From Referral to Connection: Presentation Patterns & Service Pathways In School Mental Health Referrals

Background and Purpose

Schools are a major access point for youth mental health care, yet referral systems are often framed as if students present primarily with either internalizing or externalizing concerns. This framing may obscure mixed and less clearly categorized presentations and may limit understanding of how students are connected to care after referral. This study examined: (1) how referrals were distributed across internalizing-only, externalizing-only, mixed, and neither presentations, and whether those patterns differed by demographic characteristics, school characteristics, and acute clinical risk; and (2) whether those groups were equally likely to be connected to a service pathway after referral.

Methods

This quantitative archival study analyzed 811 referral-level records from a public school district behavioral and mental health program during the 2024–2025 academic year. Referral presentations were coded from symptom checklist indicators and reviewed referral information into four mutually exclusive groups: internalizing-only, externalizing-only, mixed, and neither. Post-referral outcomes were coded from closure data as connected versus not connected to a service pathway; a secondary descriptive variable distinguished district-provided from outside/non-district pathways. Analyses included descriptive statistics, age comparisons, chi-square tests, multinomial logistic regression for referral presentation patterns, and binary logistic regression for service connection.

Results

For RQ1, referrals were distributed as 35.4% internalizing-only, 15.3% externalizing-only, 41.4% mixed, and 7.9% neither. Mean age differed significantly across groups, with internalizing-only referrals older than externalizing-only, mixed, and neither referrals. In adjusted multinomial models, externalizing-only rather than internalizing-only referrals were 73% less likely among girls than boys, 5.73 times more likely in elementary school and 3.61 times more likely in middle school than in high school, 3.87 times more likely among Black/African American students than Hispanic/Latino students, and 59% less likely when acute safety risk was present. Mixed rather than internalizing-only referrals were 56% less likely among girls, 2.83 times more likely among Black/African American students than Hispanic/Latino students, and 65% more likely when acute safety risk was present. For RQ2, 77.8% of referrals with resolved closure data were connected to a service pathway. Descriptively, connection rates were 75.7% for internalizing-only, 70.7% for externalizing-only, 83.8% for mixed, and 69.5% for neither referrals. In adjusted models, referral presentation group was not a clear independent predictor of connection. Instead, elementary referrals were 3.47 times more likely and middle school referrals were 77% more likely than high school referrals to be connected, referrals involving acute safety risk were 2.84 times more likely to be connected, and White students had 57% lower odds of connection than Hispanic/Latino students.

Conclusions and Implications

School mental health referrals were not adequately characterized by a simple internalizing-versus-externalizing distinction. Mixed presentations were the largest referral group, and referral patterns varied by gender, school level, ethnicity, and acute risk. After referral, connection to care appeared to depend more on school level and acute safety risk than on presentation group alone. Findings support more nuanced referral classification, stronger tracking of post-referral outcomes, and tighter coordination between school-based and community-based pathways to improve triage, equity, and continuity of care in school social work systems.