Correction to: Quality of life in children and adults with epidermolysis bullosa: the QoL-REB explorative study.

Social determinants of health (SDoH) account for 80% of health outcomes; yet, systematic collection of SDoH data remains inconsistent across health care settings [,]. Emergency departments, primary care clinics, and other health care environments frequently encounter patients with unmet social needs, including housing instability, food insecurity, transportation barriers, and financial hardship []. However, time constraints and workflow pressures create significant barriers to comprehensive SDoH data collection []. Development of a chatbot to capture social needs in the clinical environment has not been described. This use case involves collecting inherently sensitive information, often from vulnerable populations [,], which presents unique challenges requiring innovative methods of development and evaluation. Patients who disclose housing instability, substance use, intimate partner violence, or financial hardship may be experiencing trauma, stigma, or fear of consequences []. Therefore, effective SDoH chatbots must demonstrate cultural humility, recognize signs of distress, maintain appropriate boundaries between data collection and clinical care, and avoid perpetuating health care disparities through biased responses []. Accordingly, the present work offers a formative feasibility evaluation process using synthetic scenarios to refine the tool and rubric before any real-patient implementation; it does not establish effectiveness or safety under real-world conditions such as time pressure, low trust, low literacy, or emotional distress. Given the sensitive nature of SDoH conversations and the vulnerability of many patients with social needs, it is essential that patient-facing chatbots designed for this purpose undergo rigorous evaluation before deployment in real-world clinical environments. Deploying untested chatbots with patients risks unintended harm and can undermine patient trust and engagement with health care providers. However, methods of evaluation for this purpose must also be sufficiently feasible for health systems to incorporate into routine operations and clinical workflows.