The quality assessment of the Japanese Adverse Drug Event Report database using vigiGrade

Background The quality of adverse drug reaction reports can have a strong influence on the causality assessment or the detection of safety signals. Various methods have been used to evaluate the quality of adverse drug reaction reports. Objective We used the vigiGrade completeness score to evaluate the quality of the Japanese adverse drug event report database (JADER). Methods Among all reports in the JADER (cases received by regulatory authorities from April 2004 to November 2018), 427,797 cases that did not meet the exclusion criteria were analyzed using vigiGrade. An report was considered a 'well-documented' report if the completeness score > 0.8. For each dimension, the proportion of incomplete information was calculated. Main outcome measure Proportion of 'well-documented' reports. Results A total of 211,774 (49.5%) cases were classified as 'well-documented' reports. When classified by sender, 6240 out of 8257 cases (75.6%) from medical institutions were 'well-documented', while 205,534 out of 419,540 cases (49.0%) from pharmaceutical companies were categorized as such. 'Time-to-onset' could not be calculated in 173,027 (41.2%) cases from the pharmaceutical companies and in 1228 (14.9%) cases from medical institutions. Conclusion In the JADER, the proportion of 'well-documented' adverse drug reaction reports from medical institutions was higher than that from pharmaceutical companies, and reports from medical institutions could help to clarify generated safety signals. We also found that more than 40% of the reports from pharmaceutical companies did not include information that is considered essential for the evaluation of any causal relationship between suspected drugs and adverse drug reactions, which could pose challenges to the detection of safety signals.