a Division of Speech and Hearing Sciences, Faculty of Education , The University of Hong Kong , Hong Kong , China.
b Department of Otorhinolaryngology , Shenzhen Children's Hospital , Shenzhen , China.
Disabil Rehabil. 2018 May;40(10):1166-1175. doi: 10.1080/09638288.2017.1290698. Epub 2017 Feb 25.
Otitis media with effusion (OME) is a common middle ear disease in children. The associated conductive hearing loss is a major concern for hearing health professionals. The aim of the present study was to describe the configuration of pure tone audiograms of children with OME and to design a statistical stratification algorithm to facilitate hearing loss profiling in children with OME.
School age children with OME were recruited. Bone and air conduction thresholds were obtained using standard procedures. Hierarchical cluster analysis was employed to determine audiometric profile groups. The Mandarin Hearing in Noise Test was used to measure sentence perception in children for cluster analysis validity assessment.
Ninety-seven children (164 ears) aged between 72 months and 153 months were examined. Air conduction thresholds averaged for 500 Hz, 1000 Hz and 2000 Hz were in the range of 8.3-53.3 dB HL with a mean of 26.8 dB HL. Bone conduction thresholds were found to be influenced by middle ear pathology with a maximal elevation at 2000 Hz of 25 dB HL. Four audiometric profiles were identified. Cluster 1 contained 54 ears (32.9%) with normal or near normal hearing, Clusters 2 contained 37 ears (22.6%) with mild hearing loss, Cluster 3 included 48 ears (29.3%) and Cluster 4 included 25 ears (15.2%) with moderate hearing loss. Stability and validity of the four-cluster profiling procedure was examined and established with satisfactory results.
OME in children is associated with pure tone hearing thresholds ranging from normal to moderate hearing loss. The hierarchical clustering algorithm proved useful as a novel means of profiling hearing loss in children with OME and may assist in identifying affected children at greater risk of auditory disadvantage. Implications for rehabilitation A hierarchical cluster analysis method can be used to determine audiometric profiles in children with OME. This algorithm assists to identify children at greater risk of auditory disadvantage. Cluster groups with more elevated pure tone thresholds may be targeted for priority in clinical surveillance and medical/surgical intervention.
分泌性中耳炎(OME)是儿童常见的中耳疾病。相关的传导性听力损失是听力保健专业人员关注的主要问题。本研究的目的是描述 OME 儿童纯音听力图的形态,并设计一种统计分层算法,以方便对 OME 儿童的听力损失进行分析。
招募患有 OME 的学龄儿童。使用标准程序获得骨导和气导阈值。采用层次聚类分析确定听力图形态组。使用普通话言语测听评估儿童在噪声环境下的句子感知能力,以评估聚类分析的有效性。
共检查了 97 名(164 耳)年龄在 72 个月至 153 个月之间的儿童。500 Hz、1000 Hz 和 2000 Hz 的气导阈值平均值在 8.3-53.3 dB HL 之间,平均为 26.8 dB HL。骨导阈值受中耳病理影响,在 2000 Hz 处最大升高 25 dB HL。确定了 4 种听力图形态。第 1 组包含 54 只(32.9%)听力正常或接近正常的耳朵,第 2 组包含 37 只(22.6%)轻度听力损失的耳朵,第 3 组包含 48 只(29.3%)和第 4 组包含 25 只(15.2%)中度听力损失的耳朵。通过检查和评估,验证了四组分层程序的稳定性和有效性。
儿童 OME 与从正常到中度听力损失的纯音听力阈值相关。层次聚类算法被证明是一种有用的方法,可用于分析 OME 儿童的听力损失模式,有助于识别听力受损风险较高的儿童。
层次聚类分析方法可用于确定 OME 儿童的听力图形态。该算法有助于识别听力受损风险较高的儿童。纯音阈值较高的聚类组可能需要优先进行临床监测和医疗/手术干预。