Intended for healthcare professionals

Rapid response to:

Feature Hidden Data

Putting GlaxoSmithKline to the test over paroxetine

BMJ 2013; 347 doi: (Published 12 November 2013) Cite this as: BMJ 2013;347:f6754

Rapid Response:

Dear Editor,

Over 70% of paroxetine studies registered at (completion date before January 1st, 2011) remain unpublished. This problem is not unique to paroxetine because other blockbuster antidepressants (e.g., fluoxetine and sertraline) have not published over half of their studies.(1) More unpublished paroxetine studies can be identified analyzing data of other registers (e.g., Eudra CT, ISRCTN, Chi-CTR, among others) downloaded from WHO ICTRP.(2)

Concerns expressed by other authors, regarding the efficacy and safety of antidepressant drugs in humans, are legitimate and well-founded.(3,4) The evidence in peer-review medical journals supporting the efficacy and safety of antidepressants in humans have been cherry-picked.


1. Ramirez, Jorge H (2014): Database of human studies evaluating antidepressant drugs. figshare.

2. Ramirez, Jorge H (2014): Paroxetine: analysis of unpublished human studies (observational studies and clinical trials). figshare.

3. Website of CEP: the Council for Evidence-based Psychiatry. URL: (accessed: December 11, 2014).

4. Ioannidis JPA. Ranking antidepressants. Lancet 2009;373:1759–60; author reply 1761–2. doi:10.1016/S0140-6736(09)60974-0

Competing interests: I endorse the principles of AllTrials and The BMJ Open Data Campaign.

11 December 2014
Jorge Hernán Ramirez
Professor of Pharmacology
Universidad del Valle
Cl 4B No. 36-00