Key points are not available for this paper at this time.
Artificial intelligence (AI) tools allow to extract knowledge from big data and are increasingly used for research purposes applied to -omics, diagnostic images, complex patterns of diseases and system medicine, drug development, robotics, and other topics. The management of big data, largely made of individual clinical data, poses specific ethical challenges that must be addressed in research studies and that should be reflected in the informed consent process. Explaining the mechanisms used by AI algorithms in supporting clinical decision making may be particularly difficult because of the opacity of its process. Moreover, depending on the quality of data feeding their algorithms, AI applications may result in errors. As the General Data Protection Regulation (GDPR) includes the possibility that a patient withdraws his/her informed consent from a study, it may be challenging to update AI algorithms accordingly. On the other hand, AI tools may help support the recruitment and retention of participants in clinical trials matching eligibility criteria with individual data collected for clinical purposes in electronic health records, and improve data collection and analytics. The possibility to stream data from wearable devices offers the possibility to generate large data volumes relevant to Patient Reported Outcomes feeding AI predictive algorithms. The Covid-19 pandemic has promoted the application of digital tools and of AI in clinical trials in order to limit personal contacts. The pressure exerted by the pandemic will possibly speed up the adoption of AI solutions for clinical trials and will highlight their potential ethical implications.
Tozzi et al. (Wed,) studied this question.