Prescription entry support system and prescription entry support method

Abstract

We provide technology that appropriately supports prescription data entry tasks. [Solution] The prescription input support system includes means for acquiring prescription information, which includes information about the target patient and the prescription details of the target prescription, information about the target patient and the prescription details of the target prescription, means for acquiring dispensing fee addition item information that indicates at least applicable drug preparation fee addition items or pharmaceutical management fee items for the target prescription, or comment information for the dispensing fee statement for the target prescription, by inputting at least the acquired prescription information into an inference model including a trained machine learning model, and means for outputting drug information to be dispensed and the acquired dispensing fee addition item information or comment information corresponding to the acquired prescription information.

Description

[Technical Field]

[0001] The present invention relates to a technology for supporting prescription input work in dispensing pharmacies (including in-hospital pharmacies). [Background Art]

[0002] When a dispensing pharmacy (including in-hospital pharmacies) receives a prescription issued by a physician, the receptionist inputs various types of information related to the prescription into a computer system such as a receipt computer (re-secon). The information to be input includes, in addition to information to be described in the drug information sheet provided to the patient, the dispensing statement, and the medication notebook, information that should be described in the dispensing remuneration statement (sometimes referred to as a receipt). For example, as patient information, name, date of birth, insurer number, etc. are input; as issuer information, the name of the physician, the name of the medical institution, issue date, etc. are input; as prescription content, drug information such as the name of the drug and dosage, administration method, supplementary information for administration method, dispensing instructions, etc. are input.

[0003] The following Patent Document 1 proposes a method for supporting medication history creation work as part of such pharmacy work. In this method, patient personal information indicating the patient's attributes, medical questionnaire information indicating the patient's answers to the medical questionnaire, and prescription information indicating the content of the prescription for the patient are input into a learning model, and the learning model is caused to generate and output entry information for at least Subject (patient's subjective information) and Object (objective information) in SOAP (Subject Object Assessment Plan) medication histories.

[0004] Furthermore, Patent Document 2 discloses a method for supporting dispensing operations, which involves acquiring prescription data, changing the generic names of drugs in the acquired prescription data to names of drugs suitable for the pharmacy or patient, generating processed prescription data (dispensing instruction data) including text data, and transmitting the generated processed prescription data to a dispensing management computer and a claims processing computer. This method also discloses that drug picking errors can be reduced by performing drug picking operations based on the names of drugs used in the generated processed prescription data. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Patent No. 7507329 [Patent Document 2] Japanese Patent Publication No. 2024-86305 [Overview of the project] [Problems that the invention aims to solve]

[0006] The methods described above can support pharmacy operations such as entering patient medication history and converting and entering the names of medications to be dispensed. However, the information that needs to be entered regarding received prescriptions includes not only information for various documents given to patients, but also information for prescription fee statements (receipts), and conventional support methods cannot be said to adequately support prescription entry work.

[0007] This invention has been made in view of these circumstances and provides a technology that appropriately supports prescription entry work. [Means for solving the problem]

[0008] According to the present invention, a prescription acquisition means for acquiring prescription information that includes information about a target prescription issued for a target patient and the prescription details of the target prescription, A drug selection means for selecting candidate drugs to be dispensed for the prescription content of the target prescription based on the acquired prescription information,An inference means that inputs at least the acquired prescription information into an inference model including a trained machine learning model to acquire dispensing fee addition item information that indicates at least the drug preparation fee addition item or pharmaceutical management fee item applicable to the target prescription, or comment information for the dispensing fee statement related to the target prescription, and Selected Medications to be dispensed Candidate A prescription input support system is provided, comprising an output processing means that outputs information and the acquired dispensing fee addition item information or comment information.

[0009] Furthermore, according to the present invention, a prescription input support method is performed on one or more computers, wherein the one or more computers acquire prescription information relating to a target prescription issued for a target patient, including information about the target patient and the prescription details of the target prescription. Based on the acquired prescription information, candidate drugs to be dispensed for the prescription contents of the target prescription are selected. By inputting at least the acquired prescription information into an inference model that includes a trained machine learning model, information on additional dispensing fees that indicates at least the applicable drug preparation fee additional items or pharmaceutical management fee items for the target prescription, or comment information for the dispensing fee statement for the target prescription, Selected Medications to be dispensed Candidate A prescription input support method is provided, which includes outputting information and the acquired dispensing fee addition item information or comment information. Furthermore, according to the present invention, a computer program capable of causing one or more computers to execute the above-described prescription input support method, and a recording medium on which the computer program is stored, may also be provided. [Effects of the Invention]

[0010] The present invention provides a technology that appropriately supports prescription data entry tasks. [Brief explanation of the drawing]

[0011] [Figure 1] This diagram shows an example of a system configuration for a pharmacy operations support system. [Figure 2]This diagram conceptually illustrates an example of the software configuration for a prescription claims server. [Figure 3] This figure shows an example of an input confirmation screen. [Figure 4] This is a flowchart conceptually illustrating the prescription input support method (this support method) according to this embodiment. [Modes for carrying out the invention]

[0012] The following describes embodiments of the present invention (hereinafter sometimes referred to as "these embodiments"). Note that the embodiments listed below are illustrative, and the present invention is not limited to the configurations of the embodiments described below.

[0013] First, an overview of the prescription input support system and prescription input support method according to this embodiment will be described. The prescription input support system according to this embodiment is composed of one or more computers and includes at least prescription acquisition means, inference means, and output processing means. Each of these means may be implemented entirely on a single computer, or they may be distributed across multiple computers. The hardware configuration of the prescription input support system according to this embodiment is not limited in any way.

[0014] The prescription acquisition method acquires prescription information that includes information about the target prescription issued for the target patient, including the patient's information and the prescription details. The acquired prescription information includes, for example, the patient's information such as age and date of birth, and for example, the drug name, dosage method, and dosage form as prescription details. However, the specific content of the acquired prescription information is not limited.

[0015] The prescription acquisition means may acquire the prescription information from another computer or device via a communication means, or may acquire the prescription information by applying some kind of processing to the original data. Regarding the latter, for example, the prescription information can be generated from character data obtained by applying OCR (Optical Character Recognition) processing to a prescription image obtained by scanning or imaging a paper-based prescription. Further, the prescription information may be generated from electronic prescription data. As such, there is no particular limitation on the method of acquiring prescription information by the prescription acquisition means.

[0016] The inference means inputs at least the prescription information acquired by the prescription acquisition means into an inference model including a trained machine learning model, thereby acquiring dispensing fee addition item information indicating at least applicable drug preparation fee addition items or pharmaceutical management fee items for the target prescription, or comment information for a dispensing fee statement related to the target prescription. The inference model is configured to be capable of outputting, by inputting at least prescription information, dispensing fee addition item information indicating at least applicable drug preparation fee addition items or pharmaceutical management fee items for the target prescription, or comment information for a dispensing fee statement related to the target prescription. The information input to the inference model may be all or only a part of the prescription information acquired by the prescription acquisition means, and may further include other information. Further, the output information from the inference model may be only the dispensing fee addition item information, may be only the comment information, may be only the dispensing fee addition item information and the comment information, or may further include other information.

[0017] "Dispensing fee addition item information" indicates at least a drug preparation fee addition item or a pharmaceutical management fee item specified in the dispensing fee point table, and may indicate only the drug preparation fee addition item, may indicate only the pharmaceutical management fee item, or may indicate both of them. The dispensing fee addition item information may be an ID capable of identifying the item, or may be character string information indicating the item name. The "drug preparation additional items" are additional items under "drug preparation fees" in the "dispensing technical fees" specified in the dispensing fee point table, and include, for example, items such as sterile preparation processing additional fee, additional fee for narcotics etc., in-house preparation additional fee (for oral medication / occasional medication / external medication), measured mixing dispensing additional fee, out-of-hours additional fee, and additional fee for nighttime / holidays etc. The "pharmaceutical management fee items" are specific items under the "pharmaceutical management fees" specified in the dispensing fee point table, and include, for example, items such as additional fee for medication guidance for infants and young children, comprehensive management fee for community pharmacists, outpatient medication support fee 1, and outpatient medication support fee 2. The dispensing fee additional item information is information necessary for dispensing fee statements (receipts), dispensing statements, and the like.

[0018] The "comment information for dispensing fee statements" is information indicating one or both of a summary comment and a prescription comment, and includes at least text data (character string data) of each comment. The "summary comment" is a comment described in the summary column of a dispensing fee statement (receipt), and the "prescription comment" is a comment described in the prescription column of the receipt.

[0019] The inference model includes one or more trained machine learning models. The inference model may be configured only with a machine learning model, or may be configured with a machine learning model and a rule-based model. A rule-based model is a model that derives an output based on pre-designed rules. The inference model includes one or more trained machine learning models and only needs to be capable of outputting either or both the information on additional dispensing fees or the comment information for claims, taking prescription information as input. Its specific configuration and training method are not limited in any way. For this reason, the inference model may be constructed to output the information on additional dispensing fees or the comment information for claims through cooperation between one or more rule-based models and one or more machine learning models, or through cooperation between multiple machine learning models. For example, the inference model may be constructed to use information output from one rule-based model as input information for a machine learning model, or to use information output from one machine learning model as input information for another rule-based model, or to use information output from one machine learning model as input information for another machine learning model. In the specific examples described later, the inference model includes at least a dispensing fee inference model that outputs information on additional dispensing fee items and a comment generation model that outputs summary comment information.

[0020] Here, a "trained machine learning model" refers to a model obtained through machine learning using training data, i.e., supervised learning, and can be written as an AI (Artificial Intelligence) model, a Machine Learning (ML) model, etc. The "trained machine learning model" used in this embodiment may be a regression equation obtained by regression analysis, or a neural network model obtained by principal component analysis or deep learning, and the data structure and learning algorithm of the model are not limited. For example, the machine learning model may be realized by a combination of a computer program and parameters, or a combination of multiple functions and parameters. When the machine learning model is constructed using a neural network, and the input layer, hidden layer, and output layer are considered as a single neural network unit, it may refer to a single neural network or a combination of multiple neural networks. Furthermore, the machine learning model may consist of a combination of multiple multiple regression equations or a single multiple regression equation.

[0021] The inference model may be stored on one or more computers that constitute the prescription entry support system, or it may be stored on an external computer that is accessible via communication from those one or more computers. If the inference model is stored on the computer implementing the inference means, the inference means uses the inference model to obtain the dispensing fee addition item information or comment information output from that model. On the other hand, if the inference model is stored on a computer other than the computer implementing the inference means, the inference means sends input information to the inference model on that other computer via communication and obtains the dispensing fee addition item information or comment information output from that model via communication.

[0022] The output processing means outputs information on the medication to be dispensed corresponding to the prescription information obtained by the prescription acquisition means, and information on additional dispensing fee items or comment information obtained by the inference means. The output format of the medication information, additional dispensing fee items, and comment information is not limited. For example, the output processing means may display the information on a display unit as in the specific example described later, transmit the information to another computer, or save the information as an electronic file.

[0023] The prescription input support method according to this embodiment is executed by one or more computers, such as the prescription input support system described above. The prescription input support method includes one or more computers acquiring prescription information that includes information about the target patient and the prescription details of the target prescription, inputting at least the acquired prescription information into an inference model that includes a trained machine learning model, thereby acquiring information on additional dispensing fees that indicate at least applicable drug preparation fee additional items or pharmaceutical management fee items for the target prescription, or comment information for the dispensing fee statement for the target prescription, and outputting information on the drugs to be dispensed and the acquired information on additional dispensing fees or comment information corresponding to the acquired prescription information.

[0024] As described above, according to this embodiment, when prescription information is acquired, information on the drugs to be dispensed and information on additional dispensing fee items or comment information for claims are output. Therefore, information that previously had to be manually entered in prescription input work can be automatically acquired, thereby reducing the effort required for prescription input work. Furthermore, identifying additional dispensing fee items and comments for claims processing from prescription information requires experience and knowledge of pharmacy operations. However, by using an inference model that includes at least a pre-trained machine learning model to infer additional dispensing fee items or comments, it becomes possible to appropriately identify additional dispensing fee items or comments for claims processing regardless of experience or knowledge, thereby optimizing prescription entry work.

[0025] Hereafter, the prescription input support system and prescription input support method according to this embodiment will be described in more detail with specific examples. In this specific example, the prescription input support system according to this embodiment is implemented as a dispensing claims server 10, which will be described later.

[0026] <Pharmacy Business Support System> Figure 1 shows an example of the system configuration of pharmacy operations support system 1. The pharmacy operations support system 1 includes a pharmacy terminal 3, an accounting system 5, an inventory management system 6, an electronic patient record system 7, an in-pharmacy system 8, a prescription claims server 10, etc., all of which are interconnected via a communication network 2. Communication network 2 consists of one or more of the following: a public network such as the Internet, a WAN (Wide Area Network), a LAN (Local Area Network), a wireless communication network, etc. However, the communication methods implemented in communication network 2 are not limited in this embodiment.

[0027] Pharmacy terminal 3 is a computer operated by a pharmacist or pharmacy staff member and functions as the user interface for pharmacy operations support system 1. Pharmacy terminal 3 may be a stationary computer or a portable computer such as a tablet. Although not shown in the diagram, pharmacy terminal 3 has a CPU, memory, input / output interface (I / F), and communication unit, and is connected to user interface devices such as a display unit and input unit via the I / F. Pharmacy terminal 3 is connected to a communication network 2, enabling it to communicate with each component within the pharmacy operations support system 1. By exchanging information with these components, it outputs various display screens and accepts various user operations. As will be described in detail later, pharmacy terminal 3 enables the input of prescription information, the referencing of patient information, and the confirmation of dispensing history. In this embodiment, the pharmacy terminal 3 only needs to be able to function as a user interface for the pharmacy business support system 1, and its hardware and software configurations are not limited.

[0028] The accounting system 5 performs accounting processing related to the sale of pharmaceuticals to patients and consists of one or more computers, such as a POS (Point of Sales) terminal. For example, the accounting system 5 obtains information such as dispensing fee points from the dispensing claim server 10 and performs accounting processing. In the accounting processing, in addition to accounting slips (receipts), drug information sheets and dispensing details to be given to patients are printed. However, in this embodiment, the accounting system 5 only needs to have known functions, and its specific functions are not limited in any way.

[0029] The inventory management system 6 consists of one or more computers and manages the inventory status of pharmaceuticals in a pharmacy. For example, the inventory management system 6 can record and update information on the receipt and dispatch of pharmaceuticals, and provide real-time inventory tracking, reorder point management, and automatic ordering functions. In this embodiment, the inventory management system 6 provides pharmaceutical inventory information to the dispensing claims server 10. The specific functions of the inventory management system 6 are not limited in any way.

[0030] The electronic medical record system 7 consists of one or more computers and manages patient-specific medication history information. For example, the electronic medical record system 7 stores medication history data including prescription information, medication guidance history, side effect history, allergy information, etc. The medication history data stored in the electronic medical record system 7 may be viewable or updateable via the pharmacy terminal 3. In this embodiment, the electronic medical record system 7 provides the prescription server 10 with medical record information for each designated patient. The specific functions of the electronic medical record system 7 are not limited in any way.

[0031] The pharmacy system 8 consists of other electronic devices and equipment installed within the pharmacy, and it integrates and coordinates these devices. For example, the pharmacy system 8 includes a medicine bag printing machine, a label printer, and dispensing equipment (such as an automatic tablet packaging machine), and controls these devices. In this embodiment, the specific configuration and specific functions of the pharmacy system 8 are not limited in any way.

[0032] The dispensing claim server 10 corresponds to the prescription input support system according to the embodiment described above. The dispensing claim server 10 acquires prescription information related to prescriptions issued by physicians, and while referring to the adopted drug information and drug inventory information provided by the inventory management system 6, and the patient's medication history information provided by the electronic medical record system 7, it determines the drugs to be dispensed and generates information for various documents to be given to the patient (drug information sheet, dispensing statement, medication record book, etc.) and information for the dispensing claim statement (claim). The dispensing claim server 10 also generates, transmits, and stores the claims.

[0033] The accounting system 5, inventory management system 6, electronic patient record system 7, and in-pharmacy system 8 may be implemented using different computers or devices, or they may be implemented partially or entirely on a common computer. Furthermore, the accounting system 5, inventory management system 6, electronic medical record system 7, in-pharmacy system 8, and prescription claims server 10 may be located in the target pharmacy or may be partially located on the cloud. The following describes the detailed configuration of the prescription claims server 10.

[0034] 《Pharmacy Receipt Server》 The prescription claims server 10 is a computer, and its hardware configuration includes a CPU (Central Processing Unit) 11, memory 12, input / output interface (I / F) 13, communication unit 14, etc. CPU11 refers to what is commonly known as a processor, and in addition to general CPUs, it may also include application-specific integrated circuits (ASICs), DSPs (Digital Signal Processors), GPUs (Graphics Processing Units), etc. Memory 12 includes RAM (Random Access Memory), ROM (Read Only Memory), and auxiliary storage devices (such as hard disks).

[0035] The input / output interface 13 can be connected to user interface devices such as display devices and input devices, which are not shown in the diagram. The display device is a device that displays a screen corresponding to drawing data processed by the CPU 11, such as an LCD (Liquid Crystal Display) or CRT (Cathode Ray Tube) display. The input device is a device that accepts user input, such as a keyboard or mouse. The display device and input device may be integrated and implemented as a touch panel, and they do not need to be connected to the dispensing receipt server 10. The communication unit 14 communicates with other computers via the communication network 2 and exchanges signals with other devices such as printers. Portable recording media may also be connected to the communication unit 14. In this embodiment, the communication unit 14 is connected to the pharmacy terminal 3, accounting system 5, inventory management system 6, electronic medical record system 7, in-pharmacy system 8, etc., via the communication network 2.

[0036] However, the hardware configuration of the prescription claims server 10 is not limited to the example in Figure 1. The prescription claims server 10 may include other hardware elements not shown. Also, the number of each hardware element is not limited to the example in Figure 1. For example, the prescription claims server 10 may have multiple CPUs 11. Furthermore, the prescription claims server 10 may be implemented by multiple computers consisting of multiple enclosures.

[0037] Figure 2 is a conceptual diagram showing an example of the software configuration of the dispensing claims server 10. As shown in Figure 2, the dispensing claim server 10 includes an input processing unit 100, a dispensing master information database (DB) 110, a linkage processing unit 120, a claim processing unit 130, and the like. Each of these processing modules is implemented, for example, by the CPU 11 executing a computer program stored in memory 12. This computer program may be installed, for example, from a portable recording medium such as a CD (Compact Disc) or memory card, or from another computer on a network, via an input / output interface 13 or a communication unit 14, and stored in memory 12.

[0038] The input processing unit 100 performs the main functions of the prescription input support system according to this embodiment described above, namely generating prescription input data by executing processes to support prescription input operations. Details of the input processing unit 100 will be described later.

[0039] The dispensing master information DB110 stores master information for each drug, comment, and administration method. Each piece of master information may be standardized for each pharmacy, or it may be standardized for all pharmacies. The drug master data is the master data for drugs listed in the drug price list, and it includes the drug code, name, specifications, various classifications, etc., for each drug. Comment master information is the master information for comments to be written in the summary column of the medical claim form (summary comments) and comments to be written in the prescription column of the medical claim form (prescription comments). For each comment, the comment master information shows the comment type (summary comment or prescription comment) and text (string) data. The text data of the comments shown in the comment master information includes fixed text data and the symbol "*" which indicates the variable part, such as "* is ground in a dispensing mill and dispensed with lactose excipient in 1 packet *g" and "* sheets * days' supply per day". The dosage method master information is master information that shows the name of the dosage method and the details of the dosage method for each dosage method. For example, the dosage method name is set as text (string) data such as "Once a day after breakfast" or "Apply to *** once a day", and the dosage method details show the time of administration, number of doses, etc.

[0040] The integration processing unit 120 performs integration processing with the accounting system 5, inventory management system 6, electronic medical record system 7, and in-pharmacy system 8 within the pharmacy business support system 1. Specifically, the integration processing unit 120 obtains information necessary for the input processing unit 100 from other systems and provides the information generated by the input processing unit 100 to the other systems. For example, the integration processing unit 120 extracts desired medical record information from the electronic medical record system 7 and extracts drug inventory information from the inventory management system 6. In addition, the integration processing unit 120 updates the medical record information by sending the latest prescription information to the electronic medical record system 7, sends accounting information such as dispensing fee points to the accounting system 5, and sends information necessary for printing medicine bags to the in-pharmacy system 8.

[0041] The claims processing unit 130 stores information generated by the input processing unit 100 (such as prescription input data and dispensing fee points) for each patient, and executes a series of claims processing related to dispensing fee claims based on the stored data. The claims processing is executed at predetermined intervals (e.g., once a month), and claim data in accordance with a predetermined format for medical fee claims is generated and transmitted. In this embodiment, the claims processing unit 130 only needs to execute known claims processing, and the specific processing content is not limited.

[0042] [Input Processing Unit] The input processing unit 100 includes a prescription acquisition unit 101, a drug selection unit 102, an inference processing unit 103, a correction processing unit 104, an output processing unit 105, etc., and generates prescription input data. The input processing unit 100 generates prescription input data based on the information obtained by the prescription acquisition unit 101, the drug selection unit 102, the inference processing unit 103, and the correction processing unit 104. At this time, the input processing unit 100 further performs calculations of dispensing fee points, etc. The prescription input data thus generated is displayed on the pharmacy terminal 3 by the output processing unit 105, as will be described in detail later, and is confirmed after being checked by the pharmacist.

[0043] The prescription acquisition unit 101 acquires prescription information that includes information about the target patient and the prescription details of the target prescription, which are information about the target patient issued for the target patient. In other words, the prescription acquisition unit 101 corresponds to the prescription acquisition means described above. The contents of the prescription information obtained are as described above and are not limited to those described above, but to add to that, they mainly include patient information, prescription issuance information, and prescription details. Examples of patient information include the patient's name, date of birth, gender, insurer number, and insured person's card symbol / number. Examples of prescription issuance information include the name of the physician who issued the prescription, the name, address, and telephone number of the medical institution, the date the prescription was issued, and the prescription's validity period. Examples of prescription contents include the prescription remarks section, dosage form information (oral, as needed, topical, etc.), dispensing quantity (number of days of administration, number of administrations, total amount administered, etc.), administration method (three times a day after each meal, etc.), supplementary information on administration method (supplements and details on administration method (in the shoulder, in the right eye, two drops at a time, every other day, etc.), dispensing instructions (unit-dose packaging, crushing, etc.), etc.), drug information (drug name (brand name, generic name), dosage / unit (daily dose and its unit (3 capsules, etc.)), single dose, etc.), supplementary drug information (supplements (one tablet in the morning, one tablet at noon, two tablets at night, etc.), dispensing instructions (generic drug substitution not permitted, patient requests original drug, etc.)), etc.). The prescription acquisition unit 101 acquires prescription information from prescriptions issued for the target patient, the target patient's health insurance card or medical record, initial questionnaire, medication record book, etc. The method for acquiring prescription information by the prescription acquisition unit 101 is as described above.

[0044] The drug selection unit 102 selects candidate drugs to be dispensed for the prescription content of the target prescription based on the necessary information in the prescription information obtained by the prescription acquisition unit 101. The drugs selected here are referred to as "candidate drugs" because they will be finalized after confirmation by a pharmacist. Furthermore, the prescription information used by the drug selection department 102 includes patient information and prescription details. Patient information includes, for example, the patient's date of birth, and prescription details include, for example, information in the prescription remarks section, dosage form information, dispensing quantity, administration method information, supplementary administration method information, drug information, and supplementary drug information. If the prescription in question includes the names of drugs used by the pharmacy, the drug candidates should be selected based on the prescription details. However, if the prescription includes the generic name or active ingredient name of a drug, or if it includes the name of a drug not used by the pharmacy, the drug selection unit 102 will select drug candidates to be dispensed that correspond to the prescription details of the prescription.

[0045] The drug selection unit 102 may further refer to patient dispensing preference information, past prescription information, and drug information adopted by the pharmacy, in addition to the prescription information, to select drug candidates. Patient dispensing preference information includes information about the dispensing method and form preferred by the patient, such as preferences for unit-dose packaging, halving tablets, or crushing, as well as preferences for brand-name or generic drugs. Past prescription information refers to the dispensing history of the target patient, including details such as unit-dose packaging, crushing, splitting tablets in half, and mixing. If the target patient is a new patient, dispensing history information from other patients may be used. The drug information indicates the drugs stocked at the pharmacy, showing the drug code and supply status for each drug. The supply status indicated in the drug information shows whether the drug is out of supply due to manufacturing delays, etc. If a drug is out of supply, it may be necessary to prepare the drug. Patient dispensing preference information can be obtained and retained through initial questionnaires given to patients or by inquiring about their preferences at that time. Past prescription information can be extracted based on medication history information obtained from the electronic medical record system 7. Adopted drug information can be extracted from drug master information stored in the dispensing master information DB 110 and inventory information obtained from the inventory management system 6.

[0046] Based on this information, the drug selection unit 102 selects candidate drugs to be dispensed by converting generic names to adopted drugs and converting original drugs to generic drugs in the prescription details of the target prescription. However, the selection of candidate drugs by the drug selection unit 102 can be achieved using known methods such as those disclosed in the above-mentioned Patent Document 2, and the specific selection method is not limited.

[0047] The inference processing unit 103 obtains various information as prescription input data by having an inference model, which includes a pre-trained machine learning model, perform inference. In this specific example, the inference model includes a dispensing fee inference model, a prescription order classification model, and a comment generation model, each containing a pre-trained machine learning model. Therefore, the inference processing unit 103 corresponds to the inference means described above, and as described above, by inputting prescription information into the inference model, it obtains dispensing fee surcharge item information or comment information for dispensing fee statements (receipts) that indicate at least the drug preparation fee surcharge items or pharmaceutical management fee items applicable to the target prescription. In this specific example, the inference processing unit 103 can obtain both the dispensing fee surcharge item information and the comment information.

[0048] More specifically, the inference processing unit 103 obtains the dispensing fee addition item information by inputting at least the necessary information from the prescription information obtained by the prescription acquisition unit 101 and the information of the drug candidate selected by the drug selection unit 102 into the machine learning model of the dispensing fee inference model. The dispensing fee addition item information obtained in this case includes information on drug preparation fee addition items and specific items within pharmaceutical management fees, which can be calculated based on the dispensing activities of the pharmacy. For this reason, the inference processing unit 103 includes dispensing fee inference means. The prescription information acquired by the prescription acquisition unit 101 includes drug information described in the target prescription, and among the drug candidates selected by the drug selection unit 102, there may be drugs with names different from the drug names indicated in the prescription details of the target prescription. Therefore, by using the drug information included in the original prescription details and the converted drug information as input information for the machine learning model of the dispensing fee inference model, the estimation accuracy of the dispensing fee addition item information in the machine learning model can be improved.

[0049] Furthermore, in this specific example, the machine learning model of the dispensing fee inference model may be constructed to output dispensing fee addition item information that indicates applicable drug preparation fee addition items and pharmaceutical management fee items for the target prescription, in addition to the necessary information in the prescription information acquired by the prescription acquisition unit 101 and the drug candidate information selected by the drug selection unit 102, as well as patient dispensing preference information, past prescription information, and drug master information from the dispensing master information DB 110. In this case, the machine learning model of the dispensing fee inference model may be constructed to output information on compounding fee addition items (oral medications), (as-needed medications), (topical medications) and weighing and mixing dispensing addition items within the drug preparation fee addition items, and information on specific items that can be calculated based on the dispensing activities of the pharmacy within the pharmaceutical management fee items (outpatient medication support fee 2, etc.). The prescription information required as input to the machine learning model includes patient information and prescription details. Patient information may include, for example, the patient's date of birth, and prescription details may include, for example, information in the prescription remarks section, dosage form information, administration instructions, supplementary administration instructions, drug information, and supplementary drug information.

[0050] By using machine learning models to estimate information on additional dispensing fee items such as the additional fee for compounded preparations, the additional fee for weighed and mixed dispensing, and the outpatient medication support fee 2, it is possible to estimate the items necessary for dispensing fees even if they are not explicitly stated on the prescription. For example, regarding a prescription that reads "XX ointment 0.05% 15.0g, [generic] XX ointment 0.3% 15.0g, apply 2-3 times a day, generic substitution not permitted, mix above, for itchy areas, hands," even if the instruction "mixing and preparing" is not explicitly stated, it is possible to infer the "weighing and mixing preparation fee (soft and hard ointments)" as a weighing and mixing dispensing fee item. Furthermore, regarding a prescription that states "[General] XX Syrup 50% 1.1g, [General] XX 0.6% 1.8g 3 times a day (after each meal)", even if no dispensing instructions are written, it is possible to infer the "Measurement and Mixing Preparation Fee (Powder / Granules)" as a charge for measurement and mixing preparation. Furthermore, regarding a prescription that states "[General] XX Tablets 30mg 0.5 tablets once a day after breakfast," even if no dispensing instructions are written, it is possible to infer the compounding surcharge item, "Compounding Surcharge (Divided) (Oral Medication) (Tablets, etc.)." Furthermore, for prescriptions that list the quantities of multiple tablets and state "Dose once a day after breakfast, packaged in one dose," it is possible to infer that the "Outpatient Medication Support Fee 2 (for 43 days or more)" item applies.

[0051] The machine learning model for this dispensing fee inference model is trained using training data consisting of pairs of input information and correct output information. However, the specific machine learning method and model configuration are not limited. Furthermore, as mentioned above, the dispensing fee inference model may include a rule-based model in addition to the machine learning model. For example, rule-based models may be used for dispensing fee items where the number of applicable cases is small, as it may be difficult to collect training data and sufficient learning may not be possible. For example, among the drug preparation fee surcharge items, items such as sterile preparation processing surcharges and narcotics surcharges, as well as some surcharge items among pharmaceutical management fee items, may be inferred using the rule-based model of the dispensing fee inference model.

[0052] The comment generation model included in the inference model may include a trained machine learning model and may consist solely of that machine learning model, or it may consist of a machine learning model and a rule-based model. The comment generation model only needs to be able to output at least summary comment information as comment information for medical claims, and may be constructed to be able to output both summary comments and prescription comments.

[0053] In this specific example, the comment generation model, which includes a trained machine learning model, is constructed to output summary comment information for the claim form related to the target prescription, using necessary information from the prescription information acquired by the prescription acquisition unit 101, information on drug candidates selected by the drug selection unit 102, and information on additional dispensing fee items obtained using the aforementioned dispensing fee inference model as input information. The inference processing unit 103 then acquires summary comment information for the claim form related to the target prescription by inputting at least the prescription information, drug candidate information, and dispensing fee additional item information into this comment generation model. In other words, the inference processing unit 103 also includes a comment acquisition means. In this case, the prescription information input to the comment generation model only needs to include drug information, and it is more preferable that it includes patient information and prescription details. Specifically, patient information may include, for example, the patient's date of birth, and prescription details may include, for example, information in the prescription remarks section, dosage form information, administration information, supplementary administration information, drug information, and supplementary drug information.

[0054] Here, "summary comments" refer to comments written in the summary section of the medical claim form, as described above. There are two types: those for which a combination of standard text and a code (a code for the medical claim processing system (sometimes referred to as a medical claim code)) is predetermined by the Ministry of Health, Labour and Welfare (Social Insurance Medical Fee Payment Fund) (hereinafter referred to as summary comments (with code)), and those for which no standard text is predetermined but which must be written in the summary section of the medical claim form (hereinafter referred to as summary comments (without code)). The comment generation model described above can be configured to output summary comment information that includes either summary comment (with code) or summary comment (without code), or both. Summary comments (with code) can be inferred to include those related to drug preparation fees (oral medications), compounding fees, specific drug management guidance fees 2, inhalation drug guidance fees, post-dispensing drug management guidance fees 1, post-dispensing drug management guidance fees 2, and situations where dispensing is performed based on a prescription containing more than 63 patches with analgesic and anti-inflammatory efficacy. Summary comments (without code) can be inferred to include those related to outpatient medication support fees 2, intended use such as gargles and nicotine dependence, and PPI-related matters.

[0055] The comment generation model may be configured to output summary comment information including the text data of the summary comment and the receipt code assigned to it for summary comments (with a code), and to output summary comment information including the text data of the summary comment and a specific code indicating that there is no receipt code for summary comments (without a code). In this case, the inference processing unit 103 obtains summary comment information in which the text data of the summary comment and the receipt code corresponding to that summary comment are associated for summary comments (with a code) that are predetermined and assigned to a receipt code. For example, regarding summary comments (with codes) related to the additional charge for compounded medications, it is possible to obtain summary comment information associated with the claim code "830100438" and the standard comment text "Reason for calculation (additional charge for compounded medications);******", summary comment information associated with the claim code "830100908" and the standard comment text "Name of drug for which the necessary quantity for dispensing could not be secured (additional charge for compounded medications);******", and summary comment information associated with the claim code "820101255" and the standard comment text "Unavoidable circumstances for which the necessary quantity for dispensing could not be secured (additional charge for compounded medications); Problems with the supply of pharmaceuticals".

[0056] Furthermore, separate comment generation models may be provided for generating summary comments (with codes) and for generating summary comments (without codes). In this case, the former comment generation model may be constructed to output summary comment information including the text data of the summary comment (with code) and the corresponding receipt code, while the latter comment generation model may be constructed to output summary comment information including only the text data of the summary comment (without a code). Furthermore, if the comment generation model consists of a machine learning model and a rule-based model, some of the summary comments (with codes) may be generated by the machine learning model and the rest by the rule-based model, and some of the summary comments (without codes) may be generated by the machine learning model and the rest by the rule-based model. For example, summary comment information related to the self-compounding fee or when dispensing is performed based on a prescription that specifies more than 63 patches of transdermal patches with analgesic and anti-inflammatory effects may be generated by the machine learning model, while summary comment information related to the drug adjustment fee (oral medication) or the specific drug management guidance fee 2 may be generated by the rule-based model. Also, summary comment information related to the purpose of use for nicotine dependence may be generated by the machine learning model, while summary comment information related to the outpatient medication support fee 2 or PPI-related summary comments (without codes) may be generated by the rule-based model.

[0057] In this specific example, the comment generation model may be constructed to include a separate model for generating prescription comment information. The model for generating prescription comment information may be constructed, for example, as a rule-based model, and may take prescription information, as well as the dosage master information and comment master information from the dispensing master information DB110, as input information, and output information on dosage comments, which are prescription comments. This model for generating prescription comment information may, for example, perform a process to standardize dosages using a rule-based approach and a process to generate information on dosage comments. Note that the standardization of dosages and the generation of dosage comments may be performed using a machine learning model.

[0058] The prescription instruction classification model included in the inference model is a trained machine learning model that takes prescription information as input and is constructed to output prescription instruction information indicating whether or not there is an instruction to package the medication in unit doses, split the tablet in half, or crush the medication for each prescription included in the prescription contents of the target prescription. The inference processing unit 103 obtains prescription instruction information indicating whether or not there is an instruction to package the medication in unit doses, split the tablet in half, or crush the medication for each prescription included in the prescription contents of the target prescription by inputting at least prescription information into such a prescription instruction classification model. In other words, the inference processing unit 103 also includes a means for obtaining prescription instructions.

[0059] The prescription information entered into the prescription instruction classification model only needs to include the prescription details, such as information in the prescription remarks section, information on how to take the medication, supplementary information on how to take the medication, and supplementary information on the medication. Furthermore, the prescription instruction information output from the prescription instruction classification model may indicate, for each prescription, one or more of the following: whether or not there is an instruction to package the tablets together, whether or not there is an instruction to split the tablets in half, or whether or not there is an instruction to crush the tablets. In addition, it may also indicate, for the prescription as a whole, one or more of the following: whether or not there is an instruction to package the tablets together, whether or not there is an instruction to split the tablets in half, or whether or not there is an instruction to crush the tablets. By using prescription instruction information obtained from this prescription instruction classification model, it is possible to obtain dispensing instructions without missing any, which may be overlooked if only the information written on the prescription is considered, thereby reducing dispensing errors.

[0060] Furthermore, the inference models used by the inference processing unit 103 (pharmacy reimbursement inference model, prescription instruction classification model, and comment generation model) may, in addition to the input information described above, also take necessary master information from the various master information in the dispensing master information DB 110 as input information, or may be constructed to hold such master information. For example, the comment generation model for summary comments may take drug master information and comment master information from the dispensing master information DB 110 as input information, or may be constructed to hold such master information. In addition, the machine learning model for the comment generation model for summary comments (with codes) may take necessary information from prescription information, drug candidate information, and dispensing reimbursement addition item information, as well as drug information adopted by the pharmacy, drug master information and comment master information from the dispensing master information DB 110 as input information.

[0061] The correction processing unit 104 applies correction processing to the information obtained by the inference processing unit 103. Because the inference processing unit 103 uses a machine learning model for the inference model, it may sometimes obtain information that is not optimal. For example, the dispensing fee addition item information output from the machine learning model of the dispensing fee inference model may include items that cannot be calculated together, or it may include the item with the lower fee point among items that cannot be calculated together. In such cases, the summary comment generated by the comment generation model using such dispensing fee addition item information as input information may also be an inappropriate summary comment. Therefore, the correction processing unit 104 performs correction processing on the dispensing fee addition item information, such as deleting items that cannot be calculated together and replacing them with items that have higher fee points. It also performs correction processing on the summary comment information, such as deleting summary comments corresponding to deleted items and changing summary comments corresponding to changed items. Furthermore, the correction processing unit 104 may perform logic-based processing to add dispensing fee addition items, summary comments, and prescription comments that are not targeted for acquisition by the inference processing unit 103.

[0062] The output processing unit 105 corresponds to the output processing means described above and outputs at least the drug information to be dispensed corresponding to the prescription information obtained by the prescription acquisition unit 101 and the dispensing fee addition item information obtained by the inference processing unit 103. Furthermore, the output processing unit 105 can output not only information on drug candidates selected by the drug selection unit 102 and information on dispensing fee addition items acquired by the inference processing unit 103, but also summary comment information acquired by the inference processing unit 103. Furthermore, the output processing unit 105 can also output prescription instruction information acquired by the inference processing unit 103. The output format of the output processing unit 105 is not limited to the output processing means as described above.

[0063] In this specific example, the output processing unit 105 displays the information obtained through the processing of the prescription acquisition unit 101, the drug selection unit 102, the inference processing unit 103, and the correction processing unit 104 on the display unit of the pharmacy terminal 3 in a modifiable manner. That is, the output processing unit 105 displays an input confirmation screen for the target prescription on the display unit (the display unit of the pharmacy terminal 3). The pharmacist checks this display on pharmacy terminal 3, makes changes as necessary, and finalizes the prescription input data.

[0064] The input confirmation screen includes a first display area that displays prescription information acquired by the prescription acquisition unit 101, and a second display area that displays prescription input information including at least information on drug candidates selected by the drug selection unit 102 and summary comment information acquired by the inference processing unit 103. The first and second display areas are separated and arranged so as to be comparable to each other, and the second display area can be divided into prescription units included in the prescription content. This type of input confirmation screen allows pharmacists to easily compare the original prescription details written on the issued prescription with the information on drug candidates and summary comments automatically selected and retrieved based on that information, making it easier to verify the appropriateness of the automatically selected and retrieved information.

[0065] Furthermore, the output processing unit 105 can also display an operation unit that allows the user to select whether or not to adopt each of the drug candidate information and acquired summary comment information, which are displayed in the second display area, separated by prescription unit. For example, if the user selects not to adopt on the input confirmation screen, the original prescription content may be restored, or the information may be deleted. This approach makes it easy to reject candidates if the automatically selected and acquired information is inappropriate.

[0066] Figure 3 shows an example of an input confirmation screen. The input confirmation screen illustrated in Figure 3 includes a display area GA1 that displays information about the target patient, a display area GA2 that displays information about the issuer of the target prescription, a display area GA3 that displays information about the prescription content of the target prescription (prescription content of the prescription information acquired by the prescription acquisition unit 101), a display area GA4 that displays information about the prescription input data generated by the input processing unit 100, and a display area GA5 that displays the dispensing fee items and dispensing fee points. Each display area is divided in a way that allows for comparison.

[0067] Furthermore, the display area GA4 is divided into prescription units, and has an operation section that allows the user to select whether or not to adopt the drug candidate information and summary comment information displayed for each prescription unit. In the example in Figure 3, the display area GA4 is divided into prescription units and row units, and each row has its own operation section. The control unit includes a checkbox GS1 and a change button GS2, which can be toggled by the user to indicate whether or not to adopt the information. By unchecking checkbox GS1 (not adopting the information) and pressing the change button GS2, the information corresponding to that row is changed.

[0068] Furthermore, in the example in Figure 3, the history of the patient's prescriptions for the current, previous, and the one before that is shown. Display area GA10 displays information corresponding to the prescription before last, display area GA11 displays information corresponding to the previous prescription, and display area GA12 displays information corresponding to the current prescription. Then, the information in each row of display area GA4 in display area GA12 corresponding to the current prescription is compared row by row with the information in each row of display area GA4 in display area GA11 corresponding to the previous prescription and display area GA10 corresponding to the prescription before that. If there is a discrepancy in the information in display area GA12, each row of display area GA4 is highlighted (colored). This allows pharmacists to easily identify differences between the prescription input data generated for the current prescription and the previous or the one before that by viewing the input confirmation screen, thereby reducing dispensing errors.

[0069] Furthermore, in the display area GA3, which displays information about the prescription details of the target prescription, the display may be configured to show information based on prescription instruction information obtained using a prescription instruction classification model. For example, in each prescription displayed in the display area GA3, if the prescription instruction information indicates that there is an instruction, it may be highlighted with "Unit-dose packaging instruction," "Half-tablet splitting instruction," or "Crushing instruction." This helps prevent prescriptions from being overlooked and reduces dispensing errors.

[0070] <Methods for supporting prescription data entry tasks> The details of the prescription input support method according to this embodiment (hereinafter sometimes referred to as "this support method") will be explained below with reference to Figure 4. Figure 4 is a flowchart conceptually illustrating the prescription input support method (this support method) according to this embodiment. This support method is executed by the dispensing claim server 10 described above. This support method is realized by the CPU 11 executing a computer program stored in memory 12 on the dispensing claim server 10. This computer program is installed on the dispensing claim server 10 via input / output I / F 13 or communication unit 14 from a portable recording medium such as a CD (Compact Disc) or memory card, or from another computer on the network, and stored in memory 12.

[0071] The entity executing each step of this support method is the dispensing claim server (hereinafter sometimes abbreviated as "this server") 10. However, the entity executing each step can also be described as the CPU 11 provided by the dispensing claim server 10, the input processing unit 100 implemented by the dispensing claim server 10, or each processing module included in the input processing unit 100. Each step of the support method described below includes the same processing content as the aforementioned processing modules included in the input processing unit 100 of the dispensing receipt server 10; therefore, content that is the same as described above will be omitted as appropriate.

[0072] This server 10 (prescription acquisition unit 101) acquires prescription information that includes information about the target patient and the prescription details of the target prescription, which is information about the target patient (S41). Details of the acquired prescription information are as described above. The server 10 may acquire prescription information by applying specific processing such as OCR to prescription images obtained by scanning or photographing paper prescriptions, or it may acquire prescription information from electronic prescription data received from a medical institution's system, etc. Thus, the method of acquiring prescription information in step (S41) is not limited.

[0073] Next, the server 10 (drug selection unit 102) selects candidate drugs to be dispensed for the prescription content of the target prescription (S42) based on the necessary information in the prescription information acquired in step (S41). The server 10 can select candidate drugs from the prescription information acquired in step (S41) alone, or it may further refer to patient dispensing preference information, past prescription information, and the pharmacy's adopted drug information in addition to the said prescription information to select candidate drugs. The server 10 selects candidate drugs to be dispensed by performing conversions such as converting generic names to adopted drugs and converting original drugs to generic drugs in the prescription content of the target prescription. However, the method for selecting candidate drugs in step (S42) can be implemented using known methods, and the specific selection method is not limited.

[0074] Next, the server 10 (inference processing unit 103) acquires prescription instruction information using the prescription instruction classification model included in the inference model (S43). The prescription instruction classification model is a trained machine learning model, and the server 10 acquires prescription instruction information indicating whether or not there is an instruction to package the medication into a single dose, an instruction to split the tablet in half, or an instruction to crush the medication for each prescription included in the prescription content of the target prescription, by inputting at least prescription information into this prescription instruction classification model. At this time, the prescription information input into the prescription instruction classification model only needs to include the prescription content, and may include, for example, information in the prescription remarks column, information on how to take the medication, supplementary information on how to take the medication, and supplementary information on the medication.

[0075] Furthermore, the server 10 (inference processing unit 103) acquires information on additional dispensing fees using the dispensing fee inference model included in the inference model (S44). Details of the dispensing fee inference model and the additional dispensing fee information are as described above. For example, the server 10 obtains the dispensing fee addition item information by inputting at least the necessary information from the prescription information acquired in process (S41) and the information of the drug candidate selected in process (S42) into the trained machine learning model in the dispensing fee inference model. Alternatively, if the dispensing fee inference model includes a rule-based model in addition to the machine learning model, the server 10 may obtain a portion of the dispensing fee addition item information by inputting at least the necessary information from the prescription information acquired in process (S41) into the rule-based model.

[0076] Furthermore, the server 10 (inference processing unit 103) generates summary comment information and prescription comment information using the comment generation model included in the inference model (S45). Details of the comment generation model and the summary comment information and prescription comment information are as described above. For example, the server 10 obtains summary comment information for the dispensing fee statement (receipt) related to the target prescription by inputting at least the necessary information from the prescription information obtained in step (S41), the information of the drug candidate selected in step (S42), and the dispensing fee addition item information obtained in step (S44) into the comment generation model. Furthermore, the server 10 obtains prescription comment information indicating a dosage comment by inputting the necessary information from the prescription information obtained in step (S41), the dosage master information, and the comment master information into the model that generates prescription comment information included in the comment generation model.

[0077] As mentioned above, summary comments include summary comments (with code) and summary comments (without code). Therefore, it is preferable that the summary comment information obtained in process (S44) includes information on both summary comments (with code) and summary comments (without code), but it may also include only one of the two. Furthermore, for summary comments (with code), it is preferable that summary comment information including the text data of the summary comment and the receipt code assigned to that summary comment is obtained. Furthermore, while Figure 4 shows an example in which both summary comment information and prescription comment information are generated in process (S45), it is also possible to generate only one of them.

[0078] Subsequently, the server 10 (correction processing unit 104) applies a correction process (S46) to the dispensing fee addition item information acquired in step (S44) and the summary comment information and prescription comment information generated in step (S45). The details of this correction process are as described above. Note that the dispensing fee addition item information acquired in step (S44) and the summary comment information and prescription comment information generated in step (S45) may be correct, in which case the correction process in step (S46) is unnecessary.

[0079] The server 10 (output processing unit 105) outputs the information obtained through the above process, namely prescription information, drug candidate information, prescription instruction information, dispensing fee addition item information, summary comment information, and prescription comment information, as prescription input data. Although there are various output formats, in the example shown in Figure 4, the server 10 (output processing unit 105) displays an input confirmation screen on the display unit of the pharmacy terminal 3 that presents this prescription input data in a modifiable format (S47). Details of this input confirmation screen are as described above.

[0080] The server 10 detects whether or not a change operation has been performed on the input confirmation screen at the pharmacy terminal 3 (S48). If no change operation has been performed (S48; NO), the server confirms the prescription input data. On the other hand, if a change operation is performed (S48; YES), the server 10 determines whether reinference is necessary depending on the content of the change (S49). For example, if the selected drug candidate is changed, it may be determined that reinference is necessary. On the other hand, if an item indicated by the dispensing fee addition item information is not adopted and is simply deleted, or if a comment indicated by the summary comment information and prescription comment information is adopted and simply deleted, it may be determined that reinference is not necessary.

[0081] If server 10 determines that reinference is necessary (S49; YES), it re-executes the process from (S44) onward based on the changed information. If it determines that reinference is not necessary (S49; NO), it finalizes the prescription input data with the changes reflected. The server 10 (cooperation processing unit 120) sends the necessary information to the accounting system 5, the electronic medical record system 7, the in-pharmacy system 8, etc., based on the prescription input data confirmed as described above. As a result, the in-pharmacy system 8 performs tasks such as printing medicine bags and dispensing using dispensing equipment, the electronic medical record system 7 updates the medical record information of the target patient, and the accounting system 5 performs accounting processing for the target patient. Furthermore, this server 10 (receipt processing unit 130) stores prescription input data and dispensing fee points for each patient, and executes a series of receipt processing related to dispensing fee claims based on the stored data.

[0082] [Differentiation] The above-described embodiment is merely an example and may be partially modified as appropriate. For example, in the specific example described above, the inference model used by the inference processing unit 103 included a dispensing fee inference model, a prescription order classification model, and a comment generation model, but it may include only one or two of these.

[0083] Furthermore, the inference model may be stored in the memory 12 of the dispensing receipt server 10 where the inference processing unit 103 is implemented, or it may be provided on another computer. In the latter case, the inference processing unit 103 receives output information from the inference model by sending input information to the other computer where the inference model is stored.

[0084] Furthermore, although Figure 4 shows multiple processes in sequence, the execution order of each process in this support method is not limited to the example in Figure 4. For example, process (S43) may be executed in parallel with processes (S44) and (S45), or it may be executed after those processes.

[0085] Some or all of the above embodiments and modifications may also be specified as follows; however, the above embodiments and modifications are not limited to those described below.

[0086] <1> A prescription acquisition means for acquiring prescription information that includes information about the target patient and the prescription details of the target prescription, which is information about the target patient issued for the target patient. A dispensing fee inference means that inputs at least the acquired prescription information into an inference model including a trained machine learning model to acquire dispensing fee additional item information that indicates at least the drug preparation fee additional item or pharmaceutical management fee item applicable to the target prescription, or comment information for the dispensing fee statement related to the target prescription, An output processing means that outputs information on the medication to be dispensed corresponding to the acquired prescription information and information on the acquired dispensing fee addition items or comment information, A prescription entry support system equipped with [features / equipment]. <2> A drug selection means for selecting candidate drugs to be dispensed for the prescription contents of the target prescription based on the acquired prescription information, Furthermore, Among the selected drug candidates, there may be drugs with names different from the drug names indicated by the prescription details of the target prescription included in the prescription information. The aforementioned inference model includes a dispensing fee inference model that includes a trained machine learning model, The inference means includes a dispensing fee inference means that acquires the dispensing fee addition item information by inputting at least the acquired prescription information and the selected drug candidate information into the machine learning model of the dispensing fee inference model. <1> The prescription entry support system described above. <3> The aforementioned inference model further includes a comment generation model which includes a pre-trained machine learning model, The inference means further includes a comment acquisition means that acquires summary comment information for dispensing fee statements as comment information relating to the target prescription by inputting at least the acquired prescription information, the selected drug candidate information and the acquired dispensing fee addition item information into the comment generation model, The output processing means outputs the information of the selected drug candidates as drug information to be dispensed, the acquired dispensing fee addition item information, and the acquired summary comment information. <2> The prescription entry support system described above. <4> The aforementioned comment generation model is capable of outputting text data of summary comments and a receipt code corresponding to the summary comments predetermined by the Ministry of Health, Labour and Welfare. The comment acquisition means acquires summary comment information, which is a combination of the text data of the summary comment and the corresponding receipt code, for summary comments that are predetermined and associated with the receipt code. <3> The prescription entry support system described above. <5> The output processing means displays the input confirmation screen for the target prescription on the display unit. The input confirmation screen includes a first display area that displays the prescription details information of the acquired prescription information, and a second display area that displays prescription input information that includes at least the information of the selected drug candidates and the acquired summary comment information. The first display area and the second display area are separated and arranged in a way that allows for comparison with each other. In the second display area, the prescription units included in the prescription content are divided as follows: <3> or <4> The prescription entry support system described above. <6> The output processing means displays an operation unit that allows the user to select whether or not to adopt each of the drug candidate information and the acquired summary comment information, which are displayed in the second display area, divided into prescription units. <5> The prescription entry support system described above. <7> A prescription instruction acquisition means that inputs at least the acquired prescription information into a prescription instruction classification model, which is a pre-trained machine learning model, to acquire prescription instruction information indicating whether or not there is an instruction to package the prescriptions together, split the tablets in half, or crush them, for each prescription included in the prescription content of the target prescription. Furthermore, The output processing means further outputs the acquired prescription instruction information. <1> from <6> A prescription entry support system described in any one of the following. <8> A method for supporting prescription data entry tasks, which is performed on one or more computers, The aforementioned one or more computers A step of obtaining prescription information that includes information about the target patient and the prescription details of the target prescription, which is information about the target patient issued for the target patient. An inference step of inputting at least the acquired prescription information into an inference model including a trained machine learning model to acquire information on additional dispensing fees that indicate at least the applicable drug preparation fee additional items or pharmaceutical management fee items for the target prescription, or comment information for the dispensing fee statement for the target prescription, An output process that outputs information on the medication to be dispensed corresponding to the acquired prescription information and information on the acquired dispensing fee addition items or comment information, A method for supporting prescription entry tasks, which includes performing the following actions. <9> The aforementioned one or more computers A step of selecting candidate drugs to be dispensed for the prescription content of the target prescription based on the acquired prescription information, This further includes performing, Among the selected drug candidates, there may be drugs with names different from the drug names indicated by the prescription details of the target prescription included in the prescription information. The aforementioned inference model includes a dispensing fee inference model that includes a trained machine learning model, In the inference process, one or more computers acquire the dispensing fee addition item information by inputting at least the acquired prescription information and the selected drug candidate information into the machine learning model of the dispensing fee inference model. <8> The prescription entry support method described below. <10> The aforementioned inference model further includes a comment generation model which includes a pre-trained machine learning model, The inference process is performed by one or more computers, The process further includes inputting at least the acquired prescription information, the selected drug candidate information, and the acquired dispensing fee addition item information into the comment generation model to perform a comment acquisition step to acquire summary comment information for dispensing fee statements as comment information for the target prescription, In the output step, one or more computers output the information of the selected drug candidates as drug information to be dispensed, the acquired dispensing fee addition item information, and the acquired summary comment information. <9> The prescription entry support method described below. <11> The aforementioned comment generation model is capable of outputting text data of summary comments and a receipt code corresponding to the summary comments predetermined by the Ministry of Health, Labour and Welfare. In the comment acquisition process, one or more computers acquire summary comment information, which is a combination of the text data of the summary comment and the corresponding receipt code, for summary comments that are predetermined and associated with the receipt code. <10> The prescription entry support method described below. <12> In the output process, one or more computers display an input confirmation screen for the target prescription on the display unit. The input confirmation screen includes a first display area that displays the prescription details information of the acquired prescription information, and a second display area that displays prescription input information that includes at least the information of the selected drug candidates and the acquired summary comment information. The first display area and the second display area are separated and arranged in a way that allows for comparison with each other. In the second display area, the prescription units included in the prescription content are divided as follows: <10> or <11> The prescription entry support method described below. <13> In the output process, one or more computers display an operation unit that allows the user to select whether or not to adopt each of the drug candidate information and the acquired summary comment information, which are displayed in the second display area, divided into prescription units. <12> The prescription entry support method described below. <14> The aforementioned one or more computers A step of obtaining prescription instruction information indicating whether or not there is an instruction to package the prescription in a single dose, split the tablet in half, or crush it, by inputting at least the acquired prescription information into a prescription instruction classification model, which is a pre-trained machine learning model, This further includes performing, In the output step, one or more computers further output the acquired prescription instruction information. <8> from <13> A prescription entry support method described in any one of the following. <15> <8> from <14> A computer program capable of causing one or more computers to execute the prescription input support method described in any one of the above. [Explanation of symbols]

[0087] 1. Pharmacy Business Support System 2. Communication Network 3. Pharmacy terminal 5. Accounting System 6. Inventory Management System 7. Electronic Medical Record System 8. Pharmacy System 10. Prescription claims server 11 CPU 12 memory 13 Input / Output Interfaces 14 Communication Unit 100 Input Processing Unit 101 Prescription Acquisition Department 102 Pharmaceutical Selection Department 103 Inference Processing Unit 104 Correction Processing Unit 105 Output Processing Unit 110 Dispensing Master Information Database 120 Interoperability Processing Unit 130 Receipt Processing Unit

Claims

1. A prescription acquisition means for acquiring prescription information that includes information about the target patient and the prescription details of the target prescription, which is information about the target patient issued for the target patient. A drug selection means for selecting candidate drugs to be dispensed for the prescription content of the target prescription based on the acquired prescription information, An inference means that inputs at least the acquired prescription information into an inference model including a trained machine learning model to acquire dispensing fee addition item information that indicates at least the drug preparation fee addition item or pharmaceutical management fee item applicable to the target prescription, or comment information for the dispensing fee statement related to the target prescription, An output processing means that outputs information on the selected drug candidates to be dispensed and the acquired information on the additional dispensing fee items or comment information, A prescription entry support system equipped with [features / equipment].

2. Among the selected drug candidates, there may be a drug whose name is different from the drug name indicated by the prescription content of the target prescription included in the prescription information, The aforementioned inference model includes a dispensing fee inference model that includes a trained machine learning model, The inference means includes a dispensing fee inference means that acquires the dispensing fee addition item information by inputting at least the acquired prescription information and the selected drug candidate information into the machine learning model of the dispensing fee inference model. The prescription input support system according to claim 1.

3. The aforementioned inference model further includes a comment generation model which includes a pre-trained machine learning model, The inference means further includes a comment acquisition means that acquires summary comment information for dispensing fee statements as comment information relating to the target prescription by inputting at least the acquired prescription information, the selected drug candidate information and the acquired dispensing fee addition item information into the comment generation model, The output processing means outputs the selected drug candidate information and the acquired dispensing fee addition item information, in addition to the acquired summary comment information. The prescription input support system according to claim 2.

4. The aforementioned comment generation model is capable of outputting text data of summary comments and a receipt code corresponding to the summary comments predetermined by the Ministry of Health, Labour and Welfare. The comment acquisition means acquires summary comment information, which is a combination of the text data of the summary comment and the corresponding receipt code, for summary comments that are predetermined and associated with the receipt code. The prescription input support system according to claim 3.

5. The output processing means displays the input confirmation screen for the target prescription on the display unit. The input confirmation screen includes a first display area that displays the prescription details information of the acquired prescription information, and a second display area that displays prescription input information that includes at least the information of the selected drug candidates and the acquired summary comment information. The first display area and the second display area are separated and arranged in a way that allows for comparison with each other. In the second display area, the prescription units included in the prescription content are divided as follows: The prescription input support system according to claim 3 or 4.

6. The output processing means displays an operation unit that allows the user to select whether or not to adopt each of the drug candidate information and the acquired summary comment information, which are displayed in the second display area, divided into prescription units. The prescription input support system according to claim 5.

7. A prescription instruction acquisition means that inputs at least the acquired prescription information into a prescription instruction classification model, which is a pre-trained machine learning model, to acquire prescription instruction information indicating whether or not there is an instruction to package the prescriptions together, split the tablets in half, or crush them, for each prescription included in the prescription content of the target prescription. Furthermore, The output processing means further outputs the acquired prescription instruction information. The prescription input support system according to claim 1.

8. A method for supporting prescription data entry tasks, which is performed on one or more computers, The aforementioned one or more computers Information regarding prescriptions issued for target patients, including information about the target patient and the prescription details, is obtained. Based on the acquired prescription information, candidate drugs to be dispensed for the prescription contents of the target prescription are selected. By inputting at least the acquired prescription information into an inference model that includes a trained machine learning model, information on additional dispensing fees that indicates at least the applicable drug preparation fee additional items or pharmaceutical management fee items for the target prescription, or comment information for the dispensing fee statement for the target prescription, The system outputs information on the selected drug candidates to be dispensed and the acquired information on additional dispensing fee items or comments. A method for supporting prescription data entry tasks, including the following.

Images