Predicting neuroradiologic outcome in patients referred for audiovestibular dysfunction.

PURPOSE To relate clinical presentation and results of audiovestibular testing to neuroradiologic outcome in patients with audiovestibular dysfunction. METHODS We retrospectively reviewed the neuroimaging studies, results of audiometric and vestibular testing, and medical records of 118 patients referred for imaging over a 2-year period for evaluation of sensorineural hearing loss, dizziness, and/or vertigo, and to rule out acoustic neuroma. Patients' presentation and results of audiometric and vestibular testing were associated with either a positive or negative neuroimaging outcome. Discriminant analysis was performed to identify variables related significantly to imaging results. Two-way cross-tabulation of these significant variables was performed to assess their sensitivity and specificity in predicting imaging outcome. RESULTS Fifteen (13%) of 118 patients had neuroimaging findings related to presenting symptoms. Discriminant analysis identified vertigo, dizziness, and dysequilibrium as corresponding to negative radiologic outcome. Nonvestibulocochlear cranial nerve involvement correlated significantly with positive neuroimaging results. Of all audiovestibular testing, only vestibular testing results correlated significantly with neuroimaging outcome. In conjunction with the results of vestibular testing the symptoms and signs identified above yielded a sensitivity of 57% and specificity of 93% in predicting neuroradiologic results. In the absence of vestibular testing, sensitivity and specificity were 29% and 98%, respectively. CONCLUSIONS Clinical presentation and audiovestibular testing could not sensitively predict the outcome of neuroimaging in our cohort of patients referred for audiovestibular dysfunction.