Clinical value of computerized tomography scanning in severe head injury

Serial computerized tomography (SCT) was performed on 138 patients suffering from severe head injuries (8 or less on the Glasgow Coma Scale). Standard practice called for scans to be done upon admission (within hours of the injury), and after 1, 3, 7 days and 1 month. Subsequent CT's depended on the patient's condition. Clinical results at the time of discharge were graded according to the Glasgow Outcome Scale. Patients who died, were in a persistent vegetative state, or were severely disabled were considered to have a ''bad outcome''. On the other hand, patients who were somewhat disabled or made good recoveries were considered to have a ''good outcome''. During the serial CT scan, there were new findings (not visualized on the initial CT but appearing on subsequent ones) in 91 of the 138 patients. These new findings were classified as follows; 1) decreased density collection in the subdural space (DDC), 2) ventricular dilation (VD), 3) intracerebral hematoma (ICH), 4) intraventricular hemorrhage (IVH), 5) extracerebral hematoma (ECH), 6) edema (E), 7) infarction (I). We defined ICH, IVH, ECH, E and I as new lesions. Of the 60 patients with new lesions 12 had good outcomes and 48 had poor outcomes. There were 78 patients who did not have any new lesions, 60 with good outcomes and 18 with poor outcomes. A significant correlation was found between good outcomes and the absence of new lesions, and between bad outcomes and the development of new lesions (p < 0.01; x2 = 44.038). We conclude that SCT can help predict the outcome with severe head injury patients and may be very important in their examination and care. (J.P.N.)