[Massive transfusion predictive scores in trauma. Experience of a transfusion registry]

Abstract

Objectives: Our purpose is to validate previously described massive transfusion (MT) scoring in our Transfusion Trauma Registry.

Design: A retrospective cohort of adult trauma patients.

Setting: Trauma and Emergency Intensive Care Unit of a tertiary hospital.

Patients: Patients with severe trauma (injury severity score>15) admitted from October 2006 to July 2009.

Interventions: None.

Variables: The following MT scoring and cutoff points (CP) were evaluated: Trauma-Associated Severe Hemorrhage (TASH) CP: ≥16 and ≥18; Assessment Blood Consumption (ABC) CP: ≥2 and Emergency Transfusion Score (ETS) CP: ≥3, ≥4, ≥6. MT was defined as the transfusion of 10 units or more of packed red blood cells in the first 24 hours. We studied the sensivity (S), specifity (SP), and positive and negative predictive values (PPV, NPV), the positive and negative likehood ratios (LHR +, LHR-) and area under the receiver operating characteristic curve (ROC).

Results: A total of 568 patients were available for analysis; 77.6% were men, with a mean age of 41.16 ± 18 years and an ISS of 30 ± 13. 93.8% with blunt trauma. The overall MT rate was 18.8%. The best S was obtained with ETS ≥3 and best SP was obtained with TASH ≥18. ROC for different scores was: ABC: 0.779, ETS: 0. 784, TASH: 0.889.

Conclusion: These scales can be useful for characterizing the TM population, for excluding low-risk populations, and for attempting to be objective in hematological damage control and in supporting clinical decisions, based on fe1w and easily obtainable data.