As AI-assisted research tools continue rapidly expanding, public health researchers are increasingly faced with questions surrounding which tools are useful, how they can be integrated into evidence synthesis workflows, and what methodological and ethical considerations should guide their use. These challenges are especially relevant in interdisciplinary public health research, where fragmented terminology, inconsistent indexing, and evidence dispersed across multiple disciplines can make comprehensive literature searching particularly difficult.
This session presents a practical hybrid evidence synthesis workflow that combines traditional review methods with selectively integrated AI-assisted tools to strengthen search development, terminology refinement, citation mapping, and supplemental evidence discovery while maintaining transparency, reproducibility, and methodological rigor.
Using a scoping review on tobacco cessation interventions as a case study, the session will walk participants through each stage of the workflow. The presentation will include:
* How traditional database search strategies are initially developed and refined. * How AI-assisted tools such as Elicit and Research Rabbit can be used to identify related terminology, map citation networks, and uncover supplemental literature that may have been missed through traditional keyword searching alone. * Examples of how search terms and concepts may evolve throughout the review process in response to interdisciplinary terminology differences. * Demonstrations of practical workflow integration across stages including evidence discovery, screening support, citation organization, and documentation practices. * Discussion surrounding methodological tradeoffs, reproducibility concerns, ethical considerations, and limitations associated with integrating AI-assisted tools into evidence synthesis.
Participants will engage with example workflows and practical demonstrations and apply them to their own research question to demonstrate how hybrid approaches can support evidence synthesis across a variety of public health topic areas. The session will emphasize that AI-assisted tools are intended to complement, not replace, researchers' expertise, predefined inclusion criteria, manual screening processes, and transparent reporting practices.
Attendees will also discuss practical considerations for selecting appropriate AI-assisted tools depending on project goals, research questions, and stages of evidence synthesis. By focusing on adaptable workflows and critical evaluation rather than automation alone, the session aims to provide participants with transferable skills they can apply to their own interdisciplinary public health research environments.
This session is particularly relevant for public health researchers, graduate students, practitioners, and interdisciplinary teams conducting evidence synthesis in emerging or under-indexed topic areas where traditional search strategies alone may not fully capture the available literature.
