RESUMEN
OBJECTIVES: To compare for the first time the performance of "GFAP and UCH-L1" vs. S100B in a cohort of patients managed for mild traumatic brain injury (mTBI) according to actualized French guidelines. METHODS: A prospective study was recently carried at the Emergency Department of Clermont-Ferrand University Hospital in France. Patients with mTBI presenting a medium risk of complications were enrolled. Blood S100B and "GFAP and UCHL-1" were sampled and measured according to French guidelines. S100B was measured in patients with samples within 3â¯h of trauma (Cobas®, Roche Diagnostics), while GFAP and UCHL-1 were measured in all patients (samples <3â¯h and 3-12â¯h) using another automated assay (i-STAT® Alinity, Abbott). RESULTS: For sampling <3â¯h, serum S100B correctly identifies intracranial lesions with a specificity of 25.7â¯% (95â¯% CI; 19.5-32.6â¯%), a sensitivity of 100â¯% (95â¯% CI; 66.4-100â¯%), and a negative predictive value of 100â¯% (95â¯% CI; 92.5-100â¯%). For sampling <12â¯h, plasma "GFAP and UCH-L1" levels correctly identify intracranial lesions with a specificity of 31.7â¯% (95â¯% CI; 25.7-38.2â¯%), a sensitivity of 100â¯% (95â¯% CI; 73.5-100â¯%), and a negative predictive value of 100â¯% (95â¯% CI; 95-100â¯%). Comparison of specificities (25.7 vs. 31.7â¯%) did not reveal a statistically significant difference (p=0.16). CONCLUSIONS: We highlight the usefulness of measuring plasma "GFAP and UCH-L1" levels to target mTBI patients (sampling within 12â¯h post-injury) and optimize the reduction of CT scans.