Evaluating the Effect of Various Pure Tone Averages on Cochlear Implant Candidacy Screening.

Abstract

Objective: To provide evidence to use an extended frequency pure tone average to screen for cochlear implant evaluation candidates as recommended by the American Cochlear Implant Alliance. Additionally, to determine whether traditional low frequency, high or low frequency, high frequency, or extended frequency pure tone average most accurately predicts cochlear implant candidates based on speech perception scores from aided AzBio sentence testing or aided consonant-nucleus-consonant (CNC) testing.

Method: Adults from a tertiary care center who completed aided sentence testing during cochlear implant evaluation between 2014 and 2024 were assessed. Pure tone averages at low, high, low or high, and extended frequency ranges were evaluated with individual ear's aided AzBio or aided CNC scores as an outcome. Sensitivity, specificity, negative predictive value, and positive predictive value were assessed and compared using a paired study design.

Results: There were 675 observations from 363 patients with AzBio as an outcome variable and 665 observations from 364 patients with CNC scores as an outcome variable. High or low frequency achieved the highest sensitivity at 96.3% AzBio as an outcome and 95.6% with CNC as an outcome, which was significantly better than low frequency (p < 0.001). Low frequency achieved the best specificity at 63.6% with AzBio as an outcome 76.9% with CNC as an outcome.

Conclusion: Extended frequency pure tones resulted in greater sensitivity in predicting CI candidates compared with the traditionally used low frequency pure tone average. Screening with wider frequency values will lead to a greater number of patients being identified for a cochlear implant evaluation.

Level of evidence: Level 3 Laryngoscope, 2025.