A method and system for generating information on conflict verification of a composite cardiovascular antithrombotic drug regimen based on a disease course time window

Abstract

The application discloses a kind of based on course time window's composite cardiovascular antithrombotic drug scheme conflict verification information generation method and system, belong to medical informatization and computer data processing technical field.This method obtains and standardizes multiple-source patient antithrombotic related data, constructs patient antithrombotic time axis object;Priority determination is carried out to effective key event, determines leading event and auxiliary event, and generates course state code according to leading event and state transfer condition;Based on composite scene vector, course state code and auxiliary event superposition verification condition selection rule subset, combined with drug combination code and risk vector carries out conflict matrix matching;After conflict priority arbitration, generate main conflict, secondary conflict, comprehensive conflict level and structured review prompt;When relevant data or rule base version number changes, according to preset dependency relationship, carry out incremental re-verification to affected time axis segment and rule subset.The application is used for the review and quality control of drug scheme under composite antithrombotic scene.

Description

Technical Field

[0001] This invention belongs to the field of medical information technology and computer data processing technology, and relates to a method and system for generating conflict verification information of composite cardiovascular antithrombotic drug regimens based on disease course time windows. Background Technology

[0002] Patients with cardiovascular disease often require anticoagulant and / or antiplatelet therapy. For example, atrial fibrillation (AF), atrial flutter (AFL), acute coronary syndrome (ACS), percutaneous coronary intervention (PCI), valve replacement or repair, left ventricular thrombosis, and venous thromboembolism may all involve antithrombotic regimens of varying intensities and durations. In clinical practice, the same patient may simultaneously present with multiple antithrombotic scenarios, such as AF combined with PCI, ACS combined with AF, or requiring antithrombotic therapy after recent bleeding. In such cases, the treatment strategy depends not only on the disease diagnosis but also on factors such as the timing of the critical event, the stage of the disease, the current medication combination, platelet count (PLT), hemoglobin (Hb), renal function, history of bleeding, and follow-up plans.

[0003] Existing prescription review or rational drug use systems primarily focus on static reviews such as drug contraindications, duplicate medications, dosage, interactions, and liver and kidney function limitations. They are typically designed for single diseases, single drug categories, or single treatment scenarios, making it difficult to handle complex antithrombotic scenarios. Their main shortcomings include: difficulty in identifying the patient's disease stage based on the dates of key events such as PCI, ACS, valve surgery, bleeding, and thrombosis detection; inability to automatically switch review rules for different time windows; difficulty in determining the dominant scenario and verification path when multiple antithrombotic indications or key events overlap; difficulty in prioritizing multiple conflicts such as expired time windows, excessively strong combined antithrombotic drugs, missed antithrombotic drugs, abnormal laboratory indicators, and missed monitoring; and, most systems only provide one-time prescription reminders, lacking closed-loop management mechanisms such as review tasks, processing status, expiration reminders, operation logs, and new data-triggered re-verification.

[0004] Furthermore, relevant medical data is often scattered across electronic medical record (EMR) systems, hospital information systems (HIS), laboratory information systems (LIS), surgical / interventional registration systems, prescription systems, and pharmaceutical prescription review systems. These different systems present medications, event dates, medical orders, test results, and follow-up plans in inconsistent ways, which can easily affect the accurate identification of antithrombotic scenarios, time windows, and medication combinations. In integrated traditional Chinese and Western medicine treatment environments, patients may also use traditional Chinese medicine, prepared Chinese medicines, Chinese medicine injections, hospital-prepared formulations, or self-prepared Chinese medicines with effects such as promoting blood circulation, invigorating qi and blood, clearing the meridians, or improving circulation. Although these medications are not traditional anticoagulants or antiplatelet drugs, they may affect the bleeding risk assessment of patients undergoing combined antithrombotic therapy, those in the bleeding recovery period, those in the perioperative period, and those with abnormal hematological indicators. Existing systems typically do not include these as risk correction factors in the review process.

[0005] Therefore, it is necessary to provide a method for verifying conflicts in antithrombotic medication, which standardizes multi-source medical data, dynamically updates the patient's antithrombotic course status based on the date of key events, and generates corresponding conflict verification information by combining complex antithrombotic scenarios, current medication combinations, risk factors, and rule base version numbers. Summary of the Invention

[0006] In view of this, the present invention aims to solve the following problems: Existing prescription review systems review contraindications, interactions, dosages, and duplicate medications based on static drug information. They cannot identify the patient's antithrombotic disease status based on the dates of key events such as PCI, ACS, valve surgery, bleeding events, and thrombosis and automatically switch conflict verification rules. This makes it difficult to timely and structurally identify risks such as time window expiration, excessive combined antithrombotic therapy, missed antithrombotic therapy, missed monitoring, and combined exposure to traditional Chinese medicine and Western medicine in combined antithrombotic scenarios. Furthermore, they lack multiple conflict priority ranking, safe handling of missing data, safety protection when thrombosis and bleeding risks coexist, and an event-driven closed-loop review mechanism.

[0007] To achieve the above objectives, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for generating conflict verification information for composite cardiovascular antithrombotic drug regimens based on disease duration time windows, comprising the following steps: Acquire multi-source patient antithrombotic-related data from EMR, HIS, LIS, surgical / interventional registration system, outpatient prescription system, discharge medication system, pharmaceutical prescription review system, follow-up system, and manual verification information. The patient antithrombotic-related data includes diagnostic information, operation and surgical information, critical event dates, current antithrombotic medication regimen, laboratory test indicators, risk factors, and follow-up plans. The drug names, antithrombotic categories, critical event dates, test indicators, follow-up plans, medical order types, and data sources in the patient's antithrombotic-related data are standardized to form a unified data object; Based on the unified data object, a patient antithrombotic time axis object is constructed. The time axis object writes the event axis, medication axis, test axis, risk axis, review axis, and rule axis into the same patient-level data structure. The event axis records the PCI date, ACS onset date, stent implantation date, valve surgery or intervention date, bleeding event date, thrombosis discovery date, planned operation date, and the reliability of each key event date. The medication axis records the start and stop intervals of anticoagulants and antiplatelet drugs. The test axis records indicators such as Hb, PLT, glomerular filtration rate, and international normalized ratio, as well as their expiration dates. The risk axis records risk factors and their sources, collection time, or occurrence time. The review axis records the processing status and expiration reminders. The rule axis records the hit rule number, version number, and start / stop status. Based on the key event type, key event date, current date, and preset state transition conditions in the timeline object, the patient's key events are identified. When multiple key events exist, the dominant event and auxiliary events are determined based on event severity, event relevance, antithrombotic intensity requirements, bleeding risk correction, contraindication flags, and scenario overlap. Based on the key event type, key event date, current date, and preset state transition conditions of the dominant event, the patient is mapped to a disease progress status code. The auxiliary events serve as overlay verification conditions in subsequent rule subset selection and conflict matrix matching. The disease progress status code includes at least the following: pre-event or stable baseline status, early post-event intensified status, mid-term combined antithrombotic review status, long-term maintenance antithrombotic status, bleeding or perioperative re-verification status, and new or recurrent thrombotic event re-verification status. The disease progress status code serves as an index field for subsequent selection of state rule subsets. A composite scene vector is generated based on the diagnostic information, operation and surgical information in the unified data object and the event axis in the time axis object. Based on the composite scenario vector, the disease status code, and the superposition verification conditions corresponding to the auxiliary events, select a subset of state rules in the conflict matrix; Identify the current antithrombotic drug combination code and risk vector, and input the composite scenario vector, the disease status code, the drug combination code and the risk vector into the conflict matrix to obtain a set of hit rules. The set of hit rules includes conflict type, conflict level, trigger field, review level and review items. Based on the determination results of the aforementioned dominant and auxiliary events, the superposition verification path of the dominant and auxiliary events is determined, and the conflict priority arbitration is performed on multiple hit rules in the hit rule set to generate primary conflict, secondary conflict and comprehensive conflict level. Based on the set of hit rules and the arbitration result, structured conflict verification information is generated. The structured conflict verification information includes the triggering reason, disease status code, medication combination code, risk correction basis, hit rule number, rule base version number, review level, review items, expiration reminder, processing status and record keeping. The structured conflict verification information is a review prompt information and does not directly generate instructions to stop medication, add medication, change medication or adjust dosage. When key events, planned operation dates, antithrombotic orders, test results, or review processing status change, the current date crosses a preset status transition time threshold, or the rule base version number changes, incremental re-verification is performed on the affected patient timeline segments and corresponding rule subsets, and the disease status code, primary conflict, secondary conflict, review level, due date reminder, and log records are updated.

[0008] Furthermore, the standardization process also includes performing data source credibility scoring on key event dates, drug exposure intervals, and test results, and verifying date format, source type, time sequence, manual confirmation status, and consistency across multiple sources. Low-credibility or contradictory data are not included in the mapping of their corresponding deterministic disease course status, drug combination identification, or risk vector generation, but instead form a set of data to be verified. The update conditions for the disease progress status code include: when a new critical event is recorded, the current date crosses a preset status transition time threshold, or the review processing status changes, the system recalculates the disease progress status code and switches the corresponding status rule subset.

[0009] Furthermore, the event priority determination is based on the severity of the event, the recentity of the event, the required intensity of antithrombosis, the bleeding risk correction, the contraindication flag, and the scenario superposition relationship to calculate the priority score of each key event, select the event with the highest priority as the dominant event, and the rest as auxiliary events, and determine the superposition verification path according to the combination of the dominant event and the auxiliary events. The conflict priority arbitration calculates the conflict priority score of each hit rule based on the conflict type weight, conflict level, risk-corrected level, time window urgency, and scenario superposition coefficient. It generates primary and secondary conflicts according to priority and determines the prompt display order, review level, and closed-loop processing path.

[0010] Furthermore, the dimensions of the conflict matrix include the antithrombotic scenario, disease duration window and status, current antithrombotic drug combination, and risk factors; The generation of the risk vector includes extracting PLT, Hb, glomerular filtration rate (eGFR), liver function, international normalized ratio (INR), fecal occult blood, age, weight, history of bleeding, and recent bleeding manifestations as risk factors. For dynamic indicators of Hb and PLT, the baseline value, previous value, most recent value, absolute decrease value, and relative decrease percentage are calculated and used as the basis for adjusting the conflict level, review level, and re-examination items.

[0011] Furthermore, when key event dates are missing, medication information is missing, test indicators are expired, follow-up plans are missing, data from multiple sources are inconsistent, or the source of medication is unknown, the system does not directly generate a deterministic conflict prompt. Instead, it outputs a prompt to be supplemented, a prompt to be verified, a prompt for insufficient data, or a prompt for conservative review. After supplementing or updating the data, the system performs disease status mapping and conflict matrix verification again. When both the thrombosis risk rule and the bleeding risk rule are triggered, the system generates a compound conflict warning and raises the review level. It also lists the triggering criteria for both the thrombosis risk and the bleeding risk, and does not output one-way instructions to stop medication, add medication, change medication, or adjust dosage.

[0012] Furthermore, the structured conflict verification information also includes patient identifier, visit identifier, composite scenario, risk vector, primary conflict, secondary conflict, trigger field, and explanatory basis. After the doctor or pharmacist confirms, processes, postpones, ignores, or records the reason for continuing the original plan, the system saves the review status, review responsible person, review time, and log information, and generates a review plan and due reminder. The method supports single-case verification mode and batch quality control mode. The batch quality control mode performs batch scanning of patients in departments, wards or specific time periods, and outputs a list of high-risk antithrombotic patients, a list of patients with overdue combined antithrombotic therapy, a list of patients with missing follow-up plans, and statistics on the rate of review.

[0013] Furthermore, the current antithrombotic drug combination code also includes exposure information of antithrombotic related traditional Chinese medicine and prepared Chinese medicine. The system standardizes the name, ingredients, efficacy attributes, source of medication, start-up and stop status and evidence level of the traditional Chinese medicine and prepared Chinese medicine, and inputs them as drug exposure risk correction factors into the conflict matrix. In scenarios involving combined antithrombotic therapy, bleeding recovery period, perioperative period, or abnormal Hb / PLT dynamics, if exposure to traditional Chinese medicine or prepared Chinese medicine with properties of promoting blood circulation, removing blood stasis, clearing collaterals, or improving circulation is identified, the system will generate a risk review prompt for combined antithrombotic therapy with traditional Chinese and Western medicine, a drug source verification prompt, and a record keeping.

[0014] Secondly, the present invention provides a system for generating conflict verification information for composite cardiovascular antithrombotic medication regimens based on a disease duration time window, comprising: The data access and standardization module is used to acquire antithrombotic-related data from multiple sources of patients and perform standardization processing to form a unified data object. The time axis object construction module is used to construct a patient antithrombotic time axis object based on the unified data object. The time axis object includes at least an event axis, a medication axis, a test axis, a risk axis, a review axis, and a rule axis. The disease status mapping module is used to map patients to disease status codes based on the key event type, key event date, current date, and preset status transition conditions of the dominant event determined by the arbitration module, and to use auxiliary events as superimposed verification conditions. The composite scene recognition module is used to generate a composite scene vector based on the diagnostic information, operation and surgical information in the unified data object and the event axis in the time axis object. The rule subset selection module is used to select a subset of state rules in the conflict matrix based on the composite scenario vector, the disease status code, and the superposition verification conditions corresponding to the auxiliary events. The conflict matrix verification module is used to identify the current antithrombotic drug combination code and risk vector. The composite scenario vector, the disease status code, the drug combination code and the risk vector are input into the conflict matrix to obtain the hit rule set. The arbitration module is used to determine the event priority of multiple key events before the disease status mapping, so as to identify the dominant event and the auxiliary event, and to arbitrate the conflict priority of multiple hit rules in the hit rule set to generate the main conflict, secondary conflict and comprehensive conflict level. The output module is used to generate structured conflict verification information based on the set of hit rules and the arbitration result; The closed-loop verification module is used to determine the affected patient timeline segments and corresponding rule subsets based on the pre-established dependencies between input fields, patient timeline segments, and conflict rules when relevant data changes, and to perform incremental re-verification.

[0015] Thirdly, the present invention provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-mentioned method for generating conflict verification information of composite cardiovascular antithrombotic drug regimens based on disease duration time windows.

[0016] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the above-mentioned method for generating conflict verification information of a composite cardiovascular antithrombotic drug regimen based on a disease course time window.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention constructs a patient antithrombotic timeline object and introduces a disease state machine mapping mechanism based on key event dates. This enables the system to automatically identify the patient's antithrombotic disease status based on key event dates such as PCI, ACS, valve surgery, bleeding events, and thrombosis detection. It also selects the corresponding subset of conflict verification rules based on the status code, solving the problems of existing static prescription review systems being unable to identify disease stages and automatically switch rules, thus improving the accuracy of conflict identification in complex antithrombotic scenarios.

[0018] This invention employs a multi-key event priority determination and multi-conflict priority arbitration mechanism. When a patient has multiple key events or simultaneously triggers multiple conflict rules, the system can calculate event priorities based on event severity, event relevance, antithrombotic intensity requirements, bleeding risk correction, contraindication flags, and scenario superposition relationships. Furthermore, it calculates conflict priorities based on conflict type weights, conflict levels, risk-corrected levels, time window urgency, and scenario superposition coefficients, generating primary, secondary, and comprehensive conflict levels. This determines the order of prompt display and the level of review, avoiding the reminder fatigue and disordered prompts caused by the simple listing of reminders in existing systems.

[0019] This invention employs missing data processing and security protection mechanisms. When data such as key event dates, medication information, and test indicators are missing or inconsistent across multiple sources, the system outputs prompts for supplementation, verification, insufficient data, or conservative review, and performs re-verification after supplementation. When thrombosis risk rules and bleeding risk rules are simultaneously triggered, the system generates a compound conflict prompt and increases the review level, while listing the triggering criteria for thrombosis risk and bleeding risk. It does not output one-way medication adjustment instructions, thus improving the system's safety and usability in real clinical environments.

[0020] This invention employs an event-driven incremental re-verification mechanism. When key events, planned operation dates, antithrombotic orders, test results, or review processing status change, the current date crosses a preset status transition time threshold, or the rule base version number changes, the system performs local re-verification only on the affected patient timeline segments and corresponding rule subsets, and updates the primary conflict, secondary conflict, review level, due date reminders, and log records, forming a review closed loop. This avoids repeated full scans and improves processing efficiency.

[0021] This invention standardizes the names, ingredients, efficacy attributes, sources of medication, start-up and discontinuation status, and evidence levels of antithrombotic Chinese medicines and proprietary Chinese medicines, and uses these as drug exposure risk correction factors input into the conflict matrix. In scenarios such as combined antithrombotic therapy, bleeding recovery period, perioperative period, or dynamic abnormality of Hb / PLT, the system can generate risk review prompts for combined Chinese and Western medicine antithrombotic therapy, medication source verification prompts, and record keeping, reducing missed identifications and mechanical prompts in scenarios of integrated Chinese and Western medicine diagnosis and treatment.

[0022] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will now be described in detail with reference to the accompanying drawings, wherein: Figure 1 A flowchart illustrating the overall technical process of the method for generating conflict verification information for a composite cardiovascular antithrombotic medication regimen based on a disease duration time window, as provided in this embodiment of the invention. Figure 2 A diagram showing the components of a conflict verification information generation system for composite cardiovascular antithrombotic medication regimens based on disease duration time windows, provided in an embodiment of the present invention. Figure 3This is a schematic diagram illustrating the construction of a patient antithrombotic time axis object according to an embodiment of the present invention; Figure 4 A diagram illustrating the priority of multiple critical events and the arbitration process for multiple conflicts provided in this embodiment of the invention; Figure 5 This is a state machine diagram of the antithrombotic disease course time window provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the collision matrix verification provided in an embodiment of the present invention; Figure 7 This is a diagram illustrating the missing data, uncertain data, and security protection processing provided in the embodiments of the present invention. Figure 8 This is a schematic diagram of the verification output and review closed loop provided in an embodiment of the present invention; Figure 9 This is a schematic diagram illustrating the identification and risk correction of exposure to antithrombotic-related traditional Chinese medicines and prepared Chinese medicines provided in an embodiment of the present invention. Detailed Implementation

[0024] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0025] Example 1 like Figure 1 As shown, this embodiment of the invention provides a method for generating conflict verification information for composite cardiovascular antithrombotic drug regimens based on disease duration time windows.

[0026] like Figure 2 As shown in the figure, the conflict verification information generation system for composite cardiovascular antithrombotic medication regimens based on disease course time windows provided in this embodiment of the invention is deployed on an in-hospital server. The in-hospital server includes a processor, a memory, a communication interface, and a data bus. The processor is a multi-core central processing unit (CPU), and the memory includes a rule base storage area, a patient time axis object storage area, and a conflict verification result storage area.

[0027] The system interacts with EMR, HIS, LIS, surgical / interventional registration system, outpatient prescription system, discharge medication system, pharmaceutical prescription review system, and follow-up system via the hospital intranet. The system is implemented as an independent system, interface service, middleware service, embedded prescription review component, batch quality control platform, or rule validation engine. In terms of human-computer interaction, the system is implemented as an independent webpage system, embedded webpage, doctor workstation plugin, pharmacist prescription review system plugin, medical order creation interface prompt component, batch quality control dashboard, or mobile review page.

[0028] The system comprises a data access and standardization module, a timeline object construction module, a composite scene recognition module, a disease course status mapping module, a rule subset selection module, a conflict matrix verification module, an arbitration module, an output module, and a review closed-loop module. Each of these modules is implemented by a processor executing computer programs stored in memory, collaboratively completing multi-source data access, disease course status recognition, conflict rule matching, review information generation, record keeping, and incremental re-verification. These modules may further include functional units such as key event date extraction, medication combination identification, risk factor correction, multi-event priority determination, missing data processing, security protection, standardization of exposure to traditional Chinese medicine and prepared Chinese medicines, perioperative pre-verification, and batch quality control. These functional units can be implemented as sub-units of the aforementioned modules, or they can be implemented collaboratively by two or more modules; they are not limited to independently set modules.

[0029] like Figure 1 As shown, the method provided in this embodiment of the invention is executed by the processor of the aforementioned in-hospital server, and includes the following steps: S1, Data Acquisition.

[0030] The processor's data access and standardization module acquires basic patient information, diagnostic information, procedure and surgical information, critical event dates, current antithrombotic medication regimen, exposure information to antithrombotic-related traditional Chinese medicines and proprietary Chinese medicines, laboratory test indicators, past bleeding history, thrombotic risk factors, follow-up plans, prescription sources, and manual confirmation information. Data sources include EMR, HIS, LIS, surgical / interventional registration system, outpatient prescription system, discharge medication system, pharmaceutical prescription review system, follow-up system, and manually verified and entered information.

[0031] Input data items include: Basic information: age, gender, weight, hospital admission number, outpatient admission number, department, current date; Cardiovascular diagnosis: AF, AFL, ACS, ST-segment elevation myocardial infarction (STEMI), non-ST-segment elevation myocardial infarction (NSTEMI), unstable angina, stable coronary artery disease, valvular heart disease, heart failure, left ventricular thrombosis, venous thromboembolism (VTE), etc. Procedures or events: PCI, stent implantation, coronary artery bypass grafting (CABG), transcatheter aortic valve replacement (TAVR / TAVI), surgical valve replacement, valve repair, transcatheter edge-to-edge repair (TEER), left atrial appendage occlusion, thrombus detection, bleeding events, surgery, endoscopy, etc. Key event dates: ACS onset date, PCI date, stent implantation date, valve surgery or intervention date, bleeding date, thrombus detection date, last review date, planned follow-up date, etc. Current antithrombotic drugs include: warfarin, rivaroxaban, dabigatran, apixaban, edoxaban, aspirin, clopidogrel, ticagrelor, prasugrel, indobufen, etc. Medication attributes: dosage, frequency, start date, expected discontinuation date, number of days to take medication upon discharge, prescription source, type of doctor's order, whether it is a long-term medication, etc.; Laboratory indicators: PLT, Hb, creatinine (Cr), eGFR, liver function, INR, fecal occult blood, etc., and record the collection time and validity period of the indicators. Risks of bleeding and thrombosis: previous gastrointestinal bleeding, cerebral hemorrhage, recent melena, subcutaneous ecchymosis, hematuria, risk of falls, recurrence of thrombosis, in-stent thrombosis, etc. Protective and monitoring measures: whether proton pump inhibitors (PPIs) are used, and whether a plan has been established for follow-up examinations of complete blood count, renal function, INR, fecal occult blood, or bleeding symptoms; Manually confirmed information: doctor or pharmacist confirmation, reason for continuing the original plan, adjustment record, person responsible for review, and review time; dynamic laboratory indicators (baseline Hb and PLT upon admission, previous Hb and PLT, most recent Hb and PLT, absolute value of Hb / PLT decrease, percentage decrease, test date, test interval, and whether the preset dynamic warning threshold has been reached); information on planned procedures (type of planned procedures such as coronary angiography, PCI, CABG, TAVR / TAVI, endoscopy, puncture, radiofrequency ablation, left atrial appendage occlusion, and surgery, planned operation date, pre- and post-operative time windows, and risk level of bleeding during the procedure); information on exposure to antithrombotic-related traditional Chinese medicine and proprietary Chinese medicines (proprietary Chinese medicines, Chinese medicine injections, Chinese medicine decoction pieces, hospital-prepared preparations, patient-prepared Chinese medicines, and health care products with effects such as promoting blood circulation, removing blood stasis, invigorating qi and blood, clearing the meridians, and improving circulation).

[0032] S2, Data standardization and construction of patient antithrombotic timeline objects.

[0033] The processor's data standardization module standardizes drug names, drug categories, antithrombotic categories, names and efficacy attributes of traditional Chinese medicine and prepared Chinese medicines, critical event dates, planned operation dates, test indicators, re-examination plans, medical order types, and data sources to form a unified data object.

[0034] Standardization processes include: Drug name standardization: Mapping brand names, generic names, English names, and abbreviations to unified drug codes. For example, mapping enteric-coated aspirin tablets, Bayer aspirin, aspirin, and ASA to aspirin-like antiplatelet drugs. Antithrombotic category mapping: Standardized drugs are mapped to categories such as vitamin K antagonists (VKA), direct oral anticoagulants (DOAC), aspirin, P2Y12 inhibitors, and other antiplatelet drugs; Standardized drug combination: Based on drug category, dosage, frequency, start date, expected discontinuation date and doctor's order status, generate standardized combinations such as no antithrombotic therapy, single antiplatelet therapy (SAPT), dual antiplatelet therapy (DAPT), oral anticoagulation therapy (OAC), OAC+SAPT, OAC+DAPT, etc. Event date standardization: Standardize the dates in structured fields, surgical records, interventional records, discharge summaries, or manually entered data into a standard date format, and indicate the data source; Standardize test results: Record the indicator values, units, collection time, and validity period; Standardize follow-up plans: unify medical orders, discharge medical orders, follow-up plans or manually entered content into fields such as follow-up items, plan date, responsible person and status; Multi-source consistency verification: If the electronic medical record, outpatient prescription, discharge medication and patient self-reported medication are inconsistent, the system will generate a medication source inconsistency verification prompt; Standardization of the names and efficacy attributes of Chinese medicine and prepared Chinese medicines: Map the brand name, generic name, compound composition, decoction piece name, hospital preparation name, alias and patient-prepared Chinese medicine information of prepared Chinese medicines to a unified Chinese medicine medication target, and label the efficacy attributes such as promoting blood circulation and removing blood stasis, clearing the collaterals, invigorating qi and promoting blood circulation, and improving circulation, as well as the potential bleeding risk attributes, evidence level and data source.

[0035] Standardization also includes scoring the data source credibility of key event dates, medication exposure intervals, and test results. The system generates credibility labels based on data source, field structuring level, whether it has been confirmed by a doctor or pharmacist, whether there is consistency across multiple sources, and the data update time. Surgical / interventional registration systems, structured surgical records, or structured medical orders are set to high credibility; unstructured discharge summaries, patient self-reports, or medications brought in from outside the hospital are set to credibility requiring verification. Low-credibility or contradictory data are not included in the mapping of their corresponding deterministic disease status, medication combination identification, or risk vector generation, but instead form a set of data to be verified. like Figure 3 As shown, the processor's time axis object construction module constructs a patient antithrombotic time axis object based on standardized data objects, and writes the event axis, medication axis, test axis, risk axis, review axis and rule axis into the same patient-level data structure.

[0036] The event axis records PCI, ACS, valve surgery, structural heart disease interventions, bleeding events, thrombotic events, planned operation dates, and the reliability of their sources. The medication axis records the start and stop intervals for anticoagulants, antiplatelet drugs, and antithrombotic-related traditional Chinese medicines and proprietary Chinese medicines. The laboratory axis records indicators such as Hb, PLT, eGFR, INR, and fecal occult blood, as well as their expiration dates. The risk axis records risk factors such as past bleeding history, decreased renal function, age, weight, fall risk, and recent bleeding clues, along with their sources, collection time, or occurrence time. The review axis records the indicated treatment status, reasons for continuing the original protocol, the person responsible for review, and expiration reminders. The rule axis records the hit rule number, version number, and start / stop status.

[0037] The timeline object also includes several intermediate data objects: A composite scenario vector is generated by mapping diagnosis, operation, event axis, and medical history, such as AF+PCI, PCI combined with recent bleeding, etc. The disease status code (StateCode) is generated from the date of the critical event, the current date, and the state transition conditions, and its value ranges from Q0 to Q5. The Drug Combination Code is generated from the drug category, exposure range, and doctor's order status. The risk vector is generated by examining dynamic indicators, past medical history, renal function, bleeding clues, and exposure to traditional Chinese medicine. The ConflictResult is generated by the conflict matrix matching and arbitration mechanism.

[0038] S3, composite antithrombotic scenario recognition.

[0039] The processor's composite scene recognition module generates composite scene vectors based on diagnosis, operation, medical history, event axes, and medication axes. These vectors are used to identify single scenarios such as AF, AFL, ACS, PCI, valve replacement or repair, structural heart disease intervention, left ventricular thrombosis, VTE, and high-risk bleeding, as well as composite scenarios such as AF+PCI, PCI with recent bleeding, mechanical valve combined with PCI, left ventricular thrombosis combined with PCI, and VTE combined with PCI. This scene vector does not directly generate treatment recommendations but serves as input for subsequent rule subset selection and conflict matrix matching.

[0040] S4, Key Event Extraction and Validation.

[0041] The processor's critical event date extraction module extracts critical event dates from structured fields, surgical records, interventional registrations, discharge summaries, medical records, unstructured text, or manually entered information. These dates include PCI dates, ACS onset dates, stent implantation dates, valve surgery or interventional dates, bleeding event dates, thrombosis detection dates, planned operation dates, and last review dates. The system verifies date format, source type, time sequence, manual confirmation status, and consistency across multiple sources. Low-confidence or contradictory data is not directly mapped to a deterministic state but instead generates a verification prompt.

[0042] S5, priority determination for multiple critical events.

[0043] Specifically, when the same patient experiences multiple critical events simultaneously, such as PCI, ACS, bleeding, thrombosis, valve surgery, AF, and planned procedures, the processor's multi-event priority module calculates the event priority of each critical event. To determine the dominant events used for disease course mapping and the auxiliary events involved in overlay verification.

[0044] like Figure 4 As shown, the event priority score is expressed as follows:

[0045] Among them, W1 to W6 are preset weights or locally configured weights. System Selection The highest-scoring event is designated as the primary event, while the others are considered secondary events and included in the cumulative verification. If two event scores are close or conflict between the risks of thrombosis and bleeding, the system does not draw a single conclusion but instead outputs a compound conflict warning and increases the level of review.

[0046] The scoring methods mentioned above are not limited to linear weighting; they can also be implemented using level mapping, threshold triggering, hierarchical sorting, rule table matching, or a combination thereof. Each weight and threshold can be configured according to the rule base version number, local quality control strategy, or applicable scenario.

[0047] In one specific implementation, each evaluation item is mapped according to a preset level value; when two events have the same event priority score or the difference does not exceed a preset threshold, the dominant event is determined in sequence according to the taboo item flag, event severity and event relevance, and the remaining events are treated as auxiliary events.

[0048] For example, a patient has had atrial fibrillation (AF) for many years, underwent PCI 3 months ago, and experienced gastrointestinal bleeding 1 month ago. Currently, the patient is still using anticoagulants combined with a P2Y12 inhibitor. The system identifies AF and PCI as the primary antithrombotic needs, while also identifying recent gastrointestinal bleeding as a high-priority re-verification event. The system determines recent gastrointestinal bleeding as the primary event and PCI and AF as secondary events. Based on the recent gastrointestinal bleeding event, the system maps the patient to the Q4 bleeding recovery period re-verification state, using the time window conditions corresponding to the PCI mid-term review status as overlay verification conditions. The system outputs the main conflict and re-examination reminders related to the bleeding recovery period and combined antithrombotic review.

[0049] S6, Disease course status mapping.

[0050] The processor's disease state machine module determines events for state mapping from key events that have passed validity and reliability checks. When a patient has only one valid key event, the disease state code is determined based on the type, date, current date, and preset state transition conditions of that key event. When a patient has multiple valid key events, the dominant event and auxiliary events are determined according to the multi-key event priority determination method described in S5 above. The disease state code is determined based on the type, date, current date, and preset state transition conditions of the dominant event, and the auxiliary events are used as overlay verification conditions.

[0051] like Figure 5 As shown, the antithrombotic disease course time window state machine includes the following states: Q0, Pre-event or Stable Baseline Status: Patients have stable AF, stable coronary artery disease, or a long-term baseline antithrombotic status, and have not experienced any new PCI, ACS, valve surgery, thrombotic, or bleeding events. The Q0 status is not a risk-free status. For patients using aspirin, P2Y12 inhibitors, anticoagulants, or antithrombotic-related traditional Chinese medicines and proprietary Chinese medicines long-term, the system still generates a long-term baseline antithrombotic safety review prompt based on dynamic decline in Hb / PLT, positive fecal occult blood, decreased eGFR, or lack of follow-up plans.

[0052] Q1, Early Enhanced Post-Event Status: The early status following a non-thrombotic recalibration critical event such as ACS, PCI, stent implantation, valve replacement or repair. Systematic verification checks for missed antithrombotic events, whether the initial protocol is too strong or too weak, and whether basic monitoring is lacking. New or recurrent left ventricular thrombosis, venous thromboembolism, or in-stent thrombosis are not directly mapped to Q1, but are first mapped to Q5.

[0053] Q2, Interim Combined Antithrombotic Review Status: After the event, the system enters a phase where periodic assessments are needed to determine whether to continue combined antithrombotic therapy. The system verifies the duration of combined antithrombotic therapy, bleeding risk, follow-up examination items, and whether an expiration review reminder needs to be generated.

[0054] Q3, Long-term maintenance of antithrombotic status: When there is a long time between the critical event and the event, the system will pay attention to whether there is still an overdue or unnecessary combination of intensive antithrombotic treatment, and identify time window overdue conflicts, long-term intensive antithrombotic treatment without review, and the risk of long-term continuation of medication after discharge.

[0055] Q4. Bleeding, Surgery, Invasive Procedures, and Perioperative Re-verification Status: This includes events such as gastrointestinal bleeding, cerebral hemorrhage, significant anemia, significantly decreased platelet count (PLT), surgery, endoscopy, puncture, or planned invasive procedures such as coronary angiography, PCI, CABG, TAVR / TAVI, endoscopy, puncture, and surgery. Systematic verification of Hb / PLT dynamic changes, bleeding recovery period patterns, perioperative antithrombotic initiation and discontinuation verification, missing preoperative key tests, postoperative antithrombotic restart verification, and safety precautions when thrombotic and bleeding risks coexist.

[0056] Q5, New or Recurrent Thrombotic Event Re-verification Status: New or recurrent left ventricular thrombosis, VTE, in-stent thrombosis, etc., have occurred. The system verifies whether the antithrombotic intensity is insufficient, whether there are any omissions in previous antithrombotic protocols, and whether a follow-up time window and treatment course review reminder are needed.

[0057] The status transition conditions are as follows: When a patient in state Q0 has a new non-thrombotic re-verification key event such as PCI, ACS, stent implantation, valve surgery, or intervention, and the date of the key event passes validity and reliability verification, and the difference between the current date and the key event date is within the preset early time window T1, the patient transitions to Q1. When a patient in state Q1 has a dominant event time window that exceeds T1 but does not exceed the preset intermediate time window T2, the patient transitions to Q2. When a patient in state Q2 has a dominant event time window that exceeds T2, and no new bleeding events, perioperative events, or new or recurrent thrombotic events occur, the patient transitions to Q3. In any state, if a new bleeding event, surgery, endoscopy, or puncture record is added, and the current date falls within the preset preoperative time window before the scheduled operation date, or if the dynamic changes in Hb and PLT meet the preset re-verification trigger conditions, the patient transitions to Q4. In any state, if a new or recurrent thrombotic event such as left ventricular thrombosis, venous thromboembolism, or in-stent thrombosis occurs, and the date of the thrombotic event is valid and the reliability reaches the preset threshold, the patient transitions to Q5. After the re-verification process for state Q5 is completed, the newly occurring or recurrent thrombotic event is used as the new dominant event. The disease status is re-mapped based on the difference between the current date and the date of the thrombotic event: if the time difference is within the preset early time window T1, it switches to Q1; if it exceeds T1 but does not exceed the preset mid-term time window T2, it switches to Q2; if it exceeds T2, it switches to Q3. After the re-verification process for state Q4 is completed, it switches to Q0, Q1, Q2, or Q3 based on the type of the original dominant event, the event date, and the current date. If the same patient simultaneously meets the triggering conditions for both Q4 and Q5, the disease status code corresponding to the dominant event is determined according to event priority, and the other event is used as an auxiliary event in the overlay verification.

[0058] Status codes Q0 through Q5 are not only used as prompt text, but also as index fields for the subset of rules called in the conflict matrix. After obtaining the status code, the system first selects the subset of rules that matches the status code and the composite scenario vector, and then performs medication combination and risk vector matching. Rules not selected by the target status code are not included in this round of conflict matrix matching, thereby reducing the scanning of irrelevant rules and disordered prompts.

[0059] In one modified implementation, the preset early time window T1 and preset mid-term time window T2 in the state transition conditions are configured separately according to different key event types. For example, after PCI, T1 is set to 30 days and T2 to 12 months; after ACS onset, T1 is set to 14 days and T2 to 6 months; after valve replacement, T1 is set to 30 days and T2 to 6 months. These time window thresholds are stored in the rule base configuration area of ​​the memory and can be updated according to the rule base version number and adjusted based on hospital-localized parameters.

[0060] S7, Drug Combination Identification.

[0061] The processor's medication combination identification module identifies the current antithrombotic medication combination code based on the standardized drug category, dosage, frequency, start date, expected discontinuation date, doctor's order status, and medication source.

[0062] Combination codes include types such as no antithrombotic drugs, SAPT, DAPT, OAC, OAC+SAPT, OAC+DAPT, VKA, DOAC, special drug combinations, and combined exposure to antithrombotic-related traditional Chinese medicines or proprietary Chinese medicines.

[0063] When there are inconsistencies between inpatient medical orders, outpatient prescriptions, discharge medications, pharmaceutical prescription review records, or the patient's actual medication information, the system will simultaneously generate a medication source verification mark.

[0064] S8, Risk Factor Extraction and Weighted Adjustment.

[0065] The processor's risk factor correction module extracts risk factors such as PLT, Hb, eGFR, liver function, INR, fecal occult blood, age, weight, history of bleeding, recent melena or hematuria, fall risk, thrombosis recurrence, in-stent thrombosis, and exposure to antithrombotic Chinese medicine and proprietary Chinese medicine, and generates a risk vector.

[0066] For dynamic indicators such as Hb and PLT, the system calculates the baseline value, previous value, most recent value, absolute decrease value, and relative decrease percentage, and uses these as the basis for correcting the conflict level, review level, and review items.

[0067] The dynamic adjustment of risk factors is achieved through a deterministic hierarchical rule of "basic rule level + dynamic risk factor + processing status deweighting", which does not rely on machine learning or black box prediction models.

[0068] When PLT falls below a preset threshold or decreases significantly from baseline / previous test result in a short period, the conflict levels related to combined antithrombotic therapy, intensified antithrombotic therapy, perioperative care, and bleeding recovery are increased. When Hb falls below a threshold or decreases significantly from baseline / previous test result in a short period, the conflict level for high-risk bleeding is increased. When eGFR decreases, a reminder to review DOAC dosage, increase the risk of combined antithrombotic therapy, and renal function retest is triggered.

[0069] Example correction logic: The basic conflict level is "Medium"; if the combined PLT is below the threshold, it is raised by one level; if there is a history of gastrointestinal bleeding, it is raised by another level; if a contraindication occurs, it is directly marked as "Severe". The actual threshold and escalation rules are determined by the hospital's configuration or rule base version number.

[0070] S9, Conflict Matrix Matching.

[0071] like Figure 6 As shown, the processor's conflict matrix verification module inputs the composite scenario vector, disease status code, medication combination code, risk vector, and rule base version number into a preset conflict matrix to obtain a set of matching rules. The conflict matrix uses the composite antithrombotic scenario code, disease status code, medication combination code, risk vector code, and rule base version number as matrix keys to match and obtain the conflict type, basic level, correction level, review level, review items, explanation field, and re-verification trigger field.

[0072] Each rule in the conflict matrix is ​​configured with a rule number, applicable scenario, applicable disease stage or time window, triggering condition, risk correction condition, conflict type, basic conflict level, review action, review items, rule base version number, effective date, source of evidence, localized parameters for the hospital or department, and start / stop status. The system locates rule subsets based on the matrix key and generates structured review information based on the matrix output. The rule output is conflict verification information or review prompt information and does not constitute a specific instruction to stop medication, add medication, change medication, or adjust dosage.

[0073] The matrix keys of the conflict matrix must include at least ScenarioCode, StateCode, DrugCombinationCode, RiskVectorCode, and RuleVersion; the matrix output must include at least ConflictType, BaseLevel, AdjustedLevel, ReviewLevel, ReviewItems, TraceFields, and RecheckTrigger.

[0074] For example, rule number R004 applies to scenarios where ScenarioCode is AF+PCI, StateCode is Q2, and DrugCombinationCode is OAC+DAPT. The system adjusts the conflict level based on risk factors such as dynamic Hb decline, abnormal PLT, decreased eGFR, history of bleeding, or recent bleeding manifestations, and outputs a triple antithrombotic duration review prompt, review level, review items, explanatory basis, and re-verification trigger field. The rule number is used for rule retrieval, version management, and verification result traceability, and there is no limit to the total number of rules in the rule base or the continuity of rule numbers.

[0075] In one modified implementation, the conflict matrix employs a hierarchical matching strategy: the first layer matches the subset of state rules corresponding to the StateCode; the second layer matches the ScenarioCode and DrugCombinationCode within the subset of state rules; and the third layer matches the RiskVectorCode from the results of the second layer and performs risk correction. Hierarchical matching reduces the computational cost of a single matrix query, as the processor only needs to scan the subset of rules corresponding to the current state code instead of the entire rule base.

[0076] S10, Multi-Conflict Arbitration.

[0077] like Figure 4 As shown, when the same patient simultaneously matches multiple conflicting rules, the processor's arbitration module calculates the conflict priority. This generates primary conflict, secondary conflict, and comprehensive conflict levels.

[0078] Conflict priority scores are expressed as follows:

[0079] in, Indicates the conflict type weight. This indicates the basic conflict level score. This represents the risk-adjusted score obtained after adjusting the basic conflict level based on medication intensity, risk factors, and contraindications. This indicates the urgency score of the time window. Indicates the scene overlay coefficient. This indicates the reduced-weight score of the processed records. The system will... The highest-level conflict is designated as the primary conflict, and the remaining conflicts are arranged according to... The conflicts are ranked from highest to lowest as secondary conflicts; the overall conflict level is determined based on the basic conflict level and risk adjustment level of the primary conflict.

[0080] The above-mentioned conflict arbitration methods are not limited to linear weighting, but can also be implemented by level mapping, threshold triggering, hierarchical sorting, rule priority queue or a combination thereof; each score and threshold can be configured according to the rule base version number, local quality control strategy or applicable scenario.

[0081] The system has a mechanism to suppress duplicate alerts. When the same patient, rule number, rule base version number, and trigger field have already been processed and logged by a doctor or pharmacist within the validity period, the system reduces the intensity of the duplicate pop-up, switches to an expiration review reminder, or only sends a strong reminder when the patient's timeline data changes, the rule base version number changes, or the risk level is increased.

[0082] S11, missing data processing and security protection.

[0083] like Figure 7 As shown, when key event dates are missing, medication information is missing, test indicators are expired, follow-up plans are missing, data from multiple sources are inconsistent, or the source of traditional Chinese medicine and prepared Chinese medicine is unknown, the processor's missing data processing module does not directly generate a deterministic conflict prompt. Instead, it outputs a prompt to be supplemented, a prompt to be verified, a prompt for insufficient data, or a prompt for conservative review. After supplementing or updating the data, it performs the disease status mapping and conflict matrix verification again.

[0084] The specific methods for handling missing data include: when a critical event date is missing, outputting a "critical event date missing - to be supplemented" prompt, without temporarily determining whether the specific time window has expired; when current medication information is missing, outputting a "medication plan incomplete - to be verified" prompt; when test indicators have exceeded their expiration date, outputting a "risk factor expired - to be re-examined" prompt; when a re-examination plan is missing, outputting a monitoring omission - conflict prompt; when there is a contradiction between diagnosis and medication but insufficient data, entering a conservative review mode; when data from multiple sources is inconsistent, outputting a "medication source inconsistency - to be verified" prompt; when the source of traditional Chinese medicine and prepared Chinese medicine is unknown, outputting a "thrombotic-related traditional Chinese medicine and prepared Chinese medicine exposure to be verified" prompt.

[0085] When both the thrombosis risk rule and the bleeding risk rule are triggered simultaneously, the processor's safety protection module generates a compound conflict warning and elevates the review level. It also lists the triggering criteria for both thrombosis and bleeding risks, without outputting one-way instructions to stop medication, add medication, change medication, or adjust dosage. The system marks the patient in a two-way risk-locked state and outputs the thrombosis risk criteria, bleeding risk criteria, primary conflict, secondary conflict, and the higher-level review level.

[0086] In one modified implementation, the safety protection mechanism employs a tiered response strategy: when both thrombosis risk and bleeding risk are present and their conflict levels are both "low," the system generates a warning about a compound conflict; when at least one of the conflict levels is "medium," the system escalates the review level to joint review by a doctor and pharmacist; when at least one of the conflict levels is "high" or "severe," the system escalates the review level to review by a senior physician and generates a reminder for review within a specified timeframe. The threshold parameters of the tiered response strategy are stored in the local configuration area of ​​the memory.

[0087] S12, Verification information generation.

[0088] like Figure 8 As shown, the processor's output module generates structured conflict verification information based on the set of hit rules and the arbitration result. Output fields include patient identifier, visit identifier, composite scenario, disease status code, medication combination code, risk vector, conflict type, conflict level, primary conflict, secondary conflict, triggering reason, triggering field, explanatory basis, hit rule number, rule base version number, review level, review items, expiration reminder, processing status, review responsibility person, and record keeping. The system output is a review prompt message; it does not directly generate instructions to stop medication, add medication, change medication, cancel operations, or adjust dosage.

[0089] The system also saves a complete rule path tracking record from the data source to the final prompt output. The rule path tracking record includes the data source field, standardized mapping relationship, scene identification basis, state mapping condition, hit rule number, rule library version number and output field, which are used by doctors and pharmacists to quickly determine the source of the reminder.

[0090] S13, Review closed loop and event-driven incremental re-verification.

[0091] like Figure 8 As shown, after the doctor or pharmacist confirms, processes, postpones, ignores, or records the reason for continuing the original plan, the processor's review closed-loop module saves the review status, the person responsible for the review, the review time, and the record information, and generates a review plan and an expiration reminder.

[0092] When adding or updating key events, planned operation dates, antithrombotic orders, exposure to antithrombotic-related traditional Chinese medicine and proprietary Chinese medicines, test results such as Hb, PLT, and eGFR, review nodes, or review processing status, when the current date crosses a preset status transition time threshold, or when the rule base version number changes, the system determines the patient timeline segment and corresponding rule subset affected by the changed input field based on the pre-established dependencies between input fields, patient timeline segments, and conflict rules. Incremental re-verification is performed only on the determined patient timeline segments and corresponding rule subsets, reusing the original verification results of unaffected patient timeline segments and rule subsets, and updating the disease status code, primary conflict, secondary conflict, review level, due date reminder, and log records.

[0093] The dependencies are stored in the form of a field-time axis-rule association table, index table, or rule configuration field. For example, dependencies are established between Hb test results, Hb data segments in the test axis, Hb-related risk components in the risk vector, and conflicting rules referencing Hb values ​​or their dynamic decline rates. When the Hb test results are updated, the system updates the corresponding test axis data segments and Hb-related risk components, and performs revalidation on the conflicting rules referencing the risk components. When the updated Hb test results do not meet the disease state transition conditions, the original scene recognition results, disease state mapping results, and validation results of other irrelevant rules are reused. When the disease state transition conditions are met, the disease state code is recalculated and the corresponding subset of state rules is called.

[0094] The pseudocode-based deterministic data processing flow is as follows: Input the patient's antithrombotic timeline object T, the current date D, and the rule base version number V.

[0095] A1: Extract the set of valid key events E from the event axis of T, and filter out invalid or unverified events based on the credibility of the data source.

[0096] A2: Calculate the event priority for each critical event e. Select the dominant event and retain the auxiliary events.

[0097] A3: Output the disease status code Q based on the dominant event type, the time difference between the event date and D, and the state transition conditions.

[0098] A4: Select a rule subset based on the composite scene vector C and the status code Q. .

[0099] A5: Identify the current medication combination code M and risk vector F, input C, Q, M, and F into the conflict matrix to obtain the hit rule set H.

[0100] A6: Calculate the conflict priority for each rule in H. This generates primary conflict, secondary conflict, and comprehensive conflict levels.

[0101] A7: Outputs structured review information including trigger field, rule number, rule base version number, status code, review level, review items, processing status, and re-verification trigger conditions.

[0102] A8: When the status of events, medical orders, tests, or reviews in T changes, the current date D crosses a preset status transition time threshold, or the rule base version number V changes, determine the affected timeline segments and rule subsets, and re-execute the relevant processing in A1 to A7 for the timeline segments and rule subsets.

[0103] like Figure 9 As shown, in one embodiment, the present invention also includes a mechanism for identifying and correcting exposure to antithrombotic-related traditional Chinese medicines (TCMs) and proprietary Chinese medicines. The processor's TCM and proprietary Chinese medicine exposure standardization module standardizes the names, ingredients, efficacy attributes, potential bleeding risk attributes, evidence levels, drug sources, and start / stop status of TCMs, proprietary Chinese medicines, TCM injections, hospital-prepared preparations, and patient-prepared TCMs with efficacy attributes such as promoting blood circulation, removing blood stasis, clearing collaterals, and improving circulation, and constructs an exposure map of antithrombotic-related TCMs and proprietary Chinese medicines.

[0104] In scenarios involving combined antithrombotic therapy, intensified antithrombotic therapy, bleeding recovery, perioperative periods, or dynamic abnormalities in Hb / PLT, if exposure to traditional Chinese medicine (TCM) or prepared TCM products with properties of promoting blood circulation, removing blood stasis, clearing the meridians, or improving circulation is identified, the processor's combined TCM and Western medicine risk correction module will input this as a drug exposure-type risk correction factor into the conflict matrix, generating a combined TCM and Western medicine antithrombotic risk review prompt, medication source verification prompt, re-examination items, and record keeping. When the TCM or prepared TCM product exposure information comes from patient self-reports, external imports, in-hospital preparations, or follow-up records, and the name, dosage, frequency, source, or discontinuation date is incomplete, the system will output a prompt indicating that verification is needed or that data is insufficient, and will re-execute the disease progress mapping and conflict matrix verification after supplementation.

[0105] The treatment of exposure to traditional Chinese medicine and prepared Chinese medicine does not change the technical structure of this invention, which is based on the state machine of the antithrombotic course time window driven by the date of the critical event. Instead, it serves as an extension mechanism for drug exposure risk correction and drug source verification.

[0106] In one embodiment, the present invention also includes a perioperative pre-verification mechanism. When a patient has a planned procedure such as coronary angiography, PCI, CABG, TAVR / TAVI, endoscopy, puncture, or surgery, the processor's perioperative pre-verification module writes the planned procedure date as the date of a planned critical event into the patient's antithrombotic timeline object, and forms a preoperative time window, a procedure day time window, and a postoperative restart review time window before and after the planned procedure date. The system combines current antithrombotic medication, dynamic changes in Hb and PLT, bleeding clues, high-risk thrombotic events, and the validity period of key tests to generate perioperative antithrombotic risk review prompts, test record supplementation prompts, and closed-loop records. This module does not directly output instructions to cancel the procedure, stop medication, add medication, change medication, or adjust the dosage.

[0107] In one implementation, the method of the present invention supports both single-case verification mode and batch quality control mode. Single-case verification mode generates structured conflict verification information for a single patient in real time. Batch quality control mode uses the processor's batch quality control module to perform batch scanning of patients by department, ward, doctor group, or specific time period, outputting a list of high-risk antithrombotic patients, a list of patients with overdue combined antithrombotic therapy, a list of patients with missing follow-up plans, a list of unresolved conflicts, and statistics on the reviewed rate. Batch quality control mode supports statistics and export based on dimensions such as rule number, scenario type, conflict level, ward, doctor, pharmacist, and processing status.

[0108] This invention also provides an electronic device, including a processor, a memory, a communication interface, and a data bus. The memory stores a computer program, a rule base version number, localized configuration parameters, a patient antithrombotic timeline object, and conflict verification results. When the processor executes the computer program, it calls the rule base and the patient antithrombotic timeline object in the memory to perform the aforementioned steps of data standardization, disease progression mapping, conflict matrix matching, multi-event priority determination, multi-conflict arbitration, missing data processing, security protection, verification information generation, and closed-loop review.

[0109] This invention also provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the above-described method for generating conflict verification information for composite cardiovascular antithrombotic medication regimens based on disease duration time windows.

[0110] Example 2 This embodiment uses the review of the duration of triple antithrombotic therapy in AF+PCI patients as an example to illustrate the implementation effect of the method of the present invention.

[0111] A 72-year-old male patient, diagnosed with atrial fibrillation (AF), underwent percutaneous coronary intervention (PCI) and stent implantation on January 10, 2025. His current date is April 25, 2025. His current medication regimen is rivaroxaban, aspirin, and clopidogrel (OAC+DAPT combination). His hemoglobin (Hb) is 112 g / L, a decrease of 18 g / L from one month ago; his platelet count (PLT) is 155 × 10⁻⁶.9 / L.

[0112] After the processor constructs the patient's antithrombotic timeline object, it identifies the composite scenario as AF+PCI and maps the patient to the Q2 mid-term combined antithrombotic review status based on the difference between the PCI date and the current date (approximately 3.5 months).

[0113] The system selects the rule subset corresponding to Q2 and inputs the scenario vector, status code Q2, medication combination code OAC+DAPT, and risk vector into the conflict matrix. The matrix matches a "triple antithrombotic duration verification conflict" (rule number R004). After calculation, the primary conflict "Duration of combined antithrombotic therapy needs to be reviewed" and the secondary conflict "Hb decrease corrects for bleeding risk" are generated, and the doctor or pharmacist is prompted to review the necessity of continued combined antithrombotic therapy and supplementary follow-up plan.

[0114] In contrast, if the same patient is still using DAPT 18 months after PCI and lacks review records, the system will map it to the Q3 long-term maintenance status and identify time window overdue conflicts, generating a long-term enhanced antithrombotic necessity review prompt.

[0115] Example 3 This embodiment uses the example of recent bleeding after PCI combined with exposure to antithrombotic-related traditional Chinese medicine to illustrate the implementation effect of the method of the present invention.

[0116] A 76-year-old male patient underwent PCI and stent implantation on January 5, 2025, with the current date being June 10, 2025. His current medications are aspirin and clopidogrel. Medication records show that the patient recently self-medicated with traditional Chinese medicine for promoting blood circulation and removing blood stasis, but the source, dosage, and discontinuation date are unclear. His medical history records indicate recent gastrointestinal bleeding; the date and source of this bleeding event have been verified for validity and reliability. LIS results show an Hb level of 92 g / L.

[0117] The processor performs multi-event priority determination and calculation. After identifying bleeding-related events as the primary events and PCI events as secondary events, the patient was mapped to the Q4 bleeding recovery period re-verification state, and the Q2 mid-term combined antithrombotic review time window corresponding to the PCI event was used as the superimposed verification condition.

[0118] The system identifies the overlap between current antiplatelet drug exposure and exposure to antithrombotic-related traditional Chinese medicine. The conflict matrix hits "overlapping conflict of bleeding risk between traditional Chinese and Western medicine" (rule number R017), generating a risk review prompt for combined traditional Chinese and Western medicine antithrombotic drugs, a drug source verification prompt, and bleeding-related review items.

[0119] In addition, if the patient also has planned procedures such as an endoscopy appointment, the system will write the planned procedure date as a planned key event into the timeline and call the perioperative pre-verification rule group (R011-R016) to identify the risk of preoperative antithrombotic drug exposure and missing key tests.

[0120] Example 4 This embodiment uses the batch quality control mode to identify high-risk patients who have not received antithrombotic treatment as an example to illustrate the implementation effect of the method of the present invention.

[0121] A department set up a batch quality control task to scan inpatients and outpatients in the cardiovascular department who were taking antithrombotic medications. After the processor's batch quality control module built batch antithrombotic timeline objects, it found an example patient: female, 80 years old, with a history of atrial fibrillation (AF), still using OAC+DAPT after recent PCI; Hb decreased by 20 g / L; eGFR decreased; there were unprocessed review prompts and a lack of follow-up plans.

[0122] The processor identified the patient as belonging to the high-risk group for combined antithrombotic therapy. The conflict matrix matched the "intensified antithrombotic therapy combined with high bleeding risk without closed loop" conflict. The system added the patient to the "high-risk antithrombotic unresolved conflict list" in the quality control list. The primary conflict was "intensified antithrombotic therapy combined with high bleeding risk without closed loop in AF+PCI scenario". The secondary conflicts were "decreased renal function triggering DOAC dose review" and "missing follow-up plan". The results were output to the quality control report for departmental intervention.

[0123] To verify the recognition capability of the system of this invention, a retrospective validation was conducted using 120 desensitized real-world cases. The consensus results reached by cardiologists and clinical pharmacists were used as a reference, and the results were compared with those of ordinary static prescription review rules. The sensitivity, specificity, accuracy, F1 score, and Kappa score of the system of this invention were 90.3%, 86.2%, 88.3%, 88.9%, and 0.77, respectively, while those of ordinary static prescription review rules were 59.7%, 70.7%, 65.0%, 63.8%, and 0.30, respectively. The validation results indicate that the system of this invention has high recognition capability and consistency for events requiring review, such as time window expiration, complex antithrombotic scenarios, missed monitoring, missing dates of key events, and coexisting risks of thrombosis and bleeding. This validation was used to evaluate the system's conflict resolution and event recognition capabilities, and not to evaluate the therapeutic efficacy of specific drugs.

[0124] Furthermore, the present invention is also applicable to the following scenarios: (1) Long-term maintenance status after PCI: The patient was still using DAPT 18 months after PCI and lacked review records. The system identified time window overdue conflicts and generated a long-term review prompt for the necessity of continued enhanced antithrombotic therapy. (2) Perioperative pre-verification: When a patient is scheduled to undergo endoscopic or surgical procedures, the system writes the scheduled operation date into the timeline and calls the perioperative rule group to identify the risk of preoperative antithrombotic drug exposure and test deficiencies. (3) Inconsistent medication information from multiple sources: When outpatient prescriptions, inpatient medical orders and discharge medication records are inconsistent, the system outputs a "medication source inconsistency verification prompt" instead of directly determining a conflict. After manual confirmation by the pharmacist, the system re-enters the conflict matrix matching. The above scenarios are all executed within the framework of the methods S1 to S13 of this invention. The only difference is the specific values ​​of the status code, rule subset and risk vector.

[0125] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for generating conflict verification information for composite cardiovascular antithrombotic medication regimens based on disease duration time windows, characterized in that, Includes the following steps: Acquire multi-source patient antithrombotic-related data from electronic medical records, hospital information systems, laboratory information systems, surgical / interventional registration systems, outpatient prescription systems, discharge medication systems, pharmaceutical prescription review systems, follow-up systems, and manually verified information. The patient antithrombotic-related data includes diagnostic information, operation and surgical information, dates of critical events, current antithrombotic medication regimens, laboratory test indicators, risk factors, and follow-up plans. The drug names, antithrombotic categories, critical event dates, test indicators, follow-up plans, medical order types, and data sources in the patient's antithrombotic-related data are standardized to form a unified data object; Based on the unified data object, a patient antithrombotic timeline object is constructed. This timeline object incorporates an event axis, medication axis, laboratory test axis, risk axis, review axis, and rule axis into the same patient-level data structure. Specifically, the event axis records the dates of percutaneous coronary intervention, acute coronary syndrome onset, stent implantation, valve surgery or intervention, bleeding events, thrombosis detection, planned procedures, and the reliability of each key event date. The medication axis records the start and stop intervals for anticoagulants and antiplatelet drugs. The laboratory test axis records indicators such as hemoglobin, platelets, glomerular filtration rate, and international normalized ratio, along with their expiration dates. The risk axis records risk factors, their sources, collection times, or occurrence times. The review axis records processing status and expiration reminders. The rule axis records the hit rule number, version number, and start / stop status. Multiple key events are prioritized to determine the dominant and auxiliary events. Based on the key event type, key event date, current date, and preset state transition conditions of the dominant event, the patient is mapped to a disease progress status code, and the auxiliary events are used as overlay verification conditions. The disease progress status code includes at least the following states: pre-event or stable baseline state, early post-event intensive state, mid-term combined antithrombotic review state, long-term maintenance antithrombotic state, bleeding or perioperative re-verification state, and new or recurrent thrombotic event re-verification state. The disease progress status code serves as an index field for subsequent selection of a subset of state rules. A composite scene vector is generated based on the diagnostic information, operation and surgical information in the unified data object and the event axis in the time axis object. Based on the composite scenario vector, the disease status code, and the superposition verification conditions corresponding to the auxiliary events, select a subset of state rules in the conflict matrix; Identify the current antithrombotic drug combination code and risk vector, input the composite scenario vector, the disease status code, the drug combination code and the risk vector into the conflict matrix corresponding to the current rule base version number to obtain the hit rule set, which includes conflict type, conflict level, trigger field, review level and review item; The conflict priority of multiple hit rules in the set of hit rules is arbitrated to generate primary conflict, secondary conflict and comprehensive conflict levels; Based on the set of hit rules and the arbitration result, structured conflict verification information is generated. The structured conflict verification information includes the triggering reason, disease status code, medication combination code, risk correction basis, hit rule number, rule base version number, review level, review items, expiration reminder, processing status and record keeping. The structured conflict verification information is a review prompt information and does not directly generate instructions to stop medication, add medication, change medication or adjust dosage. When key events, planned operation dates, antithrombotic orders, test results, or review processing status change, the current date crosses a preset status transition time threshold, or the rule base version number changes, the affected patient timeline segments and corresponding rule subsets are determined based on the pre-established dependencies between input fields, patient timeline segments, and conflict rules. Incremental re-verification is performed on the determined patient timeline segments and corresponding rule subsets, the original verification results of unaffected patient timeline segments and rule subsets are reused, and the disease status code, primary conflict, secondary conflict, review level, due date reminder, and log records are updated.

2. The method according to claim 1, characterized in that: The standardization process also includes performing data source credibility scoring on key event dates, drug exposure intervals, and test results; verifying date format, source type, time sequence, manual confirmation status, and consistency across multiple sources; data with low credibility or that are contradictory are not included in the mapping of their corresponding deterministic disease course status, drug combination identification, or risk vector generation, but instead form a set of data to be verified. The update conditions for the disease progress status code include: when a new critical event is recorded, the current date crosses a preset status transition time threshold, or the review processing status changes, the system recalculates the disease progress status code and switches the corresponding status rule subset.

3. The method according to claim 1, characterized in that: The event priority determination is based on the severity of the event, the recentity of the event, the required intensity of antithrombosis, the bleeding risk correction, the contraindication flag, and the overlapping relationship of the scenarios to calculate the priority score of each key event. The event with the highest priority is selected as the dominant event, and the rest are selected as auxiliary events. The overlapping verification path is determined according to the combination of the dominant event and the auxiliary events. The conflict priority arbitration calculates the conflict priority score of each hit rule based on the conflict type weight, conflict level, risk-corrected level, time window urgency, and scenario superposition coefficient. It generates primary and secondary conflicts according to priority and determines the prompt display order, review level, and closed-loop processing path.

4. The method according to claim 1, characterized in that: The dimensions of the conflict matrix include the antithrombotic scenario, disease duration and status, current antithrombotic drug combination, and risk factors. The generation of the risk vector includes extracting platelets, hemoglobin, glomerular filtration rate, liver function, international normalized ratio, fecal occult blood, age, weight, history of bleeding and recent bleeding manifestations as risk factors. For dynamic indicators of hemoglobin and platelets, the baseline value, previous value, most recent value, absolute decrease value and relative decrease percentage are calculated and used as the basis for correcting the conflict level, review level and re-examination items.

5. The method according to claim 1, characterized in that: When key event dates are missing, medication information is missing, test indicators are expired, follow-up plans are missing, data from multiple sources are inconsistent, or the source of medication is unknown, the system does not directly generate a deterministic conflict prompt. Instead, it outputs a prompt to be supplemented, a prompt to be verified, a prompt for insufficient data, or a prompt for conservative review. After supplementing or updating the data, the system performs disease status mapping and conflict matrix verification again. When both the thrombosis risk rule and the bleeding risk rule are triggered, the system generates a compound conflict warning and raises the review level. It also lists the triggering criteria for both the thrombosis risk and the bleeding risk, and does not output one-way instructions to stop medication, add medication, change medication, or adjust dosage.

6. The method according to claim 1, characterized in that: The structured conflict verification information also includes patient identifier, visit identifier, composite scenario, risk vector, primary conflict, secondary conflict, trigger field and explanatory basis. After the doctor or pharmacist confirms, processes, postpones, ignores or records the reason for continuing the original plan, the system saves the review status, review person in charge, review time and trace information, and generates a review plan and due reminder. The method supports single-case verification mode and batch quality control mode. The batch quality control mode performs batch scanning of patients in departments, wards or specific time periods, and outputs a list of high-risk antithrombotic patients, a list of patients with overdue combined antithrombotic therapy, a list of patients with missing follow-up plans, and statistics on the rate of review.

7. The method according to claim 1, characterized in that: The current antithrombotic drug combination code also includes a combination identifier for characterizing exposure to antithrombotic-related traditional Chinese medicine and prepared Chinese medicine; The system standardizes the names, ingredients, efficacy attributes, sources of use, start-up and stop-up status, and evidence levels of the Chinese medicines and prepared Chinese medicines, and inputs the combined identifiers as drug exposure risk correction factors into the conflict matrix; In scenarios involving combined antithrombotic therapy, bleeding recovery period, perioperative period, or abnormal hemoglobin / platelet dynamics, if exposure to traditional Chinese medicine or prepared Chinese medicine with properties of promoting blood circulation, invigorating qi and blood, clearing collaterals, or improving circulation is identified, the system will generate a risk review prompt for combined antithrombotic therapy with traditional Chinese and Western medicine, a drug source verification prompt, and a record keeping.

8. A system for generating conflict verification information for composite cardiovascular antithrombotic medication regimens based on disease duration time windows, characterized in that, include: The data access and standardization module is used to acquire antithrombotic-related data from multiple sources of patients and perform standardization processing to form a unified data object. The time axis object construction module is used to construct a patient antithrombotic time axis object based on the unified data object. The time axis object includes at least an event axis, a medication axis, a test axis, a risk axis, a review axis, and a rule axis. The disease status mapping module is used to map patients to disease status codes based on the key event type, key event date, current date, and preset status transition conditions of the dominant event determined by the arbitration module, and to use auxiliary events as superimposed verification conditions. The composite scene recognition module is used to generate a composite scene vector based on the diagnostic information, operation and surgical information in the unified data object and the event axis in the time axis object. The rule subset selection module is used to select a subset of state rules in the conflict matrix based on the composite scenario vector, the disease status code, and the superposition verification conditions corresponding to the auxiliary events. The conflict matrix verification module is used to identify the current antithrombotic drug combination code and risk vector. The composite scenario vector, the disease status code, the drug combination code and the risk vector are input into the conflict matrix to obtain the hit rule set. The arbitration module is used to determine the event priority of multiple key events before the disease status mapping to identify the dominant and auxiliary events, and to arbitrate the conflict priority of multiple hit rules in the hit rule set to generate the main conflict, secondary conflict and comprehensive conflict levels. The output module is used to generate structured conflict verification information based on the set of hit rules and the arbitration result; The closed-loop verification module is used to determine the affected patient timeline segments and corresponding rule subsets based on the pre-established dependencies between input fields, patient timeline segments, and conflict rules when relevant data changes, and to perform incremental re-verification.

9. An electronic device comprising a processor and a memory, the memory storing a computer program, wherein the processor executes the computer program to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images