Accuracy of tympanometry cutoff points in children from Central-Western Brazil.

Abstract

Purpose: This study evaluated tympanometric cutoff points (tympanometric peak pressure and static compliance) for transient evoked otoacoustic emissions (TEOAEs) and determined their diagnostic accuracy.

Methods: This analytical study, based on a randomized clinical trial, included 84 children (6-42 months; median 24 months) from public childcare centers in Brazil. Associations between cutoff points and TEOAEs were verified by Fisher's exact test and Chi-square. Kappa assessed agreement between ears, and the correlation between tympanometric peak pressure and static compliance was analyzed by Spearman. Diagnostic accuracy was determined using ROC (Receiver Operating Characteristic) curves and confirmed with EpiDat 3.1.

Results: In the 84 children, 66 right ears and 73 left ears were examined; however, 42 ears were randomly selected from each side for the final analysis. A significant correlation was found between peak pressure and static compliance. ROC analysis identified -180 daPa as the most accurate cutoff (AUC = 0.781; 95% CI: 0.656-0.905; p < 0.001). EpiDat indicated a specificity of 93.9%, a sensitivity of 55.6%, and an overall accuracy of 85.7%. Static compliance <0.2 mL showed comparable performance (specificity = 98.5%, sensitivity = 50.0%, accuracy = 88.1%).

Conclusion: The cutoff point of -180 daPa combined with static compliance <0.2 mL improved screening accuracy and reduced unnecessary referrals. It provided the best balance between specificity, sensitivity, and accuracy, aligning more closely with TEOAEs results and supporting updating this tympanometric criterion in infant hearing screening.