Design: Prospective cohort study.

Setting: Federal teaching hospital.

Patients: 100 dizzy patients and 25 control subjects.

Measurements: The primary outcome was dizziness status (improved or not improved); the secondary outcomes were morbidity and health care utilization.

Results: The dizziness resolved for 18 patients, whereas the status improved for 37, stayed the same for 32, and worsened for 11, with two patients lost to follow-up. Thus, 55% of patients whose dizziness had not resolved two weeks after their initial visits improved over the subsequent 12 months. Logistic regression revealed four independent predictors of persistent dizziness at one-year follow-up: dizziness due to psychiatric causes, dysequilibrium, vertigo other than benign positional vertigo, vestibular neuronitis, or migraine (odds ratio, 6.3; 95% CI, 2.1–18.6); daily dizziness (odds ratio, 6.4; 95% CI, 2.0–21.0); dizziness worse with walking (odds ratio, 3.0; 95% CI, 1.1–9.0); and patient had initially feared a serious illness (odds ratio, 0.25; 95% CI, 0.10–0.74). These four factors could be used to classify patients as having either a high (82%), medium (47%), or low (0%) likelihood of improvement at one-year follow-up. One patient died from heart failure, and none developed a serious disease for which dizziness had been a harbinger. Dizziness was not associated with an increased number of clinic visits.

Conclusions: Among patients with a chief complaint of dizziness who are still symptomatic at two-week follow-up, more than half improve within a year. Clinical factors identify patients at higher risk for persistent dizziness.