RESUMO
BACKGROUND: S100B is a serum protein known to elevate in patients with brain injury, but it is unknown whether it can predict intracranial pathology in intoxicated patients following mild traumatic brain injury (MTBI). We performed a systematic review and meta-analysis of the English language literature to address this question. MAIN OUTCOMES AND RESULTS: Four prospective cohort trials of serum S100B levels on acutely intoxicated patients with MTBI were included in this meta-analysis. Prevalence of intracranial pathology in the pooled cohort of the intoxicated MTBI patients was 10%, lower than the 15-30% reported in the literature for the general MTBI population. Standard mean difference of serum S100B levels between patients with and without intracranial pathology on CT was 0.73 µg/L (Z = 18.33, P < 0.001). Following sensitivity analysis and hierarchical summary receiver-operating characteristic models, three remaining articles were used for pooled estimates that found that S100B had a sensitivity of 0.96 (95% CI: 0.84-1.00, I2 = 0%) and specificity of 0.63 (95% CI: 0.58-0.68, I2 = 86.8%) with a high negative predictive value (100%, 95% CI: 95.14-100, I2 = 0%) and a negative LR of 0.06 (95% CI: 0.01-0.31). CONCLUSIONS: Serum S100B levels may have utility in ruling out intracranial pathology in intoxicated patients, however more study and comparison with other serum biomarkers of brain injury are necessary before this becomes the accepted standard of care.
