Giving aspirin and ibuprofen after myocardial infarctionBMJ 2003; 327 doi: https://doi.org/10.1136/bmj.327.7427.1298 (Published 04 December 2003) Cite this as: BMJ 2003;327:1298
- Stephen E Kimmel, associate professor of medicine and biostatistics and epidemiology (, )
- Brian L Strom ()
Clinical consequences are still unknown
Aspirin is a highly effective antiplatelet agent that is used by millions of people to reduce cardiovascular morbidity and mortality.1 2However, a recent pharmacodynamic study showed that ibuprofen, a non-aspirin, non-steroidal, anti-inflammatory drug, can inhibit the antiplatelet effects of aspirin.3 This effect occurred in people who took daily ibuprofen before taking aspirin or in those taking ibuprofen regularly. Participants who took a single daily ibuprofen after aspirin did not exhibit an inhibitory effect. It is therefore possible that ibuprofen, if taken regularly or before daily aspirin, could reduce or even negate the beneficial effects of aspirin. If this interaction is clinically relevant it could have enormous public health implications because non-aspirin nonsteroidals are among the most commonly used drugs in the world.4
Designing studies to address the clinical relevance of this interaction presents some unique challenges. The ideal study would accurately measure the use of aspirin and ibuprofen, both prescription and non-prescription, and their frequency of use in the period immediately preceding the outcome. Reliance on prescription records or one time assessments of medication use at baseline are likely to be inaccurate for several reasons. Firstly, many patients obtain ibuprofen and aspirin without a prescription.5 Secondly, even among “users” of ibuprofen, many take the drug only for a short time, sporadically, and less than three times a day.6 Thirdly, adherence with prescribed aspirin is suboptimal, and many patients who are prescribed aspirin will not be taking it regularly.7 8 A final caveat is that ibuprofen itself has antiplatelet properties3 and could itself reduce the risk of myocardial infarction.9 Therefore, even if ibuprofen interferes with the antiplatelet effects of aspirin it is not clear that this would result in a net increase in risk of myocardial infarction.
With these methodological challenges in mind, three papers have recently been published that examine the clinical effects of combining ibuprofen and aspirin. Macdonald and Wei, using an electronic prescription database, found that the rates of both all cause mortality and cardiovascular mortality were higher among patients with cardiovascular disease who were prescribed aspirin and ibuprofen versus aspirin alone; diclofenac and other non-steroidals did not show this effect.10 Although the study did not measure the use of non-prescription non-steroidals, this limitation would probably have biased the results towards the null hypothesis and therefore is an unlikely explanation for their findings. By considering patients exposed to non-steroidals only if they had an active prescription for the drug, the authors tried to minimise the misclassification inherent in using a single prescription for a non-steroidal as an indicator of long term use. They did not, however, use a similar approach for aspirin prescriptions. Further acknowledged limitations of the study included the relatively small sample size and inability to adjust for potentially important confounders that are likely to be more common among users of nonsteroidals, such as higher body mass index and lower levels of physical activity.10
Kurth et al examined data from the physicians' health study.11 They found that doctors randomised to aspirin who also reported on an annual survey that they had used non-steroidals (not specifically ibuprofen) for at least five days in a one month period, but not those reporting less frequent use, had an increased risk of subsequent myocardial infarction relative to those randomised to aspirin and not reporting any non-steroidal use. This shows that frequency of non-steroidal use is an important factor. However, only six myocardial infarctions occurred among users of frequent non-steroidals in the aspirin group. Also, because actual use of aspirin was not known some of those with myocardial infarctions had probably stopped using aspirin because of their non-steroidal use.
In this issue of the BMJ Curtis et al take a different approach to studying ibuprofen and aspirin (p 1322).12 Relying on a single prescription for aspirin or a non-steroidal drug at discharge from hospital after a myocardial infarction as a marker of long term use, they found no increase in risk of death among ibuprofen plus aspirin users versus users of aspirin only. Their study has the advantages of large sample size and detailed clinical data for multivariable adjustments. However, they do not measure use of non-prescription aspirin or nonsteroidals or actual use of the prescribed drugs, both of which could mask a harmful effect of ibuprofen. Unlike the other studies they measure only a one time prescription of ibuprofen. Given the likelihood that many patients will not continue to use ibuprofen over the follow up period, the results will be biased further towards the null. The higher prevalence of smoking and dementia in users of aspirin only also shows that adherence in this group may have been worse,8 13 potentially masking harm of ibuprofen. Finally, overall mortality may be an insensitive marker for recurrent myocardial infarction. Thus, we believe that this study does not obviate the potential importance of the previous studies' findings. Nevertheless, a recent study reported as an abstract did not identify an increased risk among those receiving concomitant prescriptions for aspirin and ibuprofen versus aspirin alone.14
Therefore, although the interaction of ibuprofen with aspirin is potentially clinically important, the current level of evidence is not sufficient to make definitive recommendations for or against the use of concomitant ibuprofen for patients requiring prophylactic aspirin. Additional studies are needed that can accurately measure concomitant and regular consumption of ibuprofen and aspirin. Meanwhile, the choice of analgesics and anti-inflammatory agents for patients requiring prophylactic aspirin should be based on maximising efficacy and minimising bleeding complications.
Papers p 1322
Competing interests SEK has received funding for research from companies that manufacture non-steroidal anti-inflammatory drugs and has served as a consultant for Bayer Consumer Care. BLS has received research funding from and been a consultant to most manufacturers of non-steroidal anti-inflammatory drugs.