EFFECTIVENESS OF ARTIFICIAL INTELLIGENCE-BASED TOOLS IN MONITORING AND REDUCING BURNOUT AMONG HEALTHCARE WORKERS: A SYSTEMATIC REVIEW

Abstract

Objectives: Burnout among healthcare workers adversely affects workforce well-being, healthcare delivery, and patient outcomes. Artificial intelligence (AI) has emerged as a promising approach for monitoring burnout risk and supporting interventions to reduce occupational stress. This systematic review evaluated the effectiveness of AI-based tools in monitoring, predicting, and reducing burnout among healthcare professionals. Methods: A systematic review was conducted according to the PRISMA guidelines. Five electronic databases (Scopus, PubMed, Web of Science, CINAHL, and ProQuest) were searched for studies published between January 2014 and December 2024. Eligible studies involved healthcare professionals and evaluated AI-based approaches for burnout monitoring, prediction, or intervention. Data extraction and quality assessment were performed independently by two reviewers using the NIH Quality Assessment Tool. Results: Twenty-five studies met the inclusion criteria. AI techniques included machine learning algorithms, natural language processing, large language models, wearable-integrated monitoring systems, electronic health record-based predictive tools, and chatbot interventions. Predictive models demonstrated moderate-to-high performance, with reported AUC values ranging from 0.72 to 0.91. Chatbot interventions and EHR-integrated population health management systems showed modest but statistically significant reductions in burnout, whereas AI-assisted documentation tools reduced administrative burden and improved workflow efficiency. However, most studies relied on self-reported outcomes, had limited follow-up periods, and demonstrated considerable methodological heterogeneity. Conclusions: AI-based technologies demonstrate promising potential for identifying burnout risk and supporting interventions to reduce burnout among healthcare workers. Nevertheless, stronger longitudinal studies, standardized evaluation methods, and implementation across diverse healthcare settings are required before widespread adoption can be recommended.