226 - From "Reporter" to "Detective": A Practical Framework for AI-Assisted Evidence Synthesis in Occupational and Environmental Medicine

Session Description: The exponential growth of biomedical literature presents a significant challenge to evidence-based practice in occupational medicine. This session introduces a structured methodology utilizing grounded generative AI (Google's NotebookLM) to enhance the efficiency and rigor of literature review and analysis, ultimately improving evidence-based advice to employees and businesses. The proposed workflow involves three core stages: practitioner-led curation of a verified source corpus to prevent hallucination, iterative interrogation of the data using a hypothetico-deductive approach, and rapid validation of source integrity. This human-in-the-loop framework empowers practitioners to mitigate biases, systematically manage information overload, and deliver robust, transparently-sourced recommendations.

Educational Gaps: Current: Physicians rely on slow, manual literature reviews, struggling with information overload and cognitive bias.
Ideal: Practitioners should be able to use structured, AI-assisted workflows for rapid, targeted, and verifiable evidence synthesis to provide superior, timely, and well-supported advice.

Is this session being sponsored by an ACOEM Section or Component?: no

Has this session previously been presented?: No. Some of the principles of the material to be covered have been outlined in this blog post https://www.paulmcgovern.co.uk/post/how-i-use-ai-for-occupational-health-literature-reviews

The AOHC 2026 theme is Healthy workers. Healthy world: This session directly supports the conference theme, "Healthy workers. Healthy world," by equipping occupational medicine (OM) practitioners with a methodology to translate complex scientific evidence into actionable workplace health strategies more effectively.

The foundation of a healthy worker is clinical and strategic advice grounded in the best available evidence. With over 1.5 million papers published annually, the gap between emerging research and clinical practice is widening. Our proposed AI-assisted framework closes this gap. It allows practitioners to rapidly synthesize data on new chemical exposures, evaluate the efficacy of mental health interventions, or develop return-to-work plans for complex cases with a speed and depth previously unattainable. By ensuring advice is current, comprehensive, and accurately sourced, this method directly enhances a practitioner's ability to protect and improve the health of the individual worker and the entire workforce.

This, in turn, contributes to a healthy world. Healthy workers are the bedrock of productive, resilient, and sustainable communities. When OM practitioners operate at peak efficiency, they are better positioned to advise organizations on creating safer work environments, which reduces the burden on public healthcare systems. Furthermore, this scalable methodology allows for the rapid analysis of global health data, enabling practitioners to contribute more effectively to public health policy and the development of international occupational safety standards. By democratizing access to high-level evidence synthesis, this framework empowers the global OM community to create a world where every worker is safer and healthier.