Basal Cell Carcinoma of Skin, Melanoma (Skin Cancer), Skin Cancer, Squamous Cell Carcinoma Head and Neck Cancer (HNSCC)
Conditions
Keywords
digital phenotyping; psychological distress; social media; natural language processing; transformer; suicide prevention; psycho-oncology
Brief summary
This prospective observational study enrolls 500 skin cancer patients across five Chinese tertiary care centers. The investigators use natural language processing and a hierarchical transformer model to analyse 1.2 million social media posts (Weibo, Douyin, Xiaohongshu, WeChat) for psychological distress and suicide ideation, with prospective validation of an AI Early Warning System.
Detailed description
Skin cancer patients experience high rates of anxiety and depression, yet routine screening is inconsistent. This study collects social media content from 500 patients over 24 months (12 months retrospective + 12 months prospective). Linguistic markers are identified using a custom Skin Cancer Linguistic Inquiry and Word Count (SC-LIWC) dictionary developed through expert consensus. A hierarchical transformer network with multi-head self-attention is built, comprising word-level (RoBERTa-wwm-ext), post-level (BiGRU+attention), and user-level (8-head self-attention) encoders with platform embeddings to address cross-platform heterogeneity. The model predicts suicide ideation risk (C-SSRS score ≥2) within 12 months, achieving AUC=0.91 in validation. An AI Early Warning System (EWS) is prospectively tested: alerts are triggered at threshold 0.38, reviewed by psychologists (12.4 min/alert), and interventions (crisis hotline, online counselling, psychoeducational materials) are automatically delivered. All data collection complies with China's Personal Information Protection Law (PIPL) and uses a 5-layer de-identification protocol (direct identifier removal, SHA-256 hashing, NER redaction, timestamp shifting, role-based access control). The study also examines cross-platform performance, optimal observation windows, and false-positive patterns to guide refinement
Interventions
None listed
Sponsors
West China Hospital
Study design
Observational model
COHORT
Time perspective
PROSPECTIVE
Eligibility
Sex/Gender
ALL
Age
18 Years to 80 Years
Healthy volunteers
No
Inclusion criteria
1. Age ≥ 18 years 2. Histologically confirmed skin cancer within the preceding 60 months 3. Active user of at least one target social media platform (Weibo, Douyin, Xiaohongshu, or WeChat Channels) 4. Willingness to provide informed consent for social media data collection 5. Ability to read and write in Chinese
Exclusion criteria
1. Pre-existing major psychiatric disorder diagnosed prior to cancer diagnosis 2. Cognitive impairment preventing informed consent 3. Inability to access social media accounts 4. Concurrent participation in other digital mental health interventions
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Incidence of Suicide Ideation Risk Assessed by the Columbia-Suicide Severity Rating Scale (C-SSRS) | 24 months | Proportion of participants with C-SSRS score ≥ 2 (active suicidal ideation with some intent to act, without specific plan) at any follow-up assessment. The C-SSRS is a clinician-administered instrument with a severity subscale ranging from 0 (no suicidal ideation) to 5 (active suicidal ideation with specific plan and intent). Higher scores indicate greater suicide risk. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Depression Severity Assessed by the Patient Health Questionnaire-9 (PHQ-9) | 24 months | Mean change in PHQ-9 total score from baseline to each follow-up assessment. The PHQ-9 is a 9-item self-report questionnaire measuring depressive symptoms over the past 2 weeks. Total scores range from 0 to 27, with higher scores indicating greater depression severity (moderate-to-severe: ≥10). |
| Anxiety Severity Assessed by the Generalized Anxiety Disorder-7 (GAD-7) | 24 months | Mean change in GAD-7 total score from baseline to each follow-up assessment. The GAD-7 is a 7-item self-report questionnaire measuring generalized anxiety symptoms over the past 2 weeks. Total scores range from 0 to 21, with higher scores indicating greater anxiety severity (moderate-to-severe: ≥10). |
| Psychological Distress Assessed by the Hospital Anxiety and Depression Scale (HADS) | 24 months | Mean change in HADS total score (sum of Anxiety and Depression subscales) from baseline to each follow-up assessment. The HADS is a 14-item self-report questionnaire with two 7-item subscales (Anxiety and Depression), each ranging from 0 to 21. Higher subscale scores indicate worse psychological distress (subscale ≥15 indicates clinically significant distress). |
| Rate of Psychological Resource Utilization Following AI Early Warning System Alerts | 24 months | Proportion of high-risk participants (C-SSRS ≥ 2) who access at least one psychological resource (crisis hotline, online counseling link, or psychoeducational material) following an automated alert from the AI Early Warning System, compared to a matched historical control group (pre-EWS period). |
Countries
China
Outcome results
None listed