|Home | About | Journals | Submit | Contact Us | Français|
In the mental health arena, unlike general health, health care disparities predominate over disparities in mental health per se. Strategies to improve health care in general, such as improving access to care and improving the quality of care, would do much to eliminate mental health care disparities. However, a diverse mental health workforce, as well as provider and patient education, are important to eliminating mental health care disparities.
Mental health care disparities , defined as unfair differences in access to or quality of care according to race and ethnicity, are quite common in mental health.1 Although some studies question this consensus, 2, 3 the weight of the evidence supports the existence of serious and persistent mental health care disparities.
The purpose of this paper is to review briefly the evidence for disparities in mental health care, highlighting differences from the general health care literature. We orient our discussion around two questions relevant for public health care policy to reduce disparities:
Answers to these questions can guide steps needed to reduce disparities in mental health care. The degree to which policies need to be tailored to disparities in mental health care, or whether general health care policies can be counted upon to work is addressed.
A consensus about what constitutes a “disparity” has not been reached despite a voluminous literature on the topic. The term disparity clearly connotes an unfair difference, but measurement of this difference is far from uniform. Here, we rely on the definition employed by the Institute of Medicine (IOM) in its Unequal Treatment 1 report: a disparity is a difference in health care quality not due to differences in health care needs or preferences of the patient. As such, disparities can be rooted in inequalities in access to good providers, differences in insurance coverage, as well as stemming from discrimination by professionals in the clinical encounter.
The IOM definition is distinct from that applied by the Agency for Healthcare Quality and Research (AHRQ) in its annual National Healthcare Disparities Reports, where any difference between populations is a disparity, with no adjustment for underlying need for care. The IOM definition is also distinct from much of the research literature which adjusts disparity estimates for socio-economic and geographic variables, thus disregarding disparities associated with lack of insurance coverage, geographic access to providers, education, or income that are considered disparities within the IOM concept.4 We strongly believe that disparities should be identified only when there is need for care, and that social circumstances, such as lack of insurance, constitute disparities if they lead to poorer care when care is needed.
Disparities in mental health exhibit a decidedly different pattern from disparities in health. In general, minorities, particularly African Americans, have poorer health and health outcomes than do Whites.5 Many studies show that members of minority groups have either lower or equivalent rates of mental disorders as compared with whites. For example, both Hispanics and Blacks have lower lifetime risk of psychiatric disorders than do whites6, 7 in representative samples of the U.S. English-speaking population.
The four Consortium on Psychiatric Epidemiology Studies (CPES) studies funded by the National Institute of Mental Health fielded common core questions and unified sampling weights8 to permit comparisons across U.S. Black (African American, Caribbean) Hispanic (Puerto Rican, Cuban, Mexican, Other), and Asian (Chinese, Filipino, Vietnamese, Other) groups. These studies improve upon previous surveys because they include large enough samples to study important subgroups of the population, and they are available in Spanish and several Asian languages, as well as English. These studies yielded a surprising finding, given the higher rates of poverty among minority individuals as compared with whites in the U.S..9 With the exception of Puerto Ricans, all subgroups of minorities report lower rates of lifetime mental disorders than do White Americans. Similar advantages exist for presence of a disorder in the past year, although Latino and Black American rates are relatively close to that of white Americans10. Although not part of the CPES, the prevalence of mental disorders has also been studied in two American Indian reservation populations.11 Compared with a nationally representative sample of the U.S. population, Americans Indians were found to be at heightened risk for PTSD and alcohol dependence, but at lower risk for major depression.
Another exception may exist to the lower or equal overall prevalence of mental disorders among minorities. Blacks may have higher levels of schizophrenia, a low prevalence but very serious condition, than Whites. The results of three major community prevalence studies12–14 indicate a higher incidence of schizophrenia among African Americans than Whites, but have too few cases for stable estimates. In a prospective, birth cohort study, African Americans were found to have substantially elevated rates of schizophrenia when compared with Whites.15 Furthermore, Blacks with schizophrenia are overrepresented in state psychiatric hospitals.16 While substantial evidence exists to suggest that clinicians over diagnose schizophrenia and under diagnose mood disorders in African Americans15, 17, clinical decisions do not account for the findings above based on semi-structured survey instruments.
Rates of disorders may be inadequate when considering disparities. Hispanics and Blacks report lower risk of having a psychiatric disorder as compared with their white counterparts, but those who become ill tend to have more persistent disorders.7 Similarly, Black Americans were found to have lower rates of lifetime major depression than were their White counterparts living in similar areas, but the rates of major depression in the past year were similar across groups, indicating more persistent illness. Relative to Whites, Blacks were more likely to rate their depression as very severe and disabling.18 Although rates of disorders are not higher among minorities, psychological symptoms do tend to be higher among minorities and among the poor.19 These higher symptom levels may be important because poorer functioning has been related to subthreshhold symptoms.20 For example, maternal depressive symptoms have been related to poorer mental health and functioning in offspring.21, 22 Higher rates of symptoms among poor and minority women23, mean these women may need services to improve their functioning and subsequent outcomes of their children.
In summary, few disparities in mental health diagnoses exist. Some evidence suggests that Black Americans may have higher rates of a particularly debilitating disorder, schizophrenia, although diagnostic problems may in part account for this difference. Furthermore, minorities do have higher rates of psychological symptoms that may impair their functioning.
Turning now to the evidence on disparities in mental health care, most research comparing mental health care across groups finds evidence of disparities in access and use. As documented in “Mental Health: A Report of the Surgeon General”24 and its supplement, “Mental Health, Culture, Race and Ethnicity”19, racial and ethnic minorities have less access to mental health services than do whites, are less likely to receive needed care and are more likely to receive poor quality care when treated. Minorities in the United States are more likely than whites to delay or fail to seek mental health treatment.25–27 After entering care, minority patients are less likely than Whites to receive the best available treatments for depression and anxiety.28, 29 African Americans are more likely than Whites to terminate treatment prematurely.30 Among adults with diagnosis-based need for mental health or substance abuse care, 37.6% of Whites, but only 22.4% of Latinos and 25.0% of African Americans, receive treatment.31 This comparison is consistent with the IOM definition of disparities based on need and not controlling for socioeconomic and health system factors when comparing rates among the groups. McGuire and colleagues32 implement the IOM definition of disparities in outpatient mental health care and find overall spending for Blacks and Latinos on outpatient mental health care is about 60 percent and 75 percent of white rates, respectively, after taking into account need for care.
Some studies have begun to track trends in mental health care disparities. Three use the IOM definition of disparities. Using a national data set (MEPS), Black-White and Hispanic-White disparities in rates of any mental health care use worsened from 2000/1 to 2003/4.33 Using another nationally representative sample of service use (NAMCS), no evidence was found for progress against disparities in depression and anxiety care in either primary care or psychiatric settings over the past decade.34 Similar findings from a nationally representative sample of English-speaking individuals35, overall rates of treatment for psychiatric disorders increased between 1990 and 2003, in both years, but Blacks were only 50 percent as like to receive psychiatric treatment as whites for diseases of similar severity.
In summary, disparities exist in access to and use of mental health services for ethnic minority individuals. Furthermore, during the past decade, efforts to eliminate these disparities have not been successful in primary care or specialty psychiatric services.
Mechanisms behind disparities in health and mental health care show some differences. A broad distinction, introduced by the IOM1 is between disparities due to discriminatory behavior of providers (i.e., treating otherwise similar patients differently according to race/ethnicity) and disparities due to access, insurance, and other factors associated with the operation of the health care system. In the health care area, both geographic and provider level differences are major sources of disparity3. Provider A may be low quality for all patients and Provider B high quality for all patients, and if minorities are more likely to be seen by Provider A, these across-provider differences will account for some of disparities. Geographic-level factors can work similarly.
As in health care, mental health care disparities associated with access in general, and lack of insurance, are significant in minority communities36, but inadequate access in a poor rural community may largely be shared by everyone living there.
Provider discrimination, including bias and stereotyping on the part of providers is another source of disparities.1, 37 However, no research that we know of to date makes the empirical link between the stereotypic belief or bias and actual clinical discrimination. One recent study comes close by using innovative methods to measure stereotypes and links this to physician recommendations based on case vignettes randomized by race, but the link to actual practice remains speculative.
Provider discrimination recognizes that physicians work with another type of belief, a “prior” about the likelihood a patient has a condition, and update this prior according to the strength of information received in a clinical encounter. Even when physicians are “rational” and hold no ill will or stereotypes, different underlying assumptions about the distribution of disease or communication problems can lead to discrimination. This provider discrimination has been documented in two studies of mental health care38 in which clinicians respond with less alacrity to variation in severity of depression among minority patients than whites, implying that clinicians are less able to “read” severity among minorities.
The concept of provider discrimination ties together a number of the salient features of disparities in mental health care as compared to health care. The prevalence of mental disorders is generally lower among minorities so that a clinician’s “prior” encountering a minority patient should be that they are less likely to be a “case” in comparison to an otherwise similar white patient. If so, a more serious indication of symptoms would be necessary to cause a clinician to revise the prior enough to justify recommending treatment. In health, where minorities may on average be worse off than whites, application of population priors will tend to favor rates of treatment for minorities. Our hypothesis here, consistent with some of the fact pattern of disparities in health and mental health, is that disparities arising within the clinical encounter are more important in mental health than in health. In the case of mental disorder, where population prevalence is generally lower for minorities, and where communication/understanding may be worse, this type of provider discrimination leads to lower rates of treatment for minorities.39
Discussion of mental health policy often centers on the issue of mental health “exceptionalism:” Is mental health different from the rest of health care in some way that justifies special policy? In mental health economics, for example, higher demand response for mental health services once supported arguments for less insurance coverage. Similarly, the inadequacy of case mix adjustments for hospital care to capture expected resource use argued against use of diagnostic-related group based payment for hospital care for mental illnesses.40
In the case of mental health care disparities, appeal to exceptionalism does not seem necessary to reduce mental health care disparities. The major recommendations for eliminating health care disparities from the IOM Unequal Treatment report appear applicable to mental health care. Specifically, the report recommends that health care systems should take steps to improve access to care, ensuring that they do not disproportionately burden or restrict minority patients’ access, as well taking further steps to improve access when necessary, such as providing interpreter services. Further, they recommend that economic incentives should be considered for improving patient-provider communication and trust, as well as rewarding appropriate screening, preventive, and evidence-based clinical care. If followed, these general policies would likely improve disparities.
Increasing the proportion of racial minority providers is considered an important factor for improving health disparities. This is even more important for mental health care where ethnic minorities are even more poorly represented than in health care in general, and where diversity may make more of a difference in addressing minority patients’ concerns about trust. A more diverse workforce would likely provide not only more culturally appropriate treatment, but language skills to match those of patients. A federal commitment to the outreach and educational support necessary to build a truly diverse mental health workforce is a critical policy recommendation for decreasing disparities in mental health care.
A promising direction in mental health care picks up on our hypothesis about the relative importance of understanding of the circumstances and symptom reports from minority patients (referred to as “priors”). Mental health care is somewhat different from some of the rest of health care in its heavier reliance on understanding and communication to determine patient needs.41 Culturally appropriate education for providers is important, as well as education to patients for better understanding disease and disease management. For example, ethnic minority women are less likely to perceive a need for depression care than are their similarly poor white counterparts.42 Clearly, routine screening for depression in healthcare settings, as well as educating providers about ethnic minority patients, and educating patients about mental illness and interventions, could help to reduce disparities.
Although quality improvement strategies and steps to reduce racial/ethnic disparities do not need to be seen as alternatives, an issue of relative emphasis exists: Should quality improvement efforts be focused on low quality or should care for minorities get special attention? One argument is that the overall gaps in quality are so large that they dwarf differences among groups. Baicker and Chandra43 argue “Policies should focus on getting the rates right, rather than solely on racial differences.”
Low quality healthcare and racial/ethnic healthcare disparities are indicators of failure to attain social and healthcare objectives, in the first case for all patients and in the second case for a subgroup of patients. According to the IOM, equity is one of the 6 domains of quality of healthcare. Both low quality and disparities can be seen as manifestations of the same underlying issue. However, recent empirical research suggests that the quality of health care and disparities in health care are not invariably related. In Medicare, health plan outcomes and the magnitude of disparities are not related.44 Asch et al.2 found no racial disparity in care among a population of patients with access to care in twelve communities, despite the average low quality for patients in this study.
Should health care providers pursue one set of interventions that will address both quality and disparities or two sets of interventions--one set addressing quality and a second set addressing disparities? Two studies have looked at interventions designed to improve quality of mental health care and minority outcomes. In a large trial of quality improvement for depression in older patients, a collaborative care intervention improved care significantly more than care as usual for African American, Hispanic, and white patients similarly.45 A similar study of two quality improvement interventions for depression (psychotherapy and medications) in large managed care medical settings found that clinical outcomes at one year were greater for Latinos and African Americans than for white patients.46 Five years later, the interventions were found to improve disparities markedly by improving health outcomes and unmet need for care among Latinos and African Americans relative to whites.47
Taken together, these studies suggest that improving the quality of medical care in general is likely to benefit minority patients who have access to care. General improvement may also help to decrease disparities. In the one study showing such an effect, efforts were made to make the quality improvement interventions appropriate for ethnic minorities. Specifically, experts in mental health interventions for minorities participated in designing all of the quality improvement materials in English and Spanish. Videotaped educational materials for patients included African American and Hispanic providers and patients. Furthermore, information on cultural beliefs and ways of overcoming barriers to care for Latino and African American patients were included in all training. Finally, the psychotherapy intervention had specifically been developed for use with low-income and minority patients. Overall, quality improvement interventions have the potential to decrease disparities.
Because mental health disparities are rooted almost entirely in mental healthcare disparities, policy efforts should focus on improving access to and quality of mental healthcare for diverse Americans. For the most part, these policy efforts do not reflect exceptionalism for mental healthcare, apart from general healthcare. Specifically, policies that would result in universal coverage for mental health care would significantly improve access for ethnic minorities. Similarly, improving the quality of mental health care treatments would likely improve, but not eliminate, mental health care disparities. These quality improvement efforts would need to include screening to increase appropriate identification of disorders for minorities, as well as modest accommodations for minorities (providing language appropriate educational and treatment materials, and culturally sensitive training for providers). Of particular importance for eliminating mental health disparities for minorities, federal policies should provide the outreach and education support necessary to train a diverse workforce to meet the mental health needs of our nation.
The authors are grateful to the MacArthur Foundation for support of this research. In addition, McGuire’s research was supported by grants P50 MH073469 from the National Institute of Mental Health and P20 MD000537 from the National Center on Minority Health and Health Disparities, and Miranda’s research by Resource Centers for Minority Aging Research/Center for Health Improvement of Minority Elderly (RCMAR/CHIME) funded by the National Institute on Aging (3P03AG021684); UCLA/Drew Project EXPORT funded by the National Center for Minority Health and Health Disparities (1P20MD00148-01); and the UCLA-RAND Center for Research on Quality in Managed Care funded by the National Institute of Mental Health (MH068639-01).
Thomas G. McGuire, Department of Health Care Policy, Harvard Medical School, 180 Longwood Ave., Boston, MA 02115.
Jeanne Miranda, Center for Health Services Research, Department of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles, 10920 Wilshire Blvd, Ste 300, Los Angeles, CA 90024.