Comparison of base excess, lactate and pH predicting 72-h mortality of multiple trauma

OBJECTIVE: To compare the predictive values of base excess (BE), lactate and pH of admission arterial blood gas for 72-h mortality in patients with multiple trauma. METHODS: This was a secondary analysis based on a publicly shared trauma dataset from the Dryad database, which provided the clinical data of 3669 multiple trauma patients with ISS > = 16. The records of BE, lactate, pH and 72-h prognosis data without missing values were selected from this dataset and 2441 individuals were enrolled in the study. Logistic regression model was performed to calculate the odds ratios (ORs) of variables. Area under the curve (AUC) of receiver operating curve (ROC) was utilized to evaluate the predictive value of predictors for 72 h in-hospital mortality. Pairwise comparison of AUCs was performed using the Delong’s test. RESULTS: The statistically significant correlations were observed between BE and lactate (r = − 0.5861, p < 0.05), lactate and pH (r = − 0.5039, p < 0.05), and BE and pH (r = − 0.7433, p < 0.05). The adjusted ORs of BE, lactate and pH for 72-h mortality with the adjustment for factors including gender, age, ISS category were 0.872 (95%CI: 0.854–0.890), 1.353 (95%CI: 1.296–1.413) and 0.007 (95%CI: 0.003–0.016), respectively. The AUCs of BE, lactate and pH were 0.693 (95%CI: 0.675–0.712), 0.715 (95%CI: 0.697–0.733), 0.670 (95%CI: 0.651–0.689), respectively. CONCLUSIONS: There are significant correlations between BE, lactate and pH of the admission blood gas, all of them are independent predictors of 72-h mortality for multiple trauma. Lactate may have the best predictive value, followed by BE, and finally pH.