• Stephen M Cohn, Jorge I Arango, John G Myers, Peter P Lopez, Rachelle B Jonas, Lindsay L Waite, Michael G Corneille, Ronald M Stewart, and Daniel L Dent.
• Department of Surgery, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, Texas 78229, USA. cohn@uthscsa.edu
• Am Surg. 2009 Feb 1;75(2):133-9.

AbstractComputed tomography (CT) grading systems are often used clinically to forecast the need for interventions after abdominal trauma with solid organ injuries. We compared spleen and liver CT grading methods to determine their utility in predicting the need for operative intervention or angiographic embolization. Abdominal CT scans of 300 patients with spleen injuries, liver injuries, or both were evaluated by five trauma faculty members blinded to clinical outcomes. Studies were graded by American Association for the Surgery of Trauma criteria, a novel splenic injury CT grading system, and a novel liver injury grading system. The sensitivity and specificity of each methodology in predicting the need for intervention were calculated. The kappa statistic was used to determine interrater variability. Twenty-one per cent (39/189) of patients with splenic injuries visible on CT scans required interventions, whereas 14 per cent (21/154) of patients with liver injuries visible on CT required interventions. The overall sensitivity of all grading systems in predicting the need for surgery or angioembolization of the spleen or liver was poor; the specificity seemed to be fairly good. When evaluators were compared, the strength of agreement for the various scoring systems was only moderate. Anatomic CT grading systems are ineffective screening tools for excluding the need for operation or embolization after splenic or hepatic trauma. Although insensitive, CT is a good predictor (highly specific) of the need for intervention if certain definitive abnormalities are identified. Considerable inconsistency exists in interpretation of abdominal CT scans after trauma, even among experienced clinicians.

20 keywords...

hide…