Accuracy of dual bolus single acquisition computed tomography in the diagnosis and grading of adult traumatic splenic parenchymal and vascular injury.

Abstract

Dual Bolus Single Acquisition Computed Tomography (DBSA-CT) has been advocated for use in trauma imaging through its ability to concurrently optimise abdominal arterial and visceral parenchymal enhancement while minimising scan times and radiation dose. However, concerns have been raised regarding excessive splenic parenchymal heterogeneity using this technique. The purpose of this study is to establish and compare the accuracy of DBSA-CT to traditional sequential multiphase scanning techniques in the diagnosis and grading of traumatic adult splenic injury. All patients diagnosed with a traumatic splenic parenchymal and/or vascular injury using DBSA-CT at a Level 1 trauma centre, who then proceeded to Digital Subtraction Angiography (DSA) or surgery were used for analysis over the time period June 2014 to March 2016. Two consultant radiologists each independently retrospectively reviewed all CT examinations documenting the degree of splenic heterogeneity, splenic injury grade using the American Association for the Surgery of Trauma (AAST) system and presence of vascular injury. Individual radiologist accuracy and inter-observer variability were evaluated. Splenic parenchymal enhancement was diagnostic in all cases with, at worst, moderate splenic heterogeneity. CT AAST parenchymal injury grading correlated poorly with operative findings. Average radiologist diagnostic sensitivity was 50.0% and 38.9% for active arterial haemorrhage and pseudoaneurysm, respectively, when correlated with DSA findings. Inter-observer agreement ranged from moderate to excellent in all cases. Dual Bolus Single Acquisition Computed Tomography may lead to reduced radiologist accuracy in the diagnosis of splenic injury, particularly pseudoaneurysm; however, such conclusions remain tentative given the small patient numbers examined.