RESUMO
BACKGROUND: For patients with self-harm behaviors, the urge to hurt themselves persists after hospital discharge, leading to costly readmissions and even death. Hence, postdischarge intervention programs that reduce self-harm behavior among patients should be part of a cogent community mental health care policy. OBJECTIVE: We aimed to determine whether a combination of a self-help mobile app and volunteer support could complement treatment as usual (TAU) to reduce the risk of suicide among these patients. METHODS: We conducted a pragmatic randomized controlled trial on discharged patients aged between 18 and 45 years with self-harm episodes/suicide attempts, all of whom were recruited from 4 hospital emergency departments in Hong Kong. Participants were randomly assigned to one of three groups: (1) mobile app + TAU ("apps"), (2) mobile app + volunteer support + TAU ("volunteers"), or (3) TAU only as the control group ("TAU"). They were asked to submit a mobile app-based questionnaire during 4 measurement time points at monthly intervals. RESULTS: A total of 40 participants were recruited. Blending volunteer care with a preprogrammed mobile app was found to be effective in improving service compliance. Drawing upon the interpersonal-psychological theory of suicide, our findings suggested that a reduction in perceived burdensomeness and thwarted belongingness through community-based caring contact are linked to improvement in hopelessness, albeit a transient one, and suicide risk. CONCLUSIONS: A combination of volunteer care with a self-help mobile app as a strategy for strengthening the continuity of care can be cautiously implemented for discharged patients at risk of self-harm during the transition from the hospital to a community setting. TRIAL REGISTRATION: ClinicalTrials.gov NCT03081078; https://clinicaltrials.gov/study/NCT03081078.
RESUMO
RATIONALE: Detecting users at risk of suicide in text-based counseling services is essential to ensure that at-risk individuals are flagged and prioritized. OBJECTIVE: The objective of this study is to develop a domain knowledge-aware risk assessment (KARA) model to improve our ability of suicide detection in online counseling systems. METHODS: We obtained the largest known de-identified dataset from an emotional support system established in Hong Kong, comprising 5682 Cantonese conversations between help-seekers and counselors. Of those, 682 conversations disclosed crisis intentions of suicide. We constructed a suicide-knowledge graph, representing suicide-related domain knowledge as a computer-processible graph. Such knowledge graph was embedded into a deep learning model to improve its ability to identify help-seekers in crisis. As the baseline, a standard NLP model was applied to the same task. 80% of the study samples were randomly sampled to train model parameters. The remaining 20% were used for model validation. Evaluation metrics including precision, recall, and c-statistic were reported. RESULTS: Both KARA and the baseline achieved high precision (0.984 and 0.951, shown in Table 2) and high recall (0.942 and 0.947) towards non-crisis cases. For crisis cases, however, KARA model achieved a much higher recall than the baseline (0.870 vs 0.791). The c-statistics of KARA and the baseline were 0.815 and 0.760, respectively. CONCLUSION: KARA significantly outperformed standard NLP models, demonstrating good translational value and clinical relevance.
