Serum neuron-specific enolase as a predictor of intracranial lesions in children with head trauma: a pilot study.

OBJECTIVE To determine the reliability of serum neuron-specific enolase (NSE) levels in predicting intracranial lesions (ICL) in children with blunt head trauma (HT). METHODS A prospective pilot study was conducted of patients 0 to 18 years of age presenting to a children's hospital emergency department (ED) between December 1997 and October 1998. Children presenting within 24 hours of injury who required head computed tomography (CT) were eligible. Blood samples were obtained to measure serum NSE level. Data collected included patient demographics, historical information, Glasgow Coma Scale score (GCS), physical examination, head CT results, and outcome. Patients were assigned to one of two groups based on the head CT results (PICL; presence of intracranial lesion, or NICL; no intracranial lesion). Data were analyzed using Student's t-test and chi-square. The 95% confidence interval (95% CI) was calculated when appropriate. A receiver operating characteristic curve was constructed to determine the NSE level that yielded the highest sensitivity and specificity for predicting ICL. RESULTS Fifty patients were enrolled; 22 (45%) had abnormal head CT. No difference in demographics or mechanism of injury was observed between those with abnormal or normal CT scans. The mean GCS level was 11.9 +/- 4.2 for PICL and 13.9 +/- 2.6 for NICL (p = 0.045; 95% CI = -0.05 to -3.9). The mean NSE level was 26.7 +/- 21.4 for PICL and 17.7 +/- 7.8 for NICL (p = 0.048; 95% CI = 0.1 to 17.9). An NSE level > or = 15.3 ng/mL yielded a sensitivity of 77%, a specificity of 52%, and a negative predictive value of 74%. CONCLUSIONS These results suggest that serum NSE may be a useful screening tool for predicting ICL in children with blunt head trauma. However, the NSE alone was neither sensitive nor specific in predicting all patients with ICL.