The most common cause of preventable death in US Military Service members is hemorrhagic shock, which is managed with tourniquets, blood administration, and damage control surgery. Administration of resuscitative blood products has been done by military providers since WWII and has since become common place in civilian hospitals. In the setting of limited resources while deployed, the use of a clinical decisionmaking calculator can help mitigate and expedite the decision of activating the massive transfusion protocol (MTP). The purpose of this research was to externally validate the ability of an app-based MTP algorithm developed by Mena et al. to predict the need for a massive transfusion (MT) in a military combat casualty dataset. Using casualty data from Operation Enduring Freedom (OEF), the Mena et al. MTP algorithm calculated the predicted probability of the need to receive a MT and was evaluated by calculating the area under the receiver operating characteristic curve (AUROC), specificity, and sensitivity. The AUROC for predicting MT was 0.84 (95% CI: 0.80-0.88) with specificity of 0.96 and sensitivity of 0.25 when assuming moderate and high categories (Pr > 0.070) is positive for MT. Using OEF casualty data, the MTP algorithm correctly categorized 96.4% of non-MT casualties ‘‘very low’’ or ‘‘low’’ and 3.0% moderate and 0.7% as high probability of MT. Alternatively,24.5% of MT casualties were categorized as ‘‘high’’ or ‘‘moderate,’’ with 31.6% and 43.9%categorized as “low” or “very low” probability of MT. The algorithm showed moderate ability in predicting the need for MT in combat casualties. With a specificity of 0.96, it is particularly useful in categorizing and ruling out the need for MT. With a sensitivity of 0.25, the algorithm lacks identification of those who will need MT, categorizing 76% in “very low” and “low” probabilities who ultimately required MT. The results suggest that especially with adjustments to mechanism of injury to aid in better prediction of the need for MTP.
Read full abstract