Gaoyang Liu

Background: Virtual reality (VR) is increasingly used for adjunctive relaxation training in psychiatric care. However, evidence remains limited among hospitalized patients with depressive disorders, particularly in routine inpatient settings in China, and little is known about whether improvement varies by session frequency. Objective: This retrospective study examined whether adjunctive VR-based relaxation training was associated with changes in depressive and anxiety symptoms among inpatients with depressive disorders and whether improvement differed by session frequency. Methods: We conducted a retrospective, nonrandomized natural-group comparison using complete anonymized medical records from patients hospitalized in Lishui Second People’s Hospital between January 1 and December 31, 2022. Patients met International Classification of Diseases, Tenth Revision ( ICD-10 ) diagnostic criteria for depressive episodes or recurrent depressive disorders and were screened using predefined criteria. The analytic sample included 133 inpatients: 63 (47.4%) received adjunctive VR-based relaxation training plus usual care and 70 (52.6%) received usual care only. Usual care included pharmacotherapy and physiotherapy. The VR intervention consisted of 25-minute immersive relaxation sessions delivered approximately 3 times per week. Symptoms were assessed at admission and discharge using the 17-item Hamilton Depression Scale and Hamilton Anxiety Rating Scale. Response was defined as a reduction of 50% or more from baseline, and remission was defined as a total score of 7 or less. Baseline characteristics, outcome scores, response and remission rates, and exploratory session-frequency subgroups were compared. All analyzed variables were checked against complete medical records; no missing values were identified, and no imputation was performed. Results: The VR and control groups did not differ significantly in baseline depressive or anxiety scores. At discharge, adjunctive VR-based relaxation training was associated with lower depressive and anxiety symptom scores than usual care alone. The VR group also showed higher response rates for both depressive and anxiety symptoms and a higher anxiety remission rate, whereas depression remission was similar. Exploratory session-frequency analyses suggested that anxiety improvement may be more consistently associated with VR exposure than depression remission; however, the pattern was not strictly linear and should be interpreted cautiously because treatment frequency was linked to hospitalization duration and routine care factors. Conclusions: This study is innovative in evaluating structured VR-based relaxation training as an adjunct to routine inpatient depression care and in providing preliminary observations on session-frequency patterns in a real-world Chinese psychiatric setting. Unlike many previous VR studies conducted in noninpatient, nonclinical, or short-term experimental contexts, this study reflects everyday clinical practice among hospitalized patients with depressive disorders. The findings contribute practical evidence for integrating immersive relaxation into comprehensive inpatient care, particularly when additional anxiety relief is desired. Because the study was retrospective and nonrandomized, the findings indicate associations rather than causal effects and should be confirmed in prospective randomized controlled trials.

BACKGROUND Nonsuicidal self-injury (NSSI) is common among adolescents with depressive disorders and is associated with substantial clinical burden. Machine learning may assist in identifying multidimensional patterns associated with NSSI, although its incremental value beyond direct clinical assessment remains uncertain. OBJECTIVE This study compares 7 supervised machine learning model variants for classifying past-year NSSI status among adolescents with depressive disorders and conducts detailed analyses of interpretability, sensitivity, subgroup performance, and clinical utility using a random forest model as the focal exploratory model. METHODS We retrospectively identified 437 adolescents aged 10-19 years who received inpatient or outpatient psychiatric care at 4 hospitals in Zhejiang Province, China, between January 2021 and December 2023. A total of 410 eligible participants were included, of whom 274 reported NSSI during the preceding year. The primary analysis included 67 assessment-time sociodemographic, physiological, biochemical, and clinical-behavioral features. The sample was divided into a training set (287/410, 70%) and an untouched held-out test set (123/410, 30%) using stratified random sampling. Hyperparameters were optimized using 5-fold cross-validation within the training set. Performance was evaluated using discrimination, classification metrics, calibration, and decision-curve analysis. Random forest model behavior was examined using Gini importance, permutation importance, and Shapley Additive Explanations (SHAP). RESULTS The radial basis function support vector machine achieved the highest mean cross-validated area under the receiver operating characteristic curve (AUROC) in the training set (0.720), whereas the random forest showed the highest observed AUROC and area under the precision-recall curve (AUPRC) and the lowest Brier score in the held-out test set. The random forest achieved an AUROC of 0.681 (95% CI 0.583-0.778), an AUPRC of 0.808 (95% CI 0.741-0.874), an accuracy of 0.667 (95% CI 0.585-0.748), a sensitivity of 0.720 (95% CI 0.622-0.817), a specificity of 0.561 (95% CI 0.415-0.707), and a Brier score of 0.212 (95% CI 0.188-0.236). Its calibration slope was 1.013. Pairwise AUROC differences between the random forest and alternative models were not statistically significant after Holm correction (Holm-adjusted P values ranged from .17 to .81). The full random forest model did not outperform a suicidal-ideation-only model (AUROC 0.701), whereas excluding suicidal ideation and previous suicide attempt reduced the AUROC to 0.540. A biochemical-only model showed near-chance discrimination (AUROC 0.513). Decision-curve analysis indicated that the random forest provided greater net benefit than the assess-all and assess-none strategies across threshold probabilities of approximately 0.51-0.80. Suicidal ideation showed the most consistent importance across Gini, SHAP, and permutation analyses. CONCLUSIONS The random forest had the highest observed AUROC and AUPRC and the lowest Brier score in the held-out test set among the evaluated algorithms, although the pairwise AUROC differences were not statistically significant after correction for multiple comparisons. Suicidality-related information contributed substantially to model discrimination, whereas routinely collected biochemical variables did not demonstrate clear incremental value. Because this was a retrospective cross-sectional classification study, the findings do not establish prospective risk prediction. External and prospective temporal validation is required before clinical implementation.