Financial vulnerability among Medicare managed care enrollees.

Consumer Assessment of Health Plans Survey (CAHPS) data show that Medicare managed care plans often receive low satisfaction scores from certain vulnerable populations. This article describes findings from a qualitative study with beneficiaries about their Medicare managed care experiences. Focus groups were stratified by participant race/ethnicity and self-described health status. Yet participants did not describe their concerns in terms of their race, ethnicity, or health condition, but rather their access to financial resources. Our findings suggest that researchers consider how socioeconomics creates health care vulnerability for racial and ethnic minorities, females, people with disabilities, and other economically marginalized persons.

Introduction and Overview

Since the inception of Medicare managed care plans, CMS has sought feedback from beneficiaries about their experiences obtaining needed services under a health maintenance organization (HMO) model. In 1998, CMS launched the CAHPS®, a nationwide survey effort to collect information from Medicare beneficiaries enrolled in managed care plans about their experiences with and evaluations of their health plans. Survey results are available on the CMS Web site (www.medicare.gov) so that Medicare-eligible individuals can see how health plans available in their market are rated by their peers.

Survey results have also been used by health services researchers to get an overall sense of how well managed care plans are meeting the needs of the beneficiaries. In 2000, an assessment of CAHPS® scores along variables (Barents Group, 2000), believed to represent vulnerable subgroups revealed the following analysis:

Minority enrollees (Black, Hispanic/Latino) were less satisfied with the process of care and their access to services under Medicare managed care, although they were marginally more satisfied overall with their managed care plans than other enrollees.

Elderly beneficiaries who reported being in fair/poor health, or who reported limited independence, also gave lower ratings to most dimensions of health plan performance than did other beneficiaries.

Beneficiaries who were eligible for Medicare due to disability, rather than age, gave lower ratings of managed care plan performance across all dimensions compared to other Medicare beneficiaries.

This article presents findings from a four-city series of focus groups aimed at learning how beneficiaries' experiences had led to these lower ratings. In the first section, we detail our study methods, including our process for selecting study participants, as well as our analytic approach. We then turn to the results of our analysis, focusing in particular on beneficiaries' descriptions of situations in which they had experienced—or anticipated experiencing—difficulty getting needed medical services. In the discussion section, we talk about the implications of our findings for conceptualizing vulnerability, in particular, recognizing the role of socioeconomics in creating vulnerabilities for racial and ethnic minorities as well as other categories of persons. We conclude with a brief discussion of the limitations of our findings, as well as anticipated next steps in this area of research.

Methods

Settings and Participants

Prior to participant selection, researchers met with expert panels comprised of physicians, community service providers, and activists to discuss health issues (disease prevalence and barriers-to-care) specific to Black, Hispanic, and disabled enrollees, and to review the focus group design. The minority panel members, in particular, were quick to point out that our focus group composition had to be refined, since the research categories denoting vulnerability do not constitute mutually exclusive groups—e.g., an individual can be an ethnic minority and be disabled or in fair/poor health. At the panel members' request, and in order to discern the impact of ethnicity versus health status, we recruited for focus groups with minority enrollees who were in good/excellent health status, minorities who reported being in fair/poor health, and with White enrollees who also reported being in fair/poor health. We also recruited one group in each site of enrollees who were enrolled in Medicare because of a disability. In order to have a large enough sample from which to recruit, we selected four Medicare managed care market areas that had adequate minority representation: Birmingham, Alabama; San Antonio, Texas; Miami, Florida; and New York City. As shown in Table 1, we conducted 20 focus groups in all.

Table 1

Location and Composition of Focus Groups for Study of Financially Vulnerable Medicare Managed Care Enrollees: 2001

Location	Group Composition on Health Status	Facilitator
Birmingham, Alabama	Under Age 65 Disabled (Ethnically Mixed)	White, Experience with Disabled Population
	White Elderly Fair/Poor	White
	Black Elderly Fair/Poor	Black
	Black Elderly Good/Excellent	Black
Miami, Florida	Hispanic Elderly Fair/Poor	Bilingual Hispanic
	Hispanic Elderly Good/Excellent	Bilingual Hispanic
	White Elderly Fair/Poor	White
	Under Age 65 Disabled (Ethnically Mixed)	White, Experience with Disabled Population
San Antonio, Texas	Black Elderly Fair/Poor	Black
	Black Elderly Good/Excellent	Black
	White Elderly Fair/Poor	White
	Under 65 Disabled (Ethnically Mixed)	White, Experience with Disabled Population
	Hispanic Elderly Fair/Poor	Bilingual Hispanic
	Hispanic Elderly Good/Excellent	Bilingual Hispanic
New York City	Black Elderly Fair/Poor	Black
	Black Elderly Good/Excellent	Black
	Hispanic Elderly Fair/Poor	Bilingual Hispanic
	Hispanic Elderly Good/Excellent	Bilingual Hispanic
	White Elderly Fair/Poor	White
	Under Age 65 Disabled (Ethnically Mixed)	White, Experience with Disabled Population

SOURCE: Robins, C.S., Myers, M.A., Westat; and Heller, A., Centers for Medicare & Medicaid Services, 2001.

For the selected plan in each market area, the team created a study recruitment pool using data from the CAHPS® super sample. The super sample is used to disguise the identity of the true CAHPS® sample members from the health plans when researchers request telephone numbers. On average, a super sample of about 3,000 beneficiaries is selected for each participating health plan, with the true sample of 600 beneficiaries embedded therein. For each of the four plans our team obtained names, addresses, and telephone numbers of plan enrollees, and their date of birth and reported ethnicity. All potential participants were mailed a letter from CMS outlining the purpose, benefits, and relative risks of the study, and stating that individuals might receive a telephone call requesting their participation in a voluntary study. Recruiters at each professional focus group facility were sent these lists of beneficiary names, along with a screening form that asked respondents to verify certain key information, such as enrollment in the specified Medicare managed care plan, age, race, and ethnicity. Respondents also were asked to self-assess their health status using a 5-point scale (where 1 = poor and 5 = excellent). The category elderly in poor health was operationalized as those respondents who were over age 65 and who reported a health status of 1 or 2 (poor/fair).

With the exception of three of the groups, all sessions were held in professional focus group facilities, which were able to provide us with recruiting services, video and audio recording capabilities, transportation support, food for the participants, as well as other amenities (e.g., telephones, handicapped accessible restrooms). At three of the study sites, participants received $50 for their willingness to be a part of the study. In New York City, because of the facility director's experience recruiting for focus group participation in the market area, all beneficiaries received $100 for their participation.

Expert panel members had requested that we have culturally competent and consistent facilitation for all of the focus groups. The cultural competence literature is equivocal on the effectiveness of ethnic matching between clinicians and clients (Jerrell, 1998; Knipscheer and Kleber, 2004; Flaskerud, 1990). Marketing research, however, generally supports pairing focus groups and moderators based on the group's characteristics. Thus our study team sought to ensure that those groups in which participants were racial or ethnic minorities were moderated by an individual who was racially or ethnically similar.

We therefore contracted a Black professional facilitator to moderate each of the six groups conducted with Black beneficiaries, and a bilingual/bicultural (Mexican-American) facilitator to moderate all of the Hispanic groups. The authors have experience working directly with differently-abled populations, and thus facilitated each of the four groups with elderly White, as well as the four groups of disabled enrollees.

Although this study was initiated because of concerns about the experiences of vulnerable subgroups of Medicare managed care enrollees, we created a discussion protocol that allowed participants to define for themselves the parameters of their health care experiences. Topics included key concepts from the CAHPS® survey instrument (e.g., access to care, timely referrals), but also offered participants ample room to discuss their own reasons for why they liked, or did not like, their managed care plans, and what they might like to see done differently. The protocol contained no specific questions about plan discrimination out of concern that such questions might promote socially desirable responses; nevertheless, moderators were encouraged to use probes if respondents raised concerns about racial or ethnic bias, or specific health conditions that were not being addressed by the HMOs. All focus groups were conducted in May-June 2001.

Qualitative Analysis and Interpretation

With the participants' permission, all focus groups were both video- and audio-taped for later review by the study team. Audiotapes were transcribed and copies of the transcriptions were compared to the audiotapes to correct any errors or omissions. Each author conducted multiple readings of the transcripts to identify recurrent themes in the participants' discussions. These themes were further developed and reviewed by the authors who then met to discuss the issues that emerged from our readings, resolve questions, and refine the thematic categories.

Results

Across all of our sites, probes and specific questions about race and ethnicity suggested that while participants might have had negative experiences with a particular physician, they did not express the view that their race or ethnicity was related to their Medicare managed care experience. Moreover, health status was seen to be a significant factor in people's Medicare managed care experiences—although not in the way we had envisioned. Specifically, individuals who were in self-described good/excellent health often noted that they had had little reason to interface with their managed care plans. Consequently, they admittedly were rating their managed care plans as good on the basis of very limited experience. High CAHPS® satisfaction scores thus might indicate strong plan performance, but might also reflect that the plan had not been called on much by the respondent.

As importantly, and contrary to our expectations, participants with severe health conditions that did require significant medical intervention and insurer involvement did not consistently rate their plans as unsatisfactory. Indeed, it was the tremendous variation in their experiences and the factors they cited as contributing to their experiences that suggest the need for a more complex conceptualization of beneficiaries' Medicare managed care experiences than we had originally envisioned.

In our analysis of the discussion transcripts, three themes emerged that might further our understanding of factors that impact beneficiaries' experiences and satisfaction with their managed care plans. These themes include: (1) beneficiaries' fiscal considerations when choosing to enroll in a Medicare managed care plan; (2) how social factors play into individual life histories, leaving certain categories of individuals more financially at risk on retirement; and, (3) how community resources may have a moderating effect for at-risk beneficiaries.

Financial Considerations

As a result of the Balanced Budget Act of 1997, individuals who are eligible for Medicare can choose between enrolling in traditional fee-for-service (FFS) or traditional Medicare or select a health plan under the Medicare managed care option. The assumption behind the addition of the managed care option was that by having more choices available to them, persons eligible for Medicare could select the benefit structures that best fit their needs. Under the FFS option, beneficiaries can freely select their physicians, specialists, and medical facilities. They pay no monthly premium for Medicare Part A, but must assume responsibility for 20 percent of the cost of medical services. Medicare Part B requires a monthly premium, and covers only a core set of outpatient services (e.g., there are no outpatient pharmacy benefits). Because of this limited coverage, many beneficiaries opt to supplement their coverage with a Medigap policy that costs anywhere from $50 to more than $500 per month.

By contrast, under Medicare managed care, beneficiaries pay a monthly premium, and exchange the freedom to choose their physician or medical care facility for lower out-of-pocket costs. Managed care plans also usually offer additional benefits, such as some form of prescription drug coverage above the baseline set by FFS Medicare.

While the variations in benefits figured into beneficiaries' narratives about their plan selection processes, focus group participants emphasized that their choice of enrolling in a managed care plan was based on a combination of factors, including self-assessed medical needs, plan benefits, and, most importantly, the relative cost of those benefits compared to potential out-of-pocket medical expenses. Indeed, discussions with this purposively selected sample of Medicare managed care enrollees suggest considerably more financial vulnerability than we anticipated at the outset of the project (Shimada, Zaslavsky, and Cleary, 2003). Several participants noted, for example, that they had enrolled in a managed care plan because they could no longer pay the thousands of dollars per year ($7,000, in one case) in premiums for individual health insurance plans:

“I signed up for Blue Cross®/Blue Shield® (BC/BS) individual plan. I carried that [plan] and year after year, they were just upping their rates and upping their rates and upping their rates. And I guess…by the time I was sixty-two, the plan was costing me… $378 a quarter, which was an average of $126 a month; plus what they were taking out of my disability check. So I was paying like $178 a month for health care insurance. I had no coverage for medications through BC/BS, so my medicine I had to buy straight out. In beauty work 20 years ago, it wasn't as good as it is…beautician's work wasn't paying as much as they're paying [now]. I am just now reaching the $500 mark for monthly income.”

Echoed a beneficiary from Birmingham:

“…I retired and my insurance ran out and I needed some kind of subsidizing—[plan] was the most economical way to go.”

Other beneficiaries had enrolled in traditional Medicare when they retired, but the steady increase in supplemental premiums became prohibitive under their fixed incomes:

“Well, I had Medicare and I had a supplement plan to go along with my Medicare. Well, that insurance company that I had the plan with, they kept raising the premium and I just got to the place where I could no longer afford [it]—and even my insurance agent told me, she said, ‘You might as well drop this plan because they're going up again on it.’ …I told her, that's what I'm going to have to do because I cannot afford to pay this much for supplement insurance.”

The 80/20 cost sharing of traditional Medicare was also of concern to many beneficiaries, who recognized that even one extended hospital stay could deplete a substantial portion of their retirement savings. One gentleman in New York City reported with great relief how he was hospitalized with diverticulitis shortly after he had enrolled in a Medicare managed care plan:

“I had to have two operations and three hospital stays, which was $49,000 billed. But it's a very good [managed care] plan, because all I had to pay was $200 out of the whole business per hospital stay.”

Indeed, the ability to calculate out-of-pocket expenses was described by many focus group participants as a key reason for enrolling in a managed care plan. The premiums were affordable, and the copayments for services were also known:

“When I go to the hospital, it's $25. If I just go for an emergency, and if I just go into the doctors it's $5. And I don't have to pay nothing else if I go in the hospital, but the $25.”

An additional reason beneficiaries described switching to managed care was because traditional Medicare does not offer a prescription drug benefit. In the absence of a benefit, Medicare beneficiaries had to pay for expensive medications on a fixed income:

“I have spent so much on medicine. And I go to Eckerd's Pharmacy—and it's a good thing I did … I have to charge it on my J.C. Penney's card because J.C. Penney bought out the Eckerd's Pharmacy or something like that or they combined. But that's the only way I can get my medicine. When I got out of the hospital last year, I had to get one medicine. That one medicine alone was $140 and … I had to pay it myself. And, you know, this medicine is so outrageously expensive you can't continue to pay for medicine like that on a limited income, because you have other things to pay for.”

Thus, managed care plans that offered a prescription drug benefit—in some cases, up to $2,500 per year—held a particular appeal for beneficiaries, who could reasonably figure their out-of-pocket costs for not just medical services, but also medications.

The challenge to beneficiaries' financial calculus, however, is that the Medicare managed care market is a moving target. When the program was initiated, premiums were low and additional benefits were quite generous as plans sought to gain market share. However, by 2001 the plans were feeling the impact of their initial marketing strategy, and were subsequently beginning to shift costs back onto the consumers. Because of the beneficiaries' limited incomes (Shimada, Zaslavsky, and Cleary, 2003), they, too, were beginning to reconsider whether the product they were purchasing was still the most economical option for them:

“And at that time [when I was signing up for Medicare], all the other insurance companies were beginning to charge a premium of $10 a month and [plan] was not. So I joined and for the first year, terrific! And then the following year, they went from $1,000 a year, brand prescription name-they cut it to $500. Then the next year, they weren't paying for any brand prescription names. And the following year, you have a co-payment now if you go the hospital. For an inpatient or outpatient, you can have a copayment now. So now I'm looking for another plan.

Social History and Vulnerability

Our analysis of the transcripts suggests that concern about finances was a deciding factor for many of our focus group participants. Indeed, concerns about fixed incomes and out-of-pocket costs were raised most readily in our focus groups with Black and disabled enrollees. Female participants in all of our groups also discussed finances to a large degree, and money emerged as an important consideration in many of our Hispanic groups. As a result of these findings, we tentatively argue that race/ethnicity and health status are reasonable proxies for an individual's financial well-being. In assuming this position, we acknowledge its contentiousness—in the last thirty years, the trends in public health have been to focus on individual rather than social factors as causes of disease, health disparities, and the like. Certainly the cultural ideal is that all Americans—regardless of race, color, or class—are individuals, free to choose their occupations in order to maximize their income and wealth. Our perspective, however, is closely aligned with a group of public health scholars and medical anthropologists who are exploring how social factors may serve as fundamental causes of disease and disparity (Link and Phelan, 1992; Farmer, 1992.) This perspective assumes that individuals' life choices are strongly constrained by larger sociocultural processes: beliefs about females in the workplace; tacit assumptions about occupation and social class; clearly stated values about appropriate jobs for females, Black and Hispanic people; and often-unacknowledged wage disparities associated with those jobs. As scholars have noted, the American class system is in many respects a caste system (Dollard, 1989), where the color of one's skin or ethnic heritage dictate the economic life a person can lead. Current population survey (CPS) data, for example, indicate that the per capita income in 2001 for White individuals was $24,127; Black individuals made approximately 62 percent of this total, with a per capita income of $14, 953; and Hispanic individuals earned even lower, with per capita earnings at a mere $13,003, or 54 percent of their White individuals earnings (http://www.census.gov/hhes/ income/histinc/incperdet.html).

In addition, both White and minority females fall disproportionately into a position of financial vulnerability. As our study participants noted, females traditionally did not work in vocations that paid much money, and thus were receiving very low social security checks. And one female in New York City reminded the group, “Remember not all females worked until recently.” A report by theU.S. General Accounting Office (2003) reinforced this perspective, noting that “women have fewer years of work experience, work fewer hours per year, are less likely to work a full-time schedule, and leave the labor force for longer periods of time than men.” The report goes on to note that even when these factors are taken into account statistically, females still only earn about 80 percent of what males earn. Others have noted that the situation is even more distressing for minority females, wherein Black females earn approximately 63 percent of male wages, and Hispanic females earn only 53 percent. Hardly surprising, then, that a Hispanic female in Texas reported the following:

“I could not afford the brand name medication I was getting for cholesterol and osteoporosis. [It] used to be $15 copayment, and now it jumped up to $30 each, that's sixty bucks, and I only take home $381 from social security. So needless to say since January I haven't taken any medication because I can't afford it.”

Moreover, given the tacit-though-complex issue of social class in this country, even White males are not immune to fiscal distress, as indicated in this simple statement by a White retired laborer in New York City: “I lived on what I made.”

The result of these social factors is a lifetime of economic disparity that ultimately carries over into individuals' retired lives. Research by Crystal, Shea, and Krishnaswami (1992) has shown that people who made less during their work lives have fewer resources on retirement, leaving these individuals vulnerable to a variety of medical contingencies associated with aging. And a study conducted for the Kaiser Family Foundation (Moon, 2002) indicated the following:

“Forty percent of all [Medicare] beneficiaries have less than $12,000 in countable assets, with even higher rates reported by women (45 percent), Black (75 percent), beneficiaries in poor health (52 percent), and non-elderly beneficiaries with disabilities (74 percent).”

The cultural ideal of individualism also ignores class-based occupational hazards, and that a person's work history can impact his/her long-term health: white collar workers are at a lower risk of work-related injury than are their blue collar counterparts (Crystal, Shea, and Krishnaswami, 1992). A Hispanic male participant in the disabled group in San Antonio, Texas offered the following:

“In 1986, I met an injury at work. I used to work for a manufacturing firm and I was lifting something stainless steel and hurt my back. Ever since then I've had three back surgeries.”

While we heard distressing stories in every focus group, it was consistently the disabled beneficiaries who seemed to be experiencing the greatest medical and financial distress. In Miami, Florida, for example, our focus group participants described how even slight changes in the plan benefits structure, such as an increased copayment, were causing them to simply do without. One woman offered the following:

“I have two doctors. And both of them were on the plan but I was going there without paying any copayment—they're specialists, mental health specialists-and all of a sudden, they got this mental health network and they went off it. And now I have to pay somebody like $20. I can't afford it because I'm on a very limited income. I mean it's very difficult for me … I can't even afford $20 out of my pocket. Now I've got to pay $10 for prescriptions, and it's like squeezing blood out of a turnip because I'm on a very, very set income.”

Some of these discussants were only in their thirties, yet their work lives—and thus potential for saving money for the future—were already over. Moreover, and unlike some of their senior counterparts, these individuals already were experiencing significant and costly medical conditions. Without a lifetime of savings behind them, and facing the rest of their lives managing devastating medical conditions, these participants were likely to spend the remainder of their days hanging on a financial thread.

Community Resources As Moderators

Despite their overwhelming financial concerns, beneficiaries also described a resourcefulness that merits exploration. Specifically, although many focus group participants described a zero-sum personal financial situation, wherein out-of-pocket medical expenditures meant foregoing some other necessity, they also described a broad array of strategies for obtaining services not covered by their plans. For example, there were many beneficiaries who were strong advocates of requesting medication samples from physicians. Said one, “I believe those doctors are given those samples to give their patients, but if you don't ask for them you won't get them. But you've got to ask for them.” Some participants reported being even more directive with their physicians, “I tell you what I would do. I would go back to the doctor and ask if he could substitute it [for a cheaper medication or generic].” Still others opted to bypass their primary care provider altogether, “I have a suggestion to anybody that's a Veteran, go to the VA, and you'll get your prescription for two dollars… All you have to do is take one physical a year.”

Indeed, more than a few focus group participants reported accessing services through an alternative, non-Medicare provider, such as the Veterans' Administration hospital or their outpatient clinic. A second, frequently-cited alternative source of care was what participants referred to as the local charity hospital, i.e., a publicly-funded facility where a person's insurance status did not matter. Public hospitals or free clinics did not exist in every community, but where they were available, many seniors indicated that they took advantage of that resource.

Finally, several individuals reported receiving medical assistance—in particular, prescription support—through a State-funded program or even through Medicaid. Some States, such as New Jersey, have several well-funded prescription drug programs for seniors. Many also set qualifying levels well above (e.g., 200 percent) the Federal poverty level, thus allowing more people access to these supports and services (Safran et al., 2002).

Discussion

Our original study task was to hear from vulnerable Medicare managed care enrollees about their experiences with their managed care plans. Although the study team had some working ideas about what variables (race/ethnicity, health status) might affect beneficiaries' experiences, qualitative research allows us to determine if there are other, unanticipated factors affecting the phenomenon under study. By using a semi-structured discussion protocol that allowed focus group participants to define for themselves key determinants of their managed care experiences, we heard that a heretofore unanalyzed variable—financial well—being-was affecting many of our participants' access to health services and experiences with their health plans. As we suggested earlier, because of sociohistorical reasons, financial status may be directly correlated with both race/ethnicity and an individual's health status. It is also important to recognize, however, that there are other categories of Medicare beneficiaries who are also potentially at risk because of social processes. These categories include both White and minority females and individuals who qualify for Medicare because of a disability. Interestingly, when the CAHPS® data are analyzed along these two variables, these categories of respondents consistently give their Medicare managed care plans lower satisfaction scores (Barents Group, 2000).

Although financial vulnerability appears to be an important aspect of many of our participants' experiences, even the state of vulnerability is not a constant: shallow pockets, a lack of savings, and an unreliable Medicare managed care plan do not necessarily translate into unmet medical needs. Individuals who have the cognitive capacity or who have someone (family, friend, case worker) who can advocate on their behalf can often locate alternative sources of care and support in the community. Thus, community resources can serve as important moderators for populations that are otherwise at risk.

Conclusions

Our findings offer a differing perspective from other researchers who have discovered how racial and ethnic bias negatively affects the health status of American minorities (Virnig et al. 2002, Baicker et al. 2004, Schneider et al. 2002). Minorities in this country experience high rates of discrimination in the interpersonal setting of the clinic (Schnittker, 2004; Blendon et al., 1995, and LaVeist, Nickerson, and Bowien, 2000). What we explored, however, was not the face-to-face interaction of physician and client, but rather consumers' satisfaction with the insurance products they had purchased. In this marketplace, individuals were concerned about the value and quality of their purchase, since without deep pockets or alternative local resources, study participants were likely to have to choose between paying for medical services or basic needs (e.g., food, clothing, shelter). A gentleman in our disabled beneficiary group in New York City chose to see his own glass as half-full:

“I'm fortunate, I have my daughter-in-law [and] she does a lot for me. But what happens if you don't, if you're not knowledgeable? …What do you do? The heck with it. That's the easiest thing to do, that's what it becomes, [you] just say…To heck with it.”

There are at least two issues that emerge from our findings that suggest the need for additional research. First, if financial well-being is a critical aspect of these enrollees' experiences, researchers need to examine the CAHPS® scores in relation to respondents' socioeconomic status (SES). Unfortunately, the CAHPS® data set does not contain variables that can adequately reflect SES, such as individuals' incomes, assets, or occupations during their lifetimes. A followup quantitative study is in progress combining SES variables with CAHPS® data in order to better examine the relationship between respondent SES and CAHPS® scores. Indeed, there is only one variable in the current data set that could be used as a proxy variable for SES, and that is the educational attainment of the respondent. Interestingly, and at odds with our findings here, analysis of the CAHPS® data indicate that beneficiaries with an eighth-grade education or less reported higher levels of overall satisfaction with their Medicare managed care plans than other beneficiaries (Barents Group, 2000). This is an interesting result that merits further exploration: What is it about these individuals' experiences and expectations that might lead to a higher satisfaction rating? Did they not have insurance during their work years, thus making Medicare the first (and thus incomparable) form of health insurance they have ever had? In short, are they working off the expectation that any insurance is better than nothing at all? (Almeida, Dubay, and Ko, 2001.)

Additionally, we have pointed out that access to community resources appears to be an important mediator for otherwise financially vulnerable individuals. What is it about individuals' historical and current life circumstances that makes them more or less satisfied with their Medicare managed care than other beneficiaries? We believe that we might better understand people's satisfaction with their health plans if we learn more about the community environment in which they are trying to obtain services (Sanjek, 2000). CMS is involved in a new qualitative study in which rapid ethnographic assessments were conducted in four locations where Medicare managed care plans received markedly different satisfaction scores by beneficiaries. Initial findings from this study suggest that there are indeed important resource differences in these communities, although it is not yet clear how beneficiaries' expectations and abilities to negotiate these resources may also play into their managed care experience. We believe that further research into community resources-social capital—might help us to better understand the complex interplay between race, ethnicity, health status, and social class as Medicare managed care enrollees rate their plan experiences.