Withdrawal of medicines that cause deaths: takes longer than you think

A research article published in BMC Medicine investigates the length of time taken to withdraw dangerous drugs from the market. In this guest blog, authors Igho Onakpoya, Carl Heneghan and Jeffrey Aronson from University of Oxford discuss delays in withdrawing drugs suspected of causing deaths and how these delays could be reduced.

L-R: Jeffrey Aronson, Igho Onakpoya, Carl Heneghan
L-R: Jeffrey Aronson, Igho Onakpoya, Carl Heneghan

If a drug was suspected of causing deaths, how long would you think it should take before a drug company, or the regulator, withdraws it from the market?

To answer this question we searched scientific databases and other sources of information about adverse drug reactions to identify drugs that were withdrawn from the market in the last 60 years because they were suspected of having caused deaths. For each withdrawn drug we documented the year in which it was first introduced, the year a death attributed to its use was first reported in the scientific literature, the year it was first withdrawn, and the number of countries in which it was withdrawn at any time thereafter.

In total we identified 95 drugs withdrawn between 1957 and 2011 (82 had caused deaths at recommended doses; 13 had caused deaths after overdose). The evidence used for making the withdrawal decisions, in most cases, consisted of anecdotal reports.

The interval between the year of first use and the time to first withdrawal has shortened over time. However, the interval between the first reported death and the subsequent first withdrawal has not shortened. Obviously, there are problems with assigning causality, suggesting that there are no standardized methods for deciding whether a drug that has been suspected of causing deaths should be withdrawn from the market.

When serious adverse events occur, evaluation should be speedier than at present. In almost half of the cases, drugs that caused deaths were only withdrawn after two years had passed.

Of serious concern is that there were also discrepancies in the pattern of drug withdrawals across regulatory authorities in different geographical regions. Notably, withdrawal of medicines was much more common in Europe than in Africa.

Better international co-ordination among regulatory authorities is needed and should lead to speedier and more uniform decision-making processes when drugs are suspected of causing deaths. What we need is early formal studies when deaths are suspected to have been due to a drug. Increased transparency in the reporting of adverse events in clinical trials would also help with quicker identification of potentially dangerous drugs, and greater efforts should be made to strengthen drug monitoring systems in low- to middle-income economies.

So how long did you think it would take for a drug to be withdrawn? Well it depends on where you live and it often takes longer than you think.

View the latest posts on the On Medicine homepage