Optimal Cut-Offs for Digital Addiction Scales as Preliminary Screening Proxies for Psychological Distress Symptoms Among Vietnamese Medical Students: A Single-Centre ROC Curve Analysis

Digital technologies are central to student life, but evidence is limited on whether digital-use scales can provide useful thresholds for identifying psychological distress among medical students. This single-center cross-sectional study analyzed complete data from 600 undergraduates at a medical university in southern Vietnam. Stress, depression, and anxiety symptoms were classified using the DASS-21, while problematic smartphone use, social media use, and gaming-related problems were assessed using the SABAS, BSMAS, and a six-item GAS adaptation, respectively. Receiver operating characteristic analyses and the Youden index were used to estimate discrimination and candidate cut-offs. Non-normal DASS-21 scores were observed for stress (30.4%), depression (37.8%), and anxiety (39.0%), predominantly within the mild range. For stress, the AUCs and cut-offs were 0.75 (95% CI 0.70–0.80; ≥23.5) for SABAS, 0.67 (0.62–0.71; ≥17.5) for BSMAS, and 0.64 (0.60–0.70; ≥15.5) for GAS. For depression, the corresponding values were 0.62 (0.57–0.66; ≥21.5), 0.68 (0.63–0.73; ≥18.5), and 0.60 (0.55–0.65; ≥14.5). For anxiety, they were 0.60 (0.55–0.64; ≥20.5), 0.62 (0.57–0.67; ≥17.5) and 0.55 (0.51–0.60; ≥15.5). Sensitivity was low across thresholds (35.0–59.0%). These candidate cut-offs showed modest discrimination and should be considered only as preliminary behavioral flags prompting validated mental health assessment, not as stand-alone screening or diagnostic criteria.