The 2nd Global Evidence Summit (GES) was held in Prague, Czech Republic, from 10 to 13 September 2024. The conference focused on “Using evidence. Improving lives.”. This year, GES brought together 1824 participants from 152 countries. There are 6 plenaries, 25 special sessions, and 527 oral presentations included in the conference among 6 main domains.
One of the main themes this year is discussing the impact and application of Artificial Intelligence (AI) in evidence synthesis, including applying AI tools for producing systematic reviews, responsible use of AI, the role of AI in Evidence Synthesis, etc. The conference provides a good opportunity for researchers to understand the big picture of evidence synthesis as well as the progress and challenges in applying AI tools in the systematic review.
We presented our research, “Automating Screening of Studies for Systematic Reviews Using a Large Language Model,” at one of the “Impact of AI” oral presentation sessions, a project collaborating between the University of Bristol, Imperial College London, and the World Cancer Research Fund. We demonstrated the latest results of our study screening model on the CUP cancer inclusion/exclusion dataset.
At the conference, we noticed that using AI, especially the Large Language Model, has become a new trend in evidence synthesis and has a lot of potential. Some researchers presented their preliminary results on applying the Large Language Model in study screening and risk of bias analysis. However, the results lack consistency, and the performance can be further improved before being applied in daily evidence synthesis research. This is also what we will work on to improve in the future.