【作者】 周婷；

【导师】 王志忠；

【作者基本信息】 中南大学 ， 工程（专业学位）， 2022， 硕士

【摘要】 临床数据管理贯穿临床研究各个阶段,对试验数据的准确性和真实性起着决定性作用。针对我国临床试验数据管理系统缺乏有效的数据质量管理体系、数据质量评价标准以及不良事件的收集、分析和管理等问题,本文设计并实现了一个基于云计算的临床试验数据管理系统,主要研究内容如下:1.构建数据质量管理体系。本文通过质量手册、程序文件、作业指导书和质量记录等方式构建了数据质量管理体系,不断协调、质量监控、信息管理、质量管理体系审核和管理评审,实现质量管理体系的有效运行,保证了数据质量结果的各种因素和环节处于全面控制和管理,临床研究数据始终保持在可控和可靠的水平。2.建立数据质量评价标准。本文将试验人员、设备和数据分别建立质量评价标准,实现对临床数据的合理性、可归因性、易读性、同时性、原始性和准确性等方面进行质量评价,并采用随机森林模型,以构建的量化数据质量评价标准为输入,以数据质量是否合格为输出,实现临床数据质量总体评价准确率93.3%,数据质量评估耗时由人工的2天缩短为1秒,极大节省人力,缩短试验耗时。3.建立不良事件收集精确度。本文加强临床试验数据系统对不良事件的收集、分析和管理,将不良事件收集的精确度细分为高度精确、中度精确和低度精确的定义,明确不良事件收集和核查范围,实现保留临床研究中重要的安全性信息,佐证试验的可靠结论。4.建立基于云计算的临床试验数据管理系统本文采用云计算平台部署方式,采用C/S开发模式、Spring MVC架构和Cassandra数据库开发技术,实现了系统临床数据记录、监控、分析、中心化随机、受试者管理和不良事件管理模块,并分别从性能和功能两个维度进行测试,测试结果均表明系统符合上线需求,验证了系统的有效性。更多还原

【Abstract】 Clinical data management runs through all stages of clinical research and plays a decisive role in the accuracy and authenticity of test data.In view of the lack of effective data quality management system,data quality evaluation standards and the collection,analysis and management of adverse events in China’s clinical trial data management system,this paper designs and implements a clinical trial data management system based on cloud computing.The main research contents are as follows:1.Build a data quality management system.This paper constructs the data quality management system by means of quality manual,procedure documents,operation instructions and quality records,continuously coordinates,quality monitoring,information management,quality management system audit and management review,realizes the effective operation of the quality management system,ensures that various factors and links of data quality results are under comprehensive control and management,and clinical research data are always maintained at a controllable and reliable level.2.Establish data quality evaluation standards.In this paper,the quality evaluation standards are established for the test personnel,equipment and data respectively,so as to realize the quality evaluation of the rationality,attribution,readability,simultaneity,originality and accuracy of the clinical data.The random forest model is adopted,with the constructed quantitative data quality evaluation standard as the input and the qualified data quality as the output,so as to realize the overall evaluation accuracy rate of the clinical data quality of93.3%,The time of data quality evaluation is shortened from 2 days to 1second,which greatly saves manpower and shortens the test time.3.Establish the accuracy of adverse event collection.This paper strengthens the collection,analysis and management of adverse events in the clinical trial data system,subdivides the accuracy of adverse event collection into high accuracy,medium accuracy and low accuracy,defines the scope of adverse event collection and verification,retains important safety information in clinical research,and supports the reliable conclusions of the trial.4.Establish a clinical trial data management system based on cloud computingThis paper adopts the deployment mode of cloud computing platform,C / s development mode,Spring MVC architecture and Cassandra database development technology to realize the modules of clinical data recording,monitoring,analysis,centralized randomization,subject management and adverse event management of the system,and tests them from the two dimensions of performance and function respectively.The test results show that the system meets the on-line requirements and verify the effectiveness of the system.更多还原