This study seeks to define the most appropriate guidelines for selection of patients for magnetic resonance imaging (MRI) to exclude a vestibular schwannoma. Improved selection may reduce patient anxiety and improve resource utilization.

All MRIs of the internal auditory meatus, performed during the year 2000, were reviewed. Audiograms and symptoms were collated for all ’positive’ scans and 100 negative scans. Information was analysed using seven published protocols and other defined frequency specific criteria.

A diagnosis of vestibular schwannoma was made on 36 scans. Four criteria had a sensitivity of >95 per cent; of these the highest specificity (49 per cent) utilized an interaural difference at two adjacent frequencies of 15.dB in unilateral hearing loss and 20.dB in bilateral asymmetric loss.

Applying our best protocol would have reduced the number of scans performed from 392 to 168. The one patient with a vestibular schwannoma who was excluded had trigeminal paraesthesia, an independent indication for investigation.