Introduction: Superior canal dehiscence (SCD), characterized by the absence of bony labyrinth overlying the superior semicircular canal, can be surgically plugged with a middle fossa craniotomy. CT imaging of the temporal bone constitutes an integral element in the diagnostic workup and surgical planning. This study explores the utility of a method of manually estimating SCD size on CT imaging in predicting surgical outcomes.

Methods: A cohort study of consecutive repairs of SCD at a tertiary institution between 2011 and 2022 was conducted. Measurements of Approximate Dehiscence Area (ADA) on temporal bone CT imaging were performed by trained raters blind to clinical information. Length was estimated on coronal view by counting the number of 0.625-mm slices for which clear otic capsule aperture was appreciated. Width was measured on axial view as the diameter of the canal lumen. ADA was computed as the product of length and width. Surgical outcomes, evaluated with symptom resolution and audiometric improvement following repair, were assessed in multivariable regression models with ADA as the primary predictor and relevant history factors as covariates.

Results: Among 311 cases included, mean ADA was 2.68 mm2, and 71% of cases achieved Overall Symptom Improvement (OSI, net resolution ≥ 1 symptom). Every 1 mm2 rise in ADA predicted on average 3dB greater narrowing of low-frequency air-bone gap (β -3.0, 95% CI [-5.4, -0.7]) but 50% lower odds of OSI (aOR 0.50 [0.32, 0.78]). A model incorporating ADA and several history factors predicted OSI with a sensitivity, specificity combination as high as 85%, 73% and a PPV, NPV combination as high as 85%, 82%. (area under ROC curve 0.84 [0.76, 0.93]).

Conclusions: Greater ADA predicts more pronounced audiometric improvement but poorer symptomatic response. ADA is an efficient and practical method of estimating SCD size and demonstrated clinical utility in accurately predicting surgical outcomes.