Article Text

Download PDFPDF
PS-097 A comparative study between investigator safety data and a pharmacovigilance database: a key role of the pharmacovigilant in the relevancy of clinical trial publications
  1. V Pitance,
  2. E Blanc,
  3. A El Hachemi Dumas,
  4. D Bertram
  1. Direction de la Recherche Clinique et de l’Innovation- Hospices Civils de Lyon- 69229 Lyon Cedex 02, Rhônes-Alpes, Lyon, France


Background Safety assessment of an investigational medicinal product is an essential requirement of a clinical trial, and the Consolidated Standards of Reporting Trials group has produced recommendations to improve the communication of pharmacovigilance (PV) data in scientific publications. However, little data have been published on the quality of reporting.

Purpose To compare safety data of a clinical study reported in a published article and in the sponsor’s PV database.

Material and methods The article was randomly selected among studies in a PV database. This randomised study compared two drugs in terms of efficacy; safety was not the primary outcome. Cases of serious adverse events (SAEs) were extracted from the sponsor’s PV database. These SAEs were coded using the MedDRA dictionary according to an international terminology. SAEs presented in the article and those reported in the PV database were compared with regards to frequency, system organ class (SOC), severity, causality with treatment, expectedness and the outcome of each SAE.

Results The article reported that a total of 60 patients experienced at least 1 SAE whereas 78 patients were found in the PV database. 2 deaths were reported in the article, whereas 3 were found in the PV database. 4 SAEs in the infectious SOC failed to be reported in the article. 15 SAEs in the injury SOC were not reported in the article: 14 were a drug administration error. An additional SAE in cancer SOC was present in the PV database. 1 SAE regarding eye disorders SOC was not reported in the article. Severity, causality, expectedness and outcome were never presented in the article; data regarding all of these aspects were reported for all SAE in the PV database.

Conclusion A number of SAEs and all characterising data (ie, severity, causality, etc) were omitted in the restitution of the safety data. With the same data, the investigator in charge of writing the article did not lay out the safety elements according to the same methodology and skills as the pharmacovigilant.

No conflict of interest

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.