The algorithmic prevention: State of the knowledge and perspective

Algorithmic prevention can be defined by three criteria: (i) a collection of data on genetic, socio-economic, and behavioural factors of a citizen or patient; (ii) a prediction by algorithmic processing of the risks of disease or complications; (iii) personalised monitoring to prevent what can be prevented. This new form of prevention applies to primary prevention (before the disease occurs), secondary prevention (when the first symptoms appear), and tertiary prevention (before the onset of disease complications). It has been observed in various fields for several years, including oncology, neurology, mental health, geriatrics, and infectious disease risk, among others. Although the evidence remains limited, two trends are emerging: the satisfaction of those involved in the programmes and a reduced reliance on hospitalisation in tertiary prevention cases. Several challenges need to be addressed before considering wider application: those concerning data collection and processing, and those related to the development of these prevention programmes, particularly the organisational implications for healthcare systems and social protection. In particular, questions about adoption arise from both doctors and patients. Despite these reservations, algorithmic prevention is likely to play a role in the future, with its complementary relationship to traditional approaches in the general population becoming increasingly evident.