School-based hearing screening and diagnosis using automated and mobile health technologies

Abstract

Any loss in hearing sensitivity constitutes a major and if unidentified, an insurmountable barrier to effective learning. Current school-entry hearing screening in South Africa is unavailable to majority of children due to several challenges. This study aimed to provide research-based recommendations for clinical practice of schoolbased hearing screening. More explicitly, this study investigated the effect of hearing screening intensity (loudness) levels and the use of an immediate rescreen in reducing the referral rate in resource-limited contexts like South Africa. Furthermore, it investigated the use of a low-cost mobile smartphone application for calibrated hearing screening in schools. It also investigated the feasibility of conducting an automated diagnostic hearing evaluation on children in a school environment immediately after a refer rescreen result is obtained. Lastly, this study reported on the prevalence and nature of hearing loss in a representative urban school-aged population. A within-subject study was conducted to investigate the use of referral criteria in terms of screening intensity level and the use of a rescreen. It was found that the referral rate is reduced to 6.7% from 17% when using 25 dB HL as opposed to 20 dB HL screening intensity. Furthermore, there was a significant difference between the referral rates obtained at 20 and 25 dB HL and 20 and 30 dB HL (McNemar, p<0.01), but no significant differences between the referral rates at intensity levels 25 and 30 dB HL. In addition, an immediate rescreen also reduced the overall referral rate by more than one-third resulting in less referrals needing follow-up intervention. The use of 25 dB HL as a screening intensity together with an immediate rescreen is recommended as it reduces the referral rate significantly and will limit the burden of a screening programme on the health care resources. 1070 school-aged children in grade 1 to 3 (mean age 8 ±1.1 years) were recruited from five public schools to investigate the feasibility of a smartphone mobile application, hearScreen . No statistically significant difference in performance was obtained, with smartphone screening demonstrating equivalent sensitivity (75.0%) and specificity (98.5%) to conventional screening audiometry. Whilst referral rates were lower with the smartphone screening (3.2 vs. 4.6%) it was not significantly different (p>0.05). Smartphone screening (hearScreen ) was 12.3% faster than conventional screening (p<0.001). Sixty-two (mean age 8 ±0.9 years) of the 1070 participants were recruited to investigate the validity and time-efficiency of automated diagnostic air- and boneconduction audiometry for children in a natural school environment. With regards to the feasibility of automated diagnostic audiometry, 85.7% of air-conduction thresholds and 44.6% of bone-conduction thresholds corresponded within the normal range (15 dB HL) for manual and automated diagnostic audiometry. Both manual and automated diagnostic audiometry air- and bone-conduction thresholds exceeded 15 dB HL in 9.9% and 34.0% of thresholds respectively. For these thresholds average absolute differences for air- and bone-conduction thresholds were 6.3 dB (8.3 SD) and 2.2 dB (3.6 SD) and they corresponded within 10 dB across frequencies in 87.7% and 100.0%, respectively. There was no significant difference between manual and automated air- and bone-conduction across frequencies for these thresholds. The overall prevalence rate of hearing loss across the 1070 participants was 2.2%, with hearing loss being more prevalent in the Caucasian group (5.1%). A total of 24 children (12 male, 12 female) were diagnosed with a hearing loss (54.3% conductive, 25.7% mixed and 20.0% sensorineural). Furthermore, otoscopic examinations revealed that 6.6% of the children had cerumen and 7.4% presented with type B tympanograms which is indicative of middle ear effusion. Mobile technology together with automated diagnostic audiometry may be a feasible solution to overcome many of the prohibitive challenges faced by school-based hearing screening programs. Furthermore, prevalence data suggests that many children (2.2%) are in need of some form of follow-up service, most for medical intervention and a smaller number for audiological intervention. These findings provide valuable baseline data for realistic planning and appropriate implementation of hearing health services to ensure school-based hearing screening is employed sustainably across South Africa.