Machine learning in the EU health care context: exploring the ethical, legal and social issues

Abstract

Diagnosis and clinical decision-making based on Machine Learning technologies are showing significant advances that may change the functioning of our health care systems. They promise more effective and efficient healthcare at a lower cost. Even though evidence suggests that all these promises have yet to be demonstrated in clinical practice, it is undeniable that these technologies are already re-signifying the relationships on the health care landscape, particularly in the physician-patient relationship, which we can already redefine as a ‘physician-computer-patient relationship’. This new scenario is undoubtedly promising, but it also poses some fundamental issues that need an urgent answer. An inappropriate use of Machine Learning might involve a dramatic loss in the patients’ rights to informed consent or possible discrimination reflecting their personal circumstances. Unfortunately, the traditional principles incorporated by medical law are insufficient to face this challenge. Our most recent regulatory framework, defined by the General Regulation on Data Protection, might be useful in order to avoid this scenario since it includes the right not to be subject to a decision based solely on automated processing. In this paper, however, we argue that this legal tool is adequate but not sufficient to address the legal, ethical and social challenges that Machine Learning technologies pose to patients’ rights and health care givers’ capacities. Therefore, further development of the regulation on this topic and the development of new actors such as the Health Information Counsellors, will be necessary.

Ähnliche Arbeiten