[Diagnosis of common causes of vertigo using a structured clinical history]

The structured clinical history is the most sensitive test for diagnosing vertigo. Its diagnostic effectiveness on the first visit was analyzed and key signs and symptoms with high predictive value for common causes of vertigo were identified. One hundred outpatients who complained of dizziness or loss of balance were evaluated using a structured clinical interview. Each questionnaire was examined independently by three blinded investigators, who assigned a diagnosis and identified the elements of the history that figured most prominently in the diagnosis. The gold standard was defined as independent selection of the same diagnostic category by all three investigators. A first-visit diagnosis was obtained in 40% of patients (95% confidence interval 30-50%): 38% women and 42% men. Causes included benign positional paroxysmal vertigo (BPPV, 13 patients), headache-associated vertigo (9), Meniere disease (7), cervical vertigo (3), psychiatric dizziness (2), post-traumatic vertigo (2), vertebro-basilar transient ischemic attack (1), vestibular neuritis (1), convulsive seizure (1), and presyncope (1). The best predictors of BPPV were the precipitating mechanism (specificity [SP] 100%), positional nystagmus (sensitivity [SE] 90%, SP 63%), and the Dix-Hallpike test (SE 82%, SP 71%). Elements predictive of headache-associated vertigo were duration of the attack (minutes) and a personal history of headache (both, SP 100%). Other predictors were facial hypoesthesia (SE 92%, SP 47%) and associated neurological disease (SE 82%, SP 58%).