Affective Disorders

Affective Disorders

Article Reviews
Contents lists available at ScienceDirect

Journal of Affective Disorders

journal homepage: www.elsevier.com/locate/jad

Research paper

Racial/ethnic differences in the treatment of adolescent major depressive disorders (MDD) across healthcare providers participating in the medicaid program

Aylin Yucela, Ekere J. Essiena, Swarnava Sanyala, Osaro Mgberec, Rajender R. Aparasua, Vinod S. Bhatarab, Joy P. Alonzoa, Hua Chena,⁎

aUniversity of Houston College of Pharmacy, Houston, TX, USA bUniversity of South Dakota, Sioux Falls, SD, USA c Bureau of Epidemiology, Houston Health Department

A R T I C L E I N F O

Keywords: Major depressive disorder Access to psychiatrist Racial/ethnic variation Treatment for depression

A B S T R A C T

Background: To examine whether racial/ethnic differences in receipt of MDD treatment could be explained by the specialty of provider diagnosing the adolescent. Method: Adolescents (10–20 years-old) with≥2 MDD diagnoses were identified using 2005–2007 Medicaid data from Texas. Patients were categorized based on the types of provider who gave the initial MDD diagnosis (psychiatrist (PSY-I), social worker/psychologist (SWP-I), and primary care physician (PCP-I)). Within the sub- cohorts identified by each type of provider, patients were further divided by racial/ethnic groups. Results: Of the 13,234-new pediatric MDD cases diagnosed, 61% were SWP-I, 33% PSY-I and 6% PCP-I. Results of the analysis using general linear multi-level model showed that being first diagnosed by a psychiatrist was associated with higher chance of receiving MDD related treatment (PCP-I vs. PSY-I (OR: 0.54, 95%CI: 0.4–0.7) and SWP-I vs. PSY-I (OR: 0.17, 95%CI: 0.1–0.2)). Specifically, regarding the receipt of pharmacotherapy, an interaction effect was detected between types of identifying providers and patients’ race/ethnicity. The analysis stratified by race/ethnicity found Whites received comparable treatment regardless being PCP-Is or PSY-Is, while for Hispanics, being first identified by a PCP was associated with lower likelihood of receiving treatment as compared to being first identified by a psychiatrist. Further analysis stratified by provider types showed that a significant racial/ethnic variation in medication utilization was observed in PCP-Is, but not in PSY-Is. Conclusion: For adolescents with MDD, being first diagnosed by a psychiatrist was associated with higher treatment rate and reduced racial/ethnic variation in the utilization of pharmacotherapy.

1. Introduction

Major Depressive Disorder (MDD) affects 7.5–8.2% of adolescents every year (Cummings et al., 2011; Avenevoli et al., 2015). MDD is the third leading cause of death and has long-term negative impact on adolescents such as failure at school, substance abuse, and adult de- pression (NIH agency SAMHSA, 2008; Windfuhr et al., 2008; Hawton et al., 2012; Emslie et al., 1997). Cognitive behavior therapy and in- terpersonal therapy are generally recommended for the treatment of mild depression (Hasler et al., 2005), and antidepressant medications are recommended for moderate to severe depression treatment along with behavioral therapies. (Molly et al., 2012)

Early identification and treatment are paramount in prognosis

(Birmaher et al., 2007; Regier et al., 1993) and lower the economic burden associated with MDD (Bishop et al., 2014). However, it is ob- served that 50% of pediatric MDD patients are not identified before they reach adulthood and as a result most children, and adolescents with MDD do not receive treatmen. (Regier et al., 1993; Luppa et al., 2007; Thomas and Holzer, 2006).

Racial/ethnic differences in access to mental health care, including the care for MDD, has been well-documented in children and adoles- cents (Kim et al., 2003; IOM 2002; Mental Health 2001; Elster et al., 2003; Le Cook et al., 2013; Lau et al., 2012; Hough et al., 2002; Kataoka et al., 2002; Cummings and Druss, 2011; Zito and Safer, 2001) and this disparity can lead to a more severe, persistent and disabling disease course in minority groups when compared to Whites (Lee et al., 2014;

https://doi.org/10.1016/j.jad.2018.04.045 Received 2 January 2018; Received in revised form 9 March 2018; Accepted 4 April 2018

⁎ Corresponding author. E-mail addresses: [email protected] (A. Yucel), [email protected] (E.J. Essien), [email protected] (S. Sanyal), [email protected] (O. Mgbere),

[email protected] (R.R. Aparasu), [email protected] (V.S. Bhatara), [email protected] (J.P. Alonzo), [email protected] (H. Chen).

Journal of Affective Disorders 235 (2018) 155–161

Available online 06 April 2018 0165-0327/ © 2018 Published by Elsevier B.V.

T

http://www.sciencedirect.com/science/journal/01650327
https://www.elsevier.com/locate/jad
https://doi.org/10.1016/j.jad.2018.04.045
https://doi.org/10.1016/j.jad.2018.04.045
mailto:[email protected]
mailto:[email protected]
mailto:[email protected]
mailto:[email protected]
mailto:[email protected]
mailto:[email protected]
mailto:[email protected]
mailto:[email protected]
https://doi.org/10.1016/j.jad.2018.04.045
http://crossmark.crossref.org/dialog/?doi=10.1016/j.jad.2018.04.045&domain=pdf
Williams et al., 2007; Breslau et al., 2005). Earlier attempts to explain this disparity often focused on social economic factors such as poverty, insurance coverage, local area demographics and parents’ education (Chatterji et al., 2009; Alegria et al., 2008; Kirby et al., 2006; Zuvekas and Taliaferro, 2003; Xu et al., 2004). It has also been reported that stigma related to mental disorders in minority groups can be a barrier in access to treatment (Gresenz et al., 2009). An earlier study reported that not having a usual source of care explained a significant portion of this disparity (Schraufnagel et al., 2006). In the treatment of adult cancer patients, literature have shown that type of provider who first identifies patients were found to be the determinant for course of care and the subsequent treatment (Gage-Bouchard et al., 2014). However, it remains unknown whether the type of provider who screens and identifies youth with MDD has any impact on the subsequent care steps and whether it contributes to the racial and ethnic differences in re- ceiving MDD treatment.

Child and adolescent psychiatrists (PSYs) are the best trained pro- vider group for screening, diagnosis and treatment associated with pediatric MDD (Mental Health 2015). However, access to psychiatrists, especially for minority groups, can be problematic, primarily due to a severe shortage of the work force (Thomas et al., 2009). For Medicaid insured children, the shortage is even greater, given that many psy- chiatrists do not accept public health insurance (especially Medicaid) due to a relatively lower reimbursement rate (Bishop et al., 2014).

Primary care physicians (PCPs), and social workers and psycholo- gists (SWPs) are also frequently involved in identification and treat- ment of MDD. The important role of PCPs has been increasingly re- cognized in recent years (Cunningham et al., 2009; ODPHP. Healthy People; Muldoon et al., 2006; Olfson et al., 2014). It was found that children and adolescents under routine PCP care are more likely to receive mental health service (Petterson et al., 2014; Jones et al., 2015). PCPs have been identified as a means of increasing access to mental health care by Office of Disease Prevention and Health Promotion (ODPHP. Healthy People). Moreover, Hispanic parents from minority groups usually favor PCPs and are more likely to follow their re- commendations in medication and/or visiting specialist regarding their children’s mental health care (Brown et al., 2007). Therefore, PCPs could play an important role in reducing racial and ethnic differences for access to MDD care (Anderson et al., 2013).

Despite the advantages, many PCPs are still hesitant to screen and provide care to adolescents with depression. Barriers to depression screening cited by PCPs include lack of time and training, lack of spe- cialist to refer, and inadequate reimbursement, (Kataoka et al., 2002; Thomas et al., 2009) not having adequate knowledge and training in depression diagnosis and management was also cited by PCPs as the primary reason to avoid prescribing antidepressants, even for severely depressed teens (Thomas et al., 2009).

To better understand the implications of initial contact to pediatric MDD care, our study first characterized the cascade of the delivery steps children and adolescents went through from a new MDD diagnosis to treatment initiation, and then examined whether the racial and ethnic differences in receiving MDD care could be explained by the type of provider who first identified the MDD cases.

2. Method

2.1. Data

Institutional review board and data use approval was obtained from University of Houston and Center for Medicare & Medicaid Services to analyze Texas Medicaid Analytic eXtract (MAX) claims data from 2005 to 2007. The data consists of inpatient, outpatient and prescription claims in addition to person-level information on Medicaid eligibility and demographics.

2.2. Study population

Youth aged 10–20 were identified if they received two consecutive diagnoses of Major Depressive Disorder (ICD-9-CM: 2962.2, 296.3, 300.4, 311) (Valuck et al., 2004; Wang and Si, 2013; Fiest et al., 2014) after a six-month interval without any mental disorder diagnosis (ICD- 9-CM: 291–299, 300–310, 311–319) and treatment. Patient’s first MDD diagnosis was defined as the index diagnosis.

Patients were excluded if they did not have 13 months of continuous Medicaid eligibility (six months before and after index diagnosis and the month of index diagnosis) (Burger et al., 2018; Chen et al., 2010; Sajatovic et al., 2016) or if they ever received a bipolar disorder diag- nosis (ICD-9-CM: 296.0, 296.1, 296.4–296.8, 301.11). Patients whose index diagnosis was given by a provider with missing or unknown specialty (2%) or specialty other (10%) than PCPs, psychiatrists, or social workers/psychologists, or who were identified in a hospital or emergency room, were excluded.

2.3. Study design

A retrospective cohort design was used to observe the impact of the provider who gave the initial MDD diagnosis on the subsequent delivery steps following the diagnosis and the receipt of MDD treatment.

2.4. Exposure

Type of provider who gave the index MDD diagnosis: According to the specialty of the provider who gave the index MDD diagnosis, patients were categorized as PCP-Identified (PCP-I), PSY-Identified (PSY-I), or SWP-Identified (SWP-I). PCPs included family practice/general prac- tice, pediatrics, internal medicine, nurse, and rural health clinic regis- tered nurse. SWPs were psychologist, licensed professional counselor, licensed master social worker, or case management.

Racial/ethnic groups (resdac, 2018): Patients who met the inclusion and exclusion criteria were further categorized into one of four racial/ ethnic groups: non-Hispanic Whites (i.e. Whites), Hispanics, non-His- panic Blacks (i.e. Blacks) and other.

2.4.1. Outcome of interest PSY involvement: Identifying whether patients had MDD related

outpatient visits paid to psychiatrists following the index MDD diag- nosis after identified by PCPs or SWPs until treatment initiation.

SWP involvement: Identifying whether patients had MDD related outpatient visits paid to a SWP following the index MDD diagnosis after identified by PCP or PSY until treatment initiation.

Treatment Received: Psychotherapy and psychopharmacotherapy (Rx) are the main treatment modalities for MDD. Psychotherapy were identified based on the CPT codes of 90,804–90,815, 90,845–90,847, 90,849, 90,853, 90,857, 90,875, 90,876, 90,880, H2017-H2020, H0032, H0035-H0040, H0045, H0046. Rx was defined according to USP Medicare Model Guidelines v6.0 66, which includes selective ser- otonin reuptake inhibitors (SSRI), tricyclic antidepressants (TCA), and other antidepressants. In addition, since typical and atypical anti- psychotics can also be used to treat MDD (Wang and Si, 2013), these medications were also included. Characteristics of treatments patients received along with the index diagnosis or within 6 months post the index diagnosis was measured as: a) Receipt of any MDD treatment; b) Type of the initial MDD treatment psychotherapy vs. pharmacotherapy; and c) Time to treatment.

2.5. Statistical analysis

The involvement of PSY and SWP during follow-up care and the receipt of treatment were compared by the type of providers who gave the initial MDD diagnosis (PCP-I, SWP-I, or PSY-I) and the racial/ethnic groups (White, Hispanic, Black, or other). Pearson’s Chi-Square test and

A. Yucel et al. Journal of Affective Disorders 235 (2018) 155–161

156

one-way ANOVA test followed by Tukey test were used to group dif- ferences.

To test whether the racial/ethnic differences in pathway to MDD care differs across different providers who initiated the care, a random effect general linear multi-level model (MLM) analysis was conducted by including the interaction term between race and type of provider who initiated the care for each outcome of interest. If the interaction term in any model appeared to be statistically significant, a two-way stratified analysis was conducted to examine: i) racial/ethnic differ- ences within each provider type and ii) variations across provider types for the care of each racial/ethnic group.

Other covariates adjusted in the model were patient demographics (age and gender), Medicaid eligibility category (foster children or not); type of Medicaid insurance coverage (fee for service or not), and any hospitalization before the index diagnosis. Pediatric medical com- plexity score (PMCS) which is widely used in studies using MAX data to evaluate pediatric patients’ chronic conditions, was also included in the analysis (Simon et al., 2014). All statistical analyses were conducted using SAS v.9.3.

3. Results

3.1. Cohort demographics

13,234 new pediatric MDD cases were identified after implementing the inclusion and exclusion criteria (Fig. 1). Table 1 presents patient characteristics at baseline. The average age was approximately 14; with a higher proportion of females (54.7%). Majority of the cohort was Hispanics (53%), followed by Whites (27%) and Blacks (16%). 19.4% of the study cohort was foster children and 40% were covered under the fee for service arrangement. 63.5% of the MDD cases were first iden- tified by SWP, 33.3% by PSY and only 5.8% of the pediatric MDD cases were identified by PCPs (Table 1).

Despite PCPs only diagnosed a small fraction of MDD cases, Hispanics had a higher chance of being identified by PCPs than other racial ethnic groups. Of those identified by PCPs, 65.5% were Hispanics, whereas 53.7% and 49.4% SWP-I and PSY-I were Hispanics. In contrast, Blacks were more likely identified by SWPs (17.2%) and PSYs (16.1%) than identified by PCPs (7.2%).

3.2. Receipt of MDD treatment

Table 2a presents the receipt of treatment following the initial MDD diagnosis by the types of identifying providers and racial/ethnic groups. Overall, 69% (N=9158) of new MDD cases received a treatment within six months since the index diagnosis. 22.5% of treated patients received psychotherapy and 77.4% received pharmacotherapy. Patients first identified by PSY had higher treatment rates (88%) compared to SWP-I (58%) or PCP-I (78%) patients. Treated SWP-I patients were most likely to receive psychotherapy as the initial treatment (34.9%), while majority of treated PSY-Is (90%) and PCP-Is (87.9%) received pharmacotherapy.

There were racial/ethnic differences in the receipt of MDD treat- ment among patients identified by all types of providers with the largest variations observed among PCP-Is and the smallest among PSY-Is. Among the PCP-Is, the treatment rate of Whites was 9.6% higher than that in Hispanics. The difference was only 3% while Hispanics and Whites were both identified by PSYs. A similar pattern was observed in the receipt of pharmacotherapy for which the medication rate in Whites was 7.6% higher than Hispanics among the PCP-I, and 1% among PSY-I.

3.3. Time to MDD treatment

Table 2b presents the duration between diagnosis and treatment by the types of identifying providers and racial/ethnic groups. The average time to treatment was 38.1 days. The duration from the index diagnosis

to the receipt of treatment was the shortest among patients identified by psychiatrists (14.8 ± 28). The duration was 30% shorter than the time to treatment observed in PCP-I (23.3 ± 37) and the longest was ob- served in SWP-I (38.1 ± 42)) (p< .0001). This pattern was not only observed in the overall cohort, but also within each racial/ethnic group.

3.4. SWP and PSY involvement between the diagnosis and treatment

As presented in Table 2a, many patients had additional MDD related visits between the index diagnosis and the receipt of treatment to providers other than the ones who gave the initial MDD diagnosis. 31.4% PCP-I and 16.3% of SWP-I were also seen by a PSY. Additional

Fig. 1. Consort diagram.

A. Yucel et al. Journal of Affective Disorders 235 (2018) 155–161

157

12.2% of PCP-I and 5.8% of PSY-I paid a visit to SWP before the treatment initiation. Within those PCP-I, the involvement of SWP and PSY varied across the racial/ethnic groups. 38.8% of the Hispanic pa- tients visited a psychiatrist and 14.9% visited a SWP after the index MDD diagnosis. The proportion of patients with psychiatrist-involved follow up visits was much lower in Blacks and Whites, with the lowest observed in Whites (15% with PSY involvement and 5.6% with SWP involvement).

3.5. Multivariate analysis

A MLM was conducted to examine the association between specia- list access and racial/ethnic disparity in receiving treatment.

3.6. Likelihood of receiving any treatment

Table 3 presents the factors that affect the likelihood of receiving treatment following a new MDD diagnosis. An interaction term between race/ethnicity and provider types was not included in the final analysis because the interaction term was not statistically significant. The multilevel analysis showed that Hispanics and Blacks were 20% less likely to receive treatment compared to Whites (Hispanics-OR: 0.80, 95%CI: 0.7–0.9; Blacks-OR: 0.81, 95%CI: 0.7–0.9). Patients identified

by PCP were 46% (OR: 0.54, 95%CI: 0.4–0.7) and by SWP 83% (OR: 0.17, 95%CI: 0.1–0.2) less likely to receive treatment compared to PSY- I. Also, being a foster child (OR: 1.26, 95%CI: 1.1–1.4) or having chronic conditions (OR: 1.40, 95%CI: 1.3–1.5) was associated with increased likelihood of receiving treatment.

3.7. Likelihood of receiving pharmacotherapy

Different from the findings for the overall treatment rate, there was a significant interaction effect between race/ethnicity and identifying provider types in predicting patients’ likelihood of receiving pharma- cotherapy (p-value: PCP-I*Hispanics= 0.04, SWP-I*Hispanics= 0.03). Therefore, a stratified analysis was conducted, and Table 4 a presents the racial/ethnic variation in likelihood of receiving pharmacotherapy within each identifying provider type. As shown in Table 4 a, Hispanics were 46% less likely to receive pharmacotherapy than Whites among PCP-I [OR: 0.54 (0.42–0.71)]. However, the racial/ethnic difference between Hispanics and Whites was reduced when patients were first diagnosed by SWP [OR: 0.73 (0.62–0.81)], and the gap was nullified in the sub-cohort first diagnosed by psychiatrists [OR: 0.87 (0.72–1.11)].

Table IVb presents the relative likelihood of receiving pharma- cotherapy when patients were first diagnosed by PCPs, SWPs versus PSYs within each racial/ethnic group. Hispanics were 46% less likely to receive pharmacotherapy when identified by PCPs compared to when they were identified by PSYs [OR: 0.54 (0.42–0.71)]. However, a si- milar practice variation was not observed in Whites [OR: 0.85 (0.62–1.33)].

4. Discussion

Although guidelines recommend psychotherapy for mild depression (Hasler et al., 2005), and adjuvant use of antidepressant medications for moderate to severe depression treatment along with behavioral therapies (Molly et al., 2012), our results indicated lower rates of initial treatment with psychotherapy relative psychopharmacotherapy espe- cially in patients identified by PCPs and PSYs, consistent across race/ ethnic groups. Despite PSYs are traditionally trained to conduct beha- vioral interventions, inadequate reimbursement for psychotherapy might have discouraged PSYs to consider it as the initial MDD treatment (nytimes 2018). For primary care physicians, medication is the only treatment option unless they refer patients out to psychologists or social workers.

Similarly to that has been reported in literature (McGuire and Miranda, 2008), our results indicate that higher percentage of minority patients identified by PCPs paid additional visits to psychiatrists before receiving any treatment. A recent systematic review revealed that di- agnostic tools such as child behavior checklist (CBCL) used by PCPs in

Table 1 Patient characteristics at baseline.

PCP-I (N=777) N (%)

SWP-I (N=8048) N (%)

PSY-I (N=4409) N (%)

p-value

Gender Male 311(40%) 3707(46%) 1,973 (44.7) 0.004 Female 466(59.9%) 4341(53.6%) 2436(55.3) Race/Ethnicity White 193(24.8%) 2057(25.5%) 1326(30%) <0.0001 Black 56(7.2%) 1386(17.2%) 712(16.1%) Hispanic 509(65.5%) 4326(53.7%) 2,179(49.4%) Others 19(2.4%) 279(3.4%) 192(4.3%) Foster care 63(8.1%) 1659 (20.6%) 849(19.2%) <0.0001 Fee for services 251 (32.3%) 3213(40%) 1785(40.4%) <0.0001 Chronic conditions No chronic

condition 62(7.9%) 634(7.8%) 277(6.2%) <0.0001

Non-chronic 433(55.7%) 4224(52.7%) 2052(46.5%) Non-complex

chronic 208(26.7%) 2506(31.1%) 1596(36.1%)

Complex chronic 74(9.5%) 684(8.4%) 484(10.9%) Hospitalization

before index diagnosis

60(7.7%) 305(3.7%) 326(7.3%) <0.0001

Table 2a Racial/Ethnic group differences in service utilization between first diagnosis and first treatment among patients identified by PCP, SWP and PSY.

Patients identified by different types of providers Total Hispanics Blacks Whites Others p-value PCP-I Patients N=777 N=509 N=56 N=193 N=19

Treatment rate: n(%N) 605 (77.8%) 381(74.8%) 44(78.5%) 163(84.4%) 17(89.4%) 0.01 Psychopharmacotherapy only initiated: npc(%n) 532 (87.9%) 326(85%) 40(90.9%) 151(92.6%) 15(88.2%) 0.2 Psychotherapy only initiated: nps(%n) 72(11.9%) 53(13.9%) 5(11.3%) 12(7.3%) 2(11.7%) <0.0001 PSY involvement(%N) 244 (31.4%) 198(38.8%) 12(21.4%) 29(15%) 5(26.3%) <0.0001 SWP involvement(%N) 95 (12.2%) 76(14.9%) 5(8.9%) 11(5.6%) 3(15.7%) 0.03 SWP-I patients N=8048 N=4326 N=1386 N=2057 N=279 Treatment rate: n(%N) 4669(58%) 2422(55.9%) 791(57%) 1275(61.9%) 181(64.8%) <0.0001 Psychopharmacotherapy only initiated npc(%n) 2,953(63.2%) 1,428(58.9%) 519(65.6%) 865(67.8%) 141(77.9%) <0.0001 Psychotherapy only initiated nps(%n) 1634(34.9%) 959(39.5%) 250(31.6%) 383(30%) 42(23.2%) <0.0001 PSY involvement: (%N) 1317(16.3%) 692(15.9%) 236(17.0%) 14(0.6%) 39(13.9%) 0.6 PSY-I patients N=4409 N=2179 N=712 N=1326 N=192 Treatment rate: n(%N) 3884(88%) 1,895(86.9%) 629(88.3%) 1,193(89.9%) 167(86.9%) 0.2 Psychopharmacotherapy only initiated npc(%n) 3,496(90%) 1709(90.1%) 544(86.4%) 1087(91.1%) 156(93.4%) 0.009 Psychotherapy only initiated nps(%n) 325(8.3%) 150(7.9%) 76(12%) 88(7.3%) 11(6.5%) 0.004 SWP involvement: (%N) 257(5.8%) 130(5.9%) 42(5.8%) 79(5.9%) 6(3.1%) 0.6

A. Yucel et al. Journal of Affective Disorders 235 (2018) 155–161

158

MDD diagnosis may not accommodate cultural specific symptoms. Such uncertainties may result in specialist referral in order to assess the ac- curacy or severity of MDD diagnosis. The study also reported that minority parents are more like than White parents to disagree with clinicians’ assessments in youth mental health symptoms. This could have led to additional confirmatory visit to psychiatrist after the initial diagnostic visit to PCPs.

The primary finding of our study is that patients who received the initial MDD diagnosis from PSYs had a significantly higher treatment rate as compared to those first identified by SWPs or PCPs. In terms of pharmacotherapy, an interaction effect was detected between provider access and racial ethnicity, which implies that being identified by a psychiatrist was associated with a higher increase in receipt of psy- chopharmacotherapy among minorities, such as Hispanics than that in Whites. Evidence from literature suggests that Hispanics are generally reluctant to seek healthcare due to myriad external (insurance access) and internal (cultural and language barriers, personal doctor-patient relationships) barriers (Miranda, 2004; Machado, 2014). Both Hispanic and African American patients had lower odds of finding pharma- cotherapy as an acceptable option for their MDD treatment compared to Whites (Cooper et al., 2003; Fernandez et al., 2011), especially among the Spanish speaking Hispanics. (Stockdale et al., 2008)

To eliminate disparities in mental health care, there has been

ongoing discussions on whether access and quality improvement should focus on minorities or a general policy that designed to improve the overall access and quality of mental health care such that minorities can benefit more than, or at least similarly as, Whites. (Alegría et al., 2008; McGuire and Miranda, 2008) A recent analysis on the 2005–2014 Na- tional Ambulatory Medical Care Survey found that disparity among racial/ethnic groups persists for depression treatment in primary care settings, while it was improved or eliminated in psychiatric settin- g.61Along with this study conducted in adults, our results provide evi- dence to support that the improved access to specialist could potentially reduce disparity existed in pediatric mental health care.

Access to mental health treatment can reduce service and expense for outpatient, emergency and inpatient care in Hispanics and Blacks more than Whites (Cook et al., 2015). Health and Human Services (HHS) (Koh et al., 2011) and Agency for Healthcare Research and Quality (AHRQ) have identified these as areas to improve (ahrq 2012). Despite the fact that access to psychiatrist is associated with improved treatment rate, the severe shortage of mental health workforce and increased public awareness on mental disorders implies that many de- pression cases will continuously be diagnosed and treated in the pri- mary care setting. In addition to the lack of knowledge on psychotropic drugs and the negative perceptions on pharmacotherapy among minorities, previous research has discussed that racial/ethnic differ- ences in physician distrust (Armstrong et al., 2007) may also play a role in disparities in the treatment mental health disorders. It is likely that specialists have more time than PCPs to discuss the treatment with patients and are in general more persuasive when patients have doubt on the medications. Other studies have shown that some racial minority patients are uncomfortable in receiving treatment from different racial/ ethnic providers, and reversely, physician subconscious prejudices, also called ‘implicit bias’, can also affect the way the patients get treated (Tello, 2017). To further reduce racial/ethnic variation in disparity, more studies are needed to understand the reasons why minorities and Whites are treated differently in the primary care setting and how in- dividual (e.g. patient and provider education) and system level inter- ventions (e.g. institutional support and health policy) could address the issues.

Table 2b Racial/Ethnic group differences in time to treatment between first diagnosis and first treatment among patients identified by PCP, SWP and PSY.

Service Providers Overall Hispanics Blacks Whites Others p-value Mean # of Days ± SD Mean # of Days ± SD Mean # of Days ± SD Mean # of Days ± SD Mean # of Days ± SD

PCP-I (N=777) 23.3 ± 37 23.2 ± 36 36.01 ± 51 20.09 ± 35 19.7 ± 27 0.1 SWP-I (N=8048) 38.1 ± 42 38.9 ± 42 37.7 ± 30.6 37 ± 42 30.6 ± 37 0.06 PSY-I (N=4409) 14.8 ± 28 13.2 ± 27 17.7 ± 33 15.7 ± 28 16.5 ± 29 0.002

Table 3 Likelihood of receiving any treatment: OR (95% CI).

Specialty OR CI

Specialty- PSY-I (Reference Group: OR=1) PCP vs. PSY-I 0.54 (0.41–0.73) SWP vs. PSY-I 0.17 (0.09–0.24) Race-Whites (Reference group: OR=1) Hispanics 0.80 (0.70–0.91) Blacks 0.81 (0.73–0.91) Others 1.16 (0.92–1.53) Other covariates Age 1.02 (0.98–1.04) Gender (Ref: male) 1.01 (0.90–1.11) Foster care 1.26 (1.12–1.42) Fee for service 0.87 (0.81–0.93) Chronic conditions 1.40 (1.33–1.51) Pre-index hospitalization 1.12 (0.89–1.42)

Table 4 Likelihood of receiving Pharmacotherapy across racial ethnic groups (IV a), and across the types of providers who initiated the care (IV b).

Table IV a: Stratified by provider specialties Hispanics vs. Whites Blacks vs. Whites Others vs. Whites

Specialty OR CI OR CI OR CI PCP-I (N=777) 0.53 (0.44–0.80) 0.68 (0.33–1.42) 1.41 (0.43–5.31) SWP-I (N=8048) 0.73 (0.62–0.81) 0.77 (0.69–0.94) 1.49 (1.1–2.0) PSY-I (N=4409) 0.87 (0.72–1.11) 0.73 (0.60–0.93) 0.93 (0.58–1.41)

Table IV b: Stratified by racial/ethnic groups Race PCP-I vs. PSY-I SWP-I vs. PSY-I

OR CI OR CI Hispanics 0.54 (0.42–0.71) 0.12 (0.08–0.20) Blacks 0.70 (0.30–1.53) 0.17 (0.06–0.21) Whites 0.85 (0.62–1.33) 0.15 (0.09–0.20)

A. Yucel et al. Journal of Affective Disorders 235 (2018) 155–161

159

5. Limitations

Our study is the first to examine the impact of access to mental health specialty care on racial/ethnic differences in pediatric MDD care. Despite significance, our study has limitations. The main concern is that the type of data used in our study does not include direct measures for disease severity. To control unmeasured severity, we restricted our analysis within a newly diagnosed cohort, excluded patients with psy- chiatric comorbidities, and adjusted for prior history of hospitalization and chronic conditions. With all these efforts, patients identified by psychiatrists may still be more severe as compared to those identified by other types of providers. An alternative explanation of our study finding could be that increased treatment rate was attributed to psy- chiatrists seeing more severe patients. However, the stratified analysis by racial/ethnic group has shown that regardless being identified by PCPs or psychiatrists, Whites had comparable chance of receiving pharmacotherapy. The finding does not support there was a significant difference in depression severity between PCP-I and PSY-I, and there- fore, the different treatment rates observed between Hispanics identi- fied by PCPs versus those identified by psychiatrists was most likely due to factors beyond the disease severity. Another limitation of our study is that the extent of practice variation differs on region and time, our study using older, regional Medicaid data may not be generalizable to other populations in different geographic regions during different time frame. (Shone et al., 2003; Horvitz-Lennon et al., 2015) However, the study provides valuable baseline data for future studies that evaluate the impact of change in public policy and mental health care delivery during the past decade on racial/ethnic disparity.

6. Conclusion

In conclusion, we found that 1) there was a modest difference in overall treatment rate and in the utilization of psychopharmacotherapy across racial/ethnic groups; 2) access to psychiatrists is associated with higher overall treatment rate and higher rate of using psychopharma- cotherapy in all racial/ethnic groups. 3) Access to psychiatrists appears to significantly reduce the inequality in using psychopharmacotherapy between Hispanics and Whites.

Conflict of interest

None.

Funding

This is the preliminary or baseline study of the AHRQ funded project (Grant no. AHRQ R03 HS025251-01) “Does Access to Patient Centered Medical Home reduce Racial and Ethnic Disparities in the care for Pediatric Major Depressive Disorders?”

Acknowledgments

We hereby declare: • That the manuscript is being submitted only to Journal of affective

disorders, that it will not be submitted elsewhere while under con- sideration, that it has not been published elsewhere, and, should it be published in Journal of Affective Disorders that it will not be published elsewhere, either in similar form or verbatim, without permission of the editors.

• That all authors are responsible for reported research. • That all authors have participated in the concept and design;

analysis and interpretation of data; drafting or revising of the manu- script, and that they have approved the manuscript as submitted.

References

http://www.nytimes.com/2011/03/06/health/policy/06doctors.html?pagewanted= all).

https://www.resdac.org/cms-data/variables/raceethnicity-msis. AHRQ (Agency for Health Care Research and Quality). Activities to reduce racial and

ethnic disparities in health care. AHRQ. 2012 [cited 2016 12/11]; Available from: http://www.ahrq.gov/research/findings/factsheets/minority/disparities/index. html.

AACAP (American Academy of Child and Adolescent Psychiatry). The child and adoles- cent psychiatrist. 2015. AACAP. Available from: https://www.aacap.org/AACAP/ Families_and_Youth/Facts_for_Families/FFF-Guide/The-Child-And-Adolescent- Psychiatrist-000.aspx.

Alegría, M, et al., 2008. Disparity in depression treatment among racial and ethnic minority populations in the United States. Psychiatr. Serv 59 (11), 1264–1272.

Anderson, K.K., et al., 2013. Determinants of negative pathways to care and their impact on service disengagement in first-episode psychosis. Social Psychiatry Psychiatr. Epidemiol. (1), 125–136.

Armstrong, K., et al., 2007. Racial/ethnic differences in physician distrust in the United States. Am. J. Public Health 97 (7), 1283–1289.

Avenevoli, S., et al., 2015. Major depression in the national comorbidity survey-adoles- cent supplement: prevalence, correlates, and treatment. J. Am. Acad. Child Adolesc. Psychiatry 54 (1), 37–44.

Birmaher, B., et al., 2007. Practice parameter for the assessment and treatment of children and adolescents with depressive disorders. J. Am. Acad. Child Adolesc. Psychiatry 46 (11), 1503–1526.

Bishop, T.F., et al., 2014. Acceptance of insurance by psychiatrists and the implications for access to mental health care. JAMA Psychiatry 71 (2), 176–181.

Breslau, J., et al., 2005. Lifetime risk and persistence of psychiatric disorders across ethnic groups in the United States. Psychol Med. 35 (3), 317–327.

Brown, J.D., et al., 2007. Receiving advice about child mental health from a primary care provider: African American and Hispanic parent attitudes. Med. Care 45 (11), 1076–1082.

Burger, C.D., et al., 2018. Treatment patterns and associated health care costs before and after treatment initiation among pulmonary arterial hypertension patients in the United States. J. Manage. Care Spec. Pharm. 13, 1–9.

Chatterji, P., Alegria, M., Takeuchi, D., 2009. Racial/ethnic differences in the effects of psychiatric disorders on employment. Atl. Econ. J. 37 (3), 243–257.

Chen, S.Y., et al., 2010. Follow-up visits by provider specialty for patients with major depressive disorder initiating antidepressant treatment. Psychiatr. Serv. 61 (1), 81–85.

Cook, B.L., et al., 2015. The costs and benefits of reducing racial-ethnic disparities in mental health care. Psychiatr. Serv. 66 (4), 389–396.

Cooper, L.A., et al., 2003. The acceptability of treatment for depression among African- American, Hispanic, and white primary care patients. Med. Care 41 (4), 479–489.

Cummings, J.R., et al., 2011. Racial/ethnic differences in mental health service use among adolescents with major depression. J. Am. Acad. Child Adolesc. Psychiatry 50 (2), 160–170.

Cummings, J.R., Druss, B.G., 2011. Racial/ethnic differences in mental health service use among adolescents with major depression. J. Am. Acad. Child Adolesc. Psychiatry 50 (2), 160–170.

Cunningham, P.J., et al., 2009. Beyond parity: primary care physicians’ perspectives on access to mental health care. Health Aff (Millwood) 28 (3) 490–50.

Elster, A., et al., 2003. Racial and ethnic disparities in health care for adolescents: a systematic review of the literature. Arch. Pediatr Adolesc. Med. 157 (9), 867–874.

Emslie, G.J., et al., 1997. Recurrence of major depressive disorder in hospitalized children and adolescents. J. Am. Acad. Child Adolesc. Psychiatry 36 (6), 785–792.

Fernandez, Y.G.E., et al., 2011. Depression treatment preferences of Hispanic individuals: exploring the influence of ethnicity, language, and explanatory models. J .Am. Board Fam. Med. 24 (1), 39–50.

Fiest, K.M., et al., 2014. Systematic review and assessment of validated case definitions for depression in administrative data. BMC Psychiatry 14, 289.

Gage-Bouchard, E.A., et al., 2014. Factors influencing patient pathways for receipt of cancer care at an NCI-designated comprehensive cancer center. PLoS ONE 9 (10), e110649.

Gresenz, C.R., Rogowski, J., Escarce, J.J., 2009. Community demographics and access to health care among US Hispanics. Health Serv. Res. 44 (5 Pt 1), 1542–1562.

Hasler, G., et al., 2005. Depressive symptoms during childhood and adult obesity: the Zurich Cohort Study. Mol. Psychiatry 10 (9), 842–850.

Hawton, K., Saunders, K.E., O’Connor, R.C., 2012. Self-harm and suicide in adolescents. Lancet 379 (9834), 2373–2382.

Horvitz-Lennon, M., et al., 2015. Where You live matters: quality and racial/ethnic dis- parities in schizophrenia care in four state medicaid programs. Health Serv. Res. 50 (5), 1710–1729.

Hough, R.L., et al., 2002. Mental health services for Latino adolescents with psychiatric disorders. Psychiatr. Serv. 53 (12), 1556–1562.

IOM, 2002. Unequal Treatment: Confronting Racial And Ethnic Disparities In Health Care. Washington,D.C.

Jones, A.L., et al., 2015. Usual primary care provider characteristics of a patient-centered medical home and mental health service use. J. Gen. Intern. Med. 30 (12), 1828–1836.

Kataoka, S.H., Zhang, L., Wells, K.B., 2002. Unmet need for mental health care among US children: variation by ethnicity and insurance status. Am. J. Psychiatry 159 (9), 1548–1555.

Kim, W.J., et al., 2003. American academy of, and N. Adolescent psychiatry task force on

A. Yucel et al. Journal of Affective Disorders 235 (2018) 155–161

160

https://www.resdac.org/cms-data/variables/raceethnicity-msis
http://www.ahrq.gov/research/findings/factsheets/minority/disparities/index.html
http://www.ahrq.gov/research/findings/factsheets/minority/disparities/index.html
https://www.aacap.org/AACAP/Families_and_Youth/Facts_for_Families/FFF-Guide/The-Child-And-Adolescent-Psychiatrist-000.aspx
https://www.aacap.org/AACAP/Families_and_Youth/Facts_for_Families/FFF-Guide/The-Child-And-Adolescent-Psychiatrist-000.aspx
https://www.aacap.org/AACAP/Families_and_Youth/Facts_for_Families/FFF-Guide/The-Child-And-Adolescent-Psychiatrist-000.aspx
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0001
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0001
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0002
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0002
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0002
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0003
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0003
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0004
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0004
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0004
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0005
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0005
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0005
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0006
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0006
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0007
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0007
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0008
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0008
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0008
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0009
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0009
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0009
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0010
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0010
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0011
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0011
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0011
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0012
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0012
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0013
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0013
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0014
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0014
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0014
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0015
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0015
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0015
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0016
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0016
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0017
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0017
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0018
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0018
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0019
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0019
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0019
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0020
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0020
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0021
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0021
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0021
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0022
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0022
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0023
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0023
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0024
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0024
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0025
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0025
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0025
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0026
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0026
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0027
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0027
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0028
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0028
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0028
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0029
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0029
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0029
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0030
workforce, child and adolescent psychiatry workforce: a critical shortage and na- tional challenge. Acad. Psychiatry 27 (4), 277–282.

Kirby, J.B., Taliaferro, G., Zuvekas, S.H., 2006. Explaining racial and ethnic disparities in health care. Med. Care 44 (5 Suppl), 64–72.

Koh, H.K., Graham, G., Glied, S.A., 2011. Reducing racial and ethnic disparities: the ac- tion plan from the department of health and human services. Health Aff (Millwood) 30 (10), 1822–1829.

Lau, M., Lin, H., Flores, G., 2012. Racial/Ethnic Disparities in Health and Health Care among US Adolescents. Health Serv. Res. 47 (5), 2031–2059.

Le Cook, B., Barry, C.,L., Busch, S.,H., 2013. Racial/ethnic disparity trends in children’s mental health care access and expenditures from 2002 to 2007. Health Serv. Res. 48 (1), 129–149.

Lee, S.Y., et al., 2014. Racial and ethnic differences in depressive subtypes and access to mental health care in the United States. J. Affective Disord. 155, 130–137.

Luppa, M., et al., 2007. Cost-of-illness studies of depression: a systematic review. J Affective Disord. 98 (1-2), 29–43.

Machado, A., 2014. Why many latinos dread going to the doctor. Atlantic(May). McGuire, G.T., Miranda, J., 2008. Racial and ethnic disparities in mental health care:

evidence and policy implications. Health Aff (Millwood) 27 (2), 393–403. Mental Health, 2001. Culture, Race, and Ethnicity: A Supplement to Mental Health: A

Report of the Surgeon General. Rockville (MD). Miranda, J., 2004. Disparities in Care for depression disparities in care for depression

among primary care patients. J. Gen. Intern. Med. 19, 120–126. Molly, S., et al., 2012. Treatment of childhood and adolescent depression. Am. Fam. Phys.

86 (5), 442–448. Muldoon, L.K., et al., 2006. Primary care (PC) and primary health care (PHC). What is the

difference? Can. J. Public Health 97 (5), 409–411. NIH agency SAMHSA (Substance Abuse and Mental Health Service Administration)

National Survey on Drug Use and Health: National Findings (Office of Applied Studies, NSDUH Series H-34, DHHS Publication No. SMA 08-4343) Available from: http://archive.samhsa.gov/data/NSDUH/2k7nsduh/2k7Results.htm#8.2.1 2008.

ODPHP. Healthy People 2020: Access to Health Services. 2016 07/01/2016 [cited 2016 07/03]; Available from: https://www.healthypeople.gov/2020/topics-objectives/ topic/Access-to-Health-Services.

Olfson, M., et al., 2014. National trends in the mental health care of children, adolescents, and adults by office-based physicians. JAMA Psychiatry 71 (1), 81–90.

Petterson, S., et al., 2014. Mental health treatment in the primary care setting: patterns and pathways. Fam. Syst. Health 32 (2), 157–166.

Regier, D.A., et al., 1993. The de facto US mental and addictive disorders service system.

Epidemiologic catchment area prospective 1-year prevalence rates of disorders and services. Arch. Gen. Psychiatry 50 (2), 85–94.

Sajatovic, M., et al., 2016. Dosing patterns and medication adherence in bipolar disorder patients treated with lurasidone: a US retrospective claims database analysis. Ther. Adv. Psychopharmacol. 6 (6), 355–368.

Schraufnagel, T.J., et al., 2006. Treating minority patients with depression and anxiety: what does the evidence tell us? Gen. Hosp. Psychiatry 28 (1), 27–36.

Shone, L.P., et al., 2003. The role of race and ethnicity in the State Children’s Health Insurance Program (SCHIP) in four states: are there baseline disparities, and what do they mean for SCHIP? Pediatrics 112 (6 Pt 2), e521.

Simon, T.D., et al., 2014. Pediatric medical complexity algorithm: a new method to stratify children by medical complexity. Pediatrics 133 (6), e1647–e1654.

Stockdale, S.E., et al., 2008. Racial and ethnic disparities in detection and treatment of depression and anxiety among psychiatric and primary health care visits, 1995-2005. Med. Care 46 (7), 668–677.

Tello, M., 2017. Racism and Discrimination in Health care: Providers and Patients. Harvard Health Publishing.

Thomas, C.R., Holzer, C.E., 2006. The continuing shortage of child and adolescent psy- chiatrists. J. Am. Acad. Child Adolesc. Psychiatry 45 (9), 1023–1031.

Thomas, K.C., et al., 2009. County-level estimates of mental health professional shortage in the United States. Psychiatr. Serv. 60 (10), 1323–1328.

Valuck, R.J., et al., 2004. Antidepressant treatment and risk of suicide attempt by ado- lescents with major depressive disorder: a propensity-adjusted retrospective cohort study. CNS Drugs 18 (15), 1119–1132.

Wang, P., Si, T., 2013. Use of antipsychotics in the treatment of depressive disorders. Shanghai Arch. Psychiatry 25 (3), 134–140.

Williams, D.R., et al., 2007. Prevalence and distribution of major depressive disorder in African Americans, Caribbean blacks, and non-Hispanic whites: results from the National Survey of American Life. Arch. Gen. Psychiatry 64 (3), 305–315.

Windfuhr, K., et al., 2008. Suicide in juveniles and adolescents in the United Kingdom. J. Child Psychol. Psychiatry 49 (11), 1155–1165.

Xu, K.T., Borders, T.F., Arif, A.A., 2004. Ethnic differences in parents’ perception of participatory decision-making style of their children’s physicians. Med. Care 42 (4), 328–335.

Zito, J.M., Safer, D.J., 2001. Services and prevention: pharmacoepidemiology of anti- depressant use. Biol. Psychiatry 49 (12), 1121–1127.

Zuvekas, S.H., Taliaferro, G., 2003. Pathways to access: Health insurance, the health care delivery system, and racial/ethnic disparities, 1996-1999. Health Affairs 22 (2), 139–153.

A. Yucel et al. Journal of Affective Disorders 235 (2018) 155–161

161

http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0030
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0030
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0031
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0031
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0032
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0032
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0032
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0033
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0033
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0034
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0034
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0034
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0035
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0035
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0036
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0036
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0037
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0038
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0038
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0039
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0039
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0040
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0040
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0041
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0041
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0042
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0042
http://archive.samhsa.gov/data/NSDUH/2k7nsduh/2k7Results.htm#8.2.1
https://www.healthypeople.gov/2020/topics-objectives/topic/Access-to-Health-Services
https://www.healthypeople.gov/2020/topics-objectives/topic/Access-to-Health-Services
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0043
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0043
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0044
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0044
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0045
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0045
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0045
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0047
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0047
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0047
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0048
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0048
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0049
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0049
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0049
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0050
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0050
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0051
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0051
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0051
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0052
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0052
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0053
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0053
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0054
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0054
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0055
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0055
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0055
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0056
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0056
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0057
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0057
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0057
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0058
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0058
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0059
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0059
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0059
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0060
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0060
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0061
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0061
http://refhub.elsevier.com/S0165-0327(17)32695-2/sbref0061
Racial/ethnic differences in the treatment of adolescent major depressive disorders (MDD) across healthcare providers participating in the medicaid program
Introduction
Method
Data
Study population
Study design
Exposure