Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Health Care Poor Underserved. Author manuscript; available in PMC 2011 October 11.
Published in final edited form as:
PMCID: PMC3190591

Social Determinants of Mental Health Treatment among Haitian, African American and White Youth in Community Health Centers

Nicholas Carson, MD, FRCPC, Ben Cook, PhD, MPH, and Margarita Alegría, PhD


We examine adequate mental health treatment, emergency room (ER) use, and early treatment dropout for Haitian, African American and White youth with a psychiatric diagnosis treated in community health centers in the United States. We present associations with possible socioeconomic determinants of care. Adequate treatment was less likely among Haitian youth from areas with greater poverty and among all youth from areas with more female-headed households. Medicaid-insured youth had more ER visits, especially African Americans. The relative impact of poverty on adequate care was higher for Haitians than Whites, and the relative impact of Medicaid coverage on ER use was higher for African Americans than for Whites. Early dropout was more likely among youth who were uninsured or from areas with more female-headed households. Socioeconomic factors and insurance status were significant determinants of care. Haitians living in poverty may face barriers to mental health services relative to other racial/ethnic groups.

Keywords: Mental health services, children, adolescents, Haitian, social determinants, race/ethnicity, community mental health, census tract, survival analysis, adequate care

Most studies to date on mental health service use among ethnic groups of youth in the United States suggest that Black youth receive less treatment than Whites.15 These findings apply to both children and adolescents and control for age, sex, socioeconomic status, and severity of illness. Research to improve mental health services for minority youth is important because mental disorders often begin in youth6 and early onset is correlated with worse education, employment, and socioeconomic achievement in adulthood, particularly for Black youth (Fergusson BJP 2007; Turnbull JCCP 1990; Kessler Educational Attainment AJP 1995).7,8,9 However, none of these studies distinguishes among the various ethnicities contained within the designation Black, making it difficult to ascertain intra-ethnic differences in mental health service use. We present an analysis of mental health service use among Haitian youth, an under-studied but sizeable community in the United States.

The most recent US census counts over 400,000 Haitians, but this number is widely viewed as an underestimate.10 There are large Haitian communities in the cities of Miami, New York, Boston, and Montreal. While studies have documented differential rates of disorders and use of mental health services between African American, English-speaking African Caribbean, and Haitian adults,11,12 few studies have examined such differences among Black youth in the United States. Intra-ethnic comparisons may identify important cultural differences within broad groups that have historically been clustered together for historical or political reasons, such as individuals of African descent. US Census categories, long the traditional unit of comparison in health disparities research, do not reflect the heterogeneity found among certain urban Black communities in the United States, listed above, and the diversity of health beliefs and behaviors contained therein. White youth, admittedly a diverse group as well, are a useful comparison because research shows them to receive the highest levels of mental health services.15

There is a scant research base on the mental health of Haitian youth who immigrated to the United States and or were born or naturalized in the United States (to whom we collectively refer as “Haitian youth”). Increasing rates of suicidality have been found among girls of Caribbean descent in the United States.13 Anecdotal evidence suggests that Haitian families may not readily engage in standard psychopharmacological or long-term psychotherapeutic approaches used in mental health treatments. Rather, some have observed a primary care model of service use among Haitian American families, where attendance in care decreases once symptoms decline.14 Prior research suggests that Haitian immigrant families have among the lowest rates of health insurance in the U.S., which could lead them to be underserved by mental health services.15,16 Although not studied in Haitian American families, under-recognition by parents of their children’s mental health problems also may account for underutilization of mental health services by Black families in the United States.17 Indeed, such parents are less likely than White parents to identify their children’s behavior as a mental health problem requiring intervention,18,19 and these perceptions may be moderated by family income level.20

Beyond patient-specific variables, there are social determinants that affect mental health and service use and may be important mediators of disparities between white, African American and Haitian youth. For example, the effects of low income, lack of education, single parenthood, and being uninsured on psychological distress in minority youth are well-documented.7, 21,22 Neighborhood disadvantage may moderate family effects on conduct-disordered behavior in Black youth.23 Higher levels of mental health need have been found among uninsured and publicly-insured youth.24 Youth from ethnic minorities may experience an exaggerated impact of these social determinants on mental health due to their disproportionate exposure to social adversity and ambient hazards such as crime, violence, or physical disorder.2427 Other studies have identified correlations between neighborhood poverty and symptoms of depression, anxiety, oppositional defiance, and conduct disorder in diverse groups of adolescents.28 It is possible that the negative effects of these individual-level and neighborhood-level socioeconomic disadvantages may be exacerbated among Haitian and African American youth given that these populations may be additionally burdened with societal influences such as discrimination and acculturative stress.

The present study considers the impact of social determinants on three important indicators of mental health service quality: treatment adequacy, emergency room (ER) use, and early dropout of care for Haitian, African American, and non-Latino White youth. We use a unique health care system dataset that includes patients from 69 ZIP codes in Boston, a metropolitan area with large African American and Haitian communities. The dataset includes mental health visits from specialty mental health, primary care, and emergency settings, all delivered within the same health system by the same providers. It is well documented that youth receive treatment for mental disorders across service sectors, and primarily in school and outpatient mental health settings;29 this dataset includes both outpatient mental health and school-based mental health clinic data.

To determine adequacy of youth mental health treatment, datasets must be able to account for differences in treatment modality (e.g., psychotherapy or psychopharmacology) and in total number of visits. Evidence-based treatment of mental illness in youth sets certain expectations in these areas to meet a basic level of quality. Research datasets that can provide this level of detail may help identify important barriers to the delivery of evidence-based care to ethnic minority youth, which is the care most likely to produce benefits. We hypothesize that Haitians and African Americans will have fewer psychotherapy visits and psychopharmacology visits compared to whites. We further hypothesize that lower rates of adequate care will be seen among families living in areas of more poverty, less education, more single parents, more foreign-born, or more non-citizens, and that these differences will be greater for Haitians and African Americans than for whites.

Because Black youth served in the public health sector use crisis psychiatric services disproportionately, emergency room visits can be seen as a negative indicator of treatment adequacy.30 This overuse is more prevalent in geographic areas of high poverty.31 Previous studies have shown that the majority of ER use is for non-emergent care and could be treated more efficiently in other outpatient settings, which may be due to decreased access to outpatient mental health care by Black youth. We hypothesize that Haitian and African American youth will use more emergency services, and that this relationship will have a positive association with social determinants.

Early dropout from mental health services in the United States is more common among Black families than white families.32 Risk factors for termination are poverty, parental stress, child antisocial behavior, and adverse child-rearing practices. There is also evidence that certain cultural groups, such as Black patients, might use services in brief episodes of engagement and disengagement.33 Because these risk factors are known to be more common among those living in social adversity, and Black families are more likely to live in such conditions, we hypothesize they will be more likely to end care prematurely.

In this paper, we test the hypotheses that African American and Haitian youth without insurance and living in areas of poverty, lower levels of education, and single parent households will have less adequate care and greater ER use than white youth in similar circumstances. We also assess the risk of dropout of mental health care for the different racial/ethnic groups by area-level characteristics and insurance status, accounting for multiple episodes of care.



We used administrative data from January 1, 2005 to February 28, 2009 gathered from 69 clinics in a Boston area community health system. This is an urban safety-net health system serving a diverse range of ethnicities and socio-economic backgrounds. The system provides mental health treatment to individuals in over 22 towns in Massachusetts and New England. Until 2008, it received state support to care for a higher proportion of uninsured and “Free Care” patients than most other regional health systems in the area. Thus, this health system may be better positioned to care for the underinsured and uninsured.

Data include all mental health care visits of African American (n=351), Haitian (n=237), and White (n=3,754) youth (ages three to twenty) with a chart diagnosis of at least one mental health disorder. Prior work in this health system suggests that approximately 13% of patients receiving pediatric mental health services are between eighteen and twenty years old, and the majority of these are black patients (First author, unpublished). To avoid biasing our sample we therefore included these youth in our analyses. Other inclusion criteria were having a record of at least one outpatient mental health visit between 2005 and 2009. Primary care and ER visits were included if the primary diagnosis was a mental health or substance-related diagnosis.

To assign racial/ethnic categories, we used the race/ethnicity information from the health system administrative database. We defined Haitian youth as any children with parent-reported Haitian ethnicity or with race listed as Black and language listed as Haitian Creole. This group therefore includes first generation Haitian youth and second generation Haitian American youth. African American youth were defined as any children with Black race but not of Haitian, African, or Latino/Hispanic ethnicity, and White youth as any children reporting White race but not Haitian, African, or Latino/Hispanic ethnicity. ZIP code information obtained from this administrative episode data was merged with 2000 ZIP code area-level U.S. Census demographic data34 to link the youth to social indicators.

We used three dependent variables. The first was adequate treatment, defined as receipt of at least eight outpatient psychotherapy visits, or at least four outpatient visits of which one was a psychopharmacological visit. This definition reflects recent treatment guidelines for depression and attention-deficit/hyperactivity disorder, which recommend specific medication trials plus four mental health visits, or eight visits of relevant psychotherapy.3541 This definition has also been used in previous studies.4244 Outpatient mental health visits were identified as psychotherapy or psychopharmacology using Current Procedural Terminology (CPT) codes assigned by clinicians for each visit.45 The second dependent variable analyzed was any emergency room (ER) use for psychiatric diagnoses, considering ER use as a preventable and thus negative outcome. The third dependent variable was early dropout from treatment, considered to be the first gap in treatment of 120 days for those individuals with fewer than eight mental health outpatient visits or less than four mental health visits of which one was a psychopharmacological visit.

We assessed the association of a number of social determinants with the dependent variables and tested the interactions of these social determinants with racial/ethnic groups. The first of these was insurance status (enrollment in private insurance, Medicaid, or uninsured). Using administrative data, we were able to identify insurance status for every visit for every individual. For the analysis of adequate care and ER use, we used the last recorded insurance status. For analysis of dropout of care, insurance status was considered a time-varying variable; that is, analysis of time to dropout incorporated changes in individuals’ insurance status.

Other social determinants of health were derived from combining patients’ ZIP codes with ZIP code-level sociodemographic data from the 2000 U.S. Census, including percent female-headed household, percent foreign-born, percent living in poverty, and percent with less than a high school education. Like previous researchers, we considered these variables as proxies for community characteristics such as neighborhood disadvantage (percent poverty, percent with less than high school education), family structure (percent female-headed household), ethnic context (percent foreign-born), and concentration of poverty, and lack of economic resources (percent poverty).4648 We recognize that other studies have combined these neighborhood-level measures using algorithms or factor analysis but have chosen to assess their independent contributions to these outcomes.49,50 Like previous researchers, we adjusted for severity of illness by including a variable indicating the presence of one or more psychiatric comorbidities.51 Finally, we included gender and age (0–12, 13–18, 19+) in our regression models to account for these groups’ differential rates of need for mental health care found in previous studies.52, 53

Statistical Analyses

First, we measured unadjusted rates for our first two dependent variables, any adequate mental health care and any psychiatric ER visit. We also compared the use of psychopharmacological treatments, age, gender, severity of illness, ZIP code area-level sociodemographic characteristics, and diagnosis for White, African American, and Haitian American youth with a psychiatric diagnosis in our sample.

We next used logistic regressions to model the probability of receiving adequate care and ER use for psychiatric disorders after adjustment for racial/ethnic group, age, sex, severity, insurance status, and ZIP code area-level sociodemographic characteristics. Models also included the interactions of racial/ethnic group with variables indicating social determinants in order to assess differences-in-difference, or racial differences in the association between social determinants and adequate mental health care and ER use. To translate odds ratios into predicted probabilities, we applied the model coefficients to the sample first when African American and private insurance indicators were set to 1 or present, next when Haitian and private insurance were set to 1 or present, next when African American and Medicaid insurance were set to 1, and so on, until we generated predicted rates of adequate care and ER use for all race/ethnicity by insurance coverage combinations.

To present the association of ZIP code area characteristics with adequate care and ER use, we generated predictions for each racial/ethnic group as if all lived in a ZIP code area where there was a low percentage of female-headed households (3.7%, the 25th percentile of our data) and a high percentage of female-headed households (5.8%, the 75th percentile of our data). Similarly, predictions were calculated in areas where there was a low percentage of households living below the poverty line (6.1%, the 25th percentile of our data) and a high percentage of households living below the poverty line (14.2%, the 75th percentile of our data). This technique, named recycled predictions54, 55 or predictive margins,56 has been used in previous health services studies.5759 It is a generalization of the adjusted treatment means to nonlinear models, allowing us to compare the rates of differing insurance categories and neighborhood contexts of racial/ethnic groups on the outcomes, after adjusting the distribution of all other observable attributes. Standard errors for recycled predictions and predicted differences were derived using the bootstrap procedure60 (set at 100 replications) and differences were considered significant if their 95% bootstrap intervals did not include zero.

Differences in racial/ethnic groups, insurance, and sociodemographic variables associated with the third dependent variable, early dropout of mental health treatment, were assessed first using Kaplan-Meier survival function curves. Separate curves were created for each of the racial/ethnic groups, plotting at day-long intervals the proportion of the number of cases that dropped services over the number of cases remaining in each time period. We next used Cox proportional hazard models to model the risk of early dropout from care for the three racial/ethnic groups, adjusting for the same covariates described in the logit model mentioned above. Insurance status, interactions of race/ethnicity and insurance status, severity, and age were treated as time-varying predictors and race, gender, and ZIP code area-level characteristics were treated as time-invariant predictors. To account for the possibility that some individuals were in the midst of a treatment episode before January 1, 2005, we conducted a sensitivity analysis, which excluded the drops incurred in the first year. All individuals that had not yet terminated care were considered to be right-censored at the end of our data (February 28, 2009). We assessed the Cox proportional hazards assumption using tests of Schoenfeld61 residuals and by visual estimation of log-log plots of the estimated survival curves. We used Wald tests62 to test significance in group hazard rate differences. All analyses were conducted using Stata 10.0.63


In our sample of youth with a psychiatric diagnosis, no significant racial/ethnic differences were found in the probability of receiving adequate mental health care treatment (Table 1). African American youth were more likely to use the ER for mental health treatment. Both Haitian and African American youth were less likely to have psychopharmacology visits in comparison to Whites. African American youth were also significantly more likely than Whites to have been assigned two or more comorbid mental health diagnoses at one of their mental health visits.

Table 1
Sample characteristics - includes all non-Latino Whites, African-Americans, and Haitian children age 3 to 20 seen at Boston area health care system hospitals and clinics between January 1, 2005 and February 28, 2009 (n=4342)

In contrast, Haitian and African American youth were less likely to have private insurance and more likely to have Medicaid than White youth. In comparison with Whites, Haitian and African American youth lived in ZIP code areas with greater percentage of foreign-born residents and households living below the federal poverty level. Racial/ethnic minority youth in our sample were significantly different from Whites on a number of psychiatric diagnoses. Compared with Whites, Haitian youth were less likely to have been diagnosed with anxiety disorders, bipolar disorder, and substance-related disorders, and more likely to have been diagnosed with adjustment disorders and developmental delays. African American youth were also more likely to have a diagnosis of depressive disorder, bipolar disorder, adjustment disorder, trauma-related disorder, and psychotic disorder, and less likely to have a diagnosis of anxiety disorder and substance-related disorders than their White counterparts.

In Table 2, we present two types of differences estimated from the multivariate logistic regression: differences within race/ethnicity by insurance category and ZIP code area characteristic (percent female-headed household and percent of residents below federal poverty line; significant differences are starred) and differences-in-differences (differences by race/ethnic group and insurance/ZIP code characteristic variable; shaded if significant).

Table 2
Predicted Rates of Adequate Care and ER Use by Race/Ethnicity and Selected Social Determinants

For the first set of comparisons, we found that White youth with private insurance were more likely to receive adequate care than Medicaid-insured and uninsured White youth, and African American youth with private insurance were more likely to receive adequate care than Medicaid-covered African American youth. Living in a ZIP code area with greater percentage of female-headed households was associated with a lower likelihood of receiving adequate care for all three racial/ethnic groups. Haitian youth living in ZIP code areas with a greater percentage of households below the poverty line were less likely to have adequate care. Regarding ER use for mental health care, White youth with private insurance were less likely to use the ER than Medicaid-insured and uninsured White youth, and African American youth with private insurance were less likely to use the ER than African American Medicaid enrollees.

For the second set of comparisons, we found two significant difference-in-differences. The difference between the lower rate of adequate care for Haitians living in higher-poverty areas compared with Haitians in lower-poverty areas was significantly greater than the difference in adequate care among Whites in higher and lower-poverty areas. The difference due to the higher rate of ER use for African American Medicaid enrollees compared with African American private insurance enrollees was significantly greater than the difference between White youth in private and Medicaid insurance programs.

To assess the effect of including patients up to twenty years of age, we re-estimated the predicted rates of adequate care and ER use excluding patients older than seventeen years. This resulted in a sample of 3014 white youth, 196 Haitian youth, and 279 African American youth. This sensitivity test did not change the significance of predictions by insurance status or predictions of ER use for mental health care. It did diminish the significance of the association between living in a ZIP code area with a greater percentage of female-headed households and adequate care for African American youth, and living in ZIP code areas with a greater percentage of households below the poverty line and adequate care for Haitian youth. However, these results were likely due to a decrease in power from the smaller sample size since the differences were qualitatively similar in magnitude and direction as differences estimated in the original sample.

Figure 1 presents Kaplan-Meier curves for each racial/ethnic group, representing the different survival functions, or the different probabilities over time of not dropping out of mental health treatment. After covariate adjustment, Cox proportional hazard model results (not shown) reveal that Haitian and African American youth were no different from White youth in their overall risk of dropping out of treatment. Youth living in a ZIP code with greater percentage of female-headed households were more likely to dropout of care than youth living in ZIP codes with lower percentage of female-headed households. Contrary to expectations, living in a ZIP code with greater percentage of households with less than a high school education was associated with less likelihood of dropping out of mental health treatment. There were no significant differences among the three groups in the frequencies of having greater than one dropout, i.e. having at least two treatment episodes (African American 26.0%, Haitian 22.7%, White 20.9%).


In this study, we examined how the mental health service use patterns of African American, Haitian, and White youth with a psychiatric diagnosis varied by insurance status and other social conditions. These youth received care in a health care network with multiple treatment sites serving a wide range of metropolitan ZIP codes with diverse social conditions. This health care network has prioritized the cultural competency of its services through comprehensive interpreter services, patient resources, and clinician education. It also provides a disproportionate amount of care for individuals who are uninsured or underinsured. These efforts may help explain why three different racial/ethnic groups of youth were equally likely to receive adequate care as we defined it. Differences in care may be seen in systems that lack the infrastructure to care for families of different ethnicities or lesser socioeconomic circumstances. However, greater use of emergency psychiatric services by African American youth and less use of psychopharmacological services by Haitian and African American youth (relative to White youth) suggests there are still important differences to be explained. Sociodemographic variables such as insurance, poverty, and percent female-head households seem to be important determinants of these different patterns of service use.

For example, Haitian youth living in areas with higher-poverty levels were less likely to receive adequate care than Haitians in lower-poverty areas. This difference exceeded that between Whites in higher- and lower-poverty areas. Future studies using datasets with richer data on barriers to mental health care, immigration status, and duration of residency are needed to identify the mechanisms underlying the differential impact of these contextual conditions. In the absence of this information, we speculate that the low rates of adequate care among Haitians living in areas of high poverty may be because they are less able to tap into local mental health services than their White and African American counterparts under similar conditions. Haitian families who are poor may be more likely to include recent immigrants, a group that faces greater barriers to mental health care in form of acculturative stress, low health literacy, language barriers, immigration status or lack of supportive social networks.64 Duration of residency in the United States is another factor that has been shown to correlate with increased access to regular health care among Caribbean and African American youth.65 These barriers may be more formidable for Haitians living in areas of greater poverty.

For all racial/ethnic groups, living in a ZIP code area with a greater percentage of female-headed households was negatively associated with our first and third dependent variables: receipt of adequate care and early dropout of mental health treatment. Female-headed households are more likely to live in poverty, with less access to resources needed to support an episode of mental health treatment, such as transportation, child care, insurance, or social support.66 Longitudinal research suggests that it is not single parenthood per se that correlates with poor outcomes, but more exposure to associated psychosocial adversity, including poverty, abuse, lower parental education, parental substance abuse and crime, and lower IQ.7 Our analyses suggest that, even when controlling for living in areas of high poverty, living in areas with more female-headed households increases the probability of early dropout over time, regardless of ethnic/racial group. Though our dataset does not account for maternal mental health, there is also some evidence to suggest that non-White single mothers are more likely to experience major depression, representing another barrier to mental health treatment for youth.67 Wrap-around services and intensive case management services have proven effective in increasing engagement of parents and youth in mental health treatment.68 We expect such services may be particularly useful for female-headed households.

Another potential contributor to lower rates of adequate use may be that African American and, particularly, Haitian American youth were significantly less likely than White youth to have any medication visits. This dataset cannot provide explanations for decreased use of medication among black families. However, a number of hypotheses find support in the literature that are related to decisions by both patients and providers. From the patient side, decreased acceptance of medications by Black families may be the result of mistrust,69 preferences to avoid psychiatric medication,70 and use behavioral or religious forms of management.71 Caribbean-born and US-born Black women are more likely than other ethnic groups to endorse stigma around the diagnosis and treatment of depression,72 which may affect treatment for their children as well. Issues of trust, embarrassment, and privacy are relevant to over half of youth when asked about mental health services; with boys more likely to endorse stigma than girls.73

From the provider side, black patients find their primary care visits less participatory than White patients do.74 These barriers may be related to cultural beliefs, and culturally sensitive interventions should address the patient, provider, and health care system. For example, efforts to increase participation in care by minority patients, with sensitivity to cultural beliefs about treatment,75 may lead to better implementation of evidence-based medication recommendations.76 Provision of language-specific services, materials, and case management may also improve engagement in treatment.77

Our second dependent variable was emergency room use for mental health treatment, and we found that African American youth with Medicaid had a relatively high frequency of ER use in comparison with privately insured African American youth. Whites who were uninsured or on Medicaid also used the ER more than privately insured Whites, but these differences were not as striking as those among African Americans. This disparity in ER use by publicly insured African American youth has been observed in prior studies.78 These youth may not be able to access the outpatient services that could pre-empt urgent care use. Medicaid enrollees may be using the emergency room for primary care services during off-hours due to work demands, or because they delay needed care until symptoms become urgent. The former hypothesis finds support in comparisons of working hours for Medicaid Managed Care enrollees; those who worked more than thirty hours per week were more likely to use the emergency room.79

Our dataset is limited in its ability to identify psychopharmacological treatments that took place during primary care mental health visits. As such, we may have underestimated the frequency of medication treatments among groups who received more mental health care proportionately from primary care physicians. However, since the proportion of such visits was small across groups (less than 6%), we do not anticipate this limitation greatly affects our findings. Furthermore, the diagnostic categories reflected in Table 1 are composites of a number of related diagnoses, and thus should not be assumed to represent the community prevalence of specific disorders in these patient groups.

To our knowledge, this is the largest study of mental health service use among Haitian American youth living in the United States published to date. Our findings may be of interest to health care researchers given that, even in a network focused on providing culturally competent care, geographic contextual factors and insurance status issues were significant determinants of care. Our approach of including two comparison groups (White and African American youth) reveals heterogeneity within a group (Black youth) that research has traditionally treated as homogeneous by virtue of using U.S. Census categories. We hope these methods will spur future mental health services research for communities with a variety of ethnic groups, leading to the development of policies that address the social determinants of mental health service use by racial/ethnic groups in community settings.

Table 3
Coefficients from Cox Proportional Hazards Model Measuring Time to Early Dropout of Mental Health Treatment Adjusting for Age, Sex, Severity, and Sociodemographic Characteristics

Contributor Information

Nicholas Carson, Harvard Medical School/Cambridge Health Alliance, Psychiatry, 120 Beacon Street, 4th floor, Somerville, MA 02143, Phone: (617) 575-5269.

Ben Cook, Harvard Medical School/Cambridge Health Alliance, Psychiatry, 120 Beacon St., 4th Floor, Somerville, MA 02143, Phone: 6175038449.

Margarita Alegría, Cambridge Health Alliance, Psychiatry, 120 Beacon St., 4th floor, Somerville, MA 02143, Phone: 6175038447, Fax: 6175038430.


1. Yeh M, McCabe K, Hurlburt M, et al. Referral sources, diagnoses, and service types of youth in public outpatient mental health care: a focus on ethnic minorities. Journal of Behavioral Health Services & Research. 2002;29(1):45–60. [PubMed]
2. Kataoka SH, Zhang L, Wells KB. Unmet need for mental health care among U.S children: variation by ethnicity and insurance status. Am J Psychiatry. 2002 Sep;159(9):1548–1555. [PubMed]
3. Kodjo CM, Auinger P. Predictors for emotionally distressed adolescents to receive mental health care. The Journal of Adolescent Health. 2004 Nov;35(5):368–373. [PubMed]
4. Garland AF, Lau AS, Yeh M, McCabe KM, Hough RL, Landsverk JA. Racial and ethnic differences in utilization of mental health services among high-risk youths. The American Journal of Psychiatry. 2005 Jul;162(7):1336–1343. [PubMed]
5. Elster A, Jarosik J, VanGeest J, Fleming M. Racial and ethnic disparities in health care for adolescents: a systematic review of the literature. Archives of Pediatrics & Adolescent Medicine. 2003 Sep;157(9):867–874. [PubMed]
6. Kessler RC, Berglund P, Demler O, et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005 Jun;62(6):593–602. [PubMed]
7. Fergusson DM, Boden JM, Horwood LJ. Recurrence of major depression in adolescence and early adulthood, and later mental health, educational and economic outcomes. Br J Psychiatry. 2007 Oct;191:335–342. [PubMed]
8. Turnbull JE, George LK, Landerman R, et al. Social outcomes related to age of onset among psychiatric disorders. J Consult Clin Psychol. 1990 Dec;58(6):832–839. [PubMed]
9. Kessler RC, Foster CL, Saunders WB, et al. Social consequences of psychiatric disorders, I: Educational attainment. Am J Psychiatry. 1995 Jul;152(7) 1-26-1032. [PubMed]
10. U.S Census Bureau. Table FBP-1: profile of selected demographic and social characteristics. Washington, DC: U.S Census Bureau; 2000. Available at:
11. Neighbors HW, Caldwell C, Williams DR, et al. Race, ethnicity, and the use of services for mental disorders: results from the National Survey of American Life. Arch Gen Psychiatry. 2007 Apr;64(4):485–494. [PubMed]
12. Jackson JS, Neighbors HW, Torres M, Martin LA, Williams DR, Baser R. Use of mental health services and subjective satisfaction with treatment among Black Caribbean immigrants: results from the National Survey of American Life. Am J Public Health. 2007 Jan;97(1):60–67. [PubMed]
13. Joe S, Baser RS, Neighbors HW, et al. 12-Month and Lifetime Prevalence of Suicide Attempts Among Black Adolescents in the National Survey of American Life. J Am Acad Child Adolesc Psychiatry. 2009 Mar;48(3):271–282. [PMC free article] [PubMed]
14. Desrosiers A, St Fleurose S. Treating Haitian patients: key cultural aspects. Am J Psychother. 2002;56(4):508–521. [PubMed]
15. Carrasquillo O, Carrasquillo AI, Shea S. Health insurance coverage of immigrants living in the United States: differences by citizenship status and country of origin. Am J Public Health. 2000 Jun;90(6):917–923. [PubMed]
16. Saint-Jean G, Crandall LA. Sources and barriers to health care coverage for Haitian immigrants in Miami-Dade County, Florida. Journal of Health Care for the Poor and Underserved. 2005;16:29–41. [PubMed]
17. Roberts RE, Alegria M, Roberts CR, Chen IG. Concordance of reports of mental health functioning by adolescents and their caregivers: a comparison of European, African and Latino Americans. The Journal of Nervous and Mental Disease. 2005 Aug;193(8):528–534. [PubMed]
18. Wachtel J, Rodrigue JR, Geffken GR, Graham-Pole J, Turner C. Children Awaiting Invasive Medical Procedures: Do Children and Their Mothers Agree on Child's Level of Anxiety? 1. Journal of Pediatric Psychology. 1994;19(6):723–735. [PubMed]
19. Walton JW, Johnson SB, Algina J. Mother and child perceptions of child anxiety: Effects of race, health status and stress. Journal of Pediatric Psychology. 1999;24(1):29–39.
20. Johnson SB, Wang C. Why Do Adolescents Say They Are Less Healthy Than Their Parents Think They Are? The Importance of Mental Health Varies by Social Class in a Nationally Representative Sample. Pediatrics. 2008 February 1;121(2):e307–e313. 2008. [PubMed]
21. Wilkinson R, Marmot M, editors. Social Determinants of Health: The Solid Facts. 2nd ed. Copenhagen: WHO Regional Office for Europe; 2003.
22. Franz M, Lensche H, Schmitz N. Psychological distress and socioeconomic status in single mothers and their children in a German city. Soc Psychiatry Psychiatr Epidemiol. 2003 Feb;38(2):59–68. [PubMed]
23. Brody GH, Ge X, Kim SY, et al. Neighborhood disadvantage moderates associations of parenting and older sibling problem attitudes and behavior with conduct disorders in African American children. J Consult Clin Psychol. 2003 Apr;71(2):211–222. [PubMed]
24. Williams DR, Jackson PB. Social sources of racial disparities in health. Health Aff (Millwood) 2005 Mar–Apr;24(2):325–334. [PubMed]
25. Latkin CA, Curry AD. Stressful neighborhoods and depression: a prospective study of the impact of neighborhood disorder. Journal of Health and Social Behavior. 2003:34–44. [PubMed]
26. Ross CE, Mirowsky J. Neighborhood disadvantage, disorder, and health. Journal of Health and Social Behavior. 2001:258–276. [PubMed]
27. Clark R, Anderson NB, Clark VR, Williams DR. Racism as a stressor for African Americans. A biopsychosocial model. Am Psychol. 1999 Oct;54(10):805–816. [PubMed]
28. Aneshensel CS, Sucoff CA. The Neighborhood Context of Adolescent Mental Health. Journal of Health and Social Behavior. 1996;37(4):293–310. [PubMed]
29. Burns BJ, Costello EJ, Angold A, et al. Children's mental health service use across service sectors Health affairs (Project Hope) Fall. 1995;14(3):147–159. [PubMed]
30. Snowden LR, Masland MC, Libby AM, Wallace N, Fawley K. Racial/ethnic minority children's use of psychiatric emergency care in California's Public Mental Health System. Am J Public Health. 2008 Jan;98(1):118–124. [PubMed]
31. Chow JC, Jaffee K, Snowden L. Racial/ethnic disparities in the use of mental health services in poverty areas. American Journal of Public Health. 2003 May;93(5):792–797. [PubMed]
32. Kazdin AE, Stolar MJ, Marciano PL. Risk factors for dropping out of treatment among White and Black families. Journal of Family Psychology. 1995;9(4):402–417.
33. Bui KVT, Takeuchi DT. Ethnic minority adolescents and the use of community mental health care services. American Journal of Community Psychology. 1992;20(4):403–417. [PubMed]
34. U.S Census Bureau. U.S Census Summary File 1 (SF 1) and Summary File 3 (SF 3) 2000
35. Weisz JR, Thurber CA, Sweeney L, Proffitt VD, LeGagnoux GL. Brief treatment of mild-to-moderate child depression using primary and secondary control enhancement training. J Consult Clin Psychol. 1997 Aug;65(4):703–707. [PubMed]
36. Olfson M, Gameroff MJ, Marcus SC, Waslick BD. Outpatient treatment of child and adolescent depression in the United States. Arch Gen Psychiatry. 2003 Dec;60(12):1236–1242. [PubMed]
37. Curry JF. Specific psychotherapies for childhood and adolescent depression. Biol Psychiatry. 2001 Jun 15;49(12):1091–1100. [PubMed]
38. Birmaher B, Brent D, Bernet W, et al. Practice parameter for the assessment and treatment of children and adolescents with depressive disorders. J Am Acad Child Adolesc Psychiatry. 2007 Nov;46(11):1503–1526. [PubMed]
39. The MTA Cooperative Group. A 14-month randomized clinical trial of treatment strategies for attention-deficit/hyperactivity disorder The MTA Cooperative Group. Multimodal Treatment Study of Children with ADHD. Arch Gen Psychiatry. 1999 Dec;56(12):1073–1086. [PubMed]
40. American Academy of Pediatrics. Clinical practice guideline: treatment of the school-aged child with attention-deficit/hyperactivity disorder. Pediatrics. 2001 Oct;108(4):1033–1044. [PubMed]
41. Pliszka S. Practice parameter for the assessment and treatment of children and adolescents with attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry. 2007 Jul;46(7):894–921. [PubMed]
42. Wang PS, Demler O, Kessler RC. Adequacy of Treatment for Serious Mental Illness in the United States. Am J Public Health. 2002 January 1;92(1):92–98. 2002. [PubMed]
43. Harman JS, Edlund MJ, Fortney JC. Disparities in the adequacy of depression treatment in the United States. Psychiatric Services. 2004;55(12):1379. [PubMed]
44. Kessler R, Berglund P, Demler O, et al. The epidemiology of major depressive disorder: Results from the National Comorbidity Survey Replication (NCS-R) Journal of the American Medical Association. 2003;289(23):3095–3105. [PubMed]
45. Beebe M, Dalton JA, Espronceda M. Current Procedural Terminology, Professional Ed. American Medical Association; 2008.
46. Robert SA, Ruel E. Racial segregation and health disparities between Black and White older adults. Journals of Gerontology Series B: Psychological Sciences and Social Sciences. 2006;61(4):S203. [PubMed]
47. LeClere FB, Rogers RG, Peters K. Neighborhood social context and racial differences in women's heart disease mortality. Journal of Health and Social Behavior. 1998;39(2):91–107. [PubMed]
48. Zuvekas SH. Trends in mental health services use and spending, 1987–1996. Vol 20: Health Affairs. 2001:214–224. [PubMed]
49. Dayal HC, Sharrar C, Mangan R, Rosenwaike J, Shapiro I. Ecological correlates of cancer mortality patterns in an industrialized urban population. Journal of the National Cancer Institute. 1984;73:565–574. [PubMed]
50. Sloggett A, Joshi H. Higher mortality in deprived areas: community or personal disadvantage? British Medical Journal. 1994;309(6967):1470–1474. [PMC free article] [PubMed]
51. Harpaz-Rotem I, Leslie D, Rosenheck RA. Treatment Retention Among Children Entering a New Episode of Mental Health Care. Psychiatr Serv. 2004 September 1;55(9):1022–1028. 2004. [PubMed]
52. Kessler R, Zhao S, Katz S, et al. Past-Year Use of Outpatient Services for Psychiatric Problems in the National Comorbidity Survey American. Journal of Psychiatry. 1999;156:115–123. [PubMed]
53. Wang PS, Lane M, Olfson M, Pincus HA, Wells KB, Kessler RC. Twelve-month use of mental health services in the United States: results from the National Comorbidity Survey Replication. Archives of General Psychiatry. 2005 Jun;62(6):629–640. [PubMed]
54. StataCorp. Reference Manual (Release 9.0) College Station, TX: STATA Press; 2005.
55. Kronick R, Gilmer T. Insuring Low-Income Adults: Does Public Coverage Crowd Out Private? Health Affairs. 2002;21(1):225–239. [PubMed]
56. Graubard BI, Korn EL. Predictive margins with survey data. Biometrics. 1999 Jun;55(2):652–659. [PubMed]
57. Davern M, Rodin H, Blewett LA, Call KT. Are the Current Population Survey Uninsurance Estimates Too High? An Examination of the Imputation Process. Health Services Research. 2007 (0). [PMC free article] [PubMed]
58. Wells KB, Tang L, Miranda J, Benjamin B, Duan N, Sherbourne CD. The Effects of Quality Improvement for Depression in Primary Care at Nine Years: Results from a Randomized, Controlled Group-Level Trial. Health Services Research. 2008 (0). [PMC free article] [PubMed]
59. Blewett LA, Davidson G, Bramlett MD, Rodin H, Messonnier ML. The Impact of Gaps in Health Insurance Coverage on Immunization Status for Young Children. Health Services Research. 2008 (0). [PMC free article] [PubMed]
60. Efron B. Bootstrap Methods: Another Look at the Jackknife. The Annals of Statistics. 1979;7(1):1–26.
61. Schoenfeld D. Partial residuals for the proportional hazards regression model. Biometrika. 1982;69:239–241.
62. Korn E, Graubard B. Simultaneous Testing of Regression Coefficients with Complex Survey Data: Use of Bonferroni t Statistics. American Statistician. 1990;44(4):270–276.
63. Stata Statistical Software Release 10.0 [computer program]. Version. College Station, TX: Stata Corporation; 2008.
64. Beiser M, Hou F, Hyman I, Tousignant M. Poverty, Family Process, and the Mental Health of Immigrant Children in Canada. Am J Public Health. 2002 February 1;92(2):220–227. 2002. [PubMed]
65. Sonis J. Association between duration of residence and access to ambulatory care among Caribbean immigrant adolescents. Am J Public Health. 1998 June 1;88(6):964–966. 1998. [PubMed]
66. Angel R, Worobey JL. Single motherhood and children's health. Journal of Health and Social Behavior. 1988:38–52. [PubMed]
67. Jian Li Wang E. The difference between single and married mothers in the 12-month prevalence of major depressive syndrome, associated factors and mental health service utilization. Social Psychiatry & Psychiatric Epidemiology. 2004 Jan;39(1):26–32. [PubMed]
68. Grimes KE, Mullin B. MHSPY: A Children's Health Initiative for Maintaining At-Risk Youth in the Community. The Journal of Behavioral Health Services and Research. 2006;33(2):196–212. [PubMed]
69. Whaley AL. Cultural mistrust and mental health services for African Americans: A review and meta-analysis. Counseling Psychologist. 2001 Jul;29(4):513–531.
70. Schnittker J. Misgivings of medicine?: African Americans' skepticism of psychiatric medication. J Health Soc Behav. 2003 Dec;44(4):506–524. [PubMed]
71. Bussing R, Gary FA. Practice guidelines and parental ADHD treatment evaluations: friends or foes? Harv Rev Psychiatry. 2001 Sep–Oct;9(5):223–233. [PubMed]
72. Nadeem E, Lange JM, Edge D, et al. Does stigma keep poor young immigrant and U.S.-born black and Latina women from seeking mental health care? Psychiatr Serv. 2007;58(12):1547–1554. [PubMed]
73. Chandra A, Minkovitz CS. Stigma starts early: gender differences in teen willingness to use mental health services. Journal of Adolescent Health. 2006;38(6):754.31–754.e8. [PubMed]
74. Cooper-Patrick L, Gallo J, Gonzales J, et al. Race, Gender, and Partnership in the Patient-Physican Relationship. Journal of the American Medical Association. 1999;282(6):583–589. [PubMed]
75. Alegria M, Polo A, Gao S, et al. Evaluation of a patient activation and empowerment intervention in mental health care. Med Care. 2008 Mar;46(3):247–256. [PMC free article] [PubMed]
76. Whitley R. Cultural Competence, Evidence-Based Medicine, and Evidence-Based Practices. Psychiatr Serv. 2007 December 1;58(12):1588–1590. 2007. [PubMed]
77. Chin MH, Walters AE, Cook SC, Huang ES. Interventions to reduce racial and ethnic disparities in health care. Med Care Res Rev. 2007 Oct;64(5 Suppl):7S–28S. [PMC free article] [PubMed]
78. Snowden LR, Masland MC, Libby AM, Wallace N, Fawley K. Racial/Ethnic Minority Children's Use of Psychiatric Emergency Care in California's Public Mental Health System. Am J Public Health. 2008 January 1;98(1):118–124. 2008. [PubMed]
79. Cook BL. Effect of Medicaid managed care on racial disparities in health care access. Health Services Research. 2007;42(1):124–145. [PMC free article] [PubMed]