Blog entry written on: Analysis of reports of unintended pregnancies associated with the combined use of non-enzyme-inducing antibiotics and hormonal contraceptives (bmjebm-2020-111363)
Authors: Jeffrey K Aronson & Robin E Ferner
The British National Formulary (BNF) currently advises women to continue with their usual oral contraception when they start an antibiotic (except for enzyme inducers, e.g. rifampicin).
BNF 60 (September 2010) noted that “Some antibacterials that do not induce liver enzymes (e.g. ampicillin, doxycycline) may reduce the efficacy of combined oral contraceptives …”. But by September 2011 the advice had changed: “Latest recommendations are that no additional contraceptive precautions are required when combined oral contraceptives are used with antibacterials that do not induce liver enzymes”. This change was based on a 2011 report from the Faculty of Sexual and Reproductive Healthcare, following a 2009 WHO report. Those reports relied on studies that appeared to show no interaction, although they couldn’t always rule one out. They ignored the weaknesses of the evidence, the fact that no-one had studied unbound hormone concentrations, and the likelihood that not all women were susceptible. If 10% of women were susceptible, with an annual contraceptive failure rate of 2%, you’d need to study 200 women to have a 95% chance of detecting an interaction. The largest randomized studies had recruited 24 women.
If you were a woman of child-bearing age, keen, for whatever reason, not to become pregnant, would you want to risk pregnancy from taking an antibiotic? Even if the risk was small, why take the chance? Why not take other precautions for a week or two?
The BNF adopted the WHO’s advice, even though that advice could never be better than the evidence it was based on. How could we explore the interaction? Perhaps the UK’s Yellow Card system, through which healthcare professionals and patients can report suspected adverse drug reactions, would be informative. But did the reports include “unintended pregnancies”? We looked. They do.
First, we assumed that there is no interaction. Then we tested this assumption. We:
- counted the reported numbers of unintended pregnancies in women taking any of nine different antibiotics;
- compared them with a list of nine medicines not suspected to interact, picked at random from a hospital formulary;
- compared them with 12 enzyme-inducing medicines that everyone agrees do interact;
- chose three control outcomes (headache, cardiac arrhythmias, and congenital anomalies), and a possible confounder, diarrhoea, to compare with the primary outcome of unintended pregnancies.
The results were striking. Enzyme inducers were associated with 13 (that’s thirteen) times more reports of unintended pregnancies than the control medicines and seven times more reports of congenital anomalies. These are both known associations, confirming the usefulness of the reports. The antibiotics were associated with 7 (that’s seven) times more reports of unintended pregnancies than the controls: a clear, highly significant signal. All other results were negative. We don’t know that the reports of unintended pregnancies all came from women taking hormonal contraceptives, but that’s highly likely.
Imagine that you’re a woman of child-bearing age (perhaps you already are), keen, for whatever reason, not to become pregnant just now. What chance would you be prepared to take? Why discard even the smallest possibility that this interaction is real? It doesn’t make sense.
Some are convinced that this interaction does not exist. Some claim that extra instructions may confuse women, causing poor adherence and more contraceptive failures. This patronizes women and deprives them of the choice of what to do. Others suggest giving simple advice.
We’d say to any woman who asked us about antibiotics and the pill: “‘If you are keen not to get pregnant, it is best to avoid sex or to take extra precautions.”
If you’re in doubt, ask your prescriber for advice. And show them this blog.
Jeffrey K Aronson, University of Oxford
Robin E Ferner, University of Birmingham and University College Hospital London
COI: Neither JKA nor REF has a financial conflict of interest; they confess to never having taken an oral contraceptive, nor ever having been pregnant, although both have children.
The views and opinions expressed on this site are solely those of the original authors. They do not necessarily represent the views of the BMJ and should not be used to replace medical advice. All information on this blog is for general information, is not peer-reviewed, requires checking with original sources and should not be used to make any decisions about healthcare. No responsibility for its accuracy and correctness is assumed by us, and we disclaim all liability and responsibility arising from any reliance placed on such commentary or content by any user or visitor to the Website, or by anyone who may be informed of any of its content. Any reliance you place on the material posted on this site is therefore strictly at your own risk.