AI in Pragmatic Literature Reviews: A Scoping Review of Current Guidance and Development of a Framework for Integration
Author(s)
Grace E. Fox, PhD1, Nita Santpurkar, MSc2, Juliette Torres Ames, MSc3, Emanuele Arca`, MSc4.
1OPEN Health HEOR & Market Access, New York, NY, USA, 2OPEN Health HEOR & Market Access, Thane, India, 3OPEN Health HEOR & Market Access, London, United Kingdom, 4OPEN Health HEOR & Market Access, Rotterdam, Netherlands.
1OPEN Health HEOR & Market Access, New York, NY, USA, 2OPEN Health HEOR & Market Access, Thane, India, 3OPEN Health HEOR & Market Access, London, United Kingdom, 4OPEN Health HEOR & Market Access, Rotterdam, Netherlands.
OBJECTIVES: Pragmatic literature reviews, such as rapid and targeted reviews, offer flexibility to support evidence needs at early stages of decision-making for market access. Artificial intelligence (AI) has the potential to make these reviews more efficient while maintaining quality. This study aimed to explore current guidance for AI use in pragmatic literature reviews and propose a framework for integration.
METHODS: A scoping review was conducted in January 2025, updated in April 2025, to identify guidelines for using AI in pragmatic literature reviews. Sources included organizations and societies, such as Cochrane and the Centre for Reviews and Dissemination, and working groups, such as Responsible AI in Evidence Synthesis. Factors influencing the integration of AI into pragmatic literature review workflows were evaluated, and strategies were developed to support effective implementation.
RESULTS: Guidance on the use of AI in pragmatic literature reviews was limited, reflecting the longstanding gap in methodological guidance for such reviews. Existing recommendations were primarily developed for systematic literature reviews and do not adequately address pragmatic literature reviews involving less structured datasets and more flexible methodologies. In the absence of formal guidelines, a framework was designed for implementing AI in pragmatic literature reviews, one that mandates human oversight, aligns with review design (e.g., rapid, targeted), ensures adaptability to diverse data sources (e.g., electronic databases, gray literature), maintains transparency, enables ongoing validation, and encourages stakeholder engagement.
CONCLUSIONS: This study offers an overview of the nascent role of AI in pragmatic literature reviews. Limitations were the scarcity of tailored guidance and empirical evidence specific to pragmatic literature reviews. In light of this, we proposed a framework for integrating AI in pragmatic literature reviews, offering a structured yet adaptable approach that ensures human oversight, transparency, and stakeholder collaboration.
METHODS: A scoping review was conducted in January 2025, updated in April 2025, to identify guidelines for using AI in pragmatic literature reviews. Sources included organizations and societies, such as Cochrane and the Centre for Reviews and Dissemination, and working groups, such as Responsible AI in Evidence Synthesis. Factors influencing the integration of AI into pragmatic literature review workflows were evaluated, and strategies were developed to support effective implementation.
RESULTS: Guidance on the use of AI in pragmatic literature reviews was limited, reflecting the longstanding gap in methodological guidance for such reviews. Existing recommendations were primarily developed for systematic literature reviews and do not adequately address pragmatic literature reviews involving less structured datasets and more flexible methodologies. In the absence of formal guidelines, a framework was designed for implementing AI in pragmatic literature reviews, one that mandates human oversight, aligns with review design (e.g., rapid, targeted), ensures adaptability to diverse data sources (e.g., electronic databases, gray literature), maintains transparency, enables ongoing validation, and encourages stakeholder engagement.
CONCLUSIONS: This study offers an overview of the nascent role of AI in pragmatic literature reviews. Limitations were the scarcity of tailored guidance and empirical evidence specific to pragmatic literature reviews. In light of this, we proposed a framework for integrating AI in pragmatic literature reviews, offering a structured yet adaptable approach that ensures human oversight, transparency, and stakeholder collaboration.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
SA9
Topic
Methodological & Statistical Research, Study Approaches
Topic Subcategory
Literature Review & Synthesis
Disease
No Additional Disease & Conditions/Specialized Treatment Areas