Datum održavanja: 25. listopada 2021. u 14:00 sati u učionici O-357 uz online prijenos preko sustava Jitsi
Predavač: Andrzej Jarynowski, Instytut Badan Interdyscyplinarnych, Poljska
Naziv predavanja: Supporting understanding COVID-19 vaccines safety profiles using Social Media Listening

Abstract:

Aim: I performed exploratory study of how Social Media Listening in Slavic speaking countries using simple keywords or more sophisticated deep learning methods can be useful in pharmacosurveillance of Adverse Events (AE) of COVID-19 vaccines.

Method: Findings from two studies will be presented:

• Mild Adverse Events of Sputnik V Vaccine in Russia: Social Media Content Analysis of Telegram via Deep Learning (link to study)
• Listening Twitter about Adverse Events of the Comirnaty COVID-19 Vaccine during first weeks of immunization in Poland (link to study)

Conclusions: Infodemiology can be useful in early warning pharmacosurveillance of some AE during massive drug delivery as immunization even for single keywords search (Polish case). Adverse event extraction from embedded posts with known individuals' variables (age, gender) with application of BERT technique can give probably even more knowledge than Vaccine producers provided in their registration documentation (Russian case). Thus, I created a risk calculator for non-severe AE of Sputnik V to help people in understanding reactivity patterns of the vaccine (link).