RESUMO
BACKGROUND: S100B protein measurement in blood is proposed to exclude the presence of computed tomography (CT) lesions after minor head injury (MHI). We aimed to validate S100B as an accurate and valuable screening tool for MHI diagnosis in a large multicenter study, as well as: 1) to evaluate whether a second S100B blood level determination 3 h after the first one would be informative; 2) to compare the bioclinical performances of the two commercially available automated methods of measurement of S100B for the screening of patients. METHODS: Four thousand and thirty MHI subjects were enrolled in a prospective observational multicenter study; results for serum S100B measurement determined within 3 h after the clinical event (H0) then at H3 were compared to that of cranial CT scans performed with 6 h following the presentation to emergency department. Both the Diasorin and the Roche Diagnostics assays were systematically performed. RESULTS: Cerebral lesions on CT scan were identified with sensitivity and negative-predictive value (NPV) of 96.3% and 99.4% (Diasorin, 1 dissonant case), and of 100% and 100% (Roche Diagnostics, no dissonant case). Sensitivity and NPV at H3 appeared lower than those at H0, due to the rapid decrease in S100B levels. CONCLUSIONS: Serum S100B level on admission of patients with MHI is an accurate and useful screening tool to exclude intracranial lesions. Performing a second late S100B level determination is not informative. The two automated immunoassays appear usable in a similar manner, although the two methods are not interchangeable.