Date of Award

August 2017

Degree Type


Degree Name

Doctor of Philosophy (PhD)


Communication Sciences and Disorders


Beth A. Prieve


Acoustic leaks, Hearing Screening, Intra-subject variability, Newborns, Test-retest, Wideband Acoustic Immittance

Subject Categories

Social and Behavioral Sciences


Purpose: Absorbance is an important wideband acoustic immittance (WAI) measure to convey the mechanical-acoustic characteristics of outer/middle ear over a wide range of frequencies. The purpose of this project was to develop preliminary clinical criteria based on wideband acoustic immittance (WAI) measures to determine absorbance changes due to goodness of probe fit, and to assess intra-subject factors that contribute to the large variability in absorbance measured in newborn ears. Quantifying the effect of intra-subject factors, such as test-retest, and understanding their contribution to absorbance variability could improve the diagnostic accuracy of power absorbance in newborn middle ear measurements. This would advance the use of WAI at the time of newborn hearing screening. The following objectives were assessed: 1) To quantify variability in absorbance related to poor probe fit, and to develop probe-fit criteria. 2) To evaluate test-retest variability in absorbance with probe-reinsertions in measurements with good probe fits. 3) To quantify within-ear differences in absorbance measurements between two probe systems that are acoustically characterized using different approaches, and to assess their role in the intra-subject factors.

Method: Multiple repeated measurements were recorded in each of 98 ears from 50 full-term newborns who passed the TEOAE hearing screening test. Repeated tests were obtained using multiple probe insertions, and two probe systems; the Interacoustics’ Titan, and Mimosa Acoustics’ HearID system. Three within-ear factors were quantified; probe-fit difference, probe-reinsertion difference and differences between probe systems (Titan-HearID). Correlation analyses of probe-fit differences with WAI measures allowed for derivation of criteria for WAI measures to predict whether an absorbance probe-fit difference was caused by a poor probe fit. The criteria were used to identify poor probe-fits in the data determined by within-ear changes in power absorbance. In measurements with good probe fits, probe reinsertion differences were statistically compared between instances when ear canal debris was removed by the probe tip between insertions and when no debris was removed. Also, probe-reinsertion differences were statistically compared the between the two probe systems. The standard deviations (SD) of probe-fit differences, probe-reinsertion differences, and probe-system differences conveyed the variability associated with these factors, and were assessed in relationship to the SD of absorbance as a measure of overall variability.

Results: Criteria for probe fit were derived from twelve strong and moderate correlations between probe-fit differences and WAI measures that reflected underlying influence of acoustic leaks/ poor probe fitting. Ultimately, impedance phase of > -0.11 cycles in the 500-1000 Hz range, and absorbance of > 0.58 in the 250-1000 Hz range were determined to best predict within-ear differences in absorbance related to acoustic leaks. Poor fits identified by the criteria resulted in significant within-ear increases in absorbance by 0.08 in the 1000-6000 Hz range, and up to 0.4 in the < 1000 Hz range. The SDs associated with the probe-fit differences were greatest and similar to absorbance SD in the low frequency range. In the measurements with good probe fits, median probe reinsertion differences were generally < 0.02 and differed significantly from 0 for frequencies 500 Hz- 5000 Hz. Probe-reinsertion differences were not affected by removal of ear canal debris or use of different probe systems. Their SDs were greatest in the low frequencies and were similar to absorbance SD. Titan-HearID difference were significantly greater than 0 for frequencies around 2000 Hz, and their standard deviations were greatest and at 5000 Hz, and similar to SDs with probe-reinsertion differences for frequencies < 500 Hz Hz.

Conclusions: The findings suggest that the derived clinical criteria to eliminate measurements with poor probe fit reduced variability in absorbance in the low frequencies. Another important finding was that leaky/poor fits resulted in elevated absorbance primarily in the low frequencies, and in the diagnostically important mid frequencies. Independent from probe-fit, the intra-subject factor of test-retest with probe reinsertion was small in magnitude, but was a contributing factor in overall absorbance variability in the low frequencies. Moreover, the use of probe systems that are acoustically characterized using different approaches resulted in differences in absorbance greater than test retest for frequencies around 2000 Hz. Knowledge of the influence of intra-subject factors on absorbance, and reducing variability by controlling probe-fit are important for improving the accuracy of clinical interpretation. This could improve the diagnostic efficiency of absorbance in detecting conductive pathway alterations in the first two days of life.


SURFACE provides description only. Full text may be available to ProQuest subscribers. Please ask your Librarian for assistance.

This document is currently not available here.