In Response

We thank Carroll et al1 for their interesting investigation in the relationship between activated clotting time (ACT) and plasma anti-Xa measurements, and for their “call to action” to research the validity of ACT as a true gold standard. The flaws of the ACT test and the limitations of unfractionated heparin (UH) are discussed in our recent publication of STS/SCA/AmSECT Clinical Practice Guidelines: Anticoagulation During Cardiopulmonary Bypass,2 and we agree with the appraisal by Carroll et al.1 ACT tests vary and are dependent on variables such as sample volume, activator used, and the platform on which the test is performed. The ACT relationship with heparin level is questioned by multiple studies that showed a very large coefficient of variance among ACT tests, at baseline and especially in the heparinized state.3–5 One of these studies showed that anti-Xa levels remain constant even in the face of increasing ACT values, a fact in support of a poor correlation between ACT and heparin concentration.3 In the experiment by Carroll et al,1 a patient with a heparin level <3.5 IU/mL still had an ACT >600 seconds, and a patient with a heparin level >9.5 IU/mL had an ACT <400 seconds. To embrace the ACT as a gold standard for anticoagulation for cardiopulmonary bypass, one must ask, “What is the actual goal of heparinization?” In an ideal situation, that goal is the attainment of an unfettered degree of “reversible” anticoagulation through the anti-Xa and antithrombin activities of heparin. Thus, a heparin level alone would not be sufficient without documentation of a functional test of clotting. Similarly, a high level of functional anticoagulation alone would not ensure safety, if that anticoagulant state was not the result of heparin therapy. In the absence of a better test of high-dose heparin anticoagulation, would the conduct of cardiopulmonary bypass be safer if we recommended that all patients have anticoagulation monitored with both ACT and a heparin (or anti-Xa) level? There are certainly patient populations in whom the risks of underanticoagulation are real and may not be detected by ACT alone. These include neonates, small children, and coagulopathic patients. In these patients, a supplemental test of heparin concentration could accompany the ACT to ensure adequate drug effect. In other patient groups, it may be very difficult to achieve adequate anticoagulation using UH due to patient factors such as sepsis and thrombocytosis. In these latter patient groups, a better anticoagulant than UH or a strategy for the use of antithrombin is desired. We agree with Carroll et al1 that this is an important area for further research in cardiac surgery. By today’s regulatory standards, it is questionable whether UH and the ACT test would stand up to the rigors of clinical investigation and whether they would be approved in the current era for use in cardiac surgical procedures. Linda Shore-Lesserson, MDDepartment of AnesthesiologyNorthshore University HospitalManhasset, New York[email protected] Victor Ferraris, MDDivision of Cardiovascular and Thoracic SurgeryUniversity of KentuckyLexington, Kentucky David Fitzgerald, MPH, CCP, LPDivision of Cardiovascular PerfusionMedical University of South CarolinaCharleston, South Carolina Robert A. Baker, PhD, CCPCardiac Surgery Research and PerfusionFlinders University and Flinders Medical CentreAdelaide, South Australia, Australia