Preliminary Evaluation of the Scandinavian Guidelines for Initial Management of Minimal, Mild, and Moderate Head Injuries with Glial Fibrillary Acidic Protein

Abstract

<p>Glial fibrillary acidic protein (GFAP) has become the most promising biomarker for detecting traumatic abnormalities on head computed tomography (CT) in patients with traumatic brain injury (TBI), but most studies have not addressed the potential added value of combining the biomarker with clinical variables that confer risk for intracranial injuries. The Scandinavian Guidelines for Initial Management of Minimal, Mild, and Moderate Head Injuries in Adults were the first clinical decision rules in the field with an incorporated biomarker, the S100 astroglial calcium-binding protein B (S100B), which is used in the Mild (Low Risk) group defined by the guidelines. Our aim was to evaluate the performance of the guidelines when S100B was substituted with GFAP. The sample (<em>N</em> = 296) was recruited from the Tampere University Hospital's emergency department between November 2015 and November 2016, and there were 49 patients with available GFAP results who were stratified in the Mild (Low Risk) group (thus patients undergoing biomarker triaging). A previously reported cutoff of plasma GFAP ≥140 pg/mL was used. Within the Mild (Low Risk) group (<em>n</em> = 49), GFAP sensitivity (with 95% confidence intervals in parentheses) for detecting traumatic CT abnormalities was 1.0 (0.40–1.00), specificity 0.34 (0.19–0.53), the negative predictive value (NPV) 1.0 (0.68–1.00), and the positive predictive value (PPV) 0.16 (0.05–0.37). The sensitivity and specificity of the modified guidelines with GFAP, when applied to all imaged patients (<em>n</em> = 197) in the whole sample, were 0.94 (0.77–0.99) and 0.20 (0.15–0.28), respectively. NPV was 0.94 (0.80–0.99) and PPV 0.18 (0.13–0.25). In the Mild (Low Risk) group, none of the patients with GFAP results below 140 pg/mL had traumatic abnormalities on their head CT. These findings were derived from a small patient subgroup. Future researchers should replicate these findings in larger samples and assess whether GFAP has added or comparable value to S100B in acute TBI management.<br></p>