Comparison of ACT point-of-care measurements: repeatability and agreement

Abstract

Background: Accurate control of heparin anticoagulation is necessary during all stages of cardiopulmonary bypass (CPB). The activated clotting time, first described by Hattersley in 1966, is mostly used for determination of anticoagulation. Either celite or kaolin are used as activators. An ACT value of 480 sec is proposed to be the safe minimum level for anticoagulation during CPB. This study was designed to determine if the activated coagulation time (ACT) values of each analyser separately are repeatable, and to determine whether there exists a significant difference in ACT values measured by three different analysers: the GEM PCL (Instrumentation Laboratory), the Hemochron 801 (International Techni-dyne Corporation) and the ACT II Automated Coagulation Timer (Medtronic).Methods: All patients underwent cardiovascular surgical procedures requiring heparinisation (200 - 300 IU/kg). Blood samples for the measurement of the ACT were taken from all patients before and after heparinisation, during CPB, and after protamine administration. All samples were measured in duplicate with the three different analysers. To compare the activated clotting time data, the method described by Bland and Altman was used. The Pearson correlation coefficient was used to determine whether the differences were related to the average ACTs. p-Values < 0.05 were considered statistically significant.Results: The results showed that the three tested ACT analysers met the requirements of repeatability. The mean differences and standard deviations of the ACT values measured with the GEM PCL, the Hemochron 801, and the ACT II analyser were, respectively, -8.789 ± 37.61, -19.779±68.82, and -6.239±39.21, with p-values=0.177, 0.081 and 0.384, respectively. The Pearson correlation coefficients were too low (-0.012, -0.221 and -0.241, respectively) to show any correlation between the differences and the means. The ACT values measured with the Hemochron 801 were not significantly different from the ACT values measured with the ACT II analyser: ΔACT=-34.09 ± 146.68, with p=0.132. However, the GEM PCL did not agree with the Hemochron 801: ΔACT=-80.29 ± 143.06, with p=0.001, or the ACT II analyser: ΔACT=-119.139 ± 138.51, with p<0.001. A rather strong correlation was evident between the differences and the means measured with the GEM PCL compared with the Hemochron 801 (r=0.68) and the ACT II analyser (r=0.76).Conclusions: All analysers used celite or kaolin as activator. However, it was evident that the ACT measurements depended also on the analyser that had been chosen. A precaution that ACT values could not always be interpreted in the same way seems to be necessary.