To the Editor.
—The article by Drs Imperiale and Speroff1 contains important errors that result in a considerable understatement of the efficacy of adjusted-dose warfarin sodium in preventing PE. The 10 articles cited (Table) yield 994 patients receiving warfarin (not 864 as stated in their article) with nine definite PEs (not 23). Thus, the crude incidence of PE would be 0.91%, not 2.7%. This is almost the same incidence as with LMWH.Second, the analysis includes a series by Fordyce et al2 that used fixed minidose warfarin (1 mg daily, irrespective of the prothrombin time). This method of prophylaxis was shown to be completely ineffective by Fordyce et al, is entirely different from adjusted-dose warfarin, and should be put in a group by itself. With this series excluded, the crude incidence of proven PE decreases to just 0.43%, half that with LMWH.These errors completely alter the findings
Chell J, Clark D, Howard PW. Prophylaxis Against Venous Thromboembolism After Hip Surgery. JAMA. 1995;273(4):287-288. doi:10.1001/jama.1995.03520280031031