benchmark" rate became the beneficiary premium. For the period March–December 2004, the MMA changed county capitation rates by establishing a FFS per capita cost minimum capitation rate, raising floor rates, and establishing a minimum update of the greater of the national Medicare expenditure growth percentage or 2 percent. Changes in county rates can have a substantial effect on premiums and benefits (Pope et al., 2006). The phase-in of risk adjustment payment continued, with 75 percent of plan payments risk adjusted in 2006 and 100 percent in 2007 and after. New types of plans were created by the MMA or earlier BBA, including local and regional PPOs, Special Needs Plans (SNPs), and Private FFS (PFFS) plans. With these new plan types becoming increasingly prevalent in MA, it is important to compare the premiums and benefits they offer to beneficiaries.

This section analyzes the premiums and benefits of Medicare Advantage plans in 2006. The introduction of Part D in 2006 fundamentally changed plans' prescription drug benefits; also, drug benefits are now priced separately from non-drug benefits. A second incomparability is that because only contract-level enrollment weighting data were available, our earlier work studied trends in "basic" (lowest premiums within contract) HMO plans. Plan-level enrollment data are now available, hence our 2006 analysis incorporates all MA plans.

Because of the incomparabilities between the 2000 to 2005 and 2006 data, we present only the 2006 data in our tables. We do make some tentative comparisons in the text of the 2000 to 2005 and 2006 results to try to infer major trends between 2005 and 2006. The lack of consistency between the 2006 and earlier data should be kept in mind when assessing these comparisons.

We begin this section with an analysis of MA plan premiums, in Section 4.1. Section 4.2 analyzes the structure of MA plans' Part D prescription drug benefits for 2006. Section 4.3 then considers other benefits and cost sharing of MA plans in 2006.

4.1 Premiums

With the introduction of Part D, MA plans offering prescription drug benefits now charge two premiums, for Part C benefits (corresponding to Medicare FFS Parts A and B benefits) and for Part D prescription drug coverage. Some MA plans offer only Part C benefits, and only have a Part C premium (which may be zero). A beneficiary enrolling in MA does not have to take Part D coverage, but if the person does enroll in Part D, it must be through their plan. The only exception is if the beneficiary is enrolled in a PFFS plan not offering Part D, in which case the beneficiary can obtain Part D through a standalone prescription drug plan (PDP). As described in more detail later, nearly 90 percent of MA enrollees take up Part D through their MA plans.

We discuss Part C, Part D, and Parts C+D (total) premiums. The latter two premiums are tabulated for the subset of plans that incorporate the Part D benefit. Because the sample of plans differs, the sum of the Part C and Part D premiums does not exactly equal the Parts C+D premium. Most premiums we present are weighted by plan enrollment and reflect average premiums charged to enrollees. We also discuss national unweighted average premiums by plan type, which reflect plan offers not taking account of plan enrollment. Some enrollees may receive assistance in paying Medicare Advantage premiums (e.g., the Part D low income subsidies), thus the premium amounts reflect plan charges, not necessarily enrollee out of pocket payments.

We first discuss premiums by plan type in Section 4.1.1, then the range of premiums paid by MA enrollees in Section 4.1.2. We examine geographic variation in premiums in Section 4.1.3. Section 4.1.4 considers plans that reduce the Medicare Part B premium.

4.1.1 By Plan Type

Table 4-1 presents national average enrollment-weighted and unweighted Part C, Part D, and combined Parts C+D 2006 premiums by MA plan type. The enrollment-weighted premiums were lower, indicating that beneficiaries disproportionately enrolled in lower-premium MA plans. The enrollment-weighted average monthly MA Part C premium in 2006 was nearly $20 ($19.71), the average Part D premium among plans offering Part D was more than $10 ($11.63), and the total (Parts C+D) premium was slightly more than $30 ($30.43), or $365 per year.

Table 4-1
Mean monthly premiums of Medicare Advantage plans by plan type, 2006

Plan type	Enrollment-weighted (average enrollee premium)			Unweighted (average plan offer)
	Part C	Part D1	Parts C+D1	Part C	Part D1	Parts C+D1
1 For plans offering Part D. 2 HMO includes HMO POS and PSO plans. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Total	$19.71	$11.63	$30.43	$26.25	$18.80	$44.62
Open Access Plan	21.00	10.67	30.86	29.37	18.41	48.38
HMO2	21.68	10.34	30.84	24.02	15.46	39.87
Local PPO	45.83	23.43	68.33	44.53	27.11	71.69
Regional PPO	13.48	12.81	26.85	40.66	20.87	66.88
PFFS	9.96	7.28	14.8	30.48	18.7	48.46
SNP	0.51	24.23	24.74	1.87	21.01	22.89
Institutional	0.00	27.17	27.17	0.00	23.13	23.13
Dual	0.51	24.00	24.51	0.91	19.66	20.56
Chronic	10.80	35.93	46.72	27.27	42.88	70.15

As shown by the unweighted premiums reflecting plan offers, HMOs were priced the lowest among open access plans. PPO premiums were nearly twice as high as HMO premiums. PPOs are typically a higher-cost product than HMOs because they offer greater freedom of provider choice, out of network coverage, and less utilization management. PFFS plan pricing was intermediate between HMOs and PPOs. PFFS plans do not have defined provider networks and Enrollees are covered for services from any provider willing to accept the payment terms of the PFFS plan. As such, PFFS offers even greater freedom of provider choice than PPOs, choice that should be nearly equivalent to that available in traditional Medicare FFS.

The average SNP Part C premium was nearly zero, reflecting limited ability to pay among the dual eligibles who comprise most SNP enrollees. The Part C premium for chronic disease SNPs, however, was higher and more similar to the premiums of other MA plans. The average SNP Part D premium was only slightly above the MA average on a plan offer (unweighted) basis, but more than double the average accounting for enrollment. The higher average SNP Part D premium did not impose a greater out of pocket cost burden on many SNP enrollees because Part D low income assistance defrayed all or part of the premium for most of them (including all dual eligibles, who comprised most SNP enrollees). The chronic disease SNPs charged a higher Part D premium than other SNPs and than other MA plans.

Table 4-2 shows the percent of enrollees in zero premium plans by MA plan type. Over half (52.8%) of MA enrollees received their Part C and Part D benefits at no extra charge beyond the Medicare Part B premium. The proportion of enrollees in open access plans paying neither a Part C nor a Part D premium varied from a high of 65 percent for PFFS plans to a low of 16 percent for local PPOs. Very few SNP enrollees paid a Part C premium, but almost all were charged a Part D premium. As noted, Part D low income assistance presumably defrayed some or all of the SNP enrollees' Part D premium.

Table 4-2
Percent of Medicare Advantage enrollees in zero premium plans by plan type, 2006

Plan type	Part C	Part D1	Parts C+D1
1 For plans offering Part D. 2 HMO includes HMO POS and PSO plans. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Total	61.9%	57.1%	52.8%
Open Access Plans	59.4	61.3	56.7
HMO2	59.6	63.3	58.3
Local PPO	18.2	23.1	15.7
Regional PPO	50.4	37.5	35.8
PFFS	72.4	66.4	65.3
SNP	98.8	1.9	1.9

4.1.2 Enrollment by Premium Range

Table 4-3 shows the distribution of MA enrollees by Part C, Part D, and Parts C+D premium range. Among open access plans, there was a large concentration at zero and then a relatively uniform distribution of enrollees among the Part C premium ranges below $100. Almost all the Part D enrollment was in plans with premiums below $50. A significant fraction of MA enrollees were paying a substantial total (Parts C+D) premium. Over one-fifth (21 percent) were paying a monthly total premium of $75 or greater, and 8 percent were paying $100 or more each month. The SNP Part C premium was almost always zero, and the Part D and total premiums were overwhelmingly greater than zero, but less than $50.

Table 4-3
Percent of enrollees in Medicare Advantage plans, by premium range, 2006

Monthly premium	Part C	Part D1	Parts C+D1
1 For plans offering Part D. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Open Access Plans
$0	59.4%	61.3%	56.7%
>0–24.99	9.1	19.7	4.6
25–49.99	10.6	17.2	10.3
50–74.99	11.1	1.4	7.4
75–99.99	8.0	0.2	13.0
100+	1.7	0.2	8.0
SNP
$0	98.78	1.94	1.94
>0–24.99	0.36	58.89	58.88
25–49.99	0.47	39.14	38.21
50–74.99	0.38	0.00	0.10
75–99.99	0.00	0.00	0.83
100+	0.02	0.03	0.05

Table 4-3 adds to the information from Tables 4-1 and 4-2 by showing that although the majority of MA enrollees were in zero premium plans, and the average premium was modest, some MA enrollees were paying substantial premiums. This is consistent with what we found for MA enrollees in 2005 (Pope et al., 2006).

4.1.3 By Urbanicity and Region

Table 4-4 shows enrollment-weighted 2006 average MA premiums by urbanicity. Premiums in different urban and rural categories may be affected by several factors, including MA benchmark amounts, differences in plan types or benefits offered and chosen, the payment discounts plans can obtain from providers, beneficiary income levels and demand for extra benefits, and degree of competition among plans. Urban premiums were lower than rural premiums, but not by a large amount. The average total (Parts C+D) urban premium was $30.16 compared to $35.80 in rural areas. Within urban areas, enrollees in medium and smaller urban areas paid higher premiums than enrollees in large urban areas. Total premiums in small urban areas ($46.73) were nearly twice as great as in large urban areas ($27.48), and were the highest of any urban or rural category. Within rural areas, enrollees in counties adjacent to urban counties paid moderately higher average premiums than enrollees in non-adjacent counties. Enrollees in non-adjacent rural counties paid the second-lowest average MA total premium ($29.14), almost as low as the average total premium enrollees in large urban areas paid ($27.48). Findings from Table 4-5, percent of MA enrollees in zero premium plans by urbanicity, largely mirror those of Table 4-4.

Table 4-4
Mean monthly premiums of Medicare Advantage plans by urbanicity, 2006

Urbanicity	Part C	Part D1	Parts C+D1
1For plans offering Part D. NOTES: Weighted by plan enrollment. Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System and Medicare Beneficiary Database July 2006 data.
Urban	$19.41	$11.44	$30.16
Large Urban	17.73	10.44	27.48
Medium Urban	22.00	13.56	35.47
Small Urban	29.88	17.08	46.73
Rural	22.07	14.95	35.80
Rural—Urban Adjacent	22.97	16.38	37.69
Rural—not Urban Adjacent	18.90	9.92	29.14

Table 4-5
Percent of Medicare Advantage enrollees in zero premium plans by urbanicity, 2006

Urbanicity	Part C	Part D1	Parts C+D1
1 For plans offering Part D. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System and Medicare Beneficiary Database July 2006 data.
Urban	62.6%	57.6%	53.2%
Large Urban	65.8	61.5	57.0
Medium Urban	57.6	49.1	44.7
Small Urban	43.8	37.9	33.5
Rural	53.7	45.9	44.5
Rural—Urban Adjacent	50.4	41.7	40.1
Rural—not Urban Adjacent	65.5	61.0	59.7

Table 4-6 shows enrollment-weighted average premiums by census region, and Table 4-7 presents percent of enrollees in zero premium plans by region. Regional premium differences were pronounced. Average premiums were highest in the Northeast and lowest in the South, with a range of over 4.5 to 1 ($56.05 versus $12.26). This compares to a range of less than 2 to 1 across urbanicity categories (Table 4-4). Over 7 of 10 Southern MA enrollees paid no total premium, while fewer than 1 in 4 of Northeast MA enrollees were in zero total premium plans. Midwest and West average premiums and percent in zero premium plans were between the Northeast and South and were similar to each other.

Table 4-6
Mean monthly premiums of Medicare Advantage plans by region, 2006

	Part C	Part D1	Parts C+D1
1 For plans offering Part D. Weighted by plan enrollment. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System and Medicare Beneficiary Database July 2006 data.
Northeast	$32.53	$23.57	$56.05
Midwest	19.12	11.43	29.89
South	5.64	7.16	12.26
West	24.28	8.23	31.78

Table 4-7
Percent of Medicare Advantage enrollees in zero premium plans by region, 2006

Census region	Part C	Part D1	Parts C+D1
1 For plans offering Part D. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System and Medicare Beneficiary Database July 2006 data.
Northeast	39.9%	24.4%	24.3%
Midwest	61.6	60.0	56.6
South	82.6	72.8	71.1
West	57.6	61.6	51.6

It is not immediately clear why there was such a large regional difference in average MA premiums in 2006. Southern premiums have historically been lower than those in other regions, and were substantially lower in basic HMO plans in 2005 (Pope et al., 2006). But Northeast premiums were slightly lower than Midwest or West premiums in 2005, compared to substantially higher in 2006. Average MA premiums appear to have grown significantly in the Northeast from 2005 to 2006, while other regions showed much less change.

The factors cited above to explain differences by urbanicity—MA benchmark amounts, differences in types of plans and benefits offered and chosen, beneficiary ability to pay for enhanced benefits, provider payment discounts, and competition among plans—surely played a role in explaining regional differences. Variations in the Part C premium, which showed a nearly 6 to 1 range across regions, contributed more to total premium differences than did the 3 to 1 range in the Part D premium (Table 4-6). The Northeast had an unusually low percentage of enrollees in zero premium MA-PDs, only 24 percent, compared to at least 60 percent in the other regions (Table 4-7).

4.1.4 Part B Premium Reductions

Since 2003, plans have been allowed to reduce the Medicare Part B premium as an added benefit to their enrollees. Enrollees in Part B premium reduction plans pay a lower Medicare Part B premium than they would pay if they stayed in the traditional Medicare FFS program. In 2006, the Medicare Part B premium was $88.50.

Table 4-8 shows the percentage of Medicare Advantage enrollees who had a Part B premium reduction in 2006. Overall, 2.7 percent of MA enrollees had their Part B premium reduced, 0.8 percent did not pay any Part B premium, and the average Part B premium reduction among enrollees with a reduction was $42.07. HMO enrollees, and enrollees in large and medium urban areas and in the South, were most likely to have their Part B premium reduced. Not only did more than 5 percent of Southern enrollees have their Part B premium reduced, nearly half of the Southern enrollees in a premium reduction plan did not pay any Part B premium.

Table 4-8
Part B premium reduction by Medicare Advantage plan type, urbanicity, and region, 2006
Percent of enrollees

	With any reduction	With full reduction	Mean reduction1
1 Among enrollees with a reduction. 2 HMO includes HMO POS and PSO plan types. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System and Medicare Beneficiary Database July 2006 data.
Total	2.7%	0.8%	$42.07
Plan type
HMO2	3.0	0.9	42.20
Local PPO	1.3	1.3	87.70
Regional PPO	0.0	0.0	N/A
PFFS	1.4	0.0	25.05
Urbanicity
Urban, total	2.8	0.9	42.71
Large urban	2.8	0.7	38.47
Medium urban	3.2	1.5	57.20
Small urban	1.8	0.0	24.48
Rural, total	1.0	0.1	22.78
Rural, adjacent	1.2	0.2	23.42
Rural, not adjacent	0.4	0.0	15.53
Census region
Northeast	1.3	0.0	5.16
Midwest	0.8	0.0	25.12
South	5.2	2.5	57.32
West	2.1	0.0	22.79

The percentage of enrollees in Part B premium reduction plans seems to have declined in 2006. In 2004, 7.3 percent of enrollees in basic HMO plans had their premiums reduced, falling to 5.6 percent in 2005 (Pope et al., 2006). Table 4-8 shows that in 2006, only 3.0 percent of HMO enrollees were in Part B premium reduction plans.

4.2 Prescription Drug Benefits

The implementation of the Medicare Part D drug benefit, and the establishment of Medicare Advantage prescription drug plans (MA-PDs) was the most significant change in MA in 2006. This section characterizes the prescription drug benefits that MA-PDs provided in 2006.

MA-PDs had the flexibility to offer four types of Part D benefits:

• Defined standard
  
  
• Actuarially equivalent
  
  
• Basic alternative
  
  
• Enhanced alternative

We use these categories (merging "basic alternative" into "actuarially equivalent") as one important descriptor of drug benefits offered. We also use the category of "basic" coverage, which includes defined standard, actuarially equivalent, and basic alternative plans, as a descriptor.

The defined standard Part D benefit in 2006 had a $250 deductible, 25 percent enrollee cost sharing until the enrollee reached an "initial coverage limit" of $2,250 in total covered drug expenses. There was no coverage (other than discounted prices) in the "coverage" gap from the initial coverage limit to $3,600 in TrOOP costs, and catastrophic coverage reimbursing most expenditures above $3,600 in TrOOP costs.

The two types of basic coverage that are actuarially equivalent to defined standard plans are 1) standard coverage with actuarially equivalent cost sharing and 2) basic alternative coverage. In the first variant, plans have a similar overall structure to the defined standard benefit, but the cost sharing differs from the 25 percent coinsurance under the standard defined benefit. These "actuarially equivalent" plans tend to have tiered co-payments of a low dollar amount for a generic drug and higher amounts for preferred brand-name drugs and for non-preferred brand-name drugs. Under the second variant, termed "basic alternative coverage," plans have a different overall structure of the benefit, though they must be actuarially equivalent to the standard benefit. In a basic alternative coverage design, features such as a reduction in the deductible, changes in cost-sharing, and a modification of the initial coverage limit can be combined and still provide coverage with an actuarial value equal to standard coverage.

In addition to the defined standard plans and its two actuarially equivalent variants, plans were able to offer enhanced alternative prescription coverage that exceeds standard coverage by offering supplemental benefits such as an increase in the initial coverage limit, coverage in the gap, or reduced cost sharing.

This section is organized as follows. We begin in Section 4.2.1 by analyzing MA-PDs by plan type. Section 4.2.2 discusses drug benefits by urbanicity and region. Section 4.2.3 presents data on MA-PDs' cost sharing before the initial coverage limit, Section 4.2.4 on their initial coverage limits, and Section 4.2.5 on their coverage if any in the coverage gap. Section 4.2.6 discusses characteristics of MA-PDs' formularies.

4.2.1 By Plan Type

Table 4-9 shows type of prescription drug benefit by MA plan type. About 11 percent of MA enrollees were in plans without a drug benefit. These beneficiaries may have prescription drug coverage from another source, such as a former employer, or may have declined Part D coverage. The proportion of enrollees in plans without Part D coverage is small for all plan types except PFFS plans. PFFS enrollees in plans not offering drug coverage are allowed to enroll in standalone Part D plans (PDPs). SNPs are required to provide Part D, and so have no enrollees without it.

Table 4-9
Prescription drug benefits by Medicare Advantage plan type, 2006
Percent of enrollees with Part D benefit type for each MA plan type

MA Plan type	Total	None	Basic			Enhanced Alternative
			Total	Defined standard	Actuarially equivalent1
1 Includes actuarially equivalent and basic alternative plan types. 2 HMO includes HMO POS and PSO plans. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
All	100.0%	10.6%	27.0%	5.9%	21.1%	62.4%
Open access plans	100.0	11.3	23.0	1.9	21.1	65.7
HMO2	100.0	7.2	26.2	2.1	24.1	66.6
Local PPO	100.0	10.0	36.9	0.6	36.3	53.1
Regional PPO	100.0	7.7	12.4	8.4	4.0	79.9
PFFS	100.0	35.0	1.9	0.2	1.7	63.1
SNP	100.0	0.0	85.8	65.0	20.8	14.2

Twenty-seven percent of MA enrollees were in MA-PDs offering basic coverage, most of which was an actuarially equivalent variant rather than defined standard. Basic coverage—particularly defined standard—was especially prevalent among SNP enrollees, but the Part D low income subsidy generally exempted most SNP enrollees from the cost sharing and coverage gap in these plans except for the statutorily mandated copayment amounts. Basic coverage—nearly all actuarially equivalent—was more prevalent than average among local PPO enrollees.

Enhanced coverage was the most common Part D benefit in all plan types except SNPs. A majority of enrollees in each non-SNP plan type had enhanced coverage. Overall, 62 percent of MA enrollees enjoyed enhanced coverage in 2006.

There is no doubt that the implementation of Part D substantially improved MA enrollees' prescription drug coverage. MA plans' drug benefits had improved subsequent to the higher MA capitation rates mandated by the MMA beginning in early 2004. But in 2005, 25 percent of enrollees in basic HMO plans had no prescription drug coverage, and 36 percent had coverage only for generic drugs (Pope et al., 2006). Moreover, the drug benefit was typically capped at relatively low dollar amounts (e.g., $600).

4.2.2 By Urbanicity and Region

Table 4-10 shows type of prescription drug benefit by urbanicity. A much higher percentage of rural than urban MA enrollees were in plans without a Part D benefit. This reflects the prevalence of PFFS plans in rural areas, which were not required to offer a prescription drug benefit. Among enrollees in MA-PDs, the distribution of benefit type did not vary markedly by urbanicity.

Table 4-10
Prescription drug benefits of Medicare Advantage enrollees by urbanicity, 2006
Percent of enrollees with Part D benefit type for each urbanicity category

Urbanicity	Total	None	Basic			Enhanced Alternative
			Total	Defined standard	Actuarially equivalent1
1 Includes actuarially equivalent and basic alternative plan types. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System and Medicare Beneficiary Database July 2006 data.
Urban	100.0%	8.8%	27.9%	5.9%	22.0%	63.3%
Large Urban	100.0	6.4	27.4	6.8	20.7	66.2
Medium Urban	100.0	12.9	28.5	3.8	24.7	58.6
Small Urban	100.0	23.0	31.4	4.1	27.3	45.6
Rural	100.0	29.9	21.4	4.6	16.8	48.7
Rural—Urban Adjacent	100.0	29.8	24.0	5.4	18.6	46.2
Rural—Not Urban Adjacent	100.0	30.3	12.4	1.8	10.7	57.3

Table 4-11 shows Part D benefit type by census region. MA enrollees in the Northeast and the Midwest were more likely to be in a plan without a drug benefit than Southern or Western MA enrollees. Northeastern enrollees were more likely to have only basic drug coverage, while Southern enrollees were especially likely to have enhanced coverage.

Table 4-11
Prescription drug benefits of Medicare Advantage enrollees by region, 2006
Percent of enrollees with Part D benefit type for each region

Census region	Total	None	Basic			Enhanced Alternative
			Total	Defined standard	Actuarially equivalent1
1 Includes actuarially equivalent and basic alternative plan types. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System and Medicare Beneficiary Database July 2006 data.
Northeast	100.0%	17.1%	41.0%	8.7%	32.2%	41.9%
Midwest	100.0	15.0	19.2	2.6	16.6	65.8
South	100.0	8.3	15.0	4.1	10.9	76.7
West	100.0	6.0	33.5	6.9	26.6	60.6

4.2.3 Cost Sharing Before the Initial Coverage Limit

Table 4-12 shows the cost sharing structure of MA-PDs before the initial coverage limit, by type of drug benefit. The vast majority (86%) of MA-PD enrollees paid no deductible. Virtually no enrollees in enhanced alternative plans paid a deductible, and most in actuarially equivalent plans did not. All enrollees in defined standard coverage paid the $250 deductible, but they were a small minority of MA-PD enrollees. With the exception of defined standard plans (which used only a 25 percent coinsurance tier), most plans used both copayment and coinsurance tiers.

Table 4-12
Cost sharing before the initial coverage limit, by type of Medicare Advantage prescription drug plan, 2006
Percent of enrollees in each Part D benefit type with specified cost sharing

	Total	Type of prescription drug plan
		Defined standard	Actuarially equivalent1	Enhanced alternative
1 Includes actuarially equivalent and basic alternative plan types. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Deductible
Total	100.0%	100.0%	100.0%	100.0%
Zero	86.0	0.0	71.1	99.1
Reduced	2.4	0.0	9.5	0.2
$250	11.6	100.0	19.4	0.7
Cost sharing structure before the initial coverage limit
Total	100.0%	100.0%	100.0%	100.0%
No cost sharing	0.8	0.0	0.0	1.1
25% coinsurance amount	6.6	100.0	0.0	0.0
One or more groups of cost sharing	92.6	0.0	100.0	98.9
# of copayment tiers
Total	100.0%	NA	100.0%	100.0%
None	8.5	NA	1.2	2.3
1	3.7	NA	2.3	4.5
2	41.6	NA	40.0	46.0
3	43.5	NA	49.1	45.7
4	2.0	NA	6.4	0.8
5+	0.7	NA	0.9	0.7
# of coinsurance tiers
Total	100.0%	100.0%	100.0%	100.0%
None	26.3	0.0	50.4	20.7
1	46.8	100.0	25.1	49.2
2	26.3	0.0	24.1	29.5
3	0.5	0.0	0.4	0.6
4	0.0	0.0	0.0	0.1

Tables 4-13 and 4-14 present more detail on the drug tiering design and cost sharing amounts. Table 4-13 cross-tabulates the number of copayment tiers by the number of coinsurance tiers. It shows that the most common tiering structure (28.9% of MA-PD enrollees) was three copayment tiers and one coinsurance tier. In these plans, enrollees paid a copayment for drugs in the first three tiers and were assessed coinsurance for drugs in the fourth tier. Typically, tier 1 was generic drugs, tier 2 was preferred brand drugs, tier 3 was nonpreferred drugs, and tier 4 included specialty drugs (high-priced and unique drugs and biologicals).

Table 4-13
Drug tiers of Medicare Advantage prescription drug plans, 2006
Percent of enrollees by number of copayment and coinsurance tiers

# copayment tiers	# coinsurance tiers
	Total	None	1	2	3
NOTES: Cost sharing is before the initial coverage limit. Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Total	100.0%	26.3%	46.8%	26.3%	0.5%
none	8.5	1.0	7.3	0.0	0.1
1	3.7	0.6	2.4	0.7	0.0
2	41.6	15.3	8.2	17.8	0.3
3	43.5	7.5	28.9	7.2	0.0
4	2.0	1.6	0.0	0.4	0.0
5+	0.7	0.2	0.0	0.4	0.1

Table 4-14 tabulates median (weighted by plan enrollment) copayments and coinsurance percentages for the most common MA-PD drug tiering designs. Copayments are for a 30-day drug supply at in-network retail pharmacies. Over 90 percent of MA-PD enrollees faced one of the cost sharing structures reported in Table 4-14. The most common copayment for tier 1 drugs (typically generic drugs) was $5, but it ranged as high as $20 in some plan designs. The copayment for tier 2 drugs (usually preferred brand) was typically about $30, but ranges from $20 to $40. Plans with a third tier (usually nonpreferred drugs) typically charged a $50 to $60 copayment for drugs in this tier. The cost sharing percentage for drugs in a coinsurance tier (usually specialty, injectable, or expensive drugs) was typically 25 or 33 percent, but ranged up to 50 percent in a second coinsurance tier.

Table 4-14
Common cost sharing structures in Medicare Advantage prescription drug plans, 2006
Median copayments¹ or coinsurance by drug tier

	Cost sharing structure
	3 copay/ 1 coinsur.	2 copay/ 2 coinsur.	2 copay/ 0 coinsur.	2 copay/ 1 coinsur.	3 copay/ 0 coinsur.	0 copay/ 1 coinsur.	3 copay/ 2 coinsur.
1 For a 30-day supply from a retail pharmacy. NOTES: Medians are weighted by plan enrollment. This cost sharing is before the initial coverage limit. Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories.
% enrollment	28.9	17.8	15.3	8.2	7.5	7.3	7.2
Copayment tiers (typical drugs)
1 (generics)	$  5.00	$  8.50	$10.00	$20.00	$  5.00	--	$  5.00
2 (preferred brand)	28.00	26.70	30.00	40.00	20.00	--	25.00
3 (non-preferred)	58.00	--	--	--	50.00	--	50.00
Coinsurance tiers (typical drugs)
1 (specialty)	25%	33%	--	25%	--	25%	25%
2 (injectables)	--	50	--	--	--	--	25

4.2.4 Initial Coverage Limit

Table 4-15 characterizes the initial coverage limit in MA-PDs. SNPs are excluded. Over three-quarters of MA-PD enrollees across all Part D benefit types were in plans with the standard $2,250 initial coverage limit. About 13 percent had a lower, and about 10 percent a higher, initial coverage limit.

Table 4-15
Initial coverage limit in Medicare Advantage prescription drug plans, by plan and geographic characteristics, 2006
Percent of enrollees with specified initial coverage limits

	Total	Initial coverage limit
		< $2,250	$2,250	> $2,250
1 HMO includes HMO POS and PSO plans. NOTES: Excludes SNPs (they are not included in the CMS landscape file). Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 and CMS Landscape 2006 data.
Total	100.0%	13.4%	77.0%	9.5%
Benefit type
Defined standard	100.0	0.0	100.0	0.0
Actuarially equivalent	100.0	16.8	82.9	0.3
Enhanced	100.0	12.7	74.4	12.9
Plan type
HMO1	100.0	14.7	74.0	11.3
Local PPO	100.0	13.6	82.0	4.4
Regional PPO	100.0	1.9	98.1	0.0
PFFS	100.0	5.4	94.6	0.0
Part D premium range
$0	100.0	11.5	74.1	14.4
>0-24.99	100.0	11.2	87.0	1.8
25-49.99	100.0	23.0	76.7	0.3
50-74.99	100.0	3.1	76.9	20.1
75-99.99	100.0	25.1	63.2	11.7
100+	100.0	89.2	0.0	10.8
Urbanicity
Urban	100.0	13.9	75.9	10.2
Large Urban	100.0	13.4	73.4	13.1
Medium Urban	100.0	15.3	81.9	2.8
Small Urban	100.0	13.9	85.9	0.3
Rural	100.0	7.3	92.3	0.4
Rural, Adjacent	100.0	7.4	92.2	0.4
Rural, Nonadjacent	100.0	7.0	92.7	0.3
Region
Northeast	100.0	15.8	82.6	1.6
Midwest	100.0	11.8	87.2	1.1
South	100.0	14.1	73.7	12.2
West	100.0	12.0	72.3	15.7

Among drug benefit types, all enrollees in defined standard plans and most enrollees in actuarially equivalent and enhanced plans had the standard $2,250 initial coverage limit. About 17 percent of MA-PD enrollees in actuarially equivalent plans had a lower initial coverage limit. These enrollees' plans lowered the initial coverage limit to keep the actuarial value of the plan equal to standard coverage while reducing other cost sharing, such as eliminating the deductible. About 13 percent of enhanced plan enrollees had a lower than $2,250 initial coverage limit and 13 percent had a higher limit. A higher initial coverage limit is one way to enhance the standard Part D benefit, because it delays the drug spending level at which an enrollee enters the coverage gap. By plan type, almost all PFFS and regional PPO enrollees had the standard initial coverage limit. About 14 percent of local PPO enrollees had a reduced initial coverage limit, reflecting the prevalence of the actuarially equivalent MA-PD benefit in this plan type. The smallest proportion among plan types, but still a majority, of HMO MA-PD enrollees had the standard initial coverage limit. About 15 percent had a lower limit and 11 percent a higher limit, reflecting both actuarially equivalent and enhanced coverage.

Considering initial coverage limit by Part D premium range, almost all MA-PD enrollees are in the three lowest premium categories (Table 4-3). In the three lowest premium categories in Table 4-15, only zero premium plans had a non-negligible percentage of enrollees (14%) with a higher-than-standard initial coverage limit. All three categories, especially $25 to $49.99, had some enrollees in plans with a lower-than-standard initial coverage limit. MA-PDs priced at $75 or higher were more likely to have non-standard initial coverage limits, but very few beneficiaries enrolled in those plans.

About three-quarters of MA-PD enrollees in urban areas were in plans with the standard $2,250 initial coverage limit, but over 90 percent of rural enrollees were in such plans. Only in large urban areas were a non-negligible proportion (13 percent) of MA-PD enrollees in plans with an higher-than-standard initial coverage limit. A small, but non-negligible, proportion of MA-PD enrollees were in plans with an initial coverage limit less than $2,250 across all urbanicity categories, especially in urban areas.

Virtually no Northeastern or Midwestern MA-PD enrollees were in plans that raised their initial coverage limit above the standard amount. A small, but non-negligible proportion of Southern and Western MA-PD enrollees were in plans that enhanced their initial coverage limits. The proportion of enrollees in plans with reduced initial coverage limits did not vary markedly across areas. Overall, Southern and Western MA-PD enrollees were less likely to be in plans with the standard initial coverage limit than Northeastern or Midwestern enrollees.

4.2.5 Gap Coverage

Medicare Part D plans, as one form of enhancement to the standard Part D benefit, may offer coverage in the coverage gap. Table 4-16 shows that about 27 percent of MA-PD enrollees were in plans with some form of gap coverage in 2006.¹¹ Overwhelmingly, gap coverage was for generic drugs only (84% of all enrollees with gap coverage had it for generics only).

Table 4-16
Gap coverage in Medicare Advantage prescription drug plans, by plan and geographic characteristics, 2006
Percent of enrollees with specified gap coverage

	Total	Gap coverage
		None	Generics only	Generics and brand
1 HMO includes HMO POS and PSO plans. NOTES: Excludes SNPs (they are not included in the CMS landscape file). Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 and CMS Landscape 2006 data.
Total, open enrollment plans	100.0%	73.1%	22.7%	4.2%
Benefit type
Defined standard	100.0	100.0	0.0	0.0
Actuarially equivalent	100.0	100.0	0.0	0.0
Enhanced alternative	100.0	63.5	30.9	5.7
Plan type
HMO1	100.0	69.2	25.9	4.9
Local PPO	100.0	72.6	25.0	2.4
Regional PPO	100.0	96.2	3.8	0.0
PFFS	100.0	100.0	0.0	0.0
Part D premium range
$0	100.0	70.0	25.1	4.9
>0-24.99	100.0	87.0	11.3	1.7
25-49.99	100.0	72.3	25.0	2.8
50-74.99	100.0	32.5	56.9	10.6
75-99.99	100.0	74.9	0.0	25.1
100+	100.0	10.8	0.0	89.2
Urbanicity
Urban	100.0	71.8	23.9	4.3
Large Urban	100.0	67.3	28.9	3.8
Medium Urban	100.0	82.5	10.9	6.6
Small Urban	100.0	90.1	8.1	1.8
Rural	100.0	91.7	6.6	1.7
Rural, Adjacent	100.0	90.0	7.8	2.1
Rural, Nonadjacent	100.0	97.3	2.5	0.1
Region
Northeast	100.0	79.7	20.3	0.0
Midwest	100.0	91.0	8.1	0.9
South	100.0	64.9	27.8	7.3
West	100.0	69.1	25.8	5.1

As required by law, gap coverage was offered only in enhanced alternative benefit plans. No PFFS and very few regional PPO MA-PD enrollees had gap coverage. About one-quarter of HMO and local PPO MA-PD enrollees had gap coverage for generic drugs, but less than 5 percent had brand drug gap coverage. Among the three lower Part D premium range categories where virtually all MA-PD enrollees were concentrated, zero premium plans offered the most gap coverage. MA-PD plans with monthly premiums of $50 or more offered more extensive gap coverage than lower-priced plans, but only a very small proportion of MA-PD enrollees chose these plans.

Urban MA-PD enrollees were much more likely to have gap coverage than rural enrollees (28 percent versus 8 percent). MA-PD enrollees in large urban areas were most likely to have gap coverage for generics only, but enrollees in medium urban areas were most likely to have gap coverage for generics and brand drugs.

Gap overage varied considerably by region. Only 9 percent of Midwestern MA-PD enrollees had any gap coverage, compared to 35 percent of Southern enrollees. The South also had the highest percentage of brand and generics gap coverage, over 7 percent, compared to none in the Northeast and less than 1 percent in the Midwest. Overall, Southern and Western MA-PD enrollees had more generous gap coverage than Northeastern or especially Midwestern enrollees.

4.2.6 Formulary Characteristics

MA-PD drug formularies, or lists of covered drugs, are an important aspect of the Part D benefit. The formulary can affect access to drugs, along with other aspects of a plan's drug utilization management strategies such as non-formulary exceptions, step therapy, and medication management programs. Table 4-17 presents several statistics from CMS' Landscape file relating to MA-PDs' formularies, and other drug management strategies. These statistics are

• Mean number of top 100 drugs on MA-PDs' formularies
  
  
• Mean number of top 100 drugs requiring prior authorization
  
  
• Mean number of top 100 drugs with cost sharing below $20
  
  
• Mean number of drugs on a plan's formulary

Table 4-17
Selected characteristics of Medicare Advantage prescription drug plan formularies, by plan and geographic characteristics, 2006
Enrollment-weighted plan mean of number of drugs

	# top 100 drugs			# of drugs on formulary
	On formulary	w/prior authorization	w/cost sharing <$20
1 HMO includes HMO POS and PSO plans. NOTES: Excludes SNPs (they are not included in the CMS landscape file). Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 and CMS Landscape 2006 data.
Total	91.3	8.2	64.5	2,137
Benefit type
Defined standard	90.9	10.2	N/A	1,725
Actuarially equivalent	89.4	10.2	63.0	2,025
Enhanced alternative	91.9	7.4	65.0	2,185
Plan type
HMO1	89.9	8.2	65.8	2,187
Local PPO	96.2	9.1	65.0	2,089
Regional PPO	98.9	8.0	56.2	1,712
PFFS	97.2	6.2	57.3	2,568
Part D premium range
$0	91.1	7.4	63.8	2,173
>0-24.99	89.4	11.1	60.9	2,056
25-49.99	93.9	6.1	70.1	2,133
50-74.99	96.7	21.8	75.6	1,824
75-99.99	93.8	4.3	82.2	2,161
100+	80.4	13.9	76.6	1,642
Urbanicity
Urban	91.0	8.2	64.9	2,143
Large Urban	90.4	8.2	65.5	2,166
Medium Urban	92.4	8.2	63.2	2,095
Small Urban	93.8	8.4	63.1	2,016
Rural	95.7	8.0	59.8	2,049
Rural, Adjacent	95.2	7.9	60.4	2,138
Rural, Nonadjacent	97.3	8.4	57.6	1,756
Region
Northeast	93.9	8.1	69.5	2,469
Midwest	97.8	6.6	58.9	2,261
South	95.0	7.7	68.7	2,262
West	83.1	9.4	59.7	1,748

All statistics are plan averages weighted by plan enrollment.

The average number of top 100 drugs on the formulary did not vary significantly by drug benefit type, but enhanced MA-PDs required prior authorization for fewer of the top 100 drugs on average than non-enhanced plans. Enhanced plans also listed a larger number of drugs on their formularies, especially as compared to defined standard plans.

HMOs covered fewer of the top 100 drugs on their formularies than other plan types, but had less than $20 cost sharing for a larger number of these drugs. PFFS plans required prior authorization for the fewest of the top 100 drugs. MA-PDs with premiums of $25 or more had lower cost sharing for the top 100 drugs than lower-priced plans. On average, regional PPOs listed the fewest number of drugs on their formularies while PFFS plans listed the most drugs on their formularies.

Urban MA-PD enrollees' plans covered on average fewer of the top 100 drugs than rural enrollees' plans, but the urban plans charged less than $20 cost sharing for a larger number of the top 100 drugs and listed slightly more total drugs on their formularies. Western MA-PD plans appear to have had more restrictive drug access policies than other regions. Western plans on average listed considerably fewer of the top 100 drugs and fewer total drugs on their formularies, and required prior authorization for a larger number of drugs. They also had cost sharing below $20 for the second lowest number of top 100 drugs among regions.

4.3 Other Benefits and Cost Sharing

This section turns from MA plans' Part D drug benefits to consideration of other benefit and cost sharing policies of MA plans in 2006. Section 4.3.1 discusses supplemental benefits offered by MA plans, Section 4.3.2 considers cost sharing policies, and Section 4.3.3 analyzes out of pocket cost maximums.

4.3.1 Supplemental Benefits

MA plans can supplement the standard Medicare FFS Parts A and B benefit package by including additional benefits in their plans. Table 4-18 shows the percentage of MA enrollees who enjoyed selected mandatory supplemental benefits by plan type. "Supplemental" means that the benefits supplement the standard Medicare FFS Part A/B benefits. "Mandatory" means that the benefits were included as part of a plan's basic benefit package.¹²

Table 4-18
Selected mandatory supplemental benefits in Medicare Advantage plans, 2006
Percent of enrollees with benefit

	Total	HMO	Local PPO	Regional PPO	PFFS
1 Includes eye exams and glasses/contact lenses. 2 Includes routine hearing tests. 3 Includes prophylaxis (cleaning). 4 Includes routine foot care. 5 Includes routine care. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Vision1	83.3%	94.6%	77.1%	59.6%	20.3%
Hearing exam2	64.6	63.1	51.4	59.6	78.4
Dental3	32.7	36.0	39.2	66.5	6.3
Podiatrist4	27.1	30.1	43.6	36.9	1.9
Chiropractic5	6.3	7.0	13.8	0.0	0.0

The most common of the supplemental benefits considered is vision coverage (eye exams and glasses), which over 80 percent of MA enrollees had in 2006. About two-thirds of MA enrollees had coverage for hearing exams, one-third dental coverage, about one-quarter coverage for podiatry, and 6 percent for chiropractic treatment. The percentages of MA enrollees with these benefits in 2006 are roughly comparable to the percentages of basic HMO enrollees with the benefits in 2005 (Pope et al., 2006). Among plan types, HMO enrollees were most likely to have vision coverage, local PPO enrollees were mostly likely to have podiatry and chiropractic coverage, regional PPO enrollees were most likely to have dental coverage, and PFFS enrollees were most likely to have hearing exam coverage.

4.3.2 Cost Sharing

Table 4-19 shows the percentage of MA enrollees who faced cost sharing of the indicated amounts for selected services in 2006, by plan type. Most MA enrollees faced copayments of $5 to $15 for primary care physician visits. Eleven percent had no primary care copayment while 8 percent paid up to $25. Copayments for specialist physician visits were higher. The most common amounts were in the $15.01 to $25 range, and one-quarter of MA enrollees paid more than $25 per specialist visit. Emergency department copayments were almost always about $50. More than 85 percent of MA enrollees faced copayments or coinsurance for hospital services, either acute inpatient admissions, or outpatient care. About three-quarters were charged copayments or coinsurance for X-ray and clinical laboratory services.

Table 4-19
Cost sharing for selected services in Medicare Advantage plans, 2006
Percent of enrollees with cost sharing category

	Total	HMO	Local PPO	PFFS	Regional PPO
1 Sums to 100.0% across copayment categories. Some plans also had coinsurance for certain services. 2 Does not include any applicable deductibles. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and US territories. SOURCE: RTI analysis of CMS Health Plan Management System July 2006 data.
Primary care physician visit copayment
Total1	100.0%	100.0%	100.0%	100.0%	100.0%
$0	10.6	12.5	4.1	3.6	0.0
$0.01-$5	18.5	20.4	20.7	9.6	1.8
$5.01-$10	37.0	37.4	41.1	27.3	96.2
$10.01-$15	25.8	20.4	21.9	58.0	1.6
$15.01-$25	8.0	9.2	12.2	1.6	0.4
more than $25	0.2	0.2	0.0	0.0	0.0
Coinsurance	3.0	3.7	0.2	0.2	0.0
Specialist physician visit copayment
Total1	100.0%	100.0%	100.0%	100.0%	100.0%
$0	3.5	3.7	1.5	3.6	0.0
$0.01-$5	3.0	3.0	12.5	0.0	0.9
$5.01-$10	16.2	17.2	12.3	11.4	21.7
$10.01-$15	12.5	12.8	19.4	9.6	0.8
$15.01-$25	39.4	45.4	27.5	11.0	36.1
$25.01-$35	24.6	16.8	26.9	64.5	40.4
$35.01-$50	0.8	1.1	0.1	0.0	0.0
Coinsurance	3.8	4.7	0.2	0.2	0.0
Emergency room visit copayment
Total	100.0%	100.0%	100.0%	100.0%	100.0%
$0	1.2	1.4	0.1	0.0	0.0
$0.01-$20	0.0	0.0	0.0	0.0	0.0
$20.01-$40	5.2	4.9	11.8	3.7	3.9
$40.01-$50	93.6	93.7	88.2	96.3	96.1
Any cost sharing (either copayment or coinsurance)2
Acute hospital admission	87.1	85.2	85.9	97.5	99.0
Hospital outpatient	86.3	84.8	70.4	100.0	97.0
X-ray services	75.8	72.0	66.2	100.0	97.0
Laboratory services	76.3	73.3	62.3	96.8	97.0

The Table 4-19 copayments and percentages with cost sharing are roughly comparable for most services to the cost sharing faced by enrollees in basic HMO plans in 2005 (Pope et al., 2005). However, the percentage of MA enrollees with any cost sharing for hospital outpatient and laboratory services in 2006 was substantially higher than the percentage of basic HMO enrollees with cost sharing for these services in 2005. This difference could be the result of rising cost sharing for these services in 2006, and the inclusion of all types of MA plans in the 2006 analysis.

Cost sharing tended to be higher in PFFS plans than in other MA plan types. For example, the largest percentage of PFFS enrollees paid primary care visit copayments of $10.01 to $15 and specialist copayments of $25.01 to $35, rather than $5.01 to $10 and $15.01 to $25, respectively. All PFFS enrollees paid cost sharing for hospital outpatient and X-ray services, and almost all for acute hospital admissions and laboratory services.

4.3.3 Out of Pocket Cost Maximums

Out of pocket (OOP) cost sharing maximums offer MA enrollees protection against high medical expenses, especially beneficiaries who are in poorer health status and use more health services. This "stop loss" coverage, which is not available in the traditional FFS Medicare program, sets an upper limit on the amount an enrollee will have to pay for covered Part C benefits in a year.¹³ Tables 4-20 and 4-21 provide analysis of MA plans' and enrollees' OOP cost maximums in 2006. About 42 percent of MA enrollees had an OOP maximum (Table 4-20). About 15 percent had a maximum that applied to all covered services. One-quarter had a maximum that did not apply to all covered services, but that included hospital inpatient acute care, the largest medical expense category. Less than 2 percent of enrollees had a maximum that did not include hospital inpatient acute services. The most common OOP maximum was in the $2,001 to $3,000 range (typically $3,000), and most maximums ranged from $1,000 to $5,000.

Table 4-20
Out of pocket (OOP) maximums in Medicare Advantage plans, 2006
Percent of enrollees

NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS July 2006 Health Plan Management System data.
Has OOP maximum	41.9%
OOP maximum applies to all covered services	15.3
some covered services excluded	26.5
inpatient hospital acute included	25.0
inpatient hospital acute excluded	1.6
OOP maximum amount
$1-1,000	2.7
$1,001-2,000	7.1
$2,001-3,000	16.1
$3,001-4,000	5.1
$4,001-5,000	10.7
$5,001+	0.1

Table 4-21
Out of pocket (OOP) maximums in Medicare Advantage plans, by plan type, urbanicity, and region, 2006

	Enrollees w/OOP maximum
	%	Median1
1 Enrollment-weighted plan median. 2 HMO includes HMO POS and PSO plans. NOTE: Excludes employer-only and Part B-only plans. Excludes Puerto Rico and the US territories. SOURCE: RTI analysis of CMS July 2006 Health Plan Management System and Medicare Beneficiary Database data.
Total	41.9%	$3,000
Plan type
HMO2	33.3	3,000
Local PPO	54.1	1,500
Regional PPO	100.0	3,000
PFFS	80.0	5,000
Urbanicity
Urban	39.9	3,000
Large urban	34.3	3,000
Medium urban	53.1	3,000
Small urban	57.8	3,250
Rural	62.5	5,000
Rural, urban adjacent	58.7	4,000
Rural, not urban adjacent	75.8	5,000
Region
Northeast	13.2	2,960
Midwest	71.6	3,500
South	46.2	3,000
West	45.2	3,000

OOP maximums were least common in HMOs—only one-third of HMO enrollees had one in 2006 (Table 4-21). The MMA required regional PPOs to have OOP cost maximums; all enrollees in these plans had one. Most PFFS enrollees (80 percent) had an OOP maximum, as did about half of local PPO enrollees. Although less common in HMOs, HMO OOP maximums were typically lower ($3,000 median) than PFFS plans' maximums ($5,000 median). Local PPOs had a particularly low median OOP maximum, at $1,500.

Urban enrollees were less likely to have OOP cost maximums than rural enrollees, but when they existed urban maximums were typically lower. OOP cost maximums were most common in the least urbanized areas, and least common in the most urbanized areas. This is due to the dominance of HMOs, which were least likely to have maximums, in large urban areas, and of PFFS plans, which were most likely to have them, in rural areas. Regionally, most Midwestern enrollees had OOP cost maximums, but few Northeastern enrollees did. Nearly half of Southern and Western enrollees were in plans with a maximum.

SECTION 5
ENROLLMENT

In this section, we present results from our descriptive analysis of Medicare Advantage (MA) enrollment. We analyze trends for the time period 2000–2006, and present a detailed cross-sectional analysis of MA enrollment in 2006.¹⁴

5.1 Medicare Advantage Enrollment Trends 2000–2006

Our analysis sample for examining enrollment trends is all MA enrollment during 2000-2006. As shown in the Table 5-1, total MA enrollment has varied over the last decade. In 2000 enrollment was 6.3 million with a penetration rate of 15.8 percent, but then enrollment dropped to a low of 4.7 million enrollees in 2003, with a penetration rate of only 11.5 percent.¹⁵ However, recently Medicare health plan enrollment has rebounded, as is evident in 2006, which has an enrollment of 6.8 million and a penetration rate of 15.7 percent, similar to 2000 levels. As a percentage of total MA enrollment, local coordinated care plans (HMOs and local PPOs)¹⁶ have decreased from virtually 100 percent in 2000 to 86.9 percent in 2006. This market share was picked up by PFFS plans, which in 2000 had close to zero enrollment, but by 2006 had an enrollment of 802,068 and an MA market share of 11.8 percent.

Table 5-1
Medicare Advantage¹ enrollment, 2000–2006

Year2	Total		HMO/Local PPO3		PFFS		Regional PPO
	Enrollment	% Medicare Eligibles4	Enrollment	% Total Enrollment	Enrollment	% Total Enrollment	Enrollment	% Total Enrollment
1 For our focused analysis of 2006 MA enrollment (Tables 5-3 to 5-7), we exclude employer-only plans, Part B-only plans, and enrollment in Puerto Rico and U.S. Territories. However, due to data limitations, for our trend analysis of 2000-2006 MA enrollment (Tables 5-1 and 5-2), we present unrestricted MA enrollment counts. Therefore, 2006 MA enrollment counts and penetration rates in Tables 5-1 and 5-2 will necessarily be higher than 2006 MA enrollment counts and penetration rates in Tables 5-3 to 5-7. 2 Enrollment as of a given month in the year: December for 2000; July for 2001-2005; August for 2006. 3 Due to data limitations, we were not able to separate out HMO and local PPO. Includes HMO POS and PSO plans. For years 2003-2005, PPO demo plans included. 4 Medicare eligibles for 2000-2005 from Medicare & Medicaid Statistical Supplement. Medicare eligibles for 2006 imputed using-growth rate in Medicare population from 2004 to 2005 applied to 2005, as derived from the Statistical Supplement (1.8 percent). SOURCE: RTI analysis of 2000-2006 Medicare Monthly Contract Reports.
2000	6,261,727	15.8	6,260,549	100.0	1,178	0.0	–	–
2001	5,620,524	14.0	5,603,322	99.7	17,202	0.3	–	–
2002	4,990,083	12.3	4,966,779	99.5	23,304	0.5	–	–
2003	4,713,041	11.5	4,690,032	99.5	23,009	0.5	–	–
2004	4,774,125	11.4	4,736,768	99.2	37,357	0.8	–	–
2005	5,186,809	12.2	5,067,086	97.7	119,723	2.3	–	–
2006	6,813,397	15.7	5,921,837	86.9	802,068	11.8	89,492	1.3

Table 5-2 shows clearly the magnitudes of the recent increases in MA enrollment. Between 2005 and 2006, there has been an increase in MA enrollment of 1.6 million, with 854,751 of this increase for local coordinated care plans, and 682,345 for PFFS plans. There are several factors that might explain these increases in MA enrollment. One likely key factor is higher MA payments, and in particular payments to plans operating in areas where MA benchmarks are based on urban or rural "floor" rates. The creation of floor rates, originally established in the BBA and subsequently expanded to include urban floors, helped to make MA plan options more widely available to Medicare beneficiaries, by allowing plans in areas that previously had little or no MA availability to offer additional benefits to enrollees.

Table 5-2
Change in Medicare Advantage¹ enrollment, 2001–2006

Year2	Total		HMO/Local PPO3		PFFS		Regional PPO
	Change Enrollment4	% Change Enrollment4	Change Enrollment4	% Change Enrollment4	Change Enrollment4	% Change Enrollment4	Change Enrollment4	% Change Enrollment4
1 For our focused analysis of 2006 MA enrollment (Tables 5-3 to 5-7), we exclude employer-only plans, Part B-only plans, and enrollment in Puerto Rico and U.S. Territories. However, due to data limitations, for our trend analysis of 2000–2006 MA enrollment (Tables 5-1 and 5-2), we present unrestricted MA enrollment counts. Therefore, 2006 MA enrollment counts and penetration rates in Tables 5-1 and 5-2 will necessarily be higher than 2006 MA enrollment counts and penetration rates in Tables 5-3 to 5-7. 2 Enrollment as of a given month in the year: December for 2000; July for 2001–2005; August for 2006. 3 Due to data limitations, we were not able to separate out HMO and local PPO. Includes HMO POS and PSO plans. For years 2003-2005, PPO demo plans included. 4 Change from prior year. SOURCE: RTI analysis of 2000-2006 Medicare Monthly Contract Reports.
2001	-641,203	-10.2	-657,227	-10.5	16,024	1360.3	–	–
2002	-630,441	-11.2	-636,543	-11.4	6,102	35.5	–	–
2003	-277,042	-5.6	-276,747	-5.6	-295	-1.3	–	–
2004	61,084	1.3	46,736	1.0	14,348	62.4	–	–
2005	412,684	8.6	330,318	7.0	82,366	220.5	–	–
2006	1,626,588	31.4	854,751	16.9	682,345	569.9	89,492	–

In addition to payments, the implementation of Medicare Part D in 2006 could have had an impact on MA enrollment. For example, many MA plans used some of the "rebate" dollars from bidding below their Part C benchmark to lower their Part D premiums and enhance their Part D benefits. Also, the implementation of Part D allowed MA organizations to cross-market their Part C and their Part D plans, and as a result increase their Part C enrollment. On the other hand, the fact that FFS beneficiaries are allowed to enroll in standalone drug plans could have had a negative effect on MA enrollment because health plan enrollment was no longer necessary to obtain a Medicare drug benefit. Finally, greater availability of plans in all areas, and of plan types offering less restrictive access to providers, likely enhanced MA enrollment in 2006.

PFFS enrollment rose substantially between 2005 and 2006, by 682,345 beneficiaries. PFFS enrollment grew by nearly as many beneficiaries as HMO enrollment, despite starting from a much smaller base than HMOs, which have historically dominated the MA market. Regional PPOs added another 89,492 in MA enrollment in 2006.

5.2 Medicare Advantage Enrollment in 2006

Our analysis sample for monitoring 2006 MA enrollment was beneficiaries enrolled in an MA plan (HMO¹⁷, local PPO, regional PPO, PFFS), excluding employer-only plan enrollment, Part B-only plan enrollment, and enrollment in Puerto Rico and U.S. Territories. We examined a point in time sample, specifically, all beneficiaries enrolled on July 1, 2006 as indicated in the Medicare Beneficiary Database.

MA Enrollment Distributions. Table 5-3 shows 2006 MA enrollment overall, and by plan type, urbanicity, census region, age, sex, and Medicaid dual eligibility. As shown in the table, MA enrollment in 2006 was 5.5 million, with a penetration rate of 14.2 percent. Although HMO was still the dominant player in MA, together PFFS and PPOs (local and regional) had about 20 percent of the MA market share. Among MA enrollees, 91.4 percent resided in urban areas, with only 8.6 percent in rural areas. At 16.6 percent, the percentage of beneficiaries residing in urban areas that take-up MA was much higher than for rural beneficiaries (5.7 percent). The South and West each had the highest number of MA enrollees among census regions, with 1.7 million each. However, among Medicare beneficiaries eligible for MA in the West census region, the take-up rate for MA was about twice that of the South census region (22.1 versus 11.9 percent).

Table 5-3
Medicare Advantage¹ enrollment by plan, geographic, and beneficiary characteristics, 2006²

	Enrollment	% of Total Enrollment	% of Total Eligibles3	% of Subpopulation Eligibles3
1 For our focused analysis of 2006 MA enrollment (Tables 5-3 to 5-7), we exclude employer-only plans, Part B-only plans, and enrollment in Puerto Rico and U.S. Territories. However, due to data limitations, for our trend analysis of 2000-2006 MA enrollment (Tables 5-1 and 5-2), we present unrestricted MA enrollment counts. Therefore, 2006 MA enrollment counts and penetration rates in Tables 5-1 and 5-2 will necessarily be higher than 2006 MA enrollment counts and penetration rates in Tables 5-3 to 5-7. 2 Enrollment figures as of July 1, 2006. 3 Medicare Advantage eligibles calculated using Medicare Denominator File. Eligibles restricted to beneficiaries with Part A and B. 4 Includes HMO, POS, and PSO plans. 5 Urbanicity is undefined in our analysis for a few counties (e.g., certain counties in Alaska), and therefore the sum of Urban and Rural enrollment is slightly less than Total enrollment. SOURCE: RTI Analysis of Medicare Beneficiary Database and 2006 Medicare Denominator File.
Total Medicare Advantage	5,518,099	100	14.2	–
Plan Type
HMO4	4,389,569	79.5	11.3	–
Local PPO	273,797	5.0	0.7	–
Regional PPO	86,409	1.6	0.2	–
PFFS	768,324	13.9	2.0	–
Urbanicity5
Urban	5,041,727	91.4	13.0	16.6
Large Urban	3,599,166	65.2	9.3	19.8
Medium Urban	1,151,799	20.9	3.0	14.5
Small Urban	290,762	5.3	0.8	7.0
Rural	475,001	8.6	1.2	5.7
Rural Urban-Adjacent	369,328	6.7	1.0	6.7
Rural Non-Adjacent	105,673	1.9	0.3	3.7
Census Region
Northeast	1,297,236	23.5	3.3	17.3
Midwest	805,511	14.6	2.1	8.9
South	1,725,749	31.3	4.5	11.9
West	1,689,603	30.6	4.4	22.1
Age
Under 65	664,904	12.0	1.7	10.0
65-74	2,492,802	45.2	6.4	15.4
75-84	1,808,565	32.8	4.7	15.4
85 and older	551,828	10.0	1.4	13.3
Sex
Male	2,319,362	42.0	6.0	13.8
Female	3,198,737	58.0	8.3	4.16
Dual Eligibility
Medicaid	793,643	14.4	2.0	11.8
Non-Medicaid	4,724,456	85.6	12.2	14.8

The youngest elderly group (age 65–74) made up the highest percentage of MA enrollment (45.2 percent), with the age 75–84 group having 32.8 percent of MA enrollment. Note also that the MA take-up rate among these age groups was somewhat higher than among the oldest Medicare beneficiaries (age 85 or older) and the Medicare beneficiaries eligible by disability (age 0–64). Beneficiaries dually eligible for Medicare and Medicaid accounted for 14.4 percent of MA enrollees, but had a lower take-up rate for MA than do non-Medicaid enrollees.

Plan Type by Urbanicity and Census Region. Table 5-4 lists MA enrollment counts for plan types by urbanicity and census region, and Table 5-5 lists the row and column percentages. As shown in the tables, among the 4.4 million HMO enrollees, only 4.0 percent (column percentage in Table 5-5) resided in rural areas. This can be contrasted with the 768,324 PFFS enrollees, where 34.1 percent reside in rural areas. Nearly two-thirds of PFFS enrollment was in urban areas (65.2 percent), with most of the urban PFFS enrollment in medium and small urban areas. Clearly PFFS raised MA enrollment in rural areas. The Conference Report for the MMA of 2003 cites the decline in plan participation and indicates that the immediate changes to the payment methodology for the MA program were included in the law to "encourage plan entry," adding that "The goal is to increase beneficiary choice, by increasing private plan participation in Medicare."  The MMA Conference Report also refers to bringing greater health plan choices to areas not previously served by private plans, particularly rural areas.

Table 5-4
Medicare Advantage¹ enrollment, plan type by urbanicity and census region, 2006²

	Total	HMO3	Local PPO	Regional PPO	PFFS
1 For our focused analysis of 2006 MA enrollment (Tables 5-3 to 5-7), we exclude employer-only plans, Part B-only plans, and enrollment in Puerto Rico and U.S. Territories. However, due to data limitations, for our trend analysis of 2000-2006 MA enrollment (Tables 5-1 and 5-2), we present unrestricted MA enrollment counts. Therefore, 2006 MA enrollment counts and penetration rates in Tables 5-1 and 5-2 will necessarily be higher than 2006 MA enrollment counts and penetration rates in Tables 5-3 to 5-7. 2 Enrollment figures as of July 1, 2006. 3 Includes HMO, POS, and PSO plans. 4 Urbanicity is undefined in our analysis for a few counties (e.g., certain counties in Alaska), and therefore the sum of Urban and Rural enrollment is slightly less than Total enrollment. SOURCE: RTI Analysis of Medicare Beneficiary Database.
Total Enrollment	5,518,099	4,389,569	273,797	86,409	768,324
Urbanicity4
Urban	5,041,727	4,213,386	251,361	70,921	506,059
Large Urban	3,599,166	3,276,001	149,807	38,263	135,095
Medium Urban	1,151,799	803,305	82,544	24,196	241,754
Small Urban	290,762	134,080	19,010	8,462	129,210
Rural	475,001	174,868	22,413	15,487	262,233
Rural Urban-Adjacent	369,328	160,865	19,042	11,771	177,650
Rural Non-Adjacent	105,673	14,003	3,371	3,716	84,583
Census Region
Northeast	1,297,236	1,168,678	94,761	3,736	30,061
Midwest	805,511	459,150	46,449	10,436	289,476
South	1,725,749	1,280,977	57,979	50,396	336,397
West	1,689,603	1,480,764	74,608	21,841	112,390

Table 5-5
Medicare Advantage¹ enrollment, plan type by urbanicity and census region, row and column percentages, 2006²

	Total	Row Percentages				Total	Column Percentages
		HMO3	Local PPO	Regional PPO	PFFS		HMO3	Local PPO	Regional PPO	PFFS
1 For our focused analysis of 2006 MA enrollment (Tables 5-3 to 5-7), we exclude employer-only plans, Part B-only plans, and enrollment in Puerto Rico and U.S. Territories. However, due to data limitations, for our trend analysis of 2000–2006 MA enrollment (Tables 5-1 and 5-2), we present unrestricted MA enrollment counts. Therefore, 2006 MA enrollment counts and penetration rates in Tables 5-1 and 5-2 will necessarily be higher than 2006 MA enrollment counts and penetration rates in Tables 5-3 to 5-7. 2 Enrollment figures as of July 1, 2006. 3 Includes HMO POS and PSO plans. SOURCE: RTI Analysis of Medicare Beneficiary Database.
Total	100	79.5	5.0	1.6	13.9	100	100	100	100	100
Urbanicity
Urban	100	83.6	5.0	1.4	10.0	91.4	96.0	91.8	82.1	65.9
Large Urban	100	91.0	4.2	1.1	3.8	65.2	74.7	54.7	44.3	17.6
Medium Urban	100	69.7	7.2	2.1	21.0	20.9	18.3	30.2	28.0	31.5
Small Urban	100	46.1	6.5	2.9	44.4	5.3	3.1	6.9	9.8	16.8
Rural	100	36.8	4.7	3.3	55.2	8.6	4.0	8.2	17.9	34.1
Rural Urban-Adjacent	100	43.6	5.2	3.2	48.1	6.7	3.7	7.0	13.6	23.1
Rural Non-Adjacent	100	13.3	3.2	3.5	80.0	1.9	0.3	1.2	4.3	11.0
Census Region
Northeast	100	90.1	7.3	0.3	2.3	23.5	26.6	34.6	4.3	3.9
Midwest	100	57.0	5.8	1.3	35.9	14.6	10.5	17.0	12.1	37.7
South	100	74.2	3.4	2.9	19.5	31.3	29.2	21.2	58.3	43.8
West	100	87.6	4.4	1.3	6.7	30.6	33.7	27.2	25.3	14.6

The regional PPO option was created, in part, to provide more MA options to rural beneficiaries. They drew 17.9 percent of their total enrollment from rural areas, more than double the percentage of HMOs or local PPOs, but about half the percentage of PFFS. Regional PPOs accounted for 3.3 percent of total rural MA enrollment in 2006 (Table 5-5 row percentage). Over half of rural MA enrollees were in PFFS plans, with most of the rest in HMOs. In contrast, 83.6 percent of urban MA enrollees were in HMOs, with only 10.0 percent in PFFS plans.

As listed in Tables 5-4 and 5-5, among the 1.3 million MA enrollees residing in the Northeast census region, nine out of 10 enrollees were in an HMO (90.1 percent). The West region was also dominated by HMOs, with 87.6 percent of Western enrollees (row percentage in Table 5-5). This substantially differs from the Midwest and South census regions, where a higher proportion of MA enrollees chose PFFS plans. For example, among the 805,511 Midwestern MA enrollees, 57.0 percent were in an HMO, with 35.9 percent in a PFFS plan. Given that the MA take-up rate for Midwesterners was relatively low (8.9 percent on Table 5-3), PFFS plans appeared to be an important MA option in the Midwest. Over half of regional PPO enrollment was in the South (Table 5-5 column percentage). Regional PPOs and PFFS plans reflected less than 3 percent of MA enrollment in the Northeast.

Special Needs Plan Enrollment. Table 5-6 provides SNP enrollment by MA plan type. 348,842 (6.3 percent) of MA enrollees were enrolled in a SNP, with the vast majority enrolled in a SNP serving beneficiaries dually eligible for Medicare and Medicaid. Among SNP enrollees, 93.9 percent were enrolled in a dual eligible SNP, with 5.7 percent enrolled in an institutional SNP, and 0.4 percent enrolled in a SNP for the chronically ill. About 92 percent of SNP enrollees were in HMOs. Interestingly, local PPOs had the highest percentage of their enrollment in SNPs (9.0 percent), with a relatively strong institutional SNP presence. SNFs can only be offered as a coordinated care plan; a SNP cannot be offered through the PFFS model.

Table 5-6
Special needs plan enrollment, by plan type, 2006¹

	Total	HMO2	Local PPO	Regional PPO	PFFS
1 Enrollment figures as of July 1, 2006. 2 Includes HMO POS and PSO plans. 3 For our focused analysis of 2006 MA enrollment (Tables 5-3 to 5-7), we exclude employer-only plans, Part B-only plans, and enrollment in Puerto Rico and U.S. Territories. However, due to data limitations, for our trend analysis of 2000-2006 MA enrollment (Tables 5-1 and 5-2), we present unrestricted MA enrollment counts. Therefore, 2006 MA enrollment counts and penetration rates in, Tables 5-1 and 5-2 will necessarily be higher than 2006 MA enrollment counts and penetration rates in Tables 5-3 to 5-7. SOURCE: RTI Analysis of Medicare Beneficiary Database.
Total Medicare Advantage3 Enrollment	5,518,099	4,389,569	273,797	86,409	768,324
SNP Enrollment	348,842	320,818	24,659	3,365	0
% of total enrollment	6.3	7.3	9.0	3.9	0.0
Dual Eligible SNP	327,594	317,516	6,713	3,365	
% of SNP enrollment	93.9	99.0	27.2	100.0	
Institutional SNP	19,758	1,812	17,946	0	
% of SNP enrollment	5.7	0.6	72.8	0.0	
Chronic Condition SNP	1,490	1490	0	0	
% of SNP enrollment	0.4	0.5	0.0	0.0	
Non-SNP Enrollment	5,169,257	4,068,751	249,138	83,044	768,324
% of total enrollment	93.7	92.7	91.0	96.1	100.0

Part D Plan Enrollment. Table 5-7 lists Part D enrollment statistics for MA enrollees. The vast majority of MA enrollees were enrolled in the Medicare Part D drug program. Among the 5.5 million MA enrollees, 5.1 million (92.7 percent) were enrolled in Part D, with approximately a 90 percent Part D take-up rate for each plan type. PFFS enrollees were slightly less likely to have Part D coverage than enrollees in other plan types. Almost all of the MA/Part D enrollees were enrolled in an MA-PD (96.4 percent), although 3.6 percent of MA/Part D enrollees were enrolled in a standalone PDP. PFFS plans are not required to offer a Part D plan, and if they do not, their enrollees are allowed under Part D program rules to enroll in a standalone drug plan. About 27 percent of PFFS enrollees with Part D coverage were enrolled in standalone drug plans.

Table 5-7
Part D enrollment in Medicare Advantage¹, by plan type, 2006²

	Total	HMO3	Local PPO	Regional PPO	PFFS
1 For our focused analysis of 2006 MA enrollment (Tables 5-3 to 5-7), we exclude employer-only plans, Part B-only plans, and enrollment in Puerto Rico and U.S. Territories. However, due to data limitations, for our trend analysis of 2000-2006 MA enrollment (Tables 5-1 and 5-2), we present unrestricted MA enrollment counts. Therefore, 2006 MA enrollment counts and penetration rates in Tables 5-1 and 5-2 will necessarily be higher than 2006 MA enrollment counts and penetration rates in Tables 5-3 to 5-7. 2 Enrollment figures as of July 1, 2006. 3 Includes HMO POS and PSO plans. SOURCE: RTI Analysis of Medicare Beneficiary Databases.
Total Medicare Advantage Enrollment	5,518,099	4,389,569	273,797	86,409	768,324
Part D Enrollment	5,113,964	4,106,244	249,855	79,159	678,706
% of total enrollment	92.7	93.5	91.3	91.6	88.3
MA-PD	4,931,244	4,106,244	249,855	79,159	495,986
% of Part D enrollment	96.4	100.0	100.0	100.0	73.1
PDP	182,720	0.0	0.0	0.0	182,720
% of Part D enrollment	3.6	0.0	0.0	0.0	26.9
Non-Part D Enrollment	404,135	283,325	23,942	7,250	89,618
% of total enrollment	7.3	6.5	8.7	8.4	11.7

GLOSSARY OF ACRONYMS

AAPCC	Adjusted average per capita cost
BIPA	Benefits Improvement and Protection Act of 2000
BBRA	Balanced Budget Refinements Act of 1999
CCP	Coordinated care plan
CMS	Centers for Medicare & Medicaid Services
CMS-HCC	CMS Hierarchical Condition Categories (risk adjustment system)
FFS	Fee for service
HMO	Health maintenance organization
HMO POS	Health maintenance organization, point of service (plan)
HPMS	Health Plan Management System
MA	Medicare Advantage
MA-PD	Medicare Advantage prescription drug plan
MBD	Medicare beneficiary database
MMA	Medicare Modernization Act of 2003
MSA	Medical savings account (plan)
PDP	Prescription drug plan (standalone Part D plan)
PFFS	Private fee-for-service (plan)
PIP-DCG	Principal Inpatient Diagnostic Cost Group (risk adjustment system)
PPO	Preferred provider organization
PSO	Provider sponsored organization (plan)
RTI	Research Triangle Institute
SNP	Special needs plan
TrOOP	True out of pocket costs

REFERENCES

Blum, J., R. Brown, and M. Frieder (2007). "An Examination of Medicare Private Fee-for-Service Plans." Medicare Issue Brief. Prepared for the Kaiser Family Foundation.

CMS 2005a Advance Notice of Methodological Changes for Calendar Year (CY) 2006 Medicare Advantage (MA) Payment Rates, 2/18/2005. http://www.cms.hhs.gov/MedicareAdvtgSpecRateStats/Downloads/Advance2006.pdf

Hurley, R.E., B.C. Strunk, J.M. Grossman. (2004). "Preferred Provider Organizations in Medicare: Is There an Advantage?" Issue Brief for Center for Studying Health System Change.

Kaiser Family Foundation, Daily Health Policy Report, August 9, 2006.

MedPAC. (2007). The Medicare Advantage Program and MedPAC Recommendations. Congressional Testimony of M Miller, Executive Director, MedPAC

Mello, M.M., Stearns, S.C., Norton, E.C., and Ricketts, T.C. III (June 2003). "Understanding Biased Selection in Medicare HMOs." Health Services Research 38(3):961-992.

Pope, G.C., L.M. Greenwald, D.A. Healy, J. Kautter, E. Olmsted, and N. West (September 2006). "Impact of Increased Financial Incentives to Medicare Advantage Plans." Final Report, submitted to Centers for Medicare & Medicaid Services, Contract No. 500-00-0024, T.O.#17.

COMPUTER OUTPUT TABLE

Table 3-1 Number of Medicare Advantage contracts, by plan type

Computer Output:
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\Table3_1_2006_march14.xls
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_march14.log
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_march14.do

Table 3-2 Percent of counties with at least one Medicare Advantage contract, by plan type

Computer Output:
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\Table3_1_2006_april04.xls
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_april04.log
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_april04.do

Table 3-3 Number and percentage of Medicare beneficiaries with access to a Medicare Advantage plan, by plan type

Computer Output:
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\Table3_3_2006_snp_feb20.xls
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_snp_feb20.log
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_snp_feb20.do

Table 3-4 Percent of counties with at least one Medicare Advantage contract, by plan type and urbanicity

Computer Output:
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\Table3_4_2006_snp_feb20.xls
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_snp_feb20.log
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_snp_feb20.do

Table 3-5 Percent of counties with at least one Medicare Advantage contract, by plan type and region

Computer Output:
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\Table3_5_6_2006_snp_feb20.xls
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_snp_feb20.log
H:\project\07964\017 FAMA\pgm\ykaganova\Programs\request4_no06630_2006_snp_feb20.do

Table 3-6 Percentage of beneficiaries and counties with access to Medicare Advantage plan types, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.do

Table 3-7 Percentage of beneficiaries with access to Medicare Advantage plan types, by urbanicity, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.do

Table 3-8 Percentage of beneficiaries with access to Medicare Advantage plan types, by region, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.do

Table 3-9 Percentage of beneficiaries with access to Medicare Advantage contracts, by number of contracts and urbanicity, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.do

Table 3-10 Percentage of beneficiaries with access to Medicare Advantage contracts, by number of contracts and region, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.do

Table 3-11 Number of special needs contracts and plans by plan type and target beneficiaries, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.do

Table 4-1 Mean monthly premiums of Medicare Advantage plans by plan type, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_nov06.do
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_nov06.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\request4_no06630_2006_nt_march14.do

Table 4-2 Percent of Medicare Advantage enrollees in zero premium plans by plan type, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april4.do
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april4.log

Table 4-3 Percent of enrollees in Medicare Advantage plans, by premium range, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april4.do

Table 4-4 Mean monthly premiums of Medicare Advantage plans by urbanicity, 2006

Computer Output:
K:\Project\07964\017 FAMA\common\dulisse\design report Tables 0612\enrollment\jk_enrollment_request2.log
K:\Project\07964\017 FAMA\common\dulisse\design report Tables 0612\enrollment\jk_enrollment_request2.do

Table 4-5 Percent of Medicare Advantage enrollees in zero premium plans by urbanicity, 2006

Computer Output:
K:\Project\07964\017 FAMA\common\dulisse\design report Tables 0612\enrollment\jk_enrollment_request2.log
K:\Project\07964\017 FAMA\common\dulisse\design report Tables 0612\enrollment\jk_enrollment_request2.do

Table 4-6 Mean monthly premiums of Medicare Advantage plans by region, 2006

Computer Output:
K:\Project\07964\017 FAMA\common\dulisse\design report Tables 0612\enrollment\jk_enrollment_request2.log
K:\Project\07964\017 FAMA\common\dulisse\design report Tables 0612\enrollment\jk_enrollment_request2.do

Table 4-7 Percent of Medicare Advantage enrollees in zero premium plans by region, 2006

Computer Output:
K:\Project\07964\017 FAMA\common\dulisse\design report Tables 0612\enrollment\jk_enrollment_request2.log
K:\Project\07964\017 FAMA\common\dulisse\design report Tables 0612\enrollment\jk_enrollment_request2.do

Table 4-8 Part B premium reduction by Medicare Advantage plan type, urbanicity, and region, 2006
Percent of enrollees

Computer output:
H:\project\07964\017fama\pgm\ykaganova\Programs\table_4_17.xls
H:\project\07964\017fama\pgm\ykaganova\Programs\gp_request_april30a.log
H:\project\07964\017fama\pgm\ykaganova\Programs\gp_request_april30a.do

Table 4-9 Prescription drug benefits by Medicare Advantage plan type, 2006
Percent of enrollees

Computer output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_march28.do
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_march28.log

Table 4-10 Prescription drug benefits of Medicare Advantage enrollees by urbanicity, 2006 Percent of enrollees

Computer output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april6.do
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april6.log

Table 4-11 Prescription drug benefits of Medicare Advantage enrollees by region, 2006
Percent of enrollees

Computer output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april6.do
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april6.log

Table 4-12 Cost sharing before the initial coverage limit, by type of Medicare Advantage prescription drug plan, 2006
Percent of enrollees

Computer output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april13.do
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april13.log

Table 4-13 Drug tiers of Medicare Advantage prescription drug plans, 2006
Percent of enrollees

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april6.do
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_april6.log

Table 4-14 Common cost sharing structures in Medicare Advantage prescription drug plans, 2006 Median copayments or coinsurance by drug tier

Computer output:
H:\project\07964\017fama\pgm\ykaganova\Programs\gp_request_may3.do
H:\project\07964\017fama\pgm\ykaganova\Programs\gp_request_may3.log
H:\project\07964\017fama\pgm\ykaganova\Programs\gp_request_master_may3.do
H:\project\07964\017fama\pgm\ykaganova\Programs\Table_18_cost_sharing_structures

Table 4-15 Initial coverage limit in Medicare Advantage prescription drug plans, by plan and geographic characteristics, 2006
Percent of enrollees

Computer output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\tables_gc_1_2_3_4_5_6_7.xls
H:\project\07964\017 fama\pgm\ykaganova\Programs\gp_gap_filling_request1a.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\gp_gap_filling_request1a.do

Table 4-16 Gap coverage in Medicare Advantage prescription drug plans, by plan and geographic characteristics, 2006
Percent of enrollees

Computer output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\tables_gc_1_2_3_4_5_6_7.xls
H:\project\07964\017 fama\pgm\ykaganova\Programs\gp_gap_filling_request1a.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\gp_gap_filling_request1a.do

Table 4-17 Selected characteristics of Medicare Advantage prescription drug plan formularies, by plan and geographic characteristics, 2006 Enrollment-weighted plan mean

Computer output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\tables_gc_1_2_3_4_5_6_7.xls
H:\project\07964\017 fama\pgm\ykaganova\Programs\gp_gap_filling_request1a.log
H:\project\07964\017 fama\pgm\ykaganova\Programs\gp_gap_filling_request1a.do

Table 4-18 Selected mandatory supplemental benefits in Medicare Advantage plans, 2006 Percent of enrollees with benefit

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_oct22.do
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_oct22.log

Table 4-19 Cost sharing for selected services in Medicare Advantage plans, 2006

Computer Output:
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_oct22.do
H:\project\07964\017 fama\pgm\ykaganova\Programs\bd_request3_oct22.log

Table 4-20 Out of pocket (OOP) maximums in Medicare Advantage plans, 2006
Percent of enrollees

Computer output:
H:\project\07964\017fama\pgm\ykaganova\Programs\gp_request_may1.do
H:\project\07964\017fama\pgm\ykaganova\Programs\table_4_19.xls
H:\project\07964\017fama\pgm\ykaganova\Programs\gp_request_may1.log

Table 4-21 Out of pocket (OOP) maximums in Medicare Advantage plans, by plan type, urbanicity, and region, 2006

Computer output:
H:\project\07964\017fama\pgm\ykaganova\Programs\gp_request_may1.do
H:\project\07964\017fama\pgm\ykaganova\Programs\table_4_20.xls
H:\project\07964\017fama\pgm\ykaganova\Programs\gp_request_may1.log

Table 5-1 Medicare Advantage enrollment, 2000-2006

No computer output

Table 5-2 Change in Medicare Advantage enrollment, 2001-2006

No computer output

Table 5-3 Medicare Advantage enrollment by plan, geographic, and beneficiary characteristics, 2006

Computer Output: H:\project\07964\017 fama\pgm\ykaganova\Programs\jk_enrollment_request3-7

Table 5-4 Medicare Advantage enrollment, plan type by urbanicity and census region, 2006

Computer Output: H:\project\07964\017 fama\pgm\ykaganova\Programs\jk_enrollment_request5.log

Table 5-5 Medicare Advantage enrollment, plan type by urbanicity and census region, row and column percentages, 2006

Computer Output: H:\project\07964\017 fama\pgm\ykaganova\Programs\jk_enrollment_request5.log

Table 5-6 Special needs plan enrollment, by plan type, 2006

Computer Output: H:\project\07964\017 fama\pgm\ykaganova\Programs\jk_enrollment_request6.log

Table 5-7 Part D enrollment in Medicare Advantage, by plan type, 2006

Computer Output: H:\project\07964\017 fama\pgm\ykaganova\Programs\jk_enrollment_request6.log

Endnotes

¹ All premium and benefits results are weighted by plan enrollment and thus represent the average (or median) enrolled beneficiary premium or benefits, not average plan offerings. Premiums are those charged by plans, and are not necessarily paid out of pocket by enrollees (e.g., enrollees receiving Part D low income subsidy assistance do not themselves pay the full Part D premium).

² Average Parts C+D premiums do not equal the sum of the Part C and the Part D premiums because some MA plans do not offer Part D. Part D and total premiums (Parts C+D) are for MA plans offering Part D.

³ This percentage is of all MA enrollees, including those in MA plans not offering Part D.

⁴ Copayments are for a 30-day drug supply at in-network retail pharmacies.

⁵ The basis of the Report to Congress, with some subsequent updating, is available as the Final Report of the first phase of this project (Pope et al., 2006).

⁶ A payment for a prescription drug will constitute an "incurred cost" and could count toward a beneficiary's TrOOP threshold only if the payment is made by or on behalf of the beneficiary. Assistance from a state pharmaceutical assistance program or from a charity generally will count toward the TrOOP threshold. If the beneficiary is reimbursed for the costs by insurance, a group health plan, or other third-party arrangement, the costs will not count toward the TrOOP threshold. Payments for drugs that are not included on the plan formulary also will not be counted toward the TrOOP threshold.

⁷ As noted, we also group the uncommon "provider sponsored organization" (PSO) plan type with HMOs in our analyses.

⁸ A contract is an agreement between an MA organization and CMS to offer Medicare health plans in an area. A plan is a specific benefit package offered by the MA organization. One or more plans may be offered under a single contract but each contract is limited to one plan type, e.g., HMOs, local PPOs.

⁹ In general, each contract contains plans of a single type, e.g., HMO, PPO, etc. The exception is SNPs. See Chapter 2.

¹⁰ The data source was changed because these data were not published by CMS for 2006 (and in fact data for previous years was removed from the CMS Web site.) In order to facilitate a comparison of 2006 data to 2005, the results for 2005 were recalculated using the denominator file; as a result, the 2005 results reported here differ from earlier Tables reported in Pope et al., 2006.

¹¹ Table 4-16 excludes SNPs. Beneficiaries with the Part D low income subsidy benefit may have most of their cost sharing eliminated and thus, effectively, do not face a coverage gap even if their plan has one.

¹² As opposed to "optional supplemental" benefits offered as riders with an additional premium that a plan enrollee may accept or decline.

¹³ MA Plans' OOP maximums do not pertain to enrollee OOP costs for Part D-covered drugs. Part D OOP costs are governed by a separate set of MMA-mandated rules revolving around the "true OOP cost" concept. MA plans' OOP maximums also do not apply to non-covered benefits, such as long-term care.

¹⁴ For our detailed cross-sectional analysis of MA enrollment in 2006, we were able to exclude enrollment in employer-only plans, Part B-only plans, and Puerto Rico and U.S. Territories. However, for our trend analysis of 2000-2006 MA enrollment, due to data limitations we were not able to make these exclusions, and therefore present unrestricted MA enrollment statistics for our 2000-2006 trends. Thus, 2006 MA enrollment counts and penetration rates in our 2000-2006 trend analysis will necessarily be higher than in our detailed cross-sectional analysis of MA enrollment in 2006.

¹⁵ This decline in enrollment in the early part of this decade was likely in large part a response to the Balanced Budget Act (BBA) payment changes coupled with rising medical cost inflation, which caused many plans to withdraw or contract service areas, creating "involuntary" disenrollment. In addition, BBA payment constraints combined with medical cost inflation caused many plans to raise premiums, and reduce benefits for enrollees, which also contributed to the decline in enrollment.

¹⁶ The Medicare Monthly Contract Reports data do not allow HMOs and local PPOs to be distinguished throughout this time period. Hence, we combined them for this analysis.

¹⁷ Includes HMO POS and PSO plans.