Abstract
Distortion-product oto-acoustic emissions (DPOEs) were measured in 102 normal-hearing and hearing-impaired ears to examine the sensibility, specificity, positive predictive value and negative predictive value of this test in a real clinical setting using two decision support systems. The first model only considered the presence or the absence of DPOEs in response to a 52 dB SPL stimulation intensity. The second model used two consecutive tests: i) the presence or the absence of DPOEs in response to a 52 dB SPL stimulation intensity, and ii) evaluation of the slope of each DPOE input-output function. The sensitivity of these two models varied between 1 and 0.83 whereas the specificity varied between 0.95 and 0.72 as a function of the frequency tested. The first model proved to be a very sensitive test, but the introduction of a second criterion of decision (i.e., the slope of DPOE input-output function) did not enhance the validity of the test.