Early predictors for massive transfusion in older adult severe trauma patients.

BACKGROUND

Many scoring systems for the early prediction of the need for massive transfusion (MT) have been reported; in most of these, vital signs are regarded as important. However, the validity of these scoring systems in older patients remains unclear because older trauma patients often present with normal vital signs. In this study, we investigated the effectiveness of previously described scoring systems, as well as risk factors that can provide early prediction of the need for MT in older severe trauma patients.

METHODS

We prospectively collected data from a cohort of severe trauma patients (ISS ≥16 and age ≥16years) admitted from January 2007 to March 2015. Trauma Associated Severe Hemorrhage (TASH), Assessment of Blood Consumption (ABC), and Prince of Wales Hospital (PWH) scores were compared between a younger and an older group. Furthermore, the predictors associated with MT in older severe trauma patients were assessed using multivariable logistic regression analyses.

RESULTS

The area under the curve (AUC) was significantly smaller for older group than for younger group for all three scoring systems (p<0.05). The most important risk factors to predict the need for MT were related to anatomical factors including FAST results (odds ratio (OR): 5.58, 95% confidence interval (CI): 2.10-14.99), unstable pelvic fracture (OR: 21.56, 95% CI: 6.05-90.78), and long bone open fracture of the lower limbs (OR: 12.21, 95% CI: 4.04-39.09), along with pre-injury anticoagulant agent use (OR: 5.22, 95% CI: 1.30-19.61), antiplatelet agent use (OR: 3.81, 95% CI: 1.57-9.04), lactate levels (OR: 1.20, 95% CI: 1.04-1.39) and shock index (OR: 2.67, 95% CI: 1.05-6.84). Traditional vital signs were not early risk factors.

CONCLUSION

We suggest that MT in older trauma patients should be considered on the basis of anatomical factors, pre-injury anticoagulant or antiplatelet agent use, lactate level and SI even if traditional vital signs are normal.