Teste de palavras no ruído: proposta de um instrumento para avaliação do reconhecimento de fala na rotina clínica

Abstract

This research aimed to propose an instrument to assess speech recognition in the presence of competitive noise, define its application strategy to be applied in the clinical routine, obtain evidence of criterion validity, present the reference values, and investigate the reliability of the instrument. Initially, the organization of the material that constituted the Word-with-Noise Test was performed. Subsequently, the application strategy of the instrument was defined, through the application of a pilot study. The criterion validity investigation and definition of reference values for the test was then carried out by evaluating 50 normal-hearing adult subjects and 12 subjects with hearing loss. Lastly, the reliability of the test was investigated using a test-retest strategy in a group of 45 normal-hearing adults. The proposed instrument consisted of lists of monosyllabic and disyllabic words and speech spectrum noise. The test application strategy was defined as the Speech Recognition Threshold with noise fixed at 55 dBHL. Regarding criterion validity, the instrument showed excellent ability to distinguish between normal-hearing subjects, without reports of difficulty in recognizing speech-in-noise and subjects with hearing loss, with complaints of difficulty in recognizing speech-in-noise, evidenced by extremely high Area Under the Curve values of 0.987 for the monosyllabic stimulus and 1,000 for the disyllabic stimulus. The cut-off points expressed as signal/noise ratio of 1.47 dB for the monosyllabic stimulus and -2.02 dB for the disyllabic stimulus were defined as reference values for the proposed instrument. There was a statistically significant difference between the test and retest performances. The intraclass correlation coefficients obtained were indicative of good reliability (r=0.759; p<0.001) for the monosyllabic stimulus and moderate reliability (r=0.631; p<0.001) for the disyllabic stimulus. It was concluded that it was possible to propose an instrument to assess speech-in-noise recognition, called the Word-with-Noise Test. This test presented satisfactory evidence of criterion validity, with established reference values, presenting satisfactory reliability and demonstrating itself to be fast and easy to apply and interpret the results, making it a useful tool to be employed in the audiological clinical routine.