Clinical examination of labyrinthine-defective patients out of the vertigo attack: sensitivity and specificity of three low-cost methods

Many reports have appeared in the medical literature concerning the clinical examination at the bedside of patients with vertigo and, even if few controversial opinions exist, the observation of one or more kinds of nystagmus is generally regarded as suggesting an organic aetiology. So far, the presence of nystagmus has been generally considered to be crucially important for clinicians who are daily asked to differentiate between an "organic" cause of vertigo (for example, a labyrinthine dysfunction) and a "non-organic" cause of vertigo, such as a panic disorder. Albeit, it should not be forgotten that the central nervous system is able to resolve the asymmetry of vestibulo-ocular reflexes, due to a peripheral vestibular failure, by means of compensatory mechanisms so that nystagmus is rapidly abolished after the acute attack of vertigo. In addition, visual fixation elicits sub-cortical inhibitory pathways to the vestibular nuclei so that spontaneous nystagmus is remarkably reduced by light. In order to more easily detect nystagmus, attempts have been made to minimize the interference of visual fixation by means of positive lenses (Frenzel's glasses) and light occluding masks with infrared cameras (videonystagmoscopy) which have in part replaced direct observation of the patient's eyes, albeit no systematic validation of the advantages has been reported yet. To investigate the usefulness of these 3 low-cost methods to detect nystagmus, 528 outpatients presenting peripheral vestibular hypofunction, diagnosed by a complete audiological and vestibular examination, including caloric tests, were enrolled in the present study, while 133 subjects with normal vestibular function acted as a control group. All patients and control subjects underwent a standardized clinical examination based on search for spontaneous, positioning and head-shaking nystagmus detected by direct observation of patient's eyes, Frenzel's glasses and videonystagmoscopy. Specificity of the three techniques were 35.6, 43.7 and 91.6, whilst sensitivity was 88.7, 88.7 and 84.2, respectively. Finally, discriminant analysis based on the presence/absence of at least one kind of nystagmus was computed for each technique and showed that videonystagmoscopy allowed the examiner to correctly classify both pathological and normal subjects more frequently (> 77% of cases) than the other two methods (about 50%). It is concluded that only videonystagmoscopy is an acceptable technique for screening a labyrinth defect in a population of outpatients with vertigo.