1～6月龄分泌性中耳炎患儿多频声导纳诊断研究

The Effects of Different Probe Tone and Tympanometric Admittance Measurement Methods to Otitis Media Prediction

目的探讨探测音频率及声导纳值测量方法对1～6月龄婴儿分泌性中耳炎诊断的价值。方法筛选1～6月龄正常婴儿71人（正常组142耳）、分泌性中耳炎患儿54人（中耳炎组90耳）（颞骨薄层CT证实），采用GSl2000Tympstar2型中耳分析仪分别对两组对象进行226、1000Hz探测音鼓室声导纳检测，并用Linder／Jerger法、Baldwin法、基线法测量1000Hz探测音鼓室声导纳，采用ROC曲线下面积（A）比较的方法分析四种方法对婴儿中耳炎的诊断准确性，取值范围：0≤A≤1，A〉0．5，且越接近1，诊断准确性越高；A—0．5，无诊断意义；A〈0．5则不符合实际情况。结果1～3月龄组的226Hz探测音鼓室声导纳值以及Linder／Jerger法、Baldwin法、基线法测得的1000Hz探测音鼓室声导纳值正常组分别为0．83±0．29、0．60±0．55、0．74±0．56、0．90±0．59mmho，中耳炎组分别为0．82±0．35、0．01±0．06、-0．24±0．15、～0．29±0．21mmho，ROC曲线下面积（正常组与中耳炎组同时检测，下同）分别为0．507、0．896、0．976、0．988。4～6月龄组四种方法测得鼓室声导纳值正常组分别为0．60±0．35、0．55±0．58、0．76±1．0、0．86却．72mmho，中耳炎组分别为0．36±0．24、0±0、-0．34±0．16、-0．44±0．28mmho；ROC曲线下面积分别为0．749、0．888、0．969、0．988。1000Hz探测音鼓室声导纳的ROE曲线下面积在两个年龄组均大于226Hz探测音鼓室声导纳值（P〈O．05）；基线法测得1～3月龄组的1000Hz探测音鼓室声导纳的ROC曲线下面积大于其他三种方法（P〈0．05）。基线法测得4～6月龄组的1000Hz探测音鼓室声导纳的ROC曲线下面积大于其他三种方法，但与Baldwin法比较差异无统计学意义（P〉0．05）。结论1000Hz探测音声导纳对于1～6月龄分泌性中耳炎患儿的诊断意义高于226Hz探测音声导纳；1000Hz探测音鼓室声导纳的测量方法以基线法较为合适。

Objective To explore the effects of different probe tone and tympanometric admittance measurement methods on the diagnose of otitis media prediction,and to provide a clinical reference for the selection of probe tones and measurement method to apply to newborn infants. Methods Tympanograms with 226 Hz and 1 000 Hz probe tones were obtained from normal infants （142 ears） and infants with otitis media （90 ears） . Mean values, standard deviations, the 90% range and 95% mean confidence interval were recorded as the variable to observe po- tential impacts on tympanometric admittance with 226 Hz probe tone and tympanometric peak admittance with 1 000 Hz probe tone by three different measurement methods respectively according to Linder/Jerger classification, the Baldwin classification and baseline classification adapted from Baldwin. The four indexes were tested with compara- tive analysis. The area under ROC curve simultaneous detects the normal group and otitis media group, with refer- ences to the accuracy of the detection method to disease （including specificity and sensitivity）. Results In 1--to--3month group,tympanometric admittance with 226 Hz probe tone and tympanometric peak admittance with 1 000 Hz probe tone by three different measurement methods were 0.83±0.29,0.60±0.55,0.74±0.56,0.90±0.59 in nomal infants and 0. 82±-0. 35,0. 01±0. 06,-0. 24±0.15,-0. 29±0. 21 in infants with OME,respectively. The areas under ROC curve of the four kinds of measurement methods were 0. 507,0. 896,0. 976,0. 988, respectively. In 4 to 6 month group, tympanometric admittance with 226 Hz probe tone and tympanometric peak admittance with 1 000 Hz probe tone by three different measurement methods were 01 60±0.35,0.55± 0.58,0.76 ±1.0,0.86±0.72 in nomal infants and 0.36±0.24,0.36±0.34±0. 16,-0.44 ±0.28 in infants with OME,respectively. The areas under ROC curve of the four kinds of measurement methods were 0. 749,0. 888,0. 969,0. 988 in 4 to 6 month infants,respectively. Tympanometric peak admittance with 1 000 Hz probe tone by three different measurement methods were significantly better than that with 226 Hz probe tone. The areas under ROC curve of tympanometric peak admittance with 1 000 Hz probe tone by baseline classification adapted from Baldwin were bigger than others and the differences were significant（P〈0.05）,but no difference between the Baldwin classification and baseline classification in 4 to 6 month infants（P〉0.05）. Conclusion Tympanometric peak admittance with 1 000 Hz probe tone were better than that with 226 Hz probe tone to assess otitis media in 1-6 month infants. The baseline classification adapted from Baldwin was appropriate for the measurement of tympanometric peak admittance with 1 000 Hz probe tone.