Correlates of psychiatric hospitalization in a mood-disorder-focused youth psychiatric cohort: a real-world electronic medical record study

Abstract Background Psychiatric hospitalization is a high-intensity level of care for youths with mood-related psychiatric conditions, but real-world service-utilization data from Asian tertiary-care settings remain limited. This study examined correlates of psychiatric hospitalization in a Taiwanese youth psychiatric electronic medical record cohort and described all-cause emergency department utilization as contextual information regarding broader acute healthcare contact. Methods We conducted a retrospective patient-level, period-based observational study using structured electronic medical record data from a tertiary medical center in Taiwan. The approved institutional extraction covered records from January 1, 2017 to March 31, 2023. The analytic sample included 511 patients aged 18 years or younger with eligible mood disorder diagnoses and a small personality disorder subgroup. Psychiatric hospitalization was the principal outcome. Multivariable logistic regression included demographic and diagnostic variables, index-encounter medication exposures, psychotherapy-related service exposure, and observation duration. All-cause emergency department utilization was summarized descriptively. Results The mean age was 15.7 years, and 355 patients (69.5%) were female. Psychiatric hospitalization occurred in 76 patients (14.9%), and lithium exposure at the index qualifying encounter was recorded in 28 patients (5.5%). Index-encounter antipsychotic exposure (OR = 3.03, 95% CI: 1.65–5.57), anticonvulsant mood-stabilizing medication exposure (OR = 3.25, 95% CI: 1.16–9.09), and lithium exposure (OR = 10.64, 95% CI: 3.76–30.12) were associated with hospitalization. Psychotherapy-related service exposure was also associated with hospitalization (OR = 7.00, 95% CI: 3.65–13.42). Age, sex, broad diagnostic category, and benzodiazepine anxiolytic exposure were not significantly associated with hospitalization. All-cause emergency department utilization occurred in 14 patients (2.7%); 12 also experienced psychiatric hospitalization. Conclusions Psychiatric hospitalization was more consistently associated with index-encounter treatment characteristics and period-based service contact than with broad demographic or diagnostic categories. Medication associations should be interpreted as markers of treatment and clinical complexity rather than causal medication effects, while psychotherapy-related service exposure reflects care intensity without established temporal ordering. All-cause emergency department utilization provided descriptive acute-care context rather than an independent psychiatric outcome.