Dear Sir, Without wishing to detract from the recent very useful paper on the optimal duration of anticoagulant therapy by Pinede et al., I would like to point out a limitation to one of their conclusions that clinicians should consider before modifying their practice [1]. Pinede et al. concluded that longer term anticoagulation therapy was no more likely to cause serious haemorrhagic complications than short-term therapy (3–6 weeks). However, this conclusion is based exclusively on the results of prospective controlled trials. Many patients were excluded from participation in these trials because of medical conditions that often complicate venous thromboembolism. In everyday practice clinicians often do not have the luxury of being able to withhold anticoagulation treatment in such patients. Anticoagulation therapy is also likely to be much more closely monitored in controlled trials than in routine practice. Observational studies of ‘real life’ situations give a more accurate picture of the likelihood of haemmorrhagic complications, and show that risk is related to the duration of therapy [2, 3].


  1. Top of page
  2. References

Received 10 July 2000; accepted 23 August 2000.