Medical insurance anti-fraud intelligent management system and management method thereof

Abstract

The present invention discloses an intelligent medical insurance anti-fraud management system and a management method thereof, which relate to the technical field of anti-fraud supervision management, and include the following modules: a fraud case analysis module, which obtains fraud cases to establish a case database, obtains information values ​​through each fraud history case, obtains method values ​​through each fraud history case, and obtains the fraud coefficient of each fraud history case through the information value and method value of each fraud case; obtains a screening reference range through the fraud coefficient and the historical coefficient of the medical insurance reimbursement personnel, and then obtains a similarity overlap based on the screening reference range and the fraud coefficient reference range of the medical insurance reimbursement personnel, thereby facilitating a similarity comparison between the actual situation of the medical insurance reimbursement personnel and the fraud cases in the case database, thereby being beneficial to improving the early warning effect of medical insurance fraud and also beneficial to preventing gangs from committing crimes by using similar methods to defraud medical insurance expenses.

Description

Technical Field

[0001] The present invention relates to the technical field of anti-fraud supervision management, and specifically to a medical insurance anti-fraud intelligent management system and a management method thereof. Background Art

[0002] Medical insurance fraud includes false claims, fabricated illnesses, excessive charges, and duplicate fees. These practices lead to waste and misuse of medical insurance funds, seriously undermining the fairness and sustainability of the medical insurance system.

[0003] Anti-fraud in medical insurance big data refers to the process of using large amounts of data in the medical insurance system to identify medical insurance fraud through data analysis and mining technology, and taking corresponding preventive and combative measures. Medical insurance big data anti-fraud mainly uses data mining and machine learning technology to analyze and model the data in the medical insurance system to identify doctors, medical institutions and patients who may have fraudulent behavior.

[0004] The implementation of medical insurance big data anti-fraud can effectively improve the supervision ability and anti-fraud level of medical insurance management departments, reduce the occurrence of medical insurance fraud, and provide strong guarantees for the sustainable development of the medical insurance system. However, the current big data anti-fraud supervision and management relies solely on the education and prevention of blacklisted personnel. However, this method has disadvantages. Medical insurance fraud personnel can only be marked after they commit fraud and illegal activities. There is a lag in investigation and punishment, which can easily cause losses. It is difficult to fundamentally avoid subsequent medical insurance fraud by blacklisting insured persons related to medical insurance fraud. Relevant personnel who are not marked will still use the same methods to commit medical insurance fraud. For this reason, we propose a medical insurance anti-fraud intelligent management system and its management method. Summary of the Invention

[0005] (1) Technical problems solved

[0006] In view of the deficiencies in the prior art, the present invention provides a medical insurance anti-fraud intelligent management system and a management method thereof to solve the above-mentioned problems in the prior art.

[0007] (2) Technical solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: a medical insurance anti-fraud intelligent management system, including the following modules:

[0009] The fraud case analysis module obtains fraud cases and establishes a case database. It obtains information values ​​and method values ​​from each historical fraud case, obtains the fraud coefficient of each historical fraud case based on the information value and method value of each fraud case, incorporates the fraud coefficient of each fraud case into the case database, and obtains a reference range for the fraud coefficient based on the dispersion of all fraud coefficients.

[0010] The fraud similarity analysis module obtains the fraud coefficient of the medical insurance reimbursement personnel, obtains the historical coefficient of the medical insurance reimbursement personnel, obtains the screening reference range based on the fraud coefficient and historical coefficient of the medical insurance reimbursement personnel, and obtains the similarity overlap between the screening reference range and the fraud coefficient reference range of the medical insurance reimbursement personnel;

[0011] The suspicious warning module determines whether to activate an alarm to remind management personnel based on the similarity and overlap of medical insurance reimbursement personnel, and determines whether to include the fraud coefficient of the medical insurance reimbursement personnel in the case database based on the review results of the medical insurance reimbursement personnel.

[0012] Preferably, fraud cases are obtained in the fraud case analysis module to establish a case database, specifically:

[0013] Step 1: Obtain medical insurance fraud cases through the official website of the Medical Insurance Bureau, through the court judgment document database, and through news websites and academic journals;

[0014] Step 2: Summarize all the obtained medical insurance fraud cases, delete duplicate medical insurance fraud cases, establish a case database and include all the deleted medical insurance fraud cases in the case database.

[0015] Preferably, in the fraud case analysis module, information value is obtained from each fraud history case, specifically:

[0016] Step 1: Obtain the province where the insured person in the fraud history case is located, mark all provinces in the country with different numbers, obtain the number corresponding to the province where the insured person in the fraud history case is located, mark it as the province number, obtain the city where the insured person in the fraud history case is located, mark all cities in the country with different numbers, obtain the number corresponding to the city where the insured person in the fraud history case is located, mark it as the city number, and perform a weighted sum of the province number and the city number to obtain the location information;

[0017] Step 2: Obtain the age and gender of the insured in the fraud history case. If the insured in the fraud history case is male, the gender of the insured in the fraud history case is marked as 5. If the insured in the fraud history case is female, the gender of the insured in the fraud history case is marked as 10. The insured information is obtained by taking the weighted sum of the age and gender of the insured in the fraud history case.

[0018] Step 3: Obtain the fraud amount of the insured in the fraud history case, and obtain the information value through the fraud amount, location information and insured information;

[0019] The information value is calculated as follows:

[0020]

[0021] Among them, Xn represents the information value, Je represents the fraud amount, Wx represents the location information, and Cx represents the insured person information. 、 and are weights, , , .

[0022] Preferably, in the fraud case analysis module, the mode value is obtained through each fraud history case, specifically:

[0023] Step 1: Determine whether there is any case of bed-hanging hospitalization in the fraud history case. If so, the bed-hanging hospitalization mark is 1; if not, the bed-hanging hospitalization mark is 0;

[0024] Step 2: Obtain medical bills from fraud history cases and determine whether the medical bills in the fraud history cases are suspicious bills based on paper quality, printing quality, and format standardization. If the medical bill is suspicious, the medical bill is marked as 1; if the medical bill is not suspicious, the medical bill is marked as 0;

[0025] Step 3: Obtain the ID card information and medical insurance card information used by the insured in the fraud history case, and determine whether the ID card information and the medical insurance card information used by the insured in the fraud history case are consistent. If the ID card information and the medical insurance card information are inconsistent, the medical insurance card mark is 1; if the ID card information and the medical insurance card information are consistent, the medical insurance card mark is 0;

[0026] Step 4: Obtain the method value through the bed-hanging hospitalization, medical bills and medical insurance cards of the fraud history case;

[0027] The calculation method of the mode value is as follows:

[0028]

[0029] Among them, Fs represents the mode value, Gc represents the bed-sharing hospitalization, Yp represents the medical bill, and Yb represents the medical insurance card. 、 and are weights, , , .

[0030] Preferably, the fraud coefficient is calculated in the fraud case analysis module as follows:

[0031]

[0032] Among them, Qz represents the fraud coefficient, Xn represents the information value, and Fs represents the mode value. and are weights, , .

[0033] Preferably, in the fraud case analysis module, a reference range of fraud coefficients is obtained by the dispersion of all fraud coefficients, specifically:

[0034] Step 1: Obtain all fraud coefficients, sum and average all fraud coefficients to obtain the average fraud coefficient, subtract each fraud coefficient from the average fraud coefficient to obtain multiple fraud coefficient differences, and square each fraud coefficient difference to obtain multiple fraud coefficient variances;

[0035] Step 2: Sum and average all the fraud coefficient variances to obtain the mean fraud coefficient variance, then square the mean fraud coefficient variance to obtain the standard deviation of the fraud coefficient, set the range preset threshold, sum the fraud coefficient standard deviation and the range preset threshold to obtain the maximum range value, subtract the fraud coefficient standard deviation from the range preset threshold to obtain the minimum range value, and define the range between the minimum and maximum values ​​as the fraud coefficient reference range.

[0036] Preferably, the fraud similarity analysis module obtains the historical coefficient of the medical insurance reimbursement personnel, and obtains the screening reference range based on the fraud coefficient and the historical coefficient of the medical insurance reimbursement personnel, specifically:

[0037] Step 1: Obtain the last medical insurance reimbursement date of the medical insurance reimbursement personnel, obtain the current date, subtract the current date from the last medical insurance reimbursement date of the medical insurance reimbursement personnel to obtain the interval time, set the time correlation preset threshold, and obtain the historical coefficient by multiplying the interval time by the time correlation preset threshold;

[0038] Step 2: Obtain the fraud coefficient of the medical insurance reimbursement personnel, obtain the maximum fraud coefficient by summing the fraud coefficient of the medical insurance reimbursement personnel and the historical coefficient, obtain the minimum fraud coefficient by subtracting the fraud coefficient of the medical insurance reimbursement personnel from the historical coefficient, and define the range between the minimum fraud coefficient and the maximum fraud coefficient as the screening reference range.

[0039] Preferably, in the fraud similarity analysis module, the similarity overlap is obtained by using the screening reference range of the medical insurance reimbursement personnel and the fraud coefficient reference range, specifically:

[0040] Step 1: Obtain the screening reference range of the medical insurance reimbursement personnel, obtain the fraud coefficient reference range, and determine whether the fraud coefficient reference range and the screening reference range overlap. If the fraud coefficient reference range and the screening reference range do not overlap, mark the similarity overlap as 0. If the fraud coefficient reference range and the screening reference range overlap, proceed to step 2.

[0041] Step 2: Get the minimum and maximum values ​​of the overlapping part, and obtain the overlap value by subtracting the maximum and minimum values ​​of the overlapping part. Get the maximum and minimum values ​​of the screening reference range, and obtain the reference range value by subtracting the maximum and minimum values ​​of the screening reference range. Obtain the similarity overlap by taking the quotient of the overlap value and the reference range value.

[0042] Preferably, the suspicious warning module is specifically:

[0043] Step 1: Obtain the similarity overlap of the medical insurance reimbursement personnel, set a preset threshold for the similarity overlap, and determine whether the similarity overlap of the medical insurance reimbursement personnel is greater than the preset threshold. If the similarity overlap of the medical insurance reimbursement personnel is greater than or equal to the preset threshold, an alarm is activated to remind the management personnel to re-review the information of the medical insurance reimbursement personnel;

[0044] Step 2: Obtain the results of repeated review of the information of the medical insurance reimbursement personnel by the management personnel. If the medical insurance reimbursement personnel have committed medical insurance fraud, obtain the fraud coefficient of the personnel and include the fraud coefficient in the case database.

[0045] A medical insurance anti-fraud intelligent management method includes the following steps:

[0046] S1: Obtain fraud cases and establish a case database. Obtain information values ​​and method values ​​from each historical fraud case. Obtain a fraud coefficient for each historical fraud case based on the information value and method value. Incorporate the fraud coefficient of each fraud case into the case database. Obtain a reference range for the fraud coefficient based on the dispersion of all fraud coefficients.

[0047] S2: Obtain the fraud coefficient of the medical insurance reimbursement personnel, obtain the historical coefficient of the medical insurance reimbursement personnel, obtain the screening reference range based on the fraud coefficient and the historical coefficient of the medical insurance reimbursement personnel, and obtain the similarity overlap degree based on the screening reference range and the fraud coefficient reference range of the medical insurance reimbursement personnel;

[0048] S3: Determine whether to activate an alarm to remind management personnel based on the similarity and overlap of the medical insurance reimbursement personnel, and determine whether to include the fraud coefficient of the medical insurance reimbursement personnel in the case database based on the review results of the medical insurance reimbursement personnel.

[0049] (3) Beneficial effects

[0050] The present invention provides a medical insurance anti-fraud intelligent management system and management method thereof, which has the following beneficial effects:

[0051] (1) In the fraud similarity analysis module, this scheme obtains a screening reference range through the fraud coefficient and historical coefficient of the medical insurance reimbursement personnel, and then obtains the similarity overlap based on the screening reference range and the fraud coefficient reference range of the medical insurance reimbursement personnel, so as to facilitate the similarity comparison between the actual situation of the medical insurance reimbursement personnel and the fraud cases in the case database, which is beneficial to improve the early warning effect of medical insurance fraud.

[0052] (2) This solution controls the alarm activation in the suspicious warning module through the similarity overlap of the medical insurance reimbursement personnel, so as to remind the management personnel to conduct strict review of the medical insurance reimbursement personnel, which is beneficial to avoid economic losses caused by medical insurance fraud. The review results of the medical insurance reimbursement personnel are used to determine whether the fraud coefficient of the medical insurance reimbursement personnel should be included in the case database, so as to facilitate the supplementation and update of the case database according to the real-time fraud cases, which is beneficial to improve the accuracy of the reference range of the fraud coefficient, and thus help improve the accuracy of the warning of medical insurance fraud personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a structural diagram of a medical insurance anti-fraud intelligent management system of the present invention;

[0054] Figure 2 This is a flowchart of an intelligent management method for medical insurance anti-fraud according to the present invention. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] See also Figure 1-Figure 2 The present invention provides a medical insurance anti-fraud intelligent management system, including the following modules:

[0057] The fraud case analysis module obtains fraud cases and establishes a case database. It obtains information values ​​and method values ​​from each historical fraud case, obtains the fraud coefficient of each historical fraud case based on the information value and method value of each fraud case, incorporates the fraud coefficient of each fraud case into the case database, and obtains a reference range for the fraud coefficient based on the dispersion of all fraud coefficients.

[0058] The fraud similarity analysis module obtains the fraud coefficient of the medical insurance reimbursement personnel, obtains the historical coefficient of the medical insurance reimbursement personnel, obtains the screening reference range based on the fraud coefficient and historical coefficient of the medical insurance reimbursement personnel, and obtains the similarity overlap between the screening reference range and the fraud coefficient reference range of the medical insurance reimbursement personnel;

[0059] The suspicious warning module determines whether to activate an alarm to remind management personnel based on the similarity and overlap of medical insurance reimbursement personnel, and determines whether to include the fraud coefficient of the medical insurance reimbursement personnel in the case database based on the review results of the medical insurance reimbursement personnel;

[0060] The fraud coefficient is calculated in the fraud case analysis module as follows:

[0061]

[0062] Among them, Qz represents the fraud coefficient, Xn represents the information value, and Fs represents the mode value. and are weights, , .

[0063] In this embodiment, the scheme obtains the fraud coefficient in the fraud case analysis module by using the information value and method value of the fraud case, thereby facilitating the digitization of the perpetrator information and method based on the actual situation of the fraud case, thereby facilitating the subsequent judgment of the probability of fraudulent behavior of the medical insurance reimbursement personnel based on the fraud case, thereby improving the early warning effect of medical insurance fraud;

[0064] In the fraud similarity analysis module, this solution uses the fraud coefficient and historical coefficient of the medical insurance reimbursement personnel to obtain a screening reference range. Then, based on the screening reference range of the medical insurance reimbursement personnel and the fraud coefficient reference range, it obtains the similarity overlap. This facilitates the similarity comparison between the actual situation of the medical insurance reimbursement personnel and the fraud cases in the case database, which is beneficial to improve the early warning effect of medical insurance fraud.

[0065] This solution uses the similarity and overlap of the medical insurance reimbursement personnel in the suspicious early warning module to control the alarm activation, thereby reminding management personnel to conduct strict audits on the medical insurance reimbursement personnel, which is beneficial to avoiding economic losses caused by medical insurance fraud. The audit results of the medical insurance reimbursement personnel are used to determine whether the fraud coefficient of the medical insurance reimbursement personnel should be included in the case database, so that the case database can be supplemented and updated according to the real-time fraud cases, which is beneficial to improve the accuracy of the fraud coefficient reference range and the accuracy of early warnings for medical insurance fraud personnel.

[0066] It is worth mentioning that the weight values ​​in this scheme can be obtained through the hierarchical analysis method, and the value of the preset threshold can be obtained through the weight analysis method, which will not be elaborated here.

[0067] In the fraud case analysis module, fraud cases are obtained to establish a case database, specifically:

[0068] Step 1: Obtain medical insurance fraud cases through the official website of the Medical Insurance Bureau, through the court judgment document database, and through news websites and academic journals;

[0069] Step 2: Summarize all the obtained medical insurance fraud cases, delete duplicate medical insurance fraud cases, establish a case database and include all the deleted medical insurance fraud cases in the case database.

[0070] In this embodiment, medical insurance fraud cases are obtained through the official website of the Medical Insurance Bureau, the court judgment document database, news websites, and academic journals, which is beneficial to improving the authenticity and diversity of the obtained medical insurance fraud cases, and is beneficial to predicting the similarity of medical insurance reimbursement personnel based on the medical insurance fraud cases.

[0071] In the fraud case analysis module, information values ​​are obtained from each fraud history case, specifically:

[0072] Step 1: Obtain the province where the insured person in the fraud history case is located, mark all provinces in the country with different numbers, obtain the number corresponding to the province where the insured person in the fraud history case is located, mark it as the province number, obtain the city where the insured person in the fraud history case is located, mark all cities in the country with different numbers, obtain the number corresponding to the city where the insured person in the fraud history case is located, mark it as the city number, and perform a weighted sum of the province number and the city number to obtain the location information;

[0073] Step 2: Obtain the age and gender of the insured in the fraud history case. If the insured in the fraud history case is male, the gender of the insured in the fraud history case is marked as 5. If the insured in the fraud history case is female, the gender of the insured in the fraud history case is marked as 10. The insured information is obtained by taking the weighted sum of the age and gender of the insured in the fraud history case.

[0074] Step 3: Obtain the fraud amount of the insured in the fraud history case, and obtain the information value through the fraud amount, location information and insured information;

[0075] The information value is calculated as follows:

[0076]

[0077] Among them, Xn represents the information value, Je represents the fraud amount, Wx represents the location information, and Cx represents the insured person information. 、 and are weights, , , .

[0078] In this embodiment, the information value is obtained through the fraud amount, location information and insured person information, which is beneficial to the data analysis and classification of medical insurance fraud personnel based on their information, and the fraud coefficient of the medical insurance reimbursement personnel is obtained in the fraud similarity analysis module. Since there is no fraud amount of the medical insurance reimbursement personnel, the fraud amount of the medical insurance reimbursement personnel is 0 when the fraud coefficient is calculated by the information value and the method value.

[0079] In the fraud case analysis module, the method value is obtained from each fraud history case, specifically:

[0080] Step 1: Determine whether there is any case of bed-hanging hospitalization in the fraud history case. If so, the bed-hanging hospitalization mark is 1; if not, the bed-hanging hospitalization mark is 0;

[0081] Step 2: Obtain medical bills from fraud history cases and determine whether the medical bills in the fraud history cases are suspicious bills based on paper quality, printing quality, and format standardization. If the medical bill is suspicious, the medical bill is marked as 1; if the medical bill is not suspicious, the medical bill is marked as 0;

[0082] Step 3: Obtain the ID card information and medical insurance card information used by the insured in the fraud history case, and determine whether the ID card information and the medical insurance card information used by the insured in the fraud history case are consistent. If the ID card information and the medical insurance card information are inconsistent, the medical insurance card mark is 1; if the ID card information and the medical insurance card information are consistent, the medical insurance card mark is 0;

[0083] Step 4: Obtain the method value through the bed-hanging hospitalization, medical bills and medical insurance cards of the fraud history case;

[0084] The calculation method of the mode value is as follows:

[0085]

[0086] Among them, Fs represents the mode value, Gc represents the bed-sharing hospitalization, Yp represents the medical bill, and Yb represents the medical insurance card. 、 and are weights, , , .

[0087] In this embodiment, the method value is obtained through the bed-hanging hospitalization, medical bills and medical insurance cards of the historical fraud cases, so as to facilitate the combination of the historical fraud methods of the medical insurance fraud personnel to obtain the method value, and then facilitate the digitization of the fraud methods of the historical cases, and then facilitate the subsequent prediction and comparison of the similarity of medical insurance reimbursement personnel. Whether the medical bills in the fraud history cases are suspicious bills is judged by the paper quality, printing quality and format standardization. Specifically, the quality of the paper can be judged by parameters such as the weight and toughness of the paper, the printing quality can be judged by parameters such as the color brightness of the printing ink, and the format standardization can be judged by whether there is a signature. Finally, the bill situation is obtained by the weighted summation of the paper quality, printing quality and format standardization. Finally, whether the bill is a suspicious bill is judged based on the bill situation. It can also be judged by the naked eye and touch based on experience.

[0088] In the fraud case analysis module, the reference range of the fraud coefficient is obtained by the dispersion of all fraud coefficients, specifically:

[0089] Step 1: Obtain all fraud coefficients, sum and average all fraud coefficients to obtain the average fraud coefficient, subtract each fraud coefficient from the average fraud coefficient to obtain multiple fraud coefficient differences, and square each fraud coefficient difference to obtain multiple fraud coefficient variances;

[0090] Step 2: Sum and average all the fraud coefficient variances to obtain the mean fraud coefficient variance, then square the mean fraud coefficient variance to obtain the standard deviation of the fraud coefficient, set the range preset threshold, sum the fraud coefficient standard deviation and the range preset threshold to obtain the maximum range value, subtract the fraud coefficient standard deviation from the range preset threshold to obtain the minimum range value, and define the range between the minimum and maximum values ​​as the fraud coefficient reference range.

[0091] The fraud similarity analysis module obtains the historical coefficients of medical insurance reimbursement personnel, and uses the fraud coefficients and historical coefficients of medical insurance reimbursement personnel to obtain a screening reference range, specifically:

[0092] Step 1: Obtain the last medical insurance reimbursement date of the medical insurance reimbursement personnel, obtain the current date, subtract the current date from the last medical insurance reimbursement date of the medical insurance reimbursement personnel to obtain the interval time, set the time correlation preset threshold, and obtain the historical coefficient by multiplying the interval time by the time correlation preset threshold;

[0093] Step 2: Obtain the fraud coefficient of the medical insurance reimbursement personnel, obtain the maximum fraud coefficient by summing the fraud coefficient of the medical insurance reimbursement personnel and the historical coefficient, obtain the minimum fraud coefficient by subtracting the fraud coefficient of the medical insurance reimbursement personnel from the historical coefficient, and define the range between the minimum fraud coefficient and the maximum fraud coefficient as the screening reference range.

[0094] In the fraud similarity analysis module, the similarity overlap is obtained by using the screening reference range of medical insurance reimbursement personnel and the fraud coefficient reference range, specifically:

[0095] Step 1: Obtain the screening reference range of the medical insurance reimbursement personnel, obtain the fraud coefficient reference range, and determine whether the fraud coefficient reference range and the screening reference range overlap. If the fraud coefficient reference range and the screening reference range do not overlap, mark the similarity overlap as 0. If the fraud coefficient reference range and the screening reference range overlap, proceed to step 2.

[0096] Step 2: Get the minimum and maximum values ​​of the overlapping part, and obtain the overlap value by subtracting the maximum and minimum values ​​of the overlapping part. Get the maximum and minimum values ​​of the screening reference range, and obtain the reference range value by subtracting the maximum and minimum values ​​of the screening reference range. Obtain the similarity overlap by taking the quotient of the overlap value and the reference range value.

[0097] In this embodiment, the reference range of the fraud coefficient obtained through historical cases is overlapped and compared with the screening reference range of the medical insurance reimbursement personnel, so as to facilitate the judgment of the similarity between the medical insurance reimbursement personnel and the medical insurance fraud personnel in the historical cases, thereby facilitating the analysis and prediction of the possibility of fraud by the medical insurance reimbursement personnel based on the means and circumstances of the fraud personnel in the historical cases, thereby helping to improve the accuracy of the early warning of medical insurance fraud personnel.

[0098] Suspicious warning module, specifically:

[0099] Step 1: Obtain the similarity overlap of the medical insurance reimbursement personnel, set a preset threshold for the similarity overlap, and determine whether the similarity overlap of the medical insurance reimbursement personnel is greater than the preset threshold. If the similarity overlap of the medical insurance reimbursement personnel is greater than or equal to the preset threshold, an alarm is activated to remind the management personnel to re-review the information of the medical insurance reimbursement personnel;

[0100] Step 2: Obtain the results of repeated review of the information of the medical insurance reimbursement personnel by the management personnel. If the medical insurance reimbursement personnel have committed medical insurance fraud, obtain the fraud coefficient of the personnel and include the fraud coefficient in the case database.

[0101] In this embodiment, the alarm is turned on by controlling the similarity overlap of the medical insurance reimbursement personnel, so as to remind the management personnel to conduct strict review of the medical insurance reimbursement personnel, which is beneficial to avoiding economic losses caused by medical insurance fraud. The review results of the medical insurance reimbursement personnel are used to determine whether the fraud coefficient of the medical insurance reimbursement personnel should be included in the case database, so as to facilitate the supplementation and updating of the case database according to the real-time fraud cases, which is beneficial to improve the accuracy of the fraud coefficient reference range, and thus is beneficial to improve the accuracy of the early warning of medical insurance fraud personnel.

[0102] See also Figure 1-Figure 2The present invention provides a medical insurance anti-fraud intelligent management method, comprising the following steps:

[0103] S1: Obtain fraud cases and establish a case database. Obtain information values ​​and method values ​​from each historical fraud case. Obtain a fraud coefficient for each historical fraud case based on the information value and method value. Incorporate the fraud coefficient of each fraud case into the case database. Obtain a reference range for the fraud coefficient based on the dispersion of all fraud coefficients.

[0104] S2: Obtain the fraud coefficient of the medical insurance reimbursement personnel, obtain the historical coefficient of the medical insurance reimbursement personnel, obtain the screening reference range based on the fraud coefficient and the historical coefficient of the medical insurance reimbursement personnel, and obtain the similarity overlap degree based on the screening reference range and the fraud coefficient reference range of the medical insurance reimbursement personnel;

[0105] S3: Determine whether to activate an alarm to remind management personnel based on the similarity and overlap of the medical insurance reimbursement personnel, and determine whether to include the fraud coefficient of the medical insurance reimbursement personnel in the case database based on the review results of the medical insurance reimbursement personnel.

[0106] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art will appreciate that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0107] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0108] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A medical insurance anti-fraud intelligent management system, characterized by: Includes the following modules: The fraud case analysis module obtains fraud cases to establish a case database, obtains information values ​​from each fraud history case, and the information values ​​include the fraud amount, location information and insured person information, so as to realize data analysis and classification of medical insurance fraud personnel according to their information, obtains method values ​​from each fraud history case, and the method values ​​include the bed-hanging hospitalization, medical bills and medical insurance cards of the fraud history case, so as to predict and compare the similarity of medical insurance reimbursement personnel according to their historical fraud methods, and obtains the fraud coefficient of each fraud history case through the information value and method value of each fraud case. The fraud coefficient is used to judge the probability of fraudulent behavior of medical insurance reimbursement personnel based on the fraud case, and the fraud coefficient of each fraud case is included in the case database. The reference range of the fraud coefficient is obtained through the dispersion of all fraud coefficients; The fraud similarity analysis module obtains the fraud coefficient of the medical insurance reimbursement personnel, obtains the historical coefficient of the medical insurance reimbursement personnel, obtains the screening reference range based on the fraud coefficient and historical coefficient of the medical insurance reimbursement personnel, and obtains the similarity overlap between the screening reference range and the fraud coefficient reference range of the medical insurance reimbursement personnel; The suspicious warning module determines whether to activate an alarm to remind management personnel based on the similarity and overlap of medical insurance reimbursement personnel, and determines whether to include the fraud coefficient of the medical insurance reimbursement personnel in the case database based on the review results of the medical insurance reimbursement personnel.

2. The medical insurance anti-fraud intelligent management system according to claim 1, characterized in that: In the fraud case analysis module, fraud cases are obtained to establish a case database, specifically: Step 1: Obtain medical insurance fraud cases through the official website of the Medical Insurance Bureau, through the court judgment document database, and through news websites and academic journals; Step 2: Summarize all the obtained medical insurance fraud cases, delete duplicate medical insurance fraud cases, establish a case database and include all the deleted medical insurance fraud cases in the case database.

3. The medical insurance anti-fraud intelligent management system according to claim 1, characterized in that: In the fraud case analysis module, information values ​​are obtained from each fraud history case, specifically: Step 1: Obtain the province where the insured person in the fraud history case is located, mark all provinces in the country with different numbers, obtain the number corresponding to the province where the insured person in the fraud history case is located, mark it as the province number, obtain the city where the insured person in the fraud history case is located, mark all cities in the country with different numbers, obtain the number corresponding to the city where the insured person in the fraud history case is located, mark it as the city number, and perform a weighted sum of the province number and the city number to obtain the location information; Step 2: Obtain the age and gender of the insured in the fraud history case. If the insured in the fraud history case is male, the gender of the insured in the fraud history case is marked as 5. If the insured in the fraud history case is female, the gender of the insured in the fraud history case is marked as 10. The insured information is obtained by taking the weighted sum of the age and gender of the insured in the fraud history case. Step 3: Obtain the fraud amount of the insured in the fraud history case, and obtain the information value through the fraud amount, location information and insured information; The information value is calculated as follows: Among them, Xn represents the information value, Je represents the fraud amount, Wx represents the location information, and Cx represents the insured person information. 、 and are weights, , , .

4. The medical insurance anti-fraud intelligent management system according to claim 1, characterized in that: In the fraud case analysis module, the method value is obtained from each fraud history case, specifically: Step 1: Determine whether there is any case of bed-hanging hospitalization in the fraud history case. If so, the bed-hanging hospitalization mark is 1; if not, the bed-hanging hospitalization mark is 0; Step 2: Obtain medical bills from fraud history cases and determine whether the medical bills in the fraud history cases are suspicious bills based on paper quality, printing quality, and format standardization. If the medical bill is suspicious, the medical bill is marked as 1; if the medical bill is not suspicious, the medical bill is marked as 0; Step 3: Obtain the ID card information and medical insurance card information used by the insured in the fraud history case, and determine whether the ID card information and the medical insurance card information used by the insured in the fraud history case are consistent. If the ID card information and the medical insurance card information are inconsistent, the medical insurance card mark is 1; if the ID card information and the medical insurance card information are consistent, the medical insurance card mark is 0; Step 4: Obtain the method value through the bed-hanging hospitalization, medical bills and medical insurance cards of the fraud history case; The calculation method of the mode value is as follows: Among them, Fs represents the mode value, Gc represents the bed-sharing hospitalization, Yp represents the medical bill, and Yb represents the medical insurance card. 、 and are weights, , , .

5. The medical insurance anti-fraud intelligent management system according to claim 1, characterized in that: The fraud coefficient is calculated in the fraud case analysis module as follows: Among them, Qz represents the fraud coefficient, Xn represents the information value, and Fs represents the mode value. and are weights, , .

6. The medical insurance anti-fraud intelligent management system according to claim 1, characterized in that: In the fraud case analysis module, the reference range of the fraud coefficient is obtained by the dispersion of all fraud coefficients, specifically: Step 1: Obtain all fraud coefficients, sum and average all fraud coefficients to obtain the average fraud coefficient, subtract each fraud coefficient from the average fraud coefficient to obtain multiple fraud coefficient differences, and square each fraud coefficient difference to obtain multiple fraud coefficient variances; Step 2: Sum and average all the fraud coefficient variances to obtain the mean fraud coefficient variance, then square the mean fraud coefficient variance to obtain the standard deviation of the fraud coefficient, set the range preset threshold, sum the fraud coefficient standard deviation and the range preset threshold to obtain the maximum range value, subtract the fraud coefficient standard deviation from the range preset threshold to obtain the minimum range value, and define the range between the minimum and maximum values ​​as the fraud coefficient reference range.

7. The medical insurance anti-fraud intelligent management system according to claim 1, characterized in that: The fraud similarity analysis module obtains the historical coefficients of medical insurance reimbursement personnel, and uses the fraud coefficients and historical coefficients of medical insurance reimbursement personnel to obtain a screening reference range, specifically: Step 1: Obtain the last medical insurance reimbursement date of the medical insurance reimbursement personnel, obtain the current date, subtract the current date from the last medical insurance reimbursement date of the medical insurance reimbursement personnel to obtain the interval time, set the time correlation preset threshold, and obtain the historical coefficient by multiplying the interval time by the time correlation preset threshold; Step 2: Obtain the fraud coefficient of the medical insurance reimbursement personnel, obtain the maximum fraud coefficient by summing the fraud coefficient of the medical insurance reimbursement personnel and the historical coefficient, obtain the minimum fraud coefficient by subtracting the fraud coefficient of the medical insurance reimbursement personnel from the historical coefficient, and define the range between the minimum fraud coefficient and the maximum fraud coefficient as the screening reference range.

8. The medical insurance anti-fraud intelligent management system according to claim 1, characterized in that: In the fraud similarity analysis module, the similarity overlap is obtained by using the screening reference range of medical insurance reimbursement personnel and the fraud coefficient reference range, specifically: Step 1: Obtain the screening reference range of the medical insurance reimbursement personnel, obtain the fraud coefficient reference range, and determine whether the fraud coefficient reference range and the screening reference range overlap. If the fraud coefficient reference range and the screening reference range do not overlap, mark the similarity overlap as 0. If the fraud coefficient reference range and the screening reference range overlap, proceed to step 2. Step 2: Get the minimum and maximum values ​​of the overlapping part, and obtain the overlap value by subtracting the maximum and minimum values ​​of the overlapping part. Get the maximum and minimum values ​​of the screening reference range, and obtain the reference range value by subtracting the maximum and minimum values ​​of the screening reference range. Obtain the similarity overlap by taking the quotient of the overlap value and the reference range value.

9. The medical insurance anti-fraud intelligent management system according to claim 1, characterized in that: Suspicious warning module, specifically: Step 1: Obtain the similarity overlap of the medical insurance reimbursement personnel, set a preset threshold for the similarity overlap, and determine whether the similarity overlap of the medical insurance reimbursement personnel is greater than the preset threshold. If the similarity overlap of the medical insurance reimbursement personnel is greater than or equal to the preset threshold, an alarm is activated to remind the management personnel to re-review the information of the medical insurance reimbursement personnel; Step 2: Obtain the results of repeated review of the information of the medical insurance reimbursement personnel by the management personnel. If the medical insurance reimbursement personnel have committed medical insurance fraud, obtain the fraud coefficient of the personnel and include the fraud coefficient in the case database.

10. A medical insurance anti-fraud intelligent management method, applied to a medical insurance anti-fraud intelligent management system according to any one of claims 1 to 9, characterized in that: The steps include: S1: Obtain fraud cases to establish a case database, obtain information values ​​from each fraud history case, the information values ​​include the fraud amount, location information and insured person information, so as to realize data analysis and classification of medical insurance fraud personnel based on their information, obtain method values ​​from each fraud history case, the method values ​​include the bed-hanging hospitalization, medical bills and medical insurance cards of the fraud history case, so as to predict and compare the similarity of medical insurance reimbursement personnel based on their historical fraud methods, obtain the fraud coefficient of each fraud history case through the information value and method value of each fraud case, the fraud coefficient is used to judge the probability of fraudulent behavior of medical insurance reimbursement personnel based on the fraud case, incorporate the fraud coefficient of each fraud case into the case database, and obtain the reference range of the fraud coefficient through the dispersion of all fraud coefficients; S2: Obtain the fraud coefficient of the medical insurance reimbursement personnel, obtain the historical coefficient of the medical insurance reimbursement personnel, obtain the screening reference range based on the fraud coefficient and the historical coefficient of the medical insurance reimbursement personnel, and obtain the similarity overlap degree based on the screening reference range and the fraud coefficient reference range of the medical insurance reimbursement personnel; S3: Determine whether to activate an alarm to remind management personnel based on the similarity and overlap of the medical insurance reimbursement personnel, and determine whether to include the fraud coefficient of the medical insurance reimbursement personnel in the case database based on the review results of the medical insurance reimbursement personnel.

Images