A central randomization allocation system for clinical trials

Abstract

The present application relates to the technical field of clinical trial management, and is especially a central randomization allocation system for clinical trials. The present application uses a dynamic randomization algorithm to calculate the optimal group allocation in real time according to the baseline characteristics of the enrolled subjects, dynamically adjusts the probability using the urn method, effectively ensures the balance of key factors between groups, improves the scientificity and comparability of the trial, automatically triggers precise blinding and replenishment instructions based on real-time inventory and consumption prediction, optimizes distribution by solving the minimum cost flow problem, greatly reduces drug waste and drug interruption risk, ensures dynamic balance of supply, and checks the evolution logic of syndromes in online visits to ensure the continuity and logical reliability of the TCM syndrome data, and feeds back the updated syndrome information to the randomization unit in real time, forming a closed loop of allocation and feedback.

Description

Technical Field

[0001] This invention relates to the field of clinical trial management technology, specifically to a central randomization allocation system for clinical trials. Background Technology

[0002] Given the increasing complexity and scale of multicenter clinical trials, centralized randomization systems have become a key technological support for ensuring the scientific rigor and efficiency of trials. Most current mainstream centralized randomization systems rely on Computer Telecommunication Integration (CTI) technology architectures, enabling basic functions such as participant registration, group assignment based on static algorithms (e.g., stratified block randomization) or partially dynamic algorithms (e.g., minimization), blinding and management of investigational drugs, and electronic data collection. These systems, to a certain extent, standardize processes and reduce human bias, playing a particularly important role in Western medicine clinical trials with clearly defined diseases and standardized intervention protocols. Currently, these systems are iterating towards integration and digitalization, with some systems attempting to integrate Interactive Response Modules (IWR / IWRS) and drug supply chain management modules.

[0003] However, when the above-mentioned technical solutions are transplanted to complex clinical trial scenarios such as traditional Chinese medicine that emphasize "syndrome differentiation and treatment", their inherent shortcomings become apparent. Most existing randomization algorithms are static or only include baseline covariates for consideration, making it difficult to adapt to the core characteristics of the dynamic changes in the "syndrome type" of subjects in traditional Chinese medicine trials as the treatment progresses. This makes it difficult to maintain comparability between groups at the key pathogenesis level in the later stages of the trial. Secondly, the matching drug management module is rigid and adopts a one-time batch blinding method, which cannot adaptively allocate drugs and dynamically fill blinds according to the dynamic changes in the distribution of syndrome types in each center. This can easily lead to a shortage of drugs for some syndrome types and an oversupply of other drugs, which not only affects the maintenance of blinding state but also wastes resources. In addition, the functional units within the system (such as randomization, drug management, and data collection) often operate in silos, lacking a deep collaboration and closed-loop feedback mechanism based on business logic. For example, the syndrome type change data collected during visits cannot be fed back to the randomization strategy and drug scheduling logic in real time. Summary of the Invention

[0004] The purpose of this invention is to provide a central randomization allocation system for clinical trials to address the problems mentioned in the background section.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a central randomization allocation system for clinical trials, the system comprising a trial configuration management unit, a dynamic randomization allocation unit, a drug dynamic blinding supply chain optimization unit, an online visit data acquisition unit, a data quality control and security management unit, and an emergency handling and unblinding unit.

[0006] Preferably, the dynamic randomization allocation unit includes a hierarchical factor definition module, an adaptive random algorithm engine module, a random sequence generation and encryption module, an allocation execution and interface module, and a real-time monitoring module for allocation balance.

[0007] Preferably, the adaptive random algorithm engine module includes a TCM syndrome type-subject state adaptive randomization algorithm, the specific algorithm steps of which are as follows:

[0008] Step 1: Baseline data of newly enrolled subjects are used as the group distribution matrix M for each center and syndrome type. current Baseline data includes research center ID, disease stage, TCM four diagnostic methods information, and syndrome type;

[0009] Step 2: Convert the TCM four diagnostic methods information into standard syndrome codes using the built-in rule base. Assuming there are experimental group (T) and control group (C), for the center and syndrome type (strata) to which the subject belongs, calculate the inter-group difference values ​​DT and DC if the subject is assigned to group T or group C. The difference value calculation formula uses the range method:

[0010] D=∣N T,strata -N C,strata +∣P T,strata -P C,strata |;

[0011] Where N represents the number of people and P represents the mean of the key covariate;

[0012] Step 3: Adjust the probability using the urn method. If D T <D C If the probability of random assignment is higher, such as 0.8, then group T is assigned a higher probability; otherwise, group C is assigned a higher probability.

[0013] Step 4: Randomly generate group G based on the above probabilities. assign and encrypted random seed, G assign Send to the drug dynamic blinding supply chain optimization unit.

[0014] Preferably, the drug dynamic blinding supply chain optimization unit includes a drug coding and blind management module, a central inventory intelligent monitoring module, a dynamic blinding trigger module, a logistics and distribution optimization module, and a drug distribution and recycling verification module.

[0015] Preferably, the drug dynamic blinding supply chain optimization unit receives allocation instructions and, in conjunction with inventory forecasts, automatically generates replenishment and blinding instructions. The specific algorithm steps are as follows:

[0016] Step 1: Input the assignment group G from the randomization unit. assign Real-time inventory of each center S realtime Historical enrollment rate V enroll Shedding rate Rdropout Drug shelf life (T) exp ;

[0017] Step 2: Predict the demand D for the next t days based on the moving average algorithm. pred Then we have:

[0018] D pred =(V enroll ×(1−R dropout ))×t;

[0019] Step 3: Calculate the safety stock threshold for the corresponding certificate type and group for this center:

[0020] S safe =D pred ×1.5;

[0021] Step 4: Compare with real-time inventory S realtime With S safe If S realtime safe This will trigger the blind coding and replenishment tasks;

[0022] Step 5: Use a random number generator to generate a blind sequence for the new batch of drugs and encrypt and bind it with the drug traceability code;

[0023] Step 6: Based on the urgency and logistics costs of each center, solve the minimum cost flow problem to generate the optimal delivery route;

[0024] Step 7: Output the blinding instruction, logistics delivery note, and drug dispensing verification key, and send the drug dispensing application confirmation signal and new drug blind code to the online visit data acquisition unit.

[0025] Preferably, the online visit data acquisition unit includes an electronic medical record report customization module, a mobile follow-up and data entry module, a subject compliance assistance module, a syndrome evolution tracking and logic verification module, and an instant communication and task reminder module.

[0026] Preferably, the online visit data acquisition unit collects treatment efficacy data and verifies the logic, feeds back the dynamically changing syndrome types to the randomization unit, and uses a syndrome type evolution logic verification algorithm. The specific algorithm steps are as follows:

[0027] Step 1: Enter the TCM four diagnostic methods data for the current visit and the syndrome record from the previous visit. prev And the preliminary certificate type Z calculated during this visit current_raw ;

[0028] Step 2: Use TCM diagnostic algorithms to convert the four diagnostic methods data into syndrome vectors and check Z-type patterns. current_raw With Z prev ​If the evolutionary relationship conforms to the clinical logic of traditional Chinese medicine, it is marked as abnormal.

[0029] Step 3: If the verification passes, confirm Z. current The data will be packaged, and if the verification fails, a pop-up window will prompt the researcher to verify it, and the questioning log will be recorded.

[0030] Step 4: Confirm the Z current The system packages efficacy data, outputs structured efficacy data packages, syndrome evolution tags, and logical challenge reports, and includes Z... current The feedback is sent in real time to the hierarchical factor definition module in the dynamic randomization allocation unit.

[0031] Preferably, the data quality control and security management unit includes a real-time logic verification module, a data modification trace storage module, a blind maintenance and access control module, and a test progress panoramic dashboard module.

[0032] Preferably, the emergency handling and unblinding unit includes an emergency unblinding electronic application and approval module, a controlled information query module, and an emergency event recording and reporting module.

[0033] Preferably, the test configuration management unit is responsible for initializing the test parameters and center information.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] This invention uses a dynamic randomization algorithm to calculate the optimal group allocation in real time based on the baseline characteristics of enrolled subjects. By dynamically adjusting probabilities using the urn method, it effectively ensures the balance of key factors among groups, improving the scientific rigor and comparability of the trial. Furthermore, it automatically triggers precise blinding and replenishment instructions based on real-time inventory and consumption predictions. By optimizing distribution through solving the minimum cost flow problem, it significantly reduces drug waste and the risk of drug shortages, ensuring a dynamic balance of supply. The algorithm for verifying the logic of syndrome evolution in online visits automatically checks and provides feedback on the rationality of changes in syndrome before and after visits through a rule base, ensuring the continuity and logical reliability of TCM syndrome data. It also feeds back the updated syndrome information to the randomization unit in real time, forming a closed loop from allocation to feedback. This achieves intelligent and precise closed-loop management of the entire process from randomization and drug supply to data collection, greatly reducing drug waste. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the embodiments thereof. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0037] A central randomization allocation system for clinical trials includes a trial configuration management unit, a dynamic randomization allocation unit, a drug dynamic blinding supply chain optimization unit, an online visit data acquisition unit, a data quality control and security management unit, and an emergency handling and unblinding unit. The trial configuration management unit is responsible for initializing trial parameters and central information. The dynamic randomization allocation unit, based on the baseline characteristics of enrolled subjects (including TCM syndrome types), uses an adaptive algorithm to achieve real-time and balanced group allocation and synchronizes allocation instructions to subsequent units in real time. The drug dynamic blinding supply chain optimization unit adjusts the allocation based on real-time inventory status. The system generates precise drug blinding and distribution instructions to maintain a dynamic balance of drug supply across centers. The online visit data collection unit is responsible for executing the follow-up process, collecting efficacy and safety data (including dynamically changing syndrome information), and feeding this data back to the randomization unit to guide subsequent allocation. The data quality control and security management unit implements real-time logical verification and trace management of the entire process data to ensure data integrity and blinding security. The emergency handling and unblinding unit provides a controlled unblinding channel in emergency situations. The entire system thus constructs a closed-loop structure of "allocation-supply-execution-feedback-quality control", significantly improving the efficiency and quality of the trial.

[0038] The dynamic randomized allocation unit includes a stratification factor definition module, an adaptive random algorithm engine module, a random sequence generation and encryption module, an allocation execution and interface module, and a real-time monitoring module for allocation balance. The stratification factor definition module is used to set static stratification factors (such as research center, disease stage) and dynamic stratification factors (such as TCM syndrome type). The adaptive random algorithm engine module can integrate algorithms such as minimization method and urn method, and has a TCM syndrome type-subject status adaptive randomization algorithm to calculate the optimal allocation group for new participants based on the characteristic distribution of already enrolled participants, in order to minimize the imbalance between groups. The random sequence generation and encryption module generates unpredictable random allocation sequences based on the output of the algorithm engine and encrypts and stores them. The allocation execution and interface module receives enrollment requests from the online visit unit, executes the allocation, and synchronizes the results to the drug blinding unit and the online visit unit. The real-time monitoring module for allocation balance visualizes the balance of cases between groups under each stratification factor.

[0039] The TCM syndrome type-subject status adaptive randomization algorithm is responsible for calculating the optimal group in real time based on the subject characteristics (including TCM syndrome type) to ensure inter-group balance. The specific algorithm steps are as follows:

[0040] Step 1: Baseline data of newly enrolled subjects are used as the group distribution matrix M for each center and syndrome type. current Baseline data includes research center ID, disease stage, TCM four diagnostic methods information, and syndrome type;

[0041] Step 2: Convert the TCM four diagnostic methods information into standard syndrome codes using the built-in rule base, such as Liver Qi Stagnation and Spleen Deficiency Syndrome → Code_Z01. Imbalance calculation: Assuming there are experimental group (T) and control group (C), for the subject's "center + syndrome type" strata, calculate the inter-group difference values ​​DT and DC if the subject were assigned to group T or group C. The difference value calculation formula uses the range method.

[0042] D=∣N T,strata -N C,strata +∣P T,strata -P C,strata |;

[0043] Where N represents the number of people and P represents the mean of the key covariate;

[0044] Step 3: Adjust the probability using the urn method. If D T <D C If the probability of random assignment is higher, such as 0.8, then group T is assigned a higher probability; otherwise, group C is assigned a higher probability.

[0045] Step 4: Randomly generate group G based on the above probabilities. assign and encrypted random seed, G assign Send to the drug dynamic blinding supply chain optimization unit.

[0046] The drug dynamic blinding supply chain optimization unit includes a drug coding and blind management module, a central inventory intelligent monitoring module, a dynamic blinding trigger module, a logistics and distribution optimization module, and a drug dispensing and retrieval verification module. The drug coding and blind management module generates a unique traceability code for each smallest package of drug and establishes an initial encrypted association with the trial group information (blind). The central inventory intelligent monitoring module monitors the inventory and expiration dates of drugs corresponding to different groups and syndrome types in each branch center in real time. The dynamic blinding trigger module automatically triggers a blinding instruction when the inventory of a specific drug in a center falls below a preset threshold, and links with the drug coding module to generate a new batch blind code. The logistics and distribution optimization module integrates a dynamic blinding and replenishment algorithm based on consumption prediction. It predicts future demand based on the enrollment speed and dropout rate of each center, optimizes the distribution plan and route, and reduces drug waste due to expiration. The drug dispensing and retrieval verification module performs a secondary verification by scanning the drug traceability code and subject ID during dispensing to ensure accurate medication and records the drug flow chain. This unit is connected to the dynamic randomization allocation unit through allocation result instructions and interacts with the online visit data collection unit through dispensing application and confirmation signals.

[0047] The drug dynamic blinding supply chain optimization unit, acting as the system's "logistics" unit, receives allocation instructions and, combined with inventory forecasts, automatically generates replenishment and blinding instructions. The specific algorithm steps are as follows:

[0048] Step 1: Input the assignment group G from the randomization unit.assign Real-time inventory of each center S realtime Historical enrollment rate V enroll Shedding rate R dropout Drug shelf life (T) exp ;

[0049] Step 2: Predict the demand D for the next t days based on the moving average algorithm. pred Then we have:

[0050] D pred =(V enroll ×(1−R dropout ))×t;

[0051] Step 3: Calculate the safety stock threshold for the corresponding certificate type and group for this center:

[0052] S safe =D pred ×1.5;

[0053] Step 4: Compare with real-time inventory S realtime With S safe If S realtime safe This will trigger the blind coding and replenishment tasks;

[0054] Step 5: Call the random number generator to generate a blind sequence for the new batch of drugs, and encrypt and bind it with the drug traceability code to ensure that each drug has a unique code.

[0055] Step 6: Based on the urgency and logistics costs of each center, solve the minimum cost flow problem to generate the optimal delivery route;

[0056] Step 7: Output the blinding instruction (including the new drug blind code), logistics delivery note, and drug dispensing verification key, and send the drug dispensing application confirmation signal and the new drug blind code to the online visit data acquisition unit.

[0057] ​The online visit data acquisition unit includes an electronic medical record report customization module, a mobile follow-up and data entry module, a subject compliance assistance module, a syndrome evolution tracking and logic verification module, and an instant messaging and task reminder module. The electronic medical record report customization module supports customized eCRFs according to the protocol and specifically includes structured data collection fields for the four diagnostic methods of traditional Chinese medicine. The mobile follow-up and data entry module allows researchers to complete visits via mobile devices and uploads data in real time. The subject compliance assistance module provides medication reminders and patient report outcome collection functions. The syndrome evolution tracking and logic verification module records the TCM syndromes of each visit and has a built-in TCM four diagnostic methods information structure conversion algorithm to automatically prompt and verify the logical rationality of syndrome changes before and after visits. The instant messaging and task reminder module enables collaborative communication between researchers, monitors, and drug administrators. This unit calls the interface of the dynamic randomization allocation unit when subjects are enrolled, interacts with the drug dynamic blinding unit during each visit and medication dispensing, and transmits all collected data to the data quality control and security management unit in real time.

[0058] The online visit data collection unit collects treatment efficacy data and verifies the logic, feeding back the dynamically changing syndrome types to the randomization unit. A syndrome type evolution logic verification algorithm is used, with the specific algorithm steps as follows:

[0059] Step 1: Input the TCM diagnostic data for the current visit (tongue appearance, pulse appearance, symptom score) and the syndrome record from the previous visit. prev And the preliminary certificate type Z calculated during this visit current_raw ;

[0060] Step 2: Use TCM diagnostic algorithms (such as rule-based reasoning RBR) to convert the four diagnostic methods data into syndrome vectors and check Z. current_raw With Z prev Does the evolutionary relationship conform to the clinical logic of Traditional Chinese Medicine? For example, the direct transformation from "cold syndrome" to "heat syndrome" requires an intermediate transition or special inducing factor; otherwise, it should be marked as abnormal. IFZ prev =='Cold Syndrome' ANDZ current_raw =='Fever Syndrome' AND (No History of High Fever) THENFlag='Logical Doubt';

[0061] Step 3: If the verification passes, confirm Z. current The data will be packaged, and if the verification fails, a pop-up window will prompt the researcher to verify it, and the questioning log will be recorded.

[0062] Step 4: Confirm the Z current The system packages efficacy data, outputs structured efficacy data packages, syndrome evolution tags, and logical challenge reports, and includes Z... currentThe data is fed back in real time to the stratification factor definition module in the dynamic randomization allocation unit, which is used to update the "characteristic distribution of enrolled subjects" database when the next subject is enrolled, thereby achieving closed-loop adjustment.

[0063] The data quality control and security management unit includes a real-time logic verification module, a data modification trace storage module, a blind state maintenance and access control module, and a trial progress panoramic dashboard module. The real-time logic verification module performs range, logic, and jump checks on the entered data. Any data modification must be explained, and the values ​​before and after the modification, the operator, and the timestamp are all recorded by the data modification trace storage module and uploaded to the blockchain for decentralized storage to ensure immutability. The blind state maintenance and access control module strictly controls system user permissions and isolates blind data. All operations involving grouping information require secondary authentication and log recording. The trial progress panoramic dashboard module provides project managers with a multi-dimensional real-time view of enrollment progress, data quality, drug inventory, and other aspects. This unit acts as the central nervous system of the system, monitoring the data flow from the online visit unit and the operation logs of all other units.

[0064] The emergency response and unblinding unit comprises an emergency unblinding electronic application and approval module, a controlled information query module, and an emergency event recording and reporting module. In the emergency unblinding electronic application and approval module, researchers submit emergency unblinding applications online, which are then remotely approved electronically by a designated authorized person (such as the sponsor's medical director). After approval, the controlled information query module only releases the subject's group information to the applicant researcher, without disrupting the overall system blinding. The emergency event recording and reporting module automatically generates an emergency unblinding event report, linked to all data of the subject. This unit, responding to clinical needs in a controlled manner when necessary, is a crucial component of the system's safety design.

[0065] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0066] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A central randomization allocation system for clinical trials, characterized in that: The system includes a trial configuration management unit, a dynamic randomization allocation unit, a drug dynamic blinding supply chain optimization unit, an online visit data acquisition unit, a data quality control and security management unit, and an emergency handling and unblinding unit.

2. The central randomization allocation system for clinical trials according to claim 1, characterized in that: The dynamic randomization allocation unit includes a hierarchical factor definition module, an adaptive random algorithm engine module, a random sequence generation and encryption module, an allocation execution and interface module, and a real-time monitoring module for allocation balance.

3. A central randomization allocation system for clinical trials according to claim 2, characterized in that: The adaptive random algorithm engine module includes a TCM syndrome type-subject state adaptive randomization algorithm, the specific algorithm steps of which are as follows: Step 1: Baseline data of newly enrolled subjects are used as the group distribution matrix M for each center and syndrome type. current Baseline data includes research center ID, disease stage, TCM four diagnostic methods information, and syndrome type; Step 2: Convert the TCM four diagnostic methods information into standard syndrome codes using the built-in rule base. Assuming there are experimental group (T) and control group (C), for the center and syndrome type (strata) to which the subject belongs, calculate the inter-group difference values ​​DT and DC if the subject is assigned to group T or group C. The difference value calculation formula uses the range method: D=∣N T,strata −N C,strata +∣P T,strata −P C,strata ∣; Where N represents the number of people and P represents the mean of the key covariate; Step 3: Adjust the probability using the urn method. If D T <D C If the probability of random assignment is higher, such as 0.8, then group T is assigned a higher probability; otherwise, group C is assigned a higher probability. Step 4: Randomly generate group G based on the above probabilities. assign and encrypted random seed, G assign Send to the drug dynamic blinding supply chain optimization unit.

4. A central randomization allocation system for clinical trials according to claim 1, characterized in that: The drug dynamic blinding supply chain optimization unit includes a drug coding and blind management module, a central inventory intelligent monitoring module, a dynamic blinding trigger module, a logistics and distribution optimization module, and a drug distribution and recovery verification module.

5. A central randomization allocation system for clinical trials according to claim 2, characterized in that: The drug dynamic blinding supply chain optimization unit receives allocation instructions and, in conjunction with inventory forecasts, automatically generates replenishment and blinding instructions. The specific algorithm steps are as follows: Step 1: Input the assignment group G from the randomization unit. assign Real-time inventory of each center S realtime Historical enrollment rate V enroll Shedding rate R dropout Drug shelf life (T) exp ; Step 2: Predict the demand D for the next t days based on the moving average algorithm. pred Then we have: D pred =(V enroll ×(1−R dropout ))×t; Step 3: Calculate the safety stock threshold for the corresponding certificate type and group for this center: S safe =D pred ×1.5; Step 4: Compare with real-time inventory S realtime With S safe If S realtime safe This will trigger the blind coding and replenishment tasks;​ Step 5: Use a random number generator to generate a blind sequence for the new batch of drugs and encrypt and bind it with the drug traceability code; Step 6: Based on the urgency and logistics costs of each center, solve the minimum cost flow problem to generate the optimal delivery route; Step 7: Output the blinding instruction, logistics delivery note, and drug dispensing verification key, and send the drug dispensing application confirmation signal and new drug blind code to the online visit data acquisition unit.

6. A central randomization allocation system for clinical trials according to claim 1, characterized in that: The online visit data acquisition unit includes an electronic medical record report customization module, a mobile follow-up and data entry module, a subject compliance assistance module, a syndrome evolution tracking and logic verification module, and an instant communication and task reminder module.

7. A central randomization allocation system for clinical trials according to claim 1, characterized in that: The online visit data acquisition unit collects treatment efficacy data and verifies the logic, feeding back the dynamically changing syndrome types to the randomization unit. A syndrome type evolution logic verification algorithm is used, with the specific algorithm steps as follows: Step 1: Enter the TCM four diagnostic methods data for the current visit and the syndrome record from the previous visit. prev And the preliminary certificate type Z calculated during this visit current_raw ; Step 2: Use TCM diagnostic algorithms to convert the four diagnostic methods data into syndrome vectors and check Z-type patterns. current_raw With Z prev If the evolutionary relationship conforms to the clinical logic of traditional Chinese medicine, it is marked as abnormal. Step 3: If the verification passes, confirm Z. current The data will be packaged, and if the verification fails, a pop-up window will prompt the researcher to verify it, and the questioning log will be recorded. Step 4: Confirm the Z current The system packages efficacy data, outputs structured efficacy data packages, syndrome evolution tags, and logical challenge reports, and includes Z... current The feedback is sent in real time to the hierarchical factor definition module in the dynamic randomization allocation unit.

8. A central randomization allocation system for clinical trials according to claim 1, characterized in that: The data quality control and security management unit includes a real-time logic verification module, a data modification trace storage module, a blind maintenance and access control module, and a test progress panoramic dashboard module.

9. A central randomization allocation system for clinical trials according to claim 1, characterized in that: The emergency response and unblinding unit includes an emergency unblinding electronic application and approval module, a controlled information query module, and an emergency event recording and reporting module.

10. A central randomization allocation system for clinical trials according to claim 1, characterized in that: The test configuration management unit is responsible for initializing the test parameters and central information.