耳鼻咽喉头颈外科门诊患者咽...疾病列线图模型的建立及验证_徐娟.pdf

论著 525中国医刊 2023 年 第58卷 第5期耳鼻咽喉头颈外科门诊患者咽喉反流性疾病列线图模型的建立及验证徐娟1，彭韬2，杨明1（1.湖南长沙市第三医院 耳鼻咽喉头颈外科，湖南 长沙 410015；2.湖南省人民医院/湖南师范大学附属第一医院 耳鼻咽喉头颈外科，湖南 长沙 410000）摘要：目的建立预测耳鼻咽喉头颈外科门诊患者咽喉反流性疾病（laryngopharyngeal reflux disease，LPRD）的列线图模型并进行验证。方法选择2021年1月至2022年3月在湖南长沙市第三医院耳鼻咽喉头颈外科门诊就诊的184例患者进行回顾性分析，根据有无LPRD将患者分为LPRD组（28例）与无LPRD组（156例），比较两组的一般资料及疾病相关因素，通过LASSO回归模型筛选耳鼻咽喉头颈外科门诊患者发生LPRD的影响因素，并行多因素logistic回归分析，根据多因素分析结果建立列线图模型并进行内部验证。结果两组患者在体重指数及辛辣饮食、胃食管反流病、幽门螺杆菌感染、甲状腺疾病、阻塞性睡眠呼吸暂停低通气综合征（obstructive sleep apnea-hypopnea syndrome，OSAHS）、焦虑、抑郁、慢性咳嗽、食物过敏占比方面差异均有显著性（P 0.05）。LASSO回归模型筛选出9个潜在的影响因素，分别为体重指数、辛辣饮食、胃食管反流病、幽门螺杆菌感染、甲状腺疾病、OSAHS、焦虑、抑郁、慢性咳嗽、食物过敏。对LASSO回归模型筛选的变量进行多因素logistic分析显示，辛辣饮食、胃食管反流病、幽门螺杆菌感染、OSAHS、焦虑、慢性咳嗽是耳鼻咽喉头颈外科门诊患者LPRD的独立影响因素（P0.05）。基于多因素分析结果构建列线图模型，受试者工作特征（receiver operating characteristic，ROC）曲线分析结果显示，该模型预测耳鼻咽喉头颈外科门诊患者LPRD的曲线下面积为0.934（95%CI 0.8820.987）。采用Bootstrap法重复抽样1000次，结果显示模型预测的耳鼻咽喉头颈外科门诊患者LPRD发生率与实际发生率基本吻合。H-L拟合优度检验结果显示，该列线图模型预测耳鼻咽喉头颈外科门诊患者LPRD的发生率与实际发生率比较差异无显著性。结论耳鼻咽喉头颈外科门诊患者LPRD主要受辛辣饮食、胃食管反流病、幽门螺杆菌感染等因素的影响，本研究建立的列线图模型预测耳鼻咽喉头颈外科门诊患者LPRD具有较高的准确度与区分度。关键词：耳鼻咽喉头颈外科；咽喉反流性疾病；多因素分析；列线图模型中图分类号：R766.1文献标识码：A文章编号：1008-1070（2023）05-0525-05doi:10.3969/j.issn.1008-1070.2023.05.016Establishment and validation of Nomogram model for predicting laryngopharyngeal reflux in otorhinolaryngology head and neck surgery outpatientsXu Juan1,Peng Tao2,Yang Ming1(1.Department of Otolaryngology,Head and Neck Surgery,Third Hospital of Changsha,Hunan Changsha 410015,China;2.Department of Otolaryngology,Head and Neck Surgery,Hunan Provincial Peoples Hospital,First Affiliated Hospital of Hunan Normal University,Hunan Changsha 410000,China)Abstract:Objective To establish and validate the Nomogram model for predicting laryngopharyngeal regurgitation(LPRD)in otorhinolaryngology head and neck surgery outpatients.Method 184 patients who were treated in the outpatient department of Otolaryngology Head and Neck Surgery of the Third Hospital of Changsha,Hunan Province from January 2021 to March 2022 were divided into LPRD group(28 cases)and non-LPRD group(156 cases)according to the presence or absence of LPRD.Compare the general data and disease-related factors of the two groups,screen the influencing factors of LPRD in the outpatient department of otorhinolaryngology head and neck surgery by LASSO regression model,and carry out multi-factor logistic regression analysis.According to the results of multi-factor analysis,establish a nomograph model and conduct internal validation.Result There were significant differences between the two groups in body mass index,spicy diet,gastroesophageal reflux disease,helicobacter pylori infection,thyroid disease,obstructive sleep apnea-hypopnea syndrome(OSAHS),anxiety,depression,chronic cough,and food allergy(P0.05).LASSO regression model screened 9 potential influencing factors,including body mass index,spicy diet,gastroesophageal reflux disease,helicobacter pylori infection,thyroid disease,OSAHS,anxiety,depression,chronic cough,and food allergy.Multivariate logistic analysis of the variables screened by the LASSO regression model showed that spicy diet,gastroesophageal reflux disease,Helicobacter pylori infection,OSAHS,anxiety,and chronic cough were the independent influencing factors of LPRD in otolaryngology head and neck surgery outpatients(P0.05).A nomograph model was constructed based on the results of multifactor analysis.The receiver operating characteristic(ROC)curve analysis results showed that the area under the curve of LPRD predicted by the model was 0.934(95%CI 0.8820.987).The bootstrap method was used to repeat sampling for 1000 times,and the results showed that the incidence of LPRD predicted by the model was basically consistent with the actual incidence.The H-L goodness of fit test results showed that there was no significant difference between the predicted incidence of LPRD and the actual incidence of LPRD in the outpatient department of otolaryngology head and neck surgery.Conclusion LPRD of outpatients of otolaryngology head and neck surgery is mainly affected by spicy diet,gastroesophageal reflux disease,helicobacter pylori infection and other factors.The nomograph model established in this study has high accuracy and discrimination in predicting LPRD of outpatients of otolaryngology head and neck surgery.Keywords:Otorhinolaryngology,head and neck surgery;Laryngopharyngeal reflux disease;Multiplicity analysis;Nomogram model基金项目：湖南省青年科学基金（2020JJ5303）526 论著 中国医刊 2023 年 第58卷 第5期咽喉反流也称反流性咽喉炎，多为胃内容物反流刺激咽喉黏膜所致，与多种耳鼻咽喉头颈外科疾病密切相关，由其引起的咽喉反流性疾病（laryngopharyngeal reflux disease，LPRD）因缺乏 典型的临床特征及医师对其认识不足，临床上常发生过度诊断或治疗不规范等情况1。明确耳鼻咽喉头颈外科门诊患者 LPRD 的发生风险有利于临床制订个性化的健康教育及治疗方案，进而增加患者获益。近年来研究发现，LPRD 可增加声带疾病、慢性鼻窦炎、中耳炎甚至是喉癌的发生风险，因此及时准确地预测 LPRD 具有重要的临床意义2。LPRD 诊断流程复杂且在一定程度上受主观因素的影响，为此本研究通过分析耳鼻咽喉头颈外科门诊患者 LPRD 的影响因素建立了列线图预测模型3，以期为 LPRD 的诊断及治疗提供帮助，现报道 如下。1资料与方法 1.1临床资料选择 2021 年 1 月至 2022 年 3 月在湖南长沙市第三医院耳鼻咽喉头颈外科门诊就诊的184 例患者进行回顾性分析。参照咽喉反流性疾病诊断与治疗专家共识（2015 年）4中的标准，对于存在咽喉不适、声音嘶哑、清嗓等疑似 LPRD 表现的患者在详细询问病史及喉镜检查后采用反流症状指数评分量表（r