PURPOSETo compare the effects of empirical and modified hemostatic resuscitation for liver blast injury combined with seawater immersion.METHODSThirty rabbits were subjected to liver blast injury combined with seawater immersion, and were then divided into 3 groups randomly (n = 10 each): group A (no treatment after immersion), group B (empirical resuscitation with 20 mL hydroxyethyl starch, 50 mg tranexamic acid, 25 IU prothrombin complex concentrate and 50 mg/kg body weight fibrinogen concentrate), and group C (modified resuscitation with additional 10 IU prothrombin complex concentrate and 20 mg/kg body weight fibrinogen concentrate based on group B). Blood samples were gathered at specified moments for assessment of thromboelastography, routine coagulation test, and biochemistry. Mean arterial pressure, heart rate, and survival rate were also documented at each time point. The Kolmogorov-Smirnov test was used to examine the normality of data distribution. Multigroup comparisons were conducted with one-way ANOVA.RESULTSLiver blast injury combined with seawater immersion resulted in severe coagulo-fibrinolytic derangement as indicated by prolonged prothrombin time (s) (11.53 ± 0.98 vs. 7.61 ± 0.28, p<0.001), activated partial thromboplastin time (APTT) (s) (33.48 ± 6.66 vs. 18.23 ± 0.89, p<0.001), reaction time (R) (min) (5.85 ± 0.96 vs. 2.47 ± 0.53, p<0.001), decreased maximum amplitude (MA) (mm) (53.20 ± 5.99 vs. 74.92 ± 5.76, p<0.001) and fibrinogen concentration (g/L) (1.188 ± 0.29 vs. 1.890 ± 0.32, p = 0.003), and increased D-dimer concentration (mg/L) (0.379 ± 0.32 vs. 0.051 ± 0.03, p = 0.005). Both empirical and modified hemostatic resuscitation could improve the coagulo-fibrinolytic states and organ function, as indicated by shortened APTT and R values, decreased D-dimer concentration, increased fibrinogen concentration and MA values, lower concentration of blood urea nitrogen and creatine kinase-MB in group B and group C rabbits in comparison to that observed in group A. Further analysis found that the R values (min) (4.67 ± 0.84 vs. 3.66 ± 0.98, p = 0.038), APTT (s) (23.16 ± 2.75 vs. 18.94 ± 1.05, p = 0.001), MA (mm) (60.10 ± 4.74 vs. 70.21 ± 3.01, p < 0.001), and fibrinogen concentration (g/L) (1.675 ± 0.21 vs. 1.937 ± 0.16, p = 0.013) were remarkably improved in group C than in group B at 2 h and 4 h after injury. In addition, the concentration of blood urea nitrogen (mmol/L) (24.11 ± 1.96 vs. 21.00 ± 3.78, p = 0.047) and creatine kinase-MB (U/L) (85.50 ± 13.60 vs. 69.74 ± 8.56, p = 0.013) were lower in group C than in group B at 6 h after injury. The survival rates in group B and group C were significantly higher than those in group A at 4 h and 6 h after injury (p < 0.001), however, there were no statistical differences in survival rates between group B and group C at each time point.CONCLUSIONSModified hemostatic resuscitation could improve the coagulation parameters and organ function better than empirical hemostatic resuscitation.