Ellis Gregory M, Bieber Rebecca, Davidson Alyssa, Sherlock LaGuinn, Spencer Michele, Brungart Douglas
Audiology and Speech Pathology Center, Walter Reed National Military Medical Center, Bethesda, Maryland, USA.
Po'okela Solutions, Limited liability company, Honolulu, Hawaii, USA.
Ear Hear. 2025;46(4):909-921. doi: 10.1097/AUD.0000000000001633. Epub 2025 Jan 31.
The objective of this project was to quantify the relative efficacy of a four-frequency pure-tone average in the better ear (PTA4), the Hearing subscale of the Tinnitus and Hearing Survey (THS-H), and a combination of the two in predicting speech-in-noise performance, hearing aid recommendation, and hearing aid use among United States service members (SMs).
A two-analysis retrospective study was performed. The first analysis examined the degree to which better-ear PTA4 alone, THS-H alone, and better-ear PTA4 in conjunction with THS-H predicted performance on a speech-in-noise test, the modified rhyme test. Three binomial mixed-effects models were fitted using better-ear PTA4 alone, THS-H alone, and both measures as primary predictors of interest. Age and sex were included as covariates in all models. The models were compared to one another using Chi-square goodness-of-fit tests and the best-fitting model was examined. Data from 5988 SMs were analyzed in the first analysis. The second analysis examined the degree to which better-ear PTA4 alone, THS-H alone, and better-ear PTA4 in conjunction with THS-H predicted two hearing aid-related outcomes: recommendation for hearing aids by a clinician and hearing aid use. Three receiver operating characteristic curves were fit for each question for better-ear PTA4 alone, THS-H alone, and better-ear PTA4 + THS-H. The area under the curve was bootstrapped to generate confidence intervals to compare the three measures. Data from 8001 SMs were analyzed in the second analysis.
In the first analysis, all three models explained more variance than chance; however, the better-ear PTA4 + THS-H model was a significantly better fit than either the better-ear PTA4 alone or the THS-H alone models. Significant main effects of better-ear PTA4 and THS-H indicated that proportion correct decreased as better-ear PTA4 and THS-H increased. A significant interaction was observed such that proportion correct decreased more rapidly if both better-ear PTA4 and THS-H were increasing in tandem. In the second analysis, better-ear PTA4 + THS-H showed good predictive discrimination of a prior hearing aid recommendation. For predicting hearing aid use, better-ear PTA4 was the only predictor with an area under the curve bootstrapped confidence interval that overlapped 0.50, indicating better-ear PTA4 alone is a chance predictor for hearing aid use. Both THS-H alone and better-ear PTA4 + THS-H predicted hearing aid use better than chance, but had poor discrimination overall.
Adding the THS-H to the better-ear PTA4 improves predictions of speech intelligibility in noise, has good predictive strength for hearing aid recommendations, and predicts hearing aid use better than chance. This study provides evidence for using surveys in conjunction with objective data when evaluating hearing ability and recommending interventions. These findings are relevant to discourse regarding self-administered hearing healthcare.
本项目的目的是量化较好耳的四频率纯音平均听阈（PTA4）、耳鸣与听力调查问卷的听力子量表（THS-H）以及二者结合在预测美国现役军人（SM）的噪声环境下言语表现、助听器推荐及助听器使用方面的相对效能。
进行了一项双分析回顾性研究。首次分析考察单独的较好耳PTA4、单独的THS-H以及较好耳PTA4与THS-H结合对噪声环境下言语测试（改良押韵测试）表现的预测程度。使用单独的较好耳PTA4、单独的THS-H以及这两项指标作为主要预测指标拟合了三个二项式混合效应模型。年龄和性别作为协变量纳入所有模型。使用卡方拟合优度检验对模型进行相互比较，并检验最佳拟合模型。首次分析中对5988名现役军人的数据进行了分析。第二次分析考察单独的较好耳PTA4、单独的THS-H以及较好耳PTA4与THS-H结合对两个与助听器相关结果的预测程度：临床医生对助听器的推荐以及助听器使用情况。针对单独的较好耳PTA4、单独的THS-H以及较好耳PTA4 + THS-H，为每个问题拟合三条受试者工作特征曲线。通过自抽样生成曲线下面积的置信区间以比较这三项指标。第二次分析中对8001名现役军人的数据进行了分析。
在首次分析中，所有三个模型解释的方差均高于随机水平；然而，较好耳PTA4 + THS-H模型的拟合度显著优于单独的较好耳PTA4模型或单独的THS-H模型。较好耳PTA4和THS-H的显著主效应表明，随着较好耳PTA4和THS-H升高，正确比例下降。观察到显著的交互作用，即如果较好耳PTA4和THS-H同时升高，正确比例下降得更快。在第二次分析中，较好耳PTA4 + THS-H对既往助听器推荐具有良好的预测判别力。对于预测助听器使用情况，较好耳PTA4是唯一曲线下面积自抽样置信区间与0.50重叠的预测指标，表明单独的较好耳PTA4是助听器使用情况的随机预测指标。单独的THS-H以及较好耳PTA4 + THS-H对助听器使用情况的预测均优于随机水平，但总体判别力较差。
将THS-H添加到较好耳PTA4中可改善对噪声环境下言语可懂度的预测，对助听器推荐具有良好的预测强度，且对助听器使用情况的预测优于随机水平。本研究为在评估听力能力和推荐干预措施时结合使用调查问卷与客观数据提供了证据。这些发现与关于自我管理听力保健的讨论相关。
