Authors: Varsha Mysore Athreya¹, Malcolm Slaney¹, Matthew Fitzgerald¹
¹Stanford University
Background: For nearly 100 years, the pure-tone average (PTA) has been used to summarize hearing thresholds across individual patients and large audiometric datasets. While the PTA provides a useful summary statistic, there are well-known limitations. For example, PTA overlooks the audiometric configuration and excludes some frequencies. Together, these exclusions may limit the ability of PTA to fully characterize auditory function. Recent advances in data-driven profiling suggest that more refined classifications of hearing loss can better capture the natural variability in hearing thresholds than the PTA, thereby advancing precision audiology.
Methods: Here, we use a data-driven approach to identify the most common audiometric clusters seen in an academic medical center, and then characterize speech recognition abilities in quiet and noise for each cluster. We performed K-Means clustering using 8 frequencies (250-8000 Hz) on 70,557 patients with normal hearing sensitivity and sensorineural hearing loss. To evaluate suprathreshold auditory function across these data-driven subtypes, we analyzed speech-recognition quiet (Word Recognition in Quiet; WRQ) and in noise (QuickSIN).
Results: Our results revealed 10 unique audiometric configurations across the dataset. WRQ scores remained relatively robust across most clusters, consistent with prior findings that WRQ declines only with more severe hearing loss. In contrast, QuickSIN revealed deficits even among audiograms with mild hearing loss, with greater individual variability than observed for WRQ. Furthermore, listeners with comparable PTAs but differing audiometric configurations exhibited distinct speech perception performance, particularly in noise.
Conclusions: These results highlight the limitations of PTA and demonstrate the utility of incorporating audiometric shape into clustering frameworks to enhance the resolution of hearing loss classification, supporting precision audiology and individualized rehabilitation strategies.
