WBA reference intervals: Development & application (Humes, 2023)
Purpose: U.S. national wideband absorbance (WBA) data for 17,446 ears included in the National Health and Nutrition Examination Surveys for 2015–2016 and 2017–2020 were analyzed to develop and apply normative reference intervals (RIs).
Method: Analyses used distribution-free medians and cumulative distribution functions (CDFs). Notable differences between medians were defined as those with non-overlapping 95% confidence intervals, and differences between CDFs were evaluated using Cohen’s h effect size. Strict inclusion criteria identified “healthy ears” with 1,240 ears meeting all the inclusion criteria for the reference group. RIs, WBA values corresponding to the 2.5th and 97.5th percentiles for the reference group, were established. The established RIs were then applied to the full unscreened data set to determine the prevalence of WBA values outside the RIs.
Results: WBA RIs were established for all 6- to 19-year-olds and for 20- to 69-year-olds separated into three groups: females, males, and non-Hispanic Asians. The differences among the CDFs underlying these RIs corresponded to small effect sizes. When a single RI, 0.40 < average WBA < 0.75, was applied to the full data set, about 6%–13% of ears fell outside the derived RIs. Logistic regression analyses found abnormal tympanometric results to be responsible for the extreme WBA values among the general population. Abnormal tympanometric results increased the odds of having WBA values outside the RI by ≥ 300%.
Conclusions: U.S. population data for healthy ears were used to establish Ris for WBA of about 0.40–0.75. About 6%–13% of Americans, 6–80+ years of age, had WBA values outside these RI limits.
Supplemental Material S1. WBA amplitudes at each 1/3-octave center frequency for the following percentiles: 2.5, 5, 50, 95, and 97.5 for each of several reference groups.
Humes, L. (2023). Development and application of a reference interval approach to wideband absorbance norms using U.S. population data for ages 6 to 80+ years. Journal of Speech, Language, and Hearing Research. Advance online publication. https://doi.org/10.1044/2023_JSLHR-23-00313