Legnani C, Scharrer I, Schulman S, van der Meer FJ. Increased fetal loss in women with heritable thrombophilia. More on: high thrombin generation and the risk of recurrent venous thromboembolism Recent studies have shown that high thrombin generation is associated with the risk of first [1] and recurrent venous thromboembolism (VTE) [2][3][4]. This is not unexpected as high thrombin generation is a consequence of hypercoagulability due to elevated levels of procoagulant factors or reduced levels of anticoagulant factors. Thrombin generation is measured after activation of coagulation by small amounts of tissue factor as trigger and phospholipids as platelet substitutes [5]. The ensuing thrombin generation curve is defined by many parameters, and those that are mostly used are the lag-time (i.e. the time interval from the addition of the trigger to the start of thrombin generation), the peak thrombin and the area under the thrombin generation curve, also called endogenous thrombin potential (ETP). In principle, each of the above parameters should reflect hypercoagulability. This is obvious for the peak thrombin and the ETP, but less for the lag-time. For practical purposes the lag-time may be considered as a conventional coagulation time. Now, it is well known that shortened coagulation times such as, for instance, the activated partial thromboplastin time (APTT) are risk factors for the occurrence [6] and recurrence of VTE [7,8]. Therefore, it is not surprising if shortened lag-times have recently been associated with the risk of VTE [3]. To our knowledge whilst peak thrombin and ETP have received attention as risk factors of recurrent VTE in some reports, the same attention was not devoted to the lag-time [1,2,4]. Furthermore, it is unknown whether and to what extent the individual parameters of thrombin generation when taken in combination may add clinical value to the risk assessment.To address this issue we reviewed our results of a recent study on the risk of recurrent VTE and thrombin generation [3]. In that study we had found that high thrombin generation when measured in the presence of thrombomodulin was associated with an increased risk of recurrent VTE. The hazard ratio (95% confidence interval) for recurrent VTE in the multivariate analysis was 3.19 (1.29-7.89) for the lag-time, 3.41 (1.34-8.68) for peak thrombin and 4.57 (1.70-12.2) for ETP. In the present analysis patients were allocated according to a score system that took into consideration normal/abnormal values for each of the three parameters.Values for ETP and peak thrombin were considered as normal if they fell within the limits encompassing the first and second tertile of the distribution for all patients.Likewise, values for lag-time were considered normal if they fell within the limits encompassing the second and third tertile. Accordingly, patients had been assigned score 0 if they had all the parameters that were normal, and score 1, 2 or 3 if they had at least 1, 2 or 3 parameters that were abnormal. The distribution of patients acros...