Strategies for Medicare health plans serving racial and ethnic minorities.

The Medicare Managed Care (MMC) Consumer Assessment of Health Plans Study (CAHPS) survey offers an opportunity to examine differences in health plan experiences and patterns of use of services of racial and ethnic minority beneficiaries enrolled in health plans. Analysis of the survey data and review of prior literature indicate significant health disparities and different patterns of health care use by racial and ethnic minorities. Improved measurement of health plan performance in serving minority enrollees, and development of performance improvement strategies, could have the potential to reduce the observed health disparities.

Background

The MMC CAHPS® survey was initiated by CMS in 1997 to collect information from Medicare health plan enrollees on their experiences and assessment of the performance of these plans in meeting their needs. The survey is conducted annually and results are reported and available to beneficiaries, along with information on benefits and premiums and other quality measures. The MMC CAHPS® includes measures of overall health plan performance and measures of various dimensions of performance, including quality of service, provider relationships, and access to care. A related CMS survey collects information on reasons why Medicare beneficiaries disenroll from their MMC plan and makes these results available, along with specific health plan disenrollment rates, to further assist beneficiaries and their advisors to make decisions on health plan enrollment.

While these initiatives provide useful information overall, they do not address the performance of health plans in serving populations with exceptional needs for health care or those who, because of race, ethnicity, and cultural differences may require special efforts from health plans to assist them to effectively access and use plan services. One component of the MMC CAHPS® project has focused on examining the performance of MMC plans in meeting the needs of these population groups that may require additional attention and efforts from health plans. Results of these special studies suggest that, while health plans are successful in meeting the needs of the typical enrollee, those with exceptional health care needs and enrollees who are members of certain racial/ethnic population groups may encounter greater difficulties in obtaining and using services from their health plans.

This article examines the need for and identifies strategies that could be used by CMS to measure performance of MMC plans in serving racial/ethnic minority members and develop approaches to improving quality of care for these populations. The approach that is taken for this study relies on analysis of differences in experiences reported by racial/ethnic subgroup populations enrolled in MMC plans who participated in the MMC CAHPS® to identify potential areas of concern in plan service to these populations. These identified issues then are used to develop potential additional performance measures that could be used by CMS to assess health plan service to enrollees who are members of these population groups. Several strategies for encouraging MMC plans to improve quality of care and service for their racial/ethnic minority members are suggested.

Data

The MMC CAHPS® data are used to analyze current performance of MMC plans in serving enrollees who are members of racial/ethnic minority groups. (A description of the MMC CAHPS® is available on request from the authors.) The MMC CAHPS® annual survey sample is sufficiently large to permit analysis of the experiences of black and Hispanic/Latino persons enrolled in health plans, relative to the experiences of the majority population. By combining 3 years of MMC CAHPS® data, excluding multiple responses from sample members included in more than 1 year, it is possible to also examine the experiences of Asians, Native Americans, and Native Hawaiians/Pacific Islanders who are members of MMC plans.

Respondents to the CAHPS® surveys were asked to identify their race and were asked whether they were of Hispanic or Latino origin or descent. Responses were used to create the six racial/ethnic subgroups described in Table 1, which provide sample sizes for each of the racial/ethnic groups of interest in the combined MMC CAHPS® data sets for 1997, 1998, and 1999. A classification issue arises because respondents are allowed to check more than one box for race. Approximately 98 percent of survey respondents checked only one box. However, 1 percent of respondents checked two boxes and the remaining 1 percent checked three or more boxes. For this analysis, only those who indicated a single race were included in the sample. Those respondents who did not choose any race are also excluded. Clearly, a high majority of MMC enrollees are white and the relative size of each group roughly approximates what it is in the total U.S. population. The proportion of MMC CAHPS® respondents that is white is slightly greater than the proportion that is white in the total U.S. population age 65 or over. For the minority racial/ethnic groups, the opposite is true. That is, white persons are slightly overrepresented—and minorities are slightly underrepresented—in the MMC population compared with the total aged U.S. population.

Table 1

Number and Percent of Unduplicated, Self-Indicated Single-Race and Hispanic/Latino Medicare Managed Care CAHPS® Respondents, 1997-1999

Subgroup	Sample Size	Percent
White1	244,030	86.5
Black1	16,097	5.7
Asian1	5,139	1.8
Native Hawaiian/Pacific Islander1	3,951	1.4
Native American1	865	0.3
Hispanic Latino2	12,098	4.3

Non-Hispanic/Latino.

Hispanic/Latino may be of any race—number shown is for single-race persons only.

NOTES: CAHPS® is Consumer Assessment of Health Plans Study. Numbers exclude persons of unknown Hispanic/Latino ethnicity.

SOURCE: Centers for Medicare & Medicaid Services: Data from the Medicare Managed Care CAHPS® surveys for the years 1997, 1998, and 1999.

By combining 3 years of data, the sample sizes for Asians, Native Hawaiians/Pacific Islanders and Native Americansare made large enough to permit analyses with a satisfactory degree of statistical significance, even when the groups are disaggregated by sex and other demographic characteristics.

Table 2 presents the age, sex, and geographic distribution of respondents by race and ethnicity. Relatively few of the Medicare enrollees are under age 65. The Native American group is the youngest, overall, whereas the Native Hawaiian/Pacific Islander group is the oldest. For most groups, females outnumber males; however, there are more males than females among Native American MMC CAHPS® respondents. More survey respondents are located in the West than any other region, whereas the fewest are in the Midwest. However, there are substantial differences across the racial/ethnic groups. More than two-thirds of Asians reside in the West, as do in excess of 40 percent of Native Hawaiians/Pacific Islanders and Hispanics. Only 11 percent of black survey respondents are in the West, however. Nearly one-half of black persons are in the South. The South also is the location of more than 40 percent of Hispanic persons.

Table 2

Age, Sex, and Geographic Distribution of Medicare Managed Care CAHPS® Respondents, 1997-1999

Category	White	Black	Asian	Native Hawaiian/Pacific Islander	Native American	Hispanic/Latino
	Percent
Age
Under 65 Years	5.6	*12.9	*4.3	*6.7	*15.0	*8.8
65-69 Years	28.8	*30.3	*32.3	*24.1	30.2	*31.3
70-74 Years	27.4	26.8	*29.1	27.4	26.3	*28.9
75-79 Years	20.5	*16.7	20.0	21.0	*16.5	*17.5
80 Years or Over	17.8	*13.3	*14.4	*20.8	*12.0	*13.6
Sex
Male	44.5	*39.6	43.4	*35.9	*53.1	*46.8
Female	55.5	*60.4	56.6	*64.1	*46.9	*53.2
Region
Northeast	22.9	*19.6	*8.9	21.6	*11.6	*11.0
South	27.0	*47.5	*16.5	26.8	*35.9	*41.3
Midwest	18.1	*21.6	*6.7	*11.5	14.5	*5.5
West	32.0	*11.3	*67.9	*40.1	*38.1	*42.2

p < 0.05 for difference with white group.

NOTES: CAHPS® is Consumer Assessment of Health Plans Study. Numbers exclude persons of unknown Hispanic/Latino ethnicity. Percentages may not add to 100 because of rounding.

SOURCE: Centers for Medicare & Medicaid Services: Data from the Medicare Managed Care CAHPS® surveys for the years 1997, 1998, and 1999.

Previous Research

A substantial body of literature has documented racial and ethnic differences in health care access and utilization, as well as disparities in specific treatments (Gaskin and Hoffman, 2000; Monheit and Vistnes, 2000; Waidmann and Rajan, 2000; Weinick, Zuvekas, and Cohen, 2000; Eggers and Greenberg, 2000). Most of the literature has focused on black and Hispanic/Latino persons in comparison with the white majority, in part because other racial/ethnic minority groups are a small proportion of the population and, thus, sufficient data are not available to permit examination of their health care patterns.

Financial barriers to access have been found to be a major obstacle affecting minorities' access to and use of health services (BNA Health Care Policy Report, 1999). For minority Medicare beneficiaries enrolled in MMC plans, however, financial barriers are substantially reduced since most health plans offer supplemental Medicare benefits with only modest cost-sharing requirements. For MMC enrollees, differences in accessibility and use of services may be primarily due to other non-financial barriers to access.

Some studies have suggested that black and Hispanic/Latino persons are more likely than white persons to lack the experience and skills that are necessary to obtain and effectively use health services (Robert Wood Johnson Foundation, 1998). Because members of minority groups are less likely to have been insured prior to becoming eligible for Medicare, they may have had irregular and infrequent interactions with the health care system and little experience with managed care (Kaiser Commission on Medicaid and the Uninsured, 2000). These patterns of irregular use would not provide them with the skills that are necessary to overcome managed care barriers to access to health services. As one example, a recent study found that managed care denials of emergency room (ER) claims disproportionately affect black enrollees of health plans, who were more likely to visit the ER and more likely to be denied payment for the ER visit (Lowe et al., 2001).

Other non-financial barriers to access and effective use of services include less availability of providers in minority areas, which may require greater travel times or provide more limited choices of providers in minority neighborhoods (Watson, 1997). A number of studies have also found that patient-provider relationships that are race-concordant produce higher levels of patient satisfaction, more patient involvement in care, better outcomes, and that patients were less likely to perceive discrimination when the provider was of the same race or ethnicity (Somnath et al., 2000; Gray and Stoddard, 1997; Cooper-Patrick et al., 1999; LaVeist, Diala, and Jarett, 2000; Horner, Oddone, and Matchar, 1995). These outcomes related to race concordance may reflect the fact that physicians appear to make treatment decisions that are influenced by the race, ethnicity, and sex of patients (Council on Ethical and Judicial Affairs of the American Medical Association, 1990; Schulman et al., 1999; Cooper-Patrick et al., 1999; Sheifer, Escarce, and Schulman, 2000).

In addition, difficulty in communication with providers can pose a significant barrier to access and effective use of services for Hispanic/Latino persons and other minority groups for whom English is not their primary language. The inability of a health care provider to solicit patient impressions about illness due to language difficulties makes effective care more difficult and frustrates the non-English speaking patient. Hispanic/Latino patients report that providers fail to listen to them, fail to consistently ask about other treatment regimens being used, and leave the encounter without being assured that appropriate treatment and compliance will occur (Barents Group, 1999a, b; 2001c).

These research findings suggest that racial/ethnic minority members of health plans may have greater difficulty obtaining access to needed services and/or effectively using those services, due to lack of knowledge and skills to deal effectively with managed care systems, prior patterns, and expectations of low use and reliance on ER services, cultural differences and biases of providers, and limited facility with English.

Minority Enrollees in MMC Plans

Differences in Health Status

Table 3 presents information on health status, change in health status, and the presence of a serious health condition, of MMC CAHPS® respondents by racial/ethnic group. Black and Hispanic/Latino persons and Native Americans are less likely to report that their health is excellent or very good and are more likely to report their health as fair or poor than are white, Asian, or Native Hawaiians/Pacific Islanders.

Table 3

Health Status and Conditions for Medicare Managed Care CAHPS® Enrollees, by Racial/Ethnic Group, 1997-1999

Question	White	Black	Asian	Native Hawaiian/Pacific Islander	Native American	Hispanic/Latino
	Percent
Health Status Now
Excellent or Very Good	34.4	*25.1	34.8	*38.5	31.7	*30.8
Good	38.5	*35.2	39.6	*33.6	*30.3	*34.8
Fair or Poor	27.1	*39.8	*25.6	27.9	*38.0	*34.4
Health Status 1 Year Ago
Improved	18.8	*31.3	*23.0	*37.0	*26.4	*28.6
Same	64.5	*53.2	63.5	*49.5	*52.1	*55.7
Worse	16.7	*15.4	*13.5	13.5	*21.5	*15.7
Health Condition
Heart Disease	25.0	*21.4	*16.9	*22.4	27.0	*22.3
Cancer	15.8	*10.7	*8.5	*13.3	*12.2	*9.9
Stroke	8.1	*10.7	7.8	*9.8	*12.9	8.6
COPD	5.8	*4.0	*2.9	*4.8	*8.4	*3.9
Diabetes	14.1	*27.0	*18.2	15.0	*25.1	*22.8

p < 0.05 for difference with white group.

NOTES: Numbers exclude persons of unknown Hispanic/Latino ethnicity. Percentages may not add to 100 percent because of rounding. CAHPS® is Consumer Assessment of Health Plans Study. COPD is chronic obstructive pulmonary disease.

SOURCE: Centers for Medicare & Medicaid Services: Data from the Medicare Managed Care CAHPS® surveys for the years 1997, 1998, and 1999.

When specific self-reported health conditions are examined, Native Americans more frequently report having been told by a doctor that they have heart disease, stroke, and chronic obstructive pulmonary disease than any other racial/ethnic group. Compared with white persons, all other racial/ethnic groups (except Native Hawaiians/Pacific Islanders) are more likely to report having diabetes, with black persons most likely to report diabetes. White persons are more likely to report having cancer than any of the other groups and more likely to report heart disease than any other group except Native Americans. Overall, Asians are least likely to report these serious health conditions.

A different pattern emerges when self-reported changes in health status from the previous year are examined. White persons are least likely to report that their health improved since the previous year, and are more likely than any other group—with the exception of Native Americans—to report that their health had worsened since the previous year. Since changes in health status may reflect health plan services and outcomes, the fact that all racial/ethnic minorities report higher rates of improvement in health status and (with the exception of Native Americans) lower rates of worsened health than white persons may indicate that MMC plans are providing good service to racial/ethnic minorities, in general. Nevertheless, inferences about changes in health status based on self-reports should be made with caution. There is also the possibility that, on average, minority Medicare health plan enrollees may have faced significant financial barriers to access prior to becoming Medicare-eligible and before joining the health plan. If so, then the self-reported improvements in health status could reflect the greater potential for improvement once financial barriers to access are reduced.

Conclusions about racial/ethnic differences in health status and need for services are difficult to draw based on these data. Native Americans are the only group that appears to have consistently greater needs for health care, citing poorer health status, worse health than in the previous year, and more serious health conditions than any other group. Asians, on the other hand, appear to be in overall better health than other groups, being least likely to report fair or poor health, least likely to report that their health had worsened, and reporting the lowest proportions of serious health conditions. Relative to white enrollees, however, racial/ethnic minority enrollees in MMC plans report substantially higher rates of diabetes, suggesting that health plans need to have strong diabetes identification and management programs to meet the needs of these population groups.

Health Plan Ratings

As part of the ongoing MMC CAHPS® project, analyses have been conducted of differences by race and ethnicity in experiences with MMC plans (Barents Group, 2000a, b; 2001 a, b, c, d). An overview of the findings of these analyses indicate that:

Overall, black and Hispanic/Latino persons rate their health plan as high or higher than do white persons.

On process of care measures, minority enrollees are slightly more likely (though the differences are not significant) to indicate less positive experiences and greater difficulties than are white persons. Asians, Native Hawaiians/Pacific Islanders, and Native Americans consistently rate their plans lower on all process of care measures than do white persons, although the differences are small in magnitude.

On access to care measures, minority enrollees consistently indicate greater difficulty in obtaining access to specialist physician care, therapy, and equipment than do white persons, although the magnitude of the differences is small and for most comparisons is not significant. Native American enrollees reported the greatest difficulties, among all groups, in obtaining referrals to specialists and home health care and in getting information from customer service.

Utilization of Services

The percentage of MMC enrollees in each racial/ethnic group that reported using various types of health care services is shown in Table 4. The racial/ethnic minority subgroups have very different patterns of usage, both in comparison with the white majority and across subgroups.

Table 4

Health Care Utilization for Medicare Managed Care CAHPS® Enrollees, by Racial/Ethnic Group: 1997-1999

Survey Question	White	Black	Asian	Native Hawaiian/Pacific Islander	Native American	Hispanic/Latino
	Percent
Health Care Use in Last 6 Months
No Visit to Doctor's Office	20.8	*23.3	*29.8	*30.1	*26.3	*25.4
5 or More Visits to Doctor's Office	14.7	15.8	*10.1	*12.3	17.4	14.8
Any Visit to a Specialist	56.0	*49.7	*45.3	*44.2	*48.3	*53.4
Any Hospital Inpatient Use	18.2	*19.7	*12.9	19.0	*24.9	18.1
Any Emergency Room Use	13.1	*18.1	*11.8	*18.6	*19.5	*14.0
Any Prescription Medicine Use	82.2	*77.7	*70.0	*76.3	*73.1	*77.3
Any Special Medical Equipment Use	11.0	*14.7	*7.8	*15.5	*17.7	*10.3
Any Special Therapy Use	9.9	*10.5	*7.6	10.3	11.8	10.1
Any Home Health Care Use	4.9	*8.5	*3.1	*7.7	*8.6	5.1

p < 0.05 for difference with white group.

NOTES: Numbers exclude persons of unknown Hispanic/Latino ethnicity. CAHPS® is Consumer Assessment of Health Plans Study.

SOURCE: Centers for Medicare & Medicaid Services: Data from the Medicare Managed Care CAHPS® surveys for the years 1997, 1998, and 1999.

All other racial/ethnic groups are more likely than white persons to report “No visit to doctor's office.”

Most non-white racial/ethnic groups are more frequent users of ERs compared with white persons; Asians, however, use ERs less frequently than white persons do.

White MMC enrollees are most likely to make doctor and specialist visits and to use prescription medicine.

Black persons tend to be above-average utilizers of most services, although they make fewer doctors and specialist visits than white persons do.

Asians are the lowest utilizers of all listed health care services.

Native Americans have high rates of use of hospitals, ERs, special medical equipment, special therapy, and home health care.

Native Hawaiians/Pacific Islanders don't see doctors or specialists as frequently as other subgroups, but have higher-than-average use of other health care services.

Hispanic/Latino persons are a little less likely to use most health care services than are white persons and most other racial/ethnic minority groups.

Research reported elsewhere found that, when patterns of use are examined separately for those enrollees who report having a heart condition or diabetes, the findings are very similar to the overall patterns of use by racial/ethnic group. Minority enrollees with these two conditions are less likely to have visited their primary care provider and are less likely to have had a specialist visit than are white enrollees with heart disease or with diabetes. ER use and hospitalization is higher for minority enrollees with heart disease or with diabetes than for white persons with these conditions. (Barents, 2001a).

The primary care provider is the entry point to all other health services in most health plans. The fact that minority enrollees are more likely to be hospitalized and to have ER visits may, in part, be due to their lower frequency of contact with their primary care provider and with specialist physicians. Whatever the reasons, these patterns of use suggest that most MMC plans may not be recognizing and responding to the unique needs of their enrollees who are members of racial and ethnic minority groups. To improve performance of health plans in serving these population groups, it is first necessary to be able to measure performance on specific dimensions of service to identify aspects of health plan organization and management that could be changed to achieve better outcomes.

MMC Plans to Meet Needs

Issues for Measuring Performance

While it would be desirable to measure performance of MMC plans in serving racial/ethnic minorities, designing strategies for performance measurement is not a simple task. There are a number of issues that make this task difficult at the health plan level, including:

Small numbers of minority members are enrolled in individual health plans and, as a result, a survey sample is unlikely to include more than a few members of any one minority group at the individual health plan level.

Incomplete data on racial/ethnic identity of individual Medicare beneficiaries on the Medicare enrollment database (EDB), making it difficult to ensure that broad measurement strategies would capture the universe of specific racial/ethnic group enrollees in health plans, and limited encounter data reported by health plans, which would restrict broad measurement strategies to examining hospitalization rates.

Costs of a mandated plan-specific performance measurement strategy would likely be high and—since most MMC plans have only small numbers of racial/ethnic minority group members—would likely produce limited information.

The performance measurement strategies discussed in this section were developed to address these issues and to provide feasible options that could, if adopted, produce useful information on individual plan performance in serving minority enrollees.

Strategies for Measuring Performance

There are a number of feasible strategies that would produce useful information on performance of MMC plans in serving minority enrollees. The relatively small number of enrollees of each racial/ethnic minority group in individual health plans, however, suggests that the most feasible approaches are: (1) examining performance of individual health plans in serving all minority members combined, rather than individual racial/ethnic groups; (2) measuring aggregate performance of MMC plans in serving minority members, in more detail, in order to identify common areas of concern raised by racial/ethnic minority enrollees in MMC plans; and (3) comparing performance of MMC plans in serving minorities with performance of the original Medicare Program in serving minorities. Each of these three strategies relies primarily on Medicare CAHPS® data, with some modifications.

Strategies Using Medicare CAHPS® Data

Four potential strategies for measuring performance of MMC plans in serving minority enrollees rely on Medicare CAHPS® data.

Develop individual health plan measures of performance, using MMC CAHPS® data, stratified by white non-Hispanic and other racial/ethnic groups. Racial/ethnic minority enrollees constitute about 13 percent, combined, of all MMC enrollees. Fifty percent of plans have fewer than 10 percent minority members, while 50 percent have more than 10 percent minority enrollment. This strategy is based on an assumption that there are common issues for racial/ethnic minority groups and that health plan organization and management approaches affect all racial/ethnic minority groups in a similar manner. If all respondents who were members of any minority group are combined, there may be sufficient sample size from the MMC CAHPS® to permit comparison of ratings of the plan by minority members and white non-Hispanics. On average, 13 percent of MMC health plan enrollment is non-white and non-Hispanic. Annual MMC CAHPS® sample size per plan is 600, with an average 80 percent response rate. The average plan, then, would have 62 respondents who are members of a minority group. The issue of the appropriate sample size to produce reliable estimates would need to be addressed by a statistician; however, it is likely that a substantial majority of MMC plans would have sufficient sample size to produce estimates of ratings of specific performance measures for minority enrollees. These ratings could then be compared with ratings given by white non-Hispanic persons to determine the extent to which the plan is serving its minority members comparably or whether there may be specific areas that indicate there may be problems. One caveat to this approach should be noted and considered. Asians and Native Hawaiian/Pacific Islanders appear to be substantially different in health status and use patterns from other minority racial/ethnic groups and more similar to white enrollees. Because of these differences, it might be more useful to analyze and report only findings for black and Hispanic/Latino persons, and Native Americans.

Augment the MMC CAHPS® to obtain additional information from racial/ethnic minority enrollees. Analysis of the MMC CAHPS® data, presented earlier in this article, indicates that racial/ethnic minority enrollees are slightly more likely to give lower ratings on process and access dimensions of plan performance and are less likely to have contact with primary care providers and with specialists. Information on the reasons for these differences in ratings and barriers to access and use of services has been collected under a related component of the MMC CAHPS® project that established Technical Expert Panels to advise CMS on developing strategies for performance improvement and conducted focus groups with black and Hispanic/Latino health plan enrollees to obtain insight into their direct experiences with processes and access to care (Barents Group, 2002). Results of these studies could provide useful information and guidance for designing relevant and appropriate additional questions for the MMC CAHPS®. For example, if a survey respondent reported that they had not visited their primary physician during the previous 3 months, they could be asked: “Why didn't you visit your primary physician” with alternative response categories (e.g., “Doctor didn't tell me to make an appointment,” “Tried to make an appointment, but couldn't get in,” “Too hard to get to the doctor,” etc.). Similarly, if respondents indicated that they had visited the ER, they could be asked to provide supplemental information (e.g., “Did your health plan pay for the ER visit or did you have to pay?” and “What is the reason that you went to the ER?”). Additional questions could also be developed to elicit information on whether the enrollees speak English as their primary language and, if not, whether they are able to see a doctor who speaks their language. Analyses of the responses to these additional questions, by racial/ethnic groups could provide valuable insights on issues that may pose barriers to effective use of services by minority enrollees of health plans.

Conduct analyses of the disenrollment CAHPS® data to assess differences between minority and majority disenrollees on ratings of specific dimensions of performance. The disenrollment CAHPS® data offer the possibility of identifying differences among disenrollees, by race and ethnicity, in their ratings of their health plan and in the reasons why they chose to leave the health plan. While the survey samples for the disenrollment CAHPS® and disenrollment “reasons” survey are not large enough to permit health plan-specific estimates, by race and ethnicity, the aggregate analysis could provide measures of differences in health plan performance by race and ethnicity. In turn, this performance measurement could yield information for developing strategies to improve performance of health plans in serving minority enrollees.

Conduct analyses comparing MMC CAHPS® responses and ratings with fee-for-service (FFS) CAHPS® responses and ratings. Whether MMC plans are serving minority enrollees as effectively as they serve majority enrollees and whether there are differences among MMC plans in meeting the needs of minority enrollees are both important issues. Of equal importance, however, is whether minority enrollees in health plans experience better access to and use of services than minority Medicare beneficiaries who choose to remain in original FFS Medicare. With the FFS CAHPS® survey now being fielded annually on the same schedule as the MMC CAHPS®, it would be feasible to conduct analyses that would answer the question: “Are minority Medicare beneficiaries better served in MMC plans or in original Medicare?” These comparisons could focus on ratings of overall, process, and access dimensions of performance. In addition, comparisons could be made to determine whether minority beneficiaries are more likely to visit a doctor in original Medicare or MMC plans or to have no visit, and other utilization measures that are captured in both surveys. It would also be informative to compare the likelihood of improvement or deterioration in health status for minority beneficiaries in original Medicare and in health plans. Measurement of the performance of MMC plans in meeting the needs of minority enrollees relative to the performance of original Medicare in meeting these needs, would provide context for interpreting and understanding the performance of all health plans and across plan differences.

Strategies Using Other Data Sources

To the extent that the Medicare enrollment database has been determined to have reliable and complete data on race and ethnicity of Medicare beneficiaries, it could be a tool for measuring performance of MMC health plans in serving minority enrollees. CMS is continuing to improve the racial/ethnic identifiers on the EDB and, eventually, these identifiers can be expected to be complete and accurate for most beneficiaries. Even if the EDB does not have complete data for all racial/ethnic groups, it could be used to examine some dimensions of performance for those groups for which the identifiers are relatively complete. One strategy that might be considered, that relies on the EDB, is to examine enrollment rates of minority Medicare beneficiaries in each health plan, relative to the minority population in the health plan's market area. A health plan in an area with significant minority population that has little or no minority enrollees may warrant examination by CMS to assess the reasons for the low enrollment. The health plan provider network, for example, may be geographically distributed in areas that are not convenient for minority beneficiaries to access. Customer service representatives and providers may be homogeneous and not reflective of the minority population's culture and language. Or, marketing strategies of the health plan may be developed to attract the majority population and not to attract minority enrollees (e.g., holding health fairs in majority population areas and not in minority population areas.)

Other strategies might be considered for measuring performance of health plans in serving minority enrollees, if and when reliable hospital and physician encounter data are available from health plans on an enrollee-specific basis. Given that patterns of use for minority members indicate lower use of primary care and specialist physician services and, generally, higher rates of hospitalization, the hospital and physician encounter data could be used to identify health plans that have higher or lower rates of minority member visits to physicians. In addition, analyses could be conducted to assess whether minority enrollees who see plan physicians more frequently are less likely to be hospitalized. However, it is unlikely that reliable hospital and physician encounter data will be available in the near future for use in performance measurement.

Improving Performance of MMC Plans

Issues for Improving Performance

A broad strategy for mandating or encouraging MMC plans to undertake programs that would improve the service to racial and ethnic minorities enrolled in these plans could be designed. Many States, for example, require MMC plans to recruit providers and customer service staff that are reflective of the racial/ethnic mix of their enrollees, make all materials available in all languages spoken by a minimum of 5 percent of enrollees, develop tracking systems and outreach plans to ensure that all members have initial appointments with their primary care providers, and provide cultural competency training to providers and staff.

MMC plans, however, enroll a disproportionate number of racial/ethnic minorities, relative to the proportion of minorities who are enrolled in MMC plans. Less than 50 percent of MMC plans have more than 10 percent enrollment of racial/ethnic minorities. Given this, a broad strategy that mandates MMC plans undertake significant new programs would probably not be warranted.

In addition, there would be significant costs to MMC plans if CMS were to mandate that all MMC plans meet extensive new requirements to improve performance in serving minority enrollees. Imposing new requirements and costs would likely result in some MMC plans withdrawing from Medicare participation. This consequence would not be beneficial, nor justified, since minority enrollees in Medicare plans are generally satisfied with most aspects of the care they receive from their plans. And, research has demonstrated that minority enrollees may be disproportionately affected by withdrawal of MMC plans (Laschober et al., 1999).

With these considerations in mind, the development of strategies to improve performance of MMC plans in serving their racial/ethnic minority members should focus on development of tools, by CMS, that could be voluntarily adopted by MMC plans, feedback to health plans of performance measures that provide information on how minority members are faring in specific health plans, and incentives to encourage health plans to improve aspects of process and access that would result in increased use of services and better outcomes for their minority members.

Strategies Improving Performance of MMC Plans

Development of Tools to Improve Performance

CMS could develop or support the development of tools that could be adopted voluntarily by MMC plans that want to address specific areas of performance that are most relevant for minority members. Examples of the type of tools that might be developed by CMS include:

Cultural competency training programs for health plan providers and staff.

Outreach/education programs targeted to informing minority members about the importance of routine visits to primary care providers and preventive care.

Best practices programs for addressing specific health conditions that disproportionately affect minorities.

Develop Effective Cultural Competency Training Programs

The importance of culture in health care has received increased attention in the last two decades. Its importance is particularly relevant for racial/ethnic minority patients who receive health care from systems that are largely organized and staffed with majority group members. The process of care measures that capture perceptions of doctor-patient interactions in the MMC CAHPS®, in fact, do not show significant differences by race/ethnicity However, the literature discussed earlier on race concordance and differences in patient perceptions of care suggest that cultural competency may be an important issue for improving outcomes for racial/ethnic groups, given the underrepresentation of minorities in the medical profession. Examples of negative health consequences that could result from lack of cultural awareness include missed opportunities for screening due to unfamiliarity with the prevalence of conditions among certain minority groups; failure to take into account differing responses to medication; lack of knowledge about traditional remedies leading to harmful drug interactions; and diagnostic errors resulting from miscommunication.

There is, however, an inadequate base of knowledge on effective training mechanisms to increase cultural competency, and on the impact of cultural competency training on behaviors and outcomes for minorities. A recent article reviews and synthesizes the literature on cultural competency objectives and strategies and considers the potential impact of cultural competency programs on racial and ethnic health disparities (Brach and Fraser, 2000). The authors develop a conceptual model of the potential for cultural competency to reduce disparities, using the research literature both to develop the model and to analyze whether the evidence supported it. They conclude that, theoretically, cultural competency training should result in improved outcomes. However, they point out that there is inadequate evidence on the type of cultural competency training that is effective and on the actual impacts of cultural competency training on outcomes.

There are a number of unanswered questions related to cultural competency training programs that would need to be addressed in designing an effective program. These include:

What are the important elements of a cultural competency training program that would result in increased awareness and sensitivity of health care providers and staff to cultural differences that affect health care use, compliance, and outcomes?

Is a single training program sufficient, or do separate programs need to be developed for health plan staff and for providers?

Do separate cultural competency training programs need to be developed for each racial and ethnic group (and for subgroups—e.g., Mexican-Americans and Cuban-Americans), or can a generic cultural competency training program be adequate to increase staff and provider awareness and knowledge?

Is there an impact of cultural competency training on health care use, compliance, and health outcomes of racial and ethnic minorities?

CMS could undertake or support a research agenda to address these outstanding questions and to implement pilot projects that would test the effectiveness of different strategies for increasing cultural competency within health plans. Results of this research and evaluation of the effectiveness of the pilot programs could then be used to develop cultural competency training programs, and associated materials. Then, CMS could disseminate broad information about the importance of cultural competency, its impacts on health outcomes of minorities, and make these program and training materials available to all MMC plans and encourage their use.

Increase Use of Preventive Care

Evidence presented earlier in this article indicates that racial/ethnic minorities are less likely to visit their primary care provider and more likely to report that their health status is “fair” or “poor.” Routine visits to primary care providers and preventive care could increase early detection of disease. Early detection could, in turn, reduce deterioration of health and improve health outcomes and health status over time. An effective strategy for outreach and education of minority members of health plans could result in increased routine visits to primary care providers, with associated improvements in health.

CMS has undertaken considerable research on effective outreach and educational strategies to communicate important health insurance and health information to racial/ethnic minorities enrolled in Medicare, Medicaid, and the State Child Health Insurance Program (Barents Group, 1998a, b). Synthesis of this research, and other related research, could be conducted by CMS, focusing particularly on evidence of effective strategies for communicating health information to specific populations. Once evidence on effective strategies and best practices is assembled, CMS could use this evidence to design messages, outreach strategies, and materials for a comprehensive program to communicate the importance of routine and preventive care to minority members of MMC plans. CMS could then disseminate data from the MMC CAHPS® on lower primary care provider visit rates of minority enrollees, information on health status disparities between minorities and the majority population, and the potential for health plans to reduce costs through prevention and early detection, and announce the availability to MMC plans of outreach strategies and materials designed to increase routine visits and preventive care by minorities.

Develop and Disseminate Best Practices Guidelines

CMS, either directly or through the peer review organizations, could initiate a project to review and summarize research on specific health conditions that disproportionately affect minorities and on health conditions for which outcomes of the health condition have been determined to be worse for minorities. The objectives of this research review would include:

Identifying those health conditions for which minorities are at greater risk.

Identifying health conditions for which minorities exhibit worse outcomes.

Identifying underlying factors that contribute to higher prevalence and/or worse outcomes.

Summarizing evidence of differential treatment patterns by providers that may contribute to differential outcomes.

Determining effective strategies for encouraging minorities to seek early detection of specific conditions.

Determining whether different strategies are needed for effective management of specific health conditions for minorities.

If the review of research provided a solid foundation for developing guidelines for best practices to improve early detection, effective management, and improved outcomes for minorities, then CMS could develop a series of best practices guidelines and informational white papers to disseminate to health plans (and to original Medicare providers).

If the review of research did not provide strong evidence, then CMS could consider undertaking specific research and pilot programs that would build a foundation for developing best practices guidelines. These research and pilot programs could be targeted to a limited number of health conditions that have been identified as having the greatest negative impact on health status, morbidity, and mortality of minority Medicare beneficiaries.

Provide Feedback to MMC Plans

While a number of MMC plans have few minority members, a majority (53 percent) has 10 percent or more (combined) minority enrollment. CMS could provide to those health plans with at least 10 percent minority enrollment some selected MMC CAHPS® measures and data, with national and regional comparisons of these data, broken out by (combined) minority and non-minority enrollees. For example, CMS might consider providing:

Ratings of courtesy and respect.

Ratings of doctor spending enough time.

Ratings of difficulty in access to specific services.

Proportion who did not have a physician visit in past 6 months.

Proportion who did not see a specialist physician.

Proportion who used ER services.

This information could be included in each MMC CAHPS® plan's health plan report each year, with national or regional comparisons. While the numbers of respondents in each plan who are racial/ethnic minorities may be small, for some plans the number of respondents may be sufficiently large to provide reliable estimates of differences in perceptions and experiences of minority enrollees relative to non-minorities. Narrative in the plan report section and table footnotes could distinguish for plans the reliability of the estimates.

If provided to health plans, this information might highlight for some health plans areas for improvement. In addition, it could provide CMS central office and regional office staff with data that could be used to encourage health plans to address disparities in performance, perceptions, and access/use of services.

Provide Incentives to MMC Plans

CMS could consider offering incentives to MMC plans to develop and/or implement specific practices and programs to improve performance in serving racial/ethnic minorities. These incentives could be financial (e.g., a bonus to health plans that demonstrate a higher proportion of minority members visiting a doctor this year compared with last year). Alternatively, health plans demonstrating improved performance or implementing strategies to improve performance in serving minority enrollees could be relieved from some types of regulatory/reporting requirements.

CMS could work with the managed care industry and MMC plans to develop appropriate (and feasible) incentives that would be effective in encouraging MMC plans to develop and participate in initiatives that could lead to better service and improved outcomes for minority members.

Conclusion

MMC plans, in general, are meeting the needs of their racial/ethnic minority members. Overall, most minority enrollees give their health plans good ratings that are not significantly different from the ratings given by non-minority members. Based on the MMC CAHPS® data, however, it appears that:

There is evidence that some minority enrollees have more difficulty in obtaining access to some services than do non-minority enrollees.

Minority enrollees, even when accounting for health conditions, are less likely to have any visit to their regular provider.

Minority enrollees are less likely to visit a specialist physician.

Minority enrollees are more likely to be hospitalized and to use ER services.

Minority enrollees are more likely to report fair or poor health status.

On the positive side, minority enrollees are more likely than non-minority enrollees to report that their health has improved in the past year—suggesting that MMC plans are providing beneficial care to their minority members. And, it is possible that the analysis of FFS CAHPS® data could produce findings that indicate that minority Medicare beneficiaries in managed care plans are receiving care that is more accessible and of greater quantity than minority Medicare beneficiaries in original Medicare.

Reducing racial/ethnic disparities in access to and use of health services and in health status and health outcomes is an important goal for CMS. Even if MMC plans are serving their minority members well, disparities in use of services and in health status remain. CMS could undertake additional monitoring of MMC plan performance in serving minorities and, with these data in hand, could develop a variety of programs and incentives that could reduce disparities and improve the health of minority Medicare beneficiaries. The strategies discussed in this article represent some potential avenues that CMS could consider to achieve those objectives.