Lessons learned from annotation of VAERS reports on adverse events following influenza vaccination and related to Guillain-Barré syndrome

Madhuri Sankaranarayanapillai,Su Wang,Hangyu Ji,Hsing-Yi Song,Cui Tao
DOI: https://doi.org/10.1186/s12911-023-02374-2
IF: 3.298
2024-01-07
BMC Medical Informatics and Decision Making
Abstract:Vaccine Adverse Events ReportingSystem (VAERS) is a promising resource of tracking adverse events following immunization. Medical Dictionary for Regulatory Activities (MedDRA) terminology used for coding adverse events in VAERS reports has several limitations. We focus on developing an automated system for semantic extraction of adverse events following vaccination and their temporal relationships for a better understanding of VAERS data and its integration into other applications. The aim of the present studyis to summarize the lessons learned during the initial phase of this project in annotating adverse events following influenza vaccination and related to Guillain-Barré syndrome (GBS). We emphasize on identifying the limitations of VAERS and MedDRA.
medical informatics
What problem does this paper attempt to address?
The paper aims to address the issue of semantic extraction and temporal relationships of adverse events (particularly those related to Guillain-Barré Syndrome) following influenza vaccination in the Vaccine Adverse Event Reporting System (VAERS). The authors' goal is to develop an automated system to better understand and integrate VAERS data, and to summarize the challenges encountered and lessons learned in the process. Specifically, the research team collected 282 VAERS reports recorded between 1990 and 2016, and selected reports containing at least 1100 characters for preliminary study. They annotated all adverse events following influenza vaccination using MedDRA terminology and, where possible, annotated the relevant temporal expressions. Additionally, the study emphasized the importance of recognizing the limitations of VAERS and MedDRA. Through this work, the authors hope to improve the structured interface of the VAERS system, provide more detailed temporal information, and develop systems for the automatic identification of specific vaccine-related adverse events, thereby enhancing the value of VAERS reports in research and the capability for automated meaningful information retrieval.