Background: We aimed to validate the utility of shock index (SI) in predicting the need of blood transfusion and outcomes in patients with traumatic pelvic fracture (TPF).
Materials and methods: We conducted a retrospective analysis for patients who sustained TPF between 2012 and 2016 in a level 1 trauma center. Patients were categorized into patients with low versus high SI based on the cutoff obtained from the receiver operating characteristic curves to predict mortality.
Results: A total of 966 patients sustained TPF (28.5% had SI ≥ 0.9 based on receiver operating characteristic curves) with a median age of 33 (IQR 25-47) y. Type B and C pelvic fractures significantly had higher SI. The frequency of blood transfusion use was greater in patients with high SI (P = 0.001). SI correlated significantly with Injury Severity Score (r = 0.32), Revised Trauma Score (r = -0.40), and transfused blood units (r = 0.35). Patients with high SI had prolonged hospital length of stay and higher mortality (P = 0.001). SI ≥ 0.9 showed high sensitivity and negative predictive value to identify the need of massive blood transfusion (77% and 86%, respectively) and mortality (73.5% and 98.1%, respectively). For hospital mortality, high SI had a sensitivity of 73.5%, specificity 74%, negative predictive value 98%, and negative likelihood ratio of 0.36. After adjustment for age, sex, Injury Severity Score, Glasgow Coma Scale, pelvis Abbreviated Injury Scale, blood transfusion, and Tile classification, the multivariate analysis models showed that high SI was an independent predictor of blood transfusion (odd ratio 5.6) and mortality (odd ratio 3.63).
Conclusions: SI is a potentially useful instant tool for the prediction of massive transfusion and mortality in patients with TPF. Further prospective studies are warranted to support our findings.
Keywords: Blood transfusion; Injury; Mortality; Pelvic fracture; Shock index; Surgical intervention.
Copyright © 2019 Elsevier Inc. All rights reserved.