This Article Abstract Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when Comments are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (7) Citing Articles via Google Scholar Google Scholar Articles by Simantov, E. Articles by Bruegman, S. Search for Related Content PubMed PubMed Citation Articles by Simantov, E. Articles by Bruegman, S. Related Collections Insurance Coverage Elderly

Health Tracking

MARKETWATCH

Market Failure? Individual Insurance Markets For Older Americans

Abstract

 
This study examines the viability of tax credits and nongroup markets for covering uninsured adults ages fifty to sixty-four. We find that adults in this age group covered by nongroup plans tend to be healthier and wealthier than the average for their peers, yet more of them go without care and experience high medical bills relative to their incomes. Individual-market premiums premiums rise steeply with age in most states and are well above employer-group rates. Costs are likely to be unaffordable for most uninsured older adults, even with large tax credits or in states with community rating. These findings indicate a need to include risk and age pooling to reach the uninsured in this age group.

Early retirees and working adults in their midlife years (ages fifty to sixty-four), not yet eligible for Medicare, form one group targeted for assistance. These adults are particularly vulnerable to lack of health insurance if they sever their links to the workforce and employer-based coverage because of illness, disability, early retirement, or loss of a job late in their career.² At an age of increasing health concerns, these adults are also likely to be at high risk if uninsured. Long periods without health insurance are often associated with deteriorating health and low use of medical care despite declining health.³

Nevertheless, this age group is unlikely to be eligible for Medicaid or Medicare.⁴ These persons are then left with the option of purchasing individual insurance directly in the open market—a typically expensive alternative that is often unaffordable or unavailable. Unlike employer-based coverage, eligibility and premiums in individual markets are typically determined on the basis of the risk associated with each applicant’s demographic characteristics and health status.⁵ In most areas of the country, individual markets are largely unregulated in terms of pricing or benefits.

In some states individual coverage may not be available at any price if a person has health problems. Only fifteen states require insurers in the individual market to guarantee coverage to all participants regardless of health status or claims experience.⁶ In all other states persons who are older or in poor health may be denied coverage altogether if they are not eligible under the 1996Health Insurance Portability and Accountability Act (HIPAA).⁷ Even when individual insurance is available, carriers in all but a few states can set premiums much higher for older or sicker adults in anticipation of higher risks.⁸

A better understanding of the demographics of uninsured adults ages fifty to sixty-four and the experiences of those who purchase private coverage individually will be key to evaluating alternative policy strategies. To explore these experiences and related issues, we draw on two data sources: a recent survey that targeted adults in this age group and a sample of current premiums and coverage policies prevailing in individual insurance markets in selected cities.

Data Sources And Methods

Top Data Sources And Methods Findings Conclusions And Policy... NOTES

 
Commonwealth Fund Health Care Survey. Conducted by Princeton Survey Research Associates from August through November 1999, the survey consisted of twenty-minute telephone interviews with a random, national sample of 2,000 adults, ages fifty to seventy, living in telephone households in the continental United States. Our study uses the subgroup of 1,523 persons ages fifty to sixty-four who participated in the survey. In the analysis, responses were weighted to the adult population in this age group by sex, age, race/ethnicity, education, region, and household size based on the U.S. Census Bureau’s March 1999 Current Population Survey. The weighted sample of respondents represents thirty-nine million adults ages fifty to sixty-four. The response rate for the full survey was 57 percent.⁹ The survey has an overall margin of error of ±2 percent.

To analyze and compare demographic characteristics and health care experiences of these adults by insurance status, we divided the sample into five mutually exclusive insurance categories: employer-sponsored, public, individually purchased, Veterans Affairs (VA)/military, and uninsured. Persons with more than one source of coverage were assigned to the first group in the hierarchy. Thus, adults classified as having individual insurance did not have coverage from an employer or from a public program. From each insurance category we then pulled out respondents who reported a period of time uninsured in 1999, to compare insured persons who had not been uninsured during the year. (The survey asked the insured if they had been uninsured earlier in the year.) In this analysis, the uninsured group includes anyone uninsured when surveyed plus adults who reported being uninsured at some time during the year. This categorization divided respondents as follows: employer (68 percent), public (7 percent), individually purchased (8 percent), VA/military (2 percent), and uninsured (15 percent).¹⁰ We excluded VA/military from comparison tables because of the small sample size.

Our exhibits indicate tests for statistically significant differences between each insurance group and employer-based coverage, based on chi-square analyses. We also compared characteristics of respondents with individual insurance with those of publicly insured respondents and the uninsured. All comparisons highlighted in the text were statistically significant at the p .05 level or better.

Individual-market premium rates. To assess premiums and benefits in the nongroup health insurance market, we used two Internet sites that provide comparative data for multiple states and cities across the United States.¹¹ To examine a range of markets, we selected a sample of fifteen large cities with a geographic spread designed to include each major region of the country and states with varying levels of individual-market reforms.¹² Twelve cities were in states that have not passed major individual-market reforms other than compliance with HIPAA; three were in states with more comprehensive reforms: Massachusetts, community rating within age bands, and New York and New Jersey, community rating across age groups.¹³

Rates are typically quoted by ZIP code. We selected one residential ZIP code per city and checked to make sure that the rate in the ZIP code was representative of rates prevailing in that market.¹⁴ For each of the study markets, the Internet sites included an array of benefit and cost-sharing options. Generally, the lower the premium, the higher the out-of-pocket financial risk. The least expensive policies had deductibles and out-of-pocket limits of $5,000 or more per insured person, plus limited benefits and no drug coverage. To enable comparative analysis of premiums with relatively modest or mid-range cost sharing, we specified two levels for each market: plans with deductibles of $250 or less and plans with deductibles of $1,000 per person. We also stipulated that all plans include some drug coverage. Where several carriers offered plans within the specified deductible range and benefits, we selected the least expensive plan. Where a health maintenance organization (HMO) was the least expensive plan, the sample includes plans with no deductible for in-network services.

Findings

Top Data Sources And Methods Findings Conclusions And Policy... NOTES

 
Income and health disparities. Survey responses indicate that persons ages fifty to sixty-four covered by individual or employer coverage were typically wealthier and healthier than the uninsured or those covered by public insurance (Exhibit 1). A large proportion of those covered by public insurance rated their health as fair or poor and reported limiting disabilities and chronic conditions, and only a small proportion were working when surveyed.

In general, the health and income patterns of adults with individual coverage were quite similar to those of their employer-insured peers. Age and work status were the key distinctions between these two groups. Nearly half of the individually insured were not working at the time of the survey, more than half of those because of early retirement. Reflecting early-retirement patterns, the group with individual coverage was older than the employer-insured group. Persons ages sixty to sixty-four made up a larger proportion of the individually insured group (41 percent) than did adults insured through employers or uninsured (23 percent of each, not shown in Exhibit 1).

Experiences by insurance type. Persons in the age group we studied are at generally high risk of access problems and financial burden if not well insured. Asked whether they or their spouses had a health condition requiring prescription drugs on a regular basis, nearly three out of four study participants said yes (Exhibit 2). Reliance on regular prescription medication was also high among those covered by individual (66 percent) and employer (76 percent) coverage. However, adults insured through the individual market were about half as likely as those with employer coverage were to have coverage for medications.

Comparing the two privately insured groups, reports on overall medical care bills indicate that nongroup coverage leaves adults ages fifty to sixty-four especially vulnerable to high out-of-pocket medical expenses in general. One out of four adults with individual insurance had annual out-of-pocket medical expenses of $1,000 or more. In contrast, only 15 percent of those with employer coverage reported expenses this high. Those in fair or poor health were particularly at risk: 34 percent with individual coverage had annual out-of-pocket medical expenses of $1,000 or more, compared with 22 percent of their peers with employer coverage (not shown).

Out-of-pocket premium costs were also markedly higher in the nongroup market. Without the benefit of employers paying a share of premiums, individuals purchasing their own coverage pay the full cost of premiums in addition to out-of-pocket costs for medical care. Nearly half of respondents with individual coverage reported annual out-of-pocket premiums of $2,000 or more. In contrast, only 16 percent of adults insured through employers paid that much. Reflecting these differences, adults in the individual market were more than twice as likely to say they had difficulty paying their premium bills.

In combination, the higher premiums and out-of-pocket medical bills in the individual market add up to a high cost vis-à-vis income. Based on reports of annual and monthly costs, almost half of those with individual insurance spent $3,500 or more on health care each year. In contrast, adults with employer insurance were about one-third as likely to report expenses that high. For 21 percent of all adults with individual insurance and 30 percent of the individually insured in fair or poor health, these out-of-pocket expenditures amounted to an estimated 10 percent or more of their family income (latter not shown).

Access. Persons with individual insurance also were generally more likely than those covered by employers to experience access problems due to cost (Exhibit 3). They were twice as likely as those with employer coverage were to say that they had skipped medical tests, treatment, or follow-up, or had not seen a doctor when they had a medical problem because of cost concerns. They were also more likely to leave a prescription unfilled because of costs (although this difference did not achieve statistical significance).

Individual health insurance premiums. Exhibit 4 displays the lowest premium rates that adults ages fifty to sixty-four are likely to encounter in selected markets for a comprehensive benefit plan (including prescription benefits) with either a relatively modest deductible ($250 or less) or higher deductible ($1,000).¹⁷ In all but three of the cities (New York, Newark, and Boston, with community rating, are the exceptions) the quoted premium rates reflect the likely costs of an otherwise healthy applicant, since plans reserve the right to charge (and state laws permit) higher rates based on applicants’ health history. All health history had to be disclosed on the application questionnaire. In non-community-rated states insurers also reserved the right to obtain additional medical information through extensive follow-up health screening questionnaires or by requesting medical records. Depending on state regulations, carriers also could limit benefits for preexisting conditions or require waiting periods of twelve months or longer.

Quoted premium rates for fifty- and sixty-year- olds were the same as those for a twenty-five- year-old in the two markets with full community rating (New York and New Jersey). However, as anticipated by reforms, the spreading of risk in these markets raised the rates for twenty-five-year-olds compared with those in most other states.

Given the high annual premium costs as well as front-end deductibles, plans (with the specified benefits) in all fifteen markets are likely to be priced beyond the means of those living on low or modest incomes. A sixty-year- old with income at 150 percent of poverty (about $12,000 a year) would have to spend one-third of his or her annual income for a $1,000 deductible plan and nearly half (47 percent) of annual income for a $250 deductible policy at the median rate for this age group. A fifty-year-old with similar income would spend 21 and 29 percent, respectively. Premium costs relative to income remain high for single adults age fifty or sixty with incomes at 250 percent of poverty (about $20,000 annually), ranging from 20 to 28 percent of income for sixty-year-olds and from 13 to 17 percent for fifty-year-olds. An annual income of more than $50,000 would be necessary to hold median premium costs of the $250 deductible plan ($5,688) to no more than 10 percent of income for a sixty-year-old at prevailing rates.

An adult with health problems could expect to pay even higher rates or could be denied coverage altogether. These market dynamics dynamics are likely to be most problematic for uninsured adults age fifty or older. In the survey nearly two-thirds (62 percent) of uninsured adults ages fifty to sixty-four reported household incomes at or below 250 percent of the poverty level (Exhibit 1). The uninsured with the lowest incomes also tended to be at high risk for health problems: 42 percent of the uninsured with incomes at or below 250 percent of poverty reported fair or poor health, 39 percent said that they had a limiting disability, and 66 percent said that they had been diagnosed with at least one of five chronic conditions (data not shown).

By age fifty, uninsured adults are also less likely to hope for access to group insurance through work. Thirty-nine percent in this group were not working. Another 15 percent were working part time. Nearly one-third (31 percent) reported having a disability that limited daily activities.

Conclusions And Policy Implications

Top Data Sources And Methods Findings Conclusions And Policy... NOTES

 
The individual insurance market, as currently structured, does not work well for adults ages fifty to sixty-four who lack access to coverage at group rates through work. Our findings indicate that by age fifty those who are willing and able to pay premium costs for individual insurance are likely to find themselves paying relatively high out-of-pocket costs for health care because of restricted benefits, cost sharing, and deductibles as well as high monthly premiums. Although persons ages fifty to sixty-four with individual insurance were wealthier and healthier than average for their age group, they remained at high risk for large out-of-pocket medical expenses. Despite paying more out of pocket for premiums than those with group coverage paid, persons in this age group with individual insurance were more likely to report high out-of-pocket bills for medical care and prescription drugs and to report costs as a barrier to care. Indeed, these persons had many of the symptoms of being "underinsured."

A contrast of premiums charged in the individual insurance market for this age group with rates typical in employer group plans is revealing. Today’s system of employer-based insurance—which covers two-thirds of Americans under age sixty-five—achieves a high degree of risk pooling, especially among large firms. Because the employer market brings persons of different health risks and ages together, premiums quoted for the group market reflect a form of community rating. As a result, employer-plan rates would typically be well below individual-plan rates for persons persons ages fifty to sixty-four. For example, a recent survey of employers found that the median annual premium costs of employer insurance were $2,129 for HMO single coverage and $2,520 for a preferred provider organization (PPO) plan.¹⁸ Nearly all employer plans in that study had some type of drug coverage. The average PPO plan deductible when using in-network providers was typically low, around $181.¹⁹ Even if a person age fifty or older had to pay the entire premium out of pocket, this group rate is less than half the cost of an individual plan with a $250 deductible based on the median of rates for a sixty-year-old quoted in the fifteen cities across the country (Exhibit 4).

Employer coverage also has the advantage of lower administrative expenses, so that a greater share of the premium is allocated to medical benefits.²⁰ In addition, employers almost always contribute a substantial portion of employee premiums. Given these features of job-based group insurance, adults ages fifty to sixty-four would do well to stay attached to jobs with benefits for as long as possible to take advantage of risk pooling. This includes taking up Consolidated Omnibus Budget Reconciliation Act (COBRA) options up to the maximum number of months upon leaving work.²¹ Even without employer premium sharing, the group rate is likely to be well below that available in the individual market. Not surprisingly, studies have shown that adults in this age group with access to and eligiblity for employer-group coverage typically take up the option.²²

Study findings further indicate that loss of access to the employer-based system leaves adults ages fifty or older particularly vulnerable to loss of coverage. Since for many persons, early disengagement from the work force is motivated by health problems or disability, individual coverage may be difficult to obtain while one is waiting to become eligible for Medicare.²³

Tax credits. We began this paper with a discussion of tax credits to encourage health insurance purchase. One of the main objectives of such proposals is to reduce uninsurance rates by making coverage more affordable. With tax credits, as with other subsidies, the key question is what the individual will be able to buy with the specific credit amount. A number of current legislative proposals would provide credits of $500– $1,200 for persons buying coverage in the individual market, with amounts invariant by age.²⁴ If one considers the range of premiums in our study, such credits would be small relative to the premium rates confronting midlife adults in the nongroup market. As a result, even if they were to receive a tax credit of $1,000, low-income adults (150 percent of poverty or less) in this age group would pay between one-quarter and two-fifths of their pretax income for health insurance in the open market.²⁵ Even in states with community rating, credits in the $1,000 range would pick up at most one-third of premium costs.

To ensure access to care, subsidies would need to be set high enough to purchase comprehensive benefits, including prescription drugs, and have modest deductibles, if any. At a minimum, tax credits would need to be adjusted by age and region to reflect market realities. To the extent that states adopt community rating, age rating would be less necessary, but credits would need to be higher for all age groups to make coverage affordable.

Moreover, without market reforms, uninsured midlife adults are unlikely to find a welcome reception in the individual insurance market, because of the combination of poor health and low incomes. Although some states operate high-risk pools for those denied coverage in private markets, premiums for these pools are generally set much higher than those quoted for otherwise healthy adults. Thus, without even more substantial premium assistance, risk-pool rates also are likely to be beyond the reach of the low-income uninsured adults we studied. Even with subsidies, access to risk pools is often difficult. Two state risk pools have long waiting lists; another is closed to new enrollees. High out-of-pocket costs and low limits on benefits also deter many eligible persons from enrolling.²⁶

Our study underscores the importance of risk pooling and broad-based participation for viable insurance markets. The profile of those currently in the individual market and examination of data on market rates highlight the extent to which individual markets remain problematic for those with low incomes or poor health. Our findings confirm those of other recent studies that found that selection problems remain even in states with more comprehensive reforms. ²⁷ Given basic concerns about affordability and risk pooling, tax credit proposals that seek to reach all age groups must be pursued in concert with access and rating reforms that minimize or prohibit discrimination based on individual health status or that provide group-coverage options.²⁸

Editor's Notes

 
At this writing, Elisabeth Simantov was senior research analyst at the Commonwealth Fund in New York City. She is now with the Center for Home Care Policy and Research at the Visiting Nurse Service of New York. Cathy Schoen is vice-president of research and evaluation at Commonwealth; Stephanie Bruegman is a program assistant there.

This paper grew out of analysis of the Commonwealth Fund Survey for the fund’s Task Force on the Future of Health Insurance. The authors thank Karen Davis and two anonymous reviewers for their helpful comments. The views expressed are those of the authors and should not be attributed to the Commonwealth Fund, its directors and officers, or task force members.

NOTES

Top Data Sources And Methods Findings Conclusions And Policy... NOTES

 

1. J. Gruber and L. Levitt, "Tax Subsidies for Health Insurance: Costs and Benefits," Health Affairs (Jan/Feb 2000): 72–85; and P. Fronstin, The Working Uninsured: Campaign 2000 Proposals from the Presidential Candidates, EBRI Issue Brief no. 224 (Washington: Employee Benefit Research Institute, August 2000).
2. J. Budetti et al., Risks for Midlife Americans: Getting Sick, Becoming Disabled, or Losing a Job and Health Coverage (New York: Commonwealth Fund, January 2000); P. Loprest and C. Uccello, Uninsured Older Adults: Implications for Changing Medicare Eligibility (New York: Commonwealth Fund, April 1998); and D.G. Shea, P.F. Short, and M.P. Powell, "Betwixt and Between: Targeting Coverage Reforms to Those Approaching Medicare," Health Affairs (Jan/Feb 2001): 219–230.
3. G.A. Jensen, D.A. Weydker, and J.R. Gabel, Older Adults and Their Health Insurance (Washington: AARP, March 1999).
4. Medicaid restricts eligibility for adults under age sixty-five to those with dependent children or with mental or physical disabilities as well as low incomes. Persons under age sixty-five are eligible for Medicare only if they are disabled or have kidney failure. Disabled persons are eligible after a two-year waiting period if receiving Social Security disability benefits. Committee on Ways and Means, U.S. House of Representatives, Overview of Entitlement Programs: 1994 Green Book (Washington: U.S. Government Printing Office, 1994), 123, 783–788.
5. D.J. Chollet and A.M. Kirk, Understanding Individual Health Insurance Markets (Washington: Henry J. Kaiser Family Foundation, 1998).
6. Guaranteed-issue states include Maine, Massachusetts, New Hampshire, New Jersey, New York, Vermont, Washington, Idaho, Iowa, Maryland, Nevada, North Carolina, Ohio, South Dakota, and Utah. Eight of these only require guarantee of a basic product. S. Laudicina and K. Pardo, State Legislative Health Care and Insurance Issues: 1999 Survey of Plans (Washington: Blue Cross Blue Shield Association, December 1999).
7. HIPAA guarantees those leaving group coverage access to coverage in the individual market. The guarantee applies to those who have at least eighteen months of aggregate credible coverage, were recently in a group plan and without a break of more than sixty-three days, and have exhausted COBRA or conversion coverage. Health Care Financing Administration, "The Health Insurance Portability and Accountability Act of 1996 (HIPAA) Page," no date, <www.hcfa.gov/hipaa/hipaahm.htm> (4 January 2001).
8. Only New York and New Jersey require full community rating. Fifteen other states limit rate variation. In other states insurance carriers may set premiums based on health risks, restrict coverage of preexisting conditions, or (in non-guaranteed- issue states) deny coverage altogether (except for HIPAA). Laudicina and Pardo, State Legislative Health Care and Insurance Issues.
9. The survey screened by age. The response rate calculation is the product of contact, screening, and completion rates. Of the residential numbers in the sample, more than 74 percent were contacted, 77 percent completed screening, and 99 percent of eligible respondents completed the interview, for a final response rate of 57 percent.
10. Among those with public insurance, 78 percent reported Medicare and 21 percent, Medicaid.
11. The two primary sources were eHealthInsurance. com, <www.ehealthinsurance.com>, and Quotesmith.com, <www.quotesmith.com>. In New York and New Jersey we also used two state Web sites: the New York State Insurance Department site, <www.ins.state.ny.us>, and the National Association of Insurance Commissioners’ New Jersey page, <www.naic.org/nj>.
12. We examined other major cities in the sampled states and inmost states found premiums similar to those quoted for the studied cities.
13. New York and New Jersey require guaranteed issue and full community rating. Massachusetts rates can be adjusted for age within a 2:1 ratio. Michigan, Pennsylvania, and Washington, D.C., hold open enrollment periods but allow rates to fluctuate with health risk. The other states in the study sample rely on HIPAA rules.
14. We compared the selected ZIP code rates with other ZIP codes and found that rates were comparable in 85–90 percent of all other ZIP codes in each city. Where different, rates in other ZIP codes were within ±10 percent of our rate range.
15. Health and income differences between adults with individual and public insurance or the uninsured were statistically significant at p .001.
16. Results of the multivariate analyses (logistic regressions) are available from the authors. E-mail a request to <elisabeth.simantov{at}vnsny.org>.
17. Unless otherwise indicated, deductibles apply to in-network benefits. Out-of-network cost sharing was substantially higher. Details for each city are available from the authors.
18. Jon Gabel, Health Research and Educational Trust, personal communication, 8 February 2001. Data based on the Kaiser Family Foundation/ HRET 2000 Survey.
19. J. Gabel et al., "Job-Based Health Insurance in 2000: Premiums Rise Sharply while Coverage Grows," Health Affairs (Sep/Oct 2000): 144–151.
20. N. Tapay and J. Feder, Insurance Market Reforms and the Individual Insurance Marketplace: Implications for Coverage Expansions (Washington: Henry J. Kaiser Family Foundation, October 1999).
21. Unlike HIPAA, which does not place limits on premiums, COBRA allows those leaving jobs with benefits to pay premiums at 102 percent of the group rate.
22. L. Duchon et al., Listening to Workers: Findings from the Commonwealth Fund 1999 National Survey of Workers’ Health Insurance (New York: Commonwealth Fund, January 2000).
23. Jensen et al., Older Adults and Their Health Insurance. Also see K.A. Couch, "Late Life Job Displacement," Gerontologist (February 1998): 7–17.
24. Heritage Foundation, A Guide to Tax Credits for the Uninsured (Washington: Heritage Foundation, 5 May 2000).
25. Authors’ calculation based on the median premium in the fifteen cities for a sixty-year-old.
26. California and Illinois have waiting lists; Florida is closed to new applicants. L. Achman and D. Chollet, "An Overview of State High-Risk Pools: Enrollment, Benefits, and Funding" (Mathematica Draft Report to the Commonwealth Fund, January 2001).
27. See the case study of New Jersey, which finds that people who can afford to purchase individual insurance and do so are, on average, healthier than those who choose not to enroll. K. Swartz and D.W. Garnick, "Can Adverse Selection Be Avoided in a Market for Individual Health Insurance?" Medical Care Research and Review (September 1999): 373–388.
28. K. Swartz, Markets for Individual Health Insurance: Can We Make Them Work with Incentives to Purchase Insurance? (New York: Commonwealth Fund, December 2000).

                      What's this?

This article has been cited by other articles:

				J. M. McWilliams, E. Meara, A. M. Zaslavsky, and J. Z. Ayanian Differences in Control of Cardiovascular Disease and Diabetes by Race, Ethnicity, and Education: U.S. Trends From 1999 to 2006 and Effects of Medicare Coverage Ann Intern Med, April 21, 2009; 150(8): 505 - 515. [Abstract] [Full Text] [PDF]

				C. Schoen, K. Davis, and S. R. Collins Building Blocks For Reform: Achieving Universal Coverage With Private And Public Group Health Insurance Health Aff., May 1, 2008; 27(3): 646 - 657. [Abstract] [Full Text] [PDF]

				J. M. McWilliams, E. Meara, A. M. Zaslavsky, and J. Z. Ayanian Health of Previously Uninsured Adults After Acquiring Medicare Coverage JAMA, December 26, 2007; 298(24): 2886 - 2894. [Abstract] [Full Text] [PDF]

				J. M. McWilliams, E. Meara, A. M. Zaslavsky, and J. Z. Ayanian Use of Health Services by Previously Uninsured Medicare Beneficiaries N. Engl. J. Med., July 12, 2007; 357(2): 143 - 153. [Abstract] [Full Text] [PDF]