- Email: [email protected]
Do Health Care Ratings Differ by Race or Ethnicity?
Joint Commission
Journal on Quality and Safety
Performance Measures
Do Health Care Ratings Differ by Race or Ethnicity?
Mark G. Haviland, PhD Leo S. Morales, MD, PhD Steven P. Reise, PhD Ron D. Hays, PhD
lthough health status indicators generally show marked improvement for racial/minority group members, substantial differences among groups remain.1 Moreover, recent studies have documented disparities in perceptions of health care by race and ethnicity. In particular, there is growing evidence that Asians and Pacific Islanders perceive their health care more negatively than whites and those in other racial and ethnic subgroups.2–5 There is also evidence that Hispanics, especially those who speak little or no English, have more negative experiences with health care than others.3,6–7 Studies of differences in perceptions of care by different racial and ethic groups are particularly timely for several reasons. First, the Institute of Medicine (IOM), in its report on racial and ethnic health care disparities, has noted the paucity of research on disadvantaged populations, especially groups other than nonAfrican Americans.8 Moreover, the IOM has called for research on health care access and use by patients’ race/ethnicity. Second, the National Quality Forum has recommended in its report, Improving Healthcare Quality for Minority Patients, analyses of existing health care quality measures by race/ethnicity, as well as studies to help us better understand the state of racial/ethnic health care disparities.9 Finally, the findings of the present study may be useful to the Agency for Healthcare Research and Quality as it prepares its report (due in September 2003) on health care in priority populations, which include racial/ethnic minorities.10 This study was designed to examine differences in health care experiences by race and ethnicity in a nationally representative sample that includes substantial numbers of minority respondents, including American Indians/Alaska Natives. We analyzed the 1998 National
A
134
Article-at-a-Glance Background: There is growing evidence that Asians and Pacific Islanders perceive their health care more negatively than whites and other racial and ethnic subgroups. This study of differences in health care experiences by race and ethnicity was the first to use nationally representative data. Method: Data from the 1998 National Research Corporation Healthcare Market Guide® survey were analyzed. A total of 120,855 respondents were included in the study. Four global satisfaction ratings (overall health plan satisfaction, medical care satisfaction, recommend plan to others, and intent to switch plans) and four composite measures (access to care, providers’ delivery of care, customer service, and cost/benefits of care) were examined. Results: Nonwhite survey respondents—particularly those in the other/multiracial and Asian/Pacific Islander groups—rated their health plan coverage and medical care lower than whites. Discussion: The results of this study are consistent with those of other recent (and comparable) studies in which these racial and ethnic groups are represented. Most strikingly consistent, however, are the lower ratings of Asians/Pacific Islanders and the comparable (and higher) ratings (compared to whites) of African Americans. Why Asians/Pacific Islanders are considerably less satisfied with their medical care than all other racial and ethnic groups in the United States needs to be explored. Access to care and quality of care improvement efforts should be directed at all ethnic minority groups, particularly for limited English-speaking, other/multiracial, Hispanic, and Asian/Pacific Islander subgroups.
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Research Corporation (NRC) Healthcare Market Guide® (HCMG®) survey11 to evaluate racial and ethnic differences in ratings of health care among African American, American Indian/Alaska Native, Asian/Pacific Islander, Hispanic, and other/multiracial subgroups.
Method HCMG® Survey The NRC has been conducting the NRC HCMG® since 1987. Each year content areas are expanded, survey questions are structured to meet objectives set by health care marketing directors and strategic planners, and surveys are then pretested in field conditions. Health plan satisfaction items are derived from the National Committee for Quality Assurance’s Annual Member Health Care Survey, Version 1.0.12
Sample The 1998 NRC HCMG® survey asked consumers to assess their health plans, personal physicians, and local hospitals and health systems. The survey was mailed in March 1998 to 250,000 households in the 48 contiguous states, and returns were closed in May 1998. The sample—drawn from a panel of respondents recruited and maintained by an opinion research firm—was selected to match U.S. Census demographics on geographic region, household size, age of head of household, population density, and annual household income. A total of 163,641 surveys were returned (respondents were the primary health care decision maker in each household). The analytic sample included the 120,855 respondents who answered at least two of the four global satisfaction measures and at least half of the items on each of the four composite satisfaction scales (described below). Missing values were replaced with means of available item responses. If a respondent had a missing value on one of the global items, it was replaced with the average of the other three global items; similarly, the composite scales were constructed by averaging non-missing items together. Of the 33,156 respondents for whom data were missing, only 2,625 had more than three missing data points. The sample for multivariate analyses was 98,204; respondents with missing case-mix (adjustment) data were not included.
HCMG® Survey Measures The dependent variables in this study were four global rating questions and four composite (multiple-item) scales. Global questions. The first two global questions— “All things considered, how satisfied are you with your health plan?” and “All things considered, how satisfied are you with the medical care you received from your health plan?”—were rated on a 7-point scale, which ranged from 1 (completely dissatisfied) to 7 (completely satisfied). The third global question—“Would you recommend your health plan to family or friends if they needed care?”—was rated on a 4-point scale ranging from 1 (definitely not) to 4 (definitely yes). The fourth global question—“Do you intend to switch to another health plan when you next have an opportunity?”—was rated on a 4-point scale ranging from 1 (definitely yes) to 4 (definitely no). Multiple-item composite questions. The multipleitem composite questions assessed access to care (5 items), care delivered by the provider (10 items), health plan customer service (2 items), and cost/benefits of care (4 items). Each item was assessed using a 5-point poor to excellent rating scale. Independent variables. Race/ethnicity was the main independent variable. Respondents were categorized as Hispanic (n = 3,168) if they answered yes to the question “Are you of Spanish/Hispanic origin or descent?” (regardless of race). Respondents who answered no or had missing data for this question were categorized according to their answers to a second race/ethnicity question: black or African American (n = 7,134), American Indian or Alaska Native (n = 606), Asian or Pacific Islander (n = 944), white (n = 103,218), another race or multiracial (n = 649), or missing (n = 5,136). Case-mix (adjustment) variables included age, sex, perceived health status, education, and health plan type.
Data Analysis We compared survey respondents with the analytic sample and the multivariate sample on the case mix and race/ethnicity variables. Then we cross-tabulated race/ethnicity with each of the other independent (case-mix) variables and tested the significance of associations by using a chi-square statistic. We estimated
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
135
Joint Commission
Journal on Quality and Safety
Cronbach’s alpha13 to assess internal consistency reliability of the multiple-item composites. We computed mean scores by race/ethnicity on each of the four global ratings, all of the individual items within each of the four composite measures, and each of the four composite measures. Mean differences were tested for statistical significance with one-way analysis of variance. We used linear regression to assess differences in global and composite rating scores between whites and members of the other race/ethnic groups, controlling for age, sex, perceived health status, education, and health plan type. For these analyses, n = 98,204 (respondents for whom complete case-mix data were available). We estimated a separate ordinary least-squares-regression model for each global and composite measure.
Results All survey respondents (n = 163,641) and those in the analytic sample (n = 120,855) and the multivariate sample (n = 98,204) looked very similar in terms of age, sex, educational attainment, self-rated health status, insurance type, and race/ethnicity (Table 1, p 137). Hence, the loss of cases due to missing data did not unduly affect the composition of the sample. Across race/ethnicity, the differences in the distributions of all case-mix variables were statistically significant (2, p < .01). The distributions by racial and ethnic group are shown in Table 2 (p 138). Internal consistency reliability estimates were adequate for each of the four composite scores: 0.88 for access to care, 0.95 for providers’ delivery of care, 0.73 for plan customer service, and 0.87 for cost/benefits of care. Means and standard deviations for the 4 global rating questions, the 21 individual items, and the 4 composite measures (total of scale items divided by number of items in the respective scales) are shown in Table 3 (pp 139–140). All ratings were above scale midpoints. All racial and ethnic group mean differences (global, individual, and composite) were significant (one-way analysis of variance models, p < .01). In general, persons in the other/multiracial and Asian/Pacific Islander groups gave the lowest ratings. The regression results, including standardized beta coefficients and robust standard errors, are shown in Table 4 (pp 141–143). On several measures, the differences
136
between whites and minority group members were statistically significant (p 143). For example, on the global measures, African Americans gave higher ratings than whites on overall health plan satisfaction and recommend health plan. Respondents in the other/multiracial group rated all four measures lower than whites. Asians/Pacific Islanders rated medical care satisfaction lower and were more likely to intend to switch health plans than whites. On the composite measures, other/multiracial group members and Asians/Pacific Islanders rated their health plans and medical care lower than whites on all four measures. African Americans rated providers’ delivery of care lower and plan customer service higher than whites. American Indians/Alaska Natives and Hispanics rated access to care and cost/benefits of care lower; Hispanics also rated cost/benefits of care lower than whites. Ratings of health care tended to be more positive among those who were older and for those in better health. Those at the highest educational level (> high school) tended to be more negative about several aspects of care, but the least educated (< high school) were less satisfied with access to care, providers’ delivery of care, and cost/benefits of care. Persons with traditional fee-for-service health insurance were satisfied with most aspects of care but were less satisfied with plan customer service and cost/benefits of care than those in health maintenance organization plans.
Discussion As in previous studies, consumers in this study gave generally high ratings of their health plans and medical care. Respondents in the other/multiracial group and Asians/Pacific Islanders, however, were more negative in their assessments than their white counterparts. African Americans gave ratings that were comparable to, and in some instances higher than, those of whites. Finally, Hispanics and American Indians/Alaska Natives rated aspects of their medical care and health plans lower than whites. These results are consistent with those of other recent (and comparable) studies in which these racial and ethnic groups are represented.2,3 Most strikingly consistent, however, are the lower ratings of Asians/Pacific Islanders and the comparable (and higher) ratings (compared to whites) of African Americans.
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 1. Case Mix and Race/Ethnicity Distributions for Respondents, Analytic Sample, and Multivariate Sample*
* Percentages do not always add up to 100.0 because of rounding. POS, point of service; PPO, preferred provider organization; HMO, health maintenance organization.
This study is important for several reasons. First, in contrast to previous studies, it used nationally representative data. All previous studies of racial and ethnic differences in patients’ reports about care used either regional or practice-based samples or national but not representative samples.2,4,7 Second, the NRC HCMG® survey included relatively large numbers of minority respondents. Most previous studies of racial and ethnic differences in patients’ experiences with care evaluated
differences between blacks and whites and in a few cases, included small samples of Asians/Pacific Islanders. This study included 944 Asians/Pacific Islander and 606 American Indian/Native Alaskan respondents. Third, this study is consistent with all previous studies that have included Asians/Pacific Islanders in showing worse reports and lower ratings of care by members of this minority group. Collectively, these studies suggest that Asians/Pacific Islanders are
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
137
Joint Commission
Journal on Quality and Safety
Table 2. Sample Characteristics by Race/Ethnicity, 1998 National Research Corporation Healthcare Market Guide® Survey*
* Percentages do not always add up to 100.0 because of rounding. POS, point of service; PPO, preferred provider organization; HMO, health maintenance organization.
138
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 3. Global and Composite Satisfaction Measures by Race/Ethnicity: Means and Standard Deviations (N = 120,855)
continued
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
139
Joint Commission
Journal on Quality and Safety
Table 3. Global and Composite Satisfaction Measures by Race/Ethnicity: Means and Standard Deviations (N = 120,855) (continued)
140
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 4. Regression Results: Significant (p < .05) Standardized Beta Coefficients and Robust Standard Errors*
continued
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
141
Joint Commission
Journal on Quality and Safety
Table 4. Regression Results: Significant (p < .05) Standardized Beta Coefficients and Robust Standard Errors* (continued)
142
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 4. Regression Results: Significant (p < .05) Standardized Beta Coefficients and Robust Standard Errors* (continued)
* POS, point of service; PPO, preferred provider organization; HMO, health maintenance organization.
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
143
Joint Commission
Journal on Quality and Safety
considerably less satisfied with their medical care than all other racial and ethnic groups in the United States. More research is needed to understand the underlying reasons for this disparity. Available evidence, however, suggests that researchers should evaluate potential response bias, language barriers, and Asian/Pacific Islander subgroups (for example, at the very least, disaggregating Asians and Pacific Islanders, as MurrayGarcia, et al suggest14). Among the limitations of the study is sample representativeness. Although the overall response rate of 65% was acceptable, we are not certain that the 163,641 respondents were a good representation of the 250,000 survey recipients. Moreover, we were able to use only a subset of 120,855 respondents (74% of all respondents and 48% of those mailed surveys) who answered two of the four global satisfaction questions and at least half of the items on each of the four composite satisfaction scales. Although the loss of cases due to missing data appears not to have unduly affected sample composition (as shown in Table 1), one, nevertheless, must be cautious about generalizing these findings. Next, all questionnaires were administered in English, which may have inflated ratings within Asian/Pacific Islander and Hispanic subgroups by excluding potential survey respondents with limited Englishlanguage ability. Also, as Murray-Garcia et al14 have noted, systematic differences in the ways in which racial ethnic minorities respond to surveys may not be inconsequential. African American and Hispanic respondents, for example, may more often than whites choose extreme response options; those of Asian descent, however, are more likely to choose response categories in the middle of such scales. Another limitation is that the study explored patients’ experiences with care, not the technical quality of care; thus, answers to survey questions were based on subjective judgments. Finally, insofar as this was a study of adults with health insurance, ratings were likely to have been higher than would be obtained among the uninsured. For example, access to care has been found, as expected, to be worse for those without health insurance than for those with insurance.15 Given that racial and ethnic minorities’ health plan and medical care ratings tend to be high—in both the present study and the Morales et al2 study—it is tempting to conclude that disparities in health care are inconsequential.
144
These ratings may, in fact, reflect solid health status gains1; on the other hand, gaps in coverage and care persist,1,16 and the majority of Americans are unaware of these disparities.16 To reach the Department of Health and Human Services goal of eliminating racial and ethnic disparities in health care by the year 2010,17 public awareness of these disparities needs to increase, and programs to improve care, developed and implemented. Available data suggest that access to care and quality of care improvement efforts should be directed at all ethnic minority groups, particularly for limited-Englishspeaking, other/multiracial, Hispanic, and Asian/Pacific Islander subgroups. Efforts to improve access to care may require extended office hours, providing transportation to and from physicians’ offices, and employing bilingual office staff to answer phones and make appointments. Efforts to improve the quality of medical care may require increasing the diversity of provider staff at hospitals and clinics, providing professional interpreters for medical encounters, and providing well-translated patient education materials in multiple languages. Research is needed to better characterize individuals within Asian and other subgroups. Intragroup differences may be at least as important as the intergroup differences described in this article. Unfortunately, race/ethnicity data are rarely collected at a sufficiently fine level of specificity—and that was the case with these NRC HCMG® data. Further surveys need to be designed to capture this information. J Leo S. Morales was supported by a Robert Wood Johnson Minority Medical Faculty Development Program Fellowship. We thank Michael D. Hays of the National Research Corporation for providing access to the data.
Mark G. Haviland, PhD, is Professor, Department of Psychiatry, Loma Linda University School of Medicine, Loma Linda, California. Leo S. Morales, MD, PhD, is Assistant Professor, Division of General Internal Medicine and Health Services Research, David Geffen School of Medicine at University of California at Los Angeles (UCLA). Steven P. Reise, PhD, is Professor, Department of Psychology, UCLA. Ron D. Hays, PhD, is Professor, Division of General Internal Medicine and Health Services Research, David Geffen School of Medicine at UCLA. Please address correspondence to Mark G. Haviland, PhD, [email protected].
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
References 1. Keppel KG, Pearcy JN, Wagener DK: Trends in Racial and EthnicSpecific Rates for the Health Status Indicators: United States, 1990–98. Healthy People Statistical Notes, No. 23. Hyattsville, MD: National Center for Health Statistics, 2000. 2. Morales LS, et al: Differences in CAHPS adult survey reports and ratings by race and ethnicity: An analysis of the National CAHPS Benchmarking Data 1.0. Health Serv Res 36:595–617, 2001. 3. Weech-Maldonado R, et al: Racial and ethnic differences in parents’ assessments of pediatric care in Medicaid managed care. Health Serv Res 36:575–594, 2001. 4. Snyder R, et al: Access to medical care reported by Asians and Pacific Islanders in a west coast physician group association. Med Care Res Rev 57:196–215, 2000. 5. Meredith LS, Sui AL: Variation and quality of self-report health data: Asians and Pacific Islanders compared with other ethnic groups. Med Care 33:1120–1131, 1995. 6. Fiscella K, et al: Disparities in health care by race, ethnicity, and language among the insured: Findings from a national sample. Med Care 40:52–59, 2002. 7. Morales LS, et al: Are Latinos less satisfied with communication from health care providers? J Gen Intern Med 14:409–417, 1999. 8. Institute of Medicine: Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC: National Academies Press, 2002.
9. National Quality Forum: Improving Healthcare Quality for Minority Patients. National Forum for Health Care Quality Measurement and Reporting, Washington, DC: 2002. 10. National Healthcare Disparities Report. Fact Sheet. AHRQ Publication No. 03-P007. Agency for Healthcare Research and Quality, Rockville, MD, Oct 2002. (www.ahrq.gov/news/nhdrfact.htm; last accessed Jan 24, 2003). 11. National Research Corporation: NRC Healthcare Market Guide Survey. Lincoln, NE, 1998. 12. Kippen LS, Strasser S, Joshi M: Improving the quality of the NCQA (National Committee for Quality Assurance) Annual Member Health Care Survey, Version 1.0. Am J Manag Care 3:719–730, 1997. 13. Cronbach LJ: Coefficient alpha and the internal structure of tests. Psychometrika 16:297–334, 1951. 14. Murray-Garcia JL, et al: Racial and ethnic differences in a patient survey: Patients’ values, ratings, and reports regarding physician primary care performance in a large health maintenance organization. Med Care 38:300–310, 2000. 15. Weissman JS, Epstein AM: The insurance gap: Does it make a difference? Annu Rev Public Health 14:243–270, 1993. 16. Lillie-Blanton M, et al: Race, ethnicity, and the health care system: Public perceptions and experiences. Med Care Res Rev 57:218–235, 2000. 17. U.S. Deparment of Health and Human and Services: Healthy People 2010. Washington, DC, 2000. (www.healthypeople.gov/publications; last accessed Jan 24, 2003).
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
145
Journal on Quality and Safety
Performance Measures
Do Health Care Ratings Differ by Race or Ethnicity?
Mark G. Haviland, PhD Leo S. Morales, MD, PhD Steven P. Reise, PhD Ron D. Hays, PhD
lthough health status indicators generally show marked improvement for racial/minority group members, substantial differences among groups remain.1 Moreover, recent studies have documented disparities in perceptions of health care by race and ethnicity. In particular, there is growing evidence that Asians and Pacific Islanders perceive their health care more negatively than whites and those in other racial and ethnic subgroups.2–5 There is also evidence that Hispanics, especially those who speak little or no English, have more negative experiences with health care than others.3,6–7 Studies of differences in perceptions of care by different racial and ethic groups are particularly timely for several reasons. First, the Institute of Medicine (IOM), in its report on racial and ethnic health care disparities, has noted the paucity of research on disadvantaged populations, especially groups other than nonAfrican Americans.8 Moreover, the IOM has called for research on health care access and use by patients’ race/ethnicity. Second, the National Quality Forum has recommended in its report, Improving Healthcare Quality for Minority Patients, analyses of existing health care quality measures by race/ethnicity, as well as studies to help us better understand the state of racial/ethnic health care disparities.9 Finally, the findings of the present study may be useful to the Agency for Healthcare Research and Quality as it prepares its report (due in September 2003) on health care in priority populations, which include racial/ethnic minorities.10 This study was designed to examine differences in health care experiences by race and ethnicity in a nationally representative sample that includes substantial numbers of minority respondents, including American Indians/Alaska Natives. We analyzed the 1998 National
A
134
Article-at-a-Glance Background: There is growing evidence that Asians and Pacific Islanders perceive their health care more negatively than whites and other racial and ethnic subgroups. This study of differences in health care experiences by race and ethnicity was the first to use nationally representative data. Method: Data from the 1998 National Research Corporation Healthcare Market Guide® survey were analyzed. A total of 120,855 respondents were included in the study. Four global satisfaction ratings (overall health plan satisfaction, medical care satisfaction, recommend plan to others, and intent to switch plans) and four composite measures (access to care, providers’ delivery of care, customer service, and cost/benefits of care) were examined. Results: Nonwhite survey respondents—particularly those in the other/multiracial and Asian/Pacific Islander groups—rated their health plan coverage and medical care lower than whites. Discussion: The results of this study are consistent with those of other recent (and comparable) studies in which these racial and ethnic groups are represented. Most strikingly consistent, however, are the lower ratings of Asians/Pacific Islanders and the comparable (and higher) ratings (compared to whites) of African Americans. Why Asians/Pacific Islanders are considerably less satisfied with their medical care than all other racial and ethnic groups in the United States needs to be explored. Access to care and quality of care improvement efforts should be directed at all ethnic minority groups, particularly for limited English-speaking, other/multiracial, Hispanic, and Asian/Pacific Islander subgroups.
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Research Corporation (NRC) Healthcare Market Guide® (HCMG®) survey11 to evaluate racial and ethnic differences in ratings of health care among African American, American Indian/Alaska Native, Asian/Pacific Islander, Hispanic, and other/multiracial subgroups.
Method HCMG® Survey The NRC has been conducting the NRC HCMG® since 1987. Each year content areas are expanded, survey questions are structured to meet objectives set by health care marketing directors and strategic planners, and surveys are then pretested in field conditions. Health plan satisfaction items are derived from the National Committee for Quality Assurance’s Annual Member Health Care Survey, Version 1.0.12
Sample The 1998 NRC HCMG® survey asked consumers to assess their health plans, personal physicians, and local hospitals and health systems. The survey was mailed in March 1998 to 250,000 households in the 48 contiguous states, and returns were closed in May 1998. The sample—drawn from a panel of respondents recruited and maintained by an opinion research firm—was selected to match U.S. Census demographics on geographic region, household size, age of head of household, population density, and annual household income. A total of 163,641 surveys were returned (respondents were the primary health care decision maker in each household). The analytic sample included the 120,855 respondents who answered at least two of the four global satisfaction measures and at least half of the items on each of the four composite satisfaction scales (described below). Missing values were replaced with means of available item responses. If a respondent had a missing value on one of the global items, it was replaced with the average of the other three global items; similarly, the composite scales were constructed by averaging non-missing items together. Of the 33,156 respondents for whom data were missing, only 2,625 had more than three missing data points. The sample for multivariate analyses was 98,204; respondents with missing case-mix (adjustment) data were not included.
HCMG® Survey Measures The dependent variables in this study were four global rating questions and four composite (multiple-item) scales. Global questions. The first two global questions— “All things considered, how satisfied are you with your health plan?” and “All things considered, how satisfied are you with the medical care you received from your health plan?”—were rated on a 7-point scale, which ranged from 1 (completely dissatisfied) to 7 (completely satisfied). The third global question—“Would you recommend your health plan to family or friends if they needed care?”—was rated on a 4-point scale ranging from 1 (definitely not) to 4 (definitely yes). The fourth global question—“Do you intend to switch to another health plan when you next have an opportunity?”—was rated on a 4-point scale ranging from 1 (definitely yes) to 4 (definitely no). Multiple-item composite questions. The multipleitem composite questions assessed access to care (5 items), care delivered by the provider (10 items), health plan customer service (2 items), and cost/benefits of care (4 items). Each item was assessed using a 5-point poor to excellent rating scale. Independent variables. Race/ethnicity was the main independent variable. Respondents were categorized as Hispanic (n = 3,168) if they answered yes to the question “Are you of Spanish/Hispanic origin or descent?” (regardless of race). Respondents who answered no or had missing data for this question were categorized according to their answers to a second race/ethnicity question: black or African American (n = 7,134), American Indian or Alaska Native (n = 606), Asian or Pacific Islander (n = 944), white (n = 103,218), another race or multiracial (n = 649), or missing (n = 5,136). Case-mix (adjustment) variables included age, sex, perceived health status, education, and health plan type.
Data Analysis We compared survey respondents with the analytic sample and the multivariate sample on the case mix and race/ethnicity variables. Then we cross-tabulated race/ethnicity with each of the other independent (case-mix) variables and tested the significance of associations by using a chi-square statistic. We estimated
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
135
Joint Commission
Journal on Quality and Safety
Cronbach’s alpha13 to assess internal consistency reliability of the multiple-item composites. We computed mean scores by race/ethnicity on each of the four global ratings, all of the individual items within each of the four composite measures, and each of the four composite measures. Mean differences were tested for statistical significance with one-way analysis of variance. We used linear regression to assess differences in global and composite rating scores between whites and members of the other race/ethnic groups, controlling for age, sex, perceived health status, education, and health plan type. For these analyses, n = 98,204 (respondents for whom complete case-mix data were available). We estimated a separate ordinary least-squares-regression model for each global and composite measure.
Results All survey respondents (n = 163,641) and those in the analytic sample (n = 120,855) and the multivariate sample (n = 98,204) looked very similar in terms of age, sex, educational attainment, self-rated health status, insurance type, and race/ethnicity (Table 1, p 137). Hence, the loss of cases due to missing data did not unduly affect the composition of the sample. Across race/ethnicity, the differences in the distributions of all case-mix variables were statistically significant (2, p < .01). The distributions by racial and ethnic group are shown in Table 2 (p 138). Internal consistency reliability estimates were adequate for each of the four composite scores: 0.88 for access to care, 0.95 for providers’ delivery of care, 0.73 for plan customer service, and 0.87 for cost/benefits of care. Means and standard deviations for the 4 global rating questions, the 21 individual items, and the 4 composite measures (total of scale items divided by number of items in the respective scales) are shown in Table 3 (pp 139–140). All ratings were above scale midpoints. All racial and ethnic group mean differences (global, individual, and composite) were significant (one-way analysis of variance models, p < .01). In general, persons in the other/multiracial and Asian/Pacific Islander groups gave the lowest ratings. The regression results, including standardized beta coefficients and robust standard errors, are shown in Table 4 (pp 141–143). On several measures, the differences
136
between whites and minority group members were statistically significant (p 143). For example, on the global measures, African Americans gave higher ratings than whites on overall health plan satisfaction and recommend health plan. Respondents in the other/multiracial group rated all four measures lower than whites. Asians/Pacific Islanders rated medical care satisfaction lower and were more likely to intend to switch health plans than whites. On the composite measures, other/multiracial group members and Asians/Pacific Islanders rated their health plans and medical care lower than whites on all four measures. African Americans rated providers’ delivery of care lower and plan customer service higher than whites. American Indians/Alaska Natives and Hispanics rated access to care and cost/benefits of care lower; Hispanics also rated cost/benefits of care lower than whites. Ratings of health care tended to be more positive among those who were older and for those in better health. Those at the highest educational level (> high school) tended to be more negative about several aspects of care, but the least educated (< high school) were less satisfied with access to care, providers’ delivery of care, and cost/benefits of care. Persons with traditional fee-for-service health insurance were satisfied with most aspects of care but were less satisfied with plan customer service and cost/benefits of care than those in health maintenance organization plans.
Discussion As in previous studies, consumers in this study gave generally high ratings of their health plans and medical care. Respondents in the other/multiracial group and Asians/Pacific Islanders, however, were more negative in their assessments than their white counterparts. African Americans gave ratings that were comparable to, and in some instances higher than, those of whites. Finally, Hispanics and American Indians/Alaska Natives rated aspects of their medical care and health plans lower than whites. These results are consistent with those of other recent (and comparable) studies in which these racial and ethnic groups are represented.2,3 Most strikingly consistent, however, are the lower ratings of Asians/Pacific Islanders and the comparable (and higher) ratings (compared to whites) of African Americans.
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 1. Case Mix and Race/Ethnicity Distributions for Respondents, Analytic Sample, and Multivariate Sample*
* Percentages do not always add up to 100.0 because of rounding. POS, point of service; PPO, preferred provider organization; HMO, health maintenance organization.
This study is important for several reasons. First, in contrast to previous studies, it used nationally representative data. All previous studies of racial and ethnic differences in patients’ reports about care used either regional or practice-based samples or national but not representative samples.2,4,7 Second, the NRC HCMG® survey included relatively large numbers of minority respondents. Most previous studies of racial and ethnic differences in patients’ experiences with care evaluated
differences between blacks and whites and in a few cases, included small samples of Asians/Pacific Islanders. This study included 944 Asians/Pacific Islander and 606 American Indian/Native Alaskan respondents. Third, this study is consistent with all previous studies that have included Asians/Pacific Islanders in showing worse reports and lower ratings of care by members of this minority group. Collectively, these studies suggest that Asians/Pacific Islanders are
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
137
Joint Commission
Journal on Quality and Safety
Table 2. Sample Characteristics by Race/Ethnicity, 1998 National Research Corporation Healthcare Market Guide® Survey*
* Percentages do not always add up to 100.0 because of rounding. POS, point of service; PPO, preferred provider organization; HMO, health maintenance organization.
138
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 3. Global and Composite Satisfaction Measures by Race/Ethnicity: Means and Standard Deviations (N = 120,855)
continued
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
139
Joint Commission
Journal on Quality and Safety
Table 3. Global and Composite Satisfaction Measures by Race/Ethnicity: Means and Standard Deviations (N = 120,855) (continued)
140
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 4. Regression Results: Significant (p < .05) Standardized Beta Coefficients and Robust Standard Errors*
continued
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
141
Joint Commission
Journal on Quality and Safety
Table 4. Regression Results: Significant (p < .05) Standardized Beta Coefficients and Robust Standard Errors* (continued)
142
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
Table 4. Regression Results: Significant (p < .05) Standardized Beta Coefficients and Robust Standard Errors* (continued)
* POS, point of service; PPO, preferred provider organization; HMO, health maintenance organization.
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
143
Joint Commission
Journal on Quality and Safety
considerably less satisfied with their medical care than all other racial and ethnic groups in the United States. More research is needed to understand the underlying reasons for this disparity. Available evidence, however, suggests that researchers should evaluate potential response bias, language barriers, and Asian/Pacific Islander subgroups (for example, at the very least, disaggregating Asians and Pacific Islanders, as MurrayGarcia, et al suggest14). Among the limitations of the study is sample representativeness. Although the overall response rate of 65% was acceptable, we are not certain that the 163,641 respondents were a good representation of the 250,000 survey recipients. Moreover, we were able to use only a subset of 120,855 respondents (74% of all respondents and 48% of those mailed surveys) who answered two of the four global satisfaction questions and at least half of the items on each of the four composite satisfaction scales. Although the loss of cases due to missing data appears not to have unduly affected sample composition (as shown in Table 1), one, nevertheless, must be cautious about generalizing these findings. Next, all questionnaires were administered in English, which may have inflated ratings within Asian/Pacific Islander and Hispanic subgroups by excluding potential survey respondents with limited Englishlanguage ability. Also, as Murray-Garcia et al14 have noted, systematic differences in the ways in which racial ethnic minorities respond to surveys may not be inconsequential. African American and Hispanic respondents, for example, may more often than whites choose extreme response options; those of Asian descent, however, are more likely to choose response categories in the middle of such scales. Another limitation is that the study explored patients’ experiences with care, not the technical quality of care; thus, answers to survey questions were based on subjective judgments. Finally, insofar as this was a study of adults with health insurance, ratings were likely to have been higher than would be obtained among the uninsured. For example, access to care has been found, as expected, to be worse for those without health insurance than for those with insurance.15 Given that racial and ethnic minorities’ health plan and medical care ratings tend to be high—in both the present study and the Morales et al2 study—it is tempting to conclude that disparities in health care are inconsequential.
144
These ratings may, in fact, reflect solid health status gains1; on the other hand, gaps in coverage and care persist,1,16 and the majority of Americans are unaware of these disparities.16 To reach the Department of Health and Human Services goal of eliminating racial and ethnic disparities in health care by the year 2010,17 public awareness of these disparities needs to increase, and programs to improve care, developed and implemented. Available data suggest that access to care and quality of care improvement efforts should be directed at all ethnic minority groups, particularly for limited-Englishspeaking, other/multiracial, Hispanic, and Asian/Pacific Islander subgroups. Efforts to improve access to care may require extended office hours, providing transportation to and from physicians’ offices, and employing bilingual office staff to answer phones and make appointments. Efforts to improve the quality of medical care may require increasing the diversity of provider staff at hospitals and clinics, providing professional interpreters for medical encounters, and providing well-translated patient education materials in multiple languages. Research is needed to better characterize individuals within Asian and other subgroups. Intragroup differences may be at least as important as the intergroup differences described in this article. Unfortunately, race/ethnicity data are rarely collected at a sufficiently fine level of specificity—and that was the case with these NRC HCMG® data. Further surveys need to be designed to capture this information. J Leo S. Morales was supported by a Robert Wood Johnson Minority Medical Faculty Development Program Fellowship. We thank Michael D. Hays of the National Research Corporation for providing access to the data.
Mark G. Haviland, PhD, is Professor, Department of Psychiatry, Loma Linda University School of Medicine, Loma Linda, California. Leo S. Morales, MD, PhD, is Assistant Professor, Division of General Internal Medicine and Health Services Research, David Geffen School of Medicine at University of California at Los Angeles (UCLA). Steven P. Reise, PhD, is Professor, Department of Psychology, UCLA. Ron D. Hays, PhD, is Professor, Division of General Internal Medicine and Health Services Research, David Geffen School of Medicine at UCLA. Please address correspondence to Mark G. Haviland, PhD, [email protected].
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
Joint Commission
Journal on Quality and Safety
References 1. Keppel KG, Pearcy JN, Wagener DK: Trends in Racial and EthnicSpecific Rates for the Health Status Indicators: United States, 1990–98. Healthy People Statistical Notes, No. 23. Hyattsville, MD: National Center for Health Statistics, 2000. 2. Morales LS, et al: Differences in CAHPS adult survey reports and ratings by race and ethnicity: An analysis of the National CAHPS Benchmarking Data 1.0. Health Serv Res 36:595–617, 2001. 3. Weech-Maldonado R, et al: Racial and ethnic differences in parents’ assessments of pediatric care in Medicaid managed care. Health Serv Res 36:575–594, 2001. 4. Snyder R, et al: Access to medical care reported by Asians and Pacific Islanders in a west coast physician group association. Med Care Res Rev 57:196–215, 2000. 5. Meredith LS, Sui AL: Variation and quality of self-report health data: Asians and Pacific Islanders compared with other ethnic groups. Med Care 33:1120–1131, 1995. 6. Fiscella K, et al: Disparities in health care by race, ethnicity, and language among the insured: Findings from a national sample. Med Care 40:52–59, 2002. 7. Morales LS, et al: Are Latinos less satisfied with communication from health care providers? J Gen Intern Med 14:409–417, 1999. 8. Institute of Medicine: Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC: National Academies Press, 2002.
9. National Quality Forum: Improving Healthcare Quality for Minority Patients. National Forum for Health Care Quality Measurement and Reporting, Washington, DC: 2002. 10. National Healthcare Disparities Report. Fact Sheet. AHRQ Publication No. 03-P007. Agency for Healthcare Research and Quality, Rockville, MD, Oct 2002. (www.ahrq.gov/news/nhdrfact.htm; last accessed Jan 24, 2003). 11. National Research Corporation: NRC Healthcare Market Guide Survey. Lincoln, NE, 1998. 12. Kippen LS, Strasser S, Joshi M: Improving the quality of the NCQA (National Committee for Quality Assurance) Annual Member Health Care Survey, Version 1.0. Am J Manag Care 3:719–730, 1997. 13. Cronbach LJ: Coefficient alpha and the internal structure of tests. Psychometrika 16:297–334, 1951. 14. Murray-Garcia JL, et al: Racial and ethnic differences in a patient survey: Patients’ values, ratings, and reports regarding physician primary care performance in a large health maintenance organization. Med Care 38:300–310, 2000. 15. Weissman JS, Epstein AM: The insurance gap: Does it make a difference? Annu Rev Public Health 14:243–270, 1993. 16. Lillie-Blanton M, et al: Race, ethnicity, and the health care system: Public perceptions and experiences. Med Care Res Rev 57:218–235, 2000. 17. U.S. Deparment of Health and Human and Services: Healthy People 2010. Washington, DC, 2000. (www.healthypeople.gov/publications; last accessed Jan 24, 2003).
March 2003 Volume 29 Number 3 Copyright 2003 Joint Commission on Accreditation of Healthcare Organizations
145