Asymmetric Hearing Loss as a Population Screening Signal for Vestibular Schwannoma: A Number-Needed-To-Scan Analysis...

Asymmetric sensorineural hearing loss (SNHL) is the principal audiometric trigger for magnetic resonance imaging (MRI) to exclude vestibular schwannoma (VS). Although multiple audiometric referral criteria exist, the population-level burden of applying them has not been quantified. To estimate how many adults in the general United States population would be flagged for MRI by each of the 5 recognized audiometric referral criteria per incident VS, and to place that burden in the context of established screening programs and of current clinically indicated imaging.

We pooled adult audiometry from the National Health and Nutrition Examination Survey (NHANES; 1999-2000 through 2017-March 2020; n = 14,293) and estimated the survey-weighted prevalence of 5 recognized MRI-referral criteria for asymmetric SNHL by age and sex. Age-specific VS incidence was obtained from a VS-specific Surveillance, Epidemiology, and End Results (SEER) extraction (site C72.4, histology ICD-O-3 9560, benign behavior; 12 registries, 2004-2022; n = 12,958). We calculated a Number-Needed-to-Scan (NNScan), defined as the number of adults flagged per incident VS per year.

Age-adjusted VS incidence ranged from 1.4 to 1.8 per 100,000 person-years (median 1.7; crude 1.8). The prevalence of a referral-positive audiogram increased with age (American Academy of Otolaryngology-Head and Neck Surgery [AAO-HNS] criterion, 2.2% at ages 20-29 vs. 8.9% at ages 60-69) and was consistently higher in men than women (6.2% vs. 3.9%; P < 0.001). NNScan was lowest among adults aged 70-79 years (≈1260 per incident VS with the AAO-HNS criterion) and ranged up to ≈54,200 in younger adults and with more sensitive criteria. For comparison, published diagnostic yields of MRI in symptomatic, clinically referred patients (1.6%-3.7%) correspond to ≈27-64 scans per VS, a 20-fold to 47-fold difference.

Audiometric asymmetry is far more common than VS, so testing unselected adults in order to select candidates for MRI is an inefficient use of imaging resources: even in the most favorable age group and with the most specific criterion, approximately 1260 MRI examinations would be required per incident tumor. These findings argue against population-level audiometric screening for VS, but they do not diminish the value of the same criteria in symptomatic, clinically selected patients, in whom the pre-test probability is one to 2 orders of magnitude higher.