Reporting adverse events in published and unpublished studies

Around the world / 30 January 2017

Photo: Adhy Savala, Unsplash

A recent article aims to quantify the under-reporting of adverse events (AEs) in published medical literature, compared with unpublished sources.


The article, “Reporting of Adverse Events in Published and Unpublished Studies of Health Care Interventions: A Systematic Review”, published in PLOS Medicine and written by Su Golder, Yoon K. Loke, Kath Wright, and Gill Norman, reviewed 28 studies that provided a comparison of information on AEs according to whether the study was published or unpublished.

The researchers found that fewer published studies reported information on AEs compared with unpublished studies; that both a lower number of AEs and a narrower range of named AEs are reported in published studies compared to unpublished studies; and that the inclusion of unpublished data in meta-analyses leads to more precise conclusions on harms.

The authors’ assessment was that there is a substantial amount of unpublished data on AEs which may be “hidden” from health- care providers and patients in the form of unpublished data. They suggest that it is not possible to have a complete understan- ding of the safety profile of an intervention at the time therapeutic decisions are made by physicians and patients.

An important limitation of the signal detection work that we do at UMC is the lack of “denominator data” when identifying potential safety concerns from spontaneous data. In other words, we are unable to estimate how frequently a specific adverse event occurs when it arises in the post-marketing period as we do not know how many people have taken the drug in question. It would be very interesting to investigate if those signals identified in early post-marketing use have been previously observed in clinical trials but not reported.

There is a growing movement calling for full disclosure and unrestricted access to clinical trial data, as exemplified in campaigns such as AllTrials. Given the potential for unpublished data to be an important source of information on harm from medicines, UMC fully supports the authors’ call for a new policy and action to make adverse event data readily accessible to the public.

Read more:

You may also like


Egypt uses active community engagement to raise awareness of herbals in healthcare

To raise awareness of phytovigilance in healthcare, ISoP Egypt created an outreach proposal centred on sharing information, experiences, and anecdotes of the safety of herbals.

Around the world / 26 June 2024

Celebrating 40 years of making medicines safer in Spain

This year, the Spanish pharmacovigilance system turns 40. This time has seen much success, with plans on sustaining that success in the coming four decades.

Around the world / 21 May 2024

ISoP Africa gets a fresh start with its first chapter meeting in Uganda

The meeting brought together nearly 300 experts to discuss advancing pharmacovigilance practice and local data-driven decision-making on the continent.

Around the world / 16 September 2024