RESUMO
The correspondence between distortion product otoacoustic emissions (DPOAE intensities) and pure tone threshold audiometrics (PTTA) was studied. In 140 human ears with normal hearing or cochlear hearing loss, PTTA and DPOAE were recorded at 70 dB SPL (L70dpgram). The results served to develop a mathematical model for predicting PTTA from DPOAE. It was easily concluded that several DPOAE measures are needed to obtain reliable predictions of PTTA. In our simple model, four PTTA variables (PTTA at the frequencies of 500 Hz and 1, 2 and 4 kHz) can be predicted from a six-point L70dpgram (resolution of 2 points per octave) with reasonable accuracy. This model predicts not only numerical values, but also the diagnosis of degree of hearing loss with a high sensitivity and specificity (around 90%). It is the first clinically useful model of objective audiometrics to be published. In contrast, the L75 dpgram performed poorly in predicting PTTA and hearing loss categories.