The Development Of Central Randomization Network System For Multi-Center Randomized Controlled Trials

Objective: The central randomization network system for multi-center randomized controlled trials(RCTs) was developed, which based on computer technology and network platform, and the minimization method was nested in system, in order to provide safety, convenient randomization technology tools for multi-center RCTs.Methods: The minimization method was selected Pocock and Simon standard deviation method, which commonly used and have better balance than others. C# and SQL were used to program by the aid of Microsoft Visual Studio 2015 and Microsoft SQL Server 2008. The SSL(Secure Sockets Layer) protocol was deployed in web server, and using 128-bit key encrypted between browser and web server, EVA data structure model was employed to store data in Microsoft SQL Server. Base on these, we developed a central randomization network system for multi-center RCTs.Results:1. To ensure random allocation and the balance among groups, the system needs to set some related parameters, Including bias distribution probability of allocation and the biggest difference of subjects between groups.2. Using the EVA data structure model to store data, it can accommodate multiple clinical trials without modifying the data tables.3. It can be use stratified randomization and block randomization flexibly, the allocation proportion among groups has a variety of forms, such as 1:1, 2:1, 1:1:1 and so on, which can adapt to various types of clinical trials.4. The system has simulation function,it can be use to ensure the best parameter values of bias distribution probability of allocation and the biggest difference of subjects between groups according to specific situation of different clinical trials, which can reduce the workload greatly and avoid selection bias.5. The prognosis factors not only can be classified, also can be continuous in the system. The system can automatic demarcate their level for continuous prognostic factors, and achieve good balance among groups.6. It has good advantage for the management of double blind?subjects and drug in multi-center RCTs, the system can set up the dispatch of drug number in every center dynamically, knowing the subjects who into groups and the drug remainder and consumption in real time, which can help clinical trials administrator to supplement drug without delay.7. The system sets up a multi-level management modes and authentication, it has three roles that have differing operating authorization, system administrator, clinical trials administrator and clinical trial center researchers, the role must be needs corresponding account and password to log in, which can ensure the system's operation security.8. Except Ali cloud server to provide DDOS(Distributed Denial of service)protection, Trojan Killing, brute force attacks and other security prevention, The system itself also uses SSL and ISS(Internet Information Server)to provide security prevention. Database uses different backup to backup data every 7days.Conclusion : The central randomization network system for multi-center randomized controlled trials(RCTs) integrates with the current international mainstream minimization method, it can provide good balance among groups by setting some related parameters and prognosis factors. Based on computer technology and network platform, that solves the problem of difficult organization for minimization safely and effectively, and provides a reliable technical for the use of randomization in the study, improving the efficiency of randomization greatly. The system is simple and convenient to use, has a strong applicability, particularly suitable for multi-center, randomization, blind(single/double) parallel control study.