The likelihood of detecting a significant hearing threshold shift among noise-exposed workers subjected to annual audiometric testing.

The ability of annual audiometric surveillance to detect significant deteriorations in hearing sensitivity is examined considering the time pattern of the growth of noise-induced hearing loss. The yearly rate of change in hearing level due to the effects of noise exposure and age is computed from a mathematical model put forward in the International Standard ISO 1999 (1989). The first-order derivative of hearing level associated with noise and age was compared with different criteria of significant threshold shift (STS). The results show that the potential annual variation in hearing level at 4 kHz, the audiometric frequency most affected by noise, is always less than the minimum margin of error of audiometry under exposure to 90 dBA-8 h. At 100 dBA it nearly equals the margin of error only for the very first year of exposure and for the most sensitive individuals. The rate of change in hearing threshold at 4 kHz is less than or equal to 1 dB per year for a major portion of a population exposed to an average noise level of 90 dBA-8 h (+/- 5 dB SD). For other audiometric frequencies, or for combinations of frequencies, this rate is even lower. It is also demonstrated that, using annual hearing tests, the likelihood of measuring an STS in a noise-exposed population is very small. It would be nearly equal to the likelihood of some workers being exposed to 100 dBA-8 h or more for a first year. Such results throw doubt on the validity of audiometric surveillance as a procedure for early detection of noise-induced hearing loss. They also imply that the detection by audiometric tests of true cases of significant threshold shift due to noise is excessively expensive.