Recent Trends in School-Based Mental Health Services Among Low-Income and Racial and Ethnic Minority Adolescents

Abstract

This survey study uses publicly available data from the National Survey on Drug Use and Health to evaluate school-based mental health (MH) service use among low-income and racial and ethnic minority adolescents.

In the US, youth mental health (MH) has worsened in recent years and with the COVID-19 pandemic, causing 3 health care professional organizations that serve this population to jointly declare a youth MH crisis.^1,2 Moreover, only half of youths with MH needs receive any treatment.³ Delivery of MH services in schools is a promising approach to improving MH care access, including among underserved populations.⁴

After the 2018 Parkland, Florida, school shooting, public and policy maker focus on school MH intensified in many states and communities.⁵ We evaluated whether school MH service use increased among US adolescents in 2019, following this intensified focus on school MH.

Methods

In this survey study, we used 2009-2019 public-use data from the National Survey on Drug Use and Health (NSDUH) to evaluate trends in self-reported receipt of any past-year school MH services among adolescents aged 12 to 17 years. The primary outcome captures individuals who reported talking to school social workers, school psychologists, or school counselors about problems; attending a special school for students with emotional/behavioral problems; and/or participating in a school program for students with emotional/behavioral problems in the past year. We also evaluated MH service use: (1) in nonschool settings, and (2) in schools only vs in both school and nonschool settings.⁴ The institutional review board of Emory University (Atlanta, Georgia) determined this study was not human participants research because it used publicly available, previously deidentified data, and thus informed consent was not required. This study adhered to STROBE reporting guideline.

We tested outcomes’ equivalence across years using Pearson tests. We fit logistic regression models with year-specific indicators after adjustment for predisposing, enabling, and need-related factors (Table). Trends were analyzed overall and stratified by race and ethnicity and family income to evaluate differences among underserved youth.

Table. Estimated Probability of School Mental Health Service Use Among Adolescents (2018-2019), Overall and Stratified by Family Income and Race and Ethnicity^a.

Groupb	No. (2009-2019)	Estimated probability (95% CI)		Adjusted OR, 2019 vs 2018 (95% CI)
		2018	2019
All adolescents	165 686	12.0 (11.3 to 12.8)	13.4 (12.6 to 14.2)	1.13 (1.04 to 1.23)
Family income
<100% FPL	34 702	14.2 (12.6 to 15.9)	18.0 (15.6 to 20.4)	1.32 (1.13 to 1.55)
100%-200% FPL	37 160	12.1 (10.5 to 13.7)	12.5 (10.8 to 14.1)	1.04 (0.82 to 1.30)
>200% FPL	93 824	11.2 (10.3 to 12.1)	12.2 (11.4 to 13.0)	1.10 (0.98 to 1.23)
Race and ethnicity
Hispanic	33 404	11.9 (10.5 to 13.3)	13.4 (11.5 to 15.3)	1.14 (0.94 to 1.39)
Non-Hispanic Asian	5972	9.3 (5.8 to 12.7)	9.3 (5.8 to 12.8)	1.01 (0.57 to 1.79)
Non-Hispanic Black	22 065	14.7 (12.7 to 16.6)	18.1 (16.0 to 20.2)	1.28 (1.02 to 1.61)
Non-Hispanic White	92 670	11.6 (10.7 to 12.6)	12.4 (11.5 to 13.4)	1.08 (0.97 to 1.20)
Otherc	11 575	13.4 (10.4 to 16.3)	16.8 (13.8 to 19.8)	1.31 (0.96 to 1.78)

Abbreviations: FPL, federal poverty level; OR, odds ratio.

^a Estimated probabilities and adjusted ORs estimated from logistic regression models of past-year school mental health use (2009-2019), adjusting for year indicator variables (2019 [reference]) as well as predisposing, enabling, and need-related factors. Predisposing factors included race and ethnicity (self-selected from a list with the following options: Hispanic, Non-Hispanic Asian, Non-Hispanic Black, Non-Hispanic White, other; omitted from models stratified by race and ethnicity), age (12-13 years, 14-15 years, 16-17 years), and female sex (male [reference]). Enabling characteristics included family income (<100% FPL, 100%-199% FPL, ≥200% FPL; omitted from models stratified by family income), health insurance status (any private, Medicaid or Children’s Health Insurance Program, uninsured, other), whether the adolescent lives with 2 parents (yes or no), and metropolitan area status of county of residence (large, small, nonmetropolitan). Need-related characteristics included indicators of self-rated health (fair or poor vs excellent, very good, or good), past-year mental health problems (ie, major depressive episode, suicidal thoughts or attempt), and substance use disorders (ie, illicit drug or alcohol dependence, illicit drug or alcohol abuse).

^b Number of adolescents by sex: 81 009 (48.9%) female and 84 677 (51.1%) male; by age group: 51 474 (31.0%) aged 12 and 13 years, 56 365 (34.2%) aged 14 and 15 years, and 57 847 (34.8%) aged 16 and 17 years.

^c Other race and ethnicity included American Indian or Alaska Native, Guamanian or Chamorro, Native Hawaiian, Samoan, other Pacific Islander, and other race.

To generate nationally representative estimates, weighted percentages and 95% CIs were estimated using logit transformation to adjust for complex survey design elements. Data were analyzed using Stata, version 16.1 (StataCorp LLC). Statistical significance was defined as P < .05 using 2-sided tests.

Results

The sample included 165 686 adolescents (the Table provides demographic characteristics). In this sample, the unadjusted use of any school MH services increased from 13.8% (95% CI, 13.0%-14.6%) in 2018 to 15.7% (95% CI, 14.8%-16.6%) in 2019 (relative increase, 13.6%; 95% CI, 5.4%-21.9%; P < .001) (Figure, A). Use of MH services in schools only increased from 7.5% (95% CI, 7.0%-8.0%) in 2018 to 8.9% (95% CI, 8.2%-9.5%) in 2019 (relative increase, 18.1%; 95% CI, 6.0%-30.3%; P = .003), whereas use in both school and nonschool settings did not change significantly. No single-year increases were observed in nonschool settings (Figure, B).

Figure. Use of Mental Health (MH) Services in School and Nonschool Settings Among US Adolescents Aged 12 to 17 Years, Weighted Estimates 2009-2019.

A, Use of MH services in schools reflects individuals reporting that they talked to school social workers, school psychologists, or school counselors for problems; attended a special school for students with emotional or behavioral problems; and/or participated in a school program for students with emotional/behavioral problems in the past year. B, Separate survey measures in the National Survey on Drug Use and Health were used to identify any MH service (eg, treatment and counseling for problems with behaviors or emotions that were not caused by alcohol or drugs) in the past year in each of the 5 nonschool settings represented. Error bars represent the 95% CI.

^aDay treatment facility, MH clinic, private therapist, and in-home therapist.

^bHospital or residential treatment facility.

In adjusted models, the odds of receiving any school MH services were greater in 2019 than each year from 2009 to 2018 (2019 vs 2018: odds ratio [OR], 1.13; 95% CI, 1.04-1.23) (Table). Findings were similar in schools only (2019 vs 2018: OR, 1.17; 95% CI, 1.05-1.31). In stratified models, the 2019 increase was pronounced among Black adolescents (2019 vs 2018: OR, 1.28; 95% CI, 1.02-1.61) and adolescents with family income less than 100% of the federal poverty level (2019 vs 2018: OR, 1.32; 95% CI, 1.13-1.55).

Discussion

Coincident with heightened national discourse on school MH, this survey study reveals that school MH service use increased among adolescents in 2019 vs 2018 and prior years. The 2019 increase in school MH service use was significant among adolescents who used MH services in school only. Furthermore, the increase in school MH service use was most pronounced among Black adolescents and adolescents in low-income families.

A study limitation is potential bias inherent in survey data (eg, social desirability bias), although NSDUH interviewers use methods that encourage accurate reporting.⁶ Furthermore, without geographic identifiers, our analysis cannot establish causal linkages between changes in specific state or local policy during 2018-2019 and school MH service use.

Evidence is needed about the mechanisms that link policies, school-based MH systems, and access to services. Such evidence will guide leaders in extending the 2019 gains in youth MH care access into the postpandemic US.