Assesssing Hearing Loss
Before it is possible to manage hearing loss, this loss has to be detected and assigned to the appropriate categories of conductive, sensorineural or mixed hearing loss. The basic test is the pure-tone audiogram (PTA). Other common tests are also described here, but there is a whole science of hearing tests, which is beyond the scope of this book.
THE PURE-TONE AUDIOGRAM
The aim of this test is to determine the quietest sound levels that an individual can just detect for a range of notes across the frequency keyboard. This level is called the ‘hearing threshold’, which represents the volume of sound that can just be detected at any particular pitch. The range of the test is generally spread over the five octaves from 250 Hz to 8,000 Hz. An octave is a doubling of frequency, that is, from 250 to 500 Hz then from 500 to 1,000 Hz, from 1,000 to 2,000 Hz, and so on. Although each octave is a doubling of frequency, the brain hears each octave as equal steps.
The test starts at one frequency, typically 1,000 Hz. The person being tested wears standardised headphones connected to an audiometer. The sound is directed to one ear at a time and is played at moderate volume, at say 60 dB, for several seconds. If the tone is heard, the person, who should not be facing the tester, presses a response button and holds it down while he or she is hearing the sound, so that the tester knows that there is a positive response. The volume is then reduced in 10-dB steps and the process repeated until the person no longer detects the sound. The tester then increases the volume in 5-dB steps until the sound is detected again and then down in 10-dB steps and up in 5-dB steps until the sound is detected three times out of five on the ‘ascent’. The decibel level is then entered on the audiogram chart and the process repeated for other frequencies up and down the scale, normally at 250, 500, 1,000, 2,000, 4,000 and 8,000 Hz for routine testing. If the person is involved in a claim for damages or there are abnormalities, 3,000 and 6,000 Hz and possibly other frequencies are tested as well.
Although 0 dB is the average lower limit of hearing for healthy 18 year olds, the upper limit of normal is taken as 20 dB. Lower limit hearing thresholds worse than 20 dB – that is, 30, 40, 50 or more – suggest a reduced hearing level (HL).
BONE-CONDUCTION TEST
Testing with headphones gives ‘air conduction’ (AC) levels and tests the whole system from the external ear canal to the auditory perception regions of the brain. This test does not distinguish whether a loss is conductive, sensorineural or mixed. To do this the inner ear must be tested and this is achieved by assessing ‘bone conduction’ (BC). In this test a small vibrator, which is also connected to the same audiometer, is pressed firmly against the skull usually over the mastoid prominence. Vibrations from the device enter the skull and are transmitted directly to the inner ear, bypassing the ear canal and middle ear. This was how Beethoven, who must have had a severe conductive deafness, was able to hear, by pressing his head via a stick to his piano.
The same testing process as for the pure-tone audiogram is repeated using the vibrator, giving bone-conduction thresholds that represent the hearing level in the cochlea, acoustic nerve and brainstem pathways up to the level of the auditory cortex. Again, normal bone-conduction thresholds are taken as being better than (that is, less than) the 20 dB level, with 0 dB being the average normal level.
If the air conduction shows poor hearing, but the bone conduction is normal, there is an ‘air–bone gap’ and the hearing loss is conductive. If both the bone conduction and air conduction are similarly poor and there is no significant air–bone gap then the loss is ‘sensorineural’. If there is both poor air conduction and poor bone conduction and a significant air–bone gap – that is, 15 dB or more – the loss is mixed.
There are some practical difficulties with the PTA that can make interpretation impossible if a test is not conducted with great care. The most common one occurs with bone-conduction sound applied by the vibrator to one side of the head. This sound vibrates the whole skull and is transmitted to both the inner ears with only a 5-dB reduction in sound energy occurring across the skull. Thus, if one ear is being tested and the inner ear on that side is damaged, the sound will be detected by the other good ear and ‘heard’. This gives a false impression of the hearing level in the tested ear and can result in serious errors in diagnosis. A masking sound, usually in the form of white noise – a shushing sound – has to be applied to the non-test ear to prevent this kind of mistake happening.
Another problem occurs when one ear has a large conductive loss but good bone conduction, that is, a good sensorineural reserve, and the other ear has a serious sensorineural loss. It may be impossible to get enough masking into the non-test ear with the conductive loss to make even a reasonable assessment of the true bone-conduction levels in either ear.
Once it has been decided that the loss is sensorineural, there are several techniques for deciding where the problem lies.
AUDITORY EVOKED POTENTIAL TESTING
Once the cochlea has detected an incoming sound, it converts it to an electrical signal, which travels along the nerves to the areas of the brain responsible for perception. The whole electrical journey takes only about one-fifth of a second. It is possible to detect the various electrical steps along this route by placing sensitive electrodes on the scalp and over the mastoid region. The signals occur whenever the cochlea detects sounds and so the individual’s active cooperation is not needed.
Unfortunately the auditory signals are extremely small and are swamped in the generalised electrical activity of the brain, the nerves and the muscles. To overcome this problem, we can present a short ‘click’ stimulus to the ear and then record on a computer the electrical activity from the electrodes for a short while, say 10 milliseconds – that is, ten-thousandths of a second – after. This time interval is then divided by the computer into even smaller intervals or ‘bins’. During the ten milliseconds of recording, the activity of the cochlea and the responses of the acoustic nerve are measured. The test is therefore called electrocochleography. If the recording time is lengthened the activity in different parts of the auditory pathway – the brain stem and auditory cortex – can be assessed.
As the signal arrives at the computer, it is allocated bit by bit to successive ‘bins’. The electrical signal at each point is either positive or negative and has a certain voltage, so that the bin accumulates a plus or minus value. The process is restarted with the computer returning to the first bin before the next click is made. The process is repeated again and again in this time-locked fashion.
The background electrical activity of the brain, nerves and muscles is more or less random and so, after a time, there will be about equal numbers of positives and negatives of a range of voltages arising from these structures allocated to each bin. When these random elements of the signal are all added up they more or less cancel each other out.
This leaves the auditory signal intact because, at any point of collection, the auditory signal is always positive or always negative and these values accumulate with successive time-locked samples. The whole process is called auditory evoked potential (AEP) testing. Most computer programs now have the ability to work out the signal-tonoise effect and can cancel out unduly noisy signals so that the best result is achieved in the least time. Depending on the time window during which data are collected, different parts of the auditory system can be assessed.
Electrocochleography (ECoG), brain-stem (BAEP) or cortical (CAEP) auditory evoked potentials all give different information about different parts of the auditory pathway. What is very important about this form of measurement technique is that the tester does not need the cooperation of the person to obtain a response. This whole group of tests is therefore called objective audiometry, as opposed to the term ‘subjective audiometry’ which is used for pure-tone testing and for other tests needing the person’s response. BAEP and CAEP can detect a response without the person’s help and can also be used to create the equivalent of a pure-tone audiogram without the person having to respond.
This is achieved in a similar fashion to the pure-tone audiogram described above. The testing starts with a loud click, or better still by using a short burst of a pure tone (that is, a sound of a single frequency) usually of 1,000, 2,000 or 3,000 Hz, to get a response. Then the volume is gradually reduced until the electrical response to click or the tone burst disappears. This level is taken as being close to the true hearing level (hearing threshold).
This technique is very useful in testing those people who, for some reason, are unable to produce a reliable conventional pure-tone audiogram. This might include those claiming compensation for injury who may inadvertently exaggerate their hearing loss; those who feign a hearing loss as an excuse for criminal activities; and a few people with psychological problems that show up as deafness. One big group who cannot respond is made up of babies and young children, and it is just this group in whom it is essential to be certain whether the hearing is intact and, if not, how bad it is.
DISTRACTION TECHNIQUES
Conventional hearing testing of babies who can sit up and hold up their heads has relied on what are called ‘distraction techniques’. The child sits on a parent’s lap and faces one of the testers who engages his attention. The second tester stands behind the child to one side with some means of producing sound. Nowadays, calibrated devices producing pure tones of different frequencies are in common use. When a sound is made at about a metre from the child’s ear and at the same level, the child will go still, turn his eyes to one side or even turn his whole head if he is aware of the sound. This is taken as a positive response. A failure to do this is taken as a negative response, although too much concentration on the first tester, tiredness and a variety of other factors can conspire to mean that the test does not really prove that the child has poor hearing.
Other sounds such as a spoon in a cup, rattles or squeakers can be used and are good at producing a response from children. However, with each of these test items, a range of frequencies is produced and a deaf child with residual low-tone hearing can often give a response to the low-tone elements of the test sound and thereby give a reassuring but false test result. Good teamwork between the two testers is also needed to obtain reliable results. When a child consistently fails to respond to such tests, he will require further assessment, often by some form of brainstem auditory evoked potential (BAEP) testing as described on page 33. This would need to be performed under general anaesthetic or sedation.
There are variations of the distraction test, including visual reinforcement audiometry where a correct response is ‘rewarded’ by a cuddly toy in a darkened box being lit up for a moment or two. After the age of about two and a half years, performance tests can be introduced. Here, the child performs an action such as putting a toy brick in a box in response to sound and usually with a reward for a correct answer.
OTOACOUSTIC EMISSION TESTING
Distraction techniques cannot be used with a newborn baby and routine auditory evoked potential testing as a screening procedure is simply not feasible to detect those with a moderate or more severe loss because it takes time, considerable skill and a placid, sleeping or anaesthetised baby.
You may recall, from the description of how the cochlea works, that the outer hair cells become active when the crest of the travelling wave reaches them. The outer hair cells respond in a mechanical fashion by shortening or lengthening, which somehow brings about mechanical amplification of the sound wave. This then results in the bending of the hairs of the inner hair cells and the production of cochlear nerve impulses.
Whatever the exact mechanism of the cochlear amplifier, the process is not 100 per cent efficient, and some of the mechanical energy generated by the outer hair cells travels back down the cochlea and moves the footplate of the stapes from the inside. This movement is transmitted through the ossicular chain back to the eardrum and thence to the ear canal. In the 1970s, David Kemp (now Professor of Auditory Biophysics at University College London) predicted that these cochlear ‘echoes’ should exist and eventually was able to detect them in the ear canal by using computer-averaging techniques. These signals are not, however, electrical in nature, but are sound waves coming back out of the cochlea in response to sounds going in. In a normal ear, they bear a resemblance to the in-going sound, hence the term ‘cochlear echo’.
Eventually, these cochlear echoes acquired the more descriptive title of otoacoustic emissions (OAEs). For OAEs to be present, the cochlea has to be healthy with an intact set of outer hair cells (OHCs), and the middle ear also has to be normal. In general, the hearing level has to be 30 dB or better for an OAE response to be detected. Thus, if there is a response, the hearing is normal. The test is quick, repeatable, non-invasive and does not require skilled technicians or scientists to perform it and, importantly, the equipment is no longer expensive.
The test involves introducing a gentle sound into the outer ear canal with a small probe. A fraction of a second later, the probe picks up an ‘echo’ returning from the cochlea. The echo is present in all hearing people and can be measured by computer. If the echo is absent, this may indicate that a child’s hearing mechanism is not working properly and further tests will be needed.
In many countries, routine screening of all newborn babies in hospitals is required by law and those who fail the test go on to further screening and assessment. It seems much more sensible to be aware of and manage a problem as early as possible rather than to allow it to be discovered later when the remedy may be too late to benefit the child.
Universal screening of hearing for babies has finally been accepted in principle in the UK. However, the resources to fund the implementation of a screening programme are not yet available, despite comprehensive American studies showing that the technique is effective, efficient and economical.
TYMPANOMETRY (IMPEDANCE AUDIOMETRY)
As well as being able to test the hearing level, it is often very helpful to know about the state of the middle ear. Tympanometry can help in this respect and relies on the fact that the eardrum, although good, is not a perfect absorber of sound. Sounds put into the canal are partly absorbed and partly reflected by the eardrum, and a microphone can detect this reflected sound. If an airtight probe containing a small loudspeaker, a small microphone and a fine tube connected to an air pressure pump is put into the ear canal, some of the sound-conducting properties of the middle ear can be measured.
Raising or lowering the pressure in the ear canal stretches the eardrum a little, making it less efficient so that more sound is reflected. In more technical terms, the impedance increases. The microphone can detect this change as more sound is reflected from the stretched eardrum and the change in impedance can be calculated. If the pressure in the ear canal is gently changed from normal atmospheric pressure to a slightly increased pressure, and then gradually reduced through normal to a low negative pressure, the changes in the efficiency of the eardrum can be measured continuously and plotted on a small graph. This is the tympanogram.
A normal tympanogram has a bell-shaped curve with the peak of the curve close to the normal atmospheric pressure (type A curve). A low, flat trace occurs when the middle ear is full of fluid and the eardrum is very inefficient (type B curve). If there is air in the middle ear, but the pressure is reduced because of eustachian tube inadequacy, the peak of the trace is shifted to the low-pressure region of the graph and is usually reduced in height (type C curve). This indicates that to get the eardrum into its most efficient unstretched state a low pressure in the ear canal has to be created to match the low pressure in the middle ear. This test therefore also gives us a measure of the pressure in the middle ear.
If the bones in the middle ear are broken or not connected to each other, as can happen after a head injury, then a very tall, peaked tracing is obtained. If the bones are stuck together, as in otosclerosis, the peak of the trace is often reduced in height, although still in the normal pressure region.
