Audiometric Pattern in Moderate and Severe Tympanic Membrane Retraction

Objective: To evaluate the audiometric pattern in moderate/severe retractions of the tympanic membrane and correlate it with the severity of the otoscopy findings. Study Design: Cross-sectional study. Setting: Tertiary hospital. Patients: Consecutive patients with moderate or severe tympanic membrane retraction in at least one ear (451 ears) between August 2000 and January 2019, and no surgical history or effusion (mean [standard deviation] age, 32.8 [20.2] yr; 54% female and 42.4% children). Intervention: Pure-tone audiometry. Main Outcome Measures: Air conduction (AC) and bone conduction thresholds, and air-bone gap (ABG) measured at the four-frequency pure-tone average. Results: The median in decibel hearing level (dB HL) (minimum–maximum) of the AC, BC, and ABG were 25 dB HL (0–120 dB HL), 10 dB HL (0–75 dB HL), and 12.5 dB HL (0–55 dB HL), respectively. Seventy-two percent of the ears had an ABG ≤ 20 dB HL. For severity of the retraction of pars flaccida (PF), the AC, bone conduction, and ABG were similar across groups, with a weak correlation. For the pars tensa (PT), there was a global difference in the medians of AC and ABG in terms of the degree of severity, with a moderate correlation. Retraction in PF and PT at the same time was observed in 6 4% of the ears. ABG median was lower in ears with PF retraction (6.25 dB HL) than PT retraction, isolated (15 dB HL) or not (13.75 dB HL; p < 0.05). Conclusion: The ABG pure-tone average median was higher when PT was involved. We found a significant correlation between the retraction severity and worsening of AC and ABG thresholds, only for PT.