Meredith E. Adams, Schelomo Marmor, Bevan Yueh, Robert L. Kane

Objective There is a lack of consensus regarding the indications for vestibular testing in the evaluation of dizziness and balance disorders. Geographic variation in health services utilization is associated with lack of consensus. To understand the variation in current practice, we investigated the patterns of use of vestibular testing and diagnosis codes for dizziness and balance disorders among individuals ≥65 years of age across different regions of the United States. Study Design Cross-sectional study. Setting Medicare administrative claims data. Subjects and Methods Using the Summarized Denominator file, a sample of the US population linked to the Surveillance, Epidemiology, and End Results (SEER)-Medicare files (years 2000-2010), we identified persons who were ≥65 years of age. We used multivariable analyses to determine the factors associated with vestibular testing and diagnoses. Results Of the 231,984 eligible Medicare beneficiaries, 27% were diagnosed with dizziness and balance disorders. Patterns of use of vestibular tests (eye movement recording for spontaneous nystagmus, caloric testing, and rotary chair testing) varied significantly by geographic region. Rotary chair test utilization varied most. We found significant geographic variation in vestibular testing and diagnoses after controlling for age, sex, race, Medicaid participation, and rurality. Conclusions There may be opportunities to improve the consistency and efficiency of care for dizziness and balance disorders. It will be important to define appropriate levels of vestibular diagnostic testing and which tests add sufficient value to justify the costs. Further work is needed to better characterize the causes and consequences of variation in vestibular test utilization.