<![CDATA[Occupational safety and health is a critical component in protecting workers and sustaining organizational operations, particularly in high-risk sectors such as energy, construction, and manufacturing. Conventional reactive approaches to occupational safety and health are often insufficient to prevent workplace accidents. Digital transformation has enabled the adoption of predictive analytics as a proactive strategy, utilizing historical data and machine learning algorithms to identify risks before incidents occur. This study aimed to systematically review the application of predictive analytics for the early detection of occupational accidents, mapping the methods, algorithms, and implementation outcomes across various industries. A scoping review was conducted following the PRISMA-ScR framework. Articles published between 2015 and 2025 were retrieved from IEEE Xplore, ScienceDirect, and PubMed. From 264 identified records, 9 relevant empirical studies were included in the final analysis. The review indicates that commonly applied algorithms, including Random Forest and Support Vector Machine, demonstrate high predictive performance in identifying accident risks, ergonomic hazards, and worker compliance, with accuracy levels reaching 90% or higher. The implementation of predictive analytics was associated with a reduction in workplace accidents of up to 25% and contributed to improved safety management through early warning systems. This systematic review underscores the significance of predictive analytics in transforming occupational safety and health management from a reactive to a proactive approach. The integration of big data, Internet of Things, and artificial intelligence supports the development of data-driven systems for effective accident prevention and provides practical recommendations for researchers and occupational safety and health practitioners to adopt predictive models within sustainable prevention strategies.Keywords: predictive analytics; early detection; occupational accidents; machine learning]]>
