Ethical review method, device and storage medium for clinical trial documents
Patent Information
- Application Number
- CN202610717293.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-22
- Publication Date
- 2026-08-28
AI Technical Summary
审查文件类型繁杂，审核效率较低，而且在法规频繁更新的情况下，难以保证审核法规的合规有效
利用专项审核模型获取对应文件类型的结构化专项审核意见，可以缩短审核周期，在提高文件审核效率的同时，可以保证审核法规的有效合规，且统一同类型文件审核标准，降低人工主观差异带来的审核偏差，能够提升审核结果的稳定性和精准度。
Smart Images
Figure CN122658538A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of clinical trial ethics review technology, specifically to a method, device, and storage medium for ethical review of clinical trial documents. Background Technology
[0002] Ethical review of clinical trials safeguards the rights and safety of participants and is a prerequisite for ensuring the compliant conduct of clinical trials. Therefore, before a clinical trial, the ethics committee must conduct a comprehensive review of all trial documents to verify the scientific validity, ethical rationale, and regulatory compliance of the research. Currently, the review method primarily employs a manual ethical review model, which has the following shortcomings: The types of documents to be reviewed are complex, the review efficiency is low, and with frequent updates to regulations, it is difficult to ensure compliance and effectiveness of the reviewed regulations. Summary of the Invention
[0003] In view of the deficiencies in the existing technology, the technical problem to be solved by this application is: how to improve the efficiency of clinical trial ethics review while ensuring compliance and effectiveness of review regulations.
[0004] To achieve the above objectives, in a first aspect, embodiments of this application provide a method for ethical review of clinical trial documents, the method comprising the following steps: Analyze documents involved in the ethical review of clinical trials, obtain the text content, determine the document type based on the text content, and then extract structured information; Determine the specific audit task for the current document type based on the document type; Based on a pre-established regulatory knowledge base, obtain compliance evidence that matches the specific audit task; Using structured information, special audit tasks, and compliance evidence as input, a pre-trained special audit model generates structured special audit opinions for the corresponding document types.
[0005] In conjunction with the first aspect, in one implementation method, the training method of the specialized audit model includes: After obtaining structured document information of different document types, adding special audit task and compliance basis tags, and collecting known structured special audit opinions of the corresponding structured documents, the data source is used for training. From the training data source, according to the file type, obtain the corresponding structured file information, special audit tasks and compliance basis as input parameters; use the predicted structured special audit opinions as output parameters to construct and train the special audit model corresponding to the file type. When the output accuracy of the special audit model is greater than a preset threshold, the pre-trained special audit model is obtained.
[0006] In conjunction with the first aspect, in one implementation, the process of determining the special audit task based on the document type includes: Pre-define the key review points for different document types as specific review tasks; Establish a mapping relationship between different document types and specific audit tasks; Based on the mapping relationship, determine the specific audit task for the current file type.
[0007] In conjunction with the first aspect, in one implementation method, the process of establishing the regulatory knowledge base includes: Vectorize and store regulations, guidelines, and institutional standard operating procedures to obtain a regulatory knowledge base; When an update to the regulatory data source is detected, the regulatory knowledge base is updated in real time based on the updated information.
[0008] In conjunction with the first aspect, in one implementation method, the method of obtaining compliance evidence matching the special audit task includes: Based on the context of the special audit task and the structured document information, the semantic similarity with the content stored in the regulatory knowledge base is calculated, and the regulations with the highest semantic similarity are selected as the compliance basis for matching.
[0009] In conjunction with the first aspect, in one implementation, it further includes: After extracting the structured information, two file types are predefined as a combination. After determining the verification target of the combination, the logical consistency verification of the verification target is performed based on the structured file information corresponding to the combination, and a cross-file consistency verification report is generated.
[0010] In conjunction with the first aspect, in one implementation, during the process of generating a cross-document consistency verification report, all structured special review opinions are deduplicated, merged, and marked with regulatory source information, and a structured ethics review opinion report is generated by combining the cross-document consistency verification report.
[0011] In conjunction with the first aspect, in one implementation, after generating a structured ethics review opinion report, human feedback on the structured ethics review opinion report is obtained; Based on the document type, structured document information, corresponding special audit task, compliance basis, and audit feedback, generate a new training data source for the corresponding document type; Regularly train the specialized audit model for this document type using new training data sources.
[0012] Secondly, embodiments of this application provide an ethical review device for clinical trial documents. The ethical review device for clinical trial documents includes a processor, a memory, and an ethical review program for clinical trial documents stored in the memory and executable by the processor. When the ethical review program for clinical trial documents is executed by the processor, it implements the method provided in the first aspect.
[0013] Thirdly, embodiments of this application provide a computer-readable storage medium storing an ethical review procedure for clinical trial documents, wherein when the ethical review procedure for clinical trial documents is executed, the method provided in the first aspect is implemented.
[0014] Compared with the prior art, the advantages of this application are: By using specialized audit models to obtain structured specialized audit opinions for corresponding document types, the audit cycle can be shortened. While improving the efficiency of document audit, it can also ensure the effective compliance of audit regulations, unify the audit standards for the same type of document, reduce audit bias caused by human subjective differences, and improve the stability and accuracy of audit results. Attached Figure Description
[0015] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0016] Figure 1 This is a flowchart illustrating the ethical review method for clinical trial documents in this application embodiment; Figure 2 This is a schematic diagram of the hardware structure of the ethical review device for clinical trial documents involved in the embodiments of this application. Detailed Implementation
[0017] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0018] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.
[0019] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.
[0020] Firstly, this application provides an ethical review method for clinical trial documents, referring to... Figure 1 The method includes the following steps: Analyze the documents involved in the ethical review of clinical trials, obtain the text content, determine the document type based on the text content (such as research protocol, informed consent form, investigator's manual, case report form, recruitment materials, safety report, statistical analysis plan, etc.), and then extract structured information (such as trial name, sponsor, indication, study design type, subject population, primary endpoint, risk measures, etc.). Determine the specific audit task for the current document type based on the document type; Based on a pre-established regulatory knowledge base, obtain compliance evidence that matches the specific audit task; Using structured information, specific audit tasks, and compliance evidence as input, a pre-trained specific audit model generates structured specific audit opinions for the corresponding document type. The training method for this specific audit model includes: After obtaining structured document information of different document types, adding special audit tasks and compliance basis tags, and collecting known structured special audit opinions for the corresponding structured documents (which can be obtained through manual auditing), the data is used as a training data source. From the training data source, according to the file type, obtain the corresponding structured file information, special audit tasks and compliance basis as input parameters; use the predicted structured special audit opinions as output parameters to construct and train the special audit model corresponding to the file type. When the output accuracy of the special audit model (i.e. the semantic similarity between the predicted structured special audit opinion and the known structured special audit opinion, which can be calculated by the cosine similarity method) is greater than a preset threshold (>95%), the pre-trained special audit model is obtained.
[0021] By using specialized audit models to obtain structured specialized audit opinions for corresponding document types, the audit cycle can be shortened. While improving the efficiency of document audit, it can also ensure the effective compliance of audit regulations, unify the audit standards for the same type of documents, reduce audit bias caused by human subjective differences, and improve the stability and accuracy of audit results.
[0022] In one embodiment, the process of parsing documents involved in the ethical review of clinical trials and obtaining their text content includes: Recognize file formats (such as PDF, Word, Excel, scanned documents, images, etc.); The text content is obtained using a preset corresponding scheme for the file format. For example, for scanned documents and images, optical character recognition technology is used to obtain the text content.
[0023] It should be noted that the subjects' privacy information was anonymized before the file format was identified. This is to prevent the leakage of subjects' privacy.
[0024] In one embodiment, the process for determining the current document type based on the document type in the special audit task includes: Different document types have pre-defined key review points as specific audit tasks; taking informed consent forms, research plans, recruitment materials, and safety reports as examples, the corresponding key review points include: The key aspects of reviewing an informed consent form include the completeness of information disclosure, readability of language, guarantee of voluntariness, explanation of exit mechanisms, adequacy of risk disclosure, and explanation of fees and compensation. The review of the study protocol should focus on elements such as the rationality of the scientific design, the protection measures for the subjects, the risk-benefit ratio, the rationality of the inclusion and exclusion criteria, the scientific validity of the endpoint indicators, and the safety monitoring plan. The key aspects of reviewing recruitment materials include whether there is any improper inducement, the accuracy of information, the adequacy of risk warnings, and the compliance of contact information. The review of a safety report should focus on elements such as the timeliness of adverse event reporting, the reasonableness of causal relationship determination, the completeness of safety data, and the adequacy of risk control measures.
[0025] Establish a mapping relationship between different document types and special audit tasks. The special audit task for the current document type can be determined through the mapping relationship.
[0026] By pre-establishing a mapping relationship between the key review content of different document types, the core review elements of different document types can be accurately identified, avoiding omissions in the review process and making the review more targeted and professional, thereby improving the accuracy of the review of different document types.
[0027] In one embodiment, the process for establishing the aforementioned regulatory knowledge base includes: Regulations, guidelines, and institutional standard operating procedures are vectorized, encoded, and stored to obtain a regulatory knowledge base.
[0028] Furthermore, when updates to the regulatory data source are detected, the regulatory knowledge base is updated in real time based on the updated information. This ensures the effectiveness of the regulatory knowledge base information.
[0029] Based on this, the methods for obtaining compliance evidence that matches the specific audit task include: Based on the context of the special audit task and the structured document information, the semantic similarity with the content stored in the regulatory knowledge base is calculated, and the regulations with the highest semantic similarity are selected as the compliance basis for matching.
[0030] By building a regulatory knowledge base that can be updated in real time, the audit basis can always be aligned with the latest regulatory requirements. Furthermore, by matching compliance basis through semantic similarity calculation, the traditional method of manually checking regulations clause by clause can be replaced, which can effectively improve the efficiency of compliance audit.
[0031] Furthermore, multiple specialized audit models can be audited in parallel, further improving the efficiency of auditing multiple documents.
[0032] In one embodiment, this application provides a method for ethical review of clinical trial documents, referring to... Figure 1 The method also includes the following steps: After extracting structured information, two file types are predefined as a combination. Once the verification target for this combination is determined, logical consistency checks are performed on the target based on the structured file information corresponding to that combination, and a cross-file consistency check report (i.e., a report on content inconsistencies) is generated. The combination, i.e., the corresponding verification target, includes: The study protocol and informed consent form are combined, and the verification objectives include checking whether the risk disclosure in the informed consent form is completely consistent with the risk description in the study protocol, and checking the consistency of information such as the visit process, medication plan, and examination items. The study protocol and case report form were combined, and the validation objectives included checking whether the data collection items of the case report form fully covered the primary endpoint, secondary endpoints and safety indicators defined in the study protocol. The combination of research protocol and statistical analysis plan has the following verification objectives: to check whether the analytical methods in the statistical analysis plan match the endpoint definitions and statistical assumptions in the research protocol. The combination of the researcher's handbook and the research protocol is validated to determine whether the safety data in the researcher's handbook supports the risk-benefit assessment in the research protocol.
[0033] This cross-document comparison method avoids the shortcomings of independent review of single documents, thereby ensuring the logical consistency of the entire set of documents involved in the ethical review of clinical trials.
[0034] Furthermore, after deduplication, merging, and regulatory source tracing of all structured special review opinions, a structured ethics review opinion report is generated by combining the cross-document consistency verification report. This not only facilitates archiving and regulatory verification but also provides convenience for subsequent manual review.
[0035] Furthermore, after obtaining human feedback on the structured ethics review report, a new training data source for the corresponding document type is generated based on the document type, structured document information, corresponding special review task, compliance basis, and review feedback (as known structured special review opinions).
[0036] The special audit model for this document type is trained regularly (e.g., daily, weekly, monthly) using new training data sources (see the training method for the special audit model above).
[0037] This enables the specialized audit model to adapt and grow, continuously ensuring audit efficiency and accuracy.
[0038] Secondly, this application provides an ethical review device for clinical trial documents. The ethical review device for clinical trial documents can be a personal computer (PC), a laptop computer, a server, or other devices with data processing capabilities.
[0039] Reference Figure 2 , Figure 2 This is a schematic diagram of the hardware structure of the ethics review device for clinical trial documents involved in the embodiments of this application. In this embodiment, the ethics review device for clinical trial documents may include a processor, a memory, a communication interface, and a communication bus.
[0040] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.
[0041] The communication interface includes input / output (I / O) interfaces, physical interfaces, and logical interfaces used for interconnecting devices within the ethical review equipment for clinical trial documents, as well as interfaces used for interconnecting the ethical review equipment for clinical trial documents with other devices (such as other computing devices or user devices). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user devices can be displays, keyboards, etc.
[0042] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.
[0043] The processor can be a general-purpose processor, which can call the ethics review program for clinical trial documents stored in memory and execute the ethics review method for clinical trial documents provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the ethics review program for clinical trial documents is called can be referred to in the various embodiments of the ethics review method for clinical trial documents in this application, and will not be repeated here.
[0044] Those skilled in the art will understand that Figure 2 The hardware structure shown does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.
[0045] Thirdly, embodiments of this application also provide a computer-readable storage medium.
[0046] The computer-readable storage medium of this application stores an ethical review procedure for clinical trial documents, wherein when the ethical review procedure for clinical trial documents is executed by a processor, the steps of the ethical review method for clinical trial documents as described above are implemented.
[0047] The method for implementing the ethical review procedure for clinical trial documents can be found in various embodiments of the ethical review method for clinical trial documents in this application, and will not be repeated here.
[0048] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.
[0049] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.
[0050] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.
[0051] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.
[0052] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.
[0053] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.
[0054] The above are merely specific embodiments of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the scope of the claims.
Claims
1. A method for ethical review of clinical trial documents, characterized in that, The method includes the following steps: Analyze documents involved in the ethical review of clinical trials, obtain the text content, determine the document type based on the text content, and then extract structured information; Determine the specific audit task for the current document type based on the document type; Based on a pre-established regulatory knowledge base, obtain compliance evidence that matches the specific audit task; Using structured information, special audit tasks, and compliance evidence as input, a pre-trained special audit model generates structured special audit opinions for the corresponding document types.
2. The ethical review method for clinical trial documents as described in claim 1, characterized in that: The training methods for the special audit model include: After obtaining structured document information of different document types, adding special audit task and compliance basis tags, and collecting known structured special audit opinions of the corresponding structured documents, the data source is used for training. From the training data source, according to the file type, obtain the corresponding structured file information, special audit tasks and compliance basis as input parameters; use the predicted structured special audit opinions as output parameters to construct and train the special audit model corresponding to the file type. When the output accuracy of the special audit model is greater than a preset threshold, the pre-trained special audit model is obtained.
3. The ethical review method for clinical trial documents as described in claim 1, characterized in that: The process for determining the current file type for a special audit task based on the file type includes: Pre-define the key review points for different document types as specific review tasks; Establish a mapping relationship between different document types and specific audit tasks; Based on the mapping relationship, determine the specific audit task for the current file type.
4. The ethical review method for clinical trial documents as described in claim 1, characterized in that: The process of establishing the regulatory knowledge base includes: Vectorize and store regulations, guidelines, and institutional standard operating procedures to obtain a regulatory knowledge base; When an update to the regulatory data source is detected, the regulatory knowledge base is updated in real time based on the updated information.
5. The ethical review method for clinical trial documents as described in claim 1, characterized in that: The methods for obtaining compliance evidence that matches the special audit task include: Based on the context of the special audit task and the structured document information, the semantic similarity with the content stored in the regulatory knowledge base is calculated, and the regulations with the highest semantic similarity are selected as the compliance basis for matching.
6. The ethical review method for clinical trial documents as described in claim 2, characterized in that: Also includes: After extracting the structured information, two file types are predefined as a combination. After determining the verification target of the combination, the logical consistency verification of the verification target is performed based on the structured file information corresponding to the combination, and a cross-file consistency verification report is generated.
7. The ethical review method for clinical trial documents as described in claim 6, characterized in that: During the generation of cross-document consistency verification reports, all structured special review opinions are deduplicated, merged, and marked with regulatory source information. Combined with the cross-document consistency verification reports, a structured ethics review opinion report is generated.
8. The ethical review method for clinical trial documents as described in claim 7, characterized in that: After generating a structured ethics review report, obtain human feedback on the structured ethics review report; Based on the document type, structured document information, corresponding special audit task, compliance basis, and audit feedback, generate a new training data source for the corresponding document type; Regularly train the specialized audit model for this document type using new training data sources.
9. An ethical review device for clinical trial documents, characterized in that, The ethical review device for clinical trial documents includes a processor, a memory, and an ethical review program for clinical trial documents stored in the memory and executable by the processor, wherein when the ethical review program for clinical trial documents is executed by the processor, the steps of the ethical review method for clinical trial documents as described in any one of claims 1 to 8 are implemented.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an ethical review procedure for clinical trial documents, wherein when the ethical review procedure for clinical trial documents is executed, the steps of the ethical review method for clinical trial documents as described in any one of claims 1 to 8 are implemented.