Depressive Symptoms, Emergency Care, and School Climate: An Ecological Analysis of Linked Administrative and Survey Data in New York City

TL;DR

This study explores how school climate and district policies affect adolescent mental health and emergency care use in New York City.

Contribution

The study introduces an emergency department visit intensity measure to assess depressive symptoms and school climate effects.

Findings

Depressive symptoms are strongly linked to individual and school climate indicators across all years.

There are large differences in depressive symptoms and ED use by sex.

School and district-level interventions may improve mental health outcomes for adolescents.

Abstract

As rates of depression and anxiety continue to rise, prevention, and treatment of poor mental health in adolescents is a major challenge for population health. Within the US context, a growing body of literature is examining the relationship between school climate and student mental health.

We extend the notion of school climate and mesosystem effects by creating an emergency department visit intensity, a novel indirect summary measure related to both the prevalence of depressive symptoms and district policy/resources. The intensity measure is linked to 3 years of the New York City Youth Risk Behavior Survey (2015, 2017, 2019), and we test the intensity measure in three models sequentially constructed from individual‐, school‐, and district‐level covariates across survey years with district‐level fixed effects.

We find strong evidence for a relationship between the prevalence of self‐reported depressive symptoms and individual and school climate indicators across all years, and document large differences in the prevalence of depressive symptoms and ED utilization by sex.

Within the US context, adolescents spend much of their days in a school environment; our findings suggest that interventions could focus on female adolescents' experiences as well as the school (and district) to improve school climate and depressive symptoms indirectly.

Figures