Archive Notice - OSHA Archive

NOTICE: This is an OSHA Archive Document, and may no longer represent OSHA Policy. It is presented here as historical content, for research and review purposes only.

OSHA requirements are set by statute, standards and regulations. Our interpretation letters explain these requirements and how they apply to particular circumstances, but they cannot create additional employer obligations. This letter constitutes OSHA's interpretation of the requirements discussed. Note that our enforcement guidance may be affected by changes to OSHA rules. Also, from time to time we update our guidance in response to new information. To keep apprised of such developments, you can consult OSHA's website at

August 31, 1993

Fredrik Lindgren, Ph.D. Director,
Research and Development
Cabot Safety Corporation
5457 West 79th Street
Indianapolis, Indiana 46268

Dear Dr. Lindgren:

Thank you for your letters of October 25, 1991, addressed to Mr. Gail Brinkerhoff of my staff, concerning the use of insert earphones in establishing the audiometric testing program that is required by the Occupational Safety and Health Administration's noise standard, 29 CFR 1910.95. We regret the delay in this response.

In your letter, you expressed concern regarding the conditions specified by OSHA, under which the use of insert earphones for audiometric testing would be treated as a de minimis violation of 29 CFR 1910.95(h)(2). A de minimis violation is one which has no direct or immediate relationship to safety or health. The violation is documented in the case file, but the employer is not issued a citation.

We consulted with the National Institute for Occupational Safety and Health (NIOSH) and with its assistance re-evaluated the appropriate conditions to place on the use of insert earphones giving consideration to the fact that this is a newer approach to industrial audiometric testing that has not been subjected to extensive field evaluation. As a result of this re-evaluation, OSHA has revised the conditions for which the use of insert earphones by employers would constitute a de minimis violation of 29 CFR 1910.95(h)(2). The following conditions must be implemented by employers who intend to use insert earphones in order to meet the criteria of a de minimis violation of OSHA'S noise standard. Failure to implement any of these conditions abrogates the basis of a deminimis violation, and may result in the issuance of a citation. The specific conditions are as follows:

o The audiologist, otolaryngologist, or other physician responsible for conducting the audiological testing, shall identify ear canals that prevent achievement of an acceptable fit with insert earphones, or shall assure that any technician under his/her authority who conducts audiometric testing with insert earphones has the ability to identify such ear canals.

o Insert earphones shall not be used for audiometric testing of employees with ear canal sizes that prevent achievement of an acceptable insertion depth (fit). The manufacturer's guidelines for proper use of insert earphones must be followed.

o Technicians who will be conducting audiometric tests must be trained to insert the earphones correctly into the ear canals of test subjects and to recognize conditions where ear canal size prevents achievement of an acceptable insertion depth (fit).

o The audiometer to be used with supra-aural earphones must be calibrated, by using the reference equivalent threshold pressure levels (RETSPLs) for the NBS-9A coupler presented in Table 6 of ANSI S3.6-1989. The audiometer used with insert earphones must be calibrated using the interim RETSPLs for either the HA-1 coupler or the occluded ear simulator presented in Appendix G to ANSI S3.6-1989. This acoustical calibration shall be conducted on an annual basis, and functional calibration must be conducted before each day's use of the audiometer.

o Testing shall be conducted in a room where the background ambient noise octave-band sound pressures levels meet OSHA's requirements in 29 CFR 1910.95(h)(4).

o At the time of conversion from supra-aural to insert earphones, testing must be performed with both types of earphones. The test subject must have a quiet period of at least 14 hours before testing. Hearing protectors may be used as a substitute for this requirement. The supra-aural- earphone audiogram shall be compared to the baseline audiogram, or the revised baseline audiogram if appropriate, to check for a Standard Threshold Shift (STS). In accordance with 29 CFR 1910.95(g)(7)(ii), if the audiogram shows an STS, retesting with supra-aural earphones may be performed within 30 days and the resulting audiogram adopted instead of the prior one. If retesting with supra-aural earphones is performed, then retesting with insert earphones must be performed in conjunction.

o If an STS is indicated by the test with supra-aural earphones, the audiogram may become the revised baseline audiogram per the requirements of 29.CFR 1910.95(g)(9) for all future hearing tests with supra-aural earphones. The insert-earphone audiogram will become the new reference audiogram for all future hearing tests performed with insert earphones.

o If an STS is not indicated by the test with supra-aural earphones, the baseline audiogram remains the reference audiogram for all future supra-aural earphone tests, until such time as an STS is observed.

Please bear in mind that employers must account for the workers who are in the process of developing an STS (e.g. demonstrate a 7 dB average shift), but who at the time of the conversion to insert earphones do not have a 10 dB average shift. Employers who want to use insert earphones must enter the 7 dB shift information in the employee's audiometric test records although it is not an "STS". When the next annual audiogram using insert earphones shows an average threshold shift at 2000, 3000 and 4000 Hz of 3 dB, completing the full shift (7 dB + 3 dB), employers must then label that average shift as an STS. This triggers the follow-up procedures at 29 CFR 1910.95(g)(8).

o All audiograms (both the insert and supra-aural types), calculations, pure-tone individual and average threshold shifts, full STS migrations, and audiometric acoustical calibration records, are to be preserved as records and maintained for the duration of the affected employee's employment and, upon request, are to be provided for review to employees, representatives designated by the individual employee, and the Assistant Secretary of Labor for Occupational Safety and Health, or a designee.

We hope this information is useful to you. Please feel free to contact Ms. Ruth McCully, Director, Office of Health Compliance Assistance at (202) 219-8036, should you have further questions.


Roger A. Clark, Director
Directorate of Compliance Programs

October 25, 1991

Mr. Gail Brinkerhoff
Industrial Hygienist
Office of Health Compliance OSHA,
U.S. Dept of Labor
200 Constitution Avenue, N.W.
Washington, DC 20210

Dear Mr. Brinkerhoff:

Thank you for taking the time to meet with us on September 10 and discuss the insert earphone issues. Although I can understand your viewpoint that changing over to insert earphones in industrial settings might eventually prevent detection of a Standard Threshold Shift (STS), the risk is probably overestimated. I have taken the liberty to work directly with your example and will try to break it down into relevant topics.

Specifically, you presented the following hypothetical example:


(For Figure, see printed copy)

There are a number of assumptions in order for the scenario to become a reality:

- Masking due to background noise levels must reach 10 - 15 dB at 1 kHz and at the same time produce 10 and 5 dB at 2 and 4 kHz respectively. The current literature indicates that the maximum possible masking when sound pressure levels are approaching the OSHA stipulated maximum permissible background sound levels, can reach 8.9, 11.8 and 10.5 at 1, 2, and 4 kHz respectively (Berger and Killion, 1989(1)). Hypothetically, the situation can arise when thresholds will be improved due to the reduced risk for masking when insert earphones are utilized. This should be limited to HTLs within 0 - 10 dB HL at 1, 2 and 4 kHz. The referenced work by Franks et al.(2) does not include measured masked thresholds per se, but rather computations based upon measured Real Ear at Threshold measurements and formulas outlined in the Berger & Killion paper.

FOOTNOTE(1) Berger, E. H. and Killion, M. C., 1989.  Comparison of the noise
attenuation of three audiometric earphones with additional data on masking
near threshold.  J Acoust Soc Am, 86, 1392 - 1403.

FOOTNOTE(2) Franks, J. R., Engel III, D. P. and Themann, C. L (Accepted for publication in Ear Hear) Real-ear attenuation at threshold for three audiometric headphone devices: implications for maximum permissible ambient noise level standards.

- Test frequencies 2 and 4 kHz are compromised with elevated thresholds due to masking. In reality, most test environments exhibit less of a problem with fulfilling the OSHA maximum permitted noise levels at frequencies above 500 Hz. This is due to the fact that enclosures will provide more attenuation at higher frequencies and disturbances from machinery and other sources found in industry tend to have a low frequency profile (31.5 - 500 Hz).

- Retests must be conducted with the very same equipment, test method and in the same background noise environment. In reality, a subject revealing a possible STS will be retested. It is fair to assume that the technician will try to reduce the number of positive STSs as much as possible by testing in an environment as quiet as possible. Further, by switching from the regular audiometer to the manual one typically used for retests, the difference in HTLs due to different absolute calibration and threshold determination technology might amount to more conservative thresholds than the lack of masking. Calibration differences can amount to a 6 dB discrepancy and still be in accordance with the ANSI standard! Different threshold determination technology could reduce thresholds with another 5 dB. (Bekesy vs. Hughson-Westlake technique).

- Your example further assumes that deterioration in hearing sensitivity due to noise exposure is almost equal among the three frequencies. This is not the case. Typically, 4 kHz is 3 to 5 times more sensitive than 2 kHz (ISO 1999(3)). In young, normal hearing populations with HTLs near or at absolute zero, possible STSs will be composed of threshold shifts amounting to more than 30 dB at 4 kHz. Later in life when the hearing loss is saturated at 4 kHz, then 2 kHz will be affected in a significant way. In your example, development of a NIPTS of 10 dB at 1 kHz, would in most cases be accompanied by a 40 dB loss at 4 kHz and take about 30 to 40 years of exposure to 100 dBA noise to accomplish! When the threshold effect due to age is added to 4 kHz, another 20 dB resulting in a total loss of 60 dB at 4 kHz. It is therefore fairly safe to assume that individuals with NIHL can only experience masking at one frequency.

FOOTNOTE(3) International Organization for Standardization, ISO 1999:1990
(E).  Acoustics - Determination of occupational noise exposure and estimation
of noise induced hearing impairment.

- We believe that it is important to verify an STS as early in time as possible. Due to background masking of thresholds, imprecise base line audiograms will allow a substantial loss of hearing sensitivity before a loss can be verified as an STS. In your example, a total loss over 1 - 4 kHz can be found amounting to 30 dB. However, had the subject been tested in quiet conditions, the total loss would have amounted to 55 dB. That is almost two STSs. Due to the nature of the development of noise induced hearing loss and the fact that the maximum loss tend to saturate, there are only 3 - 4 possible STSs in a worker's workable noise exposure.

- In addition, your example does not fail to verify an STS when insert earphones are utilized, but TDH-earphones do falsely indicate an STS due to background masking. The subject's actual hearing sensitivity is still 10 dB HL at 1 kHz. To assume that the background noise will increase the HTL and indicate an STS, makes the validity of the STS questionable. In the example there is a total threshold shift amounting to 55 dB but unfortunately this cannot be verified due to a compromised baseline audiogram and not because insert earphones are utilized.

- Subsequently, the vast majority of the work force exposed to hazardous noise is more aided by early detection and intervention rather than if a possible, loosely founded STS can be verified. Finally, most artificially verified STSs will most certainly be negated in the retest procedures.

It will be most appreciated if these arguments could be the foundation for removing the last four paragraphs in your letter describing the conditions for utilization of insert earphones in industrial testing.


Fredrik Lindgren, Ph.D.
Director, Research and Development

CC: Mark Tremallo