ID Number		1622
Study Type		Engineering & Physics
Model		Hearing Aid Compatability (catch all) - defining thresholds for acceptable audio interference due to mobile phone use
Details		In an early study by Levitt et al, volunteers (n = 42) wearing hearing aids tested for acceptable interference levels from mobile phone emissions. Study design included a non-RF transmitting mobile phone handset wired to deliver an audio track (A-phone) and an RF transmitting handset (I-phone) set a max power and placed at incremental distances in proximity of the hearing aid to increase or decrease incident field strength. An in-canal microphone recorded real-time audio delivered to the volunteer. The volunteers were asked to subjectively assess usability during the test. The authors report signal to noise (C/I) ratios defining 4 usability categories as follows: a) highly usable [28-32 dB C/I], b) minor limitations [20-24 dB C/I], c) major limitations [12-15 dB C/I], and d) not usable (less than ~10 dB). A nominal 22 dB value for C/I was selected in support of ANSI C63.19 hearing aid compatibility testing in response to the FCC R&O (2001). Victorian (of the Hearing Industries Association) wrote a review of the ANSI C63.19 process, including references to IEC 60118-13 and the EHIMA and Australian testing that monitors hearing aid compatibility / performance in the far field. Harkins and Kozma-Spytek additional studies using volunteers at an SHHH conference (n = 21) with usability as an endpoint. All data were generated with the intent to support ANSI C63.19 categories. A series of recent studies (e.g., J Am Acad Audiol 17:626639 2006; J Speech Lang Hear Res, Vol. 47, Pg. 1001 - 1011, 2004) report acceptable signal / noise level margins of ~7.5 dB in full time HA users vs 10-14 dB for part time HA users (less than the 20 dB margin in the current ANSI C63.19 standard). The lower acceptable S/N ratio in full time HA than part time users (matched for degree of hearing loss / hearing threshold) was interesting, especially as the acceptable noise level for part time users was most similar to that of normal hearing individuals (~14 dB). In a study by Caputa (2000), the authors placed a hearing aid near a GSM phone, either in free space or in the ear canal of a subject, and measured the acoustic interference and E/H fields in each position. H field was stronger in the ear canal than in free space, while E field was weaker in the ear canal than in free space. But acoustic interference was higher in the ear canal than in free space, corresponding to the stronger H field.
Findings		Not Applicable to Bioeffects
Status		Completed With Publication
Principal Investigator
Funding Agency		NIH, USA
Country		UNITED STATES
References		Mueller, HC et al. The Hearing J, (1990) 43:14-17 Facta, S et al. Radiat Prot Dosimetry, (2004) 111:349-353 Levitt, H et al. Sem Hearing, (2005) 26:87-98 Victorian, TA Hearing J, (2006) 51:53-60 Kozma-Spytek, L et al. J Rehabil Res Dev, (2005) 42:145-156 Levitt, H et al. Semin Hearing, (2001) 12:275-280 Srinivasan, S et al. EMC Report, Center for Wireless EMC, University of Oklahoma (Phase 2 report), (1998) :- Preves, D Semin Hearing, (2003) 24:43-62 Berger, S J Am Acad Audiol, (2001) 12:309-314 Ravn, G ATIS Incubator Solutions Prog #4 Report, (2003) :- Burwood, E National Acoustic Laboratories Ref. EB831, (1999) :- Caputa, K et al. IEEE Trans Microwave Theory Tech, (2000) 48:2148-2154 Delta Labs - Denmark, EHIMA GSM PROJECT FINAL REPORT, (1995) :- Fry, TL et al. 1999 Annual Meeting Award Competition & Biomedical Instrumentation & Technology , (1999) :- Hansen, MO et al. Scandanavian Audiology, (1996) 25:227-232 Julstrom, S et al. Submitted to the record in FCC proceeding WT Docket No. 06-203. , (2006) :- Le Strange, JR et al. National Acoustic Laboratories, NAL Report No. 131 , (1995) :- Ravindran, A et al. EMC Report, Center for Wireless EMC, University of Oklahoma (Phase 1 report), (1996) :- Schlegel, RE et al. EMC Report, Center for Wireless EMC, University of Oklahoma (Phase 3-B report), (1999) :- Schlegel, RE et al. 42nd annual Human Factors and Ergonomics Society, Chicago, IL , (1998) :- Skopec, M IEEE Trans Rehabil Eng, (1998) 6:235-239 Killion, MC The Hearing Review, (1997) 4:8-12 Yang, T et al. IEEE Antennas and Propagation., (2008) 50:51-65 Morrissey, JJ IEEE Spectrum (on line) - http://www.spectrum.ieee.org/jan09/7187 , (2009) Jan:12-(2 pages) Cohen, NL et al. Ann Otol Rhinol Laryngol , (1989) 98(Supp 142):8-11 Dorman, M et al. Ear Hear., (1991) 12:368-369 Lalwani, AK et al. Amer J Otol, (1998) 19:66-70 Sorri, M et al. Scand Audiol, (2001) 30(Suppl 52):54-56 Valimaa, T et al. Scand Audiol, (2001) 30:112-118 Cray , JW et al. Amer J Audiol, (2004) 13:200-212 Adams, JS et al. Ear Nose Throat J, (2004) 83:99-103 Tait, M et al. Otol Neurotolol, (2001) 22:47-52 Bischof, F et al. Laryngorhinootologie., (2008) 87:768-774 Mackersie, CL et al. J Am Acad Audiol, (2009) 20:109-118 Levitt, H Journal of the American Academy of Audiology., (2001) 12:322-326
Comments		The terms compatibility and usability in the 1988 HAC Act, the FCC 2003 R&O, the ANSI C63.19 standard, and present paper and others (Univ Oklahoma data) that support the C63.19 standard are ambiguous. Also in the present study, subjective evaluation of usability / intelligibility is undefined and may have introduced significant variability and bias. A set of endpoints involving task-based performance might be more appropriate. Solicitation of volunteers and performance of the study was unblinded and have also introduced significant bias (should be possible to perform the study using generic amplifiers, signal generators, and dipole antenna to keep the study blinded while also offering a more controlled exposure condition). Use of an A-phone for audio and an I-phone as an interference source did not control the variable orientation and incident angle of the I-phone that may have introduced inaccuracies with regard to the normal gain pattern of the antenna. Further, near-field effects at lower power may not have been reproduced using simple distance to decrease field strength. A controlled experimental setup as described above could overcome this limitation as well as drive higher power levels to achieve threshold values for performance degradation. The study also included volunteers with a wide range of ages, magnitude and type of hearing loss, and hearing aid types and models that may have skewed the data. It may not be possible to perform a comprehensive analysis of all combinations, but some control of variables needs to be implemented in the study design. Importantly, a 20-25 dB range was seen among respondents for the threshold of normal "usability". Further, 5 of 42 subjects in the study did not report any audible interference, and these subjects were apparently excluded from the data. In addition, ~14% were unable to achieve a not usable level of EMI and it is not clear how that data was handled. The use of a 90% value for C/I thresholds amplifies the effects of the inherent bias associated with this data. In the kozma-Spytek 2005 study, the data did not make sense - individuals got >80% words correct and suggested no or limited interference or annoyance yet rated phones "unusable".