Integrating AI Scribe Technology into Public Health Workflows: Simulation and Field Evaluation in Disease Case...

The growing integration of generative artificial intelligence (GenAI) into clinical documentation offers new opportunities to enhance public health surveillance and response. This pilot study evaluated an AI scribe for infectious disease management in two Public Health Units (PHUs) in Ontario, Canada, using a two-phase evaluation. Results show meaningful reductions in administrative workload and positive perceptions of usability and efficiency. In simulation, the scribe consistently captured clinically relevant content and filtered irrelevances. At SMDHU, scribe use was associated with a mean reduction of 12.9 minutes in documentation time compared with usual practice (p < 0.05), whereas at WDGPH it was associated with a mean increase of 10.4 minutes (p < 0.05). Key challenges included deployment delays due to technical problems, standardized PHU-specific note generation, and trust in AI-generated documentation. This study extends AI-scribe research beyond clinical care, offering one of the first empirical evaluations of AI-assisted documentation for infectious disease surveillance. Future work should focus on scaling the platform across PHUs and aligning outputs with standardized public health information systems to support a timely, data-driven response.