The activated clotting time in cardiac surgery: should Celite or kaolin be used?

Transcription

1 Interactive CardioVascular and Thoracic Surgery 24 (2017) doi: /icvts/ivw435 Advance Access publication 20 January 2017 ORIGINAL ARTICLE Cite this article as: De Vries AJ, Lansink-Hartgring AO, Fernhout F-J, Huet RCG, van den Heuvel ER. The activated clotting time in cardiac surgery: should Celite or kaolin be used? Interact CardioVasc Thorac Surg 2017;24: The activated clotting time in cardiac surgery: should Celite or kaolin be used? a b c d Adrianus J. De Vries a, *, Annemieke Oude Lansink-Hartgring b, Freek-Jan Fernhout c,rolfc.g.huet a and EdwinR.vandenHeuvel d Department of Anesthesiology, University Medical Center Groningen, University of Groningen, Groningen, Netherlands Department of Critical Care, University Medical Center Groningen, University of Groningen, Groningen, Netherlands Department of Cardiothoracic Surgery, University Medical Center Groningen, University of Groningen, Groningen, Netherlands Department of Mathematics and Computer Science, Eindhoven University of Technology, Eindhoven, Netherlands * Corresponding author. Department of Anesthesiology, University Medical Center Groningen, PO box , 9700 RB Groningen, Netherlands. Tel: ; fax: ; a.j.de.vries@umcg.nl (A.J. De Vries). ADULT CARDIAC Received 30 May 2016; received in revised form 28 November 2016; accepted 6 December 2016 Abstract OBJECTIVES: Both kaolin- and Celite-activated clotting times (ACT) are used to guide anticoagulation during cardiopulmonary bypass. It is unknown whether these methods lead to similar management procedures for anticoagulation in patients and are thus interchangeable in terms of bias, precision and variability. METHODS: We randomized 97 patients undergoing coronary artery bypass grafting or aortic valve replacement to either kaolin- or Celite-guided anticoagulation. The ACT was measured simultaneously with the other method. We administered 300 IU/kg heparin to obtain initial ACT values greater than 400 s and additional heparin in each group using the minimum value of duplicate measurements according to a predefined protocol. The primary end point was the total heparin dose and the number of heparin supplements. RESULTS: The total heparin dose per patient in the 48 Celite-guided patients was ± IU with 51 supplements and in the 49 kaolin-guided patients, ± IU (P = 0.77) with 56 supplements (P = 0.53). Postoperative thrombin generation time, fibrinolytic response time, chest tube loss and transfusion requirements were not different between the two groups. However, the methods differed in individual patients with regard to supplemental heparin (P = 0.002). Bias between methods at baseline was +10.3%, Celite being higher, and changed to a value of -12.9% at 2 h bypass. The coefficient of variation at baseline for individual patients was 2.6 times larger with kaolin than with Celite (P < 0.001). Correlation between ACT values at baseline was only 45%. CONCLUSIONS: Kaolin- and Celite-guided management of anticoagulation is clinically not different, but the methods are not interchangeable. Clinical registration number: identifier 1738 Keywords: Cardiopulmonary bypass Clinical Coagulation Aortic valve replacement Coronary artery bypass grafting INTRODUCTION The measurement of activated clotting time (ACT) using either kaolin or Celite is commonly used to manage anticoagulation for cardiopulmonary bypass (CPB). Hattersley first described the Celite ACT measurement in 1966 in an attempt to modify the coagulation time of whole blood by drawing it directly onto a contact activator [1]. Use of ACT measurements to guide heparin administration became widespread when their superiority over fixed heparin dosage schemes was demonstrated on postoperative blood loss [2]. In 1992, aprotinin, a serine protease inhibitor, was shown to prolong Celite ACT values [3] because it interferes with the kallikrein co-activation inhibiting activation of the Celite-activated contact pathway [4]. This appeared not to be the case with kaolin, which has a more negative charge; as a result, kaolin ACT has found considerable clinical application. Since then, kaolin- or Celite-based ACT measurements have been used worldwide. Despite the widespread use of these measurements, remarkably few studies have compared Celite and kaolin ACT measurements without aprotinin [5 10], and only one did so during CPB [9]. In these studies, the Hemochron (Int. Technidyne Co, Edison, NJ, USA) Celite ACT was compared with the Hemotec (Hemotec Inc. Englewood, CO, USA) kaolin ACT, but this approach introduces bias because these devices have different mechanisms of action and cannot be used interchangeably [10]. Clinical outcomes comparing Celite- and kaolin-guided management of anticoagulation during CPB have not been reported. VC The Author Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

2 550 A.J. De Vries et al. / Interactive CardioVascular and Thoracic Surgery Thus, we do not know whether the management of Celiteand kaolin-guided anticoagulation processes during CBP results in different clinical outcomes and whether the measurement methods are interchangeable between patients. We do not suggest changing from one ACT measurement method to the other in the same patient. We use only one method in each patient to guide the management of anticoagulation during CPB. The primary objective was to assess the clinical effects of Celite and kaolin anticoagulation in patients randomized to either kaolin- or Celite ACT-guided anticoagulation during CPB, using the total heparin dose per patient in each group, the number of heparin supplements and the circulating prothrombin fragments F1.2 and D-dimers after CPB as end points. The secondary objective was to assess the interchangeability of the methods by simultaneously measuring the ACT with both methods, using agreement in the decision to supplement heparin, bias, precision and variability as end points. METHODS Patients For this prospective randomized observational trial in a single university hospital in the Netherlands, we included patients scheduled for first-time elective coronary artery bypass grafting or aortic valve replacement after obtaining written informed consent. The study was approved by the ethics committee and registered ( identifier 1738). We excluded patients with known coagulopathy and patients on continuous heparin therapy. We randomized the patients using a computer-generated table to either Celite or kaolin ACT measurements to guide anticoagulation during CPB. We considered a difference in total heparin dose of 5000 IU clinical relevant. Retrospective data from our institution indicated a standard deviation of 9000 IU. With the usual assumptions of an alpha of 0.05 and a beta of 0.2, we determined that we needed a sample size of about 100 patients. Protocol The management of anticoagulation during CPB was, according to the randomization group, guided by duplicate measurements of either Celite ACT or kaolin ACT. In the patients, in whom the management of anticoagulation was guided by Celite ACT, we also measured the kaolin ACT and vice versa. These measurements were done to assess the agreement between heparin supplements in case the other measurement would have been used and to assess bias, precision and variability. We could therefore determine the interchangeability of the methods. We performed four ACT measurements (a duplicate Celite and a duplicate kaolin) for each blood sample. We took samples at the start of the operation, 3 min after administration of heparin, 5 min after the start of CPB and then every 30 min and after the reversal of heparin with protamine sulphate. We limited the number of blood samples to 8 per individual. Thus, we used a maximum of five time points on CPB over a 2-h CPB period. Before and after CPB, we took blood samples from the patient s unheparinized arterial line. During CPB we used the sample line of the heart lung machine. The protocol for the administration of heparin was as follows: We used an initial dose of porcine heparin of 300 IU/kg bodyweight (LEO Pharma, Amsterdam, The Netherlands) and an ACT value of at least 400 s for the start and maintenance of CPB. If the lowest value of the duplicate measured ACT values in each group was less than 400 s, we gave 100 IU/kg additional heparin. If this value was less than 350 s, we gave 150 IU/kg, and if this value was less than 300 s, we gave 200 IU/kg. For the measurements, we used the Hemochron Response system (Hemochron, Int Technidyne Co, Edison, NJ). All systems were serviced and approved by the manufacturer prior to the study. The measurements were performed simultaneously using two different devices, randomly chosen from a pool of seven devices. The device was prewarmed and the test tubes were at room temperature. The test tubes contained 12 mg Celite or 12 mg kaolin. When a measurement reached 1000 s, the measurement was stopped and this value was entered; in a number of these cases, a clot appeared to have formed in the test tube without capture of the magnet. Protamine was administered in a 1:1 ratio to the initial dose of heparin. Cardiopulmonary bypass The CPB circuit was not coated and consisted of roller pumps and an open venous reservoir, primed with 1000 ml Ringer s solution, 500 ml hydroxyethylstarch 6% and 5000 IU of heparin. Pump flow was 2.4 l/m 2 /min. Neither retrograde priming nor haemoconcentrators was used, resulting in an average haemodilution of 33%. Temperature was allowed to drift to 34 C. Coagulation and blood activation To determine the coagulation status and blood activation, blood samples were taken before and after the operation. Prothrombin time, activated partial thromboplastin time and the plasma levels of antithrombin III and fibrinogen were determined by routine laboratory measurements. Blood for the determination of D-dimers and F1.2 fragments was centrifuged at 1000 rpm for 10 min. The plasma was divided into aliquots and stored at -80 C until analysis. Analysis was done with a turbidimetric assay (Roche, Woerden, The Netherlands) for D-dimers and with an enzyme immunoassay (Siemens, Den Haag, The Netherlands) for prothrombin fragments F1.2. At the discretion of the attending anaesthesiologist or surgeon, some patients received tranexamic acid 2 g before the operation. End-point and statistical analysis The primary end point was the total heparin dose per patient in each group and the use of heparin during surgery (heparin supplements). Secondary end points were the D-dimer and F1.2 concentrations, postoperative chest tube loss, use of blood and blood products and length of stay in the intensive care unit. Data were analysed in two ways using intention to treat. First, an independent sample analysis with a two-sample Student t-test or a Wilcoxon rank sum test was used to determine clinical differences between the kaolin- and Celite-guided patient groups. Second, a paired data analysis was conducted to determine the interchangeability of the methods. In this analysis, we compared the duplicate Celite measurements with the duplicate kaolin measurements from each blood sample. Linear mixed models were used on the lowest of the duplicate ACT values after a logarithmic transformation, and generalized linear mixed models were applied to the use of heparin (yes/no) and the number of

3 A.J. De Vries et al. / Interactive CardioVascular and Thoracic Surgery 551 heparin supplements. Calculation of agreement was based on the intraclass correlation coefficients. Bland Altman plots were used to visualize bias and precision. RESULTS We studied 97 patients, 48 in the Celite-guided group and 49 in the kaolin-guided group. The demographic and intraoperative data are shown in Table 1. When we compared the results of the Table 1: Demographic and intraoperative data Celite (n = 48) Kaolin (n = 49) Age, years 65 ± ± 9.3 Sex, male/female 29/19 34/15 Weight, kg 81 ± ± 13.3 Height, cm 173 ± ± 7.7 Haemoglobin, mmol/l 7.8 ± ± 0.9 Platelets, 10 9 /l 243 ± ± 72 Prothrombin time, s 11.3 ± ± 0.9 APTT, s 26.7 ± ± 1.9 Fibrinogen, g/l 3.4 ± ± 0.9 Antithrombin III, % 85 ± ± 12 D-dimer, ng/ml 569 ± ± 822 F1.2 fragment, pm/l 209 ± ± 96 Preoperative medication Aspirin, n (%) 30 (63) 29 (59) LMWH, n (%) 18 (38) 15 (31) Intraoperative data CABG, n (%) 32 (67) 31 (63) AVR, n (%) 16 (33) 18 (37) Bypass time, min 103 ± ± 25 Haemoglobin, mmol/l 5.1 ± ± 0.82 Tranexamic acid, n (%) 19 (40%) 27 (55%) Red blood cells, n (%) 9 (19) 7 (14) Fresh-frozen plasma, n (%) 0 (0) 0 (0) Platelets, n (%) 0 (0) 0 (0) APTT: activated partial thromboplastin time; LMWH: low molecular weight heparin; CABG: coronary artery bypass grafting; AVR: aortic valve replacement. patients from the two treatment groups the independent samples, we found no real difference in total heparin dose per patient and in the number of heparin supplements (Table 2). There were also no differences across the two groups in concentrations of D-dimer and F1.2 fragments after CPB, in the routine postoperative coagulation assessments, in chest tube loss and in the transfusion of blood products (Table 2). However, when we compared the results between the two methods in the individual patient, the paired data analysis, we found significant differences. The average number of heparin supplements proposed for the same individuals by the two methods was significantly lower for kaolin than for Celite (P = 0.002). This finding is illustrated by the difference in probability of ACT values below 400 s between the two methods, which increased over time (Fig. 1). Furthermore, the agreement between both methods on heparin supplements in a patient gradually decreased from 55% for measurements just before the start of CPB to 17.5% after 60 min of CPB. At 90 min of CPB, the agreement increased slightly to 33%. Heparin supplements resulted in an average increase of 68 ± 92 s for the next Celite measurement and 71 ± 175 s for the next kaolin measurement. The bias and coefficient of variation between the two methods varied significantly during the operation. At baseline, the Celite measurements were on average 10.4% higher than the kaolin measurements, but during CPB the kaolin measurements had higher minimum ACT values (Fig. 2). The coefficient of variation (in percentages) was significantly smaller for Celite at baseline (P < 0.001), but this difference seemed to vanish (P > 0.05) during the operation (Fig. 3). The correlation in the minimum ACT values between kaolin and Celite varied from 0% to 59.2% during the operation. Graphically, the bias and precision of the Celite and the kaolin measurements are shown in the Bland-Altman plots in Fig. 4. The plots reflect the increase in differences between the duplicate measurements. These differences were, for Celite, 7 ± 10 s at baseline, 58 ± 72 s after heparin and 66 ± 89 s 5 min after the start of CPB. For kaolin, these differences were 9 ± 15, 70 ± 119 and 127 ± 180 s, respectively. We occasionally measured values greater than 1000 s for one of the two readings in 118 (13%) of the kaolin measurements and in 39 (4%) of the Celite measurements (P < 0.001). We analysed the occurrence of these values and found that 102 of ADULT CARDIAC Table 2: Postoperative data Celite (n = 48) Kaolin (n = 49) P-value Total heparin dose, IU ± ± Heparin supplements, n Haemoglobin, mmol/l 5.8 ± ± Platelets, x10 9 /l 156 ± ± Prothrombin time, s 12.9 ± ± Activated partial thromboplastin time, s 30.2 ± ± Fibrinogen, g/l 2.1 ± ± Antithrombin III, % 55 ± ± D-dimer, ng/ml 2836 ± ± F1.2 fragment, pm/l 1325 ± ± Postoperative blood loss, ml 613 ± ± Red blood cells, n (%) 6 (13) 7 (14) 0.22 Fresh frozen plasma, n (%) 1 (2) 2 (4) 1.00 Platelets, n (%) 1 (2) 1 (2) 1.00 Intensive care stay, days 1.4 ± ± APTT: activated partial thromboplastin time.

4 552 A.J. De Vries et al. / Interactive CardioVascular and Thoracic Surgery patients in more detail, we found that the total heparin dose and number of heparin supplements did not differ (Table 3). There were no differences in the Celite and kaolin ACT measurements after heparin and 5 min after the start of CPB. The postoperative concentration of D-dimers was lower, but this was not the case for F1.2 fragments. Postoperative chest tube loss was also lower but had no effect on blood transfusion requirements (Table 3). DISCUSSION Figure 1: The probability of finding an activated clotting time of less than 400 s in 97 patients when the management of anticoagulation is guided by Celite or kaolin measurements. The number of patients is indicated across the top of the graph and decreases due to the duration of the cardiopulmonary bypass. Figure 2: Bias between paired Celite- and kaolin-activated clotting time measurements in 97 patients at baseline, during cardiopulmonary bypass and after reversal with protamine. The bias reflects the minimum of the duplicate measured values that were used for the decision to supplement heparin. Figure 3: Coefficient of variation in duplicate Celite- and kaolin-activated clotting time measurements in 97 patients at baseline, during cardiopulmonary bypass and after reversal with protamine. these 157 measurements (65%) were apparently false readings. However, in the remainder, these values seemed coincidental with particular patients. Values greater than 1000 s did not occur in the baseline measurements or after heparin reversal. These out-of-range values did not play a dominant role clinically because we used the minimum ACT values for the decision to supplement heparin. Tranexamic acid was used in 46 patients and was not different between the groups (P = 0.126). When we examined these We found no differences in the total heparin dose or in the number of heparin supplements during coronary artery bypass grafting or aortic valve replacement when anticoagulation management during CPB was guided by either Celite or kaolin ACT measurements. We found no differences in thrombin activation and fibrinolytic response after CPB, suggesting a similar degree of anticoagulation during CPB. We also found no differences in other clinical measurements such as blood transfusion requirements and chest tube loss. These findings indicate that the management of Celite- and kaolin-guided anticoagulation led to similar clinical results in this patient group. However, when we used the paired data analysis at the individual patient level, we found significant differences in mean values and precision and only a moderate agreement in the decision to supplement heparin between the measurements. Of note was the variation in bias and coefficient of variation during surgery after logarithmic conversion of the data. This finding is a strong argument against the interchangeability of the measurement methods. Different measurement methods typically have a constant bias that can be taken into account for comparisons. In contrast, we found an increase in bias during CPB. We have no explanation for this increase, except that Celite and kaolin exert different effects on heparin anticoagulated blood through their differences in electrical charges. A related relevant point concerns the use of duplicate measurements. Few studies report the results of duplicate ACT measurements. We found values in mean differences in duplicate Celite ACT measurements and in the coefficient of variation similar to those that Gravlee et al. [8] reported in the period before CPB. During CPB, however, we found that these differences increased. Given the similarities with the findings of Gravlee et al. before CPB, we can safely assume that these larger differences are real. Different measurement devices, even those from the same manufacturer, may give different results. For every patient in this study, the attending perfusionist chose 2 devices from a pool of 7. Thus, all devices were used by chance alone in the individual patients. It seems unlikely that the devices had a measurable systemic effect on our results. The Bland-Altman plots demonstrated a better precision for the Celite ACT measurements. Bosch et al. [9], who found a variation and a coefficient of variation similar to the ones we found, made the same observation. Flom-Halvorsen et al. [11] also noted substantial variability in the Celite ACT measurements and therefore proposed duplicate measurements. The use of duplicate measurements was also promoted by Bennet and Horrow using Celite ACT measurements [12]. In these last two studies, the mean value of the duplicate measurements was used to allow for the variability in the measurements. We agree that duplicate measurements should be used, but we propose to use the lowest of the duplicate values for the decision to supplement heparin

5 A.J. De Vries et al. / Interactive CardioVascular and Thoracic Surgery 553 ADULT CARDIAC Figure 4: Bland Altman plots of 97 patients with the limits of agreement between the duplicate Celite (upper row) and kaolin (lower row) measurements at baseline, after heparin and 5 min after the start of cardiopulmonary bypass. Table 3: Effects of tranexamic acid Tranexamic acid (n = 46) No tranexamic acid (n = 51) P-value Celite ACT after heparin, s 516 ± ± Kaolin ACT after heparin, s 529 ± ± Celite ACT 5 min bypass, s 542 ± ± Kaolin ACT 5 min bypass, s 585 ± ± Bypass time, min 104 ± ± Total heparin dose, IU ± ± Heparin supplements, n D-dimer, ng/ml 1640 ± ± F1.2 fragment, pm/l 1158 ± ± Postoperative blood loss, ml 464 ± ± Red blood cells, n (%) 12 (26) 13 (25) Intensive care stay, days 2.1 ± ± ACT: activated clotting time. because it minimizes the effect of out-of-range values (greater than 1000 s). These values occurred more often in the kaolin than in the Celite measurements. Although the majority of them appeared to occur at random, a substantial number of these values could be attributed to particular patients. These out-of-range values have not been reported before and may be explained by the inhomogeneous distribution of the activator on the magnet or the support frame of the tubing that reduced the activation surface, although the samples were vigorously shaken as per the manufacturer s instructions. Capture of the magnet in the test tube is related to the volume, spread and firmness of the blood clot around the magnet and therefore requires a maximal surface area. Occasional sedimentation of the more negatively charged kaolin particles may explain the differences between the Celite and kaolin measurements. Thus, these out-of-range values are inherent in the measurement method. With increasing bypass times, the probability for a lower ACT measurement increased. We know that prothrombin is activated during CPB and that antithrombin III levels decrease, partly through binding to factor VII [13]. This situation may lead to an imbalance in the coagulation system, resulting in lower ACT measurements. The plasma half-life of heparin may also play a role, although it is known that ACT measurements do not accurately reflect plasma heparin levels [14]. A number of patients received tranexamic acid, and we analysed this group separately. Bechtel et al. [15] found no difference between Celite and kaolin measurements in patients who received tranexamic acid. Our results support their findings, but in addition, we found no difference between Celite and kaolin measurements regardless of whether tranexamic acid was or was not used. As expected, D-dimers and chest tube losses were lower, but this finding did not reduce the number of patients

6 554 A.J. De Vries et al. / Interactive CardioVascular and Thoracic Surgery who required blood transfusions. These results agree with those of Casati et al. [16]. A unique aspect of our study is that we not only had a large dataset available for the assessment of bias, precision and variability, but we could also assess the clinical effects of each ACT method using the same technology. The clinical implication is that, if one follows the same protocol for administration of heparin, independent of the monitoring method, similar results of safe anticoagulation during CPB are achieved, such as no difference in bleeding or transfusion rates and a similar degree of blood activation. In this clinical study, we noted 25 (7%) protocol violations in heparin supplementation in 18 patients. These violations occurred both in the Celite (11) and kaolin (14) groups. In 11 patients, more heparin was administered than per the protocol; in 7 patients, heparin should have been withheld. We analysed the data with intention to treat and did not exclude these patients. Moreover, the data from these patients were useful for the paired data analysis. When we re-analysed the clinical data with the exclusion of these 18 patients, the results did not change. A limitation of our study is that we did not measure anti-xa levels and heparin concentrations. However, there is currently no gold standard against which to compare the different ACT measurements. The interpretation of anti-xa levels during CPB is not clear, and heparin concentrations have also been shown to have only a weak relation to ACT values [14]. Moreover, measurement of heparin concentrations or anti-xa levels would not change our main findings. In conclusion, the management of Celite- and kaolin-guided anticoagulation processes is, from a clinical standpoint, not different in patients undergoing coronary artery bypass grafting or aortic valve replacement when the same protocol for heparin administration is used and results in safe anticoagulation during CPB. However, the methods are not interchangeable because of significant differences in bias, precision, variability, out-of-range values and the decision to supplement heparin. In this respect, Celite-guided anticoagulation appears to be more accurate and should be the preferred method. An increasing variability in bias, combined with the probability for a lower ACT measurement, is a concern for CPB periods in excess of 2 h. Conflict of interest: none declared. REFERENCES [1] Hattersly PG. Activated coagulation time of whole blood. JAMA 1966;196: [2] Babka R, Colby C, El-Etr A, Pifarre R. Monitoring of intraoperative heparinization and blood loss following cardiopulmonary bypass surgery. J Thorac Cardiovasc Surg 1977;73: [3] Wang JS, Lin CY, Hung WT, Karp RB. Monitoring of heparin-induced anticoagulation with kaolin activated clotting time in cardiac surgical patients treated with aprotinin. Anesthesiology 1992;77: [4] Dietrich W, Jochum M. Effect of Celite and kaolin on activated clotting time in the presence of aprotinin: activated clotting time is reduced by binding of aprotinin to kaolin. J Thorac Cardiovasc Surg 1995;109: [5] Doherty TM, Shavelle RM, French WJ. Reproducibility and variability of activated clotting time measurements in the cardiac catheterization laboratory. Catheter Cardiovasc Interv 2005;65: [6] Chia S, Van Cott EM, Raffel C, Jang IK. Comparison of activated clotting times obtained using Hemochron and Medtronic analysers in patients receiving anti-thrombin therapy during cardiac catheterisation. Thromb Haemost 2009;101: [7] Thenappan T, Swamy R, Shah A, Nathan S, Nichols J, Bond L et al. Interchangeability of activated clotting time value across different point of care systems. Am J Cardiol 2012;109: [8] Gravlee GP, Case LD, Angert KC, Rogers AT, Miller GS. Variability of the activated clotting time. Anesth Analg 1988;67: [9] Bosch YPJ, Ganushchak YM, De Jong DS. Comparison of ACT point-ofcare measurements: repeatability and agreement. Perfusion 2006;21: [10] Avendano A, Ferguson JJ. Comparison of Hemochron and HemoTec activated coagulation time target values during percutaneous transluminal coronary angioplasty. JACC 1994;23: [11] Flom-Halvorsen HI, Øvrum E, Abdelnoor M, Bjørnsen, Brosstad F. Assessment of heparin anticoagulation: Comparison of two commercially available methods. Ann Thorac Surg 1999;76: [12] Bennet JA, Horrow JC. Activated coagulation time: one tube or two? J Cardiothorac Vasc Anesth 1996;10: [13] Davidson SJ, Woodhams B. Factor VIIa-antithrombin complexes during cardiac surgery using cardiopulmonary bypass. Int Jnl Lab Hem 2011;33: [14] Despotis GJ, Summerfield AL, Joist JH, Goodnough LT, Santoro SA, Spitznagel E et al. Comparison of activated coagulation time and whole blood heparin measurements with laboratory plasma anti-xa heparin concentration in patients having cardiac operations. J Thorac Cardiovasc Surg 1994;108: [15] Bechtel JFM, Prosch J, Sievers H-H, Bartels C. Is the kaolin or Celite activated clotting time affected by tranexamic acid? Ann Thorac Surg 2002;74: [16] Casati V, Della Valle P, Benussi S, Franco A, Gerli C, Baili P et al. Effects of tranexamic acid on postoperative bleeding and hematochemical variables in coronary surgery: comparison between on-pump and off-pump techniques. J Thorac Cardiovasc Surg 2004;128:83 91.