Intelligent middle-end processing method and system suitable for multi-source claims review

Abstract

The present invention provides an intelligent middle-end processing method and system suitable for multi-source claim review parties. A claim declaration tree is constructed according to claim data, and the document amounts at each time node in the claim declaration tree are compared in layers to obtain amount review results; standard project data in the review party knowledge base is retrieved based on the claim type of the claim end, and the project review results are obtained by comparing the standard project data with the actual project data of the corresponding invoice node in the claim declaration tree; the actual drugs and actual doses in the claim declaration tree are determined according to the project review results, and the predicted doses are determined based on the similarity quantitative strategy and the actual drugs, and the actual doses are compared with the predicted doses to obtain dose review results; the sub-document information at each document node in the claim declaration tree is processed according to the amount review results, the project review results and the dose review results, and a claim feedback tree is generated; in response to the confirmation information of the claim feedback tree by the claim end, a claim list is generated and sent to the claim end.

Description

Technical Field

[0001] The present invention relates to data processing technology, and in particular to an intelligent middle-end processing method and system suitable for multi-source claims review. Background Art

[0002] The intelligent middle platform modularizes, components and plugs AI technologies such as deep learning, computer vision, knowledge graphs, and natural language understanding, and applies them to the middle platform. It integrates AI capabilities such as hardware computing power, algorithm training capabilities, model deployment capabilities, and basic business display capabilities, closely combines the data resources of the middle platform, and encapsulates them into an overall middle platform system. This can greatly improve the intelligent efficiency of the enterprise, and can therefore be applied in the insurance industry for intelligent claims processing.

[0003] However, the traditional claims review process often relies on manual operations, which is not only inefficient but also prone to errors. In addition, due to the diversity and complexity of data sources, the difficulty and workload of manual review will be greatly increased. At the same time, manual review is also easily affected by subjective factors, resulting in inconsistent review results.

[0004] Therefore, how to conduct comprehensive, accurate and efficient review of the diverse claims data of different claimants has become an urgent problem that needs to be solved. Summary of the invention

[0005] The present invention provides an intelligent middle-end processing method and system suitable for multi-source claims review, which can conduct a comprehensive, accurate and efficient review of the diverse claims data of different claims reviewers.

[0006] A first aspect of the present invention provides an intelligent middle-end processing method applicable to multi-source claims review, comprising:

[0007] Receive the claim data uploaded by the claim terminal, build a claim declaration tree according to the claim data, perform hierarchical comparison on the invoice amounts at each time node in the claim declaration tree, and obtain the amount review result;

[0008] Based on the claim type of the claim settlement terminal, standard project data in the review party knowledge base is retrieved, and project review results are obtained by comparing the standard project data with actual project data of the corresponding invoice node in the claim declaration tree;

[0009] Determine the actual drugs in the claim declaration tree and the actual dosage of the actual drugs according to the project review results, determine the predicted dosage based on the similarity quantitative strategy and the actual drugs, compare the actual dosage with the predicted dosage, and obtain the dosage review result;

[0010] The sub-document information at each document node in the claim declaration tree is processed according to the amount audit result, the project audit result and the dosage audit result to generate a claim feedback tree, and in response to the confirmation information of the claim feedback tree by the claim end, a claim list is generated and sent to the claim end.

[0011] Optionally, in a possible implementation of the first aspect, the receiving the claim data uploaded by the claim settlement terminal and constructing a claim declaration tree according to the claim data includes:

[0012] Receive the claim data uploaded by the claim settlement terminal, wherein the claim data includes the claim type, personnel information, medical institution and the total document information corresponding to the medical institution;

[0013] Constructing a claim node based on the personnel information, constructing an institution node based on the medical institution, and connecting the institution node with the claim node;

[0014] Obtaining a span time period of the total document information corresponding to the medical institution, and evenly dividing the span time period according to a preset number of nodes to obtain multiple sub-time periods;

[0015] Constructing a time node according to the sub-time period, connecting the time node with the organization node, dividing the total document information based on each sub-time period, and obtaining the sub-document information corresponding to the sub-time period;

[0016] Classify the sub-document information based on the document category to obtain invoice document information corresponding to the invoice category, drug document information corresponding to the drug category, and inspection document information corresponding to the inspection category;

[0017] Constructing an invoice node based on the invoice category and binding the invoice document information to the corresponding invoice node; constructing a drug node based on the drug category and binding the drug document information to the corresponding drug node;

[0018] An inspection node is constructed based on the inspection category, and the inspection document information is bound to the corresponding inspection node. The invoice node, the drug node and the inspection node are connected to the corresponding time node to generate a claim declaration tree.

[0019] Optionally, in a possible implementation of the first aspect, performing hierarchical comparison on the invoice amounts at each time node in the claim declaration tree to obtain an amount review result includes:

[0020] Counting the invoice amounts of the invoice document information corresponding to the invoice node connected to the corresponding institution node in the claim declaration tree to obtain the total invoice amount;

[0021] Counting the drug amounts corresponding to the drug document information of the drug node connected to the corresponding institution node in the claim declaration tree to obtain the total drug amount, and counting the inspection amounts corresponding to the inspection document information of the inspection node connected to the corresponding institution node in the claim declaration tree to obtain the total inspection amount;

[0022] According to the total drug amount and the total inspection amount, the total item amount is obtained, and according to the comparison information of the total invoice amount and the total item amount, a summary comparison result is obtained;

[0023] Determine the drug node and the inspection node connected to the corresponding time node in the claim declaration tree as the project node, and use the drug document information and the inspection document information corresponding to the project node as the project document information, and use the amount corresponding to the project document information as the project amount;

[0024] Obtain the project amount of the project document information corresponding to the project node connected to the time node in the claim declaration tree, and retrieve the corresponding invoice document information based on the project information and project time of the corresponding project document information to obtain the invoice amount of the corresponding invoice document information;

[0025] Obtaining a separate comparison result based on the comparison information of the invoice amount and the corresponding item amount;

[0026] Based on the summary comparison result and the individual comparison result, an amount audit result is obtained, and the amount audit result includes a normal amount result and an abnormal amount result.

[0027] Optionally, in a possible implementation of the first aspect, the retrieving standard project data in the reviewer knowledge base based on the claim type of the claim settlement end, and obtaining a project review result according to a comparison between the standard project data and actual project data of a corresponding invoice node in the claim declaration tree includes:

[0028] Based on the claim type of the claim settlement terminal, drug information and inspection information in the prescription review knowledge base are retrieved as standard item data;

[0029] Obtain drug information and inspection information in the invoice document information corresponding to the invoice node connected to each of the time nodes in the claim declaration tree as actual project data;

[0030] According to the comparison information between the standard project data and the actual project data, a project audit result is obtained, and the project audit result includes a normal project result and an abnormal project result.

[0031] Optionally, in a possible implementation of the first aspect, determining the actual drug and the actual dosage of the actual drug in the claim declaration tree according to the project review result includes:

[0032] When it is determined that the actual project data are all within the standard project data, the project review result is determined to be a normal project result, and based on the normal project result, the drug information in the actual project data is used as the actual drug, and the actual dosage of the actual drug is obtained;

[0033] When it is determined that the actual project data contains information that is not in the standard project data, the project audit result is determined to be an abnormal project result, and the information in the actual project data that is not in the standard project data is used as abnormal information;

[0034] Based on the abnormal project result, the abnormal information in the actual project data is deleted and updated to obtain updated actual project data, the drug information in the updated actual project data is used as the actual drug, and the actual dosage of the actual drug is obtained.

[0035] Optionally, in a possible implementation manner of the first aspect, determining a predicted dose based on a similarity quantification strategy and the actual drug, and comparing the actual dose with the predicted dose to obtain a dose review result includes:

[0036] Based on the actual drug, similar claims adjusters in the prescription review knowledge base are retrieved, and calculations are performed according to the personnel attributes, preference attributes, and pathological attributes of the similar claims adjusters to obtain similar similarities corresponding to the similar claims adjusters;

[0037] Sorting the same type of claims adjusters in descending order according to the same type of similarities to obtain a similarity sequence, and selecting the same type of claims adjusters in the similarity sequence according to a preset number of personnel to obtain associated claims adjusters;

[0038] The maximum dose of each actual drug corresponding to the associated claim adjuster is obtained as the initial dose, and the similarity of the same type corresponding to the associated claim adjuster is used as the associated similarity, and the associated similarity is averaged to obtain the average similarity;

[0039] When it is determined that the average similarity is greater than or equal to the benchmark similarity, the initial dose of each of the actual drugs is used as the predicted dose;

[0040] When it is determined that the average similarity is less than the reference similarity, a calculation is performed according to the reference similarity and the average similarity to obtain a magnification factor, and the initial dose is magnified based on the magnification factor to obtain a predicted dose;

[0041] The actual dose and the predicted dose are compared to obtain a dose audit result, wherein the dose audit result includes a normal dose result and an abnormal dose result.

[0042] Optionally, in a possible implementation of the first aspect, the calculating according to the personnel attributes, preference attributes and pathological attributes of the same type of claims adjusters to obtain the same type of similarity corresponding to the same type of claims adjusters includes:

[0043] The similarity of the same category is obtained by the following formula:

[0044] ,

[0045] in, is the personnel similarity, is the upper limit of the number of personnel attribute types, For the claims side The value of a person's attribute. For the same type of claims adjusters, The value of a person's attribute. For the The weight value of personnel attributes, is the preference similarity, is the upper limit of the number of preferred attribute types, For the claims side The value of the preference attribute, For the same type of claims adjusters, The value of the preference attribute, For the The weight value of preference attribute, is the pathological similarity, is the upper limit of the number of pathological attribute types, For the claims side The value of the pathological attribute, For the same type of claims adjusters, The value of the pathological attribute, For the The pathological attribute weights are For similarity of the same kind.

[0046] Optionally, in a possible implementation manner of the first aspect, when determining that the average similarity is less than a reference similarity, performing calculation according to the reference similarity and the average similarity to obtain a magnification factor, and performing a magnification processing on the initial dose based on the magnification factor to obtain a predicted dose, including:

[0047] When it is determined that the average similarity is less than the reference similarity, a similarity difference is obtained according to a difference between the reference similarity and the average similarity;

[0048] Obtaining an offset multiple according to the product of the similarity difference and the offset coefficient, and obtaining an amplification multiple based on the sum of the offset multiple and the reference multiple;

[0049] Amplifying the initial dose based on the amplification factor to obtain a predicted dose;

[0050] The predicted dose was obtained by the following formula:

[0051] ,

[0052] in, To predict the dose, For the initial dose, is the baseline similarity, is the average similarity, is the offset coefficient, The base multiple.

[0053] Optionally, in a possible implementation of the first aspect, the processing of the sub-document information at each document node in the claim declaration tree according to the amount review result, the item review result, and the dosage review result to generate a claim feedback tree includes:

[0054] When it is determined that the amount audit result is an abnormal amount result, the invoice document information corresponding to the invoice amount is used as the abnormal amount invoice document, and the project document information corresponding to the project amount is used as the abnormal project document;

[0055] When it is determined that the project audit result is an abnormal project result, public project data is obtained based on the intersection of the standard project data and the actual project data, and abnormal project data is obtained based on the difference between the actual project data and the public project data;

[0056] Determine, according to the abnormal project data, the corresponding project document information as the abnormal project document, and the corresponding invoice document information as the abnormal project invoice document;

[0057] When it is determined that the dosage audit result is an abnormal dosage result, the corresponding actual dosage is used as the abnormal dosage, and based on the abnormal dosage, the corresponding drug document information is determined as the abnormal item document, and the corresponding invoice document information is determined as the abnormal dosage invoice document;

[0058] Obtaining an abnormal invoice document according to the abnormal amount invoice document, the abnormal item invoice document and the abnormal dosage invoice document;

[0059] The abnormal invoice document and the abnormal item document are highlighted based on the preset document mark, and the corresponding abnormal item data are highlighted based on the preset item mark;

[0060] Highlighting the corresponding abnormal dose according to a preset dose mark, constructing a dose display slot on one side of the abnormal dose, and adding the corresponding predicted dose into the dose display slot;

[0061] Based on the abnormal invoice document, determine the corresponding invoice node in the claim declaration tree as the abnormal invoice node, and based on the abnormal project document, determine the corresponding project node in the claim declaration tree as the abnormal project node;

[0062] According to the abnormal invoice node and the abnormal project node, corresponding nodes in the claim declaration tree are highlighted to obtain a claim feedback tree;

[0063] The step of responding to the confirmation information of the claim settlement terminal on the claim settlement feedback tree and generating a claim settlement list and sending it to the claim settlement terminal includes:

[0064] In response to the confirmation information of the claim feedback tree by the claim settlement end, the invoice document information is deleted and updated based on the invoice document of the abnormal amount, and the data row where the abnormal item data is located in the corresponding invoice document of the abnormal item is deleted and updated;

[0065] The abnormal dose in the abnormal dose invoice document is replaced and updated according to the corresponding predicted dose to obtain updated invoice document information, the updated invoice document information is summarized to obtain a claim list and send it to the claim settlement end.

[0066] A second aspect of the present invention provides an intelligent middle-end processing system applicable to multi-source claims review, including:

[0067] A receiving module is used to receive the claim data uploaded by the claim terminal, construct a claim declaration tree according to the claim data, perform hierarchical comparison on the invoice amounts at each time node in the claim declaration tree, and obtain the amount audit result;

[0068] A comparison module, for retrieving standard project data in the reviewer knowledge base based on the claim type of the claim settlement end, and obtaining a project review result by comparing the standard project data with the actual project data of the corresponding invoice node in the claim declaration tree;

[0069] A determination module, used to determine the actual drugs in the claim declaration tree and the actual dosage of the actual drugs according to the project review result, determine the predicted dosage based on the similarity quantitative strategy and the actual drugs, compare the actual dosage with the predicted dosage, and obtain the dosage review result;

[0070] A sending module is used to process the sub-document information at each document node in the claim declaration tree according to the amount audit result, the project audit result and the dosage audit result, generate a claim feedback tree, respond to the confirmation information of the claim feedback tree by the claim end, generate a claim list and send it to the claim end.

[0071] According to a third aspect of the present invention, there is provided an electronic device, comprising: a memory, a processor and a computer program, wherein the computer program is stored in the memory, and the processor runs the computer program to execute the first aspect of the present invention and various methods that may be involved in the first aspect.

[0072] According to a fourth aspect of the present invention, a storage medium is provided, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the first aspect of the present invention and various methods that may be involved in the first aspect.

[0073] The beneficial effects of the present invention are as follows:

[0074] 1. The present invention can conduct comprehensive, accurate and efficient audits on various claims data of different claimants. The present invention can generate a corresponding claim declaration tree based on the uploaded claim data, thereby realizing automated audits, and then obtaining corresponding amount audit results, project audit results and dosage audit results, thereby realizing comprehensive and accurate claims audits. In addition, the present invention can highlight and update the claim declaration tree, which is convenient for claims personnel to check and confirm in time, thereby improving the efficiency of claims audits.

[0075] 2. The present invention can conduct a comprehensive and accurate review of the uploaded claim data, avoid the situation where the uploaded data is incorrect, and reduce the economic losses of the claim. First, the present invention can summarize the total drug amount and the total inspection amount, and then compare them with the total invoice amount of the invoice to determine the accuracy of the uploaded claim amount data, and avoid uploading irrelevant redundant invoice information to cause economic losses in the claim. Secondly, the present invention can retrieve the corresponding standard project data according to the review knowledge base, so that when reviewing the relevant project documents, the corresponding abnormal inspection items and drugs can be accurately determined to improve the accuracy of the claim amount. Finally, the present invention can prevent the claims personnel from mistakenly increasing the drug dosage in the corresponding invoice, resulting in an increase in the claim amount, so that the present invention can further review the corresponding actual drugs and the corresponding actual dosages to ensure the accuracy of the claim review results and reduce the claim losses.

[0076] 3. The present invention can predict the actual dosage of drugs based on existing data information, thereby judging the accuracy of the dosage of drugs in the uploaded document information, improving the accuracy of the audit, ensuring the authenticity and reliability of the claim results, and avoiding emergency losses in claims. Among them, the present invention can calculate the similarity between claim adjusters by comparing the physical conditions, living habits and illnesses of similar claim adjusters in the prescription review knowledge base corresponding to the actual drugs, so as to facilitate the prediction of the actual dosage of drugs of the current claim adjuster based on the actual dosage of drugs of similar claim adjusters, and then judge the authenticity of the uploaded actual dosage data, and ensure the accuracy of the data of the final audited claim amount. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 A flowchart of an intelligent middle-end processing method applicable to multi-source claims review provided by the present invention;

[0078] Figure 2 This is a schematic diagram of uploading claim data provided by the present invention;

[0079] Figure 3 A schematic diagram of a claim declaration tree provided by the present invention;

[0080] Figure 4 A schematic diagram of an abnormal dosage mark provided by the present invention;

[0081] Figure 5 This is a structural diagram of an intelligent middle-end processing system suitable for multi-source claims review provided by the present invention;

[0082] Figure 6 A schematic diagram of the hardware structure of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0083] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0084] like Figure 1 As shown, the present invention provides a flow chart of an intelligent middle platform processing method applicable to a multi-source claims reviewer, including:

[0085] S1, receiving the claim data uploaded by the claim settlement end, constructing a claim declaration tree according to the claim data, performing a hierarchical comparison on the invoice amounts at each time node in the claim declaration tree, and obtaining an amount audit result.

[0086] It can be understood that the claim end is the information terminal of the claims personnel, such as the mobile phone or computer of the patient making the claim, etc. The claim data is the relevant data information for the claim declaration, such as personnel information, disease diagnosis report, medical invoice, etc. The claim declaration tree is a structure tree that can display the claim declaration information, the time node is the node representing the time, the document amount is the amount data in the uploaded document certificate, and the amount review result is the result of the review of the amount.

[0087] It is not difficult to understand that when the insured claims adjuster makes a claim, he can select the claim application in the corresponding claim service module through the information terminal of the relevant personnel, and then upload the corresponding claim data to make a claim application. For example, Figure 2 As shown, you can fill in and upload relevant claim data information such as personal information and hospital visits, so that the system can generate a corresponding claim declaration tree based on the uploaded claim data. The claim data can also include disease diagnosis reports, medicine orders, invoices, etc., so that personnel can intuitively view relevant information through the claim declaration tree, which is convenient for the review of claim data. The corresponding document amounts can also be layered and compared according to the claim declaration tree to avoid personnel from calculating the claim amount incorrectly and ensure the accuracy of the claim data.

[0088] Through the above implementation, the present invention can obtain the amount audit result so that efficient and accurate claims processing can be carried out later.

[0089] In some embodiments, the specific implementation of step S1 (receiving the claim data uploaded by the claim settlement terminal and constructing a claim declaration tree according to the claim data) includes:

[0090] S101, receiving claim data uploaded by the claim settlement terminal, wherein the claim data includes claim type, personnel information, medical institution and total document information corresponding to the medical institution.

[0091] It is understandable that when the claims adjuster makes a claim declaration, since the claim types corresponding to different diseases are also different, the corresponding claim type can be selected according to the insurance type and disease type. The claim type is the type of claim declared, such as cancer, tumor, accidental accidents among major diseases. The personnel information is the identity information of the corresponding person, which may include name, ID number, etc. The medical institution is the medical unit that diagnoses the disease, which may be Hospital A or Hospital B. The total document information is the sum of all documents. For example, when the medical institution is Hospital A, all diagnosis forms, medical forms and payment forms in Hospital A are the total document information.

[0092] Through the above implementation, the present invention can obtain the uploaded claim data, so as to subsequently construct a corresponding claim declaration tree according to the claim data, so as to process the claim data and improve the accuracy of the claim.

[0093] S102, constructing a claim node based on the personnel information, constructing an institution node according to the medical institution, and connecting the institution node with the claim node.

[0094] It can be understood that the claim node is the node corresponding to the personnel information, and the institution node is the node corresponding to the medical institution.

[0095] It is not difficult to understand that if Figure 3 As shown, the organization node and the claim node can be connected to facilitate the subsequent generation of the corresponding claim declaration tree, wherein the corresponding claim node is constructed through personnel information, so that the claim declaration tree corresponds to the personnel information one by one.

[0096] S103, obtaining a span time period corresponding to the total document information of the medical institution, and evenly dividing the span time period according to a preset number of nodes to obtain a plurality of sub-time periods.

[0097] It is understandable that during the diagnosis and treatment of a disease, a person may be diagnosed and treated in multiple medical institutions, and may stay in each medical institution for a period of time, such as 10 days or 30 days. Therefore, in order to facilitate the review of relevant claims data and ensure the accuracy of document data, the time span can be divided according to the types of claims corresponding to different diseases.

[0098] Among them, the span time period is the diagnosis and treatment time in each medical institution. For example, when the treatment time of a person in Hospital A is from March 1 to March 30, the corresponding span time period is 30 days, that is, March 1 to March 30. The number of preset nodes is artificially set in advance according to the type of claim. The treatment cycle of different diseases is different, so the number of corresponding preset nodes may also be inconsistent. For example, the number of preset nodes corresponding to disease A is 5, and the number of preset nodes corresponding to disease B is 2. The sub-time period is the time period obtained by evenly dividing the span time period. For example, when the span time period is from March 1 to March 30, and the number of preset nodes is 3, then after even division, multiple sub-time periods are obtained, namely, from March 1 to March 10, from March 11 to March 20, and from March 21 to March 30.

[0099] S104, constructing a time node according to the sub-time period, connecting the time node with the organization node, dividing the total document information based on each sub-time period, and obtaining sub-document information corresponding to the sub-time period.

[0100] It can be understood that the sub-document information is the document corresponding to the sub-time period, and the corresponding time node is constructed according to the corresponding sub-time period, for example, Figure 3As shown, time node 1 corresponds to the sub-time period from 3.1 to 3.10, time node 2 corresponds to the sub-time period from 3.11 to 3.20, time node 3 corresponds to the sub-time period from 3.21 to 3.30, etc. At the same time, when the sub-time period is from 3.1 to 3.10, the corresponding documents from 3.1 to 3.10 will be selected from the total document information as sub-document information.

[0101] Through the above implementation, the present invention can obtain sub-document information corresponding to each sub-time period, so as to facilitate subsequent layered information comparison and ensure the accuracy of claim data.

[0102] S105, classifying the sub-document information based on the document category to obtain invoice document information corresponding to the invoice category, drug document information corresponding to the drug category, and inspection document information corresponding to the inspection category.

[0103] It is understandable that the claims data uploaded by personnel contain various types of documents, such as payment invoices, medical lists prescribed by doctors, various test reports for physical examinations, etc. Therefore, the relevant documents can be automatically classified for subsequent rapid review to improve the efficiency of claims processing.

[0104] Among them, the document category is the category of the document, for example, it can be an invoice category, a drug category, and an inspection category. The invoice category is the type corresponding to the invoice, the drug category is the document category belonging to the drug information, and the inspection category is the document category belonging to the medical inspection. The invoice document information is the invoice corresponding to the personnel payment, the drug document information is the medicine list prescribed by the doctor's diagnosis, and the inspection document information is the document for medical project inspection, such as a blood report.

[0105] Through the above implementation, the present invention can obtain the document information corresponding to each document category, so as to facilitate the subsequent comparison and review of the corresponding documents.

[0106] S106, constructing an invoice node based on the invoice category, and binding the invoice document information to the corresponding invoice node; constructing a drug node based on the drug category, and binding the drug document information to the corresponding drug node.

[0107] It can be understood that the invoice node is a node corresponding to the invoice category, and the drug node is a node corresponding to the drug category.

[0108] It is not difficult to understand that the invoice document information is bound to the invoice node, and the drug document information is bound to the drug node, so as to facilitate the subsequent review of the relevant document information according to each node.

[0109] S107, constructing an inspection node based on the inspection category, binding the inspection document information with the corresponding inspection node, connecting the invoice node, the drug node and the inspection node with the corresponding time node, and generating a claim declaration tree.

[0110] It can be understood that the inspection node is a node corresponding to the inspection category.

[0111] It is not difficult to understand that if Figure 3 As shown, node C can represent an invoice node, node D can be represented as a drug node, and node F can be represented as a check node, so that the corresponding generated invoice node, drug node, and check node are connected with the corresponding time node to obtain a claim declaration tree.

[0112] Through the above implementation, the present invention can connect the invoice node, the drug node and the inspection node with the corresponding time node to obtain a claim declaration tree.

[0113] In some embodiments, the specific implementation of step S1 (comparing the amounts of the documents at each time node in the claim declaration tree in layers to obtain the amount audit result) includes:

[0114] S108, counting the invoice amounts corresponding to the invoice document information of the invoice nodes connected to the corresponding institution nodes in the claim declaration tree to obtain the total invoice amount.

[0115] It is understandable that each invoice corresponding to the invoice document information has a corresponding invoice amount, so that the relevant invoice amounts can be counted to obtain the total invoice amount corresponding to each medical institution, so as to verify the amount data.

[0116] Among them, the invoice amount is the corresponding amount data in the invoice document, and the total invoice amount is the total invoice amount data corresponding to each medical institution.

[0117] S109, counting the drug amounts of the drug document information corresponding to the drug nodes connected to the corresponding institution nodes in the claim declaration tree to obtain the total drug amount, and counting the inspection amounts of the inspection document information corresponding to the inspection nodes connected to the corresponding institution nodes in the claim declaration tree to obtain the total inspection amount.

[0118] It is understandable that the amount corresponding to each drug can be extracted from the drug document information, so that the drug amount corresponding to the corresponding medical institution can be counted to obtain the total drug amount. At the same time, the inspection amount in the corresponding inspection document information can be counted to obtain the total inspection amount, so that the total drug amount and the total inspection amount can be summarized later, and then compared with the total invoice amount of the invoice to determine the accuracy of the uploaded claim amount data, so as to avoid uploading irrelevant and redundant invoice information and causing economic losses in claims.

[0119] Among them, the drug amount is the amount corresponding to the drug in each drug document information, the total drug amount is the total amount of the drug amount, and the inspection amount is the amount of the medical inspection. For example, if the blood antibody test costs 300 yuan, the corresponding inspection amount is 300 yuan, and the total inspection amount is the amount of all inspection items.

[0120] S110, obtaining a total item amount according to the total drug amount and the total inspection amount, and obtaining a summary comparison result according to the comparison information of the total invoice amount and the total item amount.

[0121] It can be understood that the total item amount is the sum of the total drug amount and the total inspection amount, and the summary comparison result is the result of comparing the total invoice amount and the total item amount, for example, the compared amounts are consistent or inconsistent.

[0122] For example, when the total drug amount is 8,000 yuan and the total inspection amount is 23,000 yuan, the total item amount is 8,000+23,000=31,000. When the total invoice amount is 31,000 yuan, the summary comparison result is that the amount is consistent. When the total invoice amount is 40,000 yuan, the summary comparison result is that the amount is inconsistent.

[0123] S111, determine the drug node and inspection node connected to the corresponding time node in the claim declaration tree as the project node, and use the drug document information and inspection document information corresponding to the project node as the project document information, and use the amount corresponding to the project document information as the project amount.

[0124] It can be understood that the project node is the drug node and inspection node connected to the corresponding time node, the project document information is the drug document information and inspection document information corresponding to the project node, and the project amount is the amount corresponding to the project document information.

[0125] Through the above implementation, the present invention can determine the corresponding project nodes, project document information and project amounts, so as to facilitate the subsequent comparison of data information corresponding to each time node and further determine the certainty of the document information.

[0126] S112, obtaining the project amount of the project document information corresponding to the project node connected to the corresponding time node in the claim declaration tree, and retrieving the corresponding invoice document information based on the project information and project time of the corresponding project document information to obtain the invoice amount of the corresponding invoice document information.

[0127] It is understandable that when the doctor issues the corresponding drug receipt or inspection document, the personnel will conduct corresponding inspections and obtain medicines based on the corresponding drug receipt and inspection document. When the project amounts corresponding to each time node are obtained respectively, the corresponding invoice document information can be retrieved according to the project information and project time in the corresponding project document information, thereby obtaining the invoice amount of the invoice document information, so as to facilitate the subsequent comparison of the corresponding project amount with the invoice amount, and then obtain the corresponding amount audit result.

[0128] The project information refers to the information in the project document, which may be drug information or inspection item information, and the project time refers to the time corresponding to the project document information.

[0129] It is not difficult to understand that comparing the project amount with the invoice amount of the corresponding project and time can avoid the uploaded invoice document being mixed with invoice document information that does not match the corresponding project document information, avoid the invoice amount being greater than the project amount, and improve the accuracy of the claim amount.

[0130] Through the above implementation, the present invention can obtain the project amount and invoice amount at the corresponding time node, so that the document amount corresponding to each sub-time period can be compared and reviewed later to improve the accuracy of the claim amount data.

[0131] S113, obtaining a separate comparison result based on the comparison information of the invoice amount and the corresponding item amount.

[0132] It is understood that the invoice amount is compared with the project amount to obtain a separate comparison result.

[0133] Among them, the comparison information is the information comparing the invoice amount and the project amount, and the individual comparison result is the result of individually comparing the amount information corresponding to each time node. For example, the project amount at the time nodes corresponding to March 1 to March 10 is consistent with the invoice amount.

[0134] S114, obtaining an amount audit result based on the summary comparison result and the individual comparison result, wherein the amount audit result includes a normal amount result and an abnormal amount result.

[0135] It can be understood that the amount audit result is the result of auditing the amount, the normal amount result is the result that the audited amount is normal, and the abnormal amount result is the result that the audited amount is inconsistent and abnormal.

[0136] It is not difficult to understand that by performing a double-layer comparison of the summary comparison results and the individual comparison results to obtain the amount audit results, the relevant document information that caused the abnormal amount results can be accurately determined.

[0137] S2, based on the claim type of the claim settlement end, retrieve the standard project data in the review party knowledge base, and obtain the project review result by comparing the standard project data with the actual project data of the corresponding invoice node in the claim declaration tree.

[0138] It should be noted that different types of claims, i.e. corresponding disease categories, have corresponding diagnostic items and related medications. For example, accidental fractures may include CT, anti-inflammatory drugs, analgesics, etc., but will not include dialysis items, etc. Therefore, the corresponding standard item data can be retrieved according to the review knowledge base, so that when reviewing the relevant project documents, the corresponding abnormal items and drugs can be accurately determined to improve the accuracy of the claim amount.

[0139] It can be understood that the reviewer knowledge base is a database for the reviewer to automatically review the claim data, which can be pre-configured and contains standard project data corresponding to different claim types. The standard project data is standard diagnosis and treatment information corresponding to different diseases, such as diagnosis and treatment items and medication information corresponding to different diseases. For example, the standard project data corresponding to condition A is project check 1, project check 2, project check 3, drug 1, drug 2, drug 3, and the actual project data is the project data information in the uploaded bill information. For example, the actual project data in the relevant documents uploaded by the personnel for claim type condition A contains project check 1, project check 2, project check 3, project check 4, drug 1, drug 2, drug 3, and the project review result is the review result of the data review of the diagnosis item.

[0140] Through the above implementation, the present invention can obtain the project review results, so as to obtain the accurate claim amount later.

[0141] In some embodiments, the specific implementation of step S2 (the step of retrieving standard project data in the reviewer knowledge base based on the claim type of the claim settlement terminal, and obtaining the project review result by comparing the standard project data with the actual project data of the corresponding invoice node in the claim declaration tree) includes:

[0142] S21, based on the claim type of the claim settlement terminal, retrieve the drug information and inspection information in the prescription review knowledge base as standard item data.

[0143] It can be understood that the drug information is the drug type information. For example, according to the claim type of disease A, the corresponding drug information retrieved is drug 1, drug 2, and drug 3. The inspection information is the item information of the inspection items. For example, the inspection items required to diagnose disease A include item inspection 1, item inspection 2, and item inspection 3.

[0144] Through the above implementation, the present invention can obtain corresponding standard project data, so as to facilitate subsequent comparison with actual project data, thereby obtaining corresponding project review results.

[0145] S22, obtaining drug information and inspection information in the invoice document information corresponding to the invoice node connected to each time node in the claim declaration tree as actual project data.

[0146] It can be understood that the drug information and inspection information in the invoice document information corresponding to the invoice node connected to each time node is obtained, and the drug information and inspection information are used as actual project data to facilitate subsequent review and comparison with standard project data.

[0147] S23, obtaining a project review result based on the comparison information between the standard project data and the actual project data, wherein the project review result includes a normal project result and an abnormal project result.

[0148] It can be understood that the normal project result is that the standard project data is consistent with the actual project data, indicating that the project is normal, and the abnormal project result is that the standard project data is inconsistent with the actual project data, indicating that the project is abnormal.

[0149] It is not difficult to understand that the same type of disease may have different corresponding drug information and inspection items due to different severity, but the standard project data contains all drug information and inspection items of different degrees corresponding to the disease. Therefore, when the actual project data is compared with the standard data items, if the actual project data contains project data that is not included in the standard project data, it can be said that the actual project data has abnormal project data outside the claim type, and the corresponding project review result is an abnormal project result. Conversely, when all the information in the actual project data exists in the standard project data, it can be said that the corresponding diagnosis and treatment item is indeed the diagnosis and treatment method of the corresponding disease, that is, the project review result is a normal project result.

[0150] S3, determining the actual drugs in the claim declaration tree and the actual dosage of the actual drugs according to the project review results, determining the predicted dosage based on the similarity quantitative strategy and the actual drugs, comparing the actual dosage with the predicted dosage, and obtaining the dosage review result.

[0151] It is understandable that in order to prevent personnel from mistakenly increasing the drug dosage in the corresponding invoice, thereby increasing the claim amount, the corresponding actual drugs and the corresponding actual dosage can be further reviewed to ensure the accuracy of the final claim amount. At the same time, since the symptoms of different people may differ, it is impossible to obtain accurate reference data for drug dosage based on standard project data. Therefore, this solution can make customized predictions on the corresponding drug dosage based on the similarity quantitative strategy and actual drugs, thereby improving the accuracy of the audit results.

[0152] Among them, the actual drug is the drug information of the corresponding document in the claim declaration tree, the actual dosage is the dosage of the actual drug. For example, when the actual drug is drug A, and the dosage of drug A on the document bound at the node of the claim declaration tree is 10 boxes, then the corresponding actual dosage is 10 boxes, the predicted dosage is the predicted drug dosage, and the dosage review result is the result of the drug dosage review.

[0153] Through the above implementation, the present invention completes the review of drug dosage, thereby obtaining the dosage review result, so as to facilitate the subsequent accurate claim amount.

[0154] In some embodiments, the specific implementation of step S3 (determining the actual drug in the claim declaration tree and the actual dosage of the actual drug according to the project review result) includes:

[0155] S31, when it is determined that the actual project data are all in the standard project data, the project review result is determined to be a normal project result, based on the normal project result, the drug information in the actual project data is used as the actual drug, and the actual dosage of the actual drug is obtained.

[0156] It can be understood that when the actual project data are all in the standard project data, the project review result can be obtained as a normal project result, so that the drug information in the corresponding actual project data can be used as the actual drug, and the actual dosage corresponding to the actual drug can be obtained at the same time, so as to facilitate subsequent comparison with the predicted dosage, thereby obtaining the corresponding dosage review result and improving the accuracy of the claim amount data.

[0157] S32, when it is determined that the actual project data contains information that is not in the standard project data, the project audit result is determined to be an abnormal project result, and the information in the actual project data that is not in the standard project data is used as abnormal information.

[0158] It can be understood that when the actual project data includes project inspection 1, project inspection 2, project inspection 3, drug 1, drug 2, drug 3, and drug 4, and there is drug 4 that is not in the standard project data, then the corresponding project review result is an abnormal project result, which means that there is information in the actual project data that does not belong to the diagnosis and treatment of the corresponding disease, such as drug 4, so that the drug information and / or inspection information in the actual project data that is not in the standard project data can be regarded as abnormal information.

[0159] Through the above implementation, the present invention can obtain abnormal information in the actual project data corresponding to the abnormal project result, so as to obtain the corresponding actual medicine and actual dosage information subsequently.

[0160] S33, based on the abnormal project result, the abnormal information in the actual project data is deleted and updated to obtain updated actual project data, the drug information in the updated actual project data is used as the actual drug, and the actual dosage of the actual drug is obtained.

[0161] It can be understood that after deleting the abnormal information in the actual project data corresponding to the abnormal project results, the corresponding updated actual project data can be obtained. For example, when the actual project data contains Project Check 1, Project Check 2, Project Check 3, Drug 1, Drug 2, Drug 3, and Drug 4, among which Drug 4 is abnormal information, Drug 4 can be deleted, and the updated actual project data can be Project Check 1, Project Check 2, Project Check 3, Drug 1, Drug 2, and Drug 3.

[0162] Among them, the actual drugs are the drug information in the updated actual project data. For example, when the updated actual project data is project inspection 1, project inspection 2, project inspection 3, drug 1, drug 2, drug 3, then drug 1, drug 2, drug 3 can be used as actual drugs, and the drug dosage corresponding to the actual drugs can be obtained as the actual dosage.

[0163] Through the above implementation, the present invention can obtain the actual medicine and actual dosage corresponding to the abnormal item result, so as to facilitate the subsequent comparison with the predicted dosage.

[0164] In some embodiments, the specific implementation of step S3 (determining the predicted dose based on the similarity quantitative strategy and the actual drug, comparing the actual dose with the predicted dose, and obtaining the dose review result) includes:

[0165] S34, based on the actual drug, retrieve similar claims adjusters in the prescription review knowledge base, calculate according to the personnel attributes, preference attributes and pathological attributes of the similar claims adjusters, and obtain the similarity corresponding to the similar claims adjusters.

[0166] It can be understood that the same type of claims adjusters are those who have applied for claims using the corresponding actual drugs. For example, when the actual drug is Drug 1, the personnel who also have Drug 1 can be retrieved from the prescription review knowledge base. When the claims data of A, B, and C in the prescription review knowledge base contains Drug 1, A, B, and C can be regarded as the same type of claims adjusters.

[0167] Among them, personnel attributes are relevant physical attributes of personnel, such as age, weight, etc., preference attributes are preferred information attributes in life, such as living habits, eating habits, etc., pathological attributes are symptom conditions, such as early cancer, late cancer, tumor location, etc., and similarity of the same category is the degree of similarity between the same category of claims adjusters and the current claims adjuster.

[0168] It is not difficult to understand that by comparing the physical conditions, living habits and medical conditions of similar claims adjusters corresponding to the actual drugs, the similarity between people is calculated, which makes it easier to predict the actual drug dosage of the current claims adjuster based on the actual drug dosage of similar claims adjusters, thereby judging the authenticity of the uploaded actual dosage data and ensuring the accuracy of the data of the final reviewed claim amount.

[0169] In some embodiments, the specific implementation of step S34 (the calculation based on the personnel attributes, preference attributes and pathological attributes of the same type of claims adjusters to obtain the same type of similarity corresponding to the same type of claims adjusters) includes:

[0170] S341, the similarity of the same category is obtained by the following formula,

[0171] ,

[0172] in, is the personnel similarity, is the upper limit of the number of personnel attribute types, For the claims side The value of a person's attribute. For the same type of claims adjusters, The value of a person's attribute. For the The weight value of personnel attributes, is the preference similarity, is the upper limit of the number of preferred attribute types, For the claims side The value of the preference attribute, For the same type of claims adjusters, The value of the preference attribute, For the The weight value of preference attribute, is the pathological similarity, is the upper limit of the number of pathological attribute types, For the claims side The value of the pathological attribute, For the same type of claims adjusters, The value of the pathological attribute, For the The pathological attribute weights are For similarity of the same kind.

[0173] It can be understood that different attributes have corresponding weight values, which can be pre-configured according to actual conditions, and different attributes can contain multiple types, and the greater the number, the greater the corresponding similarity will be. For example, when n is larger, the greater the number of sums, and thus the greater the personnel similarity after addition. Similarly, when m is larger, the preference similarity is greater, and when f is larger, the pathological similarity is greater.

[0174] It is not difficult to understand that when the personnel similarity, preference similarity and pathology similarity are greater, the obtained similarity of the same category will also be greater, which can indicate that the condition of the same category of claims personnel is more similar to the current claimant, which is convenient for improving the accuracy of the predicted dose data.

[0175] S35, sorting the same type of claims adjusters in descending order according to the same type of similarities to obtain a similarity sequence, and selecting the same type of claims adjusters in the similarity sequence according to a preset number of personnel to obtain associated claims adjusters.

[0176] It is understandable that there may be a large number of similar claims adjusters retrieved from the prescription review knowledge base. In order to reduce the amount of data processing, similar claims adjusters can be sorted according to similarity, and then the top few with higher similarity can be selected from the sequence after descending sorting, which can be the top three similar claims adjusters, so as to improve the accuracy of subsequent predicted doses.

[0177] Among them, the similarity sequence is a personnel sequence obtained by sorting similar claims adjusters in descending order according to similarities of the same type. The preset number of personnel is the number of pre-set selected personnel, which can be manually pre-set, such as three. The associated claims adjusters are similar claims adjusters selected in the similarity sequence.

[0178] S36, obtaining the maximum dose of each actual drug corresponding to the associated claims adjuster as the initial dose, and taking the similarity of the same type corresponding to the associated claims adjuster as the associated similarity, performing mean calculation on the associated similarity to obtain the average similarity.

[0179] It is understandable that in order to determine the authenticity of the actual dosage of the actual medicine of the current claims adjuster, the maximum dosage of each actual medicine corresponding to the selected associated claims adjuster can be used as the initial dosage to improve the accuracy of subsequent predicted dosages.

[0180] Among them, the initial dose is the maximum dose of each actual drug corresponding to the associated claims adjuster. For example, when the associated claims adjuster A uses 3 boxes of the actual drug (Drug 1), B uses 5 boxes of the actual drug (Drug 1), and C uses 2 boxes of the actual drug (Drug 1), the maximum dose corresponding to the actual drug (Drug 1) is 5 boxes, so the initial dose for Drug 1 is 5 boxes.

[0181] In addition, the associated similarity is the similarity of the same category corresponding to the associated claims adjusters. For example, when the similarity of the associated claims adjuster A is 70%, the associated similarity corresponding to A is 70%, and the average similarity is the average value of all associated similarities.

[0182] S37, when it is determined that the average similarity is greater than or equal to the reference similarity, the initial dose of each corresponding actual drug is used as the predicted dose.

[0183] It can be understood that when the average similarity is greater than or equal to the benchmark similarity, it can be explained that the symptoms, habits, etc. of the selected associated claims adjuster are very similar to those of the current claims adjuster, and the corresponding actual drug dosage will also be similar, so the corresponding initial dosage can be used as the predicted dosage.

[0184] The reference similarity is a reference value for determining the similarity, and may be preset manually.

[0185] Through the above implementation, the present invention can obtain the corresponding predicted dose when the average similarity is greater than or equal to the benchmark similarity, so as to facilitate subsequent comparison with the actual dose of the current claims adjuster and improve the accuracy of dose review.

[0186] S38, when it is determined that the average similarity is less than the reference similarity, a calculation is performed based on the reference similarity and the average similarity to obtain a magnification factor, and the initial dose is magnified based on the magnification factor to obtain a predicted dose.

[0187] It is understandable that when the average similarity is less than the baseline similarity, it means that the similarity between the selected associated claims adjuster and the current claims adjuster is low, so that the drug dosage corresponding to the associated claims adjuster and the drug dosage of the current claims adjuster may have a large deviation. Then, the baseline similarity and the average similarity can be used to calculate the amplification factor, so as to amplify the initial dose, so that the predicted dose is more accurate and consistent with the actual medication dose of the current claims adjuster.

[0188] The magnification factor is the multiple by which the initial dose is magnified.

[0189] It is not difficult to understand that the greater the difference in similarity, the greater the deviation of the corresponding medication dosage may be, thereby increasing the corresponding magnification factor.

[0190] Through the above implementation, the present invention can obtain a predicted dose, so as to facilitate subsequent comparison with the actual dose and improve the accuracy of dose review.

[0191] In some embodiments, the specific implementation of step S38 (when it is determined that the average similarity is less than the reference similarity, a calculation is performed based on the reference similarity and the average similarity to obtain a magnification factor, and the initial dose is magnified based on the magnification factor to obtain a predicted dose) includes:

[0192] S381: When it is determined that the average similarity is less than a reference similarity, a similarity difference is obtained according to a difference between the reference similarity and the average similarity.

[0193] It can be understood that the similarity difference is the difference between the baseline similarity and the average similarity. For example, when the baseline similarity is 95% and the average similarity is 75%, the corresponding similarity difference is 95%-75%=20%.

[0194] S382, obtaining an offset multiple according to the product of the similarity difference and the offset coefficient, and obtaining an amplification multiple based on the sum of the offset multiple and the reference multiple.

[0195] It can be understood that the offset coefficient is an offset value used to represent the similarity difference, which can be preset manually, the offset multiple is the product of the similarity difference and the offset coefficient, and the benchmark multiple is the multiple value of the benchmark for amplification, which can be set manually based on actual conditions. For example, the benchmark multiple can be 1.

[0196] S383, amplifying the initial dose based on the amplification factor to obtain a predicted dose.

[0197] It is understandable that the initial dose is amplified so that a predicted dose can be obtained to facilitate subsequent comparison with the actual dose.

[0198] S384, the predicted dose is obtained by the following formula,

[0199] ,

[0200] in, To predict the dose, For the initial dose, is the baseline similarity, is the average similarity, is the offset coefficient, The base multiple.

[0201] It can be understood that when the initial dose is larger and the difference between the baseline similarity and the average similarity is larger, that is, the similarity difference is larger, it can be explained that the situation of the associated claims adjuster is quite different from that of the current claims adjuster, so the amplification factor will also increase accordingly, and the predicted dose will also increase.

[0202] S39, comparing the actual dose with the predicted dose to obtain a dose audit result, wherein the dose audit result includes a normal dose result and an abnormal dose result.

[0203] It can be understood that by comparing the actual dose with the predicted dose, when the doses are consistent or the actual dose is less than the predicted dose, a normal dose result can be obtained, and when the comparison result is that the actual dose is greater than the predicted dose, an abnormal dose result can be obtained.

[0204] Among them, the dosage audit result is the result of comparing and auditing the dosage of the drug, including normal dosage results and abnormal dosage results. The normal dosage result is the result that the actual dosage value is judged to be within the normal range, and the abnormal dosage result is the result that the actual dosage value is judged to be not within the normal drug dosage range.

[0205] Through the above implementation, the present invention can obtain the dosage audit result, so as to obtain an accurate and true claim list and make claims later.

[0206] S4, processing the sub-document information at each document node in the claim declaration tree according to the amount audit result, the project audit result and the dosage audit result, generating a claim feedback tree, responding to the claim end's confirmation information on the claim feedback tree, generating a claim list and sending it to the claim end.

[0207] It can be understood that the claim feedback tree is a structure tree for providing feedback on claim results, and the claim list is a display list with claim information.

[0208] It is not difficult to understand that the corresponding sub-document information is processed according to the amount audit results, project audit results and dosage audit results. When the audit results are abnormal, the corresponding abnormal information can be deleted and updated. After the update, the claim feedback tree is displayed on the claim end, so that the current claims personnel can confirm the claim information. When the confirmation is completed, the generated corresponding claim list can be sent to the claim end for claim processing, which can reduce the time for manual review and improve the review efficiency while increasing the authenticity and accuracy of the claims data.

[0209] In some embodiments, the specific implementation of step S4 (processing the sub-document information at each document node in the claim declaration tree according to the amount review result, the project review result and the dosage review result to generate a claim feedback tree) includes:

[0210] S401, when it is determined that the amount audit result is an abnormal amount result, the invoice document information corresponding to the invoice amount is used as the abnormal amount invoice document, and the project document information corresponding to the project amount is used as the abnormal project document.

[0211] It is understandable that when the amount audit result is an abnormal amount result, it means that the uploaded invoice document information is inconsistent with the project document information. Therefore, the corresponding document information needs to be marked for subsequent information confirmation to ensure the accuracy of the claim result.

[0212] Among them, the abnormal amount invoice document is the invoice document information corresponding to the corresponding invoice amount when the amount review result is an abnormal amount result, and the abnormal project document is the project document information corresponding to the corresponding project amount when the amount review result is an abnormal amount result.

[0213] For example, when the invoice amount is 8,000 and the item amount is 5,000, an abnormal amount result can be obtained, so that the invoice document corresponding to the invoice amount of 8,000 can be used as the abnormal amount invoice document, and the drug document and / or inspection document corresponding to the item amount of 5,000 can be used as the abnormal item document.

[0214] S402, when it is determined that the project audit result is an abnormal project result, common project data is obtained based on the intersection of the standard project data and the actual project data, and abnormal project data is obtained based on the difference between the actual project data and the common project data.

[0215] It can be understood that when the project audit result is an abnormal project result, it can be explained that there is information in the actual project that is not in the standard project data. Therefore, in order to determine the corresponding abnormal project information, the mark update is performed, and then the standard project data and the actual project data can be intersection processed to obtain the project information common to the actual project data and the standard project data, that is, the common project data. Then, based on the difference set of the actual project data and the common project data, the abnormal project data is obtained, so that the abnormal project data can be updated later.

[0216] The public project data is the project information common to the actual project data and the standard project data, which can be obtained based on the intersection of the standard project data and the actual project data, and the abnormal project data is the difference between the actual project data and the public project data.

[0217] S403: Determine, according to the abnormal project data, the corresponding project document information as the abnormal project document, and the corresponding invoice document information as the abnormal project invoice document.

[0218] It can be understood that the project document information where the abnormal project data is located is regarded as the abnormal project document, and the abnormal project invoice document is the invoice document information with the abnormal project amount. For example, when the abnormal item is drug 4, the drug document with drug 4 can be regarded as the abnormal project document, and the invoice document with the drug amount corresponding to drug 4 can be regarded as the abnormal project invoice document.

[0219] S404, when it is determined that the dosage audit result is an abnormal dosage result, the corresponding actual dosage is used as the abnormal dosage, and based on the abnormal dosage, the corresponding drug document information is determined as the abnormal item document, and the corresponding invoice document information is determined as the abnormal dosage invoice document.

[0220] It can be understood that when the dosage audit result is an abnormal dosage result, it can be explained that the actual drug dosage of the corresponding drug in the uploaded document exceeds the actual drug dosage, so that the corresponding actual dosage can be used as an abnormal dosage, and the drug document information with abnormal dosage can be used as an abnormal item document. At the same time, the corresponding invoice document information can be used as an abnormal dosage invoice document, so that when feedback is provided later, the abnormal information can be displayed and marked, which is convenient for personnel to intuitively view and confirm.

[0221] S405, obtaining an abnormal invoice document according to the abnormal amount invoice document, the abnormal item invoice document and the abnormal dosage invoice document.

[0222] It can be understood that an abnormal invoice document is an invoice document with abnormal information, including an abnormal amount invoice document, an abnormal item invoice document and an abnormal dosage invoice document.

[0223] S406: highlighting the abnormal invoice document and the abnormal project document based on the preset document mark, and highlighting the corresponding abnormal project data according to the preset project mark.

[0224] It can be understood that the preset document mark is a mark preset to highlight abnormal documents, which can be a striking color red, and the preset project mark is a mark preset to highlight abnormal project information, which can be a mark box.

[0225] It is not difficult to understand that abnormal invoice documents and abnormal project documents are highlighted through preset document marks, and abnormal project data are displayed and marked through preset project marks, so that when subsequent personnel view the claims feedback tree, they can quickly and intuitively view the corresponding abnormal information, thereby reducing confirmation time and improving confirmation feedback efficiency.

[0226] S407, highlighting the corresponding abnormal dose according to the preset dose mark, constructing a dose display slot on one side of the abnormal dose, and adding the corresponding predicted dose into the dose display slot.

[0227] It can be understood that the preset dose mark is a mark preset to highlight abnormal dose information, which may be a mark frame, and the dose display slot is a slot for displaying dose information.

[0228] It is not difficult to understand that if Figure 4 As shown, when the preset dose is marked as a mark box, the abnormal dose (the actual dose of 50 boxes corresponding to drug 1) can be marked and displayed using the preset dose mark, and a dose display slot can be constructed on one side of the abnormal dose in any blank position up, down, left, or right, so that the predicted dose can be added to the corresponding dose display slot for subsequent personnel to check and confirm, thereby ensuring the accuracy of the claims data, reducing losses, and improving claims efficiency.

[0229] S408: Based on the abnormal invoice document, determine the corresponding invoice node in the claim declaration tree as the abnormal invoice node, and according to the abnormal project document, determine the corresponding project node in the claim declaration tree as the abnormal project node.

[0230] It can be understood that the abnormal invoice node is the invoice node corresponding to the abnormal invoice document in the claim declaration tree. For example, when the invoice node corresponding to the abnormal invoice document is the invoice node No. 3.1 to 3.10 under the hospital A node in the claim declaration tree, then the invoice node is the abnormal invoice node.

[0231] Among them, the abnormal project node is the project node corresponding to the abnormal project document in the claim declaration tree.

[0232] Through the above implementation, the present invention can determine abnormal invoice nodes and abnormal project nodes to facilitate subsequent updating of the claim declaration tree, thereby generating a claim feedback tree, which is convenient for claims personnel to quickly find the marked abnormal information and check and confirm it in time.

[0233] S409, highlighting corresponding nodes in the claim declaration tree according to the abnormal invoice node and the abnormal project node to obtain a claim feedback tree.

[0234] It is understandable that the abnormal invoice nodes and abnormal project nodes in the claim declaration tree are highlighted to obtain a marked claim feedback tree, so that the claims personnel can check and confirm it in time.

[0235] In some embodiments, the specific implementation of step S4 (responding to the confirmation information of the claim settlement terminal to the claim settlement feedback tree, generating a claim settlement list and sending it to the claim settlement terminal) includes:

[0236] S410, in response to the confirmation information of the claim feedback tree by the claim settlement end, the invoice document information is deleted and updated based on the invoice document of the abnormal amount, and the data row where the abnormal item data is located in the corresponding invoice document of the abnormal item is deleted and updated.

[0237] It is understandable that after receiving the confirmation information of the claim feedback tree from the claim settlement end, the invoice document with abnormal amount in the invoice document information can be deleted and updated, and the data row information of the abnormal item data in the abnormal item invoice document can be deleted and updated.

[0238] Among them, data rows are rows and columns where data information is located.

[0239] S411, replacing and updating the abnormal dose in the abnormal dose invoice document according to the corresponding predicted dose to obtain updated invoice document information, summarizing the updated invoice document information, obtaining a claim list and sending it to the claim settlement end.

[0240] It is understandable that the confirmed predicted dose replaces the abnormal dose to obtain updated invoice document information, and then the information is summarized to obtain a claim list and send it to the claim end for accurate claim settlement.

[0241] The updated invoice document information is the invoice document information updated by replacing the abnormal dose in the abnormal dose invoice document.

[0242] See also Figure 5 , is a structural diagram of an intelligent middle-end processing system applicable to a multi-source claims reviewer provided by an embodiment of the present invention, and the intelligent middle-end processing system applicable to a multi-source claims reviewer includes:

[0243] The receiving module is used to receive the claim data uploaded by the claim terminal, construct a claim declaration tree based on the claim data, perform layered comparison on the document amounts at each time node in the claim declaration tree, and obtain the amount audit result.

[0244] The comparison module is used to retrieve the standard project data in the review party knowledge base based on the claim type of the claim end, and obtain the project review result by comparing the standard project data with the actual project data of the corresponding invoice node in the claim declaration tree.

[0245] A determination module is used to determine the actual drugs in the claim declaration tree and the actual dosage of the actual drugs according to the project review results, determine the predicted dosage based on the similarity quantitative strategy and the actual drugs, compare the actual dosage with the predicted dosage, and obtain the dosage review result.

[0246] A sending module is used to process the sub-document information at each document node in the claim declaration tree according to the amount audit result, the project audit result and the dosage audit result, generate a claim feedback tree, respond to the confirmation information of the claim feedback tree by the claim end, generate a claim list and send it to the claim end.

[0247] See also Figure 6 , is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present invention, the electronic device 60 includes: a processor 61, a memory 62 and a computer program; wherein

[0248] The memory 62 is used to store the computer program, which may also be a flash memory. The computer program is, for example, an application program, a functional module, etc. for implementing the above method.

[0249] The processor 61 is used to execute the computer program stored in the memory to implement each step performed by the device in the above method. For details, please refer to the relevant description in the above method embodiment.

[0250] Optionally, the memory 62 may be independent or integrated with the processor 61 .

[0251] When the memory 62 is a device independent of the processor 61, the device may further include:

[0252] The bus 63 is used to connect the memory 62 and the processor 61 .

[0253] The present invention also provides a readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the methods provided by the various embodiments described above.

[0254] Among them, the readable storage medium can be a computer storage medium or a communication medium. The communication medium includes any medium that facilitates the transmission of a computer program from one place to another. The computer storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer. For example, a readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application-specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). In addition, the ASIC can be located in a user device. Of course, the processor and the readable storage medium can also exist in a communication device as discrete components. The readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0255] The present invention also provides a program product, which includes an execution instruction, which is stored in a readable storage medium. At least one processor of a device can read the execution instruction from the readable storage medium, and at least one processor executes the execution instruction so that the device implements the methods provided in the above various embodiments.

[0256] In the embodiments of the above-mentioned devices, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. The general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc. The steps of the method disclosed in the present invention may be directly embodied as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0257] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent middle-end processing method suitable for multi-source claims review, characterized in that: include: Receive the claim data uploaded by the claim terminal, build a claim declaration tree according to the claim data, perform hierarchical comparison on the invoice amounts at each time node in the claim declaration tree, and obtain the amount review result; Based on the claim type of the claim settlement terminal, standard project data in the review party knowledge base is retrieved, and the project review result is obtained by comparing the standard project data with the actual project data of the corresponding invoice node in the claim declaration tree; Determine the actual drugs in the claim declaration tree and the actual dosage of the actual drugs according to the project review results, determine the predicted dosage based on the similarity quantitative strategy and the actual drugs, compare the actual dosage with the predicted dosage, and obtain the dosage review result; The sub-document information at each document node in the claim declaration tree is processed according to the amount audit result, the project audit result and the dosage audit result to generate a claim feedback tree, and in response to the confirmation information of the claim feedback tree by the claim end, a claim list is generated and sent to the claim end.

2. The method according to claim 1, characterized in that The receiving claim data uploaded by the claim settlement terminal and constructing a claim declaration tree according to the claim data include: Receive the claim data uploaded by the claim settlement terminal, wherein the claim data includes the claim type, personnel information, medical institution and the total document information corresponding to the medical institution; Constructing a claim node based on the personnel information, constructing an institution node based on the medical institution, and connecting the institution node with the claim node; Obtaining a span time period of the total document information corresponding to the medical institution, and evenly dividing the span time period according to a preset number of nodes to obtain multiple sub-time periods; Constructing a time node according to the sub-time period, connecting the time node with the organization node, dividing the total document information based on each sub-time period, and obtaining the sub-document information corresponding to the sub-time period; Classify the sub-document information based on the document category to obtain invoice document information corresponding to the invoice category, drug document information corresponding to the drug category, and inspection document information corresponding to the inspection category; Constructing an invoice node based on the invoice category and binding the invoice document information to the corresponding invoice node; constructing a drug node based on the drug category and binding the drug document information to the corresponding drug node; An inspection node is constructed based on the inspection category, and the inspection document information is bound to the corresponding inspection node. The invoice node, the drug node and the inspection node are connected to the corresponding time node to generate a claim declaration tree.

3. The method according to claim 2, characterized in that The document amounts at each time node in the claim declaration tree are compared in layers to obtain amount audit results, including: Counting the invoice amounts of the invoice document information corresponding to the invoice node connected to the corresponding institution node in the claim declaration tree to obtain the total invoice amount; Counting the drug amounts corresponding to the drug document information of the drug node connected to the corresponding institution node in the claim declaration tree to obtain the total drug amount, and counting the inspection amounts corresponding to the inspection document information of the inspection node connected to the corresponding institution node in the claim declaration tree to obtain the total inspection amount; According to the total drug amount and the total inspection amount, the total item amount is obtained, and according to the comparison information of the total invoice amount and the total item amount, a summary comparison result is obtained; Determine the drug node and the inspection node connected to the corresponding time node in the claim declaration tree as the project node, and use the drug document information and the inspection document information corresponding to the project node as the project document information, and use the amount corresponding to the project document information as the project amount; Obtain the project amount of the project document information corresponding to the project node connected to the time node in the claim declaration tree, and retrieve the corresponding invoice document information based on the project information and project time of the corresponding project document information to obtain the invoice amount of the corresponding invoice document information; Obtaining a separate comparison result based on the comparison information of the invoice amount and the corresponding item amount; Based on the summary comparison result and the individual comparison result, an amount audit result is obtained, and the amount audit result includes a normal amount result and an abnormal amount result.

4. The method according to claim 3, characterized in that The method of retrieving standard project data in the reviewer knowledge base based on the claim type of the claim settlement terminal and obtaining project review results by comparing the standard project data with the actual project data of the corresponding invoice node in the claim declaration tree includes: Based on the claim type of the claim settlement terminal, drug information and inspection information in the prescription review knowledge base are retrieved as standard item data; Obtain drug information and inspection information in the invoice document information corresponding to the invoice node connected to each of the time nodes in the claim declaration tree as actual project data; According to the comparison information between the standard project data and the actual project data, a project audit result is obtained, and the project audit result includes a normal project result and an abnormal project result.

5. The method according to claim 4, characterized in that Determining the actual drugs in the claim declaration tree and the actual dosage of the actual drugs according to the project review results includes: When it is determined that the actual project data are all within the standard project data, the project review result is determined to be a normal project result, and based on the normal project result, the drug information in the actual project data is used as the actual drug, and the actual dosage of the actual drug is obtained; When it is determined that the actual project data contains information that is not in the standard project data, the project audit result is determined to be an abnormal project result, and the information in the actual project data that is not in the standard project data is used as abnormal information; Based on the abnormal project result, the abnormal information in the actual project data is deleted and updated to obtain updated actual project data, the drug information in the updated actual project data is used as the actual drug, and the actual dosage of the actual drug is obtained.

6. The method according to claim 5, characterized in that The step of determining a predicted dose based on the similarity quantitative strategy and the actual drug, comparing the actual dose with the predicted dose, and obtaining a dose review result includes: Based on the actual drug, similar claims adjusters in the prescription review knowledge base are retrieved, and calculations are performed according to the personnel attributes, preference attributes, and pathological attributes of the similar claims adjusters to obtain similar similarities corresponding to the similar claims adjusters; Sorting the same type of claims adjusters in descending order according to the same type of similarities to obtain a similarity sequence, and selecting the same type of claims adjusters in the similarity sequence according to a preset number of personnel to obtain associated claims adjusters; The maximum dose of each actual drug corresponding to the associated claim adjuster is obtained as the initial dose, and the similarity of the same type corresponding to the associated claim adjuster is used as the associated similarity, and the associated similarity is averaged to obtain the average similarity; When it is determined that the average similarity is greater than or equal to the benchmark similarity, the initial dose of each of the actual drugs is used as the predicted dose; When it is determined that the average similarity is less than the reference similarity, a calculation is performed according to the reference similarity and the average similarity to obtain a magnification factor, and the initial dose is magnified based on the magnification factor to obtain a predicted dose; The actual dose and the predicted dose are compared to obtain a dose audit result, wherein the dose audit result includes a normal dose result and an abnormal dose result.

7. The method according to claim 6, characterized in that The calculating according to the personnel attributes, preference attributes and pathological attributes of the same type of claim adjusters to obtain the same type of similarity corresponding to the same type of claim adjusters includes: The similarity of the same category is obtained by the following formula: ， in, is the personnel similarity, is the upper limit of the number of personnel attribute types, For the claims side The value of a person's attribute. For the same type of claims adjusters, The value of a person's attribute. For the The weight value of personnel attributes, is the preference similarity, is the upper limit of the number of preferred attribute types, For the claims side The value of the preference attribute, For the same type of claims adjusters, The value of the preference attribute, For the The weight value of preference attribute, is the pathological similarity, is the upper limit of the number of pathological attribute types, For the claims side The value of the pathological attribute, For the same type of claims adjusters, The value of the pathological attribute, For the The pathological attribute weights are For similarity of the same kind.

8. The method according to claim 7, characterized in that When it is determined that the average similarity is less than the reference similarity, a calculation is performed according to the reference similarity and the average similarity to obtain a magnification factor, and the initial dose is magnified based on the magnification factor to obtain a predicted dose, including: When it is determined that the average similarity is less than the reference similarity, a similarity difference is obtained according to a difference between the reference similarity and the average similarity; Obtaining an offset multiple according to the product of the similarity difference and the offset coefficient, and obtaining an amplification multiple based on the sum of the offset multiple and the reference multiple; Amplifying the initial dose based on the amplification factor to obtain a predicted dose; The predicted dose was obtained by the following formula, ， in, To predict the dose, For the initial dose, is the baseline similarity, is the average similarity, is the offset coefficient, The base multiple.

9. The method according to claim 8, characterized in that The sub-document information at each document node in the claim declaration tree is processed according to the amount audit result, the project audit result and the dosage audit result to generate a claim feedback tree, including: When it is determined that the amount audit result is an abnormal amount result, the invoice document information corresponding to the invoice amount is used as the abnormal amount invoice document, and the project document information corresponding to the project amount is used as the abnormal project document; When it is determined that the project audit result is an abnormal project result, public project data is obtained based on the intersection of the standard project data and the actual project data, and abnormal project data is obtained based on the difference between the actual project data and the public project data; Determine, according to the abnormal project data, the corresponding project document information as the abnormal project document, and the corresponding invoice document information as the abnormal project invoice document; When it is determined that the dosage audit result is an abnormal dosage result, the corresponding actual dosage is used as the abnormal dosage, and based on the abnormal dosage, the corresponding drug document information is determined as the abnormal item document, and the corresponding invoice document information is determined as the abnormal dosage invoice document; Obtaining an abnormal invoice document according to the abnormal amount invoice document, the abnormal item invoice document and the abnormal dosage invoice document; The abnormal invoice document and the abnormal item document are highlighted based on the preset document mark, and the corresponding abnormal item data are highlighted based on the preset item mark; Highlighting the corresponding abnormal dose according to a preset dose mark, constructing a dose display slot on one side of the abnormal dose, and adding the corresponding predicted dose into the dose display slot; Based on the abnormal invoice document, determine the corresponding invoice node in the claim declaration tree as the abnormal invoice node, and based on the abnormal project document, determine the corresponding project node in the claim declaration tree as the abnormal project node; According to the abnormal invoice node and the abnormal project node, corresponding nodes in the claim declaration tree are highlighted to obtain a claim feedback tree; The step of responding to the confirmation information of the claim settlement terminal on the claim settlement feedback tree and generating a claim settlement list and sending it to the claim settlement terminal includes: In response to the confirmation information of the claim feedback tree by the claim settlement end, the invoice document information is deleted and updated based on the invoice document of the abnormal amount, and the data row where the abnormal item data is located in the corresponding invoice document of the abnormal item is deleted and updated; The abnormal dose in the abnormal dose invoice document is replaced and updated according to the corresponding predicted dose to obtain updated invoice document information, the updated invoice document information is summarized to obtain a claim list and send it to the claim settlement end.

10. An intelligent middle-end processing system suitable for multi-source claims review, characterized in that: include: A receiving module is used to receive the claim data uploaded by the claim terminal, construct a claim declaration tree according to the claim data, perform hierarchical comparison on the invoice amounts at each time node in the claim declaration tree, and obtain the amount audit result; A comparison module, for retrieving standard project data in the reviewer knowledge base based on the claim type of the claim settlement end, and obtaining a project review result by comparing the standard project data with the actual project data of the corresponding invoice node in the claim declaration tree; A determination module, used to determine the actual drugs in the claim declaration tree and the actual dosage of the actual drugs according to the project review result, determine the predicted dosage based on the similarity quantitative strategy and the actual drugs, compare the actual dosage with the predicted dosage, and obtain the dosage review result; A sending module is used to process the sub-document information at each document node in the claim declaration tree according to the amount audit result, the project audit result and the dosage audit result, generate a claim feedback tree, respond to the confirmation information of the claim feedback tree by the claim end, generate a claim list and send it to the claim end.

Images