Clinical decision rules in predicting computed tomography scan findings and need for neurosurgical intervention in...

In this study, we will compare the diagnostic values of head CT decision rules in predicting the findings of CT scans in a prospective multicenter study in university emergency departments in Iran.

The primary outcome was any traumatic lesion findings in brain CT scans, and the secondary outcomes were death, the need for mechanical ventilation, and neurosurgical intervention. Decision rules including the Canadian CT Head Rule (CCHR), New Orleans Criteria (NOC), National Institute for Health and Clinical Excellence (NICE), National Emergency X-Radiography Utilization Study (NEXUS), and Neurotraumatology Committee of the World Federation of Neurosurgical Societies (NCWFNS) were compared for the main outcomes.

In total, 434 mild TBI patients were enrolled in the study. The NCWFNS had the highest sensitivity (91.14%) and the lowest specificity (39.42%) for predicting abnormal finding in CT scan compared to other models. While the NICE obtained the lowest sensitivity (79.75%), it was associated with the highest specificity (66.67%). All model performances were improved when administered to predict neurosurgical intervention among patients with GCS 13-15. NEXUS (AUC 0.862, 95% CI 0.799-0.924) and NCWFNS (AUC 0.813, 95% CI 0.723-0.903) had the best performance among all evaluated models.

The NCWFNS and the NEXUS decision rules performed better than the CCHR and NICE guidelines for predicting any lesion in the CT imaging and neurosurgical intervention among patients with mTBI with GCS 13-15. For a subset of mTBI patients with GCS 15, the NOC criteria have higher sensitivity for abnormal CT imaging, but lower specificity and more requested CTs.