[Correlation of pure tone thresholds and hearing loss for numbers. Comparison of three calculation variations for plausibility checking in expertise]

[Speech audiometry for indication of conventional and implantable hearing aids]

The social function of the human hearing apparatus is comprehension of speech. Auditory rehabilitation aims to enhance speech perception in everyday life. Consequently, audiological evaluation contains speech perception measurement. Many speech audiometric methods have been developed in German-speaking countries, which are suitable for quantification of speech perception abilities in quiet and in noise to address specific diagnostic questions. For establishment of the indication for technical hearing systems such as hearing aids and cochlear implants, the Freiburg monosyllabic test has been employed successfully for many years. Particularly sentence tests have been suggested as a complementary measure for hearing aid indication and assessment. This paper describes the characteristics and range of application of various speech audiometric methods, and highlights the methodological limitations of the individual tests.

[Two formulas for exact calculation of hearing loss for numbers]

BACKGROUND

In Germany assessment of hearing loss for numbers is used to calculate the percentage hearing loss from speech audiometry and for plausibility checking with pure tone thresholds. It is common practice to take a graphical reading from the speech audiogram to determine the hearing loss for numbers. This study searches for a mathematical formula for the exact calculation of the hearing loss for numbers from the intelligibility values measured.

METHODS

With analytical methods on the basis of the standard curve following DIN 45626-1, two simple formulas for the calculation of the hearing loss for numbers were developed.

RESULTS

The hearing loss for numbers a(1) is calculated as a(1)=p(1)+(50-v(1)) (p(2)-p(1))/(v(2)-v(1))-18,4 for two available measured values and as a(1)=p(1)-0,13 v(1)-11,9 if only one measured value is used, with v(i )being number intelligibility in percent at the level p(i) in decibels (dB) of measurement number i. Number intelligibility of all inserted pairs of values must be between 30% and 70% because the standard curve of DIN 45626-1 runs approximately linearly only in this range. The calculated value for the hearing loss for numbers is subsequently mathematically rounded up to 5 dB as well as for the conventional graphically determined value.

CONCLUSIONS

With the presented formulas the hearing loss for numbers can be calculated exactly from the measured values of the Freiburg number test especially in matters of expertise if the conventional graphical determination of this value does not seem to be unambiguous.