Timothy C. Hain, MD
The caloric test is a part of the ENG. It reflects an attempt to discover the degree to which the vestibular system is responsive and also how symmetric the responses are, between left and right. It is a test of the lateral semicircular canals alone -- it does not assess vertical canal function or otolithic function. While not as good as we would like, the caloric test is the best that we have. Most caloric tests nowadays are done using a computerized system as shown below. The computer analyzes the caloric data, computing peak slow-phase velocity.
|Contemporary ENG system (Courtesy of ICS medical)||Balloon version of caloric (one of the variant caloric tests, see below)|
There are several variant caloric tests, but at present, the standard is the bithermal caloric test. This is based on simple common sense. All of these variants have the advantage of greater convenience, but the disadvantage of lower accuracy than the bithermal test. When water is not used directly, heat or cold is not conducted as efficiently to the ear.When fewer irrigations are used, there is less averaging of data and lower accuracy. Because of the considerable inaccuracy even for the best of the lot, the bithermal water method, we think it is imprudent to adopt more convenient but less accurate methodology.
The caloric test is ordinarily performed with the subject reclining, head inclined 30 deg from horizontal so as to make the lateral canal horizontal. Water is introduced into the ear canal on one side, either 7 deg centigrade above or below assumed body temperature. The flow rate is such that the ear rapidly equilibrates with the water. The water is stopped after 30 seconds, and nystagmus is observed, while the subject is distracted (usually with tasks such as naming of animals, counting backgwards, etc). This is sometimes called "tasking", see following. Nystagmus commonly builds for about 30 seconds, then gradually decays away over roughly 2 minutes. After a rest of at least 5 minutes, the procedure is repeated with either the opposite temperature water, or on the other side.
Eye movements are usually recorded with either EOG or a video method, such as is shown on the graphic to the right. From the peak slow-phase velocity of nystagmus four numbers are obtained -- cold right, cold left, warm right, and warm left. These four numbers are used to compute three additional numbers:
There is considerable test-retest variability. The upper limits for test-retest variation for paresis were found by Proctor et al to be 24%. The upper limits for DP was 22%. This suggests that upper limits of normal should be set at greater than these values, which do not take into account variability of the normal population, and are therefore inappropriately low.
There are many methods of distracting persons during the process of recording their nystagmus. Without distraction, responses can be suppressed which reduces validity. Tasks in which the subject produces a listing of items from memory seem reasonable and effective.
|Quiz -- i.e. "what is your age", what is your favorite color.||Less effective||Fomby et al, 1992|
|Hand-motor task, clinician directed. Touch the thumb to finger as directed by clinician.|
|Alphabet task -- third letter of alphabet following a given letter randomly selected by clinician|
|Math task, Add or substract a number given by clinician from a running total|
|Quiz task: Name colors, states in USA, cities in ...||Best task|
|Hand-motor task -- touch thumb to first finger once, 2nd twice, third three times, etc.|
|Alphabet task #2 -- every third letter in the alphabet|
|Math task #2 -- count backwards by 3's or 7's.||Less effective|
There have been several attempts made to model the caloric response. The response theoretically a combination, possibly nonlinear, of convection induced stimulation of the canal, a direct effect of temperature on the nerve, transduction responses in the mechanics of the cupula, adaptation responses in the nerve and brainstem, and other central processing effects, mainly including velocity storage. A descriptive curve-fitting approach to the response is exemplified by that of Formby et al (1992, 2000).