Temperature corrected thrombelastography in hypothermic patients

Thrombelastograms and other coagulation studies are performed at 37°C, regardless of the patient's body temperature. This prospective study of 45 patients undergoing orthotopic liver transplantation was conducted to evaluate the effect on the thrombelastogram performed at the patient's actual body temperature compared with a control thrombelastogram heated in the standard fashion to 37°C. Thrombelastograms were obtained after the induction of anesthesia and at various times throughout the operation when clinically indicated. A freshly drawn sample of the patient's blood was divided into two aliquots and run simultaneously on two thrombelastographs; one thrombelastograph was modified with a thermostat to perform the test at the patient's body temperature and the other was unmodified to serve as a control. The temperature of the patients in this study ranged from 36.9°C to 32°C. The variables of the thrombelastogram measured were: r (reaction time in minutes), r + K (coagulation time in minutes), α (coagulation rate in degrees), and MA (maximum amplitude in millimeters). Whenever the patient's body temperature was less than 37°C, statistically significant prolongation of the reaction time, coagulation time, and decrease in the clot formation rate occurred compared with control variables at 37°C. Overall means were as follows: r for control, 8.24 ± 0.28 min; r for temperature corrected, 9.32 ± 0.27 min; r + K for control, 15.4 ± 0.65 min; r + K for temperature corrected, 17.5 ± 0.81 min; and α for control, 39.8 ± 1.22°; α for temperature corrected, 37.7 ± 1.23°. The MA was not significantly different between the temperature corrected thrombelastograms and the control thrombelastograms regardless of the patient's body temperature. These changes suggest that performing thrombelastograms in the standard fashion may overestimate the quality of the coagulation system in vivo when the patient's temperature is less than 37°C.