Reference values for serial vestibular testing.

This study defines the normal limits of day-to-day variation in vestibular function. A short-acting caloric test was used and nystagmus intensity was measured by means of an interactive computerized technique. Data were analyzed from 30 normal subjects tested on three successive days and from ten of these subjects who were also tested on 11 additional occasions. Individual caloric test scores generally remained between 55% and 170% of their initial value; narrower ranges resulted when scores were averaged together. There was very little habituation of caloric responses. The upper normal limits of test-retest variation in scores for right-left difference and directional preponderance of caloric responses were 24% and 22%, respectively.