Receipt data review device, receipt data review method, and program

Abstract

In the present invention, receipt data conversion information is generated on the basis of receipt data indicating a detailed statement of medical fees, said receipt data conversion information indicating the presence or absence of medical care information items that correspond to a plurality of examination failure conditions. The receipt data conversion information and an examination failure risk predictor are used to calculate the failure risk level as based on the medical care information item or items included in the receipt data. A risk factor is specified, said risk factor indicating a medical care information item with a degree of contribution to the failure risk level calculation that is greater than or equal to a prescribed threshold value for degree of contribution. If the failure risk level is greater than or equal to a prescribed threshold value for risk level, a revision screen for the medical care information item that is a risk factor is outputted.

Description

[Technical field]

[0001] The present invention relates to a medical receipt data review device, a medical receipt data review method, program Regarding. [Background technology]

[0002] Medical institutions create receipt data showing details of medical fees, etc., and send that receipt data to an examination institution. If the examination passes, the medical institution can receive the unreceived medical fees, which are medical expenses other than the patient's share of the medical fees. If there is an error in the contents of this receipt data, the examination will not pass and the unreceived medical fees cannot be received. Therefore, it is necessary to check whether the medical fee details included in the receipt data will pass the examination. Related technologies are disclosed in Patent Document 1 and Patent Document 2.

[0003] Patent Document 1 discloses a technology that identifies the amount of medical fees that may not be paid to a medical institution by referring to the amount of medical fees. Patent Document 2 discloses a technology that makes it easy to find errors in each medical receipt. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2014-21680 A [Patent Document 2] JP 2004-265444 A Summary of the Invention [Problem to be solved by the invention]

[0005] Among the many items on receipts that are deemed non-compliant, in addition to incorrect items such as mistakes, there are also items that are not incorrect, such as when the combination of medical information items listed on the receipt, such as the relationship between the main disease name and the treatment method, may be deemed inappropriate by the examiner. Therefore, in addition to simple mistakes on receipts, it was necessary to identify other items that may be non-compliant.

[0006] Therefore, the present invention provides a medical receipt data examination device and a medical receipt data examination method that solve the above problems. program The purpose is to provide. [Means for solving the problem]

[0007] According to a first aspect of the present invention, a prescription data review device includes a conversion means for generating prescription data conversion information indicating the presence or absence of medical information items corresponding to multiple review failure conditions based on prescription data indicating a medical fee statement, a failure risk calculation means for calculating a failure risk degree based on the medical information items included in the prescription data using the prescription data conversion information and an examination failure risk predictor, a risk factor identification means for identifying risk factors indicating medical information items whose contribution degree of the medical information items to the calculation of the failure risk degree is equal to or greater than a predetermined contribution degree threshold, and a correction processing means for outputting a correction screen for the medical information items that are the risk factors when the failure risk degree is equal to or greater than a predetermined risk degree threshold.

[0008] According to a second aspect of the present invention, a prescription data review method generates prescription data conversion information indicating the presence or absence of medical information items corresponding to multiple review failure conditions based on prescription data indicating a medical fee statement, calculates a failure risk degree based on the medical information items included in the prescription data using the prescription data conversion information and an examination failure risk predictor, identifies risk factors indicating medical information items whose contribution degree of the medical information items to the calculation of the failure risk degree is equal to or greater than a predetermined contribution degree threshold, and outputs a correction screen for the medical information items that are the risk factors if the failure risk degree is equal to or greater than a predetermined risk degree threshold.

[0009] According to a third aspect of the present invention, the storage medium stores programs that cause the computer of the prescription data review device to function as: a conversion means for generating prescription data conversion information indicating the presence or absence of medical information items corresponding to multiple review failure conditions based on prescription data indicating a medical fee statement; a failure risk calculation means for using the prescription data conversion information and an examination failure risk predictor to calculate a failure risk degree based on the medical information items contained in the prescription data; a risk factor identification means for identifying risk factors indicating medical information items whose contribution degree of the medical information items to the calculation of the failure risk degree is equal to or greater than a predetermined contribution degree threshold; and a correction processing means for outputting a correction screen for the medical information items that are the risk factors when the failure risk degree is equal to or greater than a predetermined risk degree threshold. Effect of the Invention

[0010] According to the present invention, in addition to simple mistakes in medical receipts, it is possible to identify other entries that may result in disqualification from review based on past cases. [Brief description of the drawings]

[0011] [Figure 1] 1 is a diagram showing an outline of a medical treatment system according to an embodiment of the present invention. [Diagram 2] This is a hardware configuration diagram of the prescription data review device according to this embodiment. [Diagram 3]FIG. 2 is a functional block diagram of the prescription data review device according to the first embodiment. [Figure 4] FIG. 1 is a first diagram showing the processing flow of the prescription data review device according to the first embodiment. [Diagram 5] A second diagram showing the processing flow of the prescription data review device according to the first embodiment. [Figure 6] FIG. 11 is a functional block diagram of a prescription data review device according to a second embodiment. [Figure 7] This is the first diagram showing the processing flow of the prescription data review device according to the second embodiment. [Figure 8] A second diagram showing the processing flow of the prescription data review device according to the second embodiment. [Figure 9] A diagram showing the minimum configuration of the prescription data review device according to this embodiment. [Figure 10] A diagram showing the processing flow of a prescription data review device with a minimum configuration. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, a medical receipt data review device according to one embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing an outline of a medical treatment system including a medical receipt data examination device according to this embodiment. As shown in Fig. 1, the medical treatment system 100 is configured by connecting a medical receipt data review device 1 and a terminal 2 used by a user such as a medical office staff via a wireless communication network or a wired communication network. The user uses the terminal 2 to create medical receipt data such as a medical fee statement, and instructs the medical receipt data review device 1 to review the created medical receipt data. The medical receipt data review device 1 performs the review process of the medical receipt data.

[0013] Figure 2 is a diagram showing the hardware configuration of the medical receipt data review device. As shown in this diagram, the medical receipt data examination device 1 is a computer equipped with various hardware components such as a CPU (Central Processing Unit) 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, a HDD (Hard Disk Drive) 104, a communication module 105, and a database 106. The terminal 2 also has a similar hardware configuration.

[0014] FIG. 3 is a functional block diagram of the medical receipt data review device according to the first embodiment. The medical receipt data examination device 1 executes an examination program stored in advance, and thereby performs the functions of a first conversion unit 11, a risk predictor generation unit 12, an acquisition unit 13, a second conversion unit 14, a risk prediction unit 15, an output unit 16, and a risk factor correction unit 17.

[0015] The first conversion unit 11 generates medical receipt data conversion information indicating the presence or absence of medical information items corresponding to a plurality of screening failure conditions, using medical receipt data recorded in the database 106. The medical information items are information written on the medical receipt, such as the main disease name and medical treatment method. The risk predictor generation unit 12 uses machine learning to learn the relationship between the prescription data conversion information generated from the previously stored prescription data, the prescription data review failure result, and the medical information items that are the cause of the failure in the review failure result, and generates an review failure risk predictor. The acquisition unit 13 acquires the medical receipt data indicating the medical receipt to be reviewed and other information input to the medical receipt data review device 1 by the user. The second conversion unit 14, like the first conversion unit 11, uses the acquired medical receipt data to be reviewed to generate medical receipt data conversion information indicating the presence or absence of medical information items that correspond to multiple review failure conditions.

[0016] The risk prediction unit 15 calculates the failure risk level based on the medical information items included in the medical receipt data and their combinations using the medical receipt data conversion information and the examination failure risk predictor. The risk prediction unit 15 also identifies risk factors indicating medical information items whose contribution to the calculation of the failure risk level is equal to or greater than a predetermined contribution threshold. The output unit 16 outputs an examination result screen showing the calculated failure risk degree and risk factors to a monitor or the like. When the failure risk level is equal to or greater than a predetermined risk level threshold, the risk factor correction unit 17 corrects the medical information items that are risk factors based on a user instruction.

[0017] By utilizing these functions, the medical prescription data review device 1 provides technology that can identify not only simple mistakes in medical prescriptions, but also other entries that may result in disqualification for review based on past cases.

[0018] First Embodiment FIG. 4 is a first diagram showing the processing flow of the medical receipt data review device according to the first embodiment. FIG. 5 is a second diagram showing the processing flow of the medical receipt data review device according to the first embodiment. Next, the operation of the medical receipt data examination device of the first embodiment will be described in detail with reference to Figures 4 and 5.

[0019] The database 106 stores, for a plurality of past receipt data, at least, information on the relationship between the past receipt data, the examination failure result of the receipt data, and the medical information items that were the cause of the failure in the examination failure result. The first conversion unit 11 acquires information on the relationship between the past receipt data recorded in the database 106, the examination failure result of the receipt data, and the medical information items that were the cause of the failure in the examination failure result (step S101).

[0020] The first conversion unit 11 acquires a plurality of screening failure conditions from the database 106. As an example, the screening failure conditions include the relationship between the main disease name and a plurality of medical information items such as a medical treatment method that may be a screening failure method for the main disease name. The medical treatment method, which is an example of a medical information item, includes information such as medication name and medication period. When the main disease name described in the medical fee statement shown by the receipt data matches the screening failure conditions, the first conversion unit 11 identifies information on the main disease name "present." When the main disease name described in the medical fee statement shown by the receipt data matches the screening failure conditions and the medication name matches the screening failure conditions, the first conversion unit 11 identifies information on the medication name "present." When the main disease name described in the medical fee statement shown by the medical receipt data matches the examination failure condition, the first conversion unit 11 calculates the medication period based on the total amount of medication and the number of doses per day described in the medical fee statement, and if the medication period is longer than the medication period of the examination failure condition, it identifies information that is "present" for the medication period or longer. As an example, the first conversion unit 11 generates, based on the medical receipt data, medical receipt data conversion information indicating the main disease name "present or absent", the medication name "present or absent", and information "present or absent" for the medication period or longer indicated by the examination failure condition. In the medical receipt data conversion information, "present" is held as numerical information such as "1" and "0" for absence. The first conversion unit 11 identifies whether there is a description that matches multiple examination failure conditions from the medical receipt data, and generates medical receipt data conversion information indicating the presence or absence of medical information items corresponding to multiple examination failure conditions (step S102).

[0021] The first conversion unit 11 outputs the receipt data conversion information, the review failure results of the receipt data used to generate the receipt data conversion information, and information on the relationship with the medical information items that were the cause of the failure in the review failure results to the risk predictor generation unit 12. The first conversion unit 11 generates receipt data conversion information based on a large number of receipt data recorded in the database 106, and outputs to the risk predictor generation unit 12 information on the relationship between the receipt data conversion information, the review failure results of the receipt data used to generate the information, and the medical information items that were the cause of the failure in the review failure results.

[0022] The risk predictor generating unit 12 acquires a plurality of pieces of information on the relationship between the receipt data conversion information, the examination failure result of the receipt data used to generate the information, and the medical information items that are the cause of the failure in the examination failure result. The risk predictor generating unit 12 acquires a plurality of pieces of information on the relationship between the acquired receipt data conversion information, the examination failure result, and the medical information items that are the cause of the failure in the examination failure result, and performs a learning process on the plurality of pieces of information by machine learning. The risk predictor generating unit 12 performs a learning process using a method such as logistic regression or LightGBM. Note that these learning methods are examples, and other methods may be used as long as a predictor can be generated. The risk predictor generating unit 12 generates an examination failure risk predictor (examination failure risk prediction model) that calculates the degree of failure risk based on the medical information items included in the receipt data as a result of the learning process (step S103). The risk predictor generating unit 12 may generate screening failure risk predictors using a plurality of different machine learning techniques, and may accept the selection of the screening failure risk predictor having the best calculation accuracy for the failure risk degree among them. The risk predictor generating unit 12 stores data of the generated or selected screening failure risk predictor in a storage unit or the like.

[0023] By the above processing, the medical receipt data review device 1 completes the processing for reviewing the medical receipt data to be newly reviewed.

[0024] In the review process, the medical receipt data review device 1 receives a review request including medical receipt data from, for example, the terminal 2 (step S201). The acquisition unit 13 acquires the medical receipt data included in the review request. The acquisition unit 13 outputs the medical receipt data to the second conversion unit 14. The second conversion unit 14 acquires a plurality of review failure conditions from the database 106 (step S202). The second conversion unit 14 generates medical receipt data conversion information based on the medical receipt data to be reviewed by processing similar to that of the first conversion unit 11 (step S203).

[0025] Specifically, the second conversion unit 14 identifies information on the main disease name "present" when the main disease name described in the medical fee statement shown by the receipt data matches the examination failure condition. When the main disease name described in the medical fee statement shown by the receipt data matches the examination failure condition and the medication name matches the examination failure condition, the second conversion unit 14 identifies information on the medication name "present". When the main disease name described in the medical fee statement shown by the receipt data matches the examination failure condition and the medication period is calculated based on the total amount of medication and the number of doses per day described in the medical fee statement, and when the medication period is equal to or longer than the medication period of the examination failure condition, the second conversion unit 14 identifies information on the medication period or more "present". As an example, the second conversion unit 14 generates, based on the receipt data, receipt data conversion information indicating the main disease name "present or absent", the medication name "present or absent", and the information "present or absent" of the medication period or more indicated by the examination failure condition. The second conversion unit 14 identifies whether or not there is a description that matches the multiple examination failure conditions from the medical receipt data, and generates medical receipt data conversion information indicating the presence or absence of medical information items that correspond to the multiple examination failure conditions. The second conversion unit 14 outputs the generated medical receipt data conversion information to the risk prediction unit 15.

[0026] The risk prediction unit 15 acquires an examination failure risk predictor. The risk prediction unit 15 inputs the receipt data conversion information into the examination failure risk predictor, and as a result, calculates the failure risk degree (step S204). The risk prediction unit 15 also identifies risk factors indicating medical information items whose contribution degree to the calculation of the failure risk degree is equal to or greater than a predetermined contribution degree threshold (step S205). When the failure risk degree is equal to or greater than a predetermined threshold, the risk prediction unit 15 may identify risk factors indicating medical information items whose contribution degree to the calculation of the failure risk degree is equal to or greater than a predetermined contribution degree threshold.

[0027] The risk prediction unit 15 performs calculations using a known technique such as a correlation analysis technique when identifying risk factors indicating medical information items whose contribution to the calculation of the pass risk degree is equal to or greater than a predetermined contribution threshold. The risk prediction unit 15 outputs the calculated fail risk degree and information on the medical information items that contributed greatly to the calculation of the fail risk degree to the output unit 16. The risk prediction unit 15 may calculate the contribution degree for each of the multiple medical information items included in the receipt data conversion information and output it to the output unit 16.

[0028] The output unit 16 generates the examination result screen information (correction screen) (step S206). The output unit 16 may generate the examination result screen information (correction screen) showing the failure risk degree and the name of the medical information item that contributed greatly to the calculation of the failure risk degree. When the failure risk degree is less than a threshold, the output unit 16 may generate the examination result screen information (correction screen) showing only the failure risk degree. The output unit 16 may calculate the contribution degree for each of the multiple medical information items included in the medical receipt data conversion information, and generate the examination result screen information (correction screen) on which the contribution degree for each medical information item is displayed. When the failure risk degree is low, such as less than a predetermined threshold, the output unit 16 may generate the examination result screen information showing information that does not require correction of the medical receipt data.

[0029] The output unit 16 has, for example, a web server function, and uses this function to transmit the examination result screen information to the terminal 2 (step S207). The terminal 2 displays the examination result screen information on a web browser or the like. Through the above process, a user such as a medical office staff member using the terminal 2 can confirm the numerical value of the failure risk degree for the prescription data indicating the created medical fee statement, and the name of the medical information item that contributed greatly to the calculation of the failure risk degree.

[0030] According to the above process, the medical receipt data review device 1 uses a review failure risk predictor generated by using a lot of information on the relationship between past medical receipt data, the review failure results of the medical receipt data, and the medical information items that were the cause of the review failure results. Therefore, in addition to simple incorrect entries such as entry errors, when the relationship between the main disease name and the medical treatment method may be inappropriate, by displaying the failure risk degree calculated based on cases where the entry was not an error but was rejected according to past cases and the name of the medical information item that contributed greatly to the calculation of the failure risk degree, it is possible to identify entries that may be rejected according to past cases other than simple entry errors.

[0031] The review result screen information may be an interface screen that allows the user to modify the contents of the medical fee statement of the prescription data for which the user has made a review request. If the failure risk level is equal to or above a threshold, the user inputs new values ​​or information for the medical information items that contributed greatly to the calculation of the failure risk level in the change column of the prescription data indicated by the review result screen information. For example, the user may change the name of the main illness or reduce the numerical value of the dosage. The user uses terminal 2 to input an update instruction after changing the medical information items to terminal 2. Terminal 2 then transmits an update request including the changed prescription data to the prescription data review device 1.

[0032] The medical receipt data review device 1 determines whether an update request has been received (step S208). When the medical receipt data review device 1 receives an update request, the acquisition unit 13 acquires the medical receipt data included in the update request. When the medical receipt data review device 1 acquires an update request, the medical receipt data review device 1 performs the review process from step S202 onward again using the medical receipt data included in the update request. The medical receipt data review device 1 may automatically request the user to send an update request and repeat the review process until the failure risk degree becomes less than the risk degree threshold. This process is one aspect of the process in which the medical receipt data review device 1 acquires medical receipt data in which the medical information items that are risk factors have been corrected on the correction screen, and repeatedly calculates the failure risk degree based on the medical information items included in the medical receipt data until the failure risk degree becomes less than the risk degree threshold.

[0033] The output unit 16 similarly transmits the review result screen information to the terminal 2 in the second review process. The terminal 2 displays the review result screen information on a web browser or the like. The user checks the review result screen information, and if the risk of failure is low, saves the receipt data displayed in the review result screen information in the terminal 2. This allows the user to create receipt data with a low risk of failure. Thereafter, the user presents the receipt data with a low risk of failure to the doctor and receives instructions on how to correct the receipt.

[0034] According to the processing of the first embodiment described above, the risk of failure of the medical receipt data to be newly examined and risk factors indicating the medical information items whose contribution to the calculation of the risk of failure is equal to or greater than a predetermined contribution threshold are displayed based on past data. According to the processing described above, when the risk of failure is equal to or greater than a predetermined risk threshold, a correction screen for the medical information items that are risk factors is output, and correction is repeated until the risk of failure is determined to be low. With this processing, the user can use the medical receipt data examination device 1 to improve the medical receipt work by selecting in advance medical receipts that will fail examination and correcting them so that they pass examination. According to the above process, the medical receipt data examination device 1 may display each medical information item in order on the terminal 2 based on the contribution of the medical information item to the calculation of the failure risk level. This allows the problematic parts of the medical receipt data to be corrected efficiently.

[0035] Here, the processing of the above-mentioned receipt data examination device 1 will be explained using a specific example of receipt data. Note that the following specific example is merely an example, and the subject processed by this receipt data examination device 1 is not limited to "gastroesophageal reflux disease", and similar processing can be performed using receipt data of other diseases (disease names). For example, the receipt data includes drug information (medication information) and dosage information that can calculate "the administration period of XX preparation is 12 weeks", as well as a description of injury / illness information such as "gastroesophageal reflux disease" as the main disease name. Based on each of these medical information items, the receipt data examination device 1 generates receipt data conversion information indicating the presence or absence of a medical information item corresponding to the examination failure condition (1 if present, 0 if not present).

[0036] When the prescription data review device 1 expresses the presence or absence of drug information (medication information) such as "The prescription period of X formulation is 12 weeks" and illness information such as "gastroesophageal reflux disease" as the main illness name as a numerical value of 0 or 1, it can multiply these values ​​to create prescription data conversion information that includes a data item that is 1 if "The prescription period of X formulation is 12 weeks and the disease is gastroesophageal reflux disease" and 0 if not.

[0037] Using such prescription data conversion information, the risk prediction unit 15 calculates the degree of risk of failure. When the degree of risk of failure is equal to or above the threshold and is high, the output unit 16 outputs medical information items that are screening failure factors that contribute highly to the calculation of the degree of risk of failure. For example, as screening failure factors, medical information items with high contributions such as "XX formulation", "medication period 12 weeks", and "gastroesophageal reflux disease", which can be interpreted as "the administration period of XX formulation is 12 weeks and the patient has gastroesophageal reflux disease", are displayed as output results.

[0038] The risk factor correction unit 17 accepts corrections to medical information items such as "medication period 12 weeks" and "reflux esophagitis" by a correction process from the user among the medical information items with high contributions such as "XX formulation", "medication period 12 weeks", and "reflux esophagitis". In other words, this indicates that the contents of the receipt data have been corrected to "the dosage period of XX formulation is 8 weeks and the patient has reflux esophagitis" or "the dosage period of XX formulation is 12 weeks and the patient has refractory reflux esophagitis". The second conversion unit 14 then converts such corrected receipt data to generate receipt data conversion information, and the risk prediction unit 15 repeats the calculation of the failure risk degree.

[0039] If the degree of risk of failure is sufficiently low, the receipt data inspection device 1 outputs information indicating "no change required." For example, if the predicted risks of both a receipt whose contents have been amended to "the administration period of XX formulation is 8 weeks and the patient has reflux esophagitis" and a receipt whose contents have been amended to "the administration period of XX formulation is 12 weeks and the patient has refractory reflux esophagitis" are sufficiently low, the doctor checks the screen information indicating "no change required" for both receipts.

[0040] A user such as a medical secretary presents the receipt with "No changes required" to a doctor for confirmation. For example, either the receipt corrected to "The prescription period of XX formulation is 8 weeks and the patient has reflux esophagitis" or the receipt corrected to "The prescription period of XX formulation is 12 weeks and the patient has refractory reflux esophagitis" may be correct. For example, if a doctor notices the contradiction that prescription of XX formulation for more than 8 weeks cannot be given for reflux esophagitis, and the mistake was "refractory reflux esophagitis," the latter corrected receipt is determined to be the one with the corrected content. This process can prevent reexamination.

[0041] Second Embodiment FIG. 6 is a functional block diagram of the medical receipt data review device according to the second embodiment. FIG. 7 is a first diagram showing the processing flow of the medical receipt data review device according to the second embodiment. FIG. 8 is a second diagram showing the processing flow of the medical receipt data review device according to the second embodiment.

[0042] Next, the operation of the medical receipt data examination device 1 of the second embodiment will be described in detail with reference to FIGS. The medical receipt data examination device 1 according to the second embodiment further includes the functions of a conversion process correction unit 18 and a risk predictor selection unit 19 in addition to the functions described in Fig. 4. The conversion process correction unit 18 accepts changes to conditions such as medical information items in the examination failure conditions used to generate medical receipt data conversion information. The risk predictor selection unit 19 accepts the selection of an examination failure risk predictor.

[0043] Before the processing of the first embodiment described above is performed, the conversion process correction unit 18 may accept changes to conditions such as medical information items in the examination failure conditions used to generate the medical receipt data conversion information. Then, the second conversion unit 14 may convert the medical receipt data using the examination failure conditions after the medical information items are changed by the conversion process correction unit 18 to generate the medical receipt data conversion information.

[0044] As an example, a user uses the terminal 2 to access the medical receipt data review device 1. The medical receipt data review device 1 uses a web server function to transmit management screen information to the terminal 2 (step S301). The terminal 2 displays the management screen information on the web browser output to the monitor.

[0045] In the management screen, the user can specify the examination failure condition and select and change the medical information items included in the examination failure condition. For example, the user assumes that the examination failure condition includes information on the relationship between the main disease name and the medical treatment method that is inappropriate as a medical treatment method for the main disease name. The user changes the information on the medication period among the medical treatment methods in the examination failure condition in the management screen. The user may be able to specify the examination failure condition and make changes such as adding or deleting the medical information items included in the examination failure condition. The user issues an update instruction by pressing an update button or the like in the management screen. The terminal 2 transmits change instruction information including the identifier of the examination failure condition and the medical information items such as the medication period included in the medical treatment method to be changed to the prescription data review device 1.

[0046] The acquisition unit 13 of the medical receipt data review device 1 acquires change instruction information (step S302). The acquisition unit 13 outputs the change instruction information to the conversion process correction unit 18. The conversion process correction unit 18 reads the identifier of the review failure condition and the medical information items such as the medication period included in the medical treatment method to be changed from the change instruction information, and updates the medical information items included in the corresponding review failure condition recorded in the database 106 (step S303). Then, in the process of step S203 of the medical receipt data review device 1, when performing the review process shown in the first embodiment, the second conversion unit 14 generates medical receipt data conversion information using the updated review failure condition. The first conversion unit 11 may also generate medical receipt data conversion information using the updated review failure condition.

[0047] When a user accesses the medical receipt data review device 1 using the terminal 2, the risk predictor selection unit 19 may include, in the management screen information transmitted by the medical receipt data review device 1 to the terminal 2, identifiers of the medical receipt failure risk predictors generated by a plurality of different learning methods, and accuracy rate, precision rate, recall rate, F value, ROC curve, AUC (Area Under the Curve), etc., when the review is performed using each of those predictors. When the management screen information is displayed on the terminal 2, identifiers of the medical receipt failure risk predictors generated by the plurality of different learning methods, and accuracy rate, precision rate, recall rate, F value, ROC curve, AUC, etc., when the review is performed using each of those predictors may be displayed.

[0048] The user checks the identifiers of the examination failure risk predictors generated by multiple different learning methods displayed in the management screen information of terminal 2, and the accuracy rate, precision rate, recall rate, F value, ROC curve, and AUC when an examination is performed using each of those predictors, and specifies one of the examination failure risk predictors. Terminal 2 then transmits selection instruction information including the identification information of the specified examination failure risk predictor to the medical receipt data examination device 1.

[0049] The risk predictor selection unit 19 of the medical receipt data review device 1 acquires selection instruction information (step S401). The risk predictor selection unit 19 acquires identification information of the selected medical receipt failure risk predictor from the selection instruction information (step S402). The risk predictor selection unit 19 instructs the medical receipt failure risk predictor that the risk prediction unit 15 uses to calculate the failure risk degree based on the acquired identification information of the medical receipt failure risk predictor. The risk prediction unit 15 stores the selected medical receipt failure risk predictor as a predictor to be used to calculate the risk prediction degree (step S403). As a result, the risk prediction unit 15 inputs the medical receipt data conversion information acquired from the second conversion unit 14 into the selected medical receipt failure risk predictor to calculate the failure risk degree.

[0050] According to the processing of the second embodiment described above, the processing is configured to allow the user to change, delete, or add clinical information items included in the examination failure conditions as necessary. Therefore, if the examination criteria at the examination institution change, examination failure conditions can be set in accordance with the changes, and new examination failure factors can be added if the analyst wishes to add their own knowledge.

[0051] In addition, according to the processing of the second embodiment described above, it is configured so that an examination failure risk predictor can be selected based on the evaluation scale, so that an appropriate degree of failure risk can be calculated depending on the situation in which it is used.

[0052] FIG. 9 is a diagram showing the minimum configuration of a medical receipt data review device. FIG. 10 is a diagram showing the processing flow of a medical receipt data review device with a minimum configuration. The medical receipt data review device 1 includes at least a conversion means 91, a failure risk calculation means 92, a risk factor identification means 93, and a correction processing means 94. The conversion means 91 generates medical receipt data conversion information indicating the presence or absence of medical information items corresponding to a plurality of screening failure conditions based on the acquisition of medical receipt data indicating a medical fee statement (step S901). The failure risk calculation means 92 uses the medical receipt data conversion information and the examination failure risk predictor to calculate the degree of failure risk based on the medical information items included in the medical receipt data (step S902). The risk factor identifying means 93 identifies risk factors indicating medical information items whose contributions to the calculation of the failure risk degree are equal to or greater than a predetermined contribution threshold (step S903). When the failure risk degree is equal to or greater than a predetermined risk degree threshold, the correction processing means 94 outputs a correction screen for the medical information item that is a risk factor (step S904).

[0053] Each of the above-mentioned devices has a computer system inside. The steps of each of the above-mentioned processes are stored in a computer-readable storage medium in the form of a program, and the above processes are performed by the computer reading and executing this program. Here, the computer-readable storage medium refers to a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, etc. Also, the computer program may be distributed to a computer via a communication line, and the computer that receives the program may execute the program.

[0054] The program may be a program for implementing some of the above-mentioned functions, or may be a so-called differential file (differential program) that can implement the above-mentioned functions in combination with a program already recorded in the computer system. [Explanation of symbols]

[0055] 1. Receipt data review device 2. Terminal 11. First conversion unit 12. Risk predictor generator 13... Acquisition part 14 Second conversion unit 15 Risk prediction unit (failure risk calculation means, risk factor identification means) 16 Output section 17 Risk Factor Modification Section 18. Conversion process correction section 19. Risk predictor selection section

Claims

1. A conversion means for generating medical receipt data conversion information indicating the presence or absence of medical information items that match the medical information items indicated by the multiple screening failure conditions based on medical receipt data indicating a medical fee statement; a failure risk calculation means for inputting the receipt data conversion information into an examination failure risk predictor generated by learning and processing the relationship between the previously stored receipt data, the examination failure result of the receipt data, and the medical information items that are the cause of the failure in the examination failure result by machine learning, the examination failure risk predictor calculating a failure risk degree according to a relationship between a main disease name and a medical treatment method or a medication method in the medical information items, and calculating a failure risk degree based on the medical information items included in the receipt data; A risk factor identification means for identifying risk factors indicating medical information items whose contribution degree to the calculation of the failure risk degree is equal to or greater than a predetermined contribution degree threshold using a correlation analysis method; a correction processing means for outputting a correction screen for the medical information item that is the risk factor when the failure risk degree is equal to or greater than a predetermined risk degree threshold; A prescription data review device comprising:

2. The conversion means generates, based on the medical receipt data, medical receipt data conversion information indicating the presence or absence of one or more medical information items corresponding to a plurality of examination failure conditions. The medical receipt data review device according to claim 1.

3. an output means for outputting failure risk information based on the failure risk degree and a medical information item that is a risk factor when the failure risk degree is equal to or greater than the risk degree threshold; The medical receipt data review device according to claim 1 or 2, comprising:

4. The failure risk calculation means acquires medical receipt data in which the medical information items that are risk factors have been corrected on the correction screen, and repeatedly calculates a failure risk degree based on the medical information items included in the medical receipt data until the failure risk degree becomes less than the risk degree threshold value. The medical receipt data examination device according to any one of claims 1 to 3.

5. The risk predictor generating means for generating the examination failure risk predictor generates a plurality of the examination failure risk predictors for each of the learning processing methods using a plurality of different learning processing methods, A risk predictor selection means for accepting the selection of the examination failure risk predictor is provided. The medical receipt data examination device according to any one of claims 1 to 4.

6. A conversion process correction means for accepting changes to medical information items corresponding to the examination failure conditions used to generate the medical receipt data conversion information; The medical receipt data review device according to any one of claims 1 to 5, comprising:

7. A computer of a medical receipt data review device, generating, based on the medical receipt data indicating the medical fee statement, medical receipt data conversion information indicating the presence or absence of medical information items that match the medical information items indicated by the multiple examination failure conditions; inputting the receipt data conversion information into an examination failure risk predictor that is generated by learning and processing the relationship between the previously stored receipt data, the examination failure result of the receipt data, and the medical information item that is the cause of the failure in the examination failure result by machine learning, the examination failure risk predictor calculating a failure risk degree according to a relationship between a main disease name and a medical treatment method or a medication method in the medical information item, and calculating a failure risk degree based on the medical information item included in the receipt data; Identifying risk factors that indicate medical information items whose contribution to the calculation of the failure risk degree is equal to or greater than a predetermined contribution threshold using a correlation analysis method; When the failure risk level is equal to or greater than a predetermined risk level threshold, a correction screen for the medical information item that is the risk factor is output. How prescription data is reviewed.

8. The computer of the medical receipt data review device, A conversion means for generating medical receipt data conversion information indicating the presence or absence of medical information items that match the medical information items indicated by the multiple examination failure conditions based on medical receipt data indicating a medical fee statement; a failure risk calculation means for inputting the medical receipt data conversion information into an examination failure risk predictor generated by learning and processing the relationship between the medical receipt data previously stored, the examination failure result of the medical receipt data, and the medical information items that are the cause of the failure in the examination failure result by machine learning, the examination failure risk predictor calculating a failure risk degree according to a relationship between a main disease name and a medical treatment method or a medication method in the medical information items, and calculating a failure risk degree based on the medical information items included in the medical receipt data; A risk factor identification means for identifying risk factors indicating medical information items whose contribution degree to the calculation of the failure risk degree is equal to or greater than a predetermined contribution degree threshold value using a correlation analysis method; a correction processing means for outputting a correction screen for the medical information item that is the risk factor when the failure risk degree is equal to or greater than a predetermined risk degree threshold; A program that functions as a

Images