Pediatric Temporal Bone Fractures: A 10-Year Experience.

OBJECTIVE

The aim of the study was to compare the traditional and newer temporal bone fracture classification systems and their reliability in predicting serious outcomes of hearing loss and facial nerve (FN) injury.

METHODS

We queried the medical record database for hospital visits from 2002 to 2013 related to the search term temporal. A total of 1144 records were identified, and of these, 46 records with documented temporal bone fractures were reviewed for patient age, etiology and classification of the temporal bone fracture, FN examination, and hearing status. Of these records, radiology images were available for 38 patients and 40 fractures.

RESULTS

Thirty-eight patients with accessible radiologic studies, aged 10 months to 16 years, were identified as having 40 temporal bone fractures for which the otolaryngology service was consulted. Twenty fractures (50.0%) were classified as longitudinal, 5 (12.5%) as transverse, and 15 (37.5%) as mixed. Using the otic capsule sparing (OCS)/violating nomenclature, 32 (80.0%) of fractures were classified as OCS, 2 (5.0%) otic capsule violating (OCV), and 6 (15.0%) could not be classified using this system. The otic capsule was involved in 1 (5%) of the longitudinal fractures, none of the transverse fractures, and 1 (6.7%) of the mixed fractures. Sensorineural hearing loss was found in only 2 fractures (5.0%) and conductive hearing loss (CHL) in 6 fractures (15.0%). Two fractures (5.0%) had ipsilateral facial palsy but no visualized fracture through the course of the FN canal. Neither the longitudinal/transverse/mixed nor OCS/OCV classifications were predictors of sensorineural hearing loss (SNHL), CHL, or FN involvement by Fisher exact statistical analysis (for SNHL: P = 0.37 vs 0.16; for CHL: P = 0.71 vs 0.33; for FN: P = 0.62 vs 0.94, respectively).

CONCLUSIONS

In this large pediatric series, neither classification system of longitudinal/transverse/mixed nor OCS/OCV was predictive of SNHL, CHL, or FN palsy. A more robust database of audiologic results would be helpful in demonstrating this relationship.