Painkillers and risk of heart rhythm disorders

This week the British Medical Journal published a report by Schmidt and colleagues from Denmark, North Carolina and Boston, USA on "Non-steroidal anti-inflammatory drug use and risk of atrial fibrillation or flutter ..." This has lead to alarming print media and web headlines on risks of heart disease from these commonly used medicines. How worried should we be? On one level this is a general reminder that powerful medicines may have powerful harmful effects. Indeed previous studies have reported an increased risk of atrial fibrillation from NSAIDs and these medicines and the related COX-2 inhibitors have been implicated in increased risk of other types of heart disease. With that in mind, patients on these medicines who develop palpitations or other features of heart problems, including shortness of breath or ankle swelling, should consult their GP for advice. However there are some interesting aspects of the study showing the importance of looking critically behind headlines at the details of research reports. This includes considering potential sources of bias which may lead to undersestimation or exaggeration of effects; and being clear on the difference between absolute effect on risk (e.g. 17 vs. 10 per 1000) and relative risk (e.g. 70% increase). Here is a link to Donald talking about the study on BBC Radio