Calibration, Validation and Utility of ARTEMIS‐A: A Web Application for Early Identification of Mental Health Difficulties in Secondary Schools

BACKGROUND: Early identification can accelerate access to effective support. We aimed to calibrate and validate ARTEMIS-A, a computerised adaptive assessment of mental health difficulties for young people (YP), and to explore its acceptability and feasibility. METHOD: The ARTEMIS-A item bank comprises 175 questions from established standardised mental health measures. Calibration involved pupils aged 11-16 years completing overlapping sets of 40 questions. Convergent validity was evaluated by comparing calibrated ARTEMIS-A scores with the Development and Well-Being Assessment (DAWBA). Online interviews with YP, parents and school staff explored perceived benefits and harms. RESULTS: Of 3183 recruited pupils from nine schools, 2276 provided complete calibration data. Then 103 YP from six schools completed both ARTEMIS-A and the DAWBA. ARTEMIS-A administered a median of four questions per participant (min 2, max 8), with a median completion time of 67 seconds. Convergent validity with clinically rated DSM 5 DAWBA diagnoses was good; correlation was 0.55 the area under the curve was 0.81. Parents and YP considered ARTEMIS-A acceptable and feasible, highlighting its speed, simplicity, and device flexibility. CONCLUSIONS: ARTEMIS-A is potentially an efficient, flexible and cost-effective method of early identification of secondary school pupils who may require additional support for their mental health.