Development And Application Of Electronic Data Management And Statistical Analysis System In Clinical Trials

BackgroundThe role of statistics in clinical trials has been recognized by the majority ofclinicians.Whether the study design, data collection, or statistical analysis, statistically hasinfiltrated all aspects of clinical trials.Standardization of clinical trials of the building isthrough the implementation of specific standardized measures in various stages of clinicaltrials, thus ensuring the authenticity and reliability of the results of clinical trials.Domesticand foreign government departments have recognized the importance of standardized, haveintroduced a number of regulations, the principle requirements of the operation of the clinicaltrials.ndustry Association also announced a series of clinical trials, standardized instructions,the industry standard for the implementation of specific clinical trials.Randomization design phase of clinical trial research clinical investigators must face aproblem. Randomized grouping method studies have shown that the traditional randomizationmethods are increasingly unable to meet demand.Dynamic random method and centralrandomization system frequently mentioned is considered to be an important means to controlthe prognostic factors.Dynamic randomized algorithms, minimization and dynamic balancedrandomization method different emphases in the control of the prognostic factors. The balanceunder Minimization in dealing with the same importance of various prognostic factors isbetter.The balance of dynamic balanced randomization method in dealing with the hierarchyof importance of various prognostic factors is better.How effective combination of the twofull play to their superiority, to obtain the optimized random effect, it is worth the statisticsfurther explore.The aim of the clinical trial data management, data from subjects, complete and accuraterecord.Record of all operations involved in each step of the data management.Which isconducive to supervise and inspect the quality of data and test the implementationprocess.Domestic and international use of Electronic Data Capture (EDC) has developed a lotof electronic data management system.How to scientifically manage the clinical dataacquisition, entry, verification, lock, export, and work a good attempt.At present, foreign-developed data management system due to language and the price in domesticapplications is limited.Domestic development of data management systems function more orless, there are some shortcomings, especially in the standardization of data structure problemhas not been resolved.CDISC standards are becoming increasingly subject to the attention ofgovernment agencies and research institutions to be regarded as the best way to solve theproblems of data standardization.Domestic research in this area is rather late, the need forstatistics to ongoing research and practice.The statistical analysis is an important essential in the clinical trials.It is also animportant basis for judgment of drug efficacy and safety.At home and abroad on the methodof statistical analysis and content are the appropriate regulatory documents and guidanceguide.These regulations and guidelines have a strong principle.These regulations andguidelines on the form of statistical reports are not specifically regulated.Statistical staff athome and abroad, according to the characteristics of the unit, the preparation of a series ofstatistical reports and template macro, used to meet the requirements to complete thestatistical analysis report.Many large foreign companies have developed a statistical analysissystem for the standardization of the statistical analysis process.Domestic statistical analysisof most of the statistical staff independently, the lack of the necessary supervision and qualitycontrol.Development of standardized statistical analysis and template macro, using thenecessary supervision and management of network system of statistical analysis, is a questionworth considering today's National Statisticians.ObjectiveAccording to the regulations and guidelines of the standardized operation of domesticand international clinical trials and clinical trials of commonly used design types, and theactual situation of China's clinical trial data management and statistical analysis, thedevelopment of domestic clinical trials data management and statistical analysissystem.Dynamic balanced randomization method of CDISC data standards, statistical analysisof the emerging concept of programming to run the process network implementation into theclinical trial data management and statistical analysis of this study system.To provide reference for the clinical trials data management system and statistical analysis systemdevelopment and application of standardized development to explore a practical, efficient andreliable way to improve our data management and statistical analysis.MethodFamiliar with clinical trials, the central randomization, data management, statisticalanalysis of the specifications and requirements, combined with the clinical trial datacollection process, the central randomization and to collect the operational processes of thesubjects data into an electronic clinical literature searchtest data management systems,web-based operating randomized clinical trial data collection and verification of electronicmanagement.According to international norms and domestic regulatory requirements,preparation of standardized charts templates, and the preparation of each template SAS macroprogram, running SAS programs written in Web-based Statistical Analysis System completedthe statistical analysis of the electronic operational processes to ensure that clinicaltrialsstatistical analysis of standardized management.Results1. Combined with minimization and dynamic balanced randomization method, twodynamic random algorithm, the development of a central randomization system.The basicidea of this system randomization method is:Dynamic balanced randomization method tostudy the importance of the different level of prognostic factors, to ensure priority access tothe balance of the treatment groups in the hierarchy of importance of prognosticfactors.Minimization to ensure balance of the overall allocation of the treatment groups on thesame level of importance of prognostic factors.The system also takes into account theimplementation of the method of randomization in a variety of specific situations, including:the remaining drugs, a drug-free goods, more than one treatment group, the treatment groups,ranging from the number of cases.2. We developed a clinical trials data management system.The system uses the CDISCstandards to build the data structure.And the use of the system module library Managementthe eCRF required to check the indicator variable, and can easily set eCRF interface modules and elements of the system modules in the library.Single copies of entry and double entry twoentry.Double entry way, real-time data than on the way to guarantee consistent.Set threeverification method: the "logical conditions verification, source data verification, and manualverification.Doubt send, receive, answer and confirm the accuracy of the data to ensure.It canbe cCRF, questions, tables and data export, easy to review and statistical analysis.All work iscompleted, the data lock.3. Prepared in accordance with the requirements of domestic and foreign statisticalanalysis, a series of tables to meet the statistical analysis of the contents of template, and thedevelopment of SAS macro programs to ensure the standardization of the statistical results.Developed a statistical analysis of clinical trials, the network editor of the SAS program,run and double tables of comparison, to ensure the reliability of the results of statisticalanalyzes. The same time, to facilitate statistical analysis of process monitoring and qualitycontrol..ConclusionsResearch and development of clinical trials electronic data management and statisticalanalysis system and the successful completion of clinical trials central randomized grouping,data entry, data verification, questions about management, data locking and data exportwork.On the basis of standardized tables template developed SAS macro program to achievethe network to write and run SAS procedures, quality control than the function on the resultsof statistical analyzes, and statistical analysis.