Utilizing AI and Social Media Analytics to Discover Adverse Side Effects of GLP-1 Receptor Agonists
This work addresses patient safety by enabling rapid discovery of hidden adverse side effects for regulators and manufacturers, though it is incremental as it applies existing AI methods to a new domain.
The researchers tackled the problem of detecting overlooked adverse side effects (ASEs) of drugs after FDA approval by developing a digital health methodology that analyzes social media, clinical research, reports, and ChatGPT data. They successfully identified 21 potential ASEs for GLP-1 receptor agonists, including irritability and numbness.
Adverse side effects (ASEs) of drugs, revealed after FDA approval, pose a threat to patient safety. To promptly detect overlooked ASEs, we developed a digital health methodology capable of analyzing massive public data from social media, published clinical research, manufacturers' reports, and ChatGPT. We uncovered ASEs associated with the glucagon-like peptide 1 receptor agonists (GLP-1 RA), a market expected to grow exponentially to $133.5 billion USD by 2030. Using a Named Entity Recognition (NER) model, our method successfully detected 21 potential ASEs overlooked upon FDA approval, including irritability and numbness. Our data-analytic approach revolutionizes the detection of unreported ASEs associated with newly deployed drugs, leveraging cutting-edge AI-driven social media analytics. It can increase the safety of new drugs in the marketplace by unlocking the power of social media to support regulators and manufacturers in the rapid discovery of hidden ASE risks.