We thank Dr. Sanfilippo et al. for their interest in our recent study published in Anesthesiology1 about the comparison of two doses of tranexamic acid in adults undergoing cardiac surgery.
They criticize the choice of the primary outcome in our study, which was the number of patients who received at least 1 unit of blood product during the first postoperative week (including the intraoperative period).1 They would have preferred a shorter period of observation for the primary outcome because of the half-life of tranexamic acid which was discontinued at the end of surgery. We disagree with their point of view. First, transfusion at day 7 includes the first day, when the majority of the transfusions occurred: 62% of the transfused patients were transfused only on the first day, 77% of the packed erythrocytes and 71% of the fresh-frozen plasma were given during day 1. Furthermore, transfusion during the first day, including the intraoperative period is reported in table 6 and transfusion was similar between groups. Second, major surgical bleeding may have been responsible for delayed transfusion after the first 24 h. Third, several major articles about antifibrinolytic therapy used a 1-week (or even more) endpoint.2–4 Indeed, the goal of antifibrinolytic therapy is to decrease transfusion with its related complications, some being life threatening, and cost. Transfusion risk depends on the number of units or donors to which the patient is exposed during his whole hospitalization, not just during day 1. Thus, in our opinion, focusing on the first day is looking through the wrong end of the telescope. If one dose of tranexamic acid does not make a difference compared with another after the intraoperative period, then it is not superior.
Sanfilippo et al. asked for some details about postoperative procedures, such as antiplatelet therapy and anticoagulation. In fact, we forgot to mention in the article that postoperative care was adapted to the medical and surgical problems of each patient and did not differ from those usually practiced. It was indeed important for us to evaluate tranexamic acid in a “real-life situation,” once again to be sure to bring out a strong difference.
Finally, Sanfilippo et al. asked for the incidence of postoperative renal replacement therapy. Unfortunately, we did not collect these data. We would like to add that hemorrhagic events (pleural and pericardial effusions) after the first 24 h (day 2 to day 28) were rare: 12 in the low-dose group (4.3%) and 8 in the high-dose group (2.8%). The low incidence of such events attenuates the possibility that all these potential flaws may have biased our study.
The authors declare no competing interests.