Previous studies have shown that 10–30% haemodilution with crystalloid may induce a hypercoagulable state demonstrable by using the Thrombelastograph® (TEG). While most are in vitro studies, the few in vivo studies are limited by confounding surgical or ‘environmental’ factors. We conducted this randomized controlled study to evaluate the coagulation changes associated with in vivo haemodilution.
Methods
Twenty patients undergoing major hepatobiliary surgery were randomly allocated to one of two study groups. Group H (n=10) had 30% blood volume withdrawn over 30 min and replaced with saline. Group C (n=10) did not have any blood withdrawn. Blood samples were taken in both groups at 10, 20 and 30 min. Native TEG, complete blood count, coagulation profile, fibrinogen, antithrombin III, protein C and thrombin–antithrombin complex concentrations were measured.
Results
Compared with Group C, Group H patients had significantly greater shortening of r-time at 30 min (–30% vs +36%), greater shortening of k-time at all time points (–36% vs +17% at 10 min; –37% vs +44% at 20 min; –45% vs +49% at 30 min), and greater widening of α at 30 min (+71% vs +4%). The decrease in antithrombin III and other natural procoagulants and anticoagulants closely followed that of haematocrit, with the exception of thrombin–antithrombin complex.
Conclusion
In vivo haemodilution of up to 30% with saline can induce a hypercoagulable state. The mechanism remains unclear as disproportionate dilution of natural anticoagulants was not detected. Thrombin–antithrombin complex concentration remained stable despite haemodilution in Group H, which may suggest increased thrombin generation.
Declaration of interest. This study was supported by a CRCG grant from the Faculty of Medicine, the University of Hong Kong (Acc Code 337/038/0013). We also thank the Department of Surgery, the University of Hong Kong for supplying the TEG used in this study and Ms Jeff Man for her technical assistance.
