The purpose of this study was to define whether the semiquantitative analysis of hemoperitoneum increases the accuracy of early prediction of massive transfusion (MT).

A retrospective review of severe trauma patients consecutively admitted to our trauma intensive care unit between January 2005 and December 2009 was conducted. Patients diagnosed with blunt abdominal trauma who had a computed tomography scan on admission were included. The hemoperitoneum size was defined using the Federle score on computed tomography as large, moderate, or minimal/none. The association between MT (≥10 U of packed red blood cells in the first 24 h) and moderate and large sizes of hemoperitoneum was assessed using a multiple logistic model.

Of the 381 patients meeting the inclusion criteria, 270 (71%) were male; the mean age was 35.5 ± 18.2 years and mean injury severity score was 23.4 ± 17. Ninety-seven (26%) had large hemoperitoneum, 107 (28%) had moderate hemoperitoneum, and 177 (46%) had minimal/no hemoperitoneum. Eighty-three patients (22%) required MT. The positive predictive value for MT of a large hemoperitoneum was 41%, 23% for a moderate hemoperitoneum, and 10% for minimal/no hemoperitoneum (P < .001). The corresponding values for hypotensive patients were 61%, 32%, and 25%, respectively (P < .001). In the multivariate analysis model, only the large size of hemoperitoneum was significantly associated with MT (OR 6.4, 95% CI 2.9−14, P < .001, r² = 0.47).

The assessment of the size of hemoperitoneum on admission substantially improves the prediction of MT in trauma patients and should be used to trigger and guide initial haemostatic resuscitation.