Correlation of Measured vs Averaged 3kHz Pure Tone Averages

Abstract

Objective: AAO-HNS reporting guidelines for hearing results recommend a pure tone average (PTA) using thresholds at 0.5, 1, 2, and 3 kHz. 3 kHz measurements, however, are not always included in routine audiometry. We determined the magnitude of PTA errors using 3 kHz thresholds interpolated from 2 and 4 kHz measurements.

Method: Retrospective analysis of audiogram data from an academic, tertiary referral center from April to August 2010. All patients (N = 1,204) in our clinic having audiograms for any reason were eligible for inclusion. Audiograms missing threshold data at any of the frequencies evaluated were excluded.

Results: The difference between calculated and measured 3 kHz was within 5 dB, 10dB, and 20dB in 89.5%, 68.9%, and 98.6% of audiograms respectively. When the 3kHz level varied more than 10 dB from actual, the error underestimated the degree of hearing loss by a 3:1 margin (7.9% vs 2.6%). The steeper the slope of the hearing loss, the greater the divergence of the interpolated value from actual. When factored into the PTA, however, measured verses interpolated thresholds at 3 kHz showed a difference within 5 dB in 98.6%, R = 0.996, with a mean difference of 0.6 dB (SD = 1.7dB).

Conclusion: While interpolation of 3 kHz introduces a degree of error in measurement of this specific frequency, the effect on PTA is minor. These data suggest that 3 kHz thresholds can be reliably interpolated using available 2 and 4 kHz thresholds when utilized in PTAs.