CN112785256A - Real-time assessment method and system for clinical endpoint events in clinical trials - Google Patents
Real-time assessment method and system for clinical endpoint events in clinical trials Download PDFInfo
- Publication number
- CN112785256A CN112785256A CN202110050570.1A CN202110050570A CN112785256A CN 112785256 A CN112785256 A CN 112785256A CN 202110050570 A CN202110050570 A CN 202110050570A CN 112785256 A CN112785256 A CN 112785256A
- Authority
- CN
- China
- Prior art keywords
- clinical
- evaluation
- data
- review
- event
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
Abstract
The invention relates to a real-time assessment method of clinical endpoint events in clinical trials, which comprises the following steps: acquiring a project data file comprising clinical data of a subject and a clinical endpoint event table according to a test project application, and storing the project data file into a source file database; configuring an endpoint event review table of a test project according to the clinical endpoint event table, and setting reviewers; the method comprises the following steps that a reviewer obtains clinical data of a subject needing manual review in a test project, reviews an end point event review table, signs and confirms the end point event review table and submits the end point event review table to obtain a final end point event table of first review; and comparing and screening the results of the final end point event table of all the first-time reviews, and when the comparison result has the identification of the inconsistent field, establishing a review meeting by the reviewer, performing second-time review on the data of the subject marked as inconsistent in the first-time review result, and obtaining a final review result. The method effectively improves the evaluation efficiency and improves the evaluation process.
Description
Technical Field
The invention belongs to the technical field of medical treatment, and particularly relates to a real-time evaluation method and an evaluation system for a clinical endpoint event in a clinical test.
Background
The Clinical endpoint Event review Committee (CEC) is a group independent of the Clinical trial population that determines the endpoint Event expected for a trial according to a uniform review standard to see if it meets the standard specified in the trial protocol.
In large-scale, multi-center, randomized clinical trials, CEC is an independent committee planned and established during the design phase of a trial protocol, so that during the course and at the end of a trial, the expected end-point events of the trial are judged under uniform review criteria to see if they meet the criteria specified in the trial protocol. For example, in clinical trials in the area of oncology or cardiovascular, evaluation of efficacy and safety can not be obtained from simple objectively tested laboratory sheets, but relies on a number of test criteria, including: laboratory reports, electrocardiogram, CT reports, optical disks and reports for catheterization, surgical records, medical history, etc., and then the conclusion of the test endpoint is obtained after the clinical experience in the field is comprehensively considered. In this case, in the clinical trial, since the researchers at the respective study centers have inconsistent understanding of the event definition of the project end point, different experiences, and inconsistent understanding of the classification criteria, and are liable to be subject to a large error, an independent CEC review board is established for the event review. The review board consists of odd numbers of people, and generally 1 chairman and several boards are arranged, but the total number of people is ensured to be odd so as to draw a conclusion.
The CEC review process in the prior art comprises: and aiming at the clinical data of the testee, filling a clinical endpoint event table by a doctor, submitting the relevant clinical data of the testee by the doctor when the doctor judges that the CEC review is needed, providing a CEC review board for reviewing the content of the endpoint event table to obtain a final endpoint event table, and selecting the endpoint event table of the review board as a final review result when the reviewed endpoint event table is different from the endpoint event table assessed by the doctor.
However, most of the domestic CECs are under-line review, and the under-line operation is easy to cause the transfer of data and the aggregated discussion of reviewers, so that the setting of the CEC is difficult to reach the standard, the review program is unreasonable, and the CEC is similar to the nominal CEC even if the CEC is set; therefore, a regular review process is necessary with uniform review standards.
Disclosure of Invention
In order to solve the technical problems, the invention adopts a real-time assessment method and a system thereof for clinical endpoint events in clinical tests.
In order to achieve the purpose, the invention adopts the technical scheme that
A method for real-time assessment of clinical endpoint events in a clinical trial comprising the steps of:
acquiring a project data file comprising clinical data of a subject and a clinical endpoint event table according to a test project application, and storing the project data file into a source file database;
step two, configuring an endpoint event review table of the test project according to the clinical endpoint event table, and setting reviewers; wherein the reviewers consist of 1 position chairman and an even position committee;
thirdly, the reviewer obtains the clinical data of the subject needing manual review in the test project, reviews the end point event review table, signs and confirms the end point event review table and submits the end point event review table to obtain a final end point event table of the first review;
comparing and screening the results of the final end point event table of all the first evaluation, and directly obtaining evaluation results when the fields of the comparison results are consistent; when the comparison result has the identification of the inconsistent field, going to step five;
and fifthly, the reviewer establishes a review meeting, carries out secondary review on the data of the subject marked as inconsistent in the primary review result, and obtains a final review result.
An improved technical scheme is that the step five specifically comprises the following steps: establishing a review meeting by the chairman and the committee together, extracting and discussing the subject data marked as inconsistent in the first review result in an online or offline meeting mode, and obtaining a final review result; when the review committee reviews the comments disagreement, the chairman assesses to obtain a final result.
According to an improved technical scheme, in the first step, a mode of applying for data acquisition by a test project comprises the following steps: the EDC system is docked, and clinical test data and a clinical endpoint event table corresponding to the system are obtained according to the test items; or the clinical test data and the clinical endpoint event list are input into a standard form page through the human-computer interaction module.
In another improved technical scheme, the method further comprises the following steps of setting the roles of the authorities of the reviewers: when the reviewers are committees, the reviewers have the permission to view the overview of the review progress of the reviewers and the permission to modify the review content of the reviewers; when the reviewer is the chairman, the system has the permission of viewing the contents of the review table finished and submitted by the reviewer and the overview permission of viewing the review progress of the reviewer.
According to an improved technical scheme, the method further comprises the step of recording and managing the business operation of the reviewers in real time, wherein the management comprises query, addition, modification and deletion operations.
The technical improvement of the second aspect of the method for evaluating the clinical endpoint event in the clinical test adds the automatic evaluation function of the system, and specifically, the method also comprises the following steps before the step two: the system evaluates the clinical endpoint events of all subjects by establishing a machine learning model for clinical endpoint event evaluation;
the method comprises the following steps:
selecting characteristics based on disease item diagnosis standards to obtain disease item clinical evaluation standards, and constructing a machine learning model for clinical endpoint event evaluation;
extracting clinical test data of a subject of a target project in a source file database, and carrying out normalization processing on the clinical test data to form unified structured data;
selecting end point clinical evaluation event data of a plurality of target items as training sample data;
inputting subject clinical test data of the target project as test sample data based on the expert evaluation rule table data of the target project;
inputting training sample data and test sample data into a machine learning model for clinical endpoint event evaluation; after multiple times of iterative training, outputting the evaluation result of the clinical endpoint event of the subject; and when the evaluation result of the clinical endpoint time evaluated by the system is objected, setting the clinical evaluation event corresponding to the subject as a subject clinical evaluation event needing manual evaluation, and entering the third step.
An improved solution, said extracting clinical trial data comprising extracting clinical data from a structured, unstructured and/or semi-structured data source extracting clinical data records of said subject; and performing natural language processing on the unstructured data and the semi-structured data, and extracting entity and patient information through a CRF (fuzzy C) model to form structured data.
In another improved technical scheme, the inputting training sample data and test sample data into a machine learning model for clinical endpoint event assessment, and outputting the clinical endpoint event assessment result of the subject after multiple iterative training specifically includes:
respectively inputting training sample data and test sample data into N candidate machine learning submodels in a machine learning model according to a ratio of 8:2, counting output results of all the candidate machine learning submodels, and judging whether a plurality of output results are consistent according to the output results; disagreement in results is objected;
the establishment method of the candidate machine learning submodel comprises the following steps: repeating the corresponding clinical evaluation event sample set M for N times, and randomly drawing M samples each time when the clinical evaluation event sample set M is replaced to obtain N sample sets, wherein each sample set comprises M samples; and establishing N corresponding machine learning submodels for the obtained N sample sets.
The invention also discloses a real-time evaluation system of the clinical terminal event in the clinical test, which comprises a project setting subsystem and a terminal event evaluation subsystem which are connected by bidirectional data;
the project setting subsystem comprises
The project data acquisition module is configured to be docked with the EDC system and acquire clinical test data and a clinical endpoint event table of the system according to the test projects;
the source file data storage module is used for storing the acquired project data files comprising clinical data and a clinical endpoint event table into a source file database;
the evaluation personnel setting module is configured to configure an endpoint event evaluation table of the test project according to the clinical endpoint event table and set evaluation personnel at the same time; the reviewers consist of 1 position chairman and an even position committee;
and the evaluation object permission setting module is configured to set the roles of the permissions of the evaluation personnel: when the reviewers are committees, the reviewers have the permission to view the overview of the review progress of the reviewers and the permission to modify the review content of the reviewers; when the reviewer is the chairman, the reviewer has the permission to view the contents of the review table finished and submitted for review in the reviewer and the overview permission to view the review progress of the reviewer;
the review progress overview module is configured to supervise and overview the review progress of the test project by the reviewers;
the endpoint event review subsystem includes
The first evaluation module is configured to enable an evaluator to obtain clinical data of a subject needing manual evaluation in a test project, evaluate an end point event evaluation table, sign and confirm the end point event evaluation table and submit the end point event evaluation table to obtain a final end point event table of the first evaluation;
the initial evaluation result processing module is configured to compare and screen the results of the final end point event table of all the first evaluations, and when the fields of the comparison results are consistent, the evaluation results are directly obtained; when the comparison result has the identification of the inconsistent field, entering a second evaluation stage;
the second evaluation module is configured to carry out second evaluation on the data of the subjects marked as inconsistent in the first evaluation result according to the establishment of an evaluation meeting of the evaluation personnel, and obtain a final evaluation result;
and the offline review module is configured to export inconsistent subject data when the first review result is inconsistent.
In an improved technical scheme, the system further comprises an automatic clinical event assessment subsystem;
the clinical event automated assessment subsystem comprises.
The decision knowledge base establishing module is configured to arrange clinical terminal evaluation frames corresponding to different disease items based on disease item diagnosis guidelines to form an evaluation decision knowledge base;
a data processing module configured to extract clinical trial data of the subject in the source file database and form structured data;
the data input module is used for inputting the structured clinical test data into the machine model;
establishing a machine learning model configured to establish a machine learning model of a subject clinical evaluation result based on expert evaluation data and a plurality of clinical evaluation event data of the evaluation decision knowledge base;
a result output module configured to output a clinical endpoint event assessment result of the subject after a plurality of iterative trainings of the model;
a result evaluation module configured to evaluate the output bed end point event evaluation result; when the evaluation result of the clinical endpoint time evaluated by the system is objected, the clinical evaluation event of the corresponding subject is set as a subject clinical evaluation event needing manual evaluation.
Compared with the prior art, the technical scheme of the invention can obtain the following beneficial effects:
the invention provides a real-time online evaluation method for clinical endpoint events in clinical tests, which does not need online data distribution, ensures the fairness and justice of the evaluation process through the setting of evaluation objects, and directly gives a conclusion when the evaluation results are consistent; the offline discussion is performed only when the evaluation results are inconsistent, so that the evaluation efficiency is effectively improved, and the evaluation process is more complete; the method realizes CEC operation in a systematized mode. In addition, the system of the invention records and manages each operation of the reviewer in real time, and carries out systematization processing on the filing, storage and the like of the data file, thereby better perfecting the review work of the system.
The method also realizes the automatic evaluation and decision support of the system, and based on big data analysis and learning technology thereof, the method integrates a medical knowledge base and expert experience to construct an automatic evaluation machine learning model, thereby more scientifically and efficiently assisting in finishing evaluation work; and when the automatic evaluation result shows objection, the evaluation work is further perfected through the on-line evaluation of medical staff.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained based on these drawings without inventive efforts.
FIG. 1 is a flow chart of a first embodiment of a method for real-time assessment of a clinical endpoint event in a clinical trial;
FIG. 2 is a flow chart of a second embodiment of a method for real-time assessment of a clinical endpoint event in a clinical trial;
FIG. 3 is a schematic diagram of a first embodiment of a system for real-time assessment of a clinical endpoint event in a clinical trial;
FIG. 4 is a schematic diagram of a second embodiment of a system for real-time assessment of clinical endpoint events in a clinical trial.
Detailed Description
In order to make the purpose, technical solution and beneficial effects of the present application more clear and more obvious, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
An embodiment of the method for real-time assessment of clinical endpoint events in a clinical trial of the present invention is described below with reference to the accompanying drawings. As shown in figure 1, comprises
Acquiring project data files including clinical data and a clinical endpoint event table according to a test project application, and storing the project data files into a source file database;
the above data acquisition method includes: the method comprises the steps of docking an EDC system, and acquiring clinical test data and a clinical endpoint event table of the EDC system according to test items; or the clinical test data and the clinical endpoint event list are input into a standard form page through the human-computer interaction module. Wherein the clinical test data comprises date, subject number, sex, age, blood pressure, blood sugar, blood routine test result, body fluid test result, secretion test result, metabolism test result, imaging test result and the like. The set source file database provides the on-line checking function of the clinical data of the testees, and the clinical experiment data of the testees can be derived through the database.
The fields of the clinical endpoint event table are configured according to specific project conditions, and the clinical data variables are different according to different projects; for example, in the case of cardiovascular type highlighting, the project data includes "whether revascularization occurs", "whether stroke occurs", "whether unstable angina occurs" …, and the like, which are set by the occurrence of different diseases.
Step two, configuring an endpoint event review table of the test project according to the clinical endpoint event table, and setting reviewers; the number of the reviewers is ensured to be odd, and an effective conclusion can be conveniently drawn.
Thirdly, the reviewer obtains the clinical data of the subject needing manual review in the test project, reviews the end point event review table, signs and confirms the end point event review table and submits the end point event review table to obtain a final end point event table of the first review;
in order to ensure the confidentiality and the effectiveness of the evaluation process, each evaluation personnel (including a chairman and a committee) needs to input a user password and submit the user password through an electronic signature after evaluation, and then the evaluation personnel can be determined as the completion of the evaluation work.
In the step, after the committee finishes the evaluation, the evaluation result needs to be signed, and the signed evaluation result enters the next step; the unsigned review result cannot proceed to the next step.
Comparing and screening the results of the final end point event table of all the first evaluation, and directly obtaining evaluation results when the fields of the comparison results are consistent; when the comparison result has the identification of the inconsistent field, going to step five;
after each committee and chairman submit the current final end point event list, the system automatically judges whether the fields of each review result are consistent by capturing key part keywords, for example, one committee selects 'possibly relevant'; if the other committees are selected from "less relevant", the results are judged to be inconsistent.
And fifthly, the reviewer establishes a review meeting, carries out secondary review on the data of the subject marked as inconsistent in the primary review result, and obtains a final review result.
In the second evaluation stage, the evaluation test aimed by the co-establishment of the chairman and the committee is a subject test with inconsistent fields evaluated in the fourth step; the review will be done either online or offline. Wherein, the online review meeting is that the reviewers (including the chairman and the committee) carry out the second review on the data of the subjects marked as inconsistent in the first review result in a video or online teleconference mode; each committee sets forth reasons and considered elements for the review results given by the committee, and final results are given after the review and research through the review party; and when the result still has objection, the chairman has final judgment right and gives a final conclusion. And the other off-line review meeting is used for exporting the test data of the testees with inconsistent fields, the review meeting adopts an off-line meeting for secondary review, and the review rule is the same as that of the on-line review meeting.
In some examples, the method further comprises role-setting the reviewer's permissions: when the reviewers are committees, the reviewers have the permission to view the overview of the review progress of the reviewers and the permission to modify the review content of the reviewers; when the reviewer is the chairman, the system has the permission of viewing the contents of the review table finished and submitted by the reviewer and the overview permission of viewing the review progress of the reviewer. And by setting different role authorities, a RBAC authority control system is completed, so that the evaluation process is safer.
In some examples, the method further comprises recording and managing in real-time the reviewer's business operations, including query, add, modify, and delete operations. The management interface of the operation log provides a trace function for the system, and a user can directly trace information in the management interface, so that the management is more rigorous.
The method is to use the on-line reviewer to review the clinical events of the subjects to be reviewed, and when a large amount of clinical data of the subjects is faced before, a large amount of medical workers are needed to review the clinical test data of each subject for the first time. In actual operation, the clinical test data evaluation efficiency is not high and problems easily occur due to insufficient personnel, heavy workload and the like of clinicians, so the invention also analyzes and evaluates the clinical data of the testee by arranging an automatic auxiliary clinical evaluation system so as to improve the clinical work efficiency.
In the improvement of the second aspect of the method of the present invention, the second step further comprises: the system assesses the clinical endpoint events for all subjects by building a machine learning model of clinical endpoint event assessment.
As shown in fig. 2, the specific method comprises the following steps:
1. selecting characteristics based on disease item diagnosis standards to obtain disease item clinical evaluation standards, and constructing a machine learning model for clinical endpoint event evaluation;
the disease item diagnosis guideline refers to medical guideline recommendations which are summarized according to the latest version of medical guideline, and the contents of the medical guideline include: important influencing factors of diseases, diagnostic evaluation indexes, special populations, scheme types and the like; for different types of disease items, after clinical trial drugs or medical means are applied, the evaluation indexes of the body of the subject are checked and diagnosed, such as laboratory examination, imaging examination, and the like, so as to form an evaluation framework corresponding to the disease items.
2. Extracting clinical test data of a subject of a target project in a source file database, and carrying out normalization processing on the clinical test data to form unified structured data;
extracting clinical trial data comprises extracting clinical data from structured, unstructured and/or semi-structured data sources that extract clinical data records of the subject; and performing natural language processing on the unstructured data and the semi-structured data, and extracting entity and patient information through a CRF (fuzzy C) model to form structured data.
3. Selecting end point clinical evaluation event data of a plurality of target items as training sample data;
4. inputting subject clinical test data of the target project as test sample data based on the expert evaluation rule table data of the target project;
5. inputting training sample data and test sample data into a machine learning model for clinical endpoint event evaluation, and outputting a clinical endpoint event evaluation result of a subject after repeated iterative training; and when the evaluation result of the clinical endpoint time evaluated by the system is objected, setting the clinical evaluation event of the corresponding subject as a subject clinical evaluation event needing manual evaluation, and entering the third step in the first scheme.
Inputting training sample data and test sample data into a machine learning model for clinical endpoint event assessment, and outputting a clinical endpoint event assessment result of a subject after multiple iterative training specifically comprises:
respectively inputting training sample data and test sample data into N candidate machine learning submodels in a machine learning model according to a ratio of 8:2, counting output results of all the candidate machine learning submodels, and judging whether a plurality of output results are consistent according to the output results; disagreement in results is objected;
the establishment method of the candidate machine learning submodel comprises the following steps: repeating the corresponding clinical evaluation event sample set M for N times, and randomly drawing M samples each time when the clinical evaluation event sample set M is replaced to obtain N sample sets, wherein each sample set comprises M samples; and establishing N corresponding machine learning submodels for the obtained N sample sets. For example, if N is 100, if the outputs of 90 candidate machine learning sub-modules are "potentially correlated" and the outputs of 10 candidate machine learning sub-modules are "potentially correlated", the results of the candidate machine learning model outputs are objectified.
Compared with the prior art, the embodiment of the invention establishes the clinical test evaluation model of 'expert experience + real clinical data', obtains the clinical test evaluation result of the subject through the matching decision of a plurality of medical staff evaluation examples and system decision design schemes, and further can improve the efficiency and accuracy of the clinical event evaluation result. The invention can take the recommendation results of the machine learning methods as an independent element, set the reliability of the evidence, and support the final decision by combining the review of the reviewers when the structure of the objection appears, thereby improving the accuracy of the review.
An example of another aspect of the invention is a system for real-time assessment of clinical endpoint events in a clinical trial.
A system for real-time assessment of clinical endpoint events in a clinical trial as shown in fig. 3, comprising a project setup subsystem and an endpoint event review subsystem that are in bidirectional data communication;
the project setting subsystem comprises
The project data acquisition module is configured to be docked with the EDC system and acquire clinical test data and a clinical endpoint event table of the system according to the test projects;
the source file data storage module is used for storing the acquired project data files comprising clinical data and a clinical endpoint event table into a source file database;
the evaluation personnel setting module is configured to configure an endpoint event evaluation table of the test project according to the clinical endpoint event table and set evaluation personnel at the same time; the reviewers consist of 1 position chairman and an even position committee;
and the evaluation object permission setting module is configured to set the roles of the permissions of the evaluation personnel: when the reviewers are committees, the reviewers have the permission to view the overview of the review progress of the reviewers and the permission to modify the review content of the reviewers; when the reviewer is the chairman, the reviewer has the permission to view the contents of the review table finished and submitted for review in the reviewer and the overview permission to view the review progress of the reviewer;
the review progress overview module is configured to supervise and overview the review progress of the test project by the reviewers;
the endpoint event review subsystem includes
The first evaluation module is configured to enable an evaluator to obtain clinical data of a subject needing manual evaluation in a test project, evaluate an end point event evaluation table, sign and confirm the end point event evaluation table and submit the end point event evaluation table to obtain a final end point event table of the first evaluation;
the initial evaluation result processing module is configured to compare and screen the results of the final end point event table of all the first evaluations, and when the fields of the comparison results are consistent, the evaluation results are directly obtained; when the comparison result has the identification of the inconsistent field, entering a second evaluation stage;
the second evaluation module is configured to carry out second evaluation on the data of the subjects marked as inconsistent in the first evaluation result according to the establishment of an evaluation meeting of the evaluation personnel, and obtain a final evaluation result;
and the offline review module is configured to export inconsistent subject data when the first review result is inconsistent.
The multi-module implementation quality management preposition system and the whole-process online real-time management clinical test review process have the following advantages:
1) checking online data: provides the function of checking the clinical original data of the online subjects;
2) the process is more perfect: the CEC system of CCRF corporation does not give a final review result;
3) the method supports the viewing of various original data, such as audio, pictures, documents and the like;
4) the logic setting is added, so that the index folding function is realized, and the humanization is realized;
5) the working area and the submitting area are isolated, and the result is clear;
6) the function of leaving marks is achieved, and the method is more strict;
7) the system has a complete RBAC authority control system, and is safer;
8) the system has a review progress overview function, can master the review progress and performs team management;
9) complying with an industry data standard;
10) may communicate with the EDC system.
In a development of the second aspect of the system according to the invention, the system further comprises an automated clinical event assessment subsystem; as shown in fig. 4.
The clinical event automated assessment subsystem comprises
The model building module is configured to select features based on disease item diagnosis standards to obtain disease item clinical evaluation standards and build a machine learning model for clinical endpoint event evaluation;
the data processing module is configured to extract clinical test data of the subjects of the target items in the source file database, and normalize the clinical test data to form unified structured data;
the data partitioning module is configured to select endpoint clinical assessment event data of a plurality of target projects as training sample data; selecting clinical test data of a subject of a target project to be tested as test sample data;
a machine training module configured to input training sample data and test sample data into a machine learning model for clinical endpoint event assessment; after multiple times of iterative training, outputting the evaluation result of the clinical endpoint event of the subject;
a result evaluation module configured to evaluate the output bed end point event evaluation result; when the evaluation result of the clinical endpoint time evaluated by the system is objected, the clinical evaluation event of the corresponding subject is set as a subject clinical evaluation event needing manual evaluation.
It is to be understood that the systems and methods described herein in accordance with the present invention may be implemented in various forms of hardware, software, firmware, special purpose processors, or a combination thereof. In one exemplary embodiment of the invention, the systems and methods described herein are implemented in software as an application program comprising program instructions that are tangibly embodied on one or more program storage devices (e.g., floppy disk, RAM, CDROM, DVD, ROM, and flash memory) and executable by any device or machine comprising suitable architecture.
It is to be further noted that, because the system modules and method steps depicted in the accompanying figures can be implemented in software, the actual connections between the system components (or the flow of the process steps) may differ depending upon the manner in which the application is programmed. Given the above teachings, one of ordinary skill in the related art will be able to contemplate these and similar implementations or configurations of the present invention.
The above-described embodiments do not limit the scope of the present invention. Any modification, equivalent replacement, and improvement made within the spirit and principle of the above-described embodiments should be included in the protection scope of the technical solution.
Claims (10)
1. A method for real-time assessment of clinical endpoint events in a clinical trial comprising
Acquiring a project data file comprising clinical data of a subject and a clinical endpoint event table according to a test project application, and storing the project data file into a source file database;
step two, configuring an endpoint event review table of the test project according to the clinical endpoint event table, and setting reviewers; wherein the reviewers consist of 1 position chairman and an even position committee;
thirdly, the reviewer obtains the clinical data of the subject needing manual review in the test project, reviews the end point event review table, signs and confirms the end point event review table and submits the end point event review table to obtain a final end point event table of the first review;
comparing and screening the results of the final end point event table of all the first evaluation, and directly obtaining evaluation results when the fields of the comparison results are consistent; when the comparison result has the identification of the inconsistent field, going to step five;
and fifthly, the reviewer establishes a review meeting, carries out secondary review on the data of the subject marked as inconsistent in the primary review result, and obtains a final review result.
2. The method for real-time assessment of clinical endpoint events in a clinical trial according to claim 1, wherein step five is specifically: a chairman and a committee jointly establish a review meeting, and subject data marked as inconsistent in the first review result is extracted and discussed in an online or offline meeting mode to obtain a final review result; when the review committee reviews the comments disagreement, the chairman assesses to obtain a final result.
3. The method according to claim 1, wherein the first trial project application collects data by a method comprising: the EDC system is docked, and clinical test data and a clinical endpoint event table corresponding to the system are obtained according to the test items; or the clinical test data and the clinical endpoint event list are input into a standard form page through the human-computer interaction module.
4. The method for real-time assessment of a clinical endpoint event in a clinical trial of claim 1, further comprising role setting of reviewers' permissions: when the reviewers are committees, the reviewers have the permission to view the overview of the review progress of the reviewers and the permission to modify the review content of the reviewers; when the reviewer is the chairman, the system has the permission of viewing the contents of the review table finished and submitted by the reviewer and the overview permission of viewing the review progress of the reviewer.
5. The method for real-time assessment of clinical endpoint events in a clinical trial of claim 1, further comprising recording and managing in real-time the business operations of the reviewers, including query, add, modify, and delete operations.
6. The method for real-time assessment of clinical endpoint events in a clinical trial of claim 1, wherein step two is preceded by: the system evaluates the clinical endpoint events of all subjects by establishing a machine learning model for clinical endpoint event evaluation;
the method comprises the following steps:
selecting characteristics based on disease item diagnosis standards to obtain disease item clinical evaluation standards, and constructing a machine learning model for clinical endpoint event evaluation;
extracting clinical test data of a subject of a target project in a source file database, and carrying out normalization processing on the clinical test data to form unified structured data;
selecting end point clinical evaluation event data of a plurality of target items as training sample data;
inputting subject clinical test data of the target project as test sample data based on the expert evaluation rule table data of the target project;
inputting training sample data and test sample data into a machine learning model for clinical endpoint event evaluation, and outputting a clinical endpoint event evaluation result of a subject after repeated iterative training; and when the evaluation result of the clinical endpoint time evaluated by the system is objected, setting the clinical evaluation event corresponding to the subject as a subject clinical evaluation event needing manual evaluation, and entering the third step.
7. The method of real-time assessment of a clinical endpoint event in a clinical trial of claim 6, wherein the extracting clinical trial data comprises extracting clinical data from a structured, unstructured and/or semi-structured data source extracting the subject's clinical data record; and performing natural language processing on the unstructured data and the semi-structured data, and extracting entity and patient information through a CRF (fuzzy C) model to form structured data.
8. The method of claim 6, wherein the step of inputting training sample data and test sample data into a machine learning model for clinical endpoint event assessment, and outputting the assessment result of the clinical endpoint event of the subject after a plurality of iterative training specifically comprises:
respectively inputting training sample data and test sample data into N candidate machine learning submodels in a machine learning model according to a ratio of 8:2, counting output results of all the candidate machine learning submodels, and judging whether a plurality of output results are consistent according to the output results; disagreement in results is objected;
the establishment method of the candidate machine learning submodel comprises the following steps: repeating the corresponding clinical evaluation event sample set M for N times, and randomly drawing M samples each time when the clinical evaluation event sample set M is replaced to obtain N sample sets, wherein each sample set comprises M samples; and establishing N corresponding machine learning submodels for the obtained N sample sets.
9. A real-time assessment system for clinical endpoint events in clinical trials is characterized by comprising a project setting subsystem and an endpoint event review subsystem which are in bidirectional data connection;
the project setting subsystem comprises
The project data acquisition module is configured to be docked with the EDC system and acquire clinical test data and a clinical endpoint event table of the system according to the test projects;
the source file data storage module is used for storing the acquired project data files comprising clinical data and a clinical endpoint event table into a source file database;
the evaluation personnel setting module is configured to configure an endpoint event evaluation table of the test project according to the clinical endpoint event table and set evaluation personnel at the same time; the reviewers consist of 1 position chairman and an even position committee;
and the evaluation object permission setting module is configured to set the roles of the permissions of the evaluation personnel: when the reviewers are committees, the reviewers have the permission to view the overview of the review progress of the reviewers and the permission to modify the review content of the reviewers; when the reviewer is the chairman, the reviewer has the permission to view the contents of the review table finished and submitted for review in the reviewer and the overview permission to view the review progress of the reviewer;
the review progress overview module is configured to supervise and overview the review progress of the test project by the reviewers;
the endpoint event review subsystem includes
The first evaluation module is configured to enable an evaluator to obtain clinical data of a subject needing manual evaluation in a test project, evaluate an end point event evaluation table, sign and confirm the end point event evaluation table and submit the end point event evaluation table to obtain a final end point event table of the first evaluation;
the initial evaluation result processing module is configured to compare and screen the results of the final end point event table of all the first evaluations, and when the fields of the comparison results are consistent, the evaluation results are directly obtained; when the comparison result has the identification of the inconsistent field, entering a second evaluation stage;
the second evaluation module is configured to carry out second evaluation on the data of the subjects marked as inconsistent in the first evaluation result according to the establishment of an evaluation meeting of the evaluation personnel, and obtain a final evaluation result;
and the offline review module is configured to export inconsistent subject data when the first review result is inconsistent.
10. The system for real-time assessment of clinical endpoint events in a clinical trial of claim 9, wherein the system further comprises an automated clinical event assessment subsystem;
the clinical event automated assessment subsystem comprises
The model building module is configured to select features based on disease item diagnosis standards to obtain disease item clinical evaluation standards and build a machine learning model for clinical endpoint event evaluation;
the data processing module is configured to extract clinical test data of the subjects of the target items in the source file database, and normalize the clinical test data to form unified structured data;
the data partitioning module is configured to select endpoint clinical assessment event data of a plurality of target projects as training sample data; selecting clinical test data of a subject of a target project to be tested as test sample data;
a machine training module configured to input training sample data and test sample data into a machine learning model for clinical endpoint event assessment; after multiple times of iterative training, outputting the evaluation result of the clinical endpoint event of the subject;
a result evaluation module configured to evaluate the output bed end point event evaluation result; when the evaluation result of the clinical endpoint time evaluated by the system is objected, the clinical evaluation event of the corresponding subject is set as a subject clinical evaluation event needing manual evaluation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110050570.1A CN112785256A (en) | 2021-01-14 | 2021-01-14 | Real-time assessment method and system for clinical endpoint events in clinical trials |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110050570.1A CN112785256A (en) | 2021-01-14 | 2021-01-14 | Real-time assessment method and system for clinical endpoint events in clinical trials |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112785256A true CN112785256A (en) | 2021-05-11 |
Family
ID=75756688
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110050570.1A Pending CN112785256A (en) | 2021-01-14 | 2021-01-14 | Real-time assessment method and system for clinical endpoint events in clinical trials |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112785256A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113449971A (en) * | 2021-06-13 | 2021-09-28 | 上海用正医药科技有限公司 | Inspection task assignment method based on clinical test index data analysis result |
| CN117153424A (en) * | 2023-11-01 | 2023-12-01 | 北京遥领医疗科技有限公司 | Centralized curative effect evaluation method and system |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050187795A1 (en) * | 2004-02-24 | 2005-08-25 | Mary Russell | Dynamic safety monitoring in clinical trial |
| KR100794516B1 (en) * | 2007-12-03 | 2008-01-14 | 한국정보통신대학교 산학협력단 | System and method for diagnosis and clinical test selection using case based machine learning inference |
| CN107220217A (en) * | 2017-05-31 | 2017-09-29 | 北京京东尚科信息技术有限公司 | Characteristic coefficient training method and device that logic-based is returned |
| CN108734330A (en) * | 2017-04-24 | 2018-11-02 | 北京京东尚科信息技术有限公司 | Data processing method and device |
| CN109166105A (en) * | 2018-08-01 | 2019-01-08 | 中国人民解放军南京军区南京总医院 | The malignancy of tumor risk stratification assistant diagnosis system of artificial intelligence medical image |
| CN111341455A (en) * | 2020-02-10 | 2020-06-26 | 厦门茶蕊生物医药科技有限公司 | Clinical test integrated cloud platform management system, method and storage medium |
| CN111640509A (en) * | 2020-06-02 | 2020-09-08 | 山东大学齐鲁医院 | Cervical cancer postoperative recurrence risk prediction method and system |
| CN111754080A (en) * | 2020-05-28 | 2020-10-09 | 山东爱城市网信息技术有限公司 | Project review method, system, device and medium based on agile management |
| CN112071428A (en) * | 2020-09-01 | 2020-12-11 | 物卡智能科技(深圳)有限公司 | Latent disease analysis system based on health data acquisition |
-
2021
- 2021-01-14 CN CN202110050570.1A patent/CN112785256A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050187795A1 (en) * | 2004-02-24 | 2005-08-25 | Mary Russell | Dynamic safety monitoring in clinical trial |
| KR100794516B1 (en) * | 2007-12-03 | 2008-01-14 | 한국정보통신대학교 산학협력단 | System and method for diagnosis and clinical test selection using case based machine learning inference |
| CN108734330A (en) * | 2017-04-24 | 2018-11-02 | 北京京东尚科信息技术有限公司 | Data processing method and device |
| CN107220217A (en) * | 2017-05-31 | 2017-09-29 | 北京京东尚科信息技术有限公司 | Characteristic coefficient training method and device that logic-based is returned |
| CN109166105A (en) * | 2018-08-01 | 2019-01-08 | 中国人民解放军南京军区南京总医院 | The malignancy of tumor risk stratification assistant diagnosis system of artificial intelligence medical image |
| CN111341455A (en) * | 2020-02-10 | 2020-06-26 | 厦门茶蕊生物医药科技有限公司 | Clinical test integrated cloud platform management system, method and storage medium |
| CN111754080A (en) * | 2020-05-28 | 2020-10-09 | 山东爱城市网信息技术有限公司 | Project review method, system, device and medium based on agile management |
| CN111640509A (en) * | 2020-06-02 | 2020-09-08 | 山东大学齐鲁医院 | Cervical cancer postoperative recurrence risk prediction method and system |
| CN112071428A (en) * | 2020-09-01 | 2020-12-11 | 物卡智能科技(深圳)有限公司 | Latent disease analysis system based on health data acquisition |
Non-Patent Citations (1)
| Title |
|---|
| 王国强;阚红星;王宗殿;: "综合多种数据挖掘技术的糖尿病诊断系统", 电脑知识与技术, no. 23 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113449971A (en) * | 2021-06-13 | 2021-09-28 | 上海用正医药科技有限公司 | Inspection task assignment method based on clinical test index data analysis result |
| CN113449971B (en) * | 2021-06-13 | 2022-03-18 | 上海用正医药科技有限公司 | Inspection task assignment method based on clinical test index data analysis result |
| CN117153424A (en) * | 2023-11-01 | 2023-12-01 | 北京遥领医疗科技有限公司 | Centralized curative effect evaluation method and system |
| CN117153424B (en) * | 2023-11-01 | 2024-02-23 | 北京遥领医疗科技有限公司 | Centralized curative effect evaluation method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tornøe et al. | Creation of a knowledge management system for QT analyses | |
| CN112785256A (en) | Real-time assessment method and system for clinical endpoint events in clinical trials | |
| CN104331778A (en) | Intelligent management control method for clinical blood transfusion electronic information system | |
| CN104919487A (en) | Apparatus and method for executing tasks | |
| Das et al. | Deep-learning algorithm helps to standardise ATS/ERS spirometric acceptability and usability criteria | |
| CN111695836B (en) | Clinical trial online operation management and control integrated system | |
| Alves et al. | Software quality evaluation of the laboratory information system used in the santa catarina state integrated telemedicine and telehealth system | |
| Megna et al. | A new relational database including clinical data and myocardial perfusion imaging findings in coronary artery disease | |
| Fu et al. | Design and implementation of clinical LIS360 laboratory management system based on AI technology | |
| CN104699968A (en) | Intelligent medical guide system and method | |
| CN116487066A (en) | Clinical test data monitoring method, system and electronic equipment | |
| CN109887602A (en) | A kind of cardiovascular disease big data analysis system and method | |
| Aziz | Gene Therapy: Development, Design of Studies, and Approval Process | |
| Ball et al. | Interdisciplinary safety evaluation for learning and decision-making | |
| Saito | Identifying and understanding stakeholders using process mining: case study on discovering business processes that involve organizational entities | |
| Constantin et al. | How Do Algorithmic Fairness Metrics Align with Human Judgement? A Mixed-Initiative System for Contextualized Fairness Assessment | |
| CN115668178A (en) | Methods, systems, and computer program products for retrospective data mining | |
| Gupta et al. | Early experience using an online reporting system for interventional radiology procedure-related complications integrated with a digital dictation system | |
| Alyea et al. | Standardizing health-care data across an enterprise | |
| JP2021506049A (en) | Systems and methods for collaborative image processing | |
| Sirkis et al. | Using statistical process control to understand variation in computer-assisted personal interviewing data | |
| Tamanini et al. | The Neuropathological Diagnosis of the Alzheimer’s Disease under the Consideration of Verbal Decision Analysis Methods | |
| Reporting | The Pathology Milestone Project | |
| Speaker | Project FORESIGHT annual report, 2016-2017 | |
| Beerepoot | Workaround: the path from detection to improvement |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |