• Aslı Gülfer Kartal, Serkan Yılmaz, Elif Yaka, Murat Pekdemir, Hasan Tahsin Sarısoy, Mustafa Baki Çekmen, and Melih Yüksel.
• The Department of Emergency Medicine, Faculty of Medicine, Kocaeli University, Kocaeli, Turkey.
• Acad Emerg Med. 2014 Jul 1; 21 (7): 736-41.

ObjectivesVertigo is a common presenting complaint resulting from central or peripheral etiologies. Because central causes may be life-threatening, ascertaining the nature of the vertigo is crucial in the emergency department (ED). With a broad range of potential etiologies, distinguishing central causes from benign peripheral causes is a diagnostic challenge. Cranial magnetic resonance imaging (MRI) is the recommended neuroimaging method when clinical findings are ambiguous. However, MRI scanning for every patient with an uncertain diagnosis may not be efficient or possible. Therefore, to improve ED resource utilization for patients with vertigo, there is a need to identify the subset most likely to have MRI abnormalities. It has previously been shown that S100B protein provides a useful serum marker of stroke, subarachnoid hemorrhage, and traumatic brain injury. This study evaluated whether S100B levels could predict central causes of vertigo as identified by cranial MRI in the ED.MethodsThis prospective, observational study was conducted with adult patients with acute-onset vertigo (within 6 hours) in the ED of a teaching hospital in Kocaeli, Turkey. Patients with nausea or dizziness complaints without previously known vertigo or cranial pathology, and who agreed to participate in the study, were included. Patients with trauma or with neurologic findings that developed concurrent with their symptoms were excluded. Serum levels of S100B were measured with an electrochemiluminescence immunoassay kit. All subjects underwent cranial MRI. The predictors of positive MRI results were evaluated using logistic regression analysis. Sensitivity and specificity of S100B​ levels for identifying subjects with central causes of vertigo on MRI were calculated with receiver operating characteristic (ROC) curve.ResultsOf the 82 subjects included in the study, 48 (58.5%) were female, and the mean (±SD) age was 51 (±16) years. Thirty-one (37.8%) subjects had positive MRI results. Median (with interquartile range [IQR]) serum S100B levels were significantly different between MRI-negative and MRI-positive groups (median = 27.00 pg/mL, IQR = 10.00 to 44.60 vs. median = 60.94 pg/mL, IQR = 38.25 to 77.95, respectively; p = 0.04). In logistic regression analysis, subjective "he or she is spinning" (p = 0.030, odds ratio [OR] = 1.63, 95% confidence interval [CI] = 1.38 to 2.49), systolic blood pressure (sBP; p = 0.045, OR = 1.044, 95% CI = 1.021 to 1.080), and serum S100B level (p = 0.042, OR = 1.22, 95% CI = 1.018 to 1.445) were found to be independent predictors of MRI abnormalities. In the ROC analysis, S100B > 30 pg/mL predicted the clinical outcome with 83.9% sensitivity (95% CI = 66.3% to 94.5%) and 51.0% specificity (95% CI = 36.6% to 65.2%). The area under the ROC curve was 0.774 (95% CI = 0.666 to 0.881).ConclusionsTo the best of our knowledge this is the first study assessing the utility of serum S100B levels for diagnosis of acute-onset vertigo. Serum S100B levels are associated with the presence of central causes of vertigo on cranial MRI. However, serum S100B levels are not sufficiently sensitive to exclude candidates from cranial MRI.© 2014 by the Society for Academic Emergency Medicine.

39 keywords...

hide…