Natural Language Processing for Pharmacovigilance: Detecting Novel Sildenafil Adverse Events in Social Media
InfoScience Trends · 2025
DOI: 10.61882/ist.202502.07.03 ↗Kurz gefasst
Fachartikel in der Zeitschrift InfoScience Trends, 2025. Verfasst von Sobhan Maghoul und 5 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Themen: Patientensicherheit
Abstract (Originalsprache)
This study leverages natural language processing (NLP) and social media data to identify novel adverse effects of sildenafil, complementing traditional pharmacovigilance systems. We analyzed 2.92 million English-language posts from Twitter, Reddit, and health forums (2020–2025) using a multi-stage NLP pipeline. First, BERTweet classified sildenafil-related content (F1=0.925), followed by BioClinical-BERT with CRF for named-entity recognition of adverse events (strict F1=0.866). A DeBERTa-v3 model then extracted drug–adverse event relationships (F1=0.812). Statistical signal detection using proportional reporting ratios (PRR) identified 42 significant adverse event signals, including eight potentially novel effects (e.g., tinnitus [PRR=2.87] and night sweats [PRR=2.46]) and six underreported effects (e.g., nasal congestion). Posts mentioning off-label sildenafil use (11.4% of corpus) showed 1.8-fold higher adverse event reporting rates (χ²=182.4, p
Zitieren
APA
Maghoul, S., Iranshahi, S., Divkolaei, S. M. H., Naderinia, K., Mahjoub, M., & Olama, E. (2025). Natural Language Processing for Pharmacovigilance: Detecting Novel Sildenafil Adverse Events in Social Media. InfoScience Trends. https://doi.org/10.61882/ist.202502.07.03
Vancouver
Maghoul S, Iranshahi S, Divkolaei SMH, Naderinia K, Mahjoub M, Olama E. Natural Language Processing for Pharmacovigilance: Detecting Novel Sildenafil Adverse Events in Social Media. InfoScience Trends. 2025. doi:10.61882/ist.202502.07.03
BibTeX
@article{Maghoul2025,
author = {Maghoul, Sobhan and Iranshahi, Setare and Divkolaei, Seyed Majid Heydari and Naderinia, Kiana and Mahjoub, Mahdi and Olama, Ensiyeh},
title = {Natural Language Processing for Pharmacovigilance: Detecting Novel Sildenafil Adverse Events in Social Media},
journal = {InfoScience Trends},
year = {2025},
doi = {10.61882/ist.202502.07.03}
}
EUVIMED