Dizziness can be one of the most vexing symptoms for audiologists. About 15 percent of the U.S. population experiences significant dizziness every year, and about 40 percent of those patients have inner ear disorders.1-2
While many dizzy patients don't have inner ear disturbances, test batteries are designed to detect central nervous system disorders, compelling audiologists to recognize diseases that cause dizziness, such as migraines or Chiari malformation. The problem is that audiology training programs may not have taught this.
Meniere's disease is responsible for only about 0.19 percent of all dizziness, with the remaining 60 percent resulting from central nervous system, other medical conditions, and a combination of unknown or psychiatric disorders.3 (Table 1.) That leaves the inner ear as just one potential source of dizziness.
Most causes of dizziness are nonvestibular, and most test interpretations will conclude “no vestibular disturbance.” But it is almost always wrong to suggest that a patient characterized simply by “no vestibular disturbance” has central dizziness, making it extremely important to know how to interpret dizziness test batteries properly.
The inner ear has five sensors that can cause dizziness: three semicircular canals and two otolithic organs. Current diagnostic technology can only assess the function of three of these sensors: electronystagmography (ENG), videonystagmography (VNG), and rotary chair tests assess the function of the lateral canal and the saccule, and one of the otolithic organs can be tested using vestibular evoked myogenic potential (VEMP).
What we don't have at present is a utricular test for the other otolithic organ, although one is currently in development, or a test for evaluating anterior or posterior canal function. Consequently, vestibular testing cannot exclude vestibular damage. Normal vestibular testing does not necessarily mean that the patient has a normal inner ear.
Vestibular evoked myogenic potential is an exciting new test of vestibular function that assesses the saccule. The ability to measure another sensor of the inner ear adds immensely to the vestibular assessment process.
VEMP testing is in rapid evolution. It is an easy and quick test to perform that does not involve making dizzy patients even dizzier, and there are numerous technical pitfalls involving patient effort and electrode placement. Currently, the most useful VEMP parameters are those that resemble caloric testing—the size of the response from each ear and the symmetry of responses.
VEMP responses, like hearing, decline with age, and it is unusual to find substantial VEMP responses in people over age 70. VEMPs are decreased by processes that damage the inner ear such as acoustic neuroma and ototoxicity, and responses are increased by pathologies that increase sound sensitivity of the inner ear such as superior canal dehiscence. VEMPs are usually much less affected by vestibular neuritis than caloric responses, making it possible to diagnose vestibular neuritis if the response is normal in the presence of caloric weakness. Interpretations should comment on this pattern, if present. If VEMP responses are large from both ears but there is substantial asymmetry, the interpretation should mention the possibility of superior canal dehiscence.
Although VEMP is a type of auditory evoked response and can be performed on many standard evoked potential devices simply by adjusting the protocol table, the Food and Drug Administration has prohibited manufacturers of evoked potential devices from selling the protocol to their customers. This has put the FDA in the odd position of making it more difficult to practice good vestibular medicine.
ENG AND VNG
Electronystagmography and videonystagmography are two different ways of recording a battery of measures mainly designed to detect vestibular imbalance or weakness. VNG is far superior to ENG because of higher resolution, greater stability, and the ability to view ocular torsion, which accompanies benign paroxysmal positional vertigo and superior canal dehiscence.
Items in the VNG test fall into two main categories: tests of vestibular function (spontaneous nystagmus, positional nystagmus, and calorics) and tests of oculomotor function (calibration, pursuit, optokinetic nystagmus). (Table 2.) Oculomotor tests are nearly always normal because of the relative rarity of CNS-related vertigo as well as the insensitivity of oculomotor tests to identify central lesions, even if present.
Vestibular testing abstracts list the findings produced by testing software. They should include the state of the patient (medication, sleep pattern, ability to follow instructions), the method of recording (VNG or ENG), the method of caloric stimulation (water or air), and any abnormal results. Caloric tests in particular must always include spontaneous nystagmus, paresis, and total response because these numbers are fundamental to identifying vestibular disorders. VNG and ENG interpretations should also include a conclusion on the existence of a significant vestibular abnormality and a differential diagnosis.
The most common mistake made in interpreting vestibular tests is a lack of appreciation for technical error. Low total caloric responses are usually due to use of air stimuli, which is not as reliable as water irrigation. Similarly, only technical error can cause one of four bithermal irrigations to be significantly weaker than the other three. On the other hand, the usual error made in interpreting oculomotor tests is overly enthusiastic attribution of the findings to central nervous system disorders.
Additional subtests also may be included in the VNG battery. (Table 3.) While useful, they often cannot be billed because CPT codes may not exist. Head-shaking and vibration of the neck are reliable methods of corroborating a unilateral vestibular weakness, and hyperventilation-induced nystagmus can be used to document irritability of the vestibular nerve and cardiac abnormalities detected by the Valsalva maneuver. Coupled with a large video image of the eye, this is an excellent method for diagnosing superior canal dehiscence. These subtests can generally only be effectively performed using the video recording method, and for this reason, are newcomers to vestibular testing.4
ROTARY CHAIR TEST
The rotary chair test is more sensitive than caloric testing for bilateral caloric weaknesses because it can detect a reduction as little as 50 percent in vestibular function. Caloric testing has wider limits of normal, and a reduction in vestibular function of approximately 80 percent is needed to cause an abnormal total response. Rotary chair testing is the gold standard for detecting bilateral vestibular weakness for this reason.5 A test of fixation suppression is usually bundled in with rotary tests.
Rotary chairs are far more costly than ENG systems, making them generally less available. This test also has no special value for diagnosing central disorders, for which VNG testing is equally sensitive, and it is poorly suited to detect unilateral vestibular weakness, for which VNG is far superior.
Interpretation of rotary chair tests should include statements about whether the patient's vestibular function is within normal range for his age and the effectiveness of fixation suppression. If abnormal, a differential diagnosis should be offered.
Cheaper devices use active head rotation rather than whole body rotation, but these are not adequate substitutes for a whole-body rotator because they cannot evaluate low-frequency responses. Responses from these devices may also be driven by neck input and mental intention, which don't accurately assess vestibular function. Some laboratories have exploited loopholes in Medicare regulations, and billed for multiple rotatory chair tests with the backing of these devices, but this type of billing is unethical.
COMPUTERIZED DYNAMIC POSTUROGRAPHY
Computerized dynamic posturography (CDP) is not a vestibular test, and it quantifies balance by registering sway in response to a set of challenging sensory conditions. CDP has not been found to have substantial value for clinical diagnosis, but it is useful in identifying inconsistent balance patterns.6
The NeuroCom Balance Master has been studied the most, and other variants of postural assessments are not properly billable as CDP tests because they are not computerized. They also lack a measurement sensor and moving platform.7 CDP sway may be involuntary or voluntary, and procedures are available to discriminate reasonably well between the two.8 Interpretations should indicate whether performance is normal for age, and if the result is not normal, it's important to determine whether the pattern is aphysiologic. If it is, the interpretation should include a differential diagnosis, including malingering as a possible source of the pattern. Although CDP is covered by Medicare, most commercial insurance providers do not cover the test. This is just one of many hurdles that still need to be overcome in formalizing vestibular testing techniques.