1. Presently, OSHA (29 CFR 1910.95) does not require testing at 8 kHz. Without testing 8 kHz, a worker with a substantial 6-kHz notch would exhibit an audiometric configuration indistinguishable from hearing loss due to presbycusis. Although even one missed diagnosis is one too many, is this a scenario occupational hearing conservationists, audiologists, and physicians are likely to encounter? How common is a noise notch at frequencies other than 4 kHz? New data are emerging that can answer these questions. These data are part of the current National Health and Nutrition Examination Survey (NHANES) conducted by the National Center for Health Statistics with collaboration from the National Institute for Occupational Safety and Health (NIOSH) and the National Institute on Deafness and Other Communication Disorders (NIDCD). As part of the NHANES study, hearing thresholds from 500 Hz to 8 kHz are being collected on adults 20 to 69 years of age. From data collected during 1999 and 2000, 3320 hearing tests have been evaluated for the presence or absence of a notch. Of the audiograms reviewed to date, 474 persons reported a history of occupational noise exposure. From this group, notches were identified in 317 ears. From these data a 4-kHz notch was clearly not the frequency of most common occurrence. Instead, if a notch was present, it was more likely to occur at 6 kHz, particularly in females. Table 2 reports the results of this analysis for females and males, respectively. Early indicators of noise induced hearing loss: Issues for consideration Mark R. Stephenson, William J. Murphy and John R. Franks, National Institute for Occupational Safety and Health Abstract: A recent article in the NHCA Spectrum [Rabinowitz and Dobie, 20:8-11, 2003] considered several methods for identifying the early onset of noise-induced hearing loss (NIHL). Among the methods they examined was the Notch Index: the difference in the pure-tone average of 2, 3 and 4 kHz and the pure-tone average of 1 and 8 kHz. This statistic was expected to be positive for NIHL where noise has produced a threshold shift in the 2 to 4- kHz region but not adversely affected the thresholds at 1 and 8- kHz. The Notch Index was examined for a subset of 1660 subjects’ right and left audiograms collected for the National Health and Nutrition Examination Survey (NHANES IV). As proposed by Rabinowitz and Dobie, the Notch Index did not prove to be effective. However, modifications to the calculation, the inclusion of 6 kHz, or calculation of the area difference, provided better identifications of the early onset of NIHL. This paper will also discuss how these findings may be related to issues such as work-relatedness and potential gender effects. Background: The primary purpose for audiometric monitoring is to identify hearing changes while they are still small. Hopefully, these changes can be identified before they represent permanent changes, i.e., while they are temporary threshold shifts. The susceptibility to developing a hearing loss is not the same at all frequencies. Assuming that a worker’s ears are otherwise healthy, that there is no excess build-up of cerumen, and that appropriate audiometric test methods are followed, hearing changes due to noise are first detected and grow most rapidly at the higher frequencies. The American National Standards Institute (ANSI) provides data that estimate expected hearing loss as a function of noise level, years of exposure, and audiometric frequency (ANSI, 1996). Table 1 below illustrates the predicted noise-induced hearing loss for workers exposed to time-weighted average levels of 95 dBA. Table 1. Predicted noise-induced hearing loss from daily time- weighted average exposure to 95 dBA for 10, 20, 30, and 40 years (ANSI S3.44-1996). Plotting the data in Table 1 illustrates how noise-induced hearing loss is primarily observed in the higher frequencies (Figure 1). Note also that the greatest hearing loss is observed at 4 kHz, giving rise to what is commonly referred to as the 4-kHz “notch”. While these data are illustrative of the 4-kHz notch, they also demonstrate the presence of considerable noise-induced hearing loss at 3 and 6 kHz. Although noise-induced hearing loss is stereotypically char- acterized by a 4-kHz notch, the occurrence of a notch at 3 or 6 kHz is well-known. Because of the recent OSHA rule on record- ing occupational hearing loss (OSHA, 2003), the ability to deter- mine the presence or absence of a “noise notch” must be reexam- ined. OSHA’s new rule includes a provision for determining whether or not a hearing loss is work-related. Consider how a 6- kHz notch affects the interpretation of a worker’s audiogram. Discussion: The evaluation of audiograms by algorithm presents an opportunity to be proactive in the prevention of hearing loss. The analysis of the Rabinowitz and Dobie Notch Index suggests that it would not perform as well if 6 kHz is excluded from the calculation. A variant of this analysis used the area of the audiogram (with 6 kHz included Figure 4) and shifted the histogram to a more positive mean notch index. However the width of the distribution demonstrated that the area calculation was not as specific as the Notch Index determined with pure-tone averages. The inclusion of 6 kHz in the Notch Index calculation didn’t shift the distribution as much as the area calculation, but the peak was sharpened and the distribution was narrower as well. Consideration of the NHANES data suggests that if 8 kHz is not tested, the notch would be difficult to identify. The NHANES data reported in Table 2 clearly demonstrate that the majority of notches among self-reported occupationally noise-exposed subjects had notches at 6 kHz. If 8 kHz is not tested, then identification of the notch would be missed on 55% of these subjects. The issue is not identification of an STS, but rather the prevention of a noise notch becoming an STS. Without the knowledge of the recovery at 8 kHz, the notches would present as a sloping hearing loss. Conclusions: The current data represent preliminary results from the National Health and Nutrition Examination Survey. When the six-year NHANES data collection period is complete at the end of 2004, approximately 6000 hearing tests will have been administered. A more definitive analysis examining the effects of other variables (e.g. age gender and ethnicity) on the prevalence of 3, 4, and 6- kHz notches will be forthcoming following the collection period. The 6-kHz frequency is critical to effective identification of the noise-induced notch. Proactive occupational hearing conservationists should be aware that the noise notch is very likely to occur not just at 4 kHz, but at 6 kHz as well. References: American Academy of Audiology (2003) “Position Statement: Preventing noise-induced occupational hearing loss”, Reston, VA, October, 2003. ANSI (1996). “Determination of occupational noise exposure and estimation of noise- induced hearing impairment”. American National Standards Institute, S3.44-1996, New York, NY. NIOSH (1998). Revised criteria for a recommended standard – occupational noise exposure. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication 98-126. OSHA (1983). “Occupational Noise Exposure: Hearing Conservation Amendment; Final Rule”. U.S. Department of Labor, Occupational Safety and Health Administration, 29 CFR 1910.95, 48 Federal Register 9738-9734. OSHA (2002). “Occupational injury and illness recording and reporting requirements.” U.S. Department of Labor, Occupational Safety and Health Administration, 29 CFR 1904.10, Federal Register, December 17, 2002, Vol. 67, No. 242, pp. 77165-77170. Niskar AS; Kieszak SM; Holmes AE; Esteban E; Rubin C, and Brody DJ. Estimated prevalence of noise-induced hearing threshold shifts among children 6 to 19 years of age: The Third National Health and Nutrition Examination Survey, 1988-1994, United States. Pediatr. 2001; 108(1):40-43. Rabinowitz, PM, and Dobie RA. Use of the Audiometric Configuration to Determine Whether Hearing Loss is Noise-Induced: Can “Notch Criteria Help?” NHCA Spectrum 2003; 20(1):8-11. Stephenson, MR, Themann, CL, and Murphy, WJ. Thoughts on the Noise “Notch” and the Importance of Testing 8 kHz. CAOHC Update Correspondence: William J. Murphy, Ph.D., National Institute for Occupational Safety and Health, 4676 Columbia Parkway MS C-27 Cincinnati, OH 45226-1998 wjm4@cdc.gov Although the above data represent preliminary findings from the first two years of the NHANES data set, occupational hearing conservationists (OHCs) should be alert to the likely presence of a notch at frequencies other than 4 kHz. If a worker presents with a sloping high-frequency hearing loss and 8 kHz is not tested, determination of the presence of a 6-kHz notch is not possible. In fact, a 6-kHz notch is so common that OHCs may wish to reconsider which audiometric frequencies they wish to test. Thus, the question remains: if the OHC is not already doing so, should he/she start testing at 8 kHz? NIOSH (1998) and the American Academy of Audiology (2003) have recently published guidelines that recommend including 8 kHz when performing audiometric monitoring in support of a hearing conservation/hearing loss prevention program. While this represents “best practice”, OHCs may find it is nearly as effective to include 8 kHz only on certain hearing tests. For example, all audiograms that provide reference hearing threshold levels, i.e., baseline and exit audiograms should include 8 kHz. Also, when an STS is suspected, any subsequent retest or confirmatory audiogram should include 8 kHz. Notch Indicators: Rabinowitz and Dobie (2003) proposed the use of a notch index to flag the presence of a notched audiogram. Their statistic was calculated as mean(  2K,  3K,  4K ) – mean(  1K,  8K ). In their article, several other papers were examined for alternative approaches to notch identification. Of these, the approach by Niskar et al. (2001), provided a means to identify with some certainty a subpopulation of notched audiograms. Niskar proposed the following criteria for audiograms: 1. 500 and 1000 Hz less than or equal to 15 dB HL. 2. Notch frequency 15 dB above worst threshold at 0.5 & 1 kHz. 3. 8 kHz exhibits 10-dB recovery from notch frequency. Using these criteria, a subset of subjects were identified from within the NHANES data set for use in the numerical investigations described below. Frequency10 yr20 yr30 yr40 yr 5000011 10002333 2000591214 300016192223 400020232526 600014161819 Table 2. Number of occupationally noise-exposed female and male subjects having notches at selected frequencies as determined by unanimous consensus of three audiologists (N = 474 subjects self-identified as occupationally exposed). Females Left ear Females Right ear Males Left ear Males Right ear 3000 Hz121012 4000 Hz15114251 6000 Hz21355364 Figure 2. Histogram of the Notch Index using 2, 3 & 4 kHz for both right and left ears. The Notch Index was calculated for a subset of subjects from the NHANES data identified with the Niskar criteria. The histogram is nearly symmetric about 0, indicating a lack of specificity for this method. More right-ear notches are present than left-ear notches which agrees with the analysis from self-identified occupationally exposed workers in Table 2. Figure 3. Histogram of the Notch Index using 2, 3, 4 & 6 kHz for both right and left ears. The addition of the 6-kHz frequency shifts the distribution of notches to predominantly positive differences in the respective pure-tone averages, mean(  2K,  3K,  4K,  6K ) – mean(  1K,  8K ). This definition appears to provide a better selection than the original definition. Figure 4. Histogram of the Notch Index Area using 2, 3, 4 & 6 kHz for both right and left ears. The notch index definition was further modified to consider the area of the audiogram if one drew an imaginary line between the thresholds at 1 – 8 kHz and the area if all frequencies in the range 1 – 8 kHz were used. Area(  1K,  2K,  3K,  4K,  6K,   K ) – Area(  1K,  8K ). These areas are composed of trapezoidal shapes that must be weighted by the respective portions of an octave band. Like the refined Notch Index in Figure 3, the mean of the distribution is more positive, but the standard deviation is increased flattening the histogram somewhat.