Letters Section Editor: Robert M. Golub, MD, Senior Editor.
In Reply: We agree with Drs Krakow and Ortel that generalizability is important. Whether any study, be it a randomized controlled trial (RCT) or an epidemiologic study, can be applied to a given patient is a question of external validity. We recognize that RCT methodology limits external validity. However, the blanket statement that “the appropriate incidence rate ratio for bleeding is not one pooled from clinical trials, but from epidemiologic studies” is too broad.
Ost D. Continuing Anticoagulation Following Venous Thromboembolism—Reply. JAMA. 2005;294(24):3088-3089. doi:10.1001/jama.294.24.3088-b