Analysis of COVID-19 vaccine-related adverse event reports following COVID-19 vaccine BNT162b2 mRNA-1273 and Ad.26.COV2.S based on the US vaccine adverse event reporting system

Yuchen Wang
DOI: https://doi.org/10.1117/12.2692443
2023-09-07
Abstract:Safety data mostly obtained before vaccines are marketed usually come from clinical trials. There has been adverse drug reaction (ADR) reports about the escalation of vaccination-related side-effects like myocarditis and pericarditis especially in mRNA vaccine and thrombosis in adenoviral vector vaccine. Data mining are implemented to detect and evaluate the ADR signals of three vaccination of SARS-COV-2 vaccines; Pfizer/BioNTech (BNT162b2), Moderna (mRNA-1273), Jansson (Johnson & Johnson). The vaccine data extracted from Vaccine Adverse Event Reporting System (VAERS) database from December 11, 2020 to March 11, 2022. A total of 824,818 adverse cases of all three vaccines were extracted from the database for signal detection. Signal detection method of reporting odds ratio (ROR) and Bayesian confidence propagation neural network (BCPNN) were adopted based on the fourfold table and corresponding formula for data mining. Information component (IC) of COVID-19 vaccine measures the strength of the relation between the interest vaccine and the interest disease. The 95% confidence interval lower limit of expected value IC which is E(IC)-2SD were used to measure the signal strength of ADR in the database where SD is standard deviation of IC. R-Studio and Visual Studio Code are used for data management and statistical analysis. Comprehensive signal detection method was adopted to mine relevant data from VAERS and many new safety signals were detected at last, which lays a foundation for the follow-up safety studies on COVID-19 vaccine. Therefore, great importance should be attached to drug monitoring during clinical use.
Engineering,Medicine,Computer Science
What problem does this paper attempt to address?