dysthymic research
PSYCHIATRIC SERVICES 'ps.psychiatryonline.org 'November 2008 Vol. 59 No. 11 1 12 26 64 4 T o respond to the Public Health Service’s Healthy People 2010 initiative for minority popula- tions, investigators must first ascer- tain at a national level the magnitude of disparities in service utilization for depression. Despite recent advances in the treatment of mental illness and considerable efforts to improve quali- ty of and access to treatment (1), there appears to be a significant mis- match between need and treatment in the United States (2). There is con- troversy about disparities in quality of care (3) at a national level, because on that large overall scale there are few ethnic and racial disparities for some chronic conditions. Yet there is evi- dence of striking quality disparities across some groups for psychiatric conditions (4–6). Part of the discrep- ancy comes from differences in the ethnic and racial groups studied, whether studies are regional or na- tional, and whether the assessment of need for depression care used diag- nostic interviews versus screeners for depression.
Prior work on racial and ethnic dis- parities in depression treatment has been limited by the scarcity of nation- al samples that include a rich array of diagnostic and quality indicators and large numbers of non–English-speak- ing minority respondents. With this study we took advantage of a unique opportunity to estimate disparities in Disparity in Depression Treatment Among Racial and Ethnic Minority Populations in the United States M Ma ar rg ga ar ri it ta a A Al le eg gr rí ía a, , P Ph h.
.D D.
.
P Pi in nk ka a C Ch ha at tt te er rj ji i, , P Ph h.
.D D.
.
K Ke en nn ne et th h W We el ll ls s, , M M.
.D D.
., , M M.
.P P.
.H H.
.
Z Zh hu un n C Ca ao o, , P Ph h.
.D D.
.
C Ch hi ih h- -n na an n C Ch he en n, , P Ph h.
.D D.
.
D Da av vi id d T Ta ak ke eu uc ch hi i, , P Ph h.
.D D.
.
J Ja am me es s J Ja ac ck ks so on n, , P Ph h.
.D D.
.
X Xi ia ao o- -L Li i M Me en ng g, , P Ph h.
.D D.
. Dr. Alegría and Dr. Chen are affiliated with the Center for Multicultural Mental Health Research, Cambridge Health Alliance and Harvard Medical School, 120 Beacon St., 4th Floor, Somerville, MA 02143 (e-mail: [email protected]). Dr. Chatterji is with the Department of Economics, State University of New York at Albany. Dr. Wells is with the Health Services Research Center, University of California, Los Angeles. Dr. Cao is with Thomson Healthcare, Cambridge, Massachusetts. Dr. Takeuchi is with the School of Social Work, University of Washington, Seattle. Dr. Jackson is with the Institute for So- cial Research, University of Michigan, Ann Arbor. Dr. Meng is with the Department of Statistics, Harvard University, Cambridge, Massachusetts.
Objective: Prior research on racial and ethnic disparities in depression treatment has been limited by the scarcity of national samples that in- clude an array of diagnostic and quality indicators and substantial num- bers of non–English-speaking individuals from minority groups. Using nationally representative data for 8,762 persons, the authors evaluated differences in access to and quality of depression treatments between patients in racial-ethnic minority groups and non-Latino white pa- tients. Methods: Access to mental health care was assessed by past-year receipt of any mental health treatment. Adequate treatment for acute depression was defined as four or more specialty or general health provider visits in the past year plus antidepressant use for 30 days or more or eight or more specialty mental health provider visits lasting at least 30 minutes, with no antidepressant use. Results: For persons with past-year depressive disorder, 63.7% of Latinos, 68.7% of Asians, and 58.8% of African Americans, compared with 40.2% of non-Latino whites, did not access any past-year mental health treatment (signifi- cantly different at p<.001). Disparities in the likelihood of both having access to and receiving adequate care for depression were significant- ly different for Asians and African Americans in contrast to non-Latino whites. Conclusions: Simply relying on present health care systems without consideration of the unique barriers to quality care that ethnic and racial minority populations face is unlikely to affect the pattern of disparities observed. Populations reluctant to visit a clinic for depres- sion care may have correctly anticipated the limited quality of usual care. (Psychiatric Services59:1264–1272, 2008) access to and quality of depression care by using pooled data from the National Institute of Mental Health Collaborative Psychiatric Epidemiolo- gy Surveys (CPES) (7). The three studies that constitute the CPES in- clude the same measures of need and quality, include significant numbers of non–English-speaking persons be- longing to racial and ethnic minority groups, and are the most current and comprehensive available to study de- pression treatment for racial and eth- nic minority populations. Using a sys- tem cost-effectiveness framework (8, 9), we evaluated whether individuals who could benefit from depression treatment were either not treated or inadequately treated. By this frame- work, not treating people who could benefit from treatment is a missed op- portunity to improve health, and treating people who do not need care increases spending without commen- surate health effects.
Methods The CPES combined sample The University of Michigan Survey Research Center (SRC) collected data for the National Latino and Asian American Study (NLAAS), collected between 2003 and 2003 (10); the Na- tional Comorbidity Survey Replica- tion (NCS-R), collected between 2001 and 2003 (11); and the National Survey of American Life (NSAL), col- lected between 2001 and 2003 (12).
Together, these studies are known as the CPES, and they used an adapta- tion of a multiple-frame approach to estimation and inference for popula- tion characteristics (13,14). This al- lows integration of design-based analysis weights to combine data sets as though they were a single, national- ly representative study (7). Design and methodological information can be found at the CPES Web site (7).
Each study in CPES focused on collection of epidemiological data on mental disorders and service usage of the general population, with special emphasis on minority groups (15). In- terviews for the studies were con- ducted by professional interviewers from the SRC, with 92.5% of inter- views in English and 7.5% in other languages (Spanish, Mandarin, Can- tonese, Tagalog, and Vietnamese). Asdescribed in detail elsewhere (16), the NLAAS is a nationally representa- tive survey of household residents age 18 and older in the noninstitutional- ized Latino and Asian populations of the coterminous United States. The final sample included 2,554 Latinos and 2,095 Asian Americans. The weighted response rates were 73.2% for the total sample, 75.5% for Lati- nos, and 65.6% for Asians (17).
The NCS-R is a nationally repre- sentative sample with a response rate of 70.9%. Eligible respondents were English-speaking, noninstitutional- ized adults age 18 or older living in civilian housing in the coterminous United States. The NCS-R was ad- ministered in two parts. Part 1 was administered to all English-speaking respondents and included core diag- nostic assessments. A subset of part 1 respondents also completed part 2 of the survey, which included additional batteries of questions addressing service use, consequences, and other correlates of psychiatric illness and additional disorders, with measures identical to those in the NLAAS.
The NSAL is a nationally represen- tative survey of household residents in the noninstitutionalized black popula- tion and included 3,570 African Amer- icans and 1,621 black respondents of Caribbean descent. The NSAL re- sponse rate was 70.9% for the African- American sample and 77.7% for the black Caribbean sample (18). Inter- views were done in English. We used a pooled sample (N=8,762) of data from Asians and Latinos from the NLAAS, non-Latino whites from the NCS-R part 2, and African Americans from the NSAL. Race and ethnicity cate- gories were based on respondents’ self-reports to questions based on U.S.
census categories. The institutional re- view boards of all participating institu- tions approved all study procedures.
Diagnostic assessment In the CPES, the presence of psychi- atric disorders over a person’s lifetime and in the past 12 months, and the presence of subthreshold depressive disorder or minor depressive disorder were evaluated via the World Mental Health survey initiative’s World Health Organization Composite In- ternational Diagnostic Interview(WMH-CIDI) (19). Diagnoses are based on DSM-IV diagnostic systems.
Findings showed good concordance between DSM-IVdiagnoses based on the WMH-CIDI and the Structured Clinical Interview for DSM-IV Axis I Disorders (20). Using the WMH- CIDI (19), we classified the pooled sample, who responded to the NLAAS, NCS-R part 2, or NSAL, into five groups: currently depressed respondents, who met criteria for a past-year diagnosis of major depres- sion or dysthymia (N=1,082); respon- dents with subthreshold symptoms, who did not meet criteria for a past- year diagnosis of major depression or dysthymia (N=158); lifetime de- pressed respondents, who met crite- ria for lifetime major depression or dysthymia but who did not meet cri- teria for past-year depression or dys- thymia (N=1,230); respondents who met past-year criteria for disorders other than depression (N=919); and the no-need group, respondents who did not meet past-year criteria for any psychiatric or substance use disorder assessed (N=7,680).
Our main sample for estimating dis- parities in access included the 8,762 respondents who belonged to the first and fifth groups—those with current depression (N=1,082) and the no-need group (N=7,680). In models of dispar- ities in the quality of depression treat- ment, our sample was further limited to 880 respondents who used services in the past year. Sensitivity analyses for access to and quality of depression care yielded an additional 158 persons with subthreshold symptoms, whom we considered in the sensitivity analyses as respondents with depression.
Role impairment and chronic medical conditions Functional impairment was measured by the World Health Organization Psy- chiatric Disability Assessment Sched- ule (21). For the domains of cognition, mobility, self-care, and social function- ing, we asked the number of days in the past 30 when health-related or mental health–related problems re- stricted ability to carry out related tasks. We measured the number of chronic medical conditions on the ba- sis of respondents’ endorsement of any of the following over their lifetime:
PSYCHIATRIC SERVICES 'ps.psychiatryonline.org 'November 2008 Vol. 59 No. 11 1 12 26 65 5 arthritis or rheumatism, an ulcer in the stomach or intestine, cancer, high blood pressure, diabetes or high blood sugar, heart attack, stroke, asthma, tu- berculosis, any other chronic lung dis- ease, HIV infection, or AIDS.
Access to and quality of depression treatment All CPES respondents were asked the same battery of questions about past- year mental health services and use of prescription medication (name of medication, length of use in the past year, and number of days that medica- tion was used in the past month) for problems related to their emotions, nerves, substance use, energy, concen- tration, sleep, or ability to cope with stress. To define access to mental health care, we assessed whether the respondent received any mental health treatment, defined as at least one visit to a specialty mental health provider or general medical provider for mental health care in the past year. General medical sector providers included gen- eral practitioners, family doctors, nurs- es, occupational therapists, or other health professionals providing care for a mental health problem. Specialty mental health sector providers includ- ed psychiatrists, psychologists, coun- selors, social workers, or other mental health professionals seen in a mental health setting. Although no validity or reliability data are available, these measures were adapted from measures used in the NCS (22,23) and were in- cluded as core measures in all of the CPES instruments.
To characterize quality of depres- sion treatment, we conceptually drew on the Institute of Medicine (24) defi- nition of quality of care: “the degree to which health services for individuals and populations increase the likeli- hood of desired health outcomes and are consistent with current profession- al knowledge” (25). Assessment of quality of depression care was derived from respondents’ reports of past-year service use (26). Quality of treatment for acute depression was defined ac- cording to Wang and colleagues’ (26) use of a binary variable, with a score of 1, indicating adequate treatment, as- signed if a patient had four or more specialty or general health provider visits in the past year plus antidepres-sant use for 30 days or more or if the patient had eight or more specialty mental health provider visits in the past year lasting at least 30 minutes.
This measure of quality has also been used extensively in other studies of health disparities (27–29). In sensitivi- ty analyses, we recognized that some respondents in different stages of the course of their illness may have been appropriately receiving maintenance care, so we considered an alternative, broader quality indicator of four or more mental health visits in the past year with any type of formal provider.
Statistical analyses We used a two-stage regression model; the first stage estimated correlates of access to any mental health treatment in the past year, and the second esti- mated correlates of quality depression treatment in the past year for those who had received any mental health care. We estimated a short specifica- tion of this model, with adjustments only for need and correlates of need classified in the literature on racial and ethnic disparities. Adjustments were made for age and sex (6,30), number of chronic conditions, and level of impair- ment. We also estimated an extended specification of the model, which added adjustments for marital status, education, insurance, poverty, and re- gion. The poverty measure was con- structed through an income-to-needs ratio according to the definition pro- vided by the U.S. Census Bureau (31).
When household income was less than family needs (determined by family size and household income specified in the Census definition), a family was considered to be in poverty. To empha- size the resource allocation issue, we present odds ratios (ORs) that distin- guish between ethnic and racial differ- ences among those with depression and ethnic and racial differences among those without depression.
Next, results from the extended specification were used to estimate the total disparity in accessing care and receiving quality care for each mi- nority group compared with non-Lati- no whites. Using the two-stage model estimates and the distribution of co- variates from the white population, we generated predicted probabilities of accessing treatment and receivingquality treatment for each racial-eth- nic and depression subgroup. This ap- proach allowed us to answer the hypo- thetical question: what level of treat- ment would persons in this minority group receive if they had the same characteristics as non-Latino whites?
McGuire and colleagues (30) used a similar approach to compute racial and ethnic disparities in outpatient mental health expenditures. In our study, minority individuals were given the non-Latino white distribution for all covariates. We adjusted for vari- ables associated with social class, such as poverty, insurance coverage, and education, to disentangle the effect of social class variables from those of ethnicity and race.
We used the bootstrap method to obtain 95% confidence intervals (CIs) in predicting probabilities of disparity in depression care between each racial and ethnic group and the non- Latino white group. For example, we predicted the treatment that Latinos with depression would receive if they had the same distribution of covari- ates as non-Latino whites with de- pression. (32). All analyses were con- ducted with Stata 9.2 statistical soft- ware (33). Models were adjusted for sampling design through a first-order Taylor series approximation, and sig- nificance tests were performed with design-adjusted Wald tests (34–36). Results Table 1 shows that there were striking racial and ethnic differences in sam- ple characteristics, including much higher rates of poverty and lower rates of health insurance coverage among all racial and ethnic minority groups compared with non-Latino whites. Latinos and Asians were much more likely than non-Latino whites and African Americans to live in the West, and African Americans were more likely than other groups to live in the South. Current depression was more prevalent among non-Lati- no whites compared with persons in racial and ethnic minority groups. For example, the prevalence of past-year depressive disorders was 5.4% for Asians and 11.2% for non-Latino whites. Table 2 shows that among those with any depressive disorder in the past 12 months, 63.7% of Latinos, PSYCHIATRIC SERVICES 'ps.psychiatryonline.org 'November 2008 Vol. 59 No. 11 1 12 26 66 6 PSYCHIATRIC SERVICES 'ps.psychiatryonline.org 'November 2008 Vol. 59 No. 11 1 12 26 67 7 T Ta ab bl le e 1 1 Sample characteristics of 8,762 respondents to surveys in the Collaborative Psychiatric Epidemiology Surveys a Combined Non-Latino African sample whites Latinos Asians Americans (N=8,762) (N=2,834) (N=1,603) (N=1,435) (N=2,890) Characteristic M SE M SE M SE M SE M SE p Age category (%)<.001 18–34 30.6 1.2 26.6 1.6 50.1 2.1 39.3 1.8 36.4 1.5 35–49 29.6 1.0 28.8 1.3 28.9 1.4 32.5 2.0 33.3 .9 50–64 21.0 1.1 22.8 1.5 12.3 1.0 17.6 1.4 18.6 1.0 ≥65 18.8 1.1 21.8 1.5 8.7 1.0 10.6 1.9 11.7 .8 Sex (%)<.05 Male 48.7 1.0 48.8 1.4 52.8 1.7 48.7 1.6 45.4 1.0 Female 51.3 1.0 51.2 1.4 47.2 1.7 51.3 1.6 54.6 1.0 Marital status (%)<.001 Married or cohabitating 60.2 1.1 62.2 1.4 64.7 1.6 70.3 2.1 43.0 1.1 Divorced, separated, or widowed 19.3 .7 19.4 .9 14.5 1.1 8.0 1.0 25.4 .9 Never married 20.5 1.1 18.4 1.5 20.7 1.4 21.7 1.6 31.6 1.5 College education (%)<.001 No 75.9 1.1 73.1 1.5 90.2 1.2 58.3 2.1 85.8 1.2 Yes 24.1 1.1 26.9 1.5 9.8 1.2 41.7 2.1 14.2 1.2 Poverty (%)<.001 No 87.4 .6 91.2 .7 73.9 2.1 83.7 1.4 77.5 1.4 Yes 12.6 .6 8.8 .7 26.1 2.1 16.3 1.4 22.5 1.4 Type of insurance (%)<.001 Not insured 12.7 .8 8.8 .9 34.8 2.9 12.8 1.3 17.5 1.0 Private through employer 56.0 1.2 58.3 1.5 39.4 2.6 57.4 2.0 55.4 1.3 Privately purchased 4.3 .4 4.6 .5 2.8 .6 8.9 1.0 2.0 .2 Medicare 20.3 1.1 23.4 1.5 9.9 1.1 11.0 1.7 13.7 .7 Medicaid 4.2 .3 2.5 .4 11.3 1.2 5.2 .9 8.2 .8 Other 2.5 .3 2.3 .5 1.8 .4 4.7 1.0 3.2 .5 Region (%)<.001 Northeast 19.6 2.7 20.6 3.6 18.6 2.1 18.3 4.1 15.5 1.1 Midwest 24.6 1.9 28.8 2.7 9.0 1.9 8.9 2.5 17.1 1.4 South 34.0 2.4 31.6 3.2 31.0 5.4 7.8 2.1 57.4 2.2 West 21.7 2.0 19.0 2.6 41.3 4.5 65.1 4.9 9.9 .9 Number of chronic conditions (%)<.001 0 48.0 .9 45.3 1.0 66.4 2.1 60.8 1.8 45.6 1.1 1 25.6 .6 26.0 .8 19.9 1.7 25.5 1.6 27.9 .9 ≥2 26.4 .7 28.8 .9 13.7 1.3 13.7 1.1 26.5 1.0 WHO-DAS b Days out of role because of mental problems c .29 .03 .24 .04 .27 .05 .07 .02 .65 .08 <.001 Cognition score d .90 .07 .91 .09 .70 .14 .34 .07 1.18 .13 <.001 Mobility score d 4.47 .29 4.96 .38 2.60 .42 1.20 .27 4.07 .28 <.001 Self-care score d .89 .15 .95 .20 .80 .17 .20 .07 .78 .11 <.001 Social functioning score d .55 .06 .55 .08 .32 .12 .14 .05 .79 .12 <.001 Role functioning score d 9.56 .39 9.36 .50 7.13 .62 5.32 .57 13.70 .72 <.001 Disorder category (%)<.001 Any depressive disorder in the past 12 months e 10.6 .5 11.2 .6 10.8 .9 5.4 .9 8.0 .6 No depressive disorder in the past 12 months 89.4 .5 88.8 .6 89.2 .9 94.6 .9 92.0 .6 aData were from the National Latino and Asian American Study (for Latinos and Asians), the National Comorbidity Survey Replication (part 2 sample), and the National Survey of American Life (NSAL). This analysis focused on the non-Latino-white, Latino, Asian, and African-American samples, drop- ping Native Americans and persons designating their race and ethnicity as “other.” Also excluded were respondents with a 12-month subthreshold de- pressive disorder, those with a lifetime depressive disorder but no 12-month depressive disorder, those who had a nondepressive diagnosis only, as well as NSAL respondents who did not complete all diagnosis batteries. Observations with any missing values were dropped.
bWald tests were conducted for disability assessment variables from the World Health Organization Psychiatric Disability Assessment SchedulecBased on number of days within the past 30 days. Possible scores range from 0 to 30, with higher scores indicating maximum disability.dPossible scores range from 0 to 100, with higher scores indicating maximum disability.eIncludes DSM-IV–defined dysthymia and major depressive episode 68.7% of Asians, and 58.8% of African Americans, compared with 40.2% of non-Latino whites, did not access any mental health treatment in the past year (p<.001). Among re- spondents with depression, those in minority groups also were significant- ly less likely than non-Latino whites to have received adequate care in the past year (p<.001) (Table 2). Al- though, as would be expected, most individuals without depression re- ceived no treatment, 3.2% of non- Latino whites without past-year de- pression (or lifetime or subthreshold depression) made four or more provider visits and received 30 days or more of antidepressant treatment, which compares with .7% of Latinos, 1.2% of Asians, and 1.3% of African Americans (p<.001) (Table 2).
All minority groups with 12-month depressive disorder were significantly less likely than non-Latino whites to receive any mental health care, after analyses adjusted for other factors (Table 3). Similarly, after adjustment for other factors, racial or ethnic mi- nority individuals without depression were less likely than non-Latino whites without depression to receive any treatment. In sensitivity analyses, in which we included the subthreshold depression data and classified these respondents as having depression, the findings were very similar to those dis- cussed here (results not shown).
Among those with depression who accessed any care, we found that al-though there were statistically signifi- cant racial and ethnic differences in the quality of care as a whole, only the difference between African Americans and non-Latino whites was statistically significant (Table 3). That is, African Americans who used services in the prior year had appreciably lower odds of receiving adequate depression care compared with whites (OR=.24, 95% CI= 14–.42). In two alternative sensi- tivity analyses, these models were reestimated independent of antide- pressant medication with the inclusion of the subthreshold cases and with the looser definition of quality of depres- sion care—the indicator of whether respondents received at least four vis- its with any formal mental health provider in the past year. The findings were similar to the others. Estimates based on analyses of specific racial and ethnic subsamples rather than a pooled sample yielded similar findings (not shown).
Table 4 shows racial and ethnic dif- ferences in predicted probabilities of accessing treatment and receiving ad- equate treatment for depression based on the extended model for each racial-ethnic subgroup and depres- sion subgroup, if every minority group had the same distribution of covariates as the non-Latino whites.
Among non-Latino whites with de- pression, 33.4% were predicted to ac- cess treatment and receive adequate depression care, compared with 25.0% of Latinos, 18.9% of Asians,and 10.4% of African Americans (sig- nificantly different for Asians and African Americans at p<.05 and mar- ginally significant for Latinos at p<.07). Among those without depres- sion, 3.1% of non-Latino whites were predicted to access treatment and re- ceive adequate treatment for acute depression; the predicted rates were much lower among racial and ethnic minority groups (Table 4). Latinos, Asians, and African Americans with depression were on average nine to 23 percentage points less likely to ac- cess mental health treatment and re- ceive adequate depression treatment than non-Latino whites with similar observed characteristics. Discussion The results of these analyses highlight that disparities in access to and quali- ty of care for ethnic and racial minor- ity populations remain a critical issue in mental health care. All racial and ethnic minority groups were signifi- cantly less likely than non-Latino whites to receive access to any mental health treatment. The observed find- ings reflect that ethnicity and race, even after adjustment for social class–related variables, such as pover- ty, insurance coverage, and educa- tion, still had an independent effect on access to depression treatment.
Several factors could account for the problem in access for persons with minority status. First, there was sig- nificant underdetection of depression PSYCHIATRIC SERVICES 'ps.psychiatryonline.org 'November 2008 Vol. 59 No. 11 1 12 26 68 8 T Ta ab bl le e 2 2 Quality of treatment for respondents with or without a depressive disorder in the past 12 months a No treatment Inadequate treatment Adequate treatment Disorder status and racial or ethnic group N % SE % SE % SE Any depressive disorder b 1,082 Non-Latino white 581 40.2 3.1 26.8 2.8 33.0 1.9 Latino 192 63.7 4.5 13.9 2.5 22.3 4.9 Asian 78 68.7 6.8 18.1 5.0 13.1 4.6 African American 231 58.8 3.2 29.0 3.1 12.1 2.5 No depressive disorder 7,680 Non-Latino white 2,253 89.9 .6 7.0 .6 3.2 .3 Latino 1,411 96.2 .6 3.1 .5 .7 .3 Asian 1,357 96.2 .7 2.6 .6 1.2 .5 African American 2,659 94.9 .5 3.8 .4 1.3 .3 aA treatment provider could be from the general medical sector or specialty mental health sector or could be a counselor or social worker in a non–men- tal health setting. Wald tests to identify differences in each of the treatment types across the racial and ethnic groups were significant at p<.001. F tests to identify racial and ethnic differences for any of the treatment types were significant at p<.001.
bIncludes DSM-IV–defined dysthymia and major depressive episode PSYCHIATRIC SERVICES 'ps.psychiatryonline.org 'November 2008 Vol. 59 No. 11 1 12 26 69 9 T Ta ab bl le e 3 3 Logistic regression results for prediction of any depression treatment and adequate depression treatment conditional on receiving any treatment a Any depression treatment (N=994) Adequate depression treatment (N=491) Standard model b Extended model c Standard model b Extended model c Variable OR 95% CI Wald dOR 95% CI Wald dOR 95% CI Wald dOR 95% CI Wald d Race or ethnicity comparison With depression e Latino and non-Latino white .52 .31–.89 ∗ <.001 .47 .25–.88 ∗ <.001 1.26 .45–3.55 <.001 1.21 .40–3.69 <.001 Asian and non-Latino white .40 .18–.87 ∗ .33 .14–.78 ∗ .89 .29–2.69 .80 .24–2.74 African American and non-Latino white .34 .22–.52 ∗∗∗ .33 .20–.55 ∗∗∗ .22 .12–.41 ∗∗∗ .24 .14–.42 ∗∗∗ Without depression e Latino and non-Latino white .30 .19–.49 ∗∗∗ <.001 .27 .15–.47 ∗∗∗ <.001 .52 .17–1.63 .51 .17–1.55 Asian and non-Latino white .41 .27–.64 ∗∗∗ .34 .21–.55 ∗∗∗ .85 .26–2.75 .76 .24–2.40 African American and non-Latino white .37 .28–.50 ∗∗∗ .33 .23–.46 ∗∗∗ .52 .27–1.01 .56 .28–1.13 Age (reference: 18–34) 35–49 1.16 .94–1.43 1.18 .90–1.54 1.15 .70–1.89 1.14 .65–2.01 50–64 1.10 .81–1.50 1.08 .73–1.61 1.02 .57–1.82 1.01 .50–2.06 ≥65 .55 .34–.88 ∗ .20 .09–.45 ∗∗∗ .45 .22–.95 ∗ .49 .18–1.27 Female (reference: male) 1.54 1.22–1.94 ∗∗∗ 1.48 1.17–1.88 ∗∗ 1.27 .76–2.12 1.29 .79–2.12 Number of chronic conditions (reference: 0) 1 1.46 1.04–2.03 ∗ 1.43 1.03–2.01 ∗ .96 .61–1.50 .94 .61–1.44 ≥2 1.43 1.07–1.92 ∗ 1.31 .97–1.77 1.27 .81–2.02 1.30 .81–2.09 WHO–Disability Assessment Schedule Days out of role >0 2.53 1.64–3.90 ∗∗∗ 2.32 1.43–3.75 ∗∗ 1.99 1.14–3.49 ∗ 2.04 1.20–3.48 ∗∗ Cognition >0 2.19 1.52–3.14 ∗∗∗ 2.17 1.61–2.91 ∗∗∗ 1.24 .83–1.86 1.28 .84–1.93 Mobility >0 1.35 .92–1.99 1.28 .88–1.85 1.44 .93–2.24 1.51 .94–2.40 Self-care >0 .98 .54–1.76 .84 .45–1.57 .66 .35–1.23 .65 .35–1.19 Social functioning >0 1.08 .73–1.58 1.10 .76–1.59 .93 .57–1.51 .85 .53–1.37 Role functioning >0 1.31 1.02–1.68 ∗ 1.33 1.02–1.74 ∗ .94 .62–1.42 .99 .66–1.48 Marital status (reference:
married) Divorced, separated, or widowed 1.53 1.13–2.08 ∗∗ 1.16 .69–1.97 Never married 1.27 .93–1.74 .95 .54–1.70 College education (reference: no) 1.33 .99–1.77 1.41 .85–2.34 Poverty below poverty threshold (reference: above threshold) .86 .57–1.30 .79 .37–1.68 Type of insurance (reference:
not insured) Private insurance through employer 1.39 .90–2.13 1.25 .70–2.24 Privately purchased insurance 1.40 .74–2.65 1.92 .66–5.59 Medicare 4.22 1.96–9.09 ∗∗∗ 1.23 .46–3.33 Medicaid 3.89 2.22–6.83 ∗∗∗ 1.79 .95–3.38 Other 2.92 1.28–6.68 ∗ 2.38 .70–8.13 Region (reference: Northeast) Midwest .78 .56–1.10 .77 .46–1.29 South .95 .66–1.37 .82 .54–1.26 West 1.13 .82–1.57 .99 .53–1.86 aRegressions for any depression treatment were based on 8,762 observations, and regressions for adequate depression treatment were based on 880 ob- servations.
bThe standard model included covariates for race and ethnicity, age, sex, number of chronic conditions, and World Health Organization Disability As- sessment Schedule scores.
cThe extended model included covariates for race and ethnicity, age, sex, number of chronic conditions, and World Health Organization Disability As- sessment Schedule scores, plus covariates for marital status, college education, poverty, insurance status, and region.
dWald tests for joint significance were conducted to test for any differences across each characteristic for each of the treatment types.eIncludes DSM-IV–defined dysthymia and major depressive episode∗p<.05∗∗p<.01∗∗∗p<.001 among the less acculturated ethnic and racial minority groups (21). Cur- rent approaches that rely on pro- viders to detect depression to facili- tate its care may have limited effec- tiveness, given that most respondents (85%–90%) belonging to ethnic and racial minority groups had recent contact with the health care system in the past year but a majority still did not receive treatment for depression.
Helping clinicians identify depres- sion for groups with these particular characteristics might be challenging.
Data indicate that symptom presenta- tion for mental health disorders varies across racial and ethnic groups and can differ from what most clinicians are trained to expect, resulting in clinical misdiagnoses (37). For exam- ple, Latinos are more likely to soma- tize psychiatric distress or to express psychiatric illness through cultural id- ioms of distress such as ataques de nervios (38). Second, losing pay from work (39) or the stigma that sur- rounds mental illness (40) may con- strain service utilization in racial and ethnic minority communities that are subject to unstable and temporary employment and that are overrepre-sented in low-wage jobs (41). For ex- ample, people of ethnic and racial mi- nority groups have reported delays in seeking services because of inability to leave work or take time off from work because of lack of benefits (29).
Third, an important factor discourag- ing minority members from accessing mental health services was their expe- rience of mistreatment by mental health professionals (42– 45). For African Americans, Asians, and Lati- nos, mistrust of health care profes- sionals and concerns about provider competence with their ethnic-racial group may decrease their sense of comfort in talking to professionals (31–33). Fourth, minority families ap- pear less likely to recognize depres- sion (46) or may feel that they can ad- equately provide care without formal providers (47). An individual with a mental illness and in a racial or ethnic minority group may be referred into mental health care only when the burden to the family creates undue stress and disruption. Patterns of dif- ferences in referral and treatment by providers have also been posited as a potential mechanism for such access disparities (48). Fifth, a limited work-force and insufficient funds result in inadequate support for mental health services in safety net settings (49).
The Institute of Medicine committee defines the health care safety net as “those providers that organize and deliver a significant level of health care and other related services to uninsured, Medicaid, and other vul- nerable patients.” (50).
We also found that regardless of race or ethnicity, most people who accessed depression treatment received inade- quate care, with African Americans being particularly unlikely to receive adequate care. This finding can be ex- plained by qualitative analyses of re- sponses of black community members that revealed that their experience of mistreatment and social exclusion by health professionals reverberated on future utilization and on community sentiments toward the mental health system (42). Disparities resulting from barriers to effective communi- cation between racially mismatched patients and providers, particularly for African Americans, may be lead- ing to greater discordance regarding a shared understanding of disease cau- sation and effectiveness of treatments PSYCHIATRIC SERVICES 'ps.psychiatryonline.org 'November 2008 Vol. 59 No. 11 1 12 27 70 0 T Ta ab bl le e 4 4 Access to and receipt of adequate treatment by depression status and racial and ethnic group a Bootstrapped calculations (%) b Sample Depression status and racial and ethnic group estimate (%) M SE 95% CI With depression Non-Latino white 33.4 33.3 2.3 28.9 to 37.9 Latino 25.0 24.6 4.1 16.7 to 32.7 Difference: Latino and non-Latino white –8.5 –8.8 –4.7 .3 to –18.1 Asian 18.9 18.8 6.1 7.9 to 31.5 Difference: Asian and non-Latino white –14.6 –14.5 –6.4 –1.3 to –26.1 African American 10.4 10.4 2.4 6.4 to 15.7 Difference: African American and non-Latino white –23.0 –23.0 –3.2 –16.4 to –29.0 Without depression Non-Latino white 3.1 3.1 .4 2.3 to 3.9 Latino .6 .6 .2 .3 to 1.2 Difference: Latino and non-Latino white –2.5 –2.4 –.4 –1.6 to –3.3 Asian 1.0 1.0 .4 .4 to 2.0 Difference: Asian and non-Latino white –2.1 –2.1 –.5 –1.0 to –3.0 African American .8 .8 .2 .4 to 1.2 Difference: African American and non-Latino white –2.3 –2.3 –.4 –1.4 to –3.2 aIncludes DSM-IV–defined dysthymia and major depressive episode. Estimates of accessing care and receiving adequate treatment were calculated by each racial or ethnic group and by depression status. These estimates were generated from the extended-model regression specifications shown in Table 3.
bOne thousand iterations were done in bootstrap. The 95% confidence intervals were obtained with a 2.5 percentile and 97.5 percentile of bootstrapped values of predicted probability. (51) and consequently substantial con- cerns about pharmacological treat- ments, thereby exacerbating unmet need among African Americans.
There are certain limitations of this study. The cross-sectional nature of the design does not permit us to iden- tify possible causal directions of the findings. The diagnostic and service utilization data are based on self-re- ports, which may be subject to incom- plete information, particularly if pa- tients did not know whether they were being prescribed an antidepressant.
Persons from ethnic and racial minor- ity populations, because they are less likely to discuss their treatment with their provider, may be unaware that they are being treated for depression (24). A further limitation is that no psy- chometric data are available for the ac- cess or quality measures used in this study. However, as mentioned above, these measures were adopted from the NCS (22,23) and have been wide- ly used in mental health services re- search. As a result, they were included as core measures in the CPES instru- ments. Regardless, studies assessing the psychometric properties of these measures are needed. Another limita- tion of this study is that the data in the racial category “other” could not be disaggregated by ethnic subgroup or by geographic location of cities be- cause of small samples. Only certain minority groups were included in the study, but these were better defined and were represented by larger sam- ples than is the case in most national studies. Finally, after we adjusted the characteristics of the minority groups to be the same as those of the non- Latino white population, the disparity estimates were strongly model based, and therefore a different model might lead to a different estimate. Future studies will permit more fine-grained analyses of the factors linked to these disparities. Regardless of these limita- tions, the findings paint a stark, recent picture of care for depression among racial and ethnic minority populations in the United States and clearly point to areas in need of further sustained attention.
An important area for further re- search includes understanding what “depression treatments” represent when received by non-Latino whiteswithout apparent depression or other measured mental disorders. This pat- tern could represent treatment for so- cial problems or general psychologi- cal distress, overuse of depression treatments, or appropriate use of an- tidepressant medications for other medical conditions, such as fibro- myalgia, painful diabetic neuropathy, migraines, and chronic back pain (52–54). To the extent that the supply of depression treatments is limited, it may be important to consider how to best distribute those resources across populations that differ in access to quality services, especially for sicker individuals. Future research could evaluate whether use of mental health services by those with no as- sessed need for care competes with access to treatment for patients in mi- nority groups, possibly limiting their access to mental health providers. Conclusions Our findings shift the debate to devel- oping policy, practice, and community solutions to address the barriers that generate these disparities. Simply rely- ing on current systems, without consid- ering the unique barriers to high-quali- ty care that apply for underserved eth- nic and racial minority populations, is unlikely to affect the pattern of dispari- ties we observed. For example, popula- tions that have been reluctant to come to the clinic for depression care may have correctly anticipated the limited benefits of usual care. One possible point of intervention is the use of qual- ity improvement programs to increase quality of care among minority groups.
Results from a randomized clinical trial demonstrated that a practice-initiated quality improvement intervention for primary care patients with depression improved the rate of appropriate care for depression for whites and under- served minorities alike (28,55). Pro- grams such as this one provide plausi- ble strategies for combating disparities in depression care. Policy changes might include increased resources for mental health services in safety net clinics. Practice changes might include training nurses in motivational inter- viewing or in routinely implementing evidence-based quality improvement programs for depression. Community strategies might include home visits bypeer counselors to engage patients in understanding the importance of treat- ment or ancillary services (such as transportation, child care, and patient advocacy) that facilitate access to care.
Future research should focus on devel- oping and evaluating the promise of such strategies.
Acknowledgments and disclosures The data analysis conducted for this article was made possible by grant K23-DA018715-01A2 from the research division of the Robert Wood Johnson Foundation. The National Latino and Asian American Study data used in this analysis were provided by the Center for Multicultural Mental Health Research at the Cambridge Health Alliance. The project was supported by grant U01-MH-06220 from the National Insti- tute of Mental Health (NIMH). This publica- tion was also made possible by grant P50-MH- 073469-02 from NIMH. The National Survey of American Life is supported by NIMH grant U01-MH-57716 with supplemental support from the National Institutes of Health Office of Behavioral and Social Science Research and from the University of Michigan. The National Comorbidity Study Replication is supported by grant U01-MH60220 from NIMH, with sup- plemental support from the National Institute on Drug Abuse, the Substance Abuse and Mental Health Services Administration, grant 044708 from the Robert Wood Johnson Foun- dation, and the John W. Alden Trust.
The authors report no competing interests.
References 1. Wang PS, Lane M, Olfson M, et al: Twelve- month use of mental health services in the United States: results from the National Co- morbidity Survey Replication. Archives of General Psychiatry 62:629–640, 2005 2. Blanco C, Patel S, Liu L, et al: National trends in ethnic disparities in mental health care. Medical Care 45:1012–1019, 2007 3. Asch S, McGlynn E, Hogan M, et al: Com- parison of quality of care for patients in the Veterans Health Administration and patients in a national sample. Annals of Internal Med- icine 141:938–945, 2004 4. Kilbourne A, Bauer M, Han X, et al: Racial differences in the treatment of veterans with bipolar disorder. Psychiatric Services 56:
1549–1555, 2006 5. Wells K, Klap R, Koike A, et al: Ethnic dis- parities in unmet need for alcoholism, drug abuse and mental health care. American Jour- nal of Psychiatry 158:2027–2032, 2001 6. Cook B, McGuire T, Miranda J: Measuring trends in mental health care disparities, 2000–2004. Psychiatric Services 58:1533– 1539, 2007 7. Heeringa SG, Berglund P: National Institutes of Mental Health (NIMH) Collaborative Psy- chiatric Epidemiology Survey Program (CPES) Data Set: Integrated Weights and Sampling Error Codes for Design-Based Analysis. Ann Arbor, University of Michigan, Survey Research Center, June 4, 2007. Avail- able at www.icpsr.umich.edu/cocoon/cpes/us PSYCHIATRIC SERVICES 'ps.psychiatryonline.org 'November 2008 Vol. 59 No. 11 1 12 27 71 1 ing.xml?Section=Weighting 8. Alegria M, Frank R, McGuire T: Managed care and systems cost- effectiveness treate- ment for depression. Medical Care 43:1225– 1233, 2005 9. Frank R, McGuire T, Normand S, et al: The value of mental health care at the system lev- el: the case of treating depression. Health Af- fairs 18:71–87, 1999 10. Alegría M, Takeuchi D, Canino G, et al: Con- sidering context, place and culture: the Na- tional Latino and Asian American Study. In- ternational Journal of Methods in Psychiatric Research 13:208–220, 2004 11. Kessler R, Merikangas K: The National Co- morbidity Survey Replication (NCS-R). In- ternational Journal of Methods in Psychiatric Research 13:60–68, 2004 12. Jackson J, Torres M, Caldwell C, et al: The National Survey of American Life: a study of racial, ethnic and cultural influences on men- tal disorders and mental health. International Journal of Methods in Psychiatric Research 13:196–207, 2004 13. Hartley H: Multiple frame surveys, in Pro- ceedings of Social Statistics Section, Ameri- can Statistical Association. Alexandria, Va, American Statistical Association, 1962 14. Hartley H: Multiple frame methodology and selected applications. Sankhya Series C 36:
99–118, 1974 15. Colpe L, Merikangas KR, Cuthbert B, et al:
Guest editorial. International Journal of Methods in Psychiatric Research 13:193–194, 2004 16. Heeringa S, Wagner J, Torres M, et al: Sam- ple designs and sampling methods for the Collaborative Psychiatric Epidemiology Studies (CPES). International Journal of Methods in Psychiatric Research 13:221–240, 2004 17. Duan N, Alegría M, Canino G, et al: Survey conditioning in self-reported mental health service use: randomized comparison of alter- native instrument formats. Health Services Research 42:890–907, 2007 18. Neighbors H, Caldwell C, Williams D, et al:
Race, ethnicity, and the use of services for mental disorders: results from the National Survey of American Life (NSAL). Archives of General Psychiatry 64:1–10, 2007 19. Kessler RC, Ustun TB: The World Mental Health (WMH) Survey initiative version of the World Health Organization (WHO) Composite International Diagnostic Inter- view (CIDI). International Journal of Meth- ods in Psychiatric Research 13:93–121, 2004 20. First MB, Spitzer RL, Gibbon M, et al: Struc- tured Clinical Interview for DSM-IV Axis I Disorders, Patient Edition (SCID-I/P, Ver- sion 2.0, 9/98 revision). New York, New York State Research Institute, Biometrics Re- search Department, 1998 21. Rehm J, Ustun T, Saxena S, et al: On the de- velopment and psychometric testing of the WHO screening instrument to assess disable- ment in the general population. International Journal of Methods in Psychiatric Research 8:110–123, 1999 22. Elhai J, Ford J: Correlates of mental health service use intensity in the National Comor- bidity Survey and National Comorbidity Sur-vey Replication. Psychiatric Services 58:
1108–1115, 2007 23. Kessler R, Zhao S, Katz S, et al: Past-year use of outpatient services for psychiatric prob- lems in the National Comorbidity Survey.
American Journal of Psychiatry 156:115–123, 1999 24. Institute of Medicine: Unequal Treatment:
Confronting Racial and Ethnic Disparities in Health Care. Washington, DC, National Academies Press, 2002 25. Lohr K: Institute of Medicine activities relat- ed to the development of practical guidelines.
Journal of Dental Education 54:699–704, 1990 26. Wang P, Berglund P, Kessler R: Recent care of common mental disorders in the United States: prevalence and conformance with evi- dence-based recommendations. Journal of General Internal Medicine 15:284–292, 2000 27. Katon W, Von Korff M, Lin E, et al: Collabo- rative management to achieve treatment guidelines: impact on depression in primary care. JAMA 279:1026–1031, 1995 28. Wells K, Sherbourne C, Schoenbaumm M, et al: Impact of disseminating quality improve- ment programs for depression in managed primary care: a randomized controlled trial.
JAMA 283:212–220, 2000 29. HEDIS 3.0: Narrative: What’s in It and Why It Matters. Washington, DC, National Com- mittee for Quality Assurance, 1997 30. McGuire T, Alegría M, Cook B, et al: Imple- menting the Institute of Medicine definition of disparities: an application to mental health care. Health Services Research 41:1979– 2005, 2006 31. Poverty in the United States, 2001, vol 2006.
Washington, DC, US Census Bureau, 2002 32. Efron B, Tibshirani R: An Introduction to the Bootstrap. New York, Chapman and Hall, 1993 33. Stata Statistical Software Release 9.2. College Station, Tex, Stata Corp, 2007 34. Lin DY: On fitting Cox’s proportional hazards models to survey data. Biometrika 87:37–47, 2000 35. Binder DA: Fitting Cox’s proportional haz- ards models from survey data. Biometrika 79:139–147, 1992 36. Schenker N, Gentleman J: On judging the significance of differences by examining the overlap between confidence intervals. Amer- ican Statistician 55:182–186, 2001 37. Alegría M, McGuire T: Rethinking a univer- salist framework in the psychiatric symp- tom–disorder relationship. Journal of Health and Social Behavior 44:257–274, 2003 38. Guarnaccia PJ, Martinez I, Ramirez R, et al:
Are ataques de nervios in Puerto Rican chil- dren associated with psychiatric disorder?
Journal of the American Academy of Child and Adolescent Psychiatry 44:1184–1192, 2005 39. Ojeda VD, McGuire TG: Gender and racial/ethnic difference in use of outpatient mental health and substance use services by depressed adults. Psychiatric Quarterly 77:
211–222, 2006 40. Schraufnagel T, Wagner A, Miranda J, et al:
Treating minority patients with depressionand anxiety: what does the evidence tell us?
General Hospital Psychiatry 28:27–36, 2006 41. Diamant AL, Hays RD, Morales LS, et al: De- lays and unmet need for health care among adult primary care patients in a restructured urban public health system. American Journal of Public Health 94:783–789, 2004 42. Mclean C, Campbell C, Cornish F: African- Caribbean interactions with mental health services in the UK: experiences and expecta- tions of exclusion as (re)productive of health inequalities. Social Science and Medicine 56:657–669, 2003 43. Echeverry JJ: Treatment barriers accessing and accepting professional help, in Psycho- logical Interventions and Research With Lati- no Populations. Edited by Garcia JG, Zea MC. Needham Heights, Mass, Allyn and Ba- con, 1997 44. Whaley AL: Cultural mistrust and mental health services for African Americans: a re- view and meta-analysis. Counseling Psycholo- gist 29:513–531, 2001 45. Whaley AL: Cultural mistrust of white men- tal health clinicians among African Americans with severe mental illness. American Journal of Orthopsychiatry 71:252–256, 2001 46. Roberts RE, Alegría M, Ramsay RC, et al:
Mental health problems of adolescents as re- ported by their caregivers: a comparison of European, African and Latino Americans.
Journal of Behavioral Health Services and Research 32:1–13, 2005 47. Pescosolido B, Wright E, Alegría M, et al: So- cial networks and patterns of use among the poor with mental health problems in Puerto Rico. Medical Care 36:1057–1072, 1998 48. Bloche M: Race and discretion in American medicine. Yale Journal of Health Policy, Law and Ethics 1:95–131, 2001 49. Vega W, Lopez S: Priority issues in Latino mental health services research. Mental Health Services Research 3:189–200, 2001 50. Lewin ME, Altman S (eds): America’s Health Care Safety Net: Intact but Endangered.
Washington, DC, Institute of Medicine, Committee on the Changing Market, Man- aged Care, and the Future Viability of Safety Net Providers, 2000 51. Cooper-Patrick L, Gallo J, Gonzales J, et al:
Race, gender, and partnership in the patient- physician relationship. JAMA 282:583–589, 1999 52. Chong M, Hester J: Diabetic painful neu- ropathy: current and future treatment op- tions. Drugs 67:569–585, 2007 53. Wong M, Chung J, Wong T: Effects of treat- ments for symptoms of painful diabetic neu- ropathy: systematic review. British Medical Journal 335:87, 2007 E Pub 54. Atkinson J, Slater M, Capparelli E, et al: Effi- cacy of noradrenergic and serotonergic anti- depressants in chronic back pain: a prelimi- nary concentration-controlled trial. Journal of Clinical Psychopharmacology 27:135–142, 2007 55. Miranda J, Duan N, Sherbourne C, et al: Im- proving care for minorities: can quality im- provement interventions improve care and outcomes for depressed minorities? Results of a randomized, controlled trial. Health Ser- vices Research 38:613–630, 2003 PSYCHIATRIC SERVICES 'ps.psychiatryonline.org 'November 2008 Vol. 59 No. 11 1 12 27 72 2
