T 05: Compare the Quality of Case Reports Originating from Social Media with Spontaneous Case Reports by Evaluating Case Attributes

Poster Presenter

Samarth Parikh

Signal Detection, Associate Director
Janssen Pharmaceutical Companies of Johnson & Johnson
United States

Objectives

The objective of this study is to compare the quality and information content of adverse event (AE) cases from a social media source (i.e., non-Company sponsored websites) with cases from spontaneous AE reporting.

Method

Case quality will be assessed using vigiGrade algorithm (Bergvall et al., 2013); a score of >0.8 represents a well-documented case. For individual case attributes, reporting fractions will be calculated. The statistical software, SAS® JMP®, will be used to prepare the statistical outputs.

Results

In this study, four Company products with social listening programs on non-Company sponsored websites were selected. The Company’s Global Safety Database was searched for cases from those four social media programs and cases from all spontaneous sources; both valid and invalid cases, and only the initial version of cases were included. Cases were restricted to those received in the United States between January 1, 2011 and July 30, 2015. The primary endpoint is to compare the case quality between social media and spontaneous cases by evaluating the absence or presence of case attributes. The case attributes evaluated in this study include time-to-onset, indication, outcome, sex, age, dose, country, primary reporter, report type, and comments. Secondary endpoints are comparison of type of AEs (e.g., serious and non-serious, labelled and non-labelled) reported between social media and spontaneous cases. The preliminary analysis from the search results revealed differences in completeness of case attributes between social media and spontaneous cases. Many of the study attributes were less frequently reported in social media cases in comparison with spontaneous cases.

Conclusion

The preliminary results of this study revealed that many of the case attributes were less frequently reported in social media cases in comparison with spontaneous cases for the four Company products. The complete analysis will identify if there are differences in the quality of cases as per vigiGrade scoring between social media and spontaneous sources for these four selected products. Additional authors: Geoffrey Gipson, Ph.D.; Robert Kwarta Jr., Ph.D.; Natalie Gearhart, PharmD; Julia Bui, PharmD; Thang Trieu, PharmD