Diagnostic Accuracy of Mobile Health-Based Audiometry for the Screening of Hearing Loss in Adults: A Systematic Review and Meta-Analysis

A Multicenter Validity Study of Four Smartphone Hearing Test Apps in Optimized and Home Environments

OBJECTIVE

Pure tone audiometry (PTA) is the gold standard for hearing assessment. However, it requires access to specialized equipment. Smartphone audiometry applications (apps) have been developed to perform automated threshold audiometry and could allow patients to perform self-administered screening or monitoring. This study aimed to assess the validity and feasibility of patients using apps to self-assess hearing thresholds at home, with comparison to PTA.

METHODS

A multi-center, prospective randomized study was conducted amongst patients undergoing PTA in clinics. Participants were randomly allocated to one of four publicly-available apps designed to measure pure tone thresholds. Participants used an app once in optimal sound-treated conditions and a further three times at home. Ear-specific frequency-specific thresholds and pure tone average were compared using Pearson correlation coefficient. The percentage of app hearing tests with results within ±10 dB of PTA was calculated. Patient acceptability was assessed via an online survey.

RESULTS

One hundred thirty-nine participants submitted data. The results of two at-home automated smartphone apps correlated strongly/very strongly with PTA average and their frequency-specific median was within ±10 dB accuracy. Smartphone audiometry performed in sound-treated and home conditions were very strongly correlated. The apps were rated as easy/very easy to use by 90% of participants and 90% would be happy/very happy to use an app to monitor their hearing.

CONCLUSION

Judicious use of self-performed smartphone audiometry was both valid and feasible for two of four apps. It could provide frequency-specific threshold estimates at home, potentially allowing assessments of patients remotely or monitoring of fluctuating hearing loss.

LEVEL OF EVIDENCE

2 Laryngoscope, 134:2864-2870, 2024.

A machine learning screening model for identifying the risk of high-frequency hearing impairment in a general population

BACKGROUND

Hearing impairment (HI) has become a major public health issue in China. Currently, due to the limitations of primary health care, the gold standard for HI diagnosis (pure-tone hearing test) is not suitable for large-scale use in community settings. Therefore, the purpose of this study was to develop a cost-effective HI screening model for the general population using machine learning (ML) methods and data gathered from community-based scenarios, aiming to help improve the hearing-related health outcomes of community residents.

METHODS

This study recruited 3371 community residents from 7 health centres in Zhejiang, China. Sixty-eight indicators derived from questionnaire surveys and routine haematological tests were delivered and used for modelling. Seven commonly used ML models (the naive Bayes (NB), K-nearest neighbours (KNN), support vector machine (SVM), random forest (RF), eXtreme Gradient Boosting (XGBoost), boosting, and least absolute shrinkage and selection operator (LASSO regression)) were adopted and compared to develop the final high-frequency hearing impairment (HFHI) screening model for community residents. The model was constructed with a nomogram to obtain the risk score of the probability of individuals suffering from HFHI. According to the risk score, the population was divided into three risk stratifications (low, medium and high) and the risk factor characteristics of each dimension under different risk stratifications were identified.

RESULTS

Among all the algorithms used, the LASSO-based model achieved the best performance on the validation set by attaining an area under the curve (AUC) of 0.868 (95% confidence interval (CI): 0.847-0.889) and reaching precision, specificity and F-score values all greater than 80%. Five demographic indicators, 7 disease-related features, 5 behavioural factors, 2 environmental exposures, 2 hearing cognitive factors, and 13 blood test indicators were identified in the final screening model. A total of 91.42% (1235/1129) of the subjects in the high-risk group were confirmed to have HI by audiometry, which was 3.99 times greater than that in the low-risk group (22.91%, 301/1314). The high-risk population was mainly characterized as older, low-income and low-educated males, especially those with multiple chronic conditions, noise exposure, poor lifestyle, abnormal blood indices (e.g., red cell distribution width (RDW) and platelet distribution width (PDW)) and liver function indicators (e.g., triglyceride (TG), indirect bilirubin (IBIL), aspartate aminotransferase (AST) and low-density lipoprotein (LDL)). An HFHI nomogram was further generated to improve the operability of the screening model for community applications.

CONCLUSIONS

The HFHI risk screening model developed based on ML algorithms can more accurately identify residents with HFHI by categorizing them into the high-risk groups, which can further help to identify modifiable and immutable risk factors for residents at high risk of HI and promote their personalized HI prevention or intervention.

Assessing the accuracy and reliability of application-based audiometry for hearing evaluation

Pure-tone audiometry (PTA) is the gold standard for assessing hearing loss. However, traditional PTA tests require specialized equipment, trained personnel, and a soundproof environment. Recently, smartphone-based PTA tests have been developed as an alternative method for hearing assessment. The aim of this study was to validate the accuracy and reliability of a smartphone application-based audiometry test. This study was conducted to assess the performance of application-based audiometry from November 2021 to January 2022. Pure-tone thresholds were measured using a smartphone application-based PTA test and compared with results obtained using a traditional audiometer in a sound-treated booth. The smartphone application used in this study was the "Care4Ear (Care4ear, version 1.0.6, MIJ Co., Ltd.)". Hearing thresholds less than 35 dB HL were classified as group A, 35-64 dB HL as group B, and 65 dB HL or greater as group C for the classification of hearing levels. We evaluated the accuracy of smartphone audiometry for each group and compared the results of frequency-specific hearing tests. Additionally, we examined the results of smartphone audiometry in individuals (n = 27) with asymmetric hearing loss. Seventy subjects completed both conventional audiometry and smartphone application-based hearing tests. Among the ears assessed, 55.7% were classified as group A, while 25.7% and 18.6% were classified as group B and group C, respectively. The average hearing threshold obtained from conventional pure-tone audiometry was 37.7 ± 25.2 dB HL, whereas the application-based hearing test yielded thresholds of 21.0 ± 23.0 dB HL. A significant correlation (r = 0.69, p < 0.01) was found between the average hearing thresholds obtained from the application-based and conventional pure-tone audiometry tests. The application-based test achieved a 97.4% hit rate for classifying hearing thresholds as class A, but lower rates of 22.2% for class B and 38.5% for class C. Notably, a discrepancy was observed between the hearing threshold measured by the application and the conventional audiometry for the worse ear with asymmetric hearing. The smartphone-based audiometry is a feasible method for hearing evaluation especially in persons with normal hearing. In cases of hearing loss or asymmetric hearing loss, the results of the application-based audiometry may be inaccurate, limiting its diagnostic utility.