Scoping review protocol: The resource demands of implementing generative artificial intelligence in multiple choice question development in medical education

Objective: This scoping review aims to describe the extent and nature of evidence relating to the resource demands of implementing generative artificial intelligence in the development of multiple choice questions (MCQs) in medical education. Introduction: Existing studies examining the use of AI for MCQ development have predominantly focused on validity and reliability, often through comparisons with human-authored items. However, the utility and implementation of any novel educational intervention are also informed by its cost-effectiveness. Evidence relating to the resources required for AI-assisted MCQ development is therefore needed to inform decisions about its adoption in medical education. Inclusion criteria: Eligible studies will include medical students, doctors, medical schools, or medical faculties that use generative artificial intelligence to develop MCQs within the context of medical education. Studies reporting monetary and/or non-monetary resources, analyses or evaluations will be included. Methods: The databases PubMed, Scopus (Elsevier), Embase (Ovid), Web of Science (Clarivate), CINAHL (EBSCO), Education Resources Information Centre (ERIC; ProQuest), and PsycInfo (Ovid) will be searched for published studies from 2020-2025. No language restrictions will be applied. Two independent reviewers will screen titles and abstracts, followed by full-text review of eligible studies. Data extraction will be conducted and refined iteratively as required. The evidence will be synthesised and presented descriptively.