基于医疗大数据和微观模拟的宫颈癌筛查策略效果评价与优化研究

Cervical cancer is the most preventable but still threatening disease, and screening is the most effective prevention measure. The incidence and mortality of cervical cancer has been increasing in the east of China, indicating that screening strategy should be optimized. However, there is a lack of longitudinal evidence about epidemiological effect and optimization path of cervical cancer screening strategies in China. Based on the maternal and child health big data platform, this study will be conducted in Zhejiang province and Beijing by a combined methodology of cohort study and modelling study. This study will include three components: (1) The study will set up a respective population-based cohort in Zhejiang province and an ambispective screening-based cohort in Beijing. The epidemiological effect of different cervical cancer screening modalities, methods and frequency, in terms of a reduction in cervical cancer cumulative incidence and death risk, will be estimated at population level. (2) Using a Bayesian age-period-cohort model, the study aims to estimate the number of women who could be spared from cervical cancer in different screening modalities during the past 10 years and extrapolate it to 2040. (3) We will develop and calibrate microsimulation models of different hypothetical cohort based on cohort study, field studies and literature research. Then we will predict cost (ie, number of tests, colposcopies, disease detection, false-positive results), benefits (averted cancer cases and deaths), and efficiency ratios expressing the trade-off of cost vs benefits for various screening strategies. We will use sensitivity analyses to assess the impact of uncertainty and alternative assumptions on the results. This study is expected to assess and predict the epidemiological effect of various cervical cancer screening strategies for optimization under different scenarios in mainland China.

宫颈癌是最可预防但仍威胁女性生命的恶性肿瘤，筛查是最有效的预防手段。我国东部地区宫颈癌发病死亡率呈上升态势，表明需要优化筛查策略，但仍缺乏我国筛查策略流行病学效果与优化路径的纵向研究证据。本课题拟基于我系妇幼健康大数据平台，选择浙江和北京现场，采用队列研究和模型研究相结合的方法。首先基于医疗大数据建立回顾性人群队列和双向性筛查队列，研究不同筛查形式、方法和频率降低人群宫颈癌累计发病死亡风险的作用；然后在队列中通过贝叶斯年龄-时期-队列模型方法，研究不同筛查形式减少宫颈癌发病死亡的人群效果；最后在不同人群场景下进行微观模拟，采用队列研究、现场调研和文献荟萃法对关键参数进行系统量化，通过敏感性分析评估参数不确定性对模拟过程的影响，从而研究各场景下不同筛查策略的成本效益比最优值。研究预期成果将定量评价和预测不同宫颈癌筛查策略在我国人群中的流行病学效果，为在不同场景选择筛查策略优化路径提供依据。

子宫颈癌是可预防但仍严重威胁女性生命的恶性肿瘤，高效的筛查策略能够早期发现癌前病变，预防子宫颈癌的发生与死亡。我国子宫颈癌发病死亡率呈上升态势，亟需优化筛查策略，但目前仍缺乏我国人群子宫颈癌筛查策略效果与优化路径的纵向研究证据。本项目基于医疗大数据平台、全国人群体检与子宫颈癌筛查数据库和多中心临床注册研究开展以下研究内容：（1）整合HPV+细胞学联合筛查双向性队列、HPV初筛首轮筛查队列、多中心临床队列和HPV感染监测队列等多个人群队列，为研究筛查策略提供人群基础。（2）研究估计我国20岁及以上女性HPV感染率（15.0%）、CIN2+中HPV感染率（92.2%）和HPV阴性CIN2+比例（7.7%），通过趋势分析、层次聚类等多种方法揭示人群中HPV流行特征，为优化筛查策略提供基础参数。（3）在真实筛查场景中证实，HPV初筛比细胞学初筛的CIN2+检出率提高18%，阴道镜转诊率增加3%，阳性预测值增加29%，其中细胞学分流策略CIN2+检出率未显著增加，阴道镜转诊率减少24%，阳性预测值增加60%，HPV-16/18+细胞学分流策略CIN2+检出率增加28%，阴道镜转诊率增加13%，阳性预测值增加31%；不同HPV初筛分流策略的效果在35-54岁和55-64岁人群间存在异质性；HPV+细胞学联合筛查策略检出率能显著提高CIN2+检出率，但阴道镜转诊率和初筛数量明显增加；AI辅助细胞学初筛与人工阅片细胞学初筛相比灵敏度和特异度接近，可以排除99%的细胞学阴性者。（4）构建多种筛查策略，结合临床场景和微观模拟提出从准确性、成本和效率三维度评价不同策略效果的方案和策略优化路径。本研究基于中国人群纵向队列数据定量评价不同宫颈癌筛查策略的效果，探索不同场景下筛查策略的优化路径和方案，为选择适合我国人群特点的子宫颈癌筛查策略、减少子宫颈癌发病死亡提供依据。

内容获取失败，请点击重试