ABC Transfusion Model: Easy as 1, 2, 3?

NAPLES, FLA. – The Assessment of Blood Consumption score was predictive of the need for massive transfusion in 89% of major trauma patients in a comparative analysis of three popular predictive models.

Of the 373 patients admitted to a rural level I trauma center from January 2008 to December 2009, 38 (10%) received massive transfusion or at least 10 units of packed red blood cells (PRBC) within 24 hours of admission.

Of these 38 patients, 34 were predicted by the ABC score, 6 by the McLaughlin score, and only 1 by the TASH (Trauma-Associated Severe Hemorrhage) score, Dr. Nicole Krumrei said at the annual meeting of the Eastern Association for the Surgery of Trauma. This resulted in sensitivities of 89%, 15.8%, and 2.6%, respectively.

The C index was 0.86 for the ABC score, 0.56 for the McLaughlin score, and 0.51 for the TASH; specificity was 85%, 98%, and 99.7%, respectively.

Although the McLaughlin and TASH scores were highly specific, they were not predictive in the rural trauma population, and therefore they place patients at a high risk for undertriage, said Dr. Krumrei, a resident with the Mayo Clinic in Rochester, Minn. In addition, the TASH and McLaughlin scores require time-consuming blood tests and cumbersome equations.

The advantages of the ABC score are that "it uses four nonweighted parameters, it’s available within minutes of arrival to the trauma bay or possibly prearrival, [and] it’s straightforward to memorize and easy to calculate," she said.

The ABC model assigns 1 point for each "yes" to the following four elements: heart rate greater than 120 beats per minute (bpm), systolic blood pressure less than 90 mm Hg, positive results on focused assessment with sonography for trauma (FAST), and penetrating mechanism of injury. A score of 2 or more is predictive of the need for transfusion.

Dr. Krumrei pointed out that three of the four elements in the ABC score are available en route, which allows for the identification of patients requiring massive transfusion prior to arrival in the trauma bay. Of the 373 patients, 183 (49%) were interfacility transfers and 48 (13%) had penetrating injuries.

The undertriage rate with the ABC model was 11% and the overtriage rate 13% – both in the acceptable range, she said.

Mortality was significantly higher (18.4%) in patients requiring massive transfusion, compared with 5.4% in those who did not require transfusion. Massive transfusion patients were also significantly more likely to have a higher median Injury Severity Score (34 vs. 13), lower median systolic blood pressure (73.5 vs. 107 mm Hg), higher median heart rate (126 vs. 103 bpm), and lower median Glasgow Coma Scale score (5 vs. 15).

"The ABC model should be widely implemented into urban and rural trauma centers," Dr. Krumrei concluded.

All three predictive models were developed and validated in urban or military settings using studies that required only that patients receive at least one unit of PRBC during their hospital stay. The validation studies also excluded interfacility transfers, which make up a majority of trauma cases, she said. Hemorrhage is the leading cause of preventable death in trauma patients, with up to 15% of trauma patients requiring massive transfusion.

Dr. Krumrei said one of the strengths of the analysis was its diverse patient population, but acknowledged that it was limited by retrospective design, incomplete data points, and the fact that the TASH and ABC scores utilize the FAST exam, which is user dependent.

Invited discussant Dr. Jay S. Jenoff of Jefferson University Hospital in Philadelphia asked whether the results may have been affected by the use of beta-blockade and the lack of certain data necessary to calculate the TASH and McLaughlin scores.

Dr. Krumrei said that it’s unlikely the outcomes were greatly affected because beta-blockade was used in only four mass transfusion patients, and that although 46% of all patients were missing pH data, the information was available for 82% of mass transfusion patients.

When asked whether her institution uses the ABC model to counsel referring centers, Dr. Krumrei said that they do, and that they’re also working on implementing it not only within the institution, but also in its catchment area.

Dr. Krumrei, her coinvestigators, and Dr. Jenoff disclosed no conflicts of interest.