Use of artificial intelligence to prevent aggressions against health professionals

Abstract

The alarming rise in assaults against healthcare professionals is a public health and occupational issue that threatens staff well-being and care quality. Violence in this sector includes physical, verbal, and psychological aggression, posing a serious risk. Four main types of workplace violence in healthcare have been identified: External violence with no prior relationship (Type I), violence by patients against professionals (Type II, the most frequent), internal or institutional violence (Type III), and personal violence (Type IV). This issue is global, with an increasing trend and significant underreporting. Its consequences are severe at multiple levels: individually (burnout, anxiety, depression), institutionally (absenteeism, staff turnover), and in patient care quality. Artificial intelligence (AI) has emerged as a promising tool to prevent and mitigate such violence. Its applications include surveillance and monitoring systems, enhanced communication between staff and patients, workflow optimization, staff training, and predictive analysis of potentially aggressive patients. However, AI implementation presents ethical challenges related to data protection, privacy, bias risks, prediction reliability, and potential dehumanization. Addressing these concerns is crucial to ensuring safe and equitable AI use, always under human supervision. Effective prevention requires a comprehensive approach that integrates technology with organizational and educational measures.