Effects of consecutive trials and test-retest reliability of 1000-Hz tympanometry in adults

Objective: To investigate the effect of consecutive 1000-Hz tympanometry testing on admittance measures, and examine test-retest reliability. Design: Repeated measures with eight trials of 1000-Hz Y-a and B-a/G(a) tympanometry, respectively, in two ears of each subject, followed by repeated 226-Hz tympanometry. Study sample: Twenty-seven normal-hearing young adults. Results: For single-peak tympanograms, peak Y-tm and G(tm) systematically increased across trials with a mean change of 8% and 15% at Trial 8, respectively, whereas B-tm increased marginally. For notched tympanograms, peak Y-tm and B-tm decreased by 23% and 162% at Trial 8, and G(tm) in two cases also decreased (10% on average). Trial 2 and 3 contributed 50% to 70% of the total changes. Test-retest differences of subsequently acquired 1000- and 226-Hz tympanograms were smaller than previously reported. Conclusions: Consecutive testing significantly alters middle-ear admittance in 1000-Hz tympanometry. The outcome is contingent on tympanogram pattern and admittance component: Increase of peak Y-tm and G(tm) in single-peak tympanograms and decrease of all measures in notched tympanograms. The present results complement previous studies on our understanding of the mechanism underlying this effect: a decrease of middle-ear stiffness. The effect of repetitive tympanometry should be accounted for in research involving sequential testing.