血清总IgE联合呼出气一氧化氮检测对激素敏感性咳嗽的诊断价值

Diagnostic value of combined measurement of serum total IgE and fractional exhaled nitric oxide in corticosteroid-responsive cough

目的:研究血清总IgE(TIgE)和呼出气一氧化氮(FeNO)联合检测对激素敏感性咳嗽的预测价值,以筛选适合糖皮质激素治疗的慢性咳嗽患者人群。方法:本研究为回顾性队列研究。采用非随机抽样法,分析2019年1月至2022年2月同济大学附属同济医院276例接受过糖皮质激素治疗的慢性咳嗽患者的临床资料,以咳嗽消失或明显缓解作为有效判断标准,采用逐步logistic回归分析性别、年龄、咳嗽病程、咳嗽症状积分、生活质量评分以及血清TIgE和FeNO等指标或因素对糖皮质激素疗效的影响,筛选独立预测因素并建立疗效预测模型方程。结果:最终纳入的158例慢性咳嗽患者中,55.1%(87/158)确诊为激素敏感性咳嗽。与非激素敏感性咳嗽相比,激素敏感性咳嗽女性为主[62.1%(54/87)比45.1%(32/71),χ^(2)=5.13,P=0.023],咳嗽日间较轻微[3.0(0.0)分比3.0(1.0)分,Z=3.46,P=0.001],TIgE[109.0(312.7)U/ml比57.8(99.8)U/ml,Z=3.40,P=0.001]和FeNO[29.0(20.5)ppb比19.0(14.0)ppb,Z=5.42,P<0.001]水平较高。多因素回归分析显示TIgE(OR=1.85,P=0.021)和FeNO(OR=4.71,P=0.010)为激素敏感性咳嗽的独立预测因素。建立的回归方程模型对激素敏感性咳嗽具有中等预测价值(Kappa=0.59,曲线下面积为0.84,P<0.001),预测敏感度为69.2%,特异度为88.1%,阳性预测值为78.3%,阴性预测值为82.2%。结论:TIgE和FeNO对激素敏感性咳嗽有一定的识别能力,两者联合能提高识别效率。

Objective To investigate the predictive value of combined measurement of serum total IgE(TIgE)and fractional exhaled nitric oxide(FeNO)for corticosteroid-responsive cough(CRC),so as to identify patients with chronic cough suitable for corticosteroids treatment.Methods This was a retrospective study.Using non-random sampling method,the clinical data of 276 patients with chronic cough treated with corticosteroids were collected in Tongji Hospital affiliated to Tongji University from Jan.2019 to Feb.2022.With cough complete resolution or noticeable relief as therapeutic outcome criterion,the effects of gender,age,cough duration,cough symptom score,quality of life score,TIgE,and FeNO on the efficacy of corticosteroids were analyzed using stepwise logistic regression.Independent predictors were screened,and an efficacy prediction model equation was developed.Results Of the included 158 patients with chronic cough,55.1%(87/158)were diagnosed with CRC.Compared with non-CRC,CRC patients were mainly female(62.1%[54/87]vs 45.1%[32/71],χ^(2)=5.13,P=0.023),had a milder daytime cough(3.0[0.0]vs 3.0[1.0],Z=3.46,P=0.001).TIgE(109.0[312.7]U/ml vs 57.8[99.8]U/ml,Z=3.40,P=0.001),and FeNO(29.0[20.5]ppb vs 19.0[14.0]ppb,Z=5.42,P<0.001)levels were also higher in CRC patients than in non-CRC patients.Multifactorial regression analysis showed that TIgE(OR=1.85,P=0.021)and FeNO(OR=4.71,P=0.010)were independent predictors for CRC.The regression equation model had moderate predictive value for CRC(Kappa=0.59,AUC=0.84,P<0.001),with a predictive sensitivity of 69.2%,specificity of 88.1%,positive predictive value of 78.3%,and negative predictive value of 82.2%.Conclusions Serum TIgE and FeNO can identify CRC to a certain extent,and the combined measurement can significantly improve the predictive efficacy.