Assessment of the Utility of Social Media for Broad-Ranging Statistical Signal Detection in Pharmacovigilance: Results from the WEB-RADR Project

Abstract

Introduction and ObjectiveSocial media has been proposed as a possibly useful data source for pharmacovigilance signal detection. This study primarily aimed to evaluate the performance of established statistical signal detection algorithms in Twitter/Facebook for a broad range of drugs and adverse events.MethodsPerformance was assessed using a reference set by Harpaz et al., consisting of 62 US Food and Drug Administration labelling changes, and an internal WEB-RADR reference set consisting of 200 validated safety signals. In total, 75 drugs were studied. Twitter/Facebook posts were retrieved for the period March 2012 to March 2015, and drugs/events were extracted from the posts. We retrieved 4.3 million and 2.0 million posts for the WEB-RADR and Harpaz drugs, respectively. Individual case reports were extracted from VigiBase for the same period. Disproportionality algorithms based on the Information Component or the Proportional Reporting Ratio and crude post/report counting were applied in Twitter/Facebook and VigiBase. Receiver operating characteristic curves were generated, and the relative timing of alerting was analysed.ResultsAcross all algorithms, the area under the receiver operating characteristic curve for Twitter/Facebook varied between 0.47 and 0.53 for the WEB-RADR reference set and between 0.48 and 0.53 for the Harpaz reference set. For VigiBase, the ranges were 0.64–0.69 and 0.55–0.67, respectively. In Twitter/Facebook, at best, 31 (16%) and four (6%) positive controls were detected prior to their index dates in the WEB-RADR and Harpaz references, respectively. In VigiBase, the corresponding numbers were 66 (33%) and 17 (27%).ConclusionsOur results clearly suggest that broad-ranging statistical signal detection in Twitter and Facebook, using currently available methods for adverse event recognition, performs poorly and cannot be recommended at the expense of other pharmacovigilance activities.Electronic supplementary materialThe online version of this article (10.1007/s40264-018-0699-2) contains supplementary material, which is available to authorized users.

Establishing a Framework for the Use of Social Media in Pharmacovigilance in Europe

“…Although future improvements to ADR recognition in social media posts in terms of performance and coverage may revise this recommendation, social media is not expected to become a first-line signal detection tool. It may, however, serve as a useful complement in specific niche areas [ 32 ]. Where an organisation decides on the use of social media data for signal detection and/or signal validation, the reporting of ADRs in the form of ICSRs should not be required as set out in GVP Module VI [ 7 ].…”

Comparison of text processing methods in social media–based signal detection

“…The examination of SM utility in identifying AEs and SDRs has been previously performed by multiple groups, using various approaches. Some authors have concluded that SM should be further considered as a potential SD tool, while others concluded that SM shows a poor performance in AE and SDR identification . Although we count this work among the former group, we do not disagree with the core conclusions of the latter.…”

“…Some authors have concluded that SM should be further considered as a potential SD tool, 17,18 while others concluded that SM shows a poor performance in AE and SDR identification. 19,20 Although we count this work among the former group, we do not disagree with the core conclusions of the latter. Namely, we agree that SM cannot be used as a stand-alone data source and should not replace other pharmacovigilance activities.…”

Health‐related Google searches performed by parents of pediatric oncology patients

“…12 Other potential applications for symptom data include adverse event monitoring, which has been shown to be feasible in social media data including Twitter. 13 However, caution must be used when considering Google data for this purpose as substantial scientific, 14…”