Analysis · Health
What 126,815 Drug Safety Reports Can—and Cannot—Reveal
By AWEI · AI-compiled · Published · Analysis prepared · 1 source · doi.org
A blood clot appeared in 1,803 safety reports involving three JAK inhibitors. That was 1.4% of the reports, but it does not mean 1.4% of people taking these medicines developed clots. This distinction is central to interpreting a 2022 study by Léa Hoisnard and colleagues in Scientific Reports.
The researchers examined 126,815 reports naming ruxolitinib, tofacitinib or baricitinib as suspected drugs in the World Health Organization’s VigiBase database, through February 28, 2021. This was a retrospective observational analysis of submitted reports, rather than an experiment assigning treatments or a study following everyone who received them.
Their method asked whether particular events occupied a larger share of reports involving these medicines than expected from reporting across the database. A statistical threshold incorporating uncertainty identified potential safety signals. It measured disproportionate reporting, not the probability that a person would experience an event.
Signals included infections, especially herpes infections, as well as blood clots and certain tumor categories. Gastrointestinal perforation showed a significant reporting signal for tofacitinib. Cases also appeared with the other two medicines, although their signals did not meet the threshold. Absence of a signal therefore did not mean absence of reported events.
The limits go beyond missing reports. Diagnoses had limited verification, and information on timing and alternative causes was incomplete. The drugs also served different patient populations: ruxolitinib reports commonly involved blood disorders, while the other two commonly involved rheumatoid arthritis. Comparisons did not adequately account for patient characteristics or treatment duration, so they cannot establish which drug caused more harm.
The practical value is identifying questions for closer investigation. These reporting associations can guide studies with defined patient populations and longer follow-up. They cannot establish causation, calculate individual risk or reliably rank the medicines by safety.
Sources used for this article (1)
Direct links to the publisher reports used to prepare this article.
- Source 1
- Adverse events associated with JAK inhibitors in 126,815 reports from the WHO pharmacovigilance database ['Léa Hoisnard', 'Bénédicte Lebrun‐Vignes', 'Sébastien Maury', 'Matthieu Mahévas', 'Khalil El Karoui', 'Lydia Roy', 'Anissa Zarour', 'Marc Michel', 'José L. Cohen', 'Aurélien Amiot', 'Pascal Claudepie
