Otitis media with effusion in children: Cross-frequency correlation in pure tone audiometry

Different guidelines are adopted in clinics and countries to assess pure tone hearing sensitivity in children with otitis media with effusion (OME). Some guidelines specify a broad range of audiometric frequencies that must be tested and from which average thresholds determined, while others leave test frequencies unspecified. For guidelines that suggest specific frequencies there are various pure tone frequencies and frequency ranges given. The present study investigated whether (1) a full range of audiometric frequencies is required to evaluate hearing loss caused by OME in children, or if neighboring frequencies provide essentially the same threshold information, and (2) if different combinations of test frequency pure tone averaging calculations may affect decision criteria for surgical treatment. In a retrospective cohort study, right and left ear air conduction pure tone threshold data were obtained, from 125 Hz to 8 kHz, for 96 children with OME aged 4 to 12 years. Paired t-tests, correlation tests (Pearson's r, Cronbach's alpha, intraclass correlation) and absolute differences were used to examine the relationships among pure tone audiometric (PTA) frequencies for all ears with hearing loss. 168 ears were found to have OME-related hearing loss. Only the 125 Hz-250 Hz comparison showed no statistically significant difference between neighboring thresholds. However, only the 4 kHz and 8 kHz comparison showed a clinically significant mean difference of ≥ 10 dB. When viewing individual differences, comparison between 250 Hz and 500 Hz, 125 Hz and 500 Hz, and 4 kHz and 8 kHz, showed a large number of ears with clinically significant differences between test frequencies. Comparisons among low frequency 3 PTA average (500 Hz, 1 kHz, 2 kHz), high frequency 3 PTA average (1 kHz, 2 kHz, 4 kHz), and 4 frequency PTA average (500 Hz, 1 kHz, 2 kHz, 4 kHz) showed no statistically significant differences, with very strong correlations for all comparisons. In addition, for all the combinations of PTA averages, no clinically significant differences were found for the various comparisons or among individual results. Clinically, testing hearing sensitivity in the 125 Hz to 8 kHz range is worthwhile in evaluating hearing sensitivity in children with OME due to large individual variability across audiometric frequencies. However, frequencies tested for criterion averages for surgical treatments of children with OME may be restricted to 3 frequency PTA averages, either an average of 500 Hz, 1 kHz, 2 kHz or an average of 1 kHz, 2 kHz, 4 kHz, as no clinically significant differences were found using these or a 4 frequency averaging technique. For research purposes, 250 Hz can proxy for hearing thresholds at 125 Hz; and the low frequency 3 PTA average, high frequency 3 PTA average and 4 frequency PTA average may be used interchangeably, as no statistically significant differences were found among these measures.