International Consensus Framework for Ethical Integration of Patient-Reported Outcomes and Digital Biomarkers in AI Healthcare Systems (Preprint)

Abstract

<sec> <title>BACKGROUND</title> Alongside expected benefits, several ethical concerns arise from Artificial Intelligence (AI) based models. From the design to the implementation and subsequent evaluation, it is crucial to map potential ethical concerns regarding the use of AI models in healthcare. Patient-Reported Outcomes (PROs) and Digital Biomarkers (DBs) are being increasingly collected to improve patient-centered healthcare systems. However, due to the sensitive nature of this data, its processing into AI models may raise ethical concerns that should be considered. While general AI ethics frameworks exist, no international consensus has specifically addressed the unique ethical challenges of integrating PROs and DBs in AI healthcare models. </sec> <sec> <title>OBJECTIVE</title> This study aims to fill this critical gap by establishing the first international expert consensus on ethical, legal, and social considerations specifically for PROs and DBs integration in AI-based healthcare models. </sec> <sec> <title>METHODS</title> A mixed-method study was performed. Initially, a narrative review was performed to identify the main ethical considerations regarding integrating PROs and DBs in AI-based models in healthcare. Then, a modified two-round online Delphi study was conducted to reach a consensus on the recommendations for integrating PROs and DBs in AI-based models, among selected international experts. </sec> <sec> <title>RESULTS</title> The findings of the two complementary components of this study (narrative review and modified Delphi study) were organized around five core ethical principles: autonomy, beneficence, non-maleficence, justice, and transparency and accountability. The modified Delphi study achieved high consensus (≥80%) on 57 specific recommendations across these principles. Key recommendations included implementing dynamic consent models, establishing continuous model validation protocols, conducting regular impact assessments, ensuring diverse stakeholder engagement to mitigate biases, and maintaining human oversight within AI systems. </sec> <sec> <title>CONCLUSIONS</title> This study provides the first comprehensive, internationally validated ethical framework specifically designed for PROs and DBs integration in AI healthcare systems, filling a critical gap in the literature that has primarily focused on general AI ethics rather than the unique challenges posed by patient-generated health data. </sec>

Ähnliche Arbeiten