Article
Diagnostic accuracy of the initial CT in the prediction of vascular pathology in trauma-associated subarachnoid hemorrhage
Search Medline for
Authors
Published: | May 21, 2013 |
---|
Outline
Text
Objective: To determine the diagnostic accuracy of the initial CT in patients that suffered from traumatic brain injury (TBI) regarding the differentiation of spontaneous subarachnoid hemorrhage (sSAH, as the cause of trauma) and traumatic SAH (tSAH, as a result of trauma).
Method: All patients with TBI and SAH in the initial CT seen by the neurosurgical service of this university hospital during one year were retrospectively analyzed. Patients were subdivided into two categories (Type I, tSAH, and Type II, sSAH) by a medical student, a neuroradiological consultant and a neurosurgical consultant based on predefined criteria applied to the initial CT. Criteria included the pattern of blood distribution, the presence of a midline shift and the correlation of SAH and trauma-associated impact zone. All patients underwent angiography as part of the initial emergency room protocol or after reevaluation of primary imaging (DSA, CTA, MRA). Sensitivity, specificity, positive and negative predictive value (PPV, NPV) and diagnostic accuracy were determined by reviewers blinded to angiographic results. STARD criteria for diagnostic studies were applied.
Results: Of n = 307 patients with TBI seen in 12 months, n = 154 (50%) harbored a trauma-associated SAH. In n = 44, angiography was not available for review. N = 110 patients with angiography (high resolution contrast CT n = 5, CTA n = 91, MRA n = 8 and DSA n = 14) were categorized into ether Type I (n = 78, 71%) or Type II (n = 32, 29%). No patient classified as Type I harbored a vascular pathology. In patients classified as Type II three intracranial aneurysms were detected as the probable cause of trauma-associated SAH. This results in a sensitivity of 100%, a specificity of 73%, PPV 9%, NPV 100%. Overall diagnostic accuracy was 24%.
Conclusions: The standardized interpretation of the initial CT is adequate as a screening test in trauma patients to exclude the presence of an underlying ruptured aneurysm as the bleeding source (high sensitivity, high NPV). The diagnostic accuracy of such an approach would be low.