Crowdsourcing Adverse Events Associated With Monoclonal Antibodies Targeting Calcitonin Gene-Related Peptide Signaling for Migraine Prevention: Natural Language Processing Analysis of Social Media.
Journal:
JMIR formative research
PMID:
39515814
Abstract
BACKGROUND: Clinical trials demonstrate the efficacy and tolerability of medications targeting calcitonin gene-related peptide (CGRP) signaling for migraine prevention. However, these trials may not accurately reflect the real-world experiences of more diverse and heterogeneous patient populations, who often have higher disease burden and more comorbidities. Therefore, postmarketing safety surveillance is warranted. Regulatory organizations encourage marketing authorization holders to screen digital media for suspected adverse reactions, applying the same requirements as for spontaneous reports. Real-world data from social media platforms constitute a potential venue to capture diverse patient experiences and help detect treatment-related adverse events. However, while social media holds promise for this purpose, its use in pharmacovigilance is still in its early stages. Computational linguistics, which involves the automatic manipulation and quantitative analysis of oral or written language, offers a potential method for exploring this content.