A Computerized Closed-Response Speech Test in the ENT Clinic

Abstract

Objective: 1) To assess the feasibility of a computer-based closed-response speech test (CRST) in the ENT clinic. 2) To compare a closed response speech test to standard speech audiometry.

Method: Thirty patients underwent otoscopy, tympanometry, pure-tone audiometry, speech audiometry and a closed-response speech test (CRST). For the CRST, patients were asked to choose, on a tactile computer screen, the words that were presented to their ear, amongst a choice of 4 words. Fifteen sets of 4 words were used.

Results: The CRST was well accepted by all patients, whatever their computer literacy, but took significantly longer (10 to 15 minutes per ear) than speech audiometry did. Pure transmission hearing losses showed a translation of the performance intensity function toward greater intensities, without change in slope, as in standard speech audiometry. Correlations between auditory thresholds and speech recognition thresholds were greater for the CRST (r = 0.70, P < .01) than for speech audiometry (r = 0.46, P < .05). Several cases of patients showing very similar speech audiometry performance intensity functions, but different CRST scores were identified.

Conclusion: Although the closed-set computerized test took substantially longer than speech audiometry, it was well accepted by patients. It was more sensitive to minor alterations of hearing than speech audiometry and shows promise as a longitudinal assessment tool of speech perception.