【作者】 周莹；

【导师】 郑怀信；

【作者基本信息】 郑州大学 ， 免疫学， 2025， 硕士

【摘要】 背景和目的:百日咳是由百日咳鲍特菌(Bordetella pertussis,Bp)感染引起的急性呼吸道传染病,主要通过飞沫传播。其典型临床特征为阵发性痉挛性咳嗽,发作末期常伴随特征性鸡鸣样吸气性吼声。由于病程可迁延2-3个月,故以“百日咳”命名。尽管全球范围内已普遍实施百日咳疫苗接种计划,但其发病率仍呈持续上升态势。我国百日咳流行趋势尤为显著:2004-2024年间发病率呈现波动性增长,2014年后进入快速上升期;至2024年,全国报告病例数呈暴发式增长,全年发病例数已达2023年12.94倍。值得注意的是,当前临床实践中非典型病例比例显著增加,导致漏诊率与误诊率居高不下。这一现象不仅延长患者病程、加重疾病负担,更可能因延迟诊断引发社区传播风险。现有诊断体系面临双重瓶颈:其一,流行病学监测数据利用不足,尚未建立基于季节、年龄、性别等高危因素的分层预警机制;其二,实验室诊断依赖核酸检测,但基层机构受限于检测成本与时效性,亟需经济高效的筛查标志物。尽管既往研究提示外周血裂隙淋巴细胞可能与百日咳感染相关,但相关证据主要基于零星病例观察,缺乏系统性验证及多维度统计分析,尚未形成具有临床适用性的诊断标准。本研究通过整合多维度流行病学监测数据与实验室检测指标,旨在构建百日咳精准防控体系。解析流行特征规律:阐明百日咳的季节性分布、年龄与性别易感性差异,揭示家庭聚集性传播风险;确立诊断生物标志物:通过对比南阳市某三甲医院97例实验室确诊百日咳患者与97例匹配对照组的血常规参数(白细胞计数、淋巴细胞百分比等)、外周血裂隙淋巴细胞占比及C反应蛋白等关键生物学指标,筛选具有高敏感性与特异性的百日咳辅助诊断生物标志物组合;构建风险分层驱动的联合筛查路径:融合流行病学高危因素(季节、年龄)与实验室诊断标志物组合(裂隙淋巴细胞占比、淋巴细胞百分比及C反应蛋白),建立基层医疗机构百日咳疑似病例“风险分级-检测优先级”决策模型,实现精准资源分配与早期干预。方法:本研究基于中国疾病预防控制信息系统、国家卫生健康委员会及公共卫生科学数据中心2004-2024年百日咳病例数据,系统解析其流行病学分布特征,明确季节波动、年龄与性别易感性差异;纳入南阳市某三甲医院2024年3–7月194例研究对象(实验室确诊组97例,匹配对照组97例),检测白细胞计数(White Blood Cell,WBC)、淋巴细胞百分比(Lymphocyte percentage,LY%)、裂隙淋巴细胞占比及C反应蛋白(C-Reactive Protein,CRP)等17项实验室指标,通过Kolmogorov-Smirnov检验评估数据分布类型,正态与非正态数据分别采用独立样本t检验与Mann-Whitney U检验筛选组间差异变量,构建决策树诊断模型,并利用ROC曲线评估模型敏感度、特异度及曲线下面积(AUC)。结果:1.流行病学趋势:2004-2024年中国累计报告百日咳病例688784例,呈现显著季节性分布,4-8月为流行高峰(占比70.96%,488752/688784)。2024年病例数达494371例,为2023年同期(38,205例)的12.94倍,疫情进入暴发期。2.人群分布特征:2004-2019年全年龄段发病率逐年递增,0-1岁组为最高危人群,男性发病率高于女性。3.实验室指标差异:确诊组(97例)与对照组(97例)对比显示,裂隙淋巴细胞占比(U=499.5,P<0.001)、淋巴细胞百分比(U=3,166,P<0.001)、C反应蛋白(U=879,P<0.001)等10项指标差异具有统计学意义;嗜酸性粒细胞百分比、红细胞计数等8项指标无显著差异(P>0.05)。4.诊断模型效能:基于10个变量构建的决策树预测模型训练集总体准确率为96.3%,验证集总体准确率为88.2%,ROC曲线下面积(AUC)为0.982(95%CI:0.963-1.000),核心预测因子为裂隙淋巴细胞占比、淋巴细胞百分比及C反应蛋白(敏感度94.1%,特异度89.7%)。结论:1.时空分布特征:2004-2024年百日咳呈季节性发病,4-8月为高峰,2024年病例数达峰值。0-1岁组发病率最高,男性高于女性。2.模型创新性:构建以裂隙淋巴细胞占比>1.5%为核心的决策树预测模型,联合淋巴细胞比例及C反应蛋白指标,AUC值为0.982(95%CI:0.963-1.000),能有效预测百日咳感染风险,为后续PCR检测提供参考。更多还原

【Abstract】 Background and Objectives:Pertussis,an acute respiratory infectious disease caused by Bordetella pertussis(Bp)infection,is primarily transmitted through respiratory droplets.Its characteristic clinical manifestations include paroxysmal spasmodic cough typically followed by a distinctive inspiratory whooping sound during the terminal phase of coughing episodes.The disease derives its name"pertussis"(literally"hundred-day cough")from its protracted clinical course that may persist for 2-3 months.Despite global implementation of pertussis vaccination programs,the incidence rate continues to demonstrate a persistent upward trend.China has experienced particularly notable epidemiological progression:from 2004 to 2024,the incidence exhibited fluctuating growth patterns,entering a phase of rapid escalation post-2014.By 2024,the nation reported explosive case growth with annual incidence reaching a staggering12.94-fold increase compared to 2023.Of significant concern is the substantial rise in atypical clinical presentations observed in current clinical practice,resulting in persistently elevated rates of missed and misdiagnoses.This phenomenon not only prolongs disease duration and exacerbates clinical burden,but also heightens community transmission risks due to diagnostic delays.The current diagnostic system faces two major bottlenecks.Firstly,epidemiological surveillance data are underutilized,and a risk-stratified early warning system based on high-risk factors such as season,age,and sex has not yet been established.Secondly,while laboratory diagnostics rely on nucleic acid testing,primary healthcare institutions face constraints in testing costs and timeliness,creating an urgent need for cost-effective and efficient screening biomarkers.Although previous studies have suggested an association between lymphocytes observed in peripheral blood smears(or cleaved lymphocyte)and Bordetella pertussis infection,the evidence primarily stems from sporadic case reports.These findings lack rigorous validation and multivariate statistical analysis,and clinically applicable diagnostic criteria have yet to be established.This study integrates multifaceted epidemiological surveillance data with laboratory diagnostic indices to establish a precision control framework for pertussis.The research encompasses three key components:Epidemiological Profiling:Systematic analysis of epidemic patterns to delineate seasonal distribution characteristics,age-and sex-specific susceptibility disparities,and household cluster transmission risks of pertussis;Biomarker Identification:Comparative analysis of hematological parameters(including leukocyte count,lymphocyte percentage),peripheral blood cleaved lymphocyte proportions,and C-reactive protein(CRP)levels between 97 laboratory-confirmed pertussis cases and 97 age-/sex-matched controls from a tertiary hospital in Nanyang City,aiming to identify biomarker panels with high sensitivity and specificity for auxiliary diagnosis.Risk-Stratified Screening Protocol:Development of a joint decision model integrating epidemiological risk predictors(seasonality,age)and laboratory biomarkers(cleaved lymphocyte proportion,lymphocyte percentage,CRP)to create a"risk stratification-testing prioritization"algorithm for primary healthcare settings,enabling targeted resource allocation and early intervention.Methods:This study utilized pertussis case data from 2004 to 2024 obtained through the China Information System for Disease Control and Prevention,the National Health Commission,and the Public Health Science Data Center to systematically analyze epidemiological distribution patterns,with specific focus on seasonal fluctuations and age-and sex-specific susceptibility disparities.A total of 194 participants(97laboratory-confirmed cases and 97 age-/sex-matched controls)were enrolled from a tertiary hospital in Nanyang City between March and July 2024.Seventeen laboratory parameters were measured,including white blood cell count(WBC),lymphocyte percentage(LY%),cleaved lymphocyte proportion,and C-reactive protein(CRP).Data distribution normality was assessed using the Kolmogorov-Smirnov test.Group comparisons were performed with independent t-tests for normally distributed variables and Mann-Whitney U tests for non-parametric data.Significant differential variables were subsequently incorporated into a decision tree diagnostic model,whose performance was evaluated through receiver operating characteristic(ROC)curve analysis,with sensitivity,specificity,and area under the curve(AUC)as key metricsResults:1.Epidemiological Trends:A total of 688,784 pertussis cases were reported in China from 2004 to 2024,demonstrating marked seasonal variation with an epidemic peak during April–August(70.96%of annual cases,488,752/688,784).The 2024case count surged to 494,371,representing a 12.94-fold increase compared to the same period in 2023(38,205 cases),indicating transition to an outbreak phase.2.Population Distribution Patterns:From 2004 to 2019,incidence rates progressively increased across all age groups.The 0–1-year-old group exhibited the highest risk,with sex-specific disparities showing elevated incidence in males versus females.3.Laboratory Parameter Disparities:Comparative analysis between confirmed cases(n=97)and matched controls(n=97)revealed statistically significant differences(Mann-Whitney U test)in 10 parameters,including cleaved lymphocyte proportion(U=499.5,P<0.001),lymphocyte percentage(U=3,166,P<0.001),and C-reactive protein(U=879,P<0.001).Eight parameters(e.g.,eosinophil percentage,red blood cell count)showed no significant intergroup differences(P>0.05).4.Diagnostic Model Performance:The decision tree prediction model incorporating 10 variables achieved an overall accuracy of 96.3%in the training set and 88.2%in the validation set.ROC analysis demonstrated excellent discriminative capacity(AUC=0.982,95%CI:0.963–1.000).Core predictors included cleaved lymphocyte proportion,lymphocyte percentage,and CRP,with sensitivity of 94.1%and specificity of 89.7%.Conclusions:1.Spatiotemporal Distribution Characteristics:From 2004 to 2024,pertussis demonstrated seasonal epidemic patterns with peak incidence occurring between April and August.The year 2024 witnessed the highest case count.The 0–1-year-old age group exhibited the highest incidence rate,with male predominance over female(male-to-female ratio>1).2.Model Innovation:We developed a novel decision tree prediction model utilizing cleft lymphocytes>1.5%as the core parameter,integrated with lymphocyte proportion and C-reactive protein(CRP)levels.This model demonstrated exceptional predictive performance with an AUC of 0.982(95%CI:0.963–0.994),.The algorithm effectively stratifies pertussis infection risk prior to polymerase chain reaction(PCR)confirmation,serving as a valuable triage tool for diagnostic testing.更多还原