Abstract
BACKGROUND: While the HTA community appears well-equipped to assess preventive and diagnostic technologies, certain limitations persist in evaluating technologies designed for rare diseases, including those based on Artificial Intelligence (AI). In Europe, the EUnetHTA Core Model(®) serves as a reference for assessing preventive and diagnostic technologies. This study aims to identify key ethical, legal, and social issues related to AI-based technologies for the prevention and diagnosis of rare diseases, proposing enhancements to the Core Model. METHODS: An exploratory sequential mixed methods approach was used, integrating a PICO-guided literature review and a focus group. The review analyzed six peer-reviewed articles and compared the findings with a prior study on childhood melanoma published in this journal (Healthcare), retaining only newly identified issues. A focus group composed of experts in ethical, legal, and social domains provided qualitative insights. RESULTS: Thirteen additional issues and their corresponding questions were identified. Ethical concerns related to rare diseases included insufficient disease history knowledge, lack of robust clinical data, absence of validated efficacy tools, overdiagnosis/underdiagnosis risks, and unknown ICER thresholds. Defensive medicine was identified as a legal issue. For AI-based technologies, concerns included discriminatory outcomes, explicability, and environmental impact (ethical); accountability and reimbursement (legal); and patient involvement and job losses (social). CONCLUSIONS: Integrating these findings into the Core Model enables a comprehensive HTA of AI-based rare disease technologies. Beyond the Core Model, these issues may inform broader assessment frameworks, ensuring rigorous and ethically responsible evaluations.