基于真实世界数据的门诊变应性鼻炎患者控制水平评估及病情未控制的风险预测研究-东南大学学报医学版

基于真实世界数据的门诊变应性鼻炎患者控制水平评估及病情未控制的风险预测研究

单位：1. 临汾市人民医院耳鼻咽喉头颈外科, 山西临汾 041000;
2. 杭州师范大学附属医院耳鼻咽喉科, 浙江杭州 310015

分类号：R765.21;R195.1

出版年·卷·期（页码）：2026·45·第三期（434-445）

摘要：

目的: 基于真实世界数据评估门诊变应性鼻炎(AR)患者控制水平,研究病情未控制的风险因素,构建对应的风险预测模型并验证其效能。方法: 本研究为前瞻性研究,选取2024年6月至2025年7月于临汾市人民医院门诊就诊的AR患者529例,将于2025年3月之前就诊的336例患者作为建模组,之后的193例患者作为验证组。患者就诊后4周统计其病情控制水平,将建模组中病情未控制、控制的患者分别作为未控制组、控制组。比较未控制组、控制组一般资料,采用多因素Logistic回归模型分析门诊AR患者病情未控制的影响因素,建立风险预测模型;经受试者工作特征曲线(ROC)、校准曲线评价模型的预测效能、区分度;采用Bootstrap法(重复抽样1 000次)进行内部验证,计算一致性指数及其95%CI;采用决策曲线分析(DCA)评价模型的临床净获益。结果: 建模组完成随访316例,其中病情未控制患者127例(40.19%),验证组完成随访178例,其中未控制患者71例(39.89%);未控制组规范规避过敏原的构成比低于控制组(P＜0.05),合并哮喘、多重过敏原致敏、中/重度AR的构成比及喷嚏、鼻塞、流涕、鼻痒视觉模拟量表(VAS)评分均高于控制组(P＜0.05),AR病程长于控制组(P＜0.05),用药依从性评分低于控制组(P＜0.05);多因素Logistic回归模型分析显示,合并哮喘、多重过敏原致敏、AR病程长、中/重度AR是门诊AR患者病情未控制的危险因素(P＜0.05),规范规避过敏原、用药依从性评分高是保护因素(P＜0.05)。ROC曲线显示,该模型预测建模组病情未控制的曲线下面积(AUC)为0.951,敏感度为86.61%,特异度为94.71%,预测验证组病情未控制的AUC为0.910,敏感度为78.87%,特异度为89.72%;Bootstrap方法(重复抽样1 000次)验证结果显示,模型在建模组、验证组中的一致性指数分别为0.948(95%CI:0.931~0.964)、0.907(95%CI:0.855~0.945),乐观估计分别为0.003、0.002;Hosmer-Lemeshow检验显示,该预测模型预测建模组、验证组病情未控制的概率与实际概率比较,差异均无统计学意义(χ²=6.335,P=0.458; χ²=7.254,P=0.519);DCA显示,风险预测模型在建模组、验证组中分别在0~0.98、0~0.83范围内获取临床净获益。结论: 基于真实世界数据构建的门诊AR患者病情未控制风险预测模型,纳入合并哮喘、多重过敏原致敏、AR病程、疾病严重程度、规范规避过敏原及用药依从性因素,经验证具有良好的预测效能与区分度。

Objective: To evaluate the control level of outpatients with allergic rhinitis(AR) based on real-world data, investigate the risk factors for uncontrolled AR, and construct and validate a corresponding risk prediction model. Methods: This prospective study enrolled 529 AR patients who visited the outpatient clinic of Linfen People's Hospital from June 2024 to July 2025. A total of 336 patients who visited before March 2025 were assigned to the modeling group, while the subsequent 193 patients were assigned to the validation group. The disease control level was evaluated 4 weeks after the visit. Patients in the modeling group were divided into an uncontrolled group and a controlled group based on their disease status. General characteristics were compared between the two groups. Multivariate Logistic regression analysis was used to identify the influencing factors for uncontrolled AR and to establish the risk prediction model. The predictive efficacy and discrimination of the model were evaluated using the receiver operating characteristic(ROC) curve and calibration curve. The Bootstrap method(1 000 resamples) was employed for internal validation, and the concordance index(C-index) with its 95% confidence interval(CI) was calculated. Decision curve analysis(DCA) was applied to evaluate the clinical net benefit of the model. Results: In the modeling group, follow-up was completed for 316 cases, and 127 patients(40.19%) had uncontrolled AR. In the validation group, follow-up was completed for 178 cases, and 71 patients(39.89%) had uncontrolled AR. Compared with the controlled group, the uncontrolled group had a lower proportion of standardized allergen avoidance(P<0.05), and higher proportions of comorbid asthma, multiple allergen sensitization, and moderate/severe AR(P<0.05). Furthermore, the Visual Analogue Scale(VAS) scores for sneezing, nasal congestion, rhinorrhea, and nasal itching were significantly higher(P<0.05), the duration of AR was longer(P<0.05), and the medication compliance score was lower(P<0.05) in the uncontrolled group. Multivariate Logistic regression analysis showed that comorbid asthma, multiple allergen sensitization, long AR duration, and moderate/severe AR were risk factors for uncontrolled AR(P<0.05), while standardized allergen avoidance and a high medication compliance score were protective factors(P<0.05). The ROC curve showed that the area under the curve(AUC) of the model for predicting uncontrolled AR in the modeling group was 0.951, with a sensitivity of 86.61% and a specificity of 94.71%. The AUC in the validation group was 0.910, with a sensitivity of 78.87% and a specificity of 89.72%. The Bootstrap validation(1 000 resamples) demonstrated that the C-indices of the model were 0.948(95%CI:0.931-0.964) in the modeling group and 0.907(95%CI: 0.855-0.945) in the validation group, with optimism estimates of 0.003 and 0.002, respectively. The Hosmer-Lemeshow test revealed no statistically significant differences between the predicted and actual probabilities of uncontrolled AR in either the modeling group(χ²=6.335, P=0.458) or the validation group(χ²=7.254, P=0.519). DCA showed that the risk prediction model obtained clinical net benefits within the threshold probability ranges of 0-0.98 and 0-0.83 in the modeling group and the validation group, respectively. Conclusion: The risk prediction model for uncontrolled AR in outpatients, constructed based on real-world data and incorporating comorbid asthma, multiple allergen sensitization, AR duration, disease severity, standardized allergen avoidance, and medication compliance, demonstrates good predictive efficacy and discrimination upon validation.

参考文献：

[1] SHIN Y H,HWANG J,KWON R,et al.Global,regional,and national burden of allergic disorders and their risk factors in 204 countries and territories,from 1990 to 2019:a systematic analysis for the Global Burden of Disease Study 2019[J].Allergy,2023,78(8):2232-2254.
[2] WANG N,YAO Y,LIU Y H,et al.Allergic rhinitis in China:trends,challenges and implications over the past two decades[J].Clin Exp Allergy,2025,55(8):648-658.
[3] RODRIGUES J,PINTO J V,ALEXANDRE P L,et al.Allergic rhinitis seasonality,severity,and disease control influence anxiety and depression[J].Laryngoscope,2023,133(6):1321-1327.
[4] HANNIKAINEN P,KAHN C,TOSKALA E.Allergic rhinitis,rhinosinusitis,and asthma connections across the unified airway[J].Otolaryngol Clin N Am,2024,57(2):171-178.
[5] WISE S K,DAMASK C,ROLAND L T,et al.International consensus statement on allergy and rhinology:allergic rhinitis-2023[J].Int Forum Allergy Rhinol,2023,13(4):293-859.
[6] 鲁莲,张家熔,陈安琪.过敏性鼻炎并发哮喘患儿临床特征及预后不良的危险因素分析[J].中国妇幼保健,2025,40(9):1649-1653.
[7] 中华耳鼻咽喉头颈外科杂志编辑委员会鼻科组,中华医学会耳鼻咽喉头颈外科学分会鼻科学组.中国变应性鼻炎诊断和治疗指南(2022年,修订版)[J].中华耳鼻咽喉头颈外科杂志,2022,57(2):106-129.
[8] MORISKY D E,GREEN L W,LEVINE D M.Concurrent and predictive validity of a self-reported measure of medication adherence[J].Med Care,1986,24(1):67-74.
[9] BROEK J L,BOUSQUET J,AGACHE I,et al.Allergic rhinitis and its impact on asthma(ARIA) guidelines—2016 revision[J].J Allergy Clin Immunol,2017,140(4):950-958.
[10] OGULUR I,MITAMURA Y,YAZICI D,et al.Type 2 immunity in allergic diseases[J].Cell Mol Immunol,2025,22(3):211-242.
[11] GAO Y D,WANG Z J,OGULUR I,et al.The evolution,immunopathogenesis and biomarkers of type 2 inflammation in common allergic disorders[J].Allergy,2025,80(7):1848-1877.
[12] TOSCA M A,TRINCIANTI C,NASO M,et al.Treatment of allergic rhinitis in clinical practice[J].Curr Pediatr Rev,2024,20(3):271-277.
[13] BOUSQUET J,MELÉN E,HAAHTELA T,et al.Rhinitis associated with asthma is distinct from rhinitis alone:the ARIA-MeDALL hypothesis[J].Allergy,2023,78(5):1169-1203.
[14] LI Y T,YE Q Q,LU Y X,et al.Allergen sensitization patterns:Allergic rhinitis with multimorbidity versus alone—a real-world study[J].Clin Transl Allergy,2025,15(1):1-11.
[15] ZHANG Q,JIAO J,WANG X,et al.The role of fibroblast in chronic rhinosinusitis with nasal polyps:a key player in the inflammatory process[J].Expert Rev Clin Immunol,2025,21(2):169-179.
[16] 蔡芳宇,李颖,李春苗,等.儿童变应性鼻炎并发哮喘的影响因素及风险预测模型构建[J].中国临床医生杂志,2025,53(7):918-921.
[17] 王健艳,林叶青,王紫臣,等.潮州地区学龄期过敏性鼻炎患儿过敏原调查分析及健康教育作用的研究[J].中国耳鼻咽喉头颈外科,2024,31(9):580-584.
[18] LONG T,HU X,LIU T,et al.A nomogram of predicting healthcare-associated infections in burned children[J].Pediatr Infect Dis J,2024,43(12):1147-1151.
[19] 张璐,于涛.慢性阻塞性肺疾病合并社区获得性肺炎风险列线图的建立及验证[J].东南大学学报(医学版),2024,43(6):882-889.
[20] 李楠楠,赵芹,杨华,等.内镜黏膜下剥离术后患者中重度疼痛影响因素分析及预测模型的构建[J].现代医学,2025,53(12):1851-1858.

服务与反馈：

【文章下载】【发表评论】【查看评论】【加入收藏】

提示：您还未登录，请登录！点此登录