Abstract
PURPOSE: We aimed to analyze the surveillance reports of adverse events (AEs) due to different types of pneumococcal vaccines, in addition to detecting and validating signals of pneumococcal vaccines by comparing AEs with labels. MATERIALS AND METHODS: We analyzed the percentages of AEs according to vaccine type [pneumococcal polysaccharide vaccines (PPSVs) and pneumococcal conjugate vaccines (PCVs)] in children and adults using data from the Korea Adverse Event Reporting System (KAERS) database from 2005 to 2016. A signal was defined as an AE that met all three indices of data mining: proportional reporting ratio (PRR), reporting odds ratio (ROR), and information component (IC). We validated the detected signals by calculating sensitivity, specificity, as well as positive and negative predictive values of the signals against label information. RESULTS: Of the 39933 AE reports on vaccination, 5718 (7.0%) were related to pneumococcal vaccine. The most frequent AE after vaccination with PPSV was fever (23.9%) in children and injection-site reaction in adults. The most frequent AE after vaccination with PCV in children was pharyngitis (26.2%). In total, 13 AEs met all three indices for signal detection. Among these, hypotension, apathy, sepsis, and increased serum glutamic oxaloacetic transaminase level were not listed on vaccine labels. In validation analysis, PRR and ROR performed slightly better than IC for adults who were vaccinated with PPSVs. CONCLUSION: Overall, 13 new signals of PPSVs, including four signals not listed on the labels, were detected. Further research based on additional AE reports is required to confirm the validity of these signals for children.