PURPOSE

Computed tomography (CT) angiography emerges as a viable alternative technique for confirmation of brain death. However, evaluation criteria are not well established for demonstration of cerebral circulatory arrest. This retrospective study aimed to evaluate CT angiography scoring systems in diagnosis of brain death, review the literature, and compare interobserver agreement between different scales for the diagnosis of brain death.

METHODS

CT angiography examinations of 25 patients with a clinical diagnosis of brain death were reevaluated according to 10-, 7-, and 4-point scales. Exams were performed with a 64-slice CT scanner including unenhanced, arterial (20 s) and venous phase (60 s) scans. Subtraction images of both phases were obtained. Interobserver agreement was evaluated for the assessment of vessel opacification and diagnosis of brain death.

RESULTS

According to 10-, 7-, and 4-point scales; 13, 16, and 22 of 25 patients had full score, respectively. Using the clinical exam as the reference standard, sensitivities obtained for 10-, 7-, and 4-point scales were 52%, 64%, and 88%, respectively. Percent agreement between readers was 100% for 10- and 7-point scales and 88% for 4-point scale. Percent agreement for opacification of scale vessels was equally high for all three scales (93.6%, 93.7%, 91% for 10-, 7-, and 4-point scales, respectively).

CONCLUSION

The 4-point scale appears to be more sensitive than the 10- and 7-point scales in CT angiography evaluation for brain death. Interobserver agreement is high for all three scales when subtraction images are used.