Errors on ENG testing (under construction)

Timothy C. Hain, MD Page last modified: August 30, 2010

see also: Blunders

ENG (electronystagmography) testing is complex, and there are numerous errors that can easily arise. Unfortunately, these errors are very commonly encountered. In the author's view, this is due to a mixture of:

• Poorly designed equipment -- the ENG industry is immature and nearly all systems have serious software problems that requires considerable vigilance.
• Poorly trained testing personnel (usually audiologists). Here we favor institution of a certification process, but this is a whole other topic. A gigantic problem is that usually the overseer cannot observe what actually happened during the ENG test. Many operator errors can get "swept under the carpet"
• Inadaquete physician oversight. Often ENG's are not interpreted by anyone other than the technician who did the study. As most audiologists are untrained in neurology, this can lead to gigantic blunders. In our view, ENG's should be read by persons who have knowledge about how the brain affects eye movements.

The goal of this document is to illustrate some of the most common errors, and suggest solutions when available.

Global Errors:

Noisy ENG -- this is uninterpretable.	Terrible VNG tracking produces an uninterpretable recording.

• Noisy signal -- this is mainly a problem with ENG rather than VNG (see above).
• Tracking problems - when the camera is not "straight on", tracking can be lost, producing an equally awful recording.
• Drugged patient -- patients who come in on vestibular sedatives will produce decreased responses on testing.
• No data - -many labs omit to prove that their data is reliable. They omit the traces. This often means that they have something to hide. If you can't get a trace out of them, you often just have to repeat this uncomfortable test.

Paretic Eye fixation syndrome (see this link)

Calibration Errors

Calibration is the process of relating the signal coming from the recording apparatus to a known displacement of the eye. Calibration errors are very pernicious because they can make the entire test wrong (i.e. show too high or too low responses), and also they can also be easily hidden (i.e. operators neglect to provide an illustration of calibration.

The eye (read and green) does not reach the target (blue)	Saccade velocities are very slow (this is a technical error)

In essence, what happens during a calibration error, is that the operator asks the patient to look between two targets, and the patient does something else. This might occur because:

• Lack of cooperation
• Everything happened too fast
• Patient can't see the target -- a very common error that audiologists make when doing testing is to ignore assessing patient's vision. Of course, if the patient can't see the target, they can't be expected to follow it.
• Change in calibration between two different recordings

Saccadic test errors

Calibration error results in eyes that go "too fast". Eye drops off of computer "strip" chart.

• Blind patient
• Poor calibration
• Overly predictable protocol
• Blink artifact rejects saccadic nystagmus

• Head movement during testing

head movement during saccade testing produces a series of backup saccades following the saccade with some drift (which is the VOR).

Spontaneous nystagmus test errors

• Patient with congenital nystagmus -- lack of appreciation that need to test each eye viewing
• Patients with torsional nystagmus -- ignoring the video in favor of horizontal/vertical traces that do not register torsion.
• Poor calibration
• Light leak
• Operator error - -ignores nystagmus (especially torsional nystagmus)

Gaze evoked test errors

• Too small a gaze angle
• Different gaze angle between right and left
• Poor calibration

Pursuit test errors

• Blind patient (really !)
• Poor calibration
• Lack of appreciation of congenital nystagmus, especially latent nystagmus which has backup saccades.

Positional test errors

Because of technician error and inability to record torsion, positional testing is best done at the bedside rather than with recordings.

• Wrong position (this is the main one)
• Too slow repositioning
• Too short recording

Caloric test errors

Caloric testing is usually the most important procedure, as well as the most uncomfortable procedure for the patient. Because of the discomfort factor, it is the place where technicians often "cut corners" -- they reduce the # of irrigations, they use air rather than water, etc.

• Blocked ear canal
• Wrong temperature
  • Usually either warm or cold is wrong, resulting in an ENG that has much better responses for one or the other.
• Wrong stimulus (i.e. air rather than water)
  • Air is just not as good a stimulus as water, and it results in many "bilateral weakness" ENG's.
• Too short irrigation
• Too short interval between irrigations
  • One should wait 10 minutes between irrigations (ideally).
• Calibration error
  • This one is very hard to spot because there is no intrinsic calibration to caloric testing as there is in saccades and pursuit. One has to use other tests to figure this out.
• Lack of distraction
  • People can suppress their responses making the test worthless.
• Lack of ice calorics
  • If there is no clear response, one has to do ice.
• Too few irrigations
  • More irrigations means more accuracy.
• Poorly done scoring
• Lack of correlation between caloric testing, HSN, vibration and rotational testing (i.e not understanding that eerything has to be consistent).

Here the operator allowed the software to "find" a response where a response cannot occur (it is too early).