RWS-based platform for tracking effects of drugs on market

Abstract

The invention discloses an RWS-based post-market drug effect tracking platform, which comprises a data acquisition layer, a data processing layer and a result analysis layer, and is characterized in that the data acquisition layer is used for acquiring RWS data from different sources, the data processing layer is used for preprocessing the RWS data from different sources, constructing an original data chain based on a block chain technology, and analyzing the result of the original data chain; writing the preprocessed RWS data of different sources into an original data chain for storage; and the result analysis layer is used for acquiring the preprocessed RWS data through the original data chain, analyzing the preprocessed RWS data, constructing a drug tracking knowledge graph, and comprehensively analyzing the drug tracking knowledge graph to obtain an indication distribution result, an adverse reaction monitoring result, a treatment cost distribution result and a curative effect analysis result.

Description

Technical Field

[0001] The present invention relates to the field of real-world research technology, and in particular to a post-marketing drug effect tracking platform based on RWS. Background Art

[0002] In recent years, with the rise of medical big data technology, real-world data (RWS) has become a crucial external data source for medical R&D. These data come from multiple sources, including electronic health records, administrative claims records, health applications, and emerging channels such as public health databases. The importance of real-world research lies in its ability to provide supplementary evidence beyond clinical trials, helping medical decision-makers better understand the performance of medical products in a wide range of routine clinical settings. This research method is particularly suitable for evaluating long-term treatment effects.

[0003] The current traditional post-marketing drug monitoring system primarily relies on spontaneous reporting systems and rigorously designed observational studies, which present significant limitations. Spontaneous reporting systems passively collect information, limited by the awareness and willingness of the reporter, and are subject to significant underreporting and reporting bias. This makes it difficult to systematically identify rare or delayed adverse reactions and accurately quantify risks. Traditional observational studies are often limited in scale, costly, and time-consuming, and their data coverage and timeliness are insufficient to meet the needs of dynamically tracking the long-term effects of drugs in the entire population, particularly in specialized populations. Furthermore, pre-marketing clinical trials are typically conducted in controlled environments with relatively homogeneous patient populations and limited follow-up periods, making it difficult to fully extrapolate their conclusions to the complex and diverse real-world patient populations and long-term medication use scenarios. This results in potential long-term efficacy loss, rare or delayed adverse reactions, drug-drug interaction risks, and variability in performance under unevenly distributed healthcare resources, which cannot be captured and assessed in a timely, comprehensive, and cost-effective manner by existing monitoring methods. Therefore, a post-marketing drug performance tracking platform that integrates real-world data is needed. Summary of the Invention

[0004] In response to the above-mentioned prior art, the present invention provides a post-marketing drug effect tracking platform based on RWS, which mainly solves the technical problems existing in the above-mentioned background technology.

[0005] To achieve the above objectives, the technical solution of the embodiment of the present invention is implemented as follows: a RWS-based post-marketing drug effect tracking platform, the platform comprising a data acquisition layer, a data processing layer, and a result analysis layer. The data acquisition layer is used to obtain RWS data from different sources, and the data processing layer is used to pre-process the RWS data from different sources, build an original data chain based on blockchain technology, and write the pre-processed RWS data from different sources into the original data chain for storage;

[0006] The result analysis layer is used to obtain pre-processed RWS data through the original data chain, construct a drug tracking knowledge graph by analyzing the pre-processed RWS data, and obtain indication distribution results, adverse reaction monitoring results, treatment cost distribution results, and efficacy analysis results through comprehensive analysis of the drug tracking knowledge graph.

[0007] Optionally, the RWS data includes the hospital's clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data.

[0008] Optionally, the data collection layer specifically includes:

[0009] The first acquisition module is used to access the hospital management system and obtain the diagnosis and treatment data and laboratory examination data in the hospital electronic medical record system in real time;

[0010] The second collection module is used to access the medical insurance management system and synchronously obtain the expense details and payment records in the medical insurance database;

[0011] The third collection module is used to access the pharmaceutical company's supply chain system to obtain drug circulation data;

[0012] The fourth acquisition module is used to access a third-party survey system to obtain patient-generated health and medical data including subjective efficacy feedback and adverse reaction reports.

[0013] Optionally, the data processing layer includes:

[0014] The data cleaning submodule is used to clean the numerical RWS data, remove duplicate records, fill in missing values, and filter abnormal data;

[0015] The data standardization submodule is used to map text data into coded expressions;

[0016] Building blocks for constructing original data chains based on blockchain technology;

[0017] Node management module, used to manage nodes participating in the original data chain consensus process;

[0018] The first storage module is used to store the hospital's clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data into the off-chain database, calculate the hash value of the stored data, and record the storage path of the stored data in the off-chain database;

[0019] The second storage module is used to write the hash value and storage path of the hospital's clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data into the original data chain to update the original data chain.

[0020] Optionally, the result analysis layer specifically includes:

[0021] The graph construction module is used to analyze RWS data and build a drug tracking knowledge graph based on the analysis results;

[0022] Indication analysis module, used to count the frequency of drug-disease relationships in the drug tracking knowledge graph;

[0023] Adverse reaction monitoring module, which is used to analyze the association paths between adverse reactions, drugs, and patient characteristics through the drug tracking knowledge graph;

[0024] The treatment cost analysis module is used to link the drug-disease-cost entity chain, analyze the treatment cost structure of different indications, and compare the cost differences between medical insurance-paid and self-funded patients;

[0025] The efficacy analysis module is used to analyze and identify abnormal indicator improvement signals that are not directly related to the current indications of the drug, and to deduce potential new indications of the drug based on the abnormal indicator improvement signals.

[0026] Optionally, the graph construction module specifically includes:

[0027] The entity extraction submodule is used to extract pre-named entities from RWS data through natural language processing technology and identify the relationships between the identified entities, where the entities include drug entities, disease entities, patient entities, indication entities, adverse reaction entities, and cost entities;

[0028] A submodule is constructed to build a drug tracking knowledge graph based on the subject corresponding to each user and its relationship with other subjects;

[0029] The graph update submodule is used to regularly synchronize new data from the original data chain and incrementally update the knowledge graph.

[0030] Optionally, the efficacy analysis module specifically includes:

[0031] The abnormal indicator real-time monitoring submodule is used to filter indicators directly related to the current indications of the drug from the drug tracking knowledge graph, and mark abnormal indicator improvement signals based on pre-set indicator improvement amplitude thresholds and improvement rate thresholds;

[0032] The comparison submodule is used to call the external medical knowledge base to retrieve and analyze the association evidence between the current indication and the aforementioned abnormal indicators. If the external medical knowledge base does not record the relevant association relationship, it is determined to be an unexpected improvement signal.

[0033] Optionally, the result analysis layer further includes a visualization module, and the visualization module is used to present the analysis results to the user in a visual form.

[0034] The beneficial effects of the present invention are as follows: the RWS-based post-marketing effect tracking platform provided by the present application obtains pre-processed RWS data through the raw data chain, uses natural language processing technology to extract pre-named entities such as drug entities, disease entities, patient entities, indication entities, adverse reaction entities, and cost entities from the data, and identifies the relationship between the entities to construct a drug tracking knowledge graph. Through comprehensive analysis through the knowledge graph, the indication distribution results are obtained by statistically analyzing the frequency of occurrence of drug-disease relationships, and the association paths between adverse reactions and drug and patient characteristics are analyzed to achieve adverse reaction monitoring. The entity chain of drug-disease-cost is associated to analyze the treatment cost structure of different indications and the cost difference between medical insurance payment and self-pay patients, identify abnormal indicator improvement signals that are not directly related to the current indication of the drug, and deduce potential new indications of the drug in combination with an external medical knowledge base. By integrating multi-source real-world data such as hospital diagnosis and treatment, medical insurance payment, drug circulation and patient feedback, a three-dimensional tracking system covering the entire life cycle of the drug is formed, breaking through the limitation of traditional post-marketing monitoring relying on a single data source, and providing all-round data support for drug supervision, clinical decision-making, and R&D innovation. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a schematic diagram of the overall connection of the RWS-based post-marketing drug effect tracking platform in the embodiment of this application;

[0036] Figure 2 This is a schematic diagram of the overall connection of the data acquisition layer in the embodiment of the present application;

[0037] Figure 3 This is a schematic diagram of the overall connection of the data processing layer in an embodiment of the present application. DETAILED DESCRIPTION

[0038] The technical solution of the present invention is further elaborated in detail below in conjunction with the drawings and specific embodiments of the specification. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention. In the following description, reference is made to "some embodiments", which describes a subset of all possible embodiments, but it should be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict.

[0039] In the following description, numerous specific details are provided to provide a more thorough understanding of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced without one or more of these details. In other instances, certain technical features well known in the art are not described to avoid confusion with the present invention.

[0040] It should be understood that the present invention can be implemented in different forms and should not be interpreted as being limited to the embodiments proposed herein. On the contrary, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the present invention to those skilled in the art. And the purpose of the terms used herein is only to describe specific embodiments and is not intended to limit the present invention. When used herein, the singular forms "one", "an" and "said / the" are also intended to include plural forms, unless the context clearly indicates another way. It should also be understood that the terms "comprising" and / or "comprising" when used in this specification determine the presence of the features, integers, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, parts and / or groups. When used herein, the term "and / or" includes any and all combinations of the relevant listed items.

[0041] It should also be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be an intermediate element. When an element is referred to as being "connected to" another element, it may be directly connected to the other element or there may be an intermediate element. The terms "vertical," "horizontal," "inner," "outer," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only implementation methods.

[0042] In order to fully understand the present invention, a detailed structure will be provided in the following description to illustrate the technical solution proposed by the present invention. Optional embodiments of the present invention are described in detail below. However, in addition to these detailed descriptions, the present invention may also have other implementations.

[0043] See also Figures 1 to 3 The present invention discloses a post-marketing drug effect tracking platform based on RWS. The platform includes a data acquisition layer, a data processing layer, and a result analysis layer. The data acquisition layer is used to obtain RWS data from different sources. The data processing layer is used to pre-process the RWS data from different sources, build an original data chain based on blockchain technology, and write the pre-processed RWS data from different sources into the original data chain for storage.

[0044] The result analysis layer is used to obtain pre-processed RWS data through the original data chain, construct a drug tracking knowledge graph by analyzing the pre-processed RWS data, and obtain indication distribution results, adverse reaction monitoring results, treatment cost distribution results, and efficacy analysis results through comprehensive analysis of the drug tracking knowledge graph.

[0045] Specifically, this application discloses a post-marketing drug effect tracking platform based on RWS, in which the data collection layer is responsible for obtaining RWS data from different sources. The data types include hospital clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data. In specific implementation, by accessing the hospital management system to obtain real-time diagnosis and treatment data and laboratory test data in electronic medical records, accessing the medical insurance management system to synchronously obtain fee details and payment records in the medical insurance database, accessing the pharmaceutical company supply chain system to obtain drug circulation data, and accessing the third-party survey system to obtain patient subjective efficacy feedback and adverse reaction reports and other data, real-time and synchronous collection of multi-channel data is achieved.

[0046] The data processing layer preprocesses RWS data from various sources, performing cleaning operations on numeric data by removing duplicate records, filling missing values, and filtering out abnormal data. It also maps textual data into coded expressions for standardization. A raw data chain is constructed based on blockchain technology, storing the preprocessed RWS data in an off-chain database. The hash value of the stored data is calculated and its storage path in the off-chain database is recorded. The hash value and storage path are then written to the raw data chain, ensuring data integrity and security through the blockchain's immutability and traceability.

[0047] The results analysis layer obtains preprocessed RWS data through the raw data chain and uses natural language processing technology to extract pre-named entities such as drug entities, disease entities, patient entities, indication entities, adverse reaction entities, and cost entities from the data. It then identifies the relationships between these entities and constructs a drug tracking knowledge graph. A comprehensive analysis is conducted through the knowledge graph, and the frequency of drug-disease relationships is statistically analyzed to derive indication distribution results. The association paths between adverse reactions and drug and patient characteristics are analyzed to achieve adverse reaction monitoring. The entity chain of drug-disease-cost is linked to analyze the treatment cost structure of different indications and the difference in costs between medical insurance-paid and self-paying patients. Signals of improvement in abnormal indicators not directly related to the drug's current indication are identified and combined with external medical knowledge bases to deduce potential new indications for the drug.

[0048] As a possible implementation method, the RWS data includes the hospital's clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data.

[0049] Among them, the hospital's clinical diagnosis and treatment data refers to the patient-related information recorded electronically in the process of medical institutions providing diagnosis and treatment services, covering the patient's basic information, medical history, diagnosis results, treatment plan, medication records, examination and test reports, surgical records, nursing records, etc. For example, the diagnosis and treatment data and laboratory test data in the hospital's electronic medical record system can continuously and dynamically reflect the patient's health status.

[0050] Medical insurance and payment data is a collection of data accumulated by insurance companies in the process of handling insured persons' medical expense claims. It contains basic identity information of insured persons, date of consultation, medical institution, diagnosis code, treatment items, drug usage, medical expense amount, reimbursement ratio and self-paid portion, etc. It can reflect the medical service utilization and expense expenditure of the insured population, such as the expense details and payment records in the medical insurance database.

[0051] Drug circulation data is data obtained by accessing the pharmaceutical company's supply chain system. It mainly involves relevant information on drugs in the circulation links such as production, transportation, storage, and sales. It is used to track the flow of drugs from manufacturers to medical institutions or retail terminals to ensure the traceability of the drug supply link.

[0052] Patient-generated health and medical data are data created, recorded, collected or inferred by patients or their designated persons that are helpful in solving health problems. They include subjective efficacy feedback and adverse reaction reports obtained through third-party survey systems, as well as personal health behaviors in health survey data, such as diet, exercise, smoking, drinking, etc., health status such as whether or not they are ill, symptoms, quality of life, demand for and utilization of health services, and other information obtained through self-reporting or surveys.

[0053] Furthermore, the data collection layer specifically includes:

[0054] The first acquisition module is used to access the hospital management system and obtain the diagnosis and treatment data and laboratory examination data in the hospital electronic medical record system in real time;

[0055] The second collection module is used to access the medical insurance management system and synchronously obtain the expense details and payment records in the medical insurance database;

[0056] The third collection module is used to access the pharmaceutical company's supply chain system to obtain drug circulation data;

[0057] The fourth acquisition module is used to access a third-party survey system to obtain patient-generated health and medical data including subjective efficacy feedback and adverse reaction reports.

[0058] Specifically, the first acquisition module obtains the diagnosis and treatment data and laboratory test data in the electronic medical record in real time by connecting to the hospital management system, which can ensure the timeliness and integrity of the clinical frontline data, and provide real and accurate basic data for analyzing the use of drugs in actual diagnosis and treatment scenarios, the correlation with disease diagnosis and changes in patients' physiological indicators, so that subsequent efficacy analysis can be closely aligned with clinical practice.

[0059] The second acquisition module is connected to the medical insurance management system to synchronously obtain cost details and payment records. It can directly link drug use with medical economic data, realize quantitative analysis of treatment costs for different indications, and compare cost differences between medical insurance payments and self-paying patients. It provides data support for evaluating the health economic benefits of drugs, the coverage effect of medical insurance policies and the economic burden of patients, and assists in the evaluation of drug cost-effectiveness and optimal resource allocation.

[0060] The third acquisition module accesses the pharmaceutical company's supply chain system to obtain drug circulation data, which can fully track the entire chain of drugs from production to sales, clearly present the regional distribution, inventory dynamics and market flow of drugs, help monitor the stability and accessibility of drug supply, and promptly discover abnormal fluctuations in the circulation link, providing decision-making basis for optimizing drug supply chain management and ensuring clinical drug supply.

[0061] The fourth collection module connects to a third-party survey system to obtain patient-generated health and medical data. It can effectively collect patient subjective experience information that is difficult to cover with traditional clinical data, such as subjective efficacy feedback after medication, unexpected adverse reaction reports, etc., filling the gap in objective diagnosis and treatment data in patients' subjective feelings and long-term prognosis observations.

[0062] As a possible implementation, the data processing layer includes:

[0063] The data cleaning submodule is used to clean the numerical RWS data, remove duplicate records, fill in missing values, and filter abnormal data;

[0064] The data standardization submodule is used to map textual data into coded expressions. For example, it uses natural language processing technology to extract key entities such as drug names and disease diagnoses from unstructured text and map them to predefined standardized codes, such as ICD-10 disease codes, ATC drug codes, and MedDRA adverse reaction terms.

[0065] The building module is used to construct the original data chain based on blockchain technology. Specifically, it uses the distributed ledger technology of blockchain to package the pre-processed RWS data metadata, such as data type and collection time, with hash values ​​into blocks. After verification through the consensus mechanism, the blocks are linked in chronological order to form a chain. Each block contains the hash value of the previous block, forming an unalterable chain structure.

[0066] The node management module manages the nodes participating in the original data chain consensus process. By assigning different permissioned nodes to participants such as medical institutions, pharmaceutical companies, and medical insurance departments, and defining node read and write permissions and consensus rules through smart contracts, the module monitors node status in real time and dynamically adjusts the network topology. This ensures the stability and security of the blockchain network, preventing single node failures from impacting system operations. Furthermore, through permission isolation, it ensures access control to sensitive data, complying with data privacy regulations.

[0067] The first storage module is used to store the hospital's clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data into the off-chain database, calculate the hash value of the stored data, and record the storage path of the stored data in the off-chain database;

[0068] The second storage module is used to write the hash value and storage path of the hospital's clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data into the original data chain to update the original data chain.

[0069] As a possible implementation, the result analysis layer specifically includes:

[0070] The graph construction module is used to analyze RWS data and construct a drug tracking knowledge graph based on the analysis results. It constructs a drug tracking knowledge graph by analyzing RWS data, using natural language processing technology to extract entities such as drugs, diseases, patients, indications, adverse reactions, and costs from the data, identify the relationships between entities, such as the therapeutic association between drugs and diseases, and the causal relationship between drugs and adverse reactions, and form a structured knowledge network based on the graph database;

[0071] The indication analysis module is used to count the frequency of drug-disease relationships in the drug tracking knowledge graph. Specifically, based on the constructed drug tracking knowledge graph, the indication analysis module counts the frequency of drug-disease relationships in the graph, and uses the frequency to reflect the actual frequency of drug use in treating different diseases, thereby clarifying the main indications of the drug and its distribution range in the real world.

[0072] The adverse reaction monitoring module is used to analyze the association paths between adverse reactions and drugs and patient characteristics through the drug tracking knowledge graph. Specifically, the adverse reaction monitoring module uses the drug tracking knowledge graph to analyze the association paths between adverse reactions and drugs and patient characteristics. By mining the association between adverse reaction entities and drug entities in the knowledge graph, such as drug type, dosage, course of treatment, and patient entities such as age, gender, underlying medical history, and allergy history, it identifies high-risk factor combinations or potential causal paths for adverse reactions. For example, when a certain antihypertensive drug is used in combination with non-steroidal anti-inflammatory drugs (NSAIDs), the knowledge graph shows that the association path weight of acute kidney injury in elderly patients (aged > 65 years old) is significantly increased. The analysis found that NSAIDs may inhibit prostaglandin synthesis and weaken the protective effect of antihypertensive drugs on renal blood flow. Combined with the basis of renal dysfunction in the elderly, a high-risk combination of drug combination + age is formed, leading to an increased risk of kidney injury.

[0073] For breast cancer patients receiving chemotherapy, the knowledge graph shows that when using regimens containing anthracyclines (such as doxorubicin), patients with underlying heart diseases (such as a history of coronary heart disease) and HER2 targeted therapy (trastuzumab) have a significantly increased risk of decreased left ventricular ejection fraction (LVEF). The causal pathway points to the synergistic effect of anthracycline cardiotoxicity and trastuzumab's myocardial damage, coupled with the patient's insufficient cardiac reserve function, forming a multi-factor-driven adverse reaction risk pathway.

[0074] The treatment cost analysis module is used to associate the entity chain of drugs, diseases and costs, and analyze the treatment cost structure of different indications. Specifically, the treatment cost analysis module integrates drug usage information, disease diagnosis information and cost data by associating the drug-disease-cost entity chain in the knowledge graph, and analyzes the medical insurance costs consumed at different stages when using the drug to treat a certain indication.

[0075] The efficacy analysis module is used to analyze and identify abnormal indicator improvement signals that are not directly related to the current indications of the drug, and deduce potential new indications of the drug based on the abnormal indicator improvement signals. For example, it identifies abnormal indicator improvement signals that are not directly related to the current indications of the drug, and uses preset indicator improvement amplitude thresholds and improvement rate thresholds to screen out indicator change data that have no clear association with the existing indications of the drug from the knowledge graph, mark the abnormal improvement signals, and then call the external medical knowledge base to retrieve and analyze the correlation between the indicator and the current indication. If no known medical correlation is found, it is determined to be an unexpected improvement signal, and then combined with clinical data to deduce potential new indications that may exist for the drug, providing clues and research directions for the expansion of indications for drug research and development.

[0076] For example, knowledge graph analysis of clinical data for a certain antidepressant, currently indicated for depression, revealed that approximately 18% of patients experienced accelerated gastric emptying after taking the medication, an improvement greater than 25% as measured by gastric motility monitoring. However, depression treatment guidelines do not mention the drug's effects on gastrointestinal motility. A medical knowledge base search confirmed no relevant associations, identifying this as an unexpected improvement.

[0077] As a possible implementation, the graph construction module specifically includes:

[0078] The entity extraction submodule is used to extract pre-named entities from RWS data through natural language processing technology and identify the relationships between the identified entities, where the entities include drug entities, disease entities, patient entities, indication entities, adverse reaction entities, and cost entities;

[0079] Specifically, the entity extraction submodule uses natural language processing technology to parse unstructured text in RWS data, such as electronic medical records, medical records, and patient feedback, and uses named entity recognition, entity linking and other technologies to accurately extract drug entities, such as generic names, trade names, ingredients, disease entities such as disease names, ICD-10 codes, patient entities such as age, gender, and regional characteristics, indication entities such as indications in drug instructions, actual clinical application symptoms, adverse reaction entities such as symptom names and severity grades, and cost entities such as drug unit prices, total treatment costs, and medical insurance reimbursement ratios. At the same time, it identifies the relationships between entities, such as drug-treatment-disease, drug-cause-adverse reaction, and patient-suffering-disease, converting unstructured data into structured entity-relationship triples to provide basic data units for knowledge graph construction.

[0080] A submodule is constructed to build a drug tracking knowledge graph based on the subject corresponding to each user and its relationship with other subjects;

[0081] The construction submodule is based on the entity and relationship data output by the entity extraction submodule, and uses a graph database to build a drug tracking knowledge graph. This module uses various entities as graph nodes and the relationships between entities as graph edges. Through attribute labels such as drug dosage, disease stage, and adverse reaction occurrence time, it enriches the feature information of nodes and edges to form a multi-dimensional and multi-level knowledge network. For example, with "drug entity" as the central node, it associates "disease entity", "adverse reaction entity", "patient entity", and "cost entity" to intuitively show the full picture of drug application in the real world.

[0082] The graph update submodule is used to regularly synchronize new data from the original data chain and incrementally update the knowledge graph.

[0083] As a possible implementation, the efficacy analysis module specifically includes:

[0084] The abnormal indicator real-time monitoring submodule is used to filter indicators directly related to the current indications of the drug from the drug tracking knowledge graph, and mark abnormal indicator improvement signals based on pre-set indicator improvement amplitude thresholds and improvement rate thresholds;

[0085] The comparison submodule is used to call the external medical knowledge base to retrieve and analyze the association evidence between the current indication and the aforementioned abnormal indicators. If the external medical knowledge base does not record the relevant association relationship, it is determined to be an unexpected improvement signal.

[0086] The real-time monitoring submodule for abnormal indicators uses the entity and relationship network of the knowledge graph to automatically screen out indicators directly related to the drug's current indication, such as blood glucose indicators for hypoglycemic drugs, and excludes these indicators, focusing on other non-directly related indicators, such as blood lipids and liver function indicators. Based on pre-set indicator improvement thresholds (e.g., when a certain indicator changes beyond the normal range) and improvement rate thresholds (e.g., when a certain indicator improves with a frequency greater than 15% in a specific population), the data in the knowledge graph is analyzed. When both the improvement magnitude and the incidence rate of a certain indicator exceed the threshold, it is automatically marked as an abnormal indicator improvement signal. For example, after taking antihypertensive drugs, a patient's urine protein level unexpectedly decreases significantly, and the incidence is high in the population, suggesting the possibility of an unrecognized kidney-protective effect.

[0087] Upon receiving a signal indicating an abnormal indicator has improved, the comparison submodule automatically accesses an external medical knowledge base to retrieve known evidence linking the current drug indication to the abnormal indicator. Using natural language processing techniques, the module then parses the knowledge base literature to determine whether any relevant associations exist based on pathological mechanisms, clinical trials, or case reports. If no relevant associations are found in the knowledge base, the signal is considered an unexpected improvement. For example, if a patient's gastric motility index unexpectedly improves after taking an antidepressant, a search of existing medical data reveals no association between the drug and gastrointestinal motility, suggesting that the drug may have potential to regulate the digestive system, providing research leads for exploring new indications.

[0088] Furthermore, the result analysis layer also includes a visualization module, which is used to present the analysis results to the user in a visual form.

[0089] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. The scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A drug post-marketing effect tracking platform based on RWS, characterized by: The platform includes a data acquisition layer, a data processing layer, and a result analysis layer. The data acquisition layer is used to obtain RWS data from different sources. The data processing layer is used to pre-process the RWS data from different sources, build an original data chain based on blockchain technology, and write the pre-processed RWS data from different sources into the original data chain for storage. The result analysis layer is used to obtain pre-processed RWS data through the original data chain, construct a drug tracking knowledge graph by analyzing the pre-processed RWS data, and obtain indication distribution results, adverse reaction monitoring results, treatment cost distribution results, and efficacy analysis results through comprehensive analysis of the drug tracking knowledge graph.

2. A drug post-marketing effect tracking platform based on RWS according to claim 1, characterized in that: The RWS data includes the hospital's clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data.

3. A drug post-marketing effect tracking platform based on RWS according to claim 2, characterized in that: The data collection layer specifically includes: The first acquisition module is used to access the hospital management system and obtain the diagnosis and treatment data and laboratory examination data in the hospital electronic medical record system in real time; The second collection module is used to access the medical insurance management system and synchronously obtain the expense details and payment records in the medical insurance database; The third collection module is used to access the pharmaceutical company's supply chain system to obtain drug circulation data; The fourth acquisition module is used to access a third-party survey system to obtain patient-generated health and medical data including subjective efficacy feedback and adverse reaction reports.

4. A drug post-marketing effect tracking platform based on RWS according to claim 3, characterized in that: The data processing layer includes: The data cleaning submodule is used to clean the numerical RWS data, remove duplicate records, fill in missing values, and filter abnormal data; The data standardization submodule is used to map text data into coded expressions; Building blocks for constructing original data chains based on blockchain technology; Node management module, used to manage nodes participating in the original data chain consensus process; The first storage module is used to store the hospital's clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data into the off-chain database, calculate the hash value of the stored data, and record the storage path of the stored data in the off-chain database; The second storage module is used to write the hash value and storage path of the hospital's clinical diagnosis and treatment data, medical insurance and payment data, drug circulation data, and patient-generated health and medical data into the original data chain to update the original data chain.

5. The RWS-based drug post-marketing effect tracking platform according to claim 1, characterized in that: The result analysis layer specifically includes: The graph construction module is used to analyze RWS data and build a drug tracking knowledge graph based on the analysis results; Indication analysis module, used to count the frequency of drug-disease relationships in the drug tracking knowledge graph; Adverse reaction monitoring module, which is used to analyze the association paths between adverse reactions, drugs, and patient characteristics through the drug tracking knowledge graph; The treatment cost analysis module is used to link the drug-disease-cost entity chain, analyze the treatment cost structure of different indications, and compare the cost differences between medical insurance-paid and self-funded patients; The efficacy analysis module is used to analyze and identify abnormal indicator improvement signals that are not directly related to the current indications of the drug, and to deduce potential new indications of the drug based on the abnormal indicator improvement signals.

6. The RWS-based post-marketing drug effect tracking platform according to claim 5, characterized in that: The graph construction module specifically includes: The entity extraction submodule is used to extract pre-named entities from RWS data through natural language processing technology and identify the relationships between the identified entities, where the entities include drug entities, disease entities, patient entities, indication entities, adverse reaction entities, and cost entities; A submodule is constructed to build a drug tracking knowledge graph based on the subject corresponding to each user and its relationship with other subjects; The graph update submodule is used to regularly synchronize new data from the original data chain and incrementally update the knowledge graph.

7. The RWS-based post-marketing drug effect tracking platform according to claim 6, characterized in that: The efficacy analysis module specifically includes: The abnormal indicator real-time monitoring submodule is used to filter indicators directly related to the current indications of the drug from the drug tracking knowledge graph, and mark abnormal indicator improvement signals based on pre-set indicator improvement amplitude thresholds and improvement rate thresholds; The comparison submodule is used to call the external medical knowledge base to retrieve and analyze the association evidence between the current indication and the aforementioned abnormal indicators. If the external medical knowledge base does not record the relevant association relationship, it is determined to be an unexpected improvement signal.

8. The RWS-based drug post-marketing effect tracking platform according to claim 7, characterized in that: The result analysis layer further includes a visualization module, which is used to present the analysis results to the user in a visual form.

Images