Objective  To construct a prediction model based on early viscoelastic coagulation function and analyze its predictive value for coagulation dysfunction in heat stroke patients.

  Methods  A total of 100 heat stroke patients treated from June 2017 to October 2020 were included in this study. Early viscoelastic coagulation function indicators were collected upon admission, and the occurrence of coagulation dysfunction was recorded. Multivariate logistic regression analysis was used to analyze the correlation between early viscoelastic coagulation function indicators and the occurrence of coagulation dysfunction. Receiver operating characteristic curves(ROC) were drawn to evaluate the efficacy of the prediction model in predicting the occurrence of coagulation dysfunction in heat stroke patients.

  Results  Among the 100 heat stroke patients, 33 patients had occurrence of coagulation dysfunction. The patients with coagulation dysfunction had higher levels of C-reactive protein(CRP), prothrombin time(PT), platelet count, clot rate(CR), and platelet factor(PF), as well as lower levels of activated clotting time(ACT) and fibrinogen(FI), compared to the patients without coagulation dysfunction(all indicators, P < 0.05). Multivariate logistic regression analysis showed that elevated CRP(OR=1.616), CR(OR=1.240), and PF(OR=2.132) were risk factors for coagulation dysfunction in heat stroke patients(P < 0.05), while elevated ACT(OR=0.885) was a protective factor(P < 0.05). ROC curve analysis showed that the AUC of the prediction model established with CRP, ACT, CR, and PF was 0.929 for predicting coagulation dysfunction in heat stroke patients, with a sensitivity of 83.41% and a specificity of 79.22%.

  Conclusions  Elevated levels of ACT, CR, and PF at early stages were influencing factors for the occurrence of coagulation dysfunction in heat stroke patients. Early viscoelastic coagulation function indicators could be used for early prediction of coagulation dysfunction in heat stroke patients.