A knowledge base system for consumable intelligent auditing and monitoring

Abstract

The application provides a consumable intelligent auditing and monitoring knowledge base system, comprising: a file library, a rule library, an admission auditing rule engine, a medical order consumable intelligent auditing engine and a pre-discharge auditing engine; the file library is used for storing relevant files in the medical field; the rule library is used for storing various predetermined rules; the admission auditing rule engine is used for auditing hospitalization behaviors according to hospitalization-related rules in the rule library; the medical order consumable intelligent auditing engine is used for auditing medical orders / prescriptions issued by doctors according to consumable regulations in the file library; and the pre-discharge auditing engine is used for auditing treatment details of pre-discharge patients according to medical insurance policies in the file library. The technical scheme of the application adopts the admission auditing rule engine, the medical order consumable intelligent auditing engine and the pre-discharge auditing engine to replace manual auditing, thereby improving auditing efficiency.

Description

Technical Field

[0001] This disclosure relates to the field of medical technology, and in particular to a knowledge base system for intelligent auditing and monitoring of consumables. Background Technology

[0002] In existing technologies, the traditional review of consumables relies heavily on manual verification of each order, which is not only inefficient but also prone to omissions, especially when dealing with a large number of inpatient cases and prescriptions. Medical violations are often covert, and manual review is easily overlooked due to professional limitations, such as unfamiliarity with cross-departmental treatment guidelines or insufficient time and energy. Summary of the Invention

[0003] To overcome the problems existing in related technologies, this disclosure provides a knowledge base system for intelligent audit and monitoring of consumables to solve the above-mentioned problems.

[0004] According to a first aspect of the present disclosure, a knowledge base system for intelligent audit and monitoring of consumables is provided, including: a file library, a rule library, an admission audit rule engine, a medical order consumables intelligent audit engine, and a pre-discharge audit engine;

[0005] The aforementioned file library is used to store relevant files in the medical field, providing the knowledge needed for the admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine to make review judgments.

[0006] The aforementioned rule base is used to store various pre-defined rules, providing the rules needed for review and judgment for the aforementioned admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine;

[0007] The aforementioned admission review rule engine is used to review hospitalization behavior based on hospitalization-related rules in the rule base, and to provide feedback on the review results if violations are found in the hospitalization behavior.

[0008] The aforementioned intelligent audit engine for medical orders and consumables is used to audit doctors' medical orders / prescriptions and non-pharmaceutical medical supplies based on the consumable regulations in the aforementioned document library. If any violations are found in the aforementioned medical orders / prescriptions or non-pharmaceutical medical supplies, the audit results will be fed back.

[0009] The aforementioned pre-discharge review engine is used to review the medical treatment details of pre-discharge patients based on the medical insurance policies in the aforementioned document library, and to provide feedback on the review results if the medical treatment details of the pre-discharge patients are found to violate the medical insurance policies.

[0010] The aforementioned admission review rule engine, the aforementioned medical order and consumables intelligent review engine, and the aforementioned pre-discharge review engine work together for review.

[0011] The aforementioned admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine are also used to automatically update the rule base and formulate personalized rules.

[0012] Consumables include pharmaceuticals and non-pharmaceutical medical supplies.

[0013] In one implementation, the aforementioned admission review rule engine, the aforementioned medical order and consumable intelligent review engine, and the aforementioned pre-discharge review engine are also used to read updated medical policy documents, respectively.

[0014] Key rules in the updated healthcare policy documents were extracted using natural language processing technology.

[0015] Based on the previous healthcare policy document and the updated healthcare policy document, determine the changes to the aforementioned key rules;

[0016] Update the rule base based on the changes to the key rules mentioned above.

[0017] In one implementation, the aforementioned admission review rule engine, the aforementioned medical order and consumable intelligent review engine, and the aforementioned pre-discharge review engine extract the patient's individual characteristics and generate a patient profile during the review process.

[0018] The aforementioned individual characteristics include: age, underlying medical conditions, allergy history, and type of medical insurance.

[0019] Based on the above patient profiles, the relevant rules are adjusted to individualize the rules, resulting in personalized rules adapted to the patients.

[0020] In one implementation, a medical review data platform is also included; the medical review data platform is connected to the admission review rule engine, the medical order and consumable intelligent review engine and the pre-discharge review engine respectively.

[0021] The aforementioned medical review data platform is used to share key data with the aforementioned admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine.

[0022] If the aforementioned hospital admission review rule engine detects that a patient has a predetermined type of disease, it will send an early warning signal to the aforementioned medical review data platform.

[0023] The aforementioned medical review data platform sends additional review instructions to the aforementioned medical order and consumable intelligent review engine and the aforementioned pre-discharge review engine.

[0024] In addition to routinely reviewing prescription drugs issued by doctors, the aforementioned intelligent review engine for medical orders and consumables also conducts additional reviews of precautions for the aforementioned types of diseases, in order to facilitate the treatment of the aforementioned types of diseases.

[0025] In addition to routinely reviewing discharge costs, the pre-discharge review engine also conducts additional reviews of discharge precautions for the aforementioned types of diseases, in order to improve the treatment and rehabilitation outcomes for patients with these diseases.

[0026] In one implementation, the aforementioned file library is used to support online viewing and management of various types of files;

[0027] The aforementioned file management includes searching: searching for target files by entering one or more of the following keywords;

[0028] The above keywords include: document name, policy category, start and end dates of issuance, and document content;

[0029] After the search, a list of files is generated, and each file in the list can be exported.

[0030] In one implementation, the aforementioned rule base is used to support clinical medical staff in viewing rules;

[0031] In the rule base management list, the parameters for each rule include: rule name, application scenario, reminder type, rule meaning, number of times to skip, status, and operation; among which, the above operations include editing or disabling.

[0032] It supports search operations, allowing you to search using one or more of the following keywords: rule name, application scenario, and alert type.

[0033] In one implementation, the management of the aforementioned rule base includes: admission review, inpatient medical orders, outpatient prescriptions, and pre-discharge review;

[0034] Alert types include warnings and blockers;

[0035] The rule base management page includes settings for: rule name, activation status, application scenario, and reminder type.

[0036] In the rule base management list, the parameters for each rule include: rule name, quality control prompt, rule meaning, number of times to skip, status, operation, and reminder type. Among them, the operation includes editing and disabling.

[0037] In one implementation, the aforementioned admission review rule engine, the aforementioned medical order and consumable intelligent review engine, or the aforementioned pre-discharge review engine is further used to, for any rule, when the review detects an operation violation, invoke a pop-up warning window to issue a warning, so as to realize an alarm-type reminder; or invoke a pop-up interception window to intercept, so as to realize an interception function.

[0038] In one implementation, the key indicators of the admission review rule engine, the medical order and consumable intelligent review engine, and the pre-discharge review engine are set in the configuration management page;

[0039] The configuration management page includes a configuration unit for the percentage of warning quota usage.

[0040] The warning quota usage ratio configuration unit includes a color selection subunit and a warning quota usage ratio configuration subunit;

[0041] The configuration subunit for the proportion of early warning quota usage includes:

[0042] A normal value configuration module, which is used to configure the range of normal values;

[0043] The normal value configuration module corresponds to the first color selection box set in the color selection subunit;

[0044] The first color selection box set provides multiple color boxes for selection;

[0045] The warning configuration module is used to configure the scope of the warning configuration; the warning configuration module corresponds to the second color selection box set in the color selection subunit.

[0046] The second color selection box set provides a variety of color boxes to choose from;

[0047] A severe warning configuration module is used to configure the scope of severe warning configuration; the severe warning configuration module corresponds to the third color selection box set in the color selection subunit;

[0048] The third color selection box set includes multiple color boxes for selection;

[0049] Among them, the first color selection box is selected and determined from the first color selection box set;

[0050] Select and confirm the second color selection box in the second color selection box set;

[0051] Select and confirm the third color selection box from the set of third color selection boxes.

[0052] In one implementation, the aforementioned admission review rule engine, the aforementioned medical order and consumable intelligent review engine, or the aforementioned pre-discharge review engine is further used to, when detecting that the aforementioned key indicator falls within the range of the aforementioned normal value configuration, mark the usage percentage progress bar of the aforementioned key indicator with the first color in real time.

[0053] If the above key indicators are detected to fall within the range of the above warning configuration, the usage percentage progress bar of the above key indicators will be marked with the second color in real time, and a warning prompt will pop up.

[0054] If the above key indicators are detected to fall within the range of the above severe warning configuration, the usage percentage progress bar of the above key indicators will be marked with the third color in real time, and a warning prompt will pop up.

[0055] Calling the above payment standard makes it easier for administrators to investigate the cause.

[0056] The technical solutions provided by the embodiments of this disclosure may include the following beneficial effects:

[0057] Compared with existing technologies, the technical solution of this application, comprising an admission review rule engine, a medical order and consumables intelligent review engine, and a pre-discharge review engine, replaces manual review, improves review efficiency, and can complete multi-stage review of a single case in seconds or minutes. It also moves the review node forward, promptly blocking violations and solving the pain points of "slow, missed, and late" manual review. Information barriers exist between the consumables department and the hospital, and consumables supervision relies heavily on post-event data sampling, making it difficult to grasp the dynamics of medical behavior in real time. Within the hospital, information asymmetry between departments can also lead to difficulties in controlling the risk of violations. The admission review rule engine, the medical order and consumables intelligent review engine, and the pre-discharge review engine, through data linkage, achieve coordinated review of admission, doctor's prescription, and discharge, which is beneficial for patients to receive better medical treatment and improve their experience.

[0058] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0059] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0060] Figure 1 This is a block diagram illustrating a knowledge base system for intelligent audit and monitoring of consumables, according to an exemplary embodiment.

[0061] Figure 2 This is an example of a medical insurance document page;

[0062] Figure 3 This is a page illustrating a list of medical insurance policy documents according to an exemplary embodiment;

[0063] Figure 4 This is a policy document management page illustrated according to an exemplary embodiment;

[0064] Figure 5This is a rule base viewing page illustrated according to an exemplary embodiment;

[0065] Figure 6 This is a rule base management page illustrated according to an exemplary embodiment;

[0066] Figure 7 This is an example of an editing rules page;

[0067] Figure 8 This is a rule management page illustrated according to an exemplary embodiment;

[0068] Figure 9 This is a configuration page illustrated according to an exemplary embodiment. Detailed Implementation

[0069] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure.

[0070] It should be noted that all actions involving the acquisition of signals, information, or data in this application are carried out in compliance with the relevant data protection laws and policies of the country where the application is located, and with the authorization granted by the owner of the relevant device.

[0071] Based on this, this application proposes a knowledge base system for intelligent audit and monitoring of consumables, see appendix. Figure 1 The knowledge base system for intelligent auditing and monitoring of consumables includes:

[0072] Document Library 01, Rule Library 02, Admission Review Rule Engine 03, Medical Orders and Consumables Intelligent Review Engine 04, and Pre-Discharge Review Engine 05.

[0073] The aforementioned file library 01 is used to store relevant files in the medical field, providing the knowledge needed for the admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine to make review judgments.

[0074] The aforementioned rule base 02 is used to store various pre-defined rules, providing the rules needed for review and judgment for the aforementioned admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine.

[0075] The aforementioned consumables are various items used throughout the entire patient diagnosis and treatment process, including various medicines and non-pharmaceutical medical supplies.

[0076] The aforementioned rules can be, for example, including but not limited to, rules for the use of drugs and rules for the use of medical devices.

[0077] The review of pharmaceuticals includes, but is not limited to, verifying the validity of the drugs (e.g., whether they are expired) and the rationality of their use (e.g., whether the medication matches the patient's condition). The review of non-pharmaceutical medical supplies includes, but is not limited to, the rationality of the use of various treatment-related items (e.g., surgical equipment, auxiliary supplies, etc.). Usage statistics are also included, such as the amount of disinfectant alcohol used in operating rooms, operating tables, etc.

[0078] The aforementioned admission review rule engine 03 is used to review hospitalization behavior based on hospitalization-related rules in the rule base, and to provide feedback on the review results if violations are found in the hospitalization behavior.

[0079] The aforementioned intelligent audit engine 04 for medical orders and consumables is used to audit doctors' medical orders / prescriptions and non-pharmaceutical medical supplies based on the consumable regulations in the aforementioned knowledge base. If any violations are found in the aforementioned medical orders / prescriptions or non-pharmaceutical medical supplies, the audit results will be fed back.

[0080] In this embodiment, the aforementioned intelligent review engine 04 for medical orders and consumables reviews the medical orders / prescriptions issued by doctors and non-pharmaceutical medical supplies.

[0081] The aforementioned pre-discharge review engine 05 is used to review the treatment details of pre-discharge patients based on the consumable regulations in the aforementioned knowledge base, and to provide feedback on the review results if the treatment details of the pre-discharge patients are found to violate the consumable regulations.

[0082] The aforementioned admission review rule engine, the aforementioned medical order and consumable intelligent review engine, and the aforementioned pre-discharge review engine work together to improve the treatment effect of diseases.

[0083] The aforementioned admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine are also used to automatically update the rule base and formulate personalized rules.

[0084] The technical solution of this application includes an admission review rule engine that reviews hospitalization behavior based on consumable regulations in a rule base, which helps to identify violations of hospitalization behavior; a medical order consumable intelligent review engine that reviews doctors' medical orders / prescriptions and non-pharmaceutical medical supplies based on consumable regulations in a knowledge base, which helps to identify violations of medical orders / prescriptions and non-pharmaceutical medical supplies; and a pre-discharge review engine that reviews the treatment details of pre-discharge patients based on consumable regulations in a knowledge base, which helps to identify violations of consumable regulations in the treatment details of pre-discharge patients.

[0085] In some embodiments, the aforementioned admission review rule engine, the aforementioned medical order and consumable intelligent review engine, and the aforementioned pre-discharge review engine are also used to read updated medical policy documents.

[0086] In this embodiment, automatically reading updated medical policy documents can be achieved in the following ways:

[0087] Information is collected using web crawling technology. Existing crawling tools are developed or used, and the official website addresses to be monitored are set, such as the official website of the National Health Commission or local health commissions. The crawler periodically visits the corresponding web pages, parses the pages according to their HTML structure, and uses techniques such as XPath and CSS selectors to extract content containing policy documents. New or modified files are identified by comparing information such as file titles, publication dates, or file hash values.

[0088] Obtain data via official API interfaces. Check if the target website has an open data API interface. If so, follow the provided API documentation to programmatically send requests to retrieve the data. This method can obtain accurate structured data and reduce the burden of parsing web pages. For example, some government data open platforms, after applying for permissions according to the corresponding rules, can use APIs to obtain the latest medical policy data under specific categories as needed.

[0089] Monitor RSS or Atom feeds. Many government agency websites offer RSS or Atom subscription services. If the website being monitored has this service, a subscription module can be developed to parse the XML or JSON data from the feed to detect new content releases, receive update notifications immediately, and retrieve files.

[0090] After obtaining policy documents, if they are in non-text formats such as PDF or images, you need to convert them into text using tools such as OCR.

[0091] Key rules from the updated healthcare policy documents were extracted using natural language processing technology.

[0092] Based on the previous healthcare policy document and the updated healthcare policy document, the changes to the aforementioned key rules are determined.

[0093] In this embodiment, key rules in medical policy documents are extracted using natural language processing technology. This process needs to be implemented step by step, taking into account the characteristics of the policy text.

[0094] Step 1: Text preprocessing to standardize input data.

[0095] Policy documents may be PDF scans, Word documents, or web page texts. They need to be standardized and cleaned first to pave the way for subsequent NLP tasks.

[0096] Perform format conversion and text extraction on policy documents. For PDF / images, use OCR tools (such as Tesseract, Baidu OCR) to convert non-text into editable text; for web HTML, use BeautifulSoup to extract the main text (filtering out irrelevant content such as advertisements and navigation).

[0097] Unify the text encoding (such as UTF-8), remove garbled characters, line breaks, redundant spaces, etc., and split the text according to "chapters - articles".

[0098] Perform basic cleaning to remove meaningless stop words (such as "of, is, in"), but retain policy-specific related words (such as "shall, shall not, except, effective as of...").

[0099] Standardize professional terms. For example, "basic medical insurance" is uniformly abbreviated as "medical insurance", which can be achieved through a custom dictionary to avoid ambiguity.

[0100] Step 2: Locate key rules and lock in "rule-based texts".

[0101] Not all content in policy documents is a rule. There may be background introductions, guiding principles, etc. It is necessary to first locate rule-based sentences containing "obligations, prohibitions, rights, conditions" (such as "Medical institutions shall...", "Insured persons shall not...").

[0102] Define rule trigger words. Rules usually contain "subject (who) + action (what to do) + condition / constraint (how to do it / what is prohibited from doing)". Trigger words can be classified as follows:

[0103] Subject words: such as "medical institutions, medical insurance agencies, insured persons, designated pharmacies".

[0104] Action words: such as "shall, must, shall not, prohibit, may, need to, need to provide, need to file for record";

[0105] Constraint words: such as "as of... date, validity period, excepted circumstances, if not... overdue, be responsible for...".

[0106] Construct rule templates and use regular expressions or sentence templates to match typical rule sentences. For example:

[0107] Template 1: [Subject word] + [Action word (shall / must)] + [Specific requirement] (such as "Designated hospitals shall upload settlement data within 3 working days").

[0108] Template 2: [Action word (shall not / prohibit)] + [Subject word] + [Specific behavior] (such as "Prohibit the use of medical insurance funds for non-medical expenses").

[0109] Step 3: Extract key information and extract the core elements of the rules.

[0110] After locating rule-based sentences, it is necessary to extract core elements such as "who (subject), what (behavior), conditions (when / how to do), and consequences (what happens if not to do / do it wrong)" and convert the natural language into structured data (such as tables or JSON).

[0111] 1. Named entity recognition to locate "feature entities".

[0112] We need to define "entity types" in the healthcare policy domain, use models to identify and tag key entities in sentences, and define custom entity types including the following:

[0113] Subject: The object of policy constraints / regulation (e.g., "tertiary hospital, insured person Zhang, municipal medical insurance bureau").

[0114] Actions: Specific actions, such as "filing, reimbursement, review, and discontinuation of medication".

[0115] Time: such as "October 1, 2024, within 3 working days, annually".

[0116] Numerical values: such as "50%, 3000 yuan, 2 times".

[0117] Conditions: such as "hospital stay exceeding 15 days, and continuous payment for 6 months".

[0118] Consequences: such as "no reimbursement, a fine of 5 times the amount, and cancellation of designated supplier status".

[0119] Implementation method: Fine-tune the model using a pre-trained model (such as the BERT model), collect labeled data of medical policies (entities of some rule sentences are manually labeled), and fine-tune the model using the labeled data so that the model can learn to recognize policy-specific entities.

[0120] 2. Extract relationships to clarify the logical connections between entities.

[0121] The core of the rules is the "relationship between entities" (such as "subject-behavior", "behavior-condition", "behavior-consequence"), which needs to be extracted through the model.

[0122] Example: For the sentence "Insured individuals who have paid premiums continuously for less than 6 months are not eligible for medical insurance reimbursement," the following needs to be extracted:

[0123] Subject-behavior: Insured persons - enjoying medical insurance reimbursement benefits.

[0124] Behavior-conditions: To enjoy medical insurance reimbursement benefits, you need to have paid premiums continuously for 6 months.

[0125] Behavioral Constraints: Enjoy medical insurance reimbursement benefits - (if not satisfied) not allowed.

[0126] Implementation method: Zero-sample / few-sample extraction based on "prompt": If there is little labeled data, prompts can be given to the model (such as "What is the behavior and subject connected by 'must not' in the sentence?"), allowing the model to directly output the relationship.

[0127] Using the "entity pair + classification" approach: first list the entity pairs in the sentence (e.g., "insured persons - enjoy medical insurance reimbursement benefits"), and then use the model to determine the relationship between the two (e.g., "subject-behavior" or "condition-behavior").

[0128] Step 4: Rule structuring and update comparison, focusing on "new / changed rules".

[0129] Finally, the extracted rules need to be organized into a structured format for easy storage by the software and viewing by users, and compared with historical policies to identify the "key updated rules".

[0130] Rule structuring: Extracted elements are organized into a fixed format, such as JSON. Update comparison: The structured new rules are compared with the historical rule base. The combination of "subject + behavior + core condition" is used to determine whether a rule is new or changed. If the "subject + behavior" is entirely new, it is marked as a "new rule". If the "subject + behavior" is the same, but the "condition / constraint / consequence" has changed (e.g., "continuous payment for 3 months" changed to "6 months"), it is marked as a "rule update", and the changed fields are highlighted. The rule base is updated based on the changes to the key rules mentioned above.

[0131] In this embodiment, the aforementioned admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine can learn policies and update automatically, solving the pain point of frequent policy changes. Existing engines suffer from the problem that medical policies, such as medical insurance reimbursement rules, admission standards, and consumable usage rules, change frequently. Each policy change requires manual rewriting of rules and modification of engine configuration, which is prone to errors and delays. The technical solution of this application provides an engine with the ability to automatically parse policy text and automatically update rules.

[0132] The engine can directly "read" medical policy documents, such as PDFs issued by the National Health Commission and official announcements, and use natural language processing (NLP) to extract key rules, such as "diabetic patients must meet the requirement of fasting blood glucose ≥7.0 mmol / L for 3 consecutive days for admission"; automatically convert the extracted rules into logic that the engine can execute, without the need for manual coding; compare the old and new policies, mark the "rule change points", such as "≥7.0" before and "≥7.3" in the new policy, and automatically update the rule base in the engine.

[0133] In contrast to the passive waiting for rules to be manually introduced in related technologies, the engine in this application "actively learns policies and improves its own rules," thus solving the problem of difficult engine maintenance caused by frequent policy changes in medical scenarios.

[0134] In some embodiments, the above-mentioned admission review rule engine, the above-mentioned medical order and consumable intelligent review engine, and the above-mentioned pre-discharge review engine extract the patient's individual characteristics and generate a patient profile during the review process.

[0135] The above-mentioned individual characteristics include: age, underlying diseases, allergy history, and type of medical insurance.

[0136] Based on the above patient profiles, the relevant rules are adjusted to individualize the rules, resulting in personalized rules adapted to the patients.

[0137] In this embodiment, the aforementioned admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine can determine the patient's personalized profile and dynamically adjust relevant rules based on the patient profile, thus addressing the pain point of "personalized treatment for every patient." In related technologies, regardless of whether the patient is elderly or a child, or has underlying medical conditions, the same set of fixed rules is used for review; for example, "adult hospitalization deposit of 2000 yuan" is directly applied to children. However, in medicine, "individual differences are significant," and fixed rules are prone to omissions (for example, elderly patients may require more lenient "admission examination exemption rules"). The engine in this application has a dynamic adaptation function based on patient characteristics. During review, the engine first extracts the patient's individual characteristics (age, underlying diseases, allergy history, medical insurance type, etc.) to generate a patient profile. It automatically adds "personalized adjustment logic" to the basic rules: for example, if the basic rule is "those who have not had a CT scan in the outpatient department need to have one before admission," but if the patient is "over 80 years old and has mobility issues," the engine automatically triggers a supplementary rule for "exemption from CT scan" (which needs to be linked to the hospital's "special population treatment guidelines").

[0138] In some embodiments, a medical review data platform is also included; the medical review data platform is connected to the admission review rule engine, the medical order and consumables intelligent review engine and the pre-discharge review engine respectively.

[0139] The aforementioned medical review data platform is used to share key data with the aforementioned admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine.

[0140] If the hospital admission review rule engine detects that a patient has a predetermined type of disease, it will send an early warning signal to the medical review data platform.

[0141] The aforementioned medical review data platform sends additional review instructions to the aforementioned medical order and consumable intelligent review engine and the aforementioned pre-discharge review engine.

[0142] In addition to routinely reviewing prescription drugs issued by doctors, the aforementioned intelligent review engine for medical orders and consumables also conducts additional reviews of precautions for the aforementioned types of diseases, in order to facilitate the treatment of these diseases.

[0143] In addition to routinely reviewing discharge costs, the pre-discharge review engine also conducts additional reviews of discharge precautions for the aforementioned types of diseases, in order to improve the treatment and rehabilitation outcomes for patients with these diseases.

[0144] In this embodiment, multiple engines are "linked together to form a closed loop," thus solving the pain point of isolated links.

[0145] The engines for reviewing admission, medical orders, and discharge are often separate, but the actual medical process is interconnected. For example, the "basic medical records" at admission will affect the review of medical orders, and the "medication" in the medical orders will affect the "medical insurance reimbursement" at discharge. The isolation of these engines can easily lead to contradictions such as "admission approved, but missing information discovered during medical order review." This application uses a medical review data platform to connect the three engines and share key data. For example, the "patient allergy history" recorded at admission is automatically synchronized to the medical order engine; and the "special medications" prescribed in the medical orders are automatically synchronized to the discharge engine.

[0146] The system designs cross-process linkage rules. For example, if a patient has a history of hypertension during admission review, it automatically sends a warning signal to the medical order engine. When a doctor prescribes nifedipine (an antihypertensive drug), the medical order engine, in addition to routine dosage verification, will also check whether it matches the patient's current blood pressure value (data from the admission examination). During discharge review, if "costs exceed the limit," it automatically traces back to the "estimated costs during admission review" and the "treatment items in the medical order," indicating "which step went wrong" (e.g., a certain examination fee was omitted during admission), rather than simply stating "the cost is incorrect." This end-to-end data flow and rule linkage solves the problems of "coherence" and "data silos" in the medical process, representing a system-level innovation rather than a simple optimization of a single engine.

[0147] See appendix Figure 2-4 In some embodiments, the aforementioned file library is used to support online viewing and management of various types of files.

[0148] The aforementioned file management includes searching: searching for target files by entering one or more of the following keywords.

[0149] The keywords mentioned above include: document name, policy category, start and end dates of issuance, and document content.

[0150] In this embodiment, the user can enter one or more of the keywords mentioned above to perform a search, such as entering the file name and the publication time range. Alternatively, the user can simply enter the file name to perform a search.

[0151] After the search, a list of files is generated, and each file in the list can be exported.

[0152] In this embodiment, the files in each consumables specification list can be exported. The parameters set for each file include file name, policy category, issuance time, and operation options.

[0153] See appendix Figure 5-8 In some embodiments, the aforementioned rule base is used to support clinical medical staff in viewing rules.

[0154] In the rule base management list, each rule's parameters include: rule name, application scenario, reminder type, rule meaning, number of skips, status, and action. The action includes editing or disabling.

[0155] It supports search operations, allowing you to search using one or more of the following keywords: rule name, application scenario, and alert type.

[0156] In some embodiments, the management of the rule base includes: admission review, inpatient medical orders, outpatient prescriptions, and pre-discharge review.

[0157] The alert types include warnings and blockers.

[0158] The rule base management page includes settings for: rule name, enabled status, application scenario, and reminder type.

[0159] In the rule base management list, the parameters for each rule include: rule name, quality control prompt, rule meaning, number of times to skip, status, operation, and reminder type. Among them, the operation includes editing and disabling.

[0160] In some embodiments, the aforementioned admission review rule engine, the aforementioned medical order and consumable intelligent review engine, or the aforementioned pre-discharge review engine is further used to, for any rule, when the review detects an operation violation, invoke a pop-up warning window to issue a warning, so as to realize an alarm-type reminder; or invoke a pop-up interception window to intercept, so as to realize an interception function.

[0161] Based on the actual workflow in medical scenarios, the above rules can be implemented through "real-time engine verification and scenario triggering." Here are a few specific examples:

[0162] For example, regarding the "age restriction rule for treatment items," when a doctor selects "nebulizer therapy for children" (this item is limited to children aged 3-12) when writing a prescription in the outpatient clinic, the intelligent review engine for medical orders and consumables will automatically retrieve the age information from the patient's medical record. If the patient's actual age is 15 years old, the system will pop up a warning window at the prescription submission stage, displaying "This item is limited to children aged 3-12 years. The current patient's age does not match. Please confirm whether to continue." At this point, the reviewing nurse or doctor can manually verify the information. If there are indeed special circumstances, the warning can be ignored and the prescription can be submitted. This is how "warning-type" reminders are implemented; they do not directly prevent the operation but only indicate the risk.

[0163] Looking at the "price limit rule," when a hospitalized patient is about to be discharged, the cashier enters the "CT scan" item. The hospital's price limit for this item is 300 yuan. If the pre-discharge review engine detects that the entered amount is 350 yuan, it will directly block the operation and pop up a blocking prompt: "The price limit for this item is 300 yuan. The current charge exceeds the limit. Please adjust the amount." It cannot be submitted directly and must be modified to 300 yuan or less to pass. This is the implementation of the "interception type," which directly blocks the violation and forces correction.

[0164] The "item matching rule" is also typical in the context of inpatient medical orders. When a doctor prescribes "antiviral oral liquid" for a patient diagnosed with "hypertension", the intelligent review engine for medical orders and consumables will compare the patient's diagnosis with the drug's indications. If a mismatch is found, the medical order submission will be blocked immediately, prompting "The current drug does not match the patient's diagnosis. Please check the treatment plan". The doctor must readjust the medical order to ensure that the item matches the diagnosis before it can be successfully submitted.

[0165] The "item frequency limit rule" is as follows: for example, "blood routine examination" is limited to a maximum of 2 times per month for outpatients. If the same patient has already had 2 blood routine examinations in a month, when the doctor issues another prescription, the medical order consumables intelligent review engine will pop up a warning: "This patient's blood routine examination has reached the limit frequency (2 times) this month. Please confirm whether it is necessary to issue it." The doctor will judge based on the condition and, if necessary, can make a note of the reason before submitting. The system will record this "skipped" time, and it will be accumulated when the number of skipped times is counted in the future.

[0166] See appendix Figure 9 In some embodiments, the key indicators of the admission review rule engine, the medical order and consumable intelligent review engine, and the pre-discharge review engine are set in the configuration management page.

[0167] The configuration management page includes a configuration unit for the percentage of warning quota usage.

[0168] The warning quota usage ratio configuration unit includes a color selection subunit and a warning quota usage ratio configuration subunit.

[0169] The configuration subunit for the proportion of early warning quota usage includes:

[0170] A normal value configuration module is used to configure the range of normal values.

[0171] The normal value configuration module corresponds to the first color selection box set in the color selection subunit.

[0172] The first color selection box set has multiple color boxes to choose from.

[0173] The warning configuration module is used to configure the scope of the warning configuration; the warning configuration module corresponds to the second color selection box set in the color selection subunit.

[0174] The second color selection box set has multiple color boxes to choose from.

[0175] A severe warning configuration module is used to configure the scope of severe warning configuration; the severe warning configuration module corresponds to the third color selection box set in the color selection subunit;

[0176] The third color selection box set includes multiple color boxes for selection.

[0177] Specifically, select and determine the first color selection box from the first color selection box set.

[0178] Select the second color selection box from the second color selection box set.

[0179] Select and confirm the third color selection box from the set of third color selection boxes.

[0180] The payment standard configuration unit includes a point value / rate input box for employees; a point value / rate input box for residents; and a selection box for medical insurance settlement that distinguishes between employees and residents.

[0181] The multiplier configuration unit includes a lower limit warning value input box and a warning name input box; an upper limit warning value input box and a warning name input box.

[0182] In some embodiments, the aforementioned admission review rule engine, the aforementioned medical order and consumable intelligent review engine, or the aforementioned pre-discharge review engine is further used to, in real time, mark the usage percentage progress bar of the aforementioned key indicator with the first color when it is detected that the aforementioned key indicator falls within the range of the aforementioned normal value configuration.

[0183] If the above key indicators are detected to fall within the range of the above warning configuration, the usage percentage progress bar of the above key indicators will be marked with the second color in real time, and a warning prompt will pop up.

[0184] If the above key indicators are detected to fall within the range of the above severe warning configuration, the usage percentage progress bar of the above key indicators will be marked with the third color in real time, and a warning prompt will pop up.

[0185] Calling the above payment standard makes it easier for administrators to investigate the cause.

[0186] The following example illustrates how hospital medical insurance staff need to configure "early warning rules for the use of outpatient pooled medical insurance funds." The settings for each parameter and their associated effects are as follows on the system's "Rule Configuration" webpage:

[0187] 1. Key indicator threshold setting area.

[0188] At the top, there is a "Monitoring Indicators" drop-down box. Select "Outpatient Fund Usage Ratio" as the monitoring indicator. In the "Key Indicator Threshold Type" radio button below, select "Ratio Threshold" and enter "50%" in the "Warning Quota Usage Ratio Configuration" input box (that is, trigger a warning when the fund usage reaches 50% of the total quota).

[0189] 2. Normal value / warning value / severe warning range configuration area.

[0190] The three are arranged side by side with a "range input box" and a "color selection box".

[0191] "Normal value configuration": Enter "0-50%", click the color selection box (built-in red, yellow, green, blue and other colors), select "first color (green)", and the system will automatically record "the green indicator corresponding to this range".

[0192] "Warning value configuration": Enter "50%-80%", select "Second color (yellow)" in the color selection box, which corresponds to "Yellow label".

[0193] "Severe warning range configuration": Enter "80%-100%", select "Third color (red)" in the color selection box, corresponding to "Red indicator".

[0194] 3. Payment standards and medical insurance settlement configuration area.

[0195] Set up a "Medical Insurance Type Differentiation" module, and set up options to differentiate between "Employee Medical Insurance" and "Resident Medical Insurance".

[0196] For "Payment Standard Configuration", enter "260" in the Point Value / Rate column for employees and "201.1" in the Point Value / Rate column for residents.

[0197] 4. Upper and lower limit warning values ​​and name configuration area.

[0198] "Lower limit warning value configuration": Enter "50%", and enter "low fund usage reminder" in the "warning name" field (triggered when the usage ratio is lower than 50%, prompting "there may be insufficient medical services").

[0199] "Upper limit warning value configuration": Enter "90%" (consistent with the starting value of the severe warning range), and enter "Fund high load warning" in the "Warning name" field (linked with the severe warning; a yellow warning will pop up first when it reaches 90%, and it will automatically switch to a red severe warning after it exceeds 90%).

[0200] 5. Multiplier Configuration Area.

[0201] Set an input box for "Special Disease Multiplier", and fill in "1.2" (if the patient is a special disease, the proportion of fund use is calculated by multiplying by 1.2 times when calculating. For example, if the actual use is 75%, it is judged according to 75%×1.2 = 90%, and the early warning is triggered).

[0202] Example of linkage effect: When the system monitors that the proportion of the use of "Employee Medical Insurance Outpatient Coordination Fund" reaches 72%, because it matches the range of "50%-80%" and corresponds to yellow, the page will immediately mark the "Usage Proportion Progress Bar" of this fund as yellow and pop up the "Fund High Load Warning" prompt in the "Warning Name"; if the proportion of fund use of a special disease patient reaches 91% after being calculated by the multiplier, the progress bar will automatically switch to red (the third color), pop up the "Severe Warning" prompt, and at the same time the system will automatically associate the settlement data of the diagnosis and treatment points of this type of patient according to the "Employee Medical Insurance Payment Standard" to facilitate the administrator to check the reasons for overspending.

[0203] The setting of the above parameters realizes the complete linkage of the monitoring indicators from "numerical input" to "automatic marking + dynamic warning" through the binding of "range - color - warning name" and the distinction of medical insurance types and multipliers.

[0204] Those skilled in the art will readily conceive of other embodiments of the present disclosure after considering the specification and practicing the present disclosure. This application aims to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include known common knowledge or conventional technical means in the technical field not disclosed in the present disclosure. The specification and examples are only considered exemplary, and the true scope and spirit of the present disclosure are pointed out by the following claims.

[0205] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A knowledge base system for intelligent auditing and monitoring of consumables, characterized in that, include: Document library, rule library, admission review rule engine, medical order and consumable intelligent review engine, and pre-discharge review engine; The file library is used to store relevant files in the medical field, and to provide the knowledge needed for the admission review rule engine, the medical order and consumable intelligent review engine and the pre-discharge review engine to make review judgments. The rule base is used to store various pre-defined rules, providing the rules needed for the admission review rule engine, the medical order and consumable intelligent review engine, and the pre-discharge review engine to make review judgments. The hospital admission review rule engine is used to review hospitalization behavior based on hospitalization-related rules in the rule base, and to provide feedback on the review results if violations are found in the hospitalization behavior. The intelligent review engine for medical orders and consumables is used to review the medical orders / prescriptions issued by doctors and non-pharmaceutical medical supplies according to the consumable regulations in the document library. If any violations are found in the medical orders / prescriptions and non-pharmaceutical medical supplies, the review results will be fed back. The pre-discharge review engine is used to review the medical treatment details of pre-discharge patients according to the medical insurance policies in the document library, and to provide feedback on the review results if the medical treatment details of pre-discharge patients are found to violate the medical insurance policies. The hospital admission review rule engine, the medical order and consumables intelligent review engine, and the pre-discharge review engine work together for review. The admission review rule engine, the medical order and consumable intelligent review engine, and the pre-discharge review engine are also used to automatically update the rule base and formulate personalized rules. Consumables include pharmaceuticals and non-pharmaceutical medical supplies; The hospital admission review rule engine, the medical order and consumable intelligent review engine, and the pre-discharge review engine are also used to read updated medical policy documents, respectively. Key rules in the updated healthcare policy document are extracted using natural language processing technology; Based on the previous medical policy document and the updated medical policy document, determine the changes to the key rules; Update the rule base based on the changes to the key rules; Medical review data platform; the medical review data platform is connected to the admission review rule engine, the medical order and consumables intelligent review engine and the pre-discharge review engine respectively; The medical review data platform is used to share key data with the hospital admission review rule engine, the medical order and consumables intelligent review engine, and the pre-discharge review engine. When the hospital admission review rule engine detects that a patient has a predetermined type of disease, it sends an early warning signal to the medical review data platform. The medical review data platform sends additional review instructions to the medical order and consumables intelligent review engine and the pre-discharge review engine; In addition to routinely reviewing prescription drugs issued by doctors, the intelligent review engine for medical orders and consumables also conducts additional reviews of precautions for the predetermined type of disease, in order to facilitate the treatment of the predetermined type of disease. In addition to routinely reviewing discharge costs, the pre-discharge review engine also conducts an additional review of discharge precautions for the pre-determined type of disease, in order to improve the treatment and rehabilitation outcomes for patients with the pre-determined type of disease. The key indicators of the admission review rule engine, the medical order and consumable intelligent review engine, and the pre-discharge review engine can be set in the configuration management page; The configuration management page includes a configuration unit for the percentage of warning quota usage. The warning quota usage ratio configuration unit includes a color selection subunit and a warning quota usage ratio configuration subunit; The configuration subunit for the proportion of early warning quota usage includes: A normal value configuration module, which is used to configure the range of normal values; The normal value configuration module corresponds to the first color selection box set in the color selection subunit; The first color selection box set provides multiple color boxes for selection; The warning configuration module is used to configure the scope of the warning configuration; the warning configuration module corresponds to the second color selection box set in the color selection subunit. The second color selection box set provides a variety of color boxes to choose from; A severe warning configuration module is used to configure the scope of severe warning configuration; the severe warning configuration module corresponds to the third color selection box set in the color selection subunit; The third color selection box set includes multiple color boxes for selection; Among them, the first color selection box is selected and determined from the first color selection box set; Select and confirm the second color selection box in the second color selection box set; Select and confirm the third color selection box from the set of third color selection boxes.

2. The knowledge base system for intelligent auditing and monitoring of consumables according to claim 1, characterized in that, When the admission review rule engine, the medical order and consumable intelligent review engine, and the pre-discharge review engine review patients, they extract the individual characteristics of the patients and generate patient profiles. The individual characteristics include: age, underlying medical conditions, allergy history, and type of medical insurance. Based on the patient profile, the relevant rules are adjusted to obtain personalized rules adapted to the patient.

3. The knowledge base system for intelligent auditing and monitoring of consumables according to claim 1, characterized in that, The file library is used to support online viewing and management of various types of files; The file management includes searching: searching for target files by entering one or more of the following keywords; The keywords include: document name, policy category, start and end dates of issuance, and document content; After the search, a file list is generated, and each file in the file list supports export operations.

4. The knowledge base system for intelligent auditing and monitoring of consumables according to claim 1, characterized in that, The rule base is used to support clinical medical staff in viewing rules; In the rule base management list, the parameters for each rule include: rule name, application scenario, reminder type, rule meaning, number of skips, status, and operation; wherein, the operation includes editing or disabling; It supports search operations, allowing you to search using one or more of the following keywords: rule name, application scenario, and alert type.

5. The knowledge base system for intelligent auditing and monitoring of consumables according to claim 1, characterized in that, The management of the rule base includes: admission review, inpatient medical orders, outpatient prescriptions, and pre-discharge review; Alert types include warnings and blockers; The rule base management page includes settings for: rule name, activation status, application scenario, and reminder type. In the rule base management list, the parameters for each rule include: rule name, quality control prompt, rule meaning, number of times to skip, status, operation, and reminder type. Among them, the operation includes editing and disabling.

6. The knowledge base system for intelligent auditing and monitoring of consumables according to claim 5, characterized in that, The admission review rule engine, the medical order and consumables intelligent review engine, or the pre-discharge review engine are also used to, for any rule, when the review detects an operation violation, invoke a pop-up warning window to issue a warning, so as to realize an alarm-type reminder; or invoke a pop-up interception window to intercept, so as to realize an interception function.

7. The knowledge base system for intelligent auditing and monitoring of consumables according to claim 1, characterized in that, The hospital admission review rule engine, the medical order and consumables intelligent review engine, or the pre-discharge review engine are also used for... If the key indicator is detected to fall within the range of the normal value configuration, the usage percentage progress bar of the key indicator will be marked with the first color in real time. If the key indicator is detected to fall within the range of the warning configuration, the usage percentage progress bar of the key indicator will be marked with the second color in real time, and a warning prompt will pop up. If the key indicator is detected to fall within the range of the severe warning configuration, the usage percentage progress bar of the key indicator will be marked with the third color in real time, and a warning prompt will pop up.

Images