Suicide Risk Screening: A New Method to Identify At-Risk Students from General College Populations

University-wide suicide risk screening is essential but difficult to implement at scale because professional counselling resources are limited and conventional tools often show low predictive value in low-prevalence populations. To support transparent and clinically interpretable campus mental health workflows, this study introduces and validates the Composite Key Conditions (CKC) method, an interpretable, computationally efficient screening approach for identifying students who match counsellor-defined suicide risk categories. A multi-phase, multi-site study involving 8819 students from two universities in the development and cross-site validation phases was conducted, followed by independent external validation at a third university. In the initial wave, students completed self-report questionnaires in November 2023, after which approximately 20% were selected and invited for centralized counsellor evaluation; the assessments were conducted in December 2023. From January to June 2024, 320 students who subsequently attended psychological services (103 at University A and 217 at University B) were assessed through the cohort pathway. CKC was further evaluated in October 2025 using an independent sample of 22,205 students from a third university. The CKC method, developed from the initial phase, correctly classified all students assigned to the suicide risk category during centralized assessment (sensitivity = 1.000, specificity = 0.901). When applied across both centralized and cohort assessments (yielding 89 counsellor-defined suicide risk cases), the method maintained high performance (sensitivity = 0.888, specificity = 0.904, PPV = 0.086). In the independent validation sample, the method correctly classified 17 of the 22 students identified by the internal reporting system as meeting its high-risk criteria, with a specificity of 0.923. CKC provides an interpretable and efficient approach to suicide risk screening that is suitable for large-scale implementation in university settings. By producing transparent, auditable risk rules rather than opaque predictions, CKC is positioned to complement (not replace) counsellor judgement and to support timely triage and referral within campus mental health service systems.