Résumé
<p><b>OBJECTIVE</b>To investigate the correlation and consistency between thromboelastography(TEG) and routine coagulation tests, and to evaluate the value of the two methods in determining the blood coagulation of patients.</p><p><b>METHODS</b>The TEG, routine coagulation tests and platelet counts of 182 patients from the Intensive Care Unit(ICU) and Department of Gastroenterology in our hospital from January to September 2014 were performed and analyzed retrospectively for their correlation, Kappa identity test analysis and chi-square test, and the diagnostic sensitivity and specificity of both methods in the patients with bleeding were evaluated.</p><p><b>RESULTS</b>The TEG R time and PT, R time and APTT showed a linear dependence (P<0.01). The relationship between the TEG K value, α-Angle, MA and Fibrinogen showed a linear dependence (P<0.001). And the relationship between the TEG K value, α-Angle, MA and the platelet count were in a linear dependent way (P<0.001). The Kappa values of the TEG R time with PT and APTT were 0.038 (P>0.05) and 0.061 (P>0.05), respectively. The chi-square test values of the TEG R time with PT and APTT were 35.309 (P<0.001) and 15.848 (P<0.001), respectively. The Fibrinogen and the TEG K value, α-Angle, MA value had statistical significance (P<0.001), with a Kappa value of 0.323, 0.288 and 0.427, respectively. The chi-square test values between Fibrinogen and the TEG K value, α-Angle, MA value were not statistically significant, with X2=1.091 (P=0.296), X2=1.361 (P=0.243), X2=0.108 (P=0.742). The Kappa values of the platelet count and the TEG K value, α-Angle, MA value were 0.379, 0.208 and 0.352, respectively, which were also statistically significant difference (P<0.001). The chi-square test values between the platelet count and the TEG K value, α-Angle, MA value showed a statistically significant difference (P<0.001), with X2=37.5, X2=37.23, X2=26.630. The diagnostic sensitivity of the two methods for the patients with bleeding was less than 50%.</p><p><b>CONCLUSION</b>There was a significant correlation between some TEG parameters and routine coagulation tests, but the consistency is weak. Moreover, the diagnostic sensitivity of two methods in the patients with bleeding is low. It was concluded that the TEG cannot replace the conventional coagulation tests, and the preferable method remains uncertain which could reflect the risk of bleeding.</p>