Optimum timing of cerebral computed tomography angiography for determination of death using neurological criteria: a retrospective sensitivity cohort study

Summary Introduction Cerebral computed tomography angiography (CTA) is increasingly incorporated into national guidelines to confirm cerebral circulatory arrest during the determination of death using neurological criteria. However, implementation remains limited due to studies showing variable sensitivity. Clinical observations suggest the time elapsed from the onset of initial signs of brainstem areflexia, defined as the appearance of fixed and dilated pupils and loss of cough and gag reflexes, significantly influences CTA sensitivity. This study aimed to evaluate the impact of this time interval on test sensitivity. Methods We retrospectively assessed 118 intensive care unit patients undergoing determination of death using neurological criteria. Cerebral CTA was used to confirm cerebral circulatory arrest based on a four‐point scoring system. Patients were categorised according to the time elapsed from the onset of initial signs of brainstem areflexia to the investigation. Results Cerebral CTA did not confirm cerebral circulatory arrest in 15/118 (13%) patients, yielding an overall sensitivity of 87.3% (95%CI 79.6–92.5%). Notably, 93.3% (95%CI 66.0–99.7%) of false‐negative results occurred within 12 h of the onset of initial signs of brainstem areflexia. Patients were secondarily stratified into early (≤ 12 h) and late (> 12 h) groups, revealing a significantly higher sensitivity in the late group (67.4%, 95%CI 51.3–80.5% vs. 98.7%, 95%CI 91.8–100%, p < 0.001). Discussion The sensitivity of cerebral CTA increases with the time elapsed after the onset of initial signs of brainstem areflexia. We recommend delaying cerebral CTA until at least 12 h after the onset of initial signs of brainstem areflexia, and, in case of inconclusive results, test repetition at least 12 h later. Further work is required to confirm our findings and increase diagnostic confidence.