首页|期刊导航|中国听力语言康复科学杂志|基于婴幼儿有意义听觉整合量表建立并验证婴幼儿听损程度预测模型

基于婴幼儿有意义听觉整合量表建立并验证婴幼儿听损程度预测模型OA

Development and Validation of a Prediction Model for the Severity of Hearing Loss in Infant and Toddler Based on IT-MAIS

中文摘要英文摘要

目的 基于婴幼儿有意义听觉整合量表(infant-toddler meaningful auditory integration scale,IT-MAIS)联合客观听力学指标及围产期危险因素,建立并验证婴幼儿听力损失程度的多模态预测模型.方法 根据筛选标准纳入553例0~6月龄完成听力学诊断测试及IT-MAIS的婴幼儿.根据短纯音听性脑干反应(ABR)结果,根据听力损失程度划分为正常-轻度、中度和重度三组.分析影响听力损失不同程度的危险因素,采用单因素加多因素logistic回归筛选自变量组建预测模型并绘制列线图.采用受试者工作特征曲线评估模型的区分度;校准图验证模型校准度;决策曲线分析法探讨模型的临床效能.结果 筛选出7个预测因子构建模型形成列线图,包括高龄产妇(P=0.004)、儿童永久性听力障碍家族史(P=0.001)、存在或怀疑有与听力损失有关的综合征或遗传病(P<0.001)、畸变产物耳声发射(P<0.001)和IT-MAIS总分(P<0.001).经验证,模型区分中度及以上和重度-极重度听力损失的曲线下面积分别为0.904(95%CI:0.876~0.932)和0.954(95%CI:0.930~0.978);预测概率和实际概率之间存在较好的一致性,平均绝对值误差分别为0.014和0.011;该模型在一定范围内对于识别不同程度听力损失风险人群具有应用价值.结论 本研究建立的婴幼儿听力损失程度预测模型具有良好的区分度、预测能力和临床适用性,为早期识别听力损失患儿,促进早发现早干预提供了科学依据.

Objective To establish and validate a multimodal prediction model for the degree of hearing loss in infants and young children based on the Infant-Toddler Meaningful Auditory Integration Scale(IT‑MAIS)in combination with objective audiological indicators and perinatal risk factors.Methods According to the screening criteria,553 infants aged 0‑6 months who had completed audiological diagnostic tests and the IT‑MAIS were enrolled.Based on the results of tone‑burst auditory brainstem response(ABR),the subjects were divided into three groups according to the degree of hearing loss as the outcome variable:normal‑mild,moderate,and severe.Risk factors affecting the degree of hearing loss were analyzed.Univariate and multivariate logistic regression were used to screen independent variables,establish the prediction model,and construct a nomogram.The discrimination of the model was assessed using the receiver operating characteristic(ROC)curve;calibration was verified by calibration plots;and clinical utility was explored by decision curve analysis(DCA).Results Seven predictive factors were screened to construct the model and form a nomogram,including advanced maternal age(P=0.004),family history of permanent childhood hearing loss(P=0.001),presence or suspicion of a syndrome or genetic disorder associated with hearing loss(P<0.001),distortion product otoacoustic emissions(P<0.001),and total IT‑MAIS score(P<0.001).After validation,the area under the curve for distinguishing moderate and above hearing loss and severe‑profound hearing loss was 0.904(95%CI:0.876‑0.932)and 0.954(95%CI:0.930‑0.978),respectively.Notable consistency was observed between the predicted probabilities and the actual probabilities,with mean absolute errors of 0.014 and 0.011.Within a certain range,the model had practical value in identifying populations at risk for different degrees of hearing loss.Conclusion The prediction model for the degree of hearing loss in infants and young children established in this study demonstrates acceptable discrimination,predictive ability,and clinical applicability,providing a scientific basis for the early identification of children with hearing loss and for promoting early detection and early intervention.

尹慧琳;顾海玲;李刚;孟照莉

四川大学华西医院耳鼻咽喉头颈外科 成都 610041四川大学华西医院耳鼻咽喉头颈外科 成都 610041四川大学华西医院耳鼻咽喉头颈外科 成都 610041四川大学华西医院耳鼻咽喉头颈外科 成都 610041

医药卫生

婴幼儿有意义听觉整合量表听力损失婴幼儿预测模型

Infant-toddler meaningful auditory integration scaleHearing lossInfant and toddlerPrediction model

《中国听力语言康复科学杂志》 2026 (5)

484-491,8

四川省科学技术厅项目"2~6月龄中耳功能正常婴儿的宽频声导抗值特征初探"(2021YFS0220)

10.3969/j.issn.1672-4933.2026.05.007

评论