• Bernardo Sousa-Pinto, Manuel Marques-Cruz, Ignacio Neumann, Yuan Chi, Artur J Nowak, Marge Reinap, Mariette Awad, Monika Nothacker, Milana Trucl, Jan Brozek, Pablo Alonso-Coello, Wojtek Wiercioch, Amir Qaseem, Elie A Akl, Holger J Schünemann, and 2024 Board of Trustees of the Guidelines International Network.
• MEDCIDS - Department of Community Medicine, Information and Health Decision Sciences, and CINTESIS@RISE - Health Research Network, MEDCIDS, Faculty of Medicine, University of Porto, Porto, Portugal (B.S.-P., M.M.-C.).
• Ann. Intern. Med. 2025 Jan 28.

DescriptionArtificial intelligence (AI) has been defined by the High-Level Expert Group on AI of the European Commission as "systems that display intelligent behaviour by analysing their environment and taking actions-with some degree of autonomy-to achieve specific goals." Artificial intelligence has the potential to support guideline planning, development and adaptation, reporting, implementation, impact evaluation, certification, and appraisal of recommendations, which we will refer to as "guideline enterprise." Considering this potential, as well as the lack of guidance for the use of AI in guidelines, the Guidelines International Network (GIN) proposes a set of principles for the development and use of AI tools or processes to support the health guideline enterprise.MethodsA GIN working group on AI developed these principles, informed by the results of a scoping review and practical examples, through iterative discussion.RecommendationsEight principles were identified to adhere to when using AI in the guideline context: transparency, preplanning, additionality, credibility, ethics, accountability, compliance, and evaluation. These complementary principles are described in a comprehensive way, but they do not provide detailed instructions on how to use specific AI tools. Although these principles are expected to apply across different contexts and stages of the guideline enterprise, details on their implementation have some degree of flexibility. Guideline development groups choosing to use AI will be able to adequately implement the principles if they ensure aspects such as structured reporting on the use of AI tools, involvement of experts in AI, and allocation of funding for the adequate use of AI tools. The GIN principles may support guideline developers in the responsible and transparent use of AI to ensure trustworthy guidelines.

8 keywords...

hide…