Measuring Auditory Fitness in Locomotive Engineers: Development and Validation of a Signal Detection Test

Abstract

Objectives: To perform their job safely and effectively, locomotive engineers are required to detect auditory warning signals in the noisy work setting of a train cabin. Based on audio recordings of Dutch train cabins, we have developed a task and job-specific test for assessing the engineer’s ability to detect the two acoustic warning signals for the Dutch situation. The aim of this study was to evaluate the reliability, agreement, and construct validity of this test. Design: Two experiments were performed. In the first experiment, reliability and agreement of the signal detection test were assessed. Normally hearing individuals (N = 12) completed a signal detection test twice in 12 driving conditions. In the second experiment, construct validity was assessed. We retrospectively identified locomotive engineers, suspected of being hearing impaired, who were referred to the Amsterdam UMC for an auditory fitness for job assessment. All included engineers (N = 83) performed the signal detection test in four driving conditions, underwent tone audiometry and two speech perception in noise tests, and rated the effort and concentration it takes to detect the auditory signals. Seven a priori formulated hypotheses were tested. Results: In the first experiment, sufficient reliability and agreement were found in nine driving conditions (ICC = 0.54–0.81; standard error of measurement = 1.15–1.92), poor reliability in two driving conditions (ICC < 0.50), and poor agreement in one driving condition (standard error of measurement = 2.67 dBA). In the second experiment, the results of the signal detection test correlated moderately with the pure-tone thresholds, speech reception threshold in fluctuating noise, and engineer’s subjective rating of effort and concentration, but not with the speech reception threshold in continuous noise. According to the hypotheses, poorer test scores were obtained by hearing aid users compared with nonhearing aid users. Conclusions: The signal detection test has sufficient reliability and agreement in all but three driving conditions. This study provides evidence supporting the construct validity of the signal detection test in locomotive engineers. The moderate associations with conventional hearing tests show that the conventional hearing tests did not cover the whole construct measured with the signal detection test. The results, therefore, underpin the importance of evaluating the ability to detect auditory warning signals separately from other hearing-critical job tasks.