Background: Massive transfusion (MT) is required to resuscitate traumatically injured patients with complex derangements. Scoring systems for MT typically require laboratory values and radiological imaging that may delay the prediction of MT.
Study Design: The Trauma ALgorithm Examining the Risk of massive Transfusion (Trauma ALERT) study was an observational cohort study. Prehospital and admission ALERT scores were constructed with logistic regression of prehospital and admission vitals, and FAST examination results. Internal validation was performed with bootstrap analysis and cross-validation.
Results: The development cohort included 2,592 patients. Seven variables were included in the prehospital ALERT score: systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate (HR), respiratory rate (RR), SpO2, motor Glasgow Coma Scale (GCS) score, and penetrating mechanism. Eight variables from 2,307 patients were included in the admission ALERT score: admission SBP, HR, RR, GCS score, temperature, FAST examination result, and prehospital SBP and DBP.The area under the receiving operator characteristic curve for the prehospital and admission models were 0.754 (95% bootstrapped CI 0.735-0.794, P < 0.001) and 0.905 (95% bootstrapped CI 0.867-0.923, P < 0.001), respectively. The prehospital ALERT score had equivalent diagnostic accuracy to the ABC score (P = 0.97), and the admission ALERT score outperformed both the ABC and the prehospital ALERT scores (P < 0.0001).
Conclusion: The prehospital and admission ALERT scores can accurately predict massive transfusion in trauma patients without the use of time-consuming laboratory studies, although prospective studies need to be performed to validate these findings. Early identification of patients who will require MT may allow for timely mobilization of scarce resources and could benefit patients by making blood products available for treating hemorrhagic shock.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1097/SHK.0000000000001772 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!