Social Listening in Pharmacovigilance: Uncovering Unreported Adverse Drug Reactions With NLP

Abstract

Adverse Drug Reaction (ADR) detection and study are important for pharmacovigilance, which is the safety of medicines. Underreporting can get in the way of traditional ADR ratings. Natural Language Processing (NLP) is used in this study to look for ADRs that haven't been mentioned on social media and online health forums. The goal is to see if social listening can be used in addition to pharmacovigilance. Many types of NLP techniques, such as sentiment analysis, topic modeling, and named entity identification, were used to gather a big set of user-generated content from different websites. Names of drugs and bad reactions were looked for. Our research shows that traditional pharmacovigilance databases missed a lot of ADRs. The ADRs found through social listening were compared to those found in medical books and databases. This shows that NLP can fill in the gaps in current reporting systems. It also looks at how reliable social media data is, how hard it is to make filtering algorithms, and how users should protect their privacy and ethically use data.