ABSTRACT
Objectives
Portable, tablet‐based audiometry has been proposed as a diagnostic tool for patients unable to access traditional sound booth audiometry. However, the validity of tablet‐based audiometry for diagnostic use in pediatric patients has not been established. The objective of this study is to validate self‐administered tablet‐based diagnostic audiometry in a pediatric clinic setting.
Methods
Children 4 to 17.9 years‐old were recruited from a pediatric audiology clinic. Tablet audiometry was performed with tablet‐based software, SHOEBOX Audiometry (Ottawa, ON) at six frequencies (250–8000 Hz). Hearing thresholds from traditional pure tone audiometry in a sound booth were used as a gold‐standard comparison. Hearing loss was defined as ≥ 1 hearing threshold of > 20 dB.
Results
120 children underwent tablet‐audiometry, though not all were able to complete the test. In 948 individual frequencies tested, the mean threshold difference was −0.95 dB (95% CI –1.5 to −0.42 dB) between traditional and tablet audiogram. Most tablet‐assessed thresholds (92%, n = 873) were within 10 dB of corresponding traditional audiometry thresholds. Tablet audiometry demonstrated 92% sensitivity (95% CI 84%–97%) and 85% specificity (95% CI 77%–92%) for diagnosis of hearing loss. Test–retest reliability of tablet audiometry was high, with 93% (55/59) of test–retest hearing threshold differences falling within ±5 dB.
Conclusion
Tablet audiometry is valid and reliable for children aged 5–17 years who can successfully complete testing in a non‐soundproof clinic setting. It may serve as a tool for remote hearing testing, with potential applications in schools, clinics, and research.
Level of Evidence
2
Keywords: diagnostic audiometry, hearing loss, pediatric audiology, pure‐tone audiometry, tablet‐based audiometry
This study validated self‐administered tablet‐based audiometry in 120 children aged 4–17 years using traditional sound booth audiometry as the gold standard comparison. Tablet audiometry showed excellent agreement with traditional methods, with 92% of hearing thresholds falling within 10 dB of corresponding sound booth thresholds and a mean hearing threshold difference of only −0.95 dB between the two methods. Self‐administered tablet‐based audiometry proved valid and reliable for children aged 5–17 years in non‐soundproof clinic settings, offering potential for remote hearing testing in schools and clinics.
1. Introduction
Hearing loss (HL) affects approximately one in five children in the United States [1, 2]. In pediatric HL, timely diagnosis and intervention mitigate developmental consequences, such as language, cognitive, or social delays [1, 3, 4]. To promote early HL identification in school‐age children, various screening guidelines have been developed [5, 6, 7]. However, implementation of these guidelines is inconsistent, with no single federal mandate and school‐based programs required in only 34 states [8]. Even when HL is suspected, diagnosis relies on access to follow‐up audiometric testing.
Screening and diagnostic audiometry differ significantly in purpose. Screening tests identify individuals at risk (pass/fail outcomes), whereas diagnostic audiometry defines specific thresholds for clinical decision‐making. The gold standard to diagnose HL following a failed hearing screening in school‐age children is audiologist‐administered pure tone audiometry performed in a sound booth [1]. However, financial, geographic, and logistic limitations create barriers to access [9, 10]. Portable, tablet‐based audiometry aims to provide diagnostic tools to patients unable to undergo gold‐standard diagnostics [11].
Various mobile applications have been developed to provide software for tablet‐based audiometry, often coupled with headphone technology that attenuates ambient sound [12]. Several studies have validated tablet audiometry as a screening tool for adults and children [13, 14, 15] with other studies validating this software for HL diagnosis and monitoring in adults [16, 17, 18].
Some studies have evaluated self‐administered tablet audiometry for diagnostic use in children [19, 20, 21, 22, 23]. However, one study included only 16–17 year‐old adolescents [19], excluding younger children. Other studies included children as young as 4–5 years old, but results were combined with adult data, preventing conclusions about younger children [20, 21]. Another study included only five children [21]. Early versions of self‐administered tablet‐based audiometry have been validated in children, though exclusively in sound booth settings [22, 23].
The validity of self‐administered tablet audiometry for threshold‐based assessment in non‐soundproof environments, and its potential role beyond pass/fail screening, has not been established. Rigorous validation of these tools is essential to support their use beyond screening, including for HL monitoring and research applications. This study aims to validate self‐administered tablet‐based diagnostic pure tone audiometry compared to traditional sound booth audiometry in a pediatric clinic setting.
2. Methods
2.1. Study Design
For the purposes of this cross‐sectional study, “traditional” or “conventional” audiometry was defined as audiologist‐administered pure tone audiometry performed in a sound booth. Institutional review board approval was granted through the Washington University Human Research Protection Office. Informed parental consent was obtained. Tablet audiometry was self‐administered in a non‐soundproof clinic room and was performed immediately after traditional audiometry in most participants. This order preserved diagnostic integrity, as traditional audiometry represents the gold‐standard test used for clinical decision‐making and avoided fatigue or inattention affecting these clinical results. Although counterbalanced testing order is preferred when comparing methods, a fixed order was selected to support real‐world clinical workflows and to ensure that agreement was evaluated relative to clinically required audiograms. For test–retest reliability, a subset of participants (n = 10; 8%) who fully completed testing in their first ear underwent repeat tablet audiometry on that ear during the same session.
2.2. Inclusion and Exclusion Criteria
Participants were recruited from a pediatric audiology clinic between January 2024 and August 2024. Children aged 4–17.9 years‐old presenting for a conventional diagnostic audiogram were eligible. Children attending hearing aid clinic with stable hearing, defined as two unchanged audiograms over 2 years or with an audiogram within the past 3 months, were also eligible. Children unable to complete tablet audiometry due to cognitive disabilities were excluded.
2.3. Tablet Audiometry
Tablet audiometry utilized SHOEBOX Audiometry (SHOEBOX Ltd., Ottawa, ON) software with SHOEBOX‐calibrated RadioEar DD450 circumaural headphones for passive noise attenuation ranging from 15 dB at low frequencies to 46 dB at higher frequencies. The software monitored background noise and paused testing when overall ambient noise, not octave‐band monitoring, exceeded 25 dB SPL. Tones from 0 to 90 dB were emitted at frequencies of 250, 500, 1000, 2000, 4000, and 8000 Hz. Tablet audiometry was conducted in quiet clinic rooms located away from high traffic areas. Practical steps were taken to minimize ambient noise, including closing doors and limiting foot traffic. Thresholds were measured with the Hughson‐Westlake algorithm. The test facilitator (AE, LM) provided brief instructions on the two‐choice task (Figure S1). Touching an object on the screen triggered a sound played in the headphones. The participant sorted the object according to whether a sound was heard into the appropriate corner of the screen: sound or no sound. The process was repeated from 250 to 8000 Hz in the right ear then left ear. All tests were self‐administered. If a child's results were flagged as unreliable by the software, an alert notified the test facilitator who then reinforced the instructions and retested the frequency. If a participant repeatedly generated unreliable results or expressed the desire to end testing, testing was considered complete. Upon completion, an on‐screen audiogram was displayed (Figure S2). Facilitators were blinded to conventional pure‐tone audiometry results during tablet testing.
2.4. Diagnostic Sound Booth Audiometry
Traditional diagnostic audiometry results were extracted from the medical record after tablet audiometry.
2.5. Data Analysis
Assuming 80% sensitivity or specificity of the test, a sample size of 109 subjects would provide us with a margin of error of 7.5% around the estimate. Using a 10% incomplete/drop‐out rate, we estimated a sample of 120 subjects to provide narrow margins of error for all planned test validation statistics.
Mean differences between tablet and traditional hearing thresholds were calculated for each frequency. Additionally, the proportion of differences within ±5 and ±10 dB was calculated given that test–retest variability in pediatric patients may be as great as ±10 dB [24]. To improve generalizability, we analyzed all tablet hearing thresholds that had corresponding conventional audiogram thresholds, even if tablet audiograms were incomplete. Because thresholds < 10 dB were not tested during traditional audiometry in some young children (n = 15), corresponding tablet thresholds measured at < 10 dB in these children (24/62 tablet thresholds) were excluded from comparative analysis to avoid artificial inflation of differences between tablet and booth results. Though this exclusion may slightly overestimate tablet performance at the lowest thresholds, it affected only 2.4% of all 995 tablet thresholds obtained overall and was necessary to ensure valid comparisons based on audiometric accuracy rather than protocol differences. Masking was also occasionally used in traditional diagnostic audiometry. As masking was not utilized in tablet audiometry, any masked thresholds were excluded from analysis. Threshold agreement was assessed per frequency, treating each frequency as an independent comparison. Agreement between methods was assessed using intraclass correlation coefficients (ICCs) and visualized with Bland–Altman plots. The effect of age on agreement was examined using adjusted Pearson correlation. Threshold differences falling outside Bland–Altman limits of agreement (mean difference±1.96 × standard deviation of the differences) were classified as outliers.
Sensitivity and specificity of tablet audiometry for HL detection were calculated using traditional sound booth audiometry as the gold standard. HL was defined as any threshold > 20 dB, based on the American Academy of Pediatrics standard [4]. As HL can be asymmetric, each ear was evaluated individually. Only ears with complete tablet audiograms were included (150/230 ears tested) to ensure that HL classification was based on assessment across all frequencies. Eighty ears with incomplete tablet audiograms were excluded to avoid misclassification bias, as missing thresholds could lead to false‐negative results if the untested frequencies were those affected by HL. Cohen's kappa was calculated to evaluate diagnostic agreement.
In participants who underwent test–retest tablet audiometry of a single ear, mean initial versus repeat threshold differences and the proportion of differences within ±5 dB were calculated. ICC evaluated agreement between initial and repeat measurements.
3. Results
3.1. Participants
In total, 120 children were enrolled, with 49.2% (n = 59) being male and 50.8% (n = 61) being female. Most were white (n = 107, 89.2%) and non‐Hispanic (n = 113, 94.2%). Of the participants, 62.5% (n = 75) had HL in at least one ear. Median (range) age at testing was 10.2 (4.2–17.3) years‐old. A portion (12/120; 10%) of study participants were hearing aid clinic patients with stable HL or recent audiograms.
Of the 120 children tested, 60 completed the full tablet audiogram (all 6 frequencies in each ear), 20 completed the test in only one ear, and 30 children did not complete tablet audiometry in either ear. Test–retest assessment of a single ear was performed for 10 participants. Reasons for failure to complete tablet audiometry included inattention, unreliable test results, and desire to stop the test. Children who completed the full tablet audiogram were older (median age = 12.4 years) than children who completed testing in only one ear (median age = 7.62 years) and children who did not complete full testing in either ear (median age = 7.15 years). Detailed demographic information can be found in Table 1.
TABLE 1.
Characteristics of study participants.
| All participants (n = 120) | Both ears complete (n = 60) | One ear complete (n = 20) | No ears complete (n = 30) | Test–retest (n = 10) | |
|---|---|---|---|---|---|
| Median age in years (range) | 10.2 (4.2–17.3) | 12.4 (5.5–17.3) | 7.62 (5.3–15.8) | 7.15 (4.2–16.4) a | 13.4 (6.4–17.2) |
| Sex | |||||
| Male | 59 (49.2%) | 29 (48.3%) | 11 (55.0%) | 12 (40.0%) | 7 (70.0%) |
| Female | 61 (50.8%) | 31 (51.7%) | 9 (45.0%) | 18 (60.0%) | 3 (30.0%) |
| Ethnicity | |||||
| Non‐Hispanic | 113 (94.2%) | 58 (96.7%) | 17 (85.0%) | 29 (96.7%) | 9 (90.0%) |
| Hispanic | 3 (2.5%) | 1 (1.7%) | 1 (5.0%) | 0 (0%) | 1 (10.0%) |
| Unknown | 4 (3.3%) | 1 (1.7%) | 2 (10.0%) | 1 (3.3%) | 0 (0%) |
| Race | |||||
| White | 107 (89.2%) | 54 (90.0%) | 16 (80.0%) | 27 (90.0%) | 10 (100%) |
| Black | 7 (5.8%) | 5 (8.3%) | 1 (5.0%) | 1 (3.3%) | 0 (0%) |
| Asian | 2 (1.7%) | 0 (0%) | 1 (5.0%) | 1 (3.3%) | 0 (0%) |
| Unknown | 4 (3.3%) | 1 (1.7%) | 2 (10.0%) | 1 (3.3%) | 0 (0%) |
| Diagnosis | |||||
| Hearing loss | 75 (62.5%) | 35 (58.3%) | 14 (70.0%) | 22 (73.3%) | 4 (40.0%) |
| No hearing loss | 45 (37.5%) | 25 (41.7%) | 6 (30.0%) | 8 (26.7%) | 6 (60.0%) |
Of note, one 16‐year‐old was not able to complete either ear fully due to the test labeling results as unreliable.
3.2. Comparison of Tablet and Traditional Audiometry Hearing Thresholds
Of all hearing threshold differences, 76% fell within ±5 dB and 92% fell within ±10 dB, with Bland–Altman limits of agreement of −17.3 to 15.4 dB (Figure 1). The mean traditional minus tablet audiogram threshold difference was −0.95 dB with a 95% confidence interval (CI) of −1.5 to −0.42 dB (n = 948). There was strong agreement between tablet and traditional hearing thresholds, with ICC of 0.92 (95% CI 0.91–0.93). At individual frequencies, more than 70% of threshold differences were within ±5 dB, and more than 87% were within ±10 dB (Table 2; Figure S3). All mean differences and 95% CIs at individual frequencies fell within ±5 dB, apart from 4000 Hz in the right (−4.3 dB; 95% CI –6.1 to −2.5 dB) and left (−3.2 dB; 95% CI –5.2 to −1.2 dB) ears. ICCs between traditional and tablet audiometry were ≥ 0.89 at all frequencies, indicating high inter‐method agreement. Correlation between tablet and traditional audiogram hearing thresholds was strong (r = 0.859) and did not change significantly when adjusted for age (r = 0.861).
FIGURE 1.
Bland–Altman Plot of Sound Booth and Tablet Audiometry Hearing Threshold Differences. Average hearing threshold between sound booth and tablet audiogram is plotted against sound booth minus tablet audiogram threshold difference. Hearing threshold differences from all individual frequencies tested (250, 500, 1000, 2000, 4000, and 8000 Hz) are displayed (n = 948). The solid line represents the mean hearing threshold difference. Dotted lines represent the outer limits of the agreement interval, calculated as mean difference±1.96 × the standard deviation of the differences.
TABLE 2.
Agreement between tablet and sound booth audiometry hearing thresholds.
| Mean difference a (95% CI) | Differences within ±5 dB (%) | Differences within ±10 dB (%) | ICC b (95% CI) | |
|---|---|---|---|---|
| All frequencies (n = 948) | −0.95 dB (−1.5 to −0.42) | 718 (76%) | 873 (92%) | 0.92 (0.91–0.93) |
| Right ear | ||||
| 250 Hz (n = 85) | 0.71 dB (−0.88 to 2.3) | 60 (71%) | 79 (93%) | 0.92 (0.88–0.95) |
| 500 Hz (n = 83) | −0.66 dB (−2.4 to 1.1) | 68 (82%) | 77 (93%) | 0.91 (0.87–0.94) |
| 1000 Hz (n = 101) | −1.63 dB (−3.2 to −0.05) | 81 (80%) | 90 (89%) | 0.92 (0.88–0.95) |
| 2000 Hz (n = 96) | −0.88 dB (−2.9 to 1.1) | 75 (78%) | 88 (92%) | 0.90 (0.85–0.93) |
| 4000 Hz (n = 91) | −4.3 dB (−6.1 to −2.5) | 70 (77%) | 86 (95%) | 0.93 (0.86–0.96) |
| 8000 Hz (n = 85) | −1.2 dB (−3.3 to 0.66) | 57 (67%) | 78 (92%) | 0.92 (0.87–0.95) |
| Left ear | ||||
| 250 Hz (n = 66) | 2.3 dB (0.42 to 4.1) | 49 (74%) | 58 (88%) | 0.89 (0.82–0.94) |
| 500 Hz (n = 66) | 1.2 dB (−0.40 to 2.8) | 52 (79%) | 61 (92%) | 0.93 (0.89–0.96) |
| 1000 Hz (n = 74) | −0.20 dB (−2.2 to 1.8) | 62 (84%) | 71 (96%) | 0.91 (0.85–0.94) |
| 2000 Hz (n = 69) | −2.5 dB (−4.0 to −0.91) | 49 (71%) | 66 (96%) | 0.95 (0.90–0.97) |
| 4000 Hz (n = 65) | −3.2 dB (−5.2 to −1.2) | 49 (75%) | 59 (91%) | 0.94 (0.88–0.96) |
| 8000 Hz (n = 67) | 0.07 dB (−2.0 to 2.2) | 46 (69%) | 60 (90%) | 0.93 (0.89–0.96) |
Sound booth audiogram threshold minus tablet threshold.
Intraclass correlation coefficient.
Participants (n = 15) with outlier threshold differences falling outside of the Bland–Altman limits of agreement are described in Table S1. Nine out of the 15 outlier participants were unable to complete full tablet audiometry (7/15), had auditory processing disorder (2/15), and/or completed traditional and tablet audiometry on different days (2/15). Of the remaining outliers, 4 out of 5 had only one pure‐tone threshold difference ≥ 20 dB.
3.3. Sensitivity and Specificity for Detection of HL
In 150 ears (63 with HL) with complete tablet audiogram data available, tablet audiometry was 92% sensitive (95% CI 84%–97%) and 85% specific (95% CI 77%–92%) for detection of HL (Table 3). There was substantial agreement between traditional and tablet audiogram for HL diagnosis (Cohen's kappa = 0.76).
TABLE 3.
Sensitivity and specificity of tablet audiometry for diagnosis of hearing loss.
| N = 150 ears | Sound booth audiogram a | ||
|---|---|---|---|
| Positive | Negative | ||
| Tablet Audiogram a | Positive | 58 | 13 |
| Negative | 5 | 74 | |
|
Sensitivity b 92% (95% CI 84%–97%) |
Specificity b 85% (95% CI 77%–92%) |
||
A positive test for hearing loss was defined as any hearing threshold > 20 dB.
Sensitivity and specificity were calculated for tablet audiometry using diagnosis of hearing loss on sound booth audiogram as the gold standard.
Audiometry results for the 5 false negative and 13 false positive tablet audiograms were reviewed (Table S2). Out of the 5 false negatives, 4 had borderline normal hearing (maximum traditional audiometry hearing threshold of 25 dB). The other false negative had a single traditional audiometry hearing threshold for a mild HL (30 dB at 2000 Hz). Out of the 13 false positives, 7 came from children with borderline HL (maximum traditional audiometry hearing threshold of 20 dB) and 3 came from children with both borderline HL and diagnosed auditory processing disorder. In the remaining 2 false positives, only the 8000 Hz hearing threshold met criteria for HL on tablet audiogram.
3.4. Test–Retest Reliability
To assess test–retest reliability, 10 children underwent repeat testing on a single ear. Most test–retest hearing threshold differences were within ±5 dB (93%; n = 55) with high test–retest reliability (ICC = 0.97; 95% CI 0.95–0.98). The mean test–retest tablet threshold difference was 0.85 dB with a 95% CI of −0.13 to 1.8 dB (n = 59).
4. Discussion
Early diagnosis of HL in children is key to mitigate hearing‐related developmental consequences. However, access to conventional diagnostic pediatric audiology is often limited due to financial, geographic, and logistic limitations. Our objective was to validate SHOEBOX tablet‐based diagnostic pure tone audiometry in a non‐soundproof pediatric clinic setting. This study demonstrates that self‐administered tablet‐based audiometry correlates strongly with traditional pure tone audiometry for children aged 5–17 years in a non‐soundproof clinic environment. Results align with earlier pediatric studies but extend validation to a setting more representative of real‐world clinics and schools.
Although self‐administered diagnostic tablet audiometry has not been previously validated for children in non‐soundproof settings, early versions were validated in a sound booth [22, 23]. One prospective study of 70 children aged 3–13 years‐old demonstrated no difference in mean pure tone average between tablet‐based audiometry and audiologist‐administered play audiometry [23]. However, this study did not report individual frequency differences, omitted low (250 Hz) and high (8000 Hz) frequencies, and excluded children with poor reliability or incomplete tests [23]. In contrast, the present study included all available tablet‐sound booth threshold pairs from 250 to 8000 Hz, even from children unable to complete the full test in one sitting, improving generalizability to real‐world scenarios in which children with limited attention may utilize breaks to complete testing.
Wiseman et al. reported mean threshold differences < 10 dB between tablet‐based audiometry completed in a sound booth and conventional sound booth audiometry, with younger age linked to a larger margin of error [22]. However, definitive conclusions were limited by a small sample of 27 children, of which 4 were diagnosed with HL [22].
While these studies support tablet audiometry for HL diagnosis, their utilization of a sound booth for tablet testing severely limits generalizability. Because access to sound booths is a major barrier to HL diagnosis, validation of tablet audiometry outside soundproof environments is essential. Our results are thus applicable to non‐soundproof office environments found in clinic and school settings.
4.1. Applications
4.1.1. Remote Hearing Testing
Once validated, tablet‐based audiometry may be used as a threshold‐based assessment tool in pediatric clinic, research, and education settings. In pediatric clinics, it can serve as a quick and accessible method to triage school‐aged children for HL. Utilizing tablet audiometry may help standardize the administration of pure tone testing across clinics and providers by automating test delivery. Additionally, the lack of trained professionals to administer pure‐tone testing is a significant barrier to hearing screening in schools. Because tablet audiometry is automated, minimal training is required to administer the test, circumventing the need for trained staff and potentially increasing the efficiency of school hearing screenings [25].
4.1.2. Research Studies
Tablet‐based audiometry may allow researchers to gather consistent and reliable data across various locations away from a medical center, enabling studies to involve a broader, more diverse demographic. Currently, limited access to audiologic evaluation due to both distance and cost of testing poses barriers to enrollment in research studies. Tablet‐based audiometry may facilitate involvement of participants from remote areas or those of low socioeconomic status in research by eliminating these barriers [21, 26]. This tool may increase representation in pediatric hearing research study design and thus overall generalizability of results.
4.1.3. Routine Monitoring of Hearing
Tablet audiometry may supplement hearing monitoring for children with existing HL, or conditions/treatments predisposing them to HL (e.g., otitis media, chemotherapy, cranial radiation). Traditional follow‐up often requires separate otolaryngology and audiology visits and significant travel, which can create a burden for families. Tablet audiometry may have a role in selected pediatric settings where access to audiologic services is limited. Importantly, tablet‐based testing is not intended to replace care for children with known HL who require comprehensive hearing management through specialized audiologic expertise and equipment.
In this context, tablet audiometry may provide practical benefit by facilitating earlier identification of hearing changes that warrant audiologic evaluation. Enabling threshold‐based assessment outside of a sound booth may reduce delays in recognizing progression of HL or treatment‐related ototoxicity, thereby supporting timely referral and intervention. When used appropriately, this approach may complement established audiologic care pathways.
4.2. Limitations
While most participants completed tablet and traditional sound booth audiometry on the same day, 10% were hearing aid clinic patients tested on different days. Although we limited inclusion to hearing aid clinic patients with stable HL or a recent audiogram, fluctuations due to cerumen or other factors could have affected results. Young children ages 4–6 years‐old experienced significant fatigue during tablet audiometry, leading to incomplete results. While this limits the generalizability of our findings to this age group, certain components of our study design may have exacerbated fatigue. Because tablet audiograms occurred after conventional audiometry, children were relatively fatigued and inattentive following their initial audiogram. In a school or pediatric office setting in which children complete only tablet audiometry, young children may not fatigue as quickly. Furthermore, children who struggle with attention may be permitted testing breaks in real‐world settings. Time limitations due to the clinic schedule prevented us from offering this option. Nonetheless, further gamification of tablet audiometry could be investigated to improve feasibility of testing in young children.
Another limitation is that hearing thresholds were analyzed as independent datapoints not accounting for correlations between thresholds at different frequencies for the same child. Thus, the estimate of precision for threshold level agreement may be overstated. Future work may apply mixed‐effects models to account for within‐subject and within‐ear correlations. It is also possible that the high prevalence of HL in our sample could have inflated diagnostic performance and may not generalize to screening populations. Finally, test–retest reliability was assessed in a single session. Intersession test–retest reliability could be lower and should be evaluated in future studies.
Ambient noise represents an important limitation of tablet‐based audiometry performed outside of a sound booth. Although testing was conducted in quiet clinic rooms and automatically paused when overall ambient noise exceeded a certain decibel threshold, this approach does not account for octave‐band–specific noise and may be less protective at lower frequencies. As a result, low‐frequency thresholds may be more susceptible to masking in less controlled environments. These findings underscore the importance of careful environmental control when tablet audiometry is used for diagnostic purposes.
Tablet audiometry has limitations for initial HL diagnosis. Unlike pediatric audiologists who use behavioral strategies to keep children engaged, the automated tablet test stopped with background noise and prevents such intervention, likely contributing to incomplete tests. Children with auditory processing disorder also showed discrepancies between tablet and traditional audiogram thresholds, suggesting that tablet audiometry may not be accurate for this group. Additional studies may clarify which children are suitable for self‐administered diagnostic tablet‐based audiometry. Finally, tablet audiometry cannot perform tympanometry or aided testing. While advanced versions include bone conduction, masking, and speech audiometry capabilities, further validation is needed [27]. These limitations highlight the need for confirmatory diagnosis of HL via conventional testing at the time of initial diagnosis. However, tablet‐based pure tone diagnostic audiometry remains a promising adjunct to audiologic testing through remote pure tone testing, HL monitoring, and research applications as described above.
5. Conclusion
Self‐administered tablet‐based audiometry correlates strongly with conventional pure tone audiometry for children aged 5–17 years in a non‐soundproof clinic environment. Future studies may investigate potential applications including pure tone testing in school and clinic settings, assessment of hearing changes that prompt further audiologic evaluation, and use in research studies. Further validation is necessary before tablet‐based audiometry can stand alone to establish an initial diagnosis of HL.
Funding
Research reported in this publication was supported by the National Institute of Deafness and Other Communication Disorders (NIDCD) within the National Institutes of Health (NIH) through the “Development of Clinician/Researchers in Academic ENT” training grant, award number R25DC020706. Support was also received through the DeNardo Education & Research Foundation Grant for the MD5 Yearlong Research Program at WashU School of Medicine. The funding organizations had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the decision to submit the article for publication; or in the preparation, review, or approval of the article. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or of Washington University in St. Louis.
Conflicts of Interest
Dorina Kallogjeri, statistics editor for JAMA Otolaryngology—Head and Neck Surgery. The other authors declare no conflicts of interest.
Supporting information
Table S1: Description of outliers.
Table S2: False positives and false negatives.
Figure S1: Game interface.
Figure S2: Audiogram generated from tablet audiometry.
Figure S3: Bland–Altman plots for individual frequencies.
Acknowledgments
We thank Lauren Ragnetti, MS, Julie DuBois, MS, Susan Walsh, AuD, and Kelly Taylor, AuD for their support in patient recruitment and for allowing us to perform testing in their clinic.
Ensing A. E., Mueller L., Lee D. S., Kallogjeri D., and Lieu J. E. C., “Validation of Non‐Soundproof Tablet Audiometry in Pediatrics: Comparison With Traditional Audiometry,” The Laryngoscope 136, no. 9 (2026): 4056–4063, 10.1002/lary.70514.
The work was presented in part as a poster presentation at the Society for Ear, Nose, and Throat Advancement in Children 2024 Annual Meeting; November 7–9, 2024; Denver, Colorado, USA.
Data Availability Statement
Research data are not shared.
References
- 1. Lieu J. E. C., Kenna M., Anne S., and Davidson L., “Hearing Loss in Children: A Review,” JAMA 324, no. 21 (2020): 2195–2205, 10.1001/jama.2020.17647. [DOI] [PubMed] [Google Scholar]
- 2. Wang J., Sung V., Carew P., et al., “Prevalence of Childhood Hearing Loss and Secular Trends: A Systematic Review and Meta‐Analysis,” Academic Pediatrics 19, no. 5 (2019): 504–514, 10.1016/j.acap.2019.01.010. [DOI] [PubMed] [Google Scholar]
- 3. Ching T. Y. C., Dillon H., Button L., et al., “Age at Intervention for Permanent Hearing Loss and 5‐Year Language Outcomes,” Pediatrics 140, no. 3 (2017): e20164274, 10.1542/peds.2016-4274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Bower C., Reilly B. K., Richerson J., and Hecht J. L., “Hearing Assessment in Infants, Children, and Adolescents: Recommendations Beyond Neonatal Screening,” Pediatrics 152, no. 3 (2023): e2023063288, 10.1542/peds.2023-063288. [DOI] [PubMed] [Google Scholar]
- 5. American Academy of Audiology , Clinical Practice Guidelines: Childhood Hearing Screening (American Academy of Audiology, 2011), https://www.audiology.org/practice‐guideline/clinical‐practice‐guidelines‐childhood‐hearing‐screening/. [Google Scholar]
- 6. American Speech‐Language‐Hearing Association , Childhood Hearing Screening (American Speech‐Language‐Hearing Association, 2014), https://www.asha.org/practice‐portal/professional‐issues/childhood‐hearing‐screening/?srsltid=AfmBOorF2aczYwduILfo08wEWYLsDOoqYqjV5BtjqV9gxcoVQRZVnEG2. [Google Scholar]
- 7. Committee on Practice & Ambulatory Medicine , “2024 Recommendations for Preventive Pediatric Health Care: Policy Statement,” Pediatrics 154, no. 1 (2024): e2024067201, 10.1542/peds.2024-067201. [DOI] [PubMed] [Google Scholar]
- 8. American Speech‐Language‐Hearing Association , School Hearing Screening: State Laws and Guidelines (ASHA, 2024). [Google Scholar]
- 9. Bush M. L., Kaufman M. R., and McNulty B. N., “Disparities in Access to Pediatric Hearing Healthcare,” Current Opinion in Otolaryngology & Head and Neck Surgery 25, no. 5 (2017): 359–364, 10.1097/MOO.0000000000000388. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kingsbury S., Khvalabov N., Stirn J., et al., “Barriers to Equity in Pediatric Hearing Health Care: A Review of the Evidence,” Perspect ASHA Spec Interest Groups 7, no. 4 (2022): 1060–1071, 10.1044/2021_persp-21-00188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. McCaslin D., “JAAA Editorial: Rise of the Machines: Audiology and Mobile Devices,” Journal of the American Academy of Audiology 30, no. 3 (2019): 168–168, https://www.audiology.org/jaaa‐editorial‐rise‐of‐the‐machines‐audiology‐and‐mobile‐devices/. [DOI] [PubMed] [Google Scholar]
- 12. Oremule B., Abbas J., Saunders G., et al., “Mobile Audiometry for Hearing Threshold Assessment: A Systematic Review and Meta‐Analysis,” Clinical Otolaryngology 49, no. 1 (2024): 74–86, 10.1111/coa.14107. [DOI] [PubMed] [Google Scholar]
- 13. Saliba J., Al‐Reefi M., Carriere J. S., Verma N., Provencal C., and Rappaport J. M., “Accuracy of Mobile‐Based Audiometry in the Evaluation of Hearing Loss in Quiet and Noisy Environments,” Otolaryngology and Head and Neck Surgery 156, no. 4 (2017): 706–711, 10.1177/0194599816683663. [DOI] [PubMed] [Google Scholar]
- 14. Xiao L., Zou B., Gao L., et al., “A Novel Tablet‐Based Approach for Hearing Screening of the Pediatric Population, 516‐Patient Study,” Laryngoscope 130, no. 9 (2020): 2245–2251, 10.1002/lary.28329. [DOI] [PubMed] [Google Scholar]
- 15. Yeung J. C., Heley S., Beauregard Y., Champagne S., and Bromwich M. A., “Self‐Administered Hearing Loss Screening Using an Interactive, Tablet Play Audiometer With Ear Bud Headphones,” International Journal of Pediatric Otorhinolaryngology 79, no. 8 (2015): 1248–1252, 10.1016/j.ijporl.2015.05.021. [DOI] [PubMed] [Google Scholar]
- 16. Bastianelli M., Mark A. E., McAfee A., Schramm D., Lefrançois R., and Bromwich M., “Adult Validation of a Self‐Administered Tablet Audiometer,” Journal of Otolaryngology ‐ Head & Neck Surgery 48, no. 1 (2019): 59, 10.1186/s40463-019-0385-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Ehlert K., Heinze B., Graham M. A., and Swanepoel D., “Surveillance for Ototoxicity in Platinum‐Based Chemotherapy Using Mobile Health Audiometry With Extended High Frequencies,” Journal of Laryngology and Otology 137, no. 1 (2023): 61–67, 10.1017/S0022215122001281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Foulad A., Bui P., and Djalilian H., “Automated Audiometry Using Apple iOS‐Based Application Technology,” Otolaryngology and Head and Neck Surgery 149, no. 5 (2013): 700–706, 10.1177/0194599813501461. [DOI] [PubMed] [Google Scholar]
- 19. Bornman M., Swanepoel D. W., Jager L. B. D., and Eikelboom R. H., “Extended High‐Frequency Smartphone Audiometry: Validity and Reliability,” Journal of the American Academy of Audiology 30 (2020): 217–226, 10.3766/jaaa.17111. [DOI] [PubMed] [Google Scholar]
- 20. Thompson G. P., Sladen D. P., Borst B. J. H., and Still O. L., “Accuracy of a Tablet Audiometer for Measuring Behavioral Hearing Thresholds in a Clinical Population,” Otolaryngology and Head and Neck Surgery 153, no. 5 (2015): 838–842, 10.1177/0194599815593737. [DOI] [PubMed] [Google Scholar]
- 21. Yalamanchali S., Albert R. R., Staecker H., Nallani R., Naina P., and J Sykes K., “Evaluation of Portable Tablet‐Based Audiometry in a South Indian Population,” Indian Journal of Otolaryngology and Head & Neck Surgery 74, no. 3 (2022): 3592–3598, 10.1007/s12070-020-02094-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Wiseman K., Slotkin J., Spratford M., et al., “Validation of a Tablet‐Based Assessment of Auditory Sensitivity for Researchers,” Behavior Research Methods 55, no. 6 (2023): 2838–2852, 10.3758/s13428-022-01933-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Yeung J. C., Javidnia H., Heley S., Beauregard Y., Champagne S., and Bromwich M., “The New Age of Play Audiometry: Prospective Validation Testing of an iPad‐Based Play Audiometer,” Journal of Otolaryngology ‐ Head & Neck Surgery 42, no. 1 (2013): 21, 10.1186/1916-0216-42-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Beahan N., Kei J., Driscoll C., Charles B., and Khan A., “High‐Frequency Pure‐Tone Audiometry in Children: A Test–Retest Reliability Study Relative to Ototoxic Criteria,” Ear and Hearing 33, no. 1 (2012): 104, 10.1097/AUD.0b013e318228a77d. [DOI] [PubMed] [Google Scholar]
- 25. Perez Zarate R., Johnson L., Mogendi S., Hogue C., Johnson V., and Gazmararian J. A., “Barriers and Facilitators to School Based Health Centers: Pilot Data From Three Sites in Georgia,” Journal of School Health 90, no. 2 (2020): 107–118, 10.1111/josh.12856. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Levy D. A., Hill D. R., Bia F. J., and Feinn R. S., “Tablet‐Based Hearing Screening in Children Aged 5 to 17 in Rural Dominican Republic,” Otology & Neurotology 39, no. 7 (2018): 823–828, 10.1097/MAO.0000000000001857. [DOI] [PubMed] [Google Scholar]
- 27. SHOEBOX , SHOEBOX Consult Shoebox Ltd., accessed January 30, 2025, https://www.shoebox.md/products/shoebox‐consult/.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Description of outliers.
Table S2: False positives and false negatives.
Figure S1: Game interface.
Figure S2: Audiogram generated from tablet audiometry.
Figure S3: Bland–Altman plots for individual frequencies.
Data Availability Statement
Research data are not shared.