Validation of an automatic otoadmittance middle ear analyzer.

Tympanometry and acoustic reflex measurements were performed with an automatic otoadmittance middle ear analyzer (MEA) on 40 otologically normal children and 172 children with a history of recurrent acute otitis media or otoscopic evidence of persistent otitis media with effusion (OME), or both. For children with OME, the measurements were taken within a one-hour period prior to myringotomy. Myringotomy findings were used to validate predictive schema aimed at determining admittance values associated with OME. Predictive accuracy was quantitifed with percent sensitivity and specificity values. The results show that the MEA demonstrated suitable diagnostic predictability, but this was significantly influenced by the schema chosen to interpret obtained admittance measures. The simple presence or absence of a tympanometric peak (regardless of gradient) yielded the highest sensitivity and specificity values, whereas the manufacturer's suggested approach, with or without the acoustic reflex, was not as successful. The implications of these results are discussed.