Evaluating dizziness.

PURPOSE To conduct a structured literature synthesis on the etiology, prognosis, and diagnostic evaluation of dizziness, and to suggest a primary-care approach to evaluating this symptom. METHODS Studies were identified from MEDLINE searches (1966 through 1996) and a manual search of bibliographies from retrieved articles. Two investigators independently abstracted study data. RESULTS The most common etiologies for dizziness were peripheral vestibulopathies (35% to 55% of patients) and psychiatric disorders (10% to 25% of patients). Cerebrovascular disease (5%) and brain tumors (<1%) were infrequent. The history and physical examination led to a diagnosis in about 75% of patients. At least 10% of patients eluded diagnosis. Symptoms were usually self-limited and not associated with an increased risk of mortality. The diagnostic testing literature, which was often methodologically flawed, suggested that routine laboratory tests as well as cardiovascular and neurologic testing had a low yield in unselected patients. We could not derive evidence-based guidelines for using specialized vestibular function tests such as electronystagmography. CONCLUSIONS Dizziness is usually a benign, self-limited complaint. When a diagnosis can be made, a careful history and physical examination will usually identify the probable cause. Cardiovascular, neurologic, and laboratory testing should be guided by the clinical evaluation. Rigorous studies are needed to determine the accuracy and utility of specialized vestibular testing.