EDITOR—Kalra et al's paper looking at the efficacy of anticoagulation for stroke prevention in atrial fibrillation1 essentially complements Copland et al's work in showing that an elderly population can be given adequate anticoagulation.2

As Kalra et al performed a cohort study with no control population, the study does not allow any estimate of the effectiveness of the treatment as is suggested in the paper. Thus we do not know if there was any treatment benefit. The cohort obtained similar therapeutic control to the original trial populations, but this is not a dramatic finding given that it represents a similar population—that is, patients selected from medical outpatient clinics. This point seems to have been lost on Connolly, who in his editorial states, “They screened 2547 patients in general practice clinics.”3

As we have consistently argued, the findings of the original treatment studies and now these newer studies must be interpreted in the context of trials conducted on selected, secondary care, populations. It is still not clear how these findings translate to the general, primary care, population. We are currently undertaking a randomised controlled trial (Birmingham atrial fibrillation treatment with anticoagulation) of warfarin versus aspirin for patients aged 75 and over with atrial fibrillation who have been identified from general practice. This is a five year study and should answer the question of effectiveness within this defined population.

EDITOR—Kalra et al's paper extends our knowledge of how effectively research into stroke prevention in atrial fibrillation can translate into clinical practice.1 However, the authors studied patients recruited in district general hospitals and whose anticoagulants were controlled in hospital outpatient clinics. As a result the patients differed greatly from the whole population of patients with atrial fibrillation. The cohort studied was considerably older than that in the pooled atrial fibrillation …