Trends in missing responses to suicide-related indicators among U.S. adolescents by race/ethnicity and sex, 2001-2021

Abstract

OBJECTIVES: Missingness to suicide-related survey items may bias adolescent suicide surveillance and limit subgroup comparisons. Prevalence and trends of missingness to suicide-related indicators were examined among U.S. adolescents from 2001 to 2021. STUDY DESIGN: Trend analysis of repeated cross-sectional surveys. METHODS: Data were drawn from a nationally representative sample of U.S. high school students (N = 163,489) participating in the 2001-2021 Youth Risk Behavior Survey. Weighted prevalence of missingness was estimated for four suicide-related indicators (ideation, plan, attempt, injury) in the total sample, and subgroups stratified by race/ethnicity, sex, and both race/ethnicity and sex together. Because ideation and plan missingness were rare, regression-based annual percentage changes (APCs), and average APCs (AAPCs) were estimated for attempt and injury missingness only among the total sample and the same subgroups. RESULTS: Missingness to ideation and plan items was rare (<2%), whereas attempt (11.1%) and injury (13.9%) missingness were higher. Black, Hispanic, and AI/AN adolescents had higher attempt missingness (21.6%, 14.6%, 14.5%) than white adolescents (7.1%). Attempt missingness declined from 2001 to 2005 (APC = -7.99; 95% CI = -11.40, -4.46), then increased 3.91% annually through 2017. Black adolescents lacked a significant post-2017 decline. Injury missingness increased 2.99% annually (AAPC 95% CI = 2.28, 3.71), with a 6.49% annual increase from 2011 to 2021. Injury missingness averaged 9.7% among white adolescents compared to 25.8%, 16.9%, and 19.4% for Black, Hispanic, and AI/AN adolescents. AI/AN adolescents had the highest annual increase in injury missingness (AAPC = 5.82; 95% CI = 2.04, 9.75). Trends were similar by sex. CONCLUSIONS: Suicide attempt and injury missingness were substantial, but disproportionately impacted ethnoracially minoritized adolescents, and may obscure suicide risk. Complete data are essential to strengthen risk assessment and prevention planning.