Racial and Ethnic Disparities in Health Care for Adolescents: A Systematic Review of the Literature

Objective  To identify the extent of racial and ethnic disparities in primary care, mental health care, reproductive health care, and asthma care for adolescents independent of socioeconomic status (SES).

Data Sources  Systematic review of the scientific literature using standard bibliographic databases.

Study Selection  Inclusion criteria were (1) studies published in the past 12 years, (2) analyses included children and adolescents aged 17 years and younger, and (3) data analyzed by racial/ethnic groups while accounting for SES. A total of 203 studies were reviewed, of which 31 met the criteria for inclusion: 14 of 65 studies on primary care, 11 of 61 studies on mental health care, 2 of 50 studies on reproductive health, and 4 of 27 studies on asthma services.

Data Extraction  Data from tables in the selected studies were used to determine whether minority children and adolescents received fewer, greater, or the same health care services as white children and youth after taking into account SES.

Data Synthesis  Black youth received fewer primary care services in 8 studies, whereas in 4 studies no disparity was noted. Hispanic youth received fewer primary care services in 6 studies, whereas no disparity was noted in 5. One study did not include Hispanic subjects. In 2 studies minority youth, combined into a single category, received fewer services than did white youth. In a total of 6 studies black youth received fewer mental health services, whereas in 3 studies no disparity was noted and in 1 study black youth received a greater number of services. In 3 studies Hispanic youth received fewer mental health services, and in 3 studies there were no group differences. In 1 study, with racial and ethnic groups combined in a single category, minority children and youth received fewer mental health services than white subjects. Three studies did not include Hispanic subjects. Too few studies of reproductive and asthma care were available to draw conclusions.

Conclusions  These results suggest that racial and ethnic disparities, independent of SES, exist in selected areas of adolescent health care. More studies are needed to better understand the extent and causes of these findings.

RACIAL AND ETHNIC disparities in health care have been well described in the scientific literature and are the focus of various national initiatives. An analysis of these studies provided the foundation for a recent report from the Institute of Medicine (IOM).¹ The IOM report was one of the first major efforts to investigate health care disparities in relation to race and ethnicity, access to care, and socioeconomic status (SES). As defined by the IOM, disparities in health care are " . . . racial or ethnic differences in the quality of health care that are not due to access-related factors or clinical needs, preferences, and appropriateness of intervention."^1(p32) One of the report's key findings is that "Racial and ethnic disparities in health care exist. These disparities are consistent and extensive across a range of medical conditions and health care services, are associated with worse health outcomes, and occur independently of insurance status, income, and education. . . ."^1(p79) The IOM hypothesized that a constellation of health system, physician, and patient factors contribute to these findings. However, the report dealt almost exclusively with data from adult patients.¹ It did not address whether disparities in health care, irrespective of SES and access to care, exist for minority adolescents.

Scattered across the medical disciplines, there are published reports on differences among adolescents in the prevalence of risk factors and diseases according to health insurance status, SES, and utilization of health services.^2-9 Few reports, however, include analyses that identify the interactive effects of race and ethnicity, SES, and health insurance status on health-related factors or services. Thus, to extend the work begun by the IOM to the adolescent population, a systematic review was conducted to better understand the role that race and ethnicity, independent of SES and access to care, have on medical services for adolescents.

Comprehensive literature searches identified scientific articles on adolescents' utilization of primary care and ambulatory health care services for 3 diseases and conditions: mental health, reproductive health, and asthma. These 4 types of services were chosen because they constitute common reasons for adolescents to seek health care and because they represent a range of issues, from traditional medical problems (ie, asthma) to behavioral issues (ie, reproductive health), emotional issues, and routine care. In addition, key informant interviews were conducted with experts in health services research, adolescent health, and the 3 thematic areas relevant to the current study. Experts identified unpublished studies or confirmed that there was a lack of data for a particular area. Reference lists of articles reviewed were also used to identify additional sources of data.

The standard scientific search engines MEDLINE, ERIC, Psych Info, and PubMed were used to identify published articles in each of the areas of interest. Various combinations of keywords (eg, racial and ethnic disparities, health care services, reproductive health, family planning, Pap test, HIV, STD [sexually transmitted disease], prenatal care, asthma, mental health services, depression, attention deficit disorder, and adolescents) were used to maximize the results. The following criteria were used to select articles to review: (1) used data from the national, regional, state, or school district level, (2) analyzed data on children and adolescents aged 17 years and younger, (3) analyzed subjects by racial/ethnic groups, and (4) controlled in the analysis for health insurance status or accounted for SES by either using a homogeneous sample (eg, Medicaid population or Job Corps students) or by controlling for parental income and/or education. Finally, to ensure that data were relatively recent, the final search criteria was that articles must have been published no earlier than 1990.

Sex was not included as a factor in the systematic review because it was used only variably in the identified studies. Because relatively few study populations included sizable numbers of minority subjects other than African Americans and Hispanics, only results from these 2 minority groups are presented. The pertinent data from each study were organized in a 2 × 3 table (not shown) according to racial and ethnic groups (ie, African American and Hispanic) and results (ie, minority group received fewer health care services than non-Hispanic white youth, minority group received equal services, or minority group received greater services). Lastly, a distinction was made as to whether the sample included just youth aged 12 to 17 years ("teenagers") or all children and youth.

A research assistant (J.J.) conducted the literature searches, identified articles that met the inclusion criteria, extracted the relevant data, and created the summary tables. The senior author (A.E.) reviewed all articles from which data were used in the analysis and validated the accuracy of the information in the tables.

A total of 203 studies were identified, of which 31 met the criteria for inclusion in the analysis and were reviewed. Of the studies reviewed, 16 included analyses of data on adolescents who were of either middle- or high-school age, and the remainder did not distinguish children from adolescents. In 3 studies researchers combined data from all minority groups. No study focused exclusively on health disparities. Rather, the foci of these studies were utilization of either primary care or health care services for 1 of the 3 areas of interest.

A total of 65 articles were reviewed, of which 14 met the criteria for inclusion in the analysis (Table 1).^10-23 Each of these studies reported on at least 1 conceptually related aspect of primary care, including having had a "well" or health care visit, frequency of health care visits, and having had a usual source of care other than the emergency department.

Six studies included data from only adolescent subjects.^10-15 In 3 studies African American (black) adolescents received fewer primary care services during the preceding year than non-Hispanic white adolescents,^10,12,13 whereas in 3 studies they received equal care.^11,14,15 Including the additional 6 studies of all subjects younger than 18 years increased the number of studies that found that black subjects received less primary care than white subjects by 5 to a total of 8.^17-20,22 The number of studies that found no disparity in primary care between white and black subjects increased by 1 study¹⁶ to a total of 4. Hispanic youth received fewer primary care services than did white youth in 3 studies^10,12,15 and equal services in 3 studies.^11,13,14 When articles that included both children and adolescents were added, the number of studies that found that Hispanic subjects received fewer services increased by 3 studies^17,18,22 to a total of 6, whereas the number indicating equal primary care services increased by 2 studies^16,20 to a total of 5. There were no studies in which either black or Hispanic subjects received more primary care than white subjects.

Researchers did not distinguish among minority groups in 2 studies. In each, nonwhite children younger than 18 years were less likely than white children to have had a usual or regular source of medical care.^21,23

The studies on primary care were mostly strong methodologically. All but 2 of 14 studies used national data,^15,19 and all but 6 controlled in the analysis for service need.^{13,15,17,19,21,22} Levels of significance could not be determined from 1 study.¹² Of the remaining 7 studies, disparities in primary care were found in 4.^10,18,20,23

A total of 60 articles were reviewed, of which 11 met our criteria for inclusion in the analysis (Table 2).^24-34 Six studies reported on mental health care service utilization among white and black adolescents. In 3 of these studies, black adolescents received fewer mental health care services than did white adolescents,^26,27,30 whereas no racial group differences were noted in 2 studies^25,28 and in 1 study black youth received more services than did white youth.²⁴ When the 4 additional studies that included subjects 17 years and younger were included, the number of studies in which black youth received fewer health services than white youth increased from 3 to 6,^32-34 whereas the number reporting no racial disparity increased from 2 to 3.³¹

Of the 11 studies on adolescents, only 4 reported data on Hispanic youth. The results of 1 study²⁷ indicated that Hispanic adolescents received fewer mental health care services than did white adolescents, whereas there were no group differences in the other 3 studies.^24,28,30 When studies from all children and youth were included, the number in which disparities in mental health care services were found increased from 1 to 3,^32,33 whereas the number of studies that found no differences between white and Hispanic children increased from 2 to 3.³⁴ In 1 study on adolescent subjects, minority racial/ethnic groups were combined and found to receive fewer mental health services than white youth.²⁹

There was a fairly large variation in the quality of studies on mental health services. Six studies involved either national data sets or large, representative regional populations.^{27,29,30,32-34} Of these studies, 1 did not use multivariate analysis,²⁷ and 2 did not control or account for service need.^27,29 Significant disparities in mental health services were present in the remaining 4 studies.

Fifty studies on reproductive health services were reviewed. Only 2 studies met the inclusion criteria, and both used adolescent subjects.^35,36 Porter and Ku,³⁵ using data from the 1995 National Survey of Adolescent Males, found, after adjusting for age, health insurance status, SES, and service need, that black 15- to 19-year-olds were significantly more likely than white youth both to discuss reproductive health issues with their provider (adjusted odds ratio, 1.45; P<.05) and to have a test for a sexually transmitted disease (adjusted odds ratio, 3.65; P = .05). No group differences were found for Hispanic youth. In the other study, Schuster et al³⁶ analyzed data from 9th- to 12th-grade students from a California school district to determine the extent to which adolescents talk with a physician about sexual issues. After controlling for sex, grade, race/ethnicity, primary language spoken at home, history of sexual intercourse, and parents' education, the researchers found that Hispanic youth (adjusted odds ratio, 1.44; 95% confidence interval, 10.5-1.98), but not black youth, were more likely than white youth to report having discussed sexual behavior with their doctor.

A total of 27 studies on asthma care were identified. Only 4 studies fit the inclusion criteria for review (Table 3).^37-40 Results from 1 study on adolescent subjects indicated that black youth were significantly less likely than white youth to receive preventive asthma medication.³⁷ Hispanic youth were not included in the data set. The remaining 3 studies used subjects younger than 18 years and produced mixed results. Racial/ethnic group differences in use of preventive medications were found in 1 study³⁸ but not a second study³⁹; emergency department use for asthma was greater among black children in 1 study³⁹; in the same study, black children were less likely than were white children to have an office visit for asthma,³⁹ whereas no group difference was noted in 2 other studies.^38,40

As described in the recent IOM report, racial and ethnic disparities in health care for adults persist after accounting for access to health care and SES.¹ Our results suggest that a similar disparity might also exist for adolescents. A systematic review of the scientific literature yielded 4 findings.

First, a racial and ethnic disparity in utilization of primary care is noted when studies of adolescents and all children and youth younger than 18 years are taken together. The findings were more consistent for black than for Hispanic youth. Most studies on primary care used national data sets and carefully controlled for both family SES and health insurance status, thus adding to the strength of the findings.

Second, although the data are not as strong as with primary care, when all studies are considered, there is also an indication of racial and ethnic disparities in adolescents' use of mental health care services. Again, the data are more consistent for black than for Hispanic youth. Although these results were somewhat less generalizable than the results for primary care, disparities were found in each of the 4 most rigorous studies. The importance of any disparities for mental health care services is significant in light of results from community-based clinical epidemiological studies on mental health diagnoses among adolescents. Data from several large studies indicate that, after controlling for SES, there are no differences in the prevalence of depressive symptoms or affective disorders among black, Hispanic, and white youth.^41-44 Thus, disparities in mental health care services probably cannot be explained by variations in prevalence of mental health disorders.

Third, although too few studies are available to draw a conclusion, the results suggest that minority youth might actually receive more reproductive health services than white youth. The small number of studies on reproductive health services that controlled for SES and health insurance status along with race/ ethnicity was disappointing. As arguably the most highly publicized and prominent cause of adolescent morbidity, it was surprising that there were not more studies in which researchers tried to identify the associations among SES, access to care, and race/ethnicity and services for sexually transmitted disease, Papanicolaou smear testing, and contraception. Instead, data in most studies were reported by race/ethnicity and by SES, leaving unanswered the question of independent contributions made by either variable to health care services. As with reproductive health services, there were too few studies on adolescents' use of health care services for asthma to draw meaningful conclusions.

Finally, there is little question that an adolescent's SES has a direct impact on health status. The results of studies by several researchers clearly demonstrate the inverse relationship between SES and involvement in chronic disease risk behaviors^4,6-8 and the overall perception of health status⁴⁵ after controlling for race and ethnicity. Data from some of these and other studies also demonstrate that even after controlling for SES, race and ethnicity remain a strong predictor of adolescent health behavior.^3-5,7 The mechanisms through which SES and race and ethnicity affect adolescent health are not understood and are probably quite complex. Some of these factors, such as the context of one's neighborhood, nutrition, stress, and social support networks, are outside the health care system. Better understood is the relationship between access to care and disparities in utilization of health services. For example, Newacheck et al⁴⁶ found in a large national sample of adolescents that SES and race/ethnicity were both associated with health insurance status. Health insurance status, in turn, was strongly associated with utilization of health care services. However, our results are consistent with many studies on adults, and the findings of the IOM report: access to care explains some, but not all, of the racial and ethnic disparities in health care.¹

The IOM speculates that variations of health care services might be related to a combination of provider and patient factors and/or interaction among factors within the health care system.¹ Some factors, such as type and location of the clinic setting, general mistrust of the health care system, and concerns regarding confidentiality of the visit, deter minority youth from seeking care. Most factors suggested by the IOM, however, influence care once patients are in clinical settings. Health care system–level factors include fragmentation of services, the fact that minorities might be disproportionately enrolled in lower-cost health care programs that provide fewer services, de-selection of minority physicians by some health plans, the failure to provide a culturally diverse workforce, and the lack of linguistically diverse clinic signs and materials. Patient-level factors include cultural beliefs that lead people to use ineffective and unproved remedies and low health literacy that influences understanding of recommended management plans. Physician-level factors relate to stereotypic beliefs of minority patients, and adolescents, that might bias diagnostic and management strategies. This becomes especially problematic when physicians are faced with diagnostic or therapeutic uncertainty and have limited time to confirm the accuracy of information as it pertains to individual patients.

A review of this magnitude has several limitations. Foremost is the problem of ensuring identification of all relevant scientific articles. Although articles were identified through a systematic review of the literature and informant interviews, some key studies may have been missed. However, it is unlikely that enough were missed to obscure a significant trend in results. A second limitation is the lack of statistical analyses to determine the significance of the differences found. Although a meta-analysis of the data might have provided statistical results, the results could be misleading given the small number of studies and the wide variability in sample size, the quality of data collected, and analytical methods. Instead, this review should be viewed as exploratory, rather than explanatory. A third limitation is the small number of studies using adolescent subjects. Findings of disparities were often only noted when studies on children and youth were added to those on adolescents. Each of these studies included sizable numbers of adolescent subjects; nevertheless, it was not possible to determine the extent to which the findings were dependent on disparities among young children, rather than on disparities among adolescents. Lastly, a final limitation to the analysis is that the studies chosen for review and the data abstracted from the articles were only reviewed by the research assistant and the senior author. Having someone outside of the data gathering process review the findings might have reduced any unintended bias.

The results of this review raise important questions about the magnitude of racial and ethnic disparities in adolescent health care. The limited number of studies in which race and ethnicity are examined, independent of SES and health insurance status, are too few to draw definitive conclusions. Instead, the results provide a direction for further research. Key national data sets, such as the National Survey of Family Growth, the National Health Interview Survey, and the National Longitudinal Survey of Adolescent Health, contain valuable information that could be used to analyze the relative importance of SES, access to care, service need, and race/ethnicity on adolescent health care. More data will also be needed on causes of the disparity and strategies for change.

Article

Corresponding author and reprints: Arthur Elster, MD, Medicine and Public Health, American Medical Association, 515 N State St, Chicago, IL 60610 (e-mail: arthur_elster@ama-assn.org).

Accepted for publication April 4, 2003.

Attention is being directed at understanding the magnitude and causes of racial and ethnic disparities in health care. Studies on adult populations strongly suggest that disparities exist even after controlling for the effects of SES and access to care. These findings have been noted for a range of medical disorders.

The current review is the first comprehensive study of racial and ethnic disparities in health care for adolescents. The findings suggest that more attention should be directed at understanding disparities in this age group. Although ensuring access to care for adolescents is important, this should not cloud the need to identify and implement strategies for reducing factors other than access that cause disparities among minority youth.