Abstract
Background. The Quick (plain thromboplastins) and Owren (combined thromboplastins, with the addition of bovine plasma with vitamin K‐dependent factors depleted) prothrombin time (PT) methods are used for measuring prothrombin time; for instance, in monitoring anticoagulation therapy with vitamin K antagonists. In most Quick PT methods, the final plasma dilution is 1/3 compared to Owren methods, which have a final dilution of 1/21. The Quick PT methods are associated with larger inter‐laboratory variations than the Owren PT methods. Objectives. We investigated whether dilution of the sample or thromboplastin has any impact on the correlation between the different PT methods. Material and methods. Plasma samples were analysed undiluted and prediluted in buffer, with adsorbed bovine plasma added, and by using rabbit brain or human placenta thromboplastins. Bootstrapping (re‐sampling) was utilized to estimate the difference in correlations between PT of different reagents at different dilutions of plasma. Results and conclusions. We found the best correlations for combined reagents when the sample was prediluted 7‐fold (r = 0.95) or more, and lowest when the sample was undiluted (r = 0.67). The source of thromboplastin had only a minor impact, if any, on the PT results. It was impossible to test whether predilution of plasma samples improved the correlation between Quick PT methods due to absence of clotting in many samples. We suggest that the Owren‐style sample predilution is preferable for the harmonization of PT results and to overcome the large inter‐laboratory variations that are associated with Quick PT methods to the benefit of patients on oral anticoagulation.