Validation of automated threshold audiometry : a systematic review and meta-analysis

Abstract

The need for hearing health care services across the world far outweighs the capacity to deliver these services with the present shortage of hearing health care personnel. Automated test procedures coupled with telemedicine may assist in extending services. Automated threshold audiometry has existed for many decades; however, there has been a lack of systematic evidence supporting its clinical use. The aim of this study was to systematically review the current body of peer-reviewed publications on the validity (test-retest reliability and accuracy) of automated threshold audiometry. A meta-analysis was thereafter conducted to combine and quantify the results of individual reports so that an overall assessment of validity based on existing evidence could be made for automated threshold audiometry. A systematic literature review and meta-analysis was conducted using peerreviewed publications. A multifaceted approach, covering several databases and employing different search strategies, was utilized to ensure comprehensive coverage and crosschecking of search findings. Publications were obtained using the following three databases: Medline, SCOPUS and PubMed, and by inspecting the reference list of relevant reports. Reports were selected based according to inclusion and an exclusion criterion, thereafter data extraction was conducted. Subsequently, the meta-analysis combined and quantified data to determine the validity of automated threshold audiometry. In total, 29 articles met the inclusion criteria. The outcomes from these studies indicated that two types of automated threshold testing procedures have been utilized, the ‘method of limits’ and ‘method of adjustments’. Reported findings suggest accurate and reliable thresholds when utilizing automated audiometry. Most of the reports included data on adult populations using air conduction testing, limited data on children, bone conduction testing and the effects of hearing status on automated threshold testing were however reported. The meta-analysis revealed that test-retest reliability for automated threshold audiometry was within typical testretest reliability for manual audiometry. Furthermore, the meta-analysis showed comparable overall average differences between manual and automated air conduction audiometry (0.4 dB, 6.1 SD) compared to test-retest differences for manual (1.3 dB, 6.1 SD) and automated (0.3 dB, 6.9 SD) air conduction audiometry. Overall, no significant differences (p>0.01; Summarized Data ANOVA) were obtained in any of the comparisons between test-retest reliability (manual and automated) and accuracy. Current evidence demonstrates that automated threshold audiometry can produce an accurate measure of hearing threshold. The differences between automated and manual audiometry fall within typical test-retest and inter-tester variability. Despite its long history however, validation is still limited for (i) automated bone conduction audiometry; (ii) automated audiometry in children and difficult-to-test populations and; (iii) automated audiometry with different types and degrees of hearing loss.