Summary

Peer-reviewed publications are a foundational element in the communication of scientific evidence about biomedical research conducted by and for pharmaceutical and biotechnology companies. Current quantitative methods for measuring and analysing scholarly literature are limited in evaluating the impact of biomedical publications on driving research and informing healthcare decision-making.

Adding artificial intelligence (AI) to these methods can instigate a move from statistical evaluation to textual evaluation, at a scale that can significantly enhance the quality and understanding of the true impact that published scientific research has on scientific debate and, ultimately, its impact on the role of medicine in improving patient outcomes. In this article, we suggest using AI-powered analytics to develop new and comprehensive ways to measure the impact of a publication, not only on research but on clinical practice. We highlight approaches to measuring when scientific research actually informs changes in clinical practice.