I thank Drs. Vandenbroucke and Rosendaal for reinforcing the importance of old-fashioned history taking as a screening strategy for thrombophilic predisposition. In a litigious venue, unfortunately, the availability of a quantifiable test often means that the new technique supplants the old and becomes the standard. And after the presence of thrombophilia has been identified, doing nothing during pregnancy (a condition highly publicized as a hypercoagulable state) is risky even when doing something like anticoagulation has clearly defined risks. In light of new studies of genetic thrombophilic conditions, neither the older literature about pregnancy and thromboembolism nor the literature that deals exclusively with thromboembolic disease in nonpregnant patients is a reliable source of guidance.

There should be no debate about adequate heparinization for women who have thromboembolism during pregnancy or the postpartum period. There should also be no debate about the universal utility of prophylactic postural and mechanical procedures (such as use of elastic stockings) for pregnant women. Which pregnant women need pharmacologic prophylaxis, how much prophylaxis is necessary, and when prophylaxis should be administered need to be further elucidated. Any clinical decisions and studies must be stratified according to the personal and family histories of thromboembolic events and according to the results of evaluation for thrombogenic predisposition (Table 1).

I am not aware of any prospective studies examining the outcome in terms of prevention of thromboembolism with varying doses of heparin during gestation. I agree with Drs. Vandenbroucke and Rosendaal that, given the coagulation changes of pregnancy, a small amount of heparin should go a long way. Studies of heparin metabolism indicate that to treat active thromboembolism during pregnancy, a larger dose than expected may be necessary. However, we do not know whether this applies to prophylaxis in a patient with a personal or familial history of thromboembolic disease.