|
注册号: Registration number: |
ChiCTR2600126445 |
|
最近更新日期: Date of Last Refreshed on: |
2026-06-09 12:01:02 |
|
注册时间: Date of Registration: |
2026-06-09 00:00:00 |
|
注册号状态: |
预注册 |
|
Registration Status: |
Prospective registration |
|
注册题目: |
基于多模态电子健康数据的深静脉血栓形成风险预测与免疫调节治疗的适应性研究 |
|
Public title: |
Adaptive Research on Risk Prediction of Deep Vein Thrombosis and immunomodulatory Therapy Based on Multimodal electronic Health Data |
|
注册题目简写: |
|
|
English Acronym: |
|
|
研究课题的正式科学名称: |
基于多模态电子健康数据的深静脉血栓形成风险预测与免疫调节治疗的适应性研究 |
|
Scientific title: |
Adaptive Research on Risk Prediction of Deep Vein Thrombosis and immunomodulatory Therapy Based on Multimodal electronic Health Data |
|
研究课题代号(代码): Study subject ID: |
|
|
在二级注册机构或其它机构的注册号: The registration number of the Partner Registry or other register: |
|
申请注册联系人: |
郑娟娟 |
研究负责人: |
郑娟娟 |
|
Applicant: |
Juanjuan Zheng |
Study leader: |
Juanjuan Zheng |
|
申请注册联系人电话: Applicant telephone: |
+86 595 8558 5856 |
研究负责人电话:
Study leader's |
+86 595 8558 5856 |
|
申请注册联系人传真 : Applicant Fax: |
研究负责人传真: Study leader's fax: |
||
|
申请注册联系人电子邮件: Applicant E-mail: |
leonkillua@163.com |
研究负责人电子邮件: Study leader's E-mail: |
332100103@qq.com |
|
申请单位网址(自愿提供): Applicant website(voluntary supply): |
研究负责人网址(自愿提供): Study leader's website(voluntary supply): |
||
|
申请注册联系人通讯地址: |
福建省泉州市东街250号 |
研究负责人通讯地址: |
福建省泉州市东街250号 |
|
Applicant address: |
No. 250, Dongjie Street, Quanzhou City, Fujian Province |
Study leader's address: |
No. 250, Dongjie Street, Quanzhou City, Fujian Province |
|
申请注册联系人邮政编码: Applicant postcode: |
研究负责人邮政编码: Study leader's postcode: |
||
|
申请人所在单位: |
福建医科大学附属泉州第一医院 |
||
|
Applicant's institution: |
Quanzhou First Hospital Affiliated to Fujian Medical University |
||
|
研究负责人所在单位: |
泉州市第一医院 |
||
|
Affiliation of the Leader: |
Quanzhou First Hospital, Fujian |
||
|
是否获伦理委员会批准: |
是 |
||
|
Approved by ethic committee: |
Yes |
||
|
伦理委员会批件文号: Approved No. of ethic committee: |
泉一伦[2025]K436号 |
伦理委员会批件附件: Approved file of Ethical Committee: |
查看附件View |
|
批准本研究的伦理委员会名称: |
泉州市第一医院伦理委员会 |
||
|
Name of the ethic committee: |
Ethics Committee of Quanzhou First Hospita |
||
|
伦理委员会批准日期: Date of approved by ethic committee: |
2026-05-18 00:00:00 | ||
|
伦理委员会联系人: |
杜苗苗 |
||
|
Contact Name of the ethic committee: |
Du Miaomiao |
||
|
伦理委员会联系地址: |
福建省泉州市东街250号 |
||
|
Contact Address of the ethic committee: |
No. 250, Dongjie Street, Quanzhou City, Fujian Province |
||
|
伦理委员会联系人电话: Contact phone of the ethic committee: |
+86 595 22277157 |
伦理委员会联系人邮箱: Contact email of the ethic committee: |
aidumimi@qq.com |
|
研究实施负责(组长)单位: |
泉州市第一医院 |
||||||||||||||||||||||
|
Primary sponsor: |
Quanzhou First Hospital, Fujian |
||||||||||||||||||||||
|
研究实施负责(组长)单位地址: |
福建省泉州市东街250号 |
||||||||||||||||||||||
|
Primary sponsor's address: |
No. 250, Dongjie Street, Quanzhou City, Fujian Province |
||||||||||||||||||||||
|
试验主办单位(项目批准或申办者): Secondary sponsor: |
|
||||||||||||||||||||||
|
经费或物资来源: |
福建省自然科学基金项目 |
||||||||||||||||||||||
|
Source(s) of funding: |
Project of Natural Science Foundation of Fujian Province |
||||||||||||||||||||||
|
研究疾病: |
深静脉血栓 |
||||||||||||||||||||||
|
Target disease: |
deep venous thrombosis |
||||||||||||||||||||||
|
研究疾病代码: |
|
||||||||||||||||||||||
|
Target disease code: |
|
||||||||||||||||||||||
|
研究类型: |
观察性研究 |
||||||||||||||||||||||
|
Study type: |
Observational study |
||||||||||||||||||||||
|
研究所处阶段: |
探索性研究/预试验 | ||||||||||||||||||||||
|
Study phase: |
0 |
||||||||||||||||||||||
|
研究设计: |
连续入组 |
||||||||||||||||||||||
|
Study design: |
Sequential |
||||||||||||||||||||||
|
研究目的: |
建立符合本地人群特征的多模态电子健康数据深静脉血栓(DVT)风险预测模型,通过融合结构化与非结构化电子健康记录(EHR)数据,开发高效精准的DVT风险预测算法,实现对临床患者的早期风险评估和分层管理。 |
||||||||||||||||||||||
|
Objectives of Study: |
1. Establish a multimodal electronic health record (EHR) data-driven deep vein thrombosis (DVT) risk prediction model tailored to local population characteristicsBy integrating structured and unstructured EHR data, the study aims to develop a highly efficient and accurate DVT risk prediction algorithm to achieve early risk assessment and stratified management for clinical patients. |
||||||||||||||||||||||
|
药物成份或治疗方案详述: |
|
||||||||||||||||||||||
|
Description for medicine or protocol of treatment in detail: |
|
||||||||||||||||||||||
|
纳入标准: |
|||||||||||||||||||||||
|
Inclusion criteria |
|||||||||||||||||||||||
|
排除标准: |
1.存在活动性出血或经研究者判断存在极高出血风险(如近期有颅内出血、消化道大出血等病史)。 |
||||||||||||||||||||||
|
Exclusion criteria: |
1. Presence of active bleeding or judged by the investigator to have an extremely high risk of bleeding (e.g., a recent history of intracranial hemorrhage, massive gastrointestinal bleeding, etc.). 2. Known severe hepatic impairment (e.g., ALT/AST > 3 times the upper limit of normal) or severe renal impairment (eGFR < 30 mL/min/1.73m²). 3. Known history of allergy to any of the drugs used in this study (including standard anticoagulants and immunomodulators). 4. Pregnant or lactating women, or women with pregnancy plans during the study period. 5. Currently participating in other interventional clinical trials (within 30 days prior to enrollment). 6. Suffering from other severe, uncontrolled systemic diseases, or other conditions that the investigator believes may affect the safety of the subject or the judgment of the study results. 7. Persons without capacity or with limited capacity for civil conduct who are unable to provide informed consent. |
||||||||||||||||||||||
|
研究实施时间: Study execute time: |
从 From 2026-05-01 00:00:00至 To 2029-04-30 00:00:00 |
征募观察对象时间: Recruiting time: |
从 From 2027-01-01 00:00:00 至 To 2029-04-30 00:00:00 |
|
干预措施: Interventions: |
|
|
研究实施地点: Countries of recruitment and research settings: |
|
||||||||||||||||||||||||||||
|
测量指标: Outcomes: |
|
|
采集人体标本:
Collecting sample(s)
|
|
|
征募研究对象情况: Recruiting status: |
尚未开始 Not yet recruiting |
年龄范围: Participant age: |
|
||||||
|
性别: |
男女均可 |
Gender: |
Both |
||||||
|
随机方法(请说明由何人用什么方法产生随机序列): |
无 |
||||||||
|
Randomization Procedure (please state who generates the random number sequence and by what method): |
None |
||||||||
|
是否公开试验完成后的统计结果: Calculated Results after the Study Completed public access: |
不公开/Private |
|
盲法: |
无 |
|
Blinding: |
None |
|
是否共享原始数据: IPD sharing |
是Yes |
|
共享原始数据的方式(说明:请填入公开原始数据日期和方式,如采用网络平台,需填该网络平台名称和网址): |
以文章发表共享数据 |
|
The way of sharing IPD”(include metadata and protocol, If use web-based public database, please provide the url): |
Publish articles to share data |
|
数据采集和管理(说明:数据采集和管理由两部分组成,一为病例记录表(Case Record Form, CRF),二为电子采集和管理系统(Electronic Data Capture, EDC),如ResMan即为一种基于互联网的EDC: |
一、 病例记录表 (Case Record Form, CRF) 设定方案鉴于本研究为基于电子健康记录(EHR)的大数据建模,项目的 CRF 并非传统的手工填报表格,而是转化为标准化的数据提取字典与变量模板(Electronic-CRF, eCRF),确保收集数据的全面性、规范性和一致性。针对 5000 例DVT患者和 5000 例对照患者,CRF 的核心数据模块设计如下: 人口学与基础特征模块:年龄、性别、体重指数(BMI)、吸烟史、饮酒史等常规结构化信息。 疾病与临床状态模块: 慢性病共患史(如糖尿病、高血压、心衰、特定类型恶性肿瘤史等)。 围手术期特征(如手术科室/类型、麻醉类型、手术时长等)。 生化与检验指标模块:D-二聚体峰值、凝血酶原时间(PT)、活化部分凝血活酶时间(APTT)、血小板计数(PLT)、血红蛋白(Hb)、炎症指标(C反应蛋白 CRP、白介素-6 IL-6)等。 NLP非结构化特征提取模块:基于预训练模型(如 BERT)从入院记录、手术记录、影像报告等无结构病历文本中自动抓取并映射的高危因素,如“近期制动”、“家族血栓史”、“遗传性凝血功能缺陷”等。 临床结局模块(黄金标准):发生深静脉血栓形成(DVT)事件的时间、预后并发症(如肺栓塞 PE 等级)与生存终点。 二、 电子采集和管理系统 (Electronic Data Capture, EDC) 建立与质控本项目的数据来源于福建医科大学附属泉州市第一医院等多家合作机构(2018-2025年)。为支撑上述 eCRF 变量的高效获取,团队依托医疗大数据中心构建了定制化的、整合自然语言处理(NLP)的云端 EDC 以及自动化数据治理平台。其管线建设包含: 多源数据接口与自动化采集依托地方卫生健康委及医疗联合体合作,EDC 系统直接打通医院信息系统(HIS)、实验室信息系统(LIS)、医学影像存档与通讯系统(PACS)及医保等多平台数据接口,实现多模态数据(临床、检验、影像、生化、药物使用)的底层批量拉取。 数据脱敏与标准化处理所有传入 EDC 的数据首先经过严格的数据脱敏技术(如隐去姓名、身份证号等直接隐私标识符),确保符合科技伦理与数据安全保护要求。 采用统一数据标准对异构数据进行格式转化,消除不同医疗机构间的系统壁垒,提升平台数据的兼容性。 NLP自动化标注与人工交叉审核机制智能标注:系统中内嵌基于规则与深度学习的 NLP 工具,对非结构化特征及 DVT 事件进行自动标注并回填至由 EDC 生成的结构化数据库。 质量控制(QC):对于核心结局事件,由两名以上资深临床医生在 EDC 系统后台进行盲法人工交叉审核,确保数据标签的准确率(预实验一致率已达92%92%)。 动态数据清洗与溯源管理EDC 内置自动化逻辑校验规则,运用四分位距(I Q R IQR)法标记异常值;针对<5%<5% 的缺失数据触发多重插补(MI)程序,严重缺失(>30%>30%)则自动剔除记录。平台保留所有处理和修改的审计追踪记录(Audit Trail),确保数据的可溯源性和模型训练集的高质量。 |
|
Data collection and Management (A standard data collection and management system include a CRF and an electronic data capture: |
1. Design Scheme for the Case Record Form (CRF)Given that this study focuses on big data modeling based on electronic health records (EHR), the project’s CRF is not a traditional manually filled form. Instead, it is transformed into a standardized data extraction dictionary and variable template (Electronic-CRF, eCRF) to ensure the comprehensiveness, standardization, and consistency of the collected data. For the 5,000 DVT patients and 5,000 control patients, the core data modules of the CRF are designed as follows:Demographics and Baseline Characteristics Module: Age, gender, body mass index (BMI), smoking history, alcohol consumption history, and other routine structured information.Disease and Clinical Status Module:History of comorbidities (e.g., diabetes, hypertension, heart failure, history of specific malignant tumors).Perioperative characteristics (e.g., surgical department/type, type of anesthesia, duration of surgery).Biochemical and Laboratory Indicators Module: Peak D-dimer, prothrombin time (PT), activated partial thromboplastin time (APTT), platelet count (PLT), hemoglobin (Hb), inflammatory markers (C-reactive protein [CRP], interleukin-6 [IL-6]), etc.NLP Unstructured Feature Extraction Module: High-risk factors (e.g., “recent immobilization,” “family history of thrombosis,” “hereditary coagulation defects”) automatically extracted and mapped from unstructured medical texts such as admission records, surgical records, and imaging reports using pre-trained models (e.g., BERT).Clinical Outcomes Module (Gold Standard): Time of deep vein thrombosis (DVT) event occurrence, prognostic complications (e.g., grade of pulmonary embolism [PE]), and survival endpoints.2. Establishment and Quality Control of the Electronic Data Capture (EDC) System The data for this project originates from Quanzhou First Hospital Affiliated to Fujian Medical University and multiple cooperative institutions (from 2018 to 2025). To support the efficient acquisition of the aforementioned eCRF variables, the team has built a customized, cloud-based EDC system integrated with natural language processing (NLP) and an automated data governance platform, relying on the medical big data center. Its pipeline construction includes:Multi-Source Data Interfaces and Automated Capture Based on the cooperation with local health commissions and medical consortia, the EDC system directly integrates with multiple platforms data interfaces, including the Hospital Information System (HIS), Laboratory Information System (LIS), Picture Archiving and Communication System (PACS), and medical insurance systems. This enables the base-level batch extraction of multimodal data (clinical, laboratory, imaging, biochemical, and medication utilization).Data Desensitization and Standardization All data entering the EDC system first undergoes strict data desensitization techniques (e.g., masking direct privacy identifiers like names and ID numbers) to ensure compliance with scientific ethics and data security protection requirements.Unified data standards are applied to format heterogeneous data, eliminating system barriers among different medical institutions and improving data compatibility across the platform.Automated NLP Annotation and Manual Cross-Review Mechanism Intelligent Annotation: The system embeds NLP tools based on rules and deep learning to automatically annotate unstructured features and DVT events, backfilling them into the structured database generated by the EDC.Quality Control (QC): For core outcome events, rigorous blinded manual cross-reviews are conducted by two or more senior clinicians in the back end of the EDC system to ensure the accuracy of data labels (the preliminary experiment consistency rate has reached92%92%).Dynamic Data Cleaning and Traceability Management The EDC has built-in automated logic validation rules, utilizing the interquartile range (I Q R IQR) method to flag outliers. For variables with missing data<5%<5%, a Multiple Imputation (MI) procedure is triggered, while records with severe missingness (>30%>30%) are automatically excluded. The platform maintains an Audit Trail for all processing and modifications, ensuring data traceability and the high quality of the model training set. |
|
数据与安全监察委员会: Data and Safety Monitoring Committee: |
有/Yes |