S100B serum protein cannot predict secondary intracranial haemorrhage after mild head injury in patients with low-dose acetylsalicylic acid prophylaxis.

INTRODUCTION

The goal of this study was to investigate if S100B serum protein could predict secondary intracranial haemorrhagic events (SIHEs) after mild head injury (mHI) in patients taking low-dose acetylsalicylic acid (LDA), making routinely repeated head computed tomography (RRHCT) scans unnecessary.

METHODS

Three hundred and eight-two patients with mHI, older than 60 years and taking LDA prophylaxis were enrolled. Primary head CT and RRHCT scans within 3 and 48 hours to trauma were performed. Additionally, S100B serum protein levels were evaluated at admission and predictive power for SIHEs was analysed.

RESULTS

Fifty-nine per cent were female and the mean age of all included patients was 81.8 ± 8.9 years. In four patients SIHEs were diagnosed. Sensitivity and the negative predictive value of S100B serum protein (cut-off value 0.10 µg l(-1)) were 75.0% and 98.6%, respectively. Specificity was 19.0% and the positive predictive value 1.0% (306 false positive values). In patients without bleeding, the median S100B value was 0.18 (IQR = 0.12-0.34) and in the ones with SIHEs, the median was 0.11 (IQR = 0.10-1.16) (p > 0.05). The discriminatory power of S100B in the ROC analysis was 0.399 (95% CI = 0.079-0.720; p > 0.05).

CONCLUSION

S100B cannot be considered as an effective diagnostic tool in the prediction or exclusion of SIHE in older patients with mHIs taking LDA prophylaxis.