宽频鼓室压测量法联合Kempthorne标准在成人分泌性中耳炎诊断中的应用：一项前瞻性研究

文献信息

中文摘要

引言： 描述成人分泌性中耳炎（OME）的临床和听力学特征，并评估宽频鼓室压测量法（WBT）联合Kempthorne耳镜特征在成人OME中的诊断价值。
方法： 在Can Tho医药大学医院于2024年6月至2025年10月期间，对66只临床疑似OME、年龄≥16岁患者的耳进行了一项前瞻性描述性研究并实施手术干预。Kempthorne评分>5仅用作入组阈值。进行了临床检查、硬性耳镜检查、纯音测听、常规226 Hz鼓室压测量法和WBT。31只手术耳中，鼓膜切开术时的术中中耳积液作为阳性参考标准；其余35只非手术耳被归类为疾病阴性，但未经手术验证（部分验证的复合参考标准）。事先样本量计算目标为≥28只手术验证耳（α=0.05，把握度80%）；AUC比较的事后把握度为84.6%。评估了三种诊断策略：（1）标准226 Hz鼓室压测量法（B型或C型）；（2）单独WBT；（3）联合模型——WBT阳性加≥1个高收益Kempthorne形态学特征（气液平面和/或活动度消失），排除入组阈值。
结果： 所有入组耳的Kempthorne评分均>5。针对66耳复合参考数据集（31耳液体阳性，35耳假定液体阴性），标准鼓室压测量法的敏感性/特异性/AUC为100.0%/68.2%/0.84，单独WBT为77.4%/81.8%/0.91，联合模型为93.5%/86.4%/0.94（McNemar p=0.042）。在手术验证耳中，标准226 Hz鼓室压测量法检测出所有液体阳性病例（敏感性100.0%）。手术耳和非手术耳在关键术前指标上无显著差异（所有p>0.05）。
结论： WBT联合特定Kempthorne形态学特征在成人OME中提供了互补的形态学和吸光度信息。然而，由于标准鼓室压测量法在手术亚组中已达到100%敏感性，联合模型在特异性、NPV和AUC方面的任何表面优势主要来源于未经手术验证的参考阴性亚组，鉴于严重的部分验证偏倚，必须谨慎解读。

英文摘要

INTRODUCTION: To describe clinical and audiological characteristics and evaluate the diagnostic value of wideband tympanometry (WBT) combined with Kempthorne otoscopic features in adults with otitis media with effusion (OME).
METHODS: A prospective descriptive study with surgical intervention was conducted on 66 ears with clinically suspected OME in patients aged ≥16 years at Can Tho University of Medicine and Pharmacy Hospital from June 2024 to October 2025. Kempthorne score >5 was used solely as an eligibility threshold. Clinical examination, rigid otoscopy, pure-tone audiometry, conventional 226-Hz tympanometry, and WBT were performed. Intraoperative middle-ear fluid at myringotomy served as the positive reference standard in 31 operated ears; the remaining 35 non-operated ears were classified as disease-negative without surgical verification (composite reference standard with partial verification). An a priori sample-size calculation targeted ≥28 surgically verified ears (α=0.05, power 80%); post hoc power for AUC comparison was 84.6%. Three diagnostic strategies were evaluated: (1) standard 226-Hz tympanometry (type B or C); (2) WBT alone; (3) combined model-positive WBT plus ≥1 high-yield Kempthorne morphological feature (air-fluid level and/or absent mobility), excluding the enrollment threshold.
RESULTS: All enrolled ears had Kempthorne scores >5. Against the 66-ear composite reference dataset (31 fluid-positive, 35 assumed fluid-negative), sensitivity/specificity/AUC were 100.0%/68.2%/0.84 for standard tympanometry, 77.4%/81.8%/0.91 for WBT alone, and 93.5%/86.4%/0.94 for the combined model (McNemar p = 0.042). Among surgically verified ears, standard 226-Hz tympanometry detected all fluid-positive cases (sensitivity 100.0%). Operated and non-operated ears did not differ significantly on key preoperative indices (all p > 0.05).
CONCLUSION: WBT combined with specific Kempthorne morphological features provided complementary morphological and absorbance information in adult OME. However, because standard tympanometry already achieved 100% sensitivity in the operated subset, any apparent advantage of the combined model in specificity, NPV, and AUC is derived largely from the surgically unverified reference-negative subset and must be interpreted with caution in light of severe partial verification bias.