Bone Conduction Audiometry Measurement Variability

Two common ways in which measurement variability in bone conduction audiometry can be introduced is through air borne radiation and the occlusion effect.  Air borne radiation is the disturbance in the air caused by the bone vibrator.  This air borne radiation disturbs the air in the ear which can be detected and can affect threshold measurements above 2 kHz.  To solve the problem of air borne radiation earplugs can be used to prevent sound entering the ear.  The use of ear plugs in bone conduction audiometry introduces another form of measurement variability, and this is variability due to the occlusion effect below 2 kHz. 

Although ideally it is desirable to use ear plugs when conducting bone tests above 2 kHz (see article on air radiated sound from bone vibrators), and remove them when testing below 2 kHz, it is not clinically practical.  The alternative is to accept certain degrees of variability and reject others.  When using the popular Radioear B-71 bone vibrator, air-bone gaps (ABGs) of 10 dB or more are considered clinically significant.  

Research

Lightfoot et al (1993) conducted a study to explore the high frequency measurement variability in bone testing.  Previous studies have shown that unoccluded bone thresholds of normal hearing subjects have often been lower than the minimum available intensity (-10 dB). 

The study was conducted in the following way.  Normal hearing subjects (age 18-30) were given bone conduction threshold tests with ears unoccluded, occluded with earphones and occluded with earplugs.  What Lightfoot et al. wanted to see was how much of an improvement in threshold would occur as a result of the occlusion effect.  The following is a table of their results. 

Taken from Lightfoot et al (1993)

Mean air-bone gaps (in dB) calculated with bone conduction hearing thresholds obtained with the subjects’ ears (a) unoccluded; (b) occluded by audiometric earphones; (c) occluded by earplugs.

As the above table indicates, occlusion by either earplugs or earphones did not significantly affect the ABG at 3-4 kHz but the gaps at 6-8 kHz were significantly larger, especially at 6 kHz.  According to the authors, these gaps can partially be attributed to central masking effects (3-5 dB at most) but central masking effects are usually not frequency specific such as the case shown in the above table.  The ABG at 6-8 kHz was suspected to be mostly due to air borne radiation but was discovered to be due to a discrepancy between the air and bone conduction reference equivalent thresholds given in British Standards BS2497 (part 5) and British Standards BS6950.  Lightfoot et al. (1993) offered three reason for this discrepancy.  First, the air conduction standard according to Robinson (1988) contains an error of 5-6 dB at 6 kHz.  Second, both the air and bone standards were derived from different subject samples.  Third, British Standards BS6950 is derived from three studies which used different bone vibrator types.        

Impact on Clinical Audiology

So now the question remains, how can this information impact an Audiologist in the clinical setting?  When testing ABG, there is often a lower limit on the magnitude of the averaged ABG when assessing the significance of any conductive element.  Does this mean that the variation presented in the Lightfoot et al study is clinically non significant?  Absolutely not because an Audiologist’s diagnosis can be affected by the presence of an apparent high frequency conductive loss.  Take the following hypothetical situation for example:

The above audiogram represents a fictitious man in his late 40’s who has worked in a noisy environment for 30 years.  The ABG on this audiogram is taken from  Lightfoot et al. (1993).  With this data, the authors argued that the audiogram above can be interpreted as a strange high frequency conductive problem with no convincing evidence of noise induced hearing loss.   

According to Lightfoot et al (1993) three things together contribute to the false ABG at high frequencies shown in the audiogram above.  First is the air conduction error at 6 kHz mentioned above.  Second is the apparent improvement in  bone conduction threshold caused by air radiated sound from the bone vibrator entering the ear canal at high frequencies.  Third is that unoccluded bone conduction thresholds at 6 and 8 kHz in normal listeners is often lower than what is measurable by most audiometers ( - 10 dB HL).  The latter two factors increase the likelihood of an ABG by contributing to an apparent improvement in BC thresholds.  

Lightfoot et al (1993) recommend that bone conduction tests should be avoided at frequencies above 4 kHz until standards are revised or until bone vibrator designs are improved.   

Note:  This site is best viewed using IE 4.X or higher under 800x600 resolution