Support information providing system, support information providing device, support information providing method, and program

Abstract

To support the business of pharmacists more appropriately.SOLUTION: In a support information providing system 1, a support request unit 53 of a terminal device 10 sends a support information providing server 30 a request to generate support information for pharmaceutical guidance based on target information which is a consideration target in pharmaceutical guidance for patients and includes contents of pharmaceutical guidance the previous time for the patients and a hearing result this time. A UI display control unit 51 displays the support information transmitted from the support information providing server 30 in response to the request by the support request unit 53. A target information acquisition unit 254 of the support information providing server 30 acquires consideration target information based on the request from the support request unit 53. A problem identification unit 256 generates the support information including contents of recommended pharmaceutical guidance by making inferences based on the consideration target information. A support information providing unit 259 provides the terminal device 10 with the support information generated by the problem identification unit 256.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a support information providing system, a support information providing device, a support information providing method, and a program. [Background technology]

[0002] BACKGROUND ART Conventionally, information processing techniques for supporting the work of pharmacists in pharmacies have been known. For example, Patent Document 1 describes that data on a patient's medication history is managed in a database, and when dispensing medicine to a patient, the database is referenced to provide medication instructions based on the medication history, etc. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-122253 Summary of the Invention [Problem to be solved by the invention]

[0004] However, even if data such as patient medication history is managed in a database, there is a wide range of information that pharmacists must consider when dispensing, and it is not easy for them to appropriately consider this information when providing medication instructions and other tasks. Even if the medication prescribed to a patient remains the same, it would be desirable for the pharmacist to provide appropriate instructions based on the information they obtain from face-to-face interactions with the patient (conversation content, impressions, etc.), but providing such medication instructions requires a great deal of effort. Furthermore, because a patient's symptoms and medication usage patterns differ between their first and second visits, pharmacists are required to provide medication instructions appropriate to the patient's current condition.

[0005] An object of the present invention is to more appropriately support the work of pharmacists. [Means for solving the problem]

[0006] In order to solve the above problems, a support information providing system according to one aspect of the present invention comprises: A support information providing system including a terminal device and a server configured to be able to communicate via a network, The terminal device a support request means for transmitting to the server a request for generating support information for pharmaceutical guidance based on target information that is to be considered in pharmaceutical guidance for the patient and includes the content of the previous pharmaceutical guidance for the patient and the result of the current interview; a support information display means for displaying the support information transmitted from the server in response to a request from the support request means, The server a target information acquisition means for acquiring the target information based on a request from the support request means; an inference means for generating the support information including the content of recommended pharmaceutical guidance by making an inference based on the target information; a support information providing means for providing the support information generated by the inference means to the terminal device; Equipped with. [Effects of the Invention]

[0007] According to the present invention, it is possible to more appropriately support the work of pharmacists. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a schematic diagram showing a concept of providing support information for pharmaceutical guidance in a support information providing system according to the present embodiment. FIG. [Figure 2] 1 is a diagram showing a system configuration of a support information providing system 1 according to an embodiment of the present invention. [Figure 3] 2 is a schematic diagram showing an overview of the business support functions provided in the support information providing server 30. FIG. [Figure 4] FIG. 8 is a diagram showing the hardware configuration of an information processing device 800 that constitutes each device. [Figure 5]2 is a block diagram showing the functional configuration of a terminal device 10. FIG. [Figure 6] FIG. 10 is a schematic diagram showing an example of a patient selection screen. [Figure 7] FIG. 10 is a schematic diagram illustrating an example of a questionnaire display screen. [Figure 8] FIG. 10 is a schematic diagram showing an example of a medical interview display screen. [Figure 9] FIG. 10 is a schematic diagram showing an example of a support information display screen (problem display screen). [Figure 10] FIG. 10 is a schematic diagram showing an example of a transition screen that is displayed when a side effect is selected as a problem related to a patient. [Figure 11] FIG. 10 is a schematic diagram showing an example of a transition screen displayed when an operation to call a side effect inference function is performed. [Figure 12] FIG. 10 is a schematic diagram showing an example of a side effect inference result display screen. [Figure 13] FIG. 10 is a schematic diagram showing another example of the support information display screen (initial support information display screen). [Figure 14] FIG. 10 is a schematic diagram showing another example of the support information display screen (revisit support information display screen). [Figure 15] FIG. 2 is a block diagram showing the functional configuration of a pharmacy computer 20. [Figure 16] 10 is a schematic diagram showing data relating to medication instructions among data relating to medications stored in a medication DB 176. FIG. [Figure 17] FIG. 2 is a block diagram showing the functional configuration of the support information providing server 30. [Figure 18] 10 is a flowchart showing the flow of a first support information display process executed by the terminal device 10. [Figure 19] 10 is a flowchart showing the flow of a first support information display process executed by the terminal device 10. [Figure 20] 10 is a flowchart showing the flow of a revisit support information display process executed by the terminal device 10. [Figure 21] 10 is a flowchart showing the flow of a revisit support information display process executed by the terminal device 10. [Figure 22] 10 is a flowchart showing the flow of information management processing executed by the pharmacy computer 20. [Figure 23] 10 is a flowchart showing the flow of support information providing processing executed by the support information providing server 30. [Figure 24] 10 is a flowchart showing the flow of a reconstruction process executed by the support information providing server 30. [Figure 25] FIG. 8 is a block diagram showing the functional configuration of a stand-alone information processing device 800 having a support information providing function. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0010] FIG. 1 is a schematic diagram showing the concept of providing support information for pharmaceutical guidance in the support information providing system according to this embodiment. As shown in FIG. 1, the support information providing system according to this embodiment supports pharmaceutical instruction for a patient visiting a pharmacy by identifying candidate medication instruction contents through inference using AI (artificial intelligence) based on the prescription data issued to the patient and questionnaire result data, and presenting the candidate medication instruction contents to the pharmacist during the patient's first visit. It is also possible to identify candidate medication instruction contents through rule-based inference based on the prescription data issued to the patient and questionnaire result data during pharmaceutical instruction during the first visit. It is also possible to identify candidate medication instruction contents by inference based on the results of the initial medical interview conducted by the pharmacist with the patient (interview results and the pharmacist's impressions from the patient) in addition to the prescription data issued to the patient and questionnaire result data.

[0011] Meanwhile, when providing pharmaceutical guidance to patients at their return visits, the support information provision system uses AI-based inference to identify potential medication instruction content based on the prescription data and questionnaire results issued to the patient, as well as the details of the previous medication instruction and the results of the medical interview at the return visit, and provides these to the pharmacist to support pharmaceutical guidance. Furthermore, when providing pharmaceutical guidance to patients at their return visits, rule-based inference can also be used to identify potential medication instruction content based on the prescription data issued to the patient, questionnaire results, the details of the previous medication instruction, and the results of the medical interview at the return visit. In this case, AI-based inference uses a trained model that has learned the content of instruction to be provided from the package inserts for various medications, using input data including text data expressing the patient's complaints (such as text data generated by voice recognition of the patient's speech, text data entered by the pharmacist, or text data representing options selected by the pharmacist). Colloquial expressions, including onomatopoeia such as "my stomach hurts" or "I feel nauseous," are acceptable in the text data expressing the patient's complaints. In other words, in this embodiment, the expressions expressed by the patient during the interview can be directly converted into text data representing the patient's complaints. However, the expressions expressed by the patient may also be interpreted by a pharmacist and converted into text data as pharmaceutical expressions representing the patient's condition.

[0012] As a result, according to the support information provision system of this embodiment, at the time of the first visit, support information (candidates for medication instruction content) inferred based on prescription data and questionnaire result data (or data from the results of the initial interview) can be presented to the pharmacist, and at the time of the second or subsequent visit, support information (candidates for medication instruction content) can be presented to the pharmacist by making inferences that reflect the previous medication instruction content and the results of the patient's interview. Therefore, the support information providing system according to this embodiment can more appropriately support the work of pharmacists.

[0013] [System Configuration] FIG. 2 is a diagram showing the system configuration of a support information providing system 1 according to an embodiment of the present invention. 2, the support information providing system 1 includes a terminal device 10, a pharmacy computer 20, and a support information providing server 30, and the terminal device 10, the pharmacy computer 20, the support information providing server 30, and an external server such as a drug information database server that stores drug information including medication instructions are configured to be able to communicate with each other via a network 40 (such as the Internet). Note that the terminal device 10 and the pharmacy computer 20 may be configured to be able to communicate with each other via a private network such as a VPN (Virtual Private Network).

[0014] The terminal device 10 is operated by a pharmacist and is configured as an information processing device such as a tablet terminal or a PC (Personal Computer). The terminal device 10 displays information to support the pharmacist's work when the pharmacist interviews a patient, and accepts information input (such as input of text data, touch operation, or voice input) by the pharmacist. For example, the terminal device 10 displays the contents of the interview (interview contents) when the pharmacist interviews a patient, and accepts input of the interview results. The terminal device 10 also accepts input of the impression the pharmacist has of the patient when they meet. Furthermore, the terminal device 10 displays support information (here, a problem related to the patient and recommended medication instructions) acquired by an initial support information display process (described later) based on information about the patient including the interview results (interview results and impressions, etc.) at the time of the first visit. Furthermore, the terminal device 10 displays support information (here, problems related to the patient and recommended medication instruction content) acquired by a revisit support information display process described later, based on information about the patient including the results of the medical interview (interview results, impressions, etc.) at the time of the return visit and the patient's information including the previous instruction history. Note that problems related to the patient are pharmaceutical judgment factors when the pharmacist provides medication instruction, including observation items, instruction guidelines, or points of focus.

[0015] The pharmacy computer 20 is composed of an information processing device such as a server computer that executes processes related to the operations of the dispensing pharmacy. The pharmacy computer 20 also has a function as a receipt computer (receipt computer function) that creates medical fee statements and a function for managing patient medication histories (medical history management function), and manages various data such as patient attributes for multiple patients, multiple patient prescriptions, the history of prescribed medications (medical history), medical interview results (interview results and impressions, etc.), the contents of interviews with patients, and information about medications handled at the pharmacy. The medical history managed by the pharmacy computer 20 includes a history of medication instructions (instruction history) given to patients by pharmacists. Furthermore, the pharmacy computer 20 executes the information management process described below to transmit various pieces of information it manages to the support information providing server 30 and update the various pieces of information it manages using various data transmitted from the support information providing server 30.

[0016] The support information providing server 30 is configured with an information processing device such as a server computer having a function for supporting the work of pharmacists (work support function). The support information providing server 30 can accept support requests (requests for providing support information) from terminal devices 10 installed in multiple pharmacies, and by executing the support information providing process described below, makes inferences by referring to data on patient attributes, medical interview results, prescriptions, or previous medication instruction content (instruction history), and generates a problem related to the patient and medication instruction content (support information) recommended for the patient. The support information providing server 30 then transmits the generated support information (the problem related to the patient and the recommended medication instruction content) to the terminal device 10. The support information providing server 30 also receives various data managed by pharmacy computers 20 installed in multiple pharmacies and manages this data in a database group similar to that of the pharmacy computers 20. In other words, the support information providing server 30 acquires (partially or entirely) the information held by the pharmacy computers 20 installed in multiple pharmacies and stores it for executing the support information providing process.

[0017] FIG. 3 is a schematic diagram showing an overview of the business support functions provided in the support information providing server 30. As shown in FIG. 3, the solid line blocks represent processing, and the dashed line blocks represent data. Among the solid line blocks, the thick line blocks are blocks to which computer inference (so-called artificial intelligence) is applied in this embodiment. As shown in FIG. 3 , the support information providing server 30 extracts features from the acquired results of the prescription, patient attributes, interviews, impressions, and previous medication instruction content (medication history), and performs machine learning-based inference based on the extracted features to generate an intermediate state for outputting recommended medication instruction content. In this embodiment, the support information providing server 30 performs inference on input data including text data expressing the patient's complaints (text data obtained by voice recognition of the patient's speech, text data entered by the pharmacist, or text data expressing options selected by the pharmacist, etc.) using a trained model that has learned the content of instruction to be provided from data on package inserts for various medications. The intermediate state generated at this time is expressed using the patient's problem as an index. As described above, the patient's problem is a pharmaceutical decision-making factor when a pharmacist provides medication instruction, including observation items, instruction guidelines, or points of focus. The intermediate state corresponds to an informational embodiment of one state of the pharmacist's thought process when providing medication instruction to the patient.

[0018] In addition, in FIG. 3, the support information providing server 30 generates recommended medication instruction contents (support information) for the patient by performing rule-based inference from the generated intermediate state (i.e., the problem related to the patient). In this way, an intermediate state is generated using machine learning-based inference from various data about the patient (patient attributes, medical interview results, prescription data, etc.), and recommended medication instructions are then generated from the intermediate state, allowing inference to be made using a procedure that conforms to the thought process of a pharmacist. In addition, when inference is performed using deep learning or the like, it is generally difficult to grasp the inference process, making it difficult to determine the basis for the inference results. However, by performing two-stage inference via an intermediate state, as in this embodiment, it is possible to confirm the basis for the recommended medication instructions.

[0019] Furthermore, in this embodiment, the support information providing server 30 extracts features by natural language processing the obtained interview results, impressions, and previous medication instruction content (instruction history). When extracting features, for example, it is possible to predefine content that is determined to be a feature in the thought process of a pharmacist when providing medication instruction (for example, the patient's age or personality) as an extraction target, or to predefine conditions for extracting words that will be a feature from the patient's complaints or the previous medication instruction content. Furthermore, in this embodiment, the support information providing server 30 reconstructs the hearing contents (hearing items) by executing a reconstruction process (described later) based on the generated intermediate state (problems related to the patient) and the hearing contents (hearing items) conducted when this intermediate state was generated. This makes it possible, for example, to add hearing contents that enable the importance of each problem to be more clearly determined to the generated intermediate state (problems related to the patient). The reconstruction performed at this time is realized, for example, by using rule-based inference or machine learning-based inference.

[0020] As an example, when rule-based reasoning is used, table data is used that defines specific interview content (listening items) that are desirable to interview when the patient is in an intermediate state of various contents (problems related to the patient).The table data is then referenced to determine whether the specific interview content (listening items) defined for the generated intermediate state (problem related to the patient) was included in the actual interview content, and if not, the specific interview content is added to subsequent interviews conducted with patients with similar patient attributes, etc., and the interview content is reconstructed. Furthermore, when machine learning-based inference is used, the generated intermediate state (patient problem) is used as input, and supervised machine learning is performed using the content (question items) that a pharmacist (especially an experienced pharmacist) thinks should be interviewed with a patient who recalls the problem as training data, and the interview content is reconstructed. However, other machine learning methods (deep learning, etc.) can also be used for reconstruction.

[0021] [Hardware configuration] Next, the hardware configuration of each device in the support information providing system 1 will be described. In the support information providing system 1, each device is configured by an information processing device such as a PC, a server computer, or a tablet terminal, and the basic configurations thereof are the same.

[0022] FIG. 4 is a diagram showing the hardware configuration of an information processing device 800 that constitutes each device. As shown in FIG. 4, the information processing device 800 constituting each device includes a CPU (Central Processing Unit) 811, a ROM (Read Only Memory) 812, a RAM (Random Access Memory) 813, a bus 814, an input unit 815, an output unit 816, a memory unit 817, a communication unit 818, a drive 819, and an imaging unit 820.

[0023] The CPU 811 executes various processes according to a program recorded in the ROM 812 or a program loaded from the storage unit 817 into the RAM 813 . The RAM 813 also stores data and the like necessary for the CPU 811 to execute various processes.

[0024] The CPU 811, ROM 812, and RAM 813 are connected to one another via a bus 814. To the bus 814, an input unit 815, an output unit 816, a storage unit 817, a communication unit 818, a drive 819, and an imaging unit 820 are connected.

[0025] The input unit 815 is composed of various buttons, a microphone, etc., and inputs various information in response to instruction operations. The output unit 816 is composed of a display, a speaker, etc., and outputs images and sounds. The storage unit 817 is configured with a hard disk or a DRAM (Dynamic Random Access Memory), etc., and stores various data managed by each server. The communication unit 818 controls communication with other devices via the network 40 .

[0026] Removable media 831, such as a magnetic disk, optical disk, magneto-optical disk, or semiconductor memory, is appropriately loaded into the drive 819. A program read from the removable media 831 by the drive 819 is installed in the storage unit 817 as needed. The imaging unit 820 is configured by an imaging device equipped with a lens, an imaging element, etc., and captures a digital image of a subject. When the information processing device 800 is configured as the pharmacy computer 20 or the support information providing server 30, it is possible to omit the imaging unit 820. When the information processing device 800 is configured as a tablet terminal, it is also possible to configure the input unit 815 using a touch sensor and place it over the display of the output unit 816, thereby providing a touch panel.

[0027] [Functional configuration] Next, the functional configuration of each device in the support information providing system 1 will be described.

[0028] [Functional configuration of terminal device 10] FIG. 5 is a block diagram showing the functional configuration of the terminal device 10. As shown in FIG. As shown in FIG. 5, in the CPU 811 of the terminal device 10, a user interface display control unit (UI display control unit) 51, a target information acquisition unit 52, a support request unit 53, and a support information acquisition unit 54 function.

[0029] The UI display control unit 51 controls the display of various input / output screens (hereinafter referred to as "UI screens") in the initial support information display process and the repeat support information display process, based on information for displaying a user interface screen (hereinafter referred to as "UI information") received from the support information providing server 30. For example, the UI display control unit 51 displays (outputs to a display) a screen for selecting a patient to whom support information is to be provided in the initial support information display process and the repeat support information display process (hereinafter referred to as a "patient selection screen"), a screen for displaying the contents of a questionnaire given to the patient at the time of their first visit (hereinafter referred to as a "questionnaire display screen"), a screen for displaying the contents of interviews (questionnaire items) when a pharmacist interviews a patient (hereinafter referred to as an "interview display screen"), and a screen for displaying support information (problems related to the patient and medication instruction contents) provided by the initial support information display process and the repeat support information display process (hereinafter referred to as a "support information display screen"), etc. The UI display control unit 51 also transmits various pieces of information input on the UI screen to the support information providing server 30. For example, the UI display control unit 51 transmits to the support information providing server 30 information identifying a patient selected on the patient selection screen, questionnaire results input on the questionnaire display screen, medical interview results input on the medical interview display screen, or previous medication instruction details (medication instruction history) acquired from the medication history managed by the pharmacy computer 20, etc. The UI display control unit 51 also transmits to the support information providing server 30 the results of medication instruction (medication instruction results) performed by a pharmacist on the support information display screen in response to a confirmation operation by the pharmacist (such as an operation of the "Send to Medication History" button). The UI display control unit 51 may also transmit various pieces of information input on the UI screen (for example, patient attributes of a new patient visiting the pharmacy) together with the support information providing server 30 to the pharmacy computer 20.

[0030] FIG. 6 is a schematic diagram showing an example of the patient selection screen. 6, the patient selection screen displays a list of patients corresponding to the patient attributes managed by the pharmacy computer 20, and the pharmacist selects a patient to interview (medical interview) from this list. For new patients, new patient attributes are registered on a registration screen (not shown).

[0031] FIG. 7 is a schematic diagram showing an example of a questionnaire display screen. As shown in Figure 7, the questionnaire display screen displays the contents of the questionnaire, including multiple-choice questions, and the patient enters the answers themselves, or the pharmacist enters the answers based on the patient's answers. The questionnaire contents displayed at this time are a series of questions set in common for all patients, and are used to identify patient attributes such as the patient's name, gender, medical history, and illnesses currently being treated.

[0032] FIG. 8 is a schematic diagram showing an example of a medical interview display screen. As shown in FIG. 8, the medical interview display screen displays interview content (interview items) including multiple-choice questions, and the pharmacist inputs the interview results (interview results) based on the patient's answers. The interview content displayed at this time is a series of questions selected according to patient attributes, and is determined for each patient, each prescription content, or the number of interviews (for example, whether it is the first interview or the second interview), etc. The medical interview display screen also has a free entry field for the pharmacist, allowing the pharmacist to freely enter any information they noticed when interviewing the patient (for example, physical impressions such as "lack of energy" or "pale complexion," or personality impressions such as "nervous personality" or "judgmental personality").

[0033] FIG. 9 is a schematic diagram showing an example of a support information display screen (problem display screen). As shown in FIG. 9 , the problem display screen generally displays pharmaceutical decision factors (patient-related problems) when a pharmacist (especially an experienced pharmacist) provides medication instruction. The patient-related problems displayed at this time include observation items that the pharmacist recalls for the patient to whom the pharmacist provides medication instruction, the instruction guidelines that the pharmacist envisions for medication instruction, and indicators (points of focus) that the pharmacist recalls in the thought process that the pharmacist follows when providing medication instruction. In this embodiment, patient-related problems (observation items, points of focus, instruction guidelines, etc.) are presented after being classified into multiple categories, such as "side effects," "adherence," "changes in physical condition," "concomitant medications (interactions)," "elderly," and "lifestyle guidance." The method of determining the classification when presenting patient-related problems (the number of categories, specificity, etc.) can be set in advance, or may be changed depending on the inference results.

[0034] Furthermore, in this embodiment, the relevance (here, a percentage indicating the importance) of a patient to each category of these patient-related problems is calculated based on the patient attributes, medication history, questionnaire results, and interview results of the patient who answered the medical interview. By referring to the relevance of the presented patient-related problems, a pharmacist who receives the support information can easily recognize what perspectives to pay attention to when providing medication instructions to the patient who answered the medical interview. Furthermore, the pharmacist can recognize the difference between the pharmacist's own thinking results and the relevance of the patient-related problems presented in the support information, which can be useful for the pharmacist's self-evaluation or self-improvement. Furthermore, if the pharmacist recorded a problem regarding the patient for whom the next instruction should be given during the previous instruction given to the patient, the recorded problem regarding the patient will be displayed in an identifiable manner (by using a different background color, surrounding it with a frame, blinking, etc.) on the support information display screen shown in Figure 9.

[0035] Furthermore, when an operation to select one of the problems related to the displayed patient is performed on the support information display screen (problem display screen) shown in FIG. 9, specific information related to the problem related to the selected patient is further displayed. FIG. 10 is a schematic diagram showing an example of a transition screen that is displayed when a side effect is selected as a problem related to a patient. The transition screen example shown in Figure 10 displays a schematic diagram of the human body and the names of body elements (organs) arranged around the schematic diagram. The names of the body elements displayed here are body elements that may cause side effects, and can serve as supplementary information when pharmacists provide guidance to patients.

[0036] Furthermore, the transition screen example shown in Figure 10 displays a list of drugs that may be related to side effects and their side effects. This list displays side effects that are recommended for guidance based on general criteria when providing guidance to a patient for the first time. Each side effect column in this list also displays a slide button for the pharmacist to input whether or not they have provided an explanation. The pharmacist explains to the patient those side effects that they determine require explanation from the listed side effects, and sets the slide button to the position indicating "explained" (the right side in Figure 10). Then, when the pharmacist operates the "Instruct on the above side effects" button, the screen transitions to a display screen showing the guidance content for the side effects. In this case, the side effects set to "explained" are recorded in the medication history as having been explained to the patient by the pharmacist. The list shown in Figure 10 includes a "previous instruction" column that indicates whether or not the patient was previously instructed about the displayed side effect.

[0037] Furthermore, in the list shown in Figure 10, when a patient is receiving instruction for the second or subsequent time (return visit), side effects that are recommended for instruction based on the content of the previous instruction are displayed. For example, if the pharmacist recorded side effects (points of focus) that should be included in the next instruction during the previous instruction given to the patient, the recorded side effects are displayed in the list in Figure 10. This ensures that important instruction content is reliably passed on even when a different pharmacist from the one who gave the previous instruction gives the current instruction.

[0038] FIG. 11 is a schematic diagram showing an example of a transition screen that is displayed when an operation for calling a side effect inference function is performed. In the example transition screen shown in Figure 11, the body parts (hexagons indicated by solid lines) and symptoms (hexagons indicated by dashed lines) that are elements of side effects are listed, and one or more of these side effect elements can be selected. In the example transition screen shown in Figure 11, the pharmacist selects one or more side effect elements based on the results of the patient's medical interview and operates the ``Answer side effects'' button to perform side effect inference.Then, from among the side effects of the drug prescribed to the patient, side effects that are inferred to be related based on the results of the patient's medical interview are identified.

[0039] The inference method used in this case involves associating all combinations of side effect elements with related side effects in a table format or the like, and identifying side effects corresponding to one or more side effect elements selected by the pharmacist as the inference result. It is also possible to perform machine learning using training data in which one or more side effect elements corresponding to the patient's interview results are input and the corresponding side effects are output, and then perform inference using AI. In this case, the AI ​​infers side effects corresponding to one or more side effect elements selected by the pharmacist based on the actual patient interview results based on the results of machine learning.

[0040] FIG. 12 is a schematic diagram showing an example of a side effect inference result display screen. In the example of the inference result display screen shown in FIG. 12, the side effects of drugs A and B tablets that are presumed to be related to the side effect element "tiredness" selected by the pharmacist are listed. The pharmacist can refer to the listed side effects of each drug and determine the side effects based on the patient's interview results. In addition, in the list of side effects displayed, a slide button is displayed in each column for the pharmacist to input whether or not an explanation has been given. The pharmacist explains to the patient about side effects that are judged to require explanation from among the listed side effects, and sets the slide button to the position indicating "explained" (the right side in Figure 12).

[0041] In Figure 12, if you set the slide button to "Explained" and operate the "Next" button, the information that you have instructed (explained) the patient about the side effect that you have set to "Explained" will be carried over to the initial support information display screen or the follow-up support information display screen. In addition, the example of the inference result display screen shown in Figure 12 displays a "Question Inquiry" button for making a question inquiry. When a pharmacist refers to the list of side effects displayed on the inference result display screen and determines that it is necessary to make a question inquiry to a doctor, they can operate the "Question Inquiry" button. This displays an input screen for inputting the details of the question inquiry, allowing the pharmacist to input the details of the question inquiry. The details of the question inquiry entered on this input screen are stored as a medication history.

[0042] FIG. 13 is a schematic diagram showing another example of the support information display screen (initial support information display screen). As shown in FIG. 13, the initial support information display screen displays a list of medication instruction sentences indicating recommended medication instruction contents for the patient who has been interviewed (hereinafter referred to as the "recommended medication instruction sentence list") and a list of general medication instruction sentences for the prescribed medication (hereinafter referred to as the "general medication instruction sentence list"). The recommended medication instruction sentence list represents a list of medication instruction sentences selected based on the patient problems identified by the support information provision process and their relevance, and reflects results reflecting the observation items that pharmacists (especially experienced pharmacists) recall for patients for whom they are providing medication instruction, the guidance guidelines that pharmacists envision in providing medication instruction, and the indicators (points of view) that pharmacists recall in their thought process when providing medication instruction. In addition, the general medication instruction sentence list represents results selected in accordance with the medication to be prescribed based on the patient's prescription data. In addition, each medication instruction statement is accompanied by a check box, and once the pharmacist has checked the medication instruction statement, it is confirmed by the pharmacist (by pressing the "Send to medication history" button, etc.) and registered in the patient's medication history as a list of medication instruction statements used for medication instructions (medication instruction results).

[0043] Here, the medication instruction content that is marked as "explained" on the inference result display screen for the focus point (side effects, etc.) illustrated in Figure 12 is displayed on the initial support information display screen shown in Figure 13 with a check mark in the checkbox (i.e., indicating that medication instruction has been given) (the same applies to the repeat visit support information display screen shown in Figure 14). In addition, on the initial support information display screen in Figure 13, it is possible to perform operations to input the interview content (interview content input operation) and operations to input problems regarding the patient for whom the pharmacist should provide guidance next (next guidance content input operation), and the input content can be recorded in the medication history by the pharmacist's confirmation operation (such as operating the ``Send to medication history'' button) (the same is true for the repeat visit support information display screen shown in Figure 14).

[0044] FIG. 14 is a schematic diagram showing another example of the support information display screen (revisit support information display screen). As shown in Figure 14, the return visit support information display screen displays a list of medication instruction sentences indicating recommended medication instruction contents for the patient who has been interviewed (recommended medication instruction sentence list) and a list of general medication instruction sentences for the prescribed medication (general medication instruction sentence list). The recommended medication instruction sentence list represents a list of medication instruction sentences selected based on the problems related to the patient identified by the support information provision process and their relevance, and represents results that reflect the observation items that pharmacists (especially experienced pharmacists) recall for patients for whom they provide medication instruction, the guidance policy that pharmacists envision when providing medication instruction, and the indicators (points of view) that come to mind in the thought process that pharmacists follow when providing medication instruction.

[0045] Here, the repeat visit support information display screen shown in Figure 14 differs from the initial support information display screen shown in Figure 13 in that it displays a list of recommended medication instruction sentences, including the contents of the previous medication instruction and the results of the current medical interview (interview results, etc.) conducted based on the contents of the previous medication instruction as input data.

[0046] Specifically, the initial support information display screen shown in FIG. 13 displays the results of inference (initial support information display process) performed using prescription data and questionnaire result data as input. If an interview is conducted at the patient's first visit, the data from the interview results can also be included as input in the initial support information display process. The initial support information display screen shown in FIG. 13 displays the following inference results for the prescribed drug "M Tablets 40 mg": a first recommended medication instruction: "Side effects such as decreased urinary output and swelling may occur. Instruct the patient to consult a doctor if this occurs." and a second recommended medication instruction: "There are drugs that are contraindicated for concomitant use. Instruct the patient to inform the doctor that the drug is currently being taken if visiting another department." The first medication instruction represents instructions calling attention to a relatively frequent side effect of the prescribed drug "M Tablets 40 mg."

[0047] In contrast, the revisit support information display screen shown in Figure 14 displays the results of inference (revisit support information display processing) performed using prescription data, questionnaire result data, data on the contents of the previous medication instruction, and data on the results of the current medical interview (interview results, etc.) as input. The revisit support information display screen shown in Figure 14 displays, as inference results, a first recommended medication instruction statement for the prescribed medication "M Tablets 40 mg," which reads, "Side effects such as muscle pain and fatigue may occur. Instruct the patient to discontinue taking the medication and consult a doctor if this occurs." and a second recommended medication instruction statement, "There are medications that are contraindicated for concomitant use. Instruct the patient to inform the doctor that the patient is currently taking these medications if visiting another department." The first medication instruction statement represents instruction content that calls attention to a side effect of the prescribed medication "M Tablets 40 mg," which occurs infrequently but is relatively serious if it does occur. In the example of the follow-up support information display screen shown in Figure 14, for the prescribed drug "M Tablets 40 mg," a side effect that was not advised in the previous medication instruction content and that is related to the symptoms that the patient complained of in the current interview (such as "I had muscle pain that I don't remember having") is recommended at the top of the recommended medication instruction content. Note that the list of general medication instruction content and other displayed content are the same as those on the initial support information display screen.

[0048] Returning to FIG. 5 , the target information acquisition unit 52 acquires various pieces of information to be considered about a patient for whom medication instruction is to be given (hereinafter referred to as "information to be considered"). That is, the information to be considered is information to be considered when providing pharmaceutical instruction to a patient. For example, the target information acquisition unit 52 acquires questionnaire result data including patient attribute data for identifying the patient, such as the name, age, medical history, and illness being treated of the patient who brought the prescription; data on the medical interview results (interview results, impressions, etc.) entered by the pharmacist on the medical interview display screen; data indicating the contents of the prescription brought by the patient (prescription data); data on the contents of medication instruction given to the patient at the patient's previous visit; and the like. Note that the information acquired by the target information acquisition unit 52 is not limited to these; it is also possible to acquire data on the patient's medication history, data stored in an electronic medication notebook, and the like.

[0049] The support request unit 53 transmits a request for providing support information to the support information providing server 30 together with the consideration target information acquired by the target information acquisition unit 52 . The support information acquisition unit 54 acquires support information transmitted from the support information providing server 30 in response to the support information request transmitted by the support request unit 53. The support information acquired by the support information acquisition unit 54 includes data on the problem related to the patient, a list of medication instructions indicating the contents of recommended medication instructions (list of recommended medication instructions), and a list of general medication instructions related to prescribed drugs (list of general medication instructions).

[0050] [Functional configuration of pharmacy computer 20] FIG. 15 is a block diagram showing the functional configuration of the pharmacy computer 20. 15, a receipt management unit 151, a medication history management unit 152, and a database management unit (DB management unit) 153 function in CPU 811 of pharmacy computer 20. In addition, a patient attribute database (patient attribute DB) 171, a prescription database (prescription DB) 172, a medication history database (medication history DB) 173, a medical interview result database (medication result DB) 174, a hearing content database (hearing content DB) 175, and a drug database (drug DB) 176 are formed in storage unit 817 of pharmacy computer 20.

[0051] Patient attribute data such as the patient's address, name, age, sex, and various information representing the patient's individual characteristics are stored in the patient attribute DB 171. These patient attributes are composed of information acquired by the pharmacist through conversations with the patient, information provided by the patient in response to questionnaires, etc., and include, for example, the patient's hobbies, work, family structure, favorite foods, etc. The prescription DB 172 stores data on prescriptions issued to patients in association with information identifying each patient and the date the prescription was brought in. The medication history DB 173 stores data on the history of medications prescribed to patients (medication history) in association with information identifying each patient. The medication history also stores data on the history of medication instructions (instruction history) given to patients by pharmacists.

[0052] The medical interview result DB 174 stores the results of medical interviews conducted with patients in association with information identifying each patient and the date and time of the interview. The medical interview results include the results of interviews entered by pharmacists after interviewing patients, and impressions entered by pharmacists based on the impressions they received from patients they met. The hearing content DB 175 stores data of a list of hearing content (question items) to be asked of patients, such as "Have you forgotten to take your medicine?" or "Have you experienced any changes in your physical condition after taking your medicine?". Data on various drugs handled at pharmacies is stored in the drug DB 176. This data on drugs includes the drug name (generic name), drug code, etc., as well as the contents of medication instructions and package inserts.

[0053] FIG. 16 is a schematic diagram showing data relating to medication instructions among data relating to medications stored in the medication DB 176. As shown in FIG. As shown in FIG. 16, each drug identified by a drug code and a drug name is associated with a plurality of medication instruction sentences provided by a drug manufacturer. In FIG. 16, a serial number (instruction sentence number) is assigned to the medication instruction sentence associated with each drug. Furthermore, each medication instruction sentence is associated with a problem related to one patient or problems related to multiple patients. In this embodiment, the relevance of the patient problem to the information to be considered of the patient for which support information is to be provided is calculated by inference, which will be described later. A high relevance of the patient problem to the information to be considered means that the importance of the medication instruction sentence associated with the problem related to the patient is relatively high for the information to be considered.

[0054] The medical receipt management unit 151 acquires information necessary for a patient's medical receipt (patient attributes, prescription contents, medical insurance points, etc.) and performs the process of issuing the medical receipt. The medication history management unit 152 manages the history of drugs prescribed to a patient (medication history). For example, when a new prescription is given to a patient, the medication history management unit 152 stores data on the history of the currently prescribed drug (current medication history) in the medication history DB 173, and when a request to transmit the patient's medication history data is received from the terminal device 10, the medication history management unit 152 acquires the requested medication history data from the medication history DB 173 and transmits it to the terminal device 10.

[0055] The DB management unit 153 manages synchronization between data in various databases managed by the pharmacy computer 20 and data in various databases provided in the support information providing server 30. For example, when the DB management unit 153 receives a request to send medication history data of a patient from the support information providing server 30, the DB management unit 153 acquires the requested medication history data from the medication history DB 173 and sends it to the support information providing server 30. Furthermore, at a preset time (for example, 3:00 AM), the DB management unit 153 sends data updated in the various databases managed by the pharmacy computer 20 to the support information providing server 30, and receives data updated in the various databases in the support information providing server 30 from the support information providing server 30 and updates the various databases it manages.

[0056] [Functional configuration of the support information providing server 30] FIG. 17 is a block diagram showing the functional configuration of the support information providing server 30. As shown in FIG. 17, the CPU 811 of the support information providing server 30 functions as follows: a DB management unit 251, a user interface information generation unit (UI information generation unit) 252, a support request receiving unit 253, a target information acquisition unit 254, a feature extraction unit 255, a problem identification unit 256, a presented information evaluation unit 257, a support information generation unit 258, a support information providing unit 259, and a reconstruction execution unit 260. The storage unit 817 of the support information providing server 30 also stores a patient attribute database (patient attribute DB) 271, a prescription database (prescription DB) 272, a medication history database (medication history DB) 273, a medical interview result database (medication result DB) 274, a hearing content database (hearing content DB) 275, and a drug database (drug DB) 276. The contents stored in each database and the contents stored in pharmacy computer 20 are synchronized by DB management unit 251.

[0057] The DB management unit 251 performs management to synchronize the data of various databases provided in the support information providing server 30 with the data of various databases managed by the pharmacy computer 20. For example, the DB management unit 251 transmits data of various databases updated in the support information providing server 30 to the pharmacy computer 20 at a preset time (e.g., 3:00 AM), and receives updated data in the various databases managed by the pharmacy computer 20 from the pharmacy computer 20, and updates the various databases it manages.

[0058] The UI information generation unit 252 generates UI information for the terminal device 10 to display a UI screen and transmits the generated UI information to the terminal device 10. Specifically, the UI information generation unit 252 generates, as UI information, a frame format for displaying the UI screen and substantive content to be inserted into the format. In this embodiment, the substantive content to be inserted into the format includes, for example, patient attributes on the patient selection screen, questionnaire content on the questionnaire display screen, hearing content on the medical interview display screen, display content on a transition screen when an inference function for a patient-related problem is invoked and on a screen displaying an inference result for a patient-related problem, and support information on a support information display screen (a problem display screen, an initial support information display screen, and a follow-up support information display screen). In addition, in response to transmitting the UI information to the terminal device 10, the UI information generation unit 252 receives various information (such as information to identify the patient, questionnaire result data, prescription data, and medical interview result data) transmitted from the terminal device 10. Furthermore, the UI information generating unit 252 infers the hearing contents suitable for the patient based on the information to be considered that has been grasped up to the current medical interview, and sets the hearing contents as the substantial contents (hearing contents) to be inserted into the medical interview display screen. Note that the hearing contents displayed on the medical interview display screen are reconstructed as described later.

[0059] The support request receiving unit 253 receives a request for providing support information together with the currently input consideration subject information from the terminal device 10. The target information acquisition unit 254 acquires the consideration subject information transmitted from the terminal device 10 and the consideration subject information stored in each database. Note that the consideration subject information stored in each database is acquired when necessary, and for example, when information identifying a patient is transmitted from the terminal device 10, data on the patient attributes of the patient is acquired as one of the consideration subject information.

[0060] The feature extraction unit 255 refers to the consideration information received by the support request receiving unit 253 and the consideration information acquired from each database, and extracts features according to predefined extraction conditions. At this time, the feature extraction unit 255 performs natural language processing to extract phrases representing features included in the consideration information, or extracts feature values ​​calculated or estimated from the consideration information. For example, the feature extraction unit 255 decomposes text data representing a patient's complaint, such as text data obtained by speech recognition of the patient's speech, text data entered by the pharmacist, or text data representing the contents of previous medication instructions, into words by performing morphological analysis, and extracts feature values ​​according to predefined conditions (e.g., extracting specific parts of speech or onomatopoeia). Furthermore, the feature extraction unit 255 extracts text data representing an option selected by the pharmacist from the text data representing the patient's complaint as a feature. Furthermore, when "patient's age" is defined as a feature, the feature extraction unit 255 extracts phrases representing the patient's age included in the consideration information. In addition, when "visit interval" is defined as a feature, the feature extraction unit 255 calculates the visit interval (i.e., the difference between the last visit date and the current visit date) from the visit date included in the information to be considered.

[0061] The problem identification unit 256 uses the feature values ​​extracted by the feature extraction unit 255 as input and performs machine learning-based inference to generate an intermediate state represented by an index of a patient-related problem. In this embodiment, the problem identification unit 256 includes an inference engine constructed by machine learning to determine what pharmaceutical decision-making factors (patient-related problems) a pharmacist (especially an experienced pharmacist) would associate with a patient when recognizing the feature values. Therefore, the intermediate state generated by the problem identification unit 256 represents the result of inferring the pharmaceutical decision-making factors associated with a patient when given the target patient's information to be considered. In addition, this inference engine includes a trained model that has trained the contents of instruction to be provided in the data on package inserts for various medications from input data including text data representing the patient's complaints (such as text data obtained by voice recognition of the patient's speech, text data entered by the pharmacist, or text data representing options selected by the pharmacist). Therefore, when the problem identification unit 256 receives text data representing a patient's complaint, it can infer the contents of pharmaceutical instruction that a pharmacist would associate with the complaint. In addition, this inference engine can include data on the medication instructions (instruction history) from the patient's previous visit as input data, making it possible to infer the pharmaceutical instructions that will come to mind when the pharmacist recognizes the patient's complaints and instruction history.

[0062] Furthermore, in this embodiment, when generating a problem related to a patient as an index of an intermediate state, the problem identification unit 256 also generates the relevance of the problem related to the patient to the information under consideration. That is, the inference engine provided in the problem identification unit 256 has been machine-learned to learn what pharmaceutical judgment factors (problems related to the patient) are associated with a patient when a pharmacist (especially an experienced pharmacist) recognizes a feature, including the relevance of the problem related to the patient.

[0063] The presented information evaluation unit 257 evaluates each medication instruction sentence related to the patient's problem by performing rule-based inference based on the patient's problem identified by the problem identification unit 256 and the relevance of the patient's problem. For example, the presented information evaluation unit 257 weights each medication instruction sentence with a numerical value indicating the relevance of the patient's problem to which the medication instruction sentence is related, and determines an evaluation value for the medication instruction sentence associated with the patient's problem. Note that a weight for the medication instruction sentence may be set for each drug, and the evaluation value for the medication instruction sentence may be determined by further reflecting the numerical value indicating the relevance of the patient's problem to this weight. The evaluation value is determined in this way, and medication instruction sentences that are assigned an evaluation value equal to or greater than a preset threshold value belong to the list of recommended medication instruction sentences. On the other hand, medication instruction sentences that are assigned an evaluation value less than the preset threshold value belong to the list of general medication instruction sentences.

[0064] The support information generation unit 258 acquires data on the problem related to the patient identified by the presented information evaluation unit 257. Furthermore, the support information generation unit 258 acquires medication instruction sentences from the drug DB 276 as medication instruction content (support information) recommended for the patient based on the evaluation value of the medication instruction sentence determined by the presented information evaluation unit 257. That is, the support information generation unit 258 acquires medication instruction sentences evaluated based on the relevance of the problem related to the patient, and generates data on a list of recommended medication instruction sentences by sorting medication instruction sentences that have been assigned an evaluation value equal to or greater than a predetermined threshold in descending order of evaluation value. Furthermore, the support information generation unit 258 acquires medication instruction sentences that have been assigned an evaluation value less than a predetermined threshold in the presented information evaluation unit 257, and sorts them in a predetermined order to generate data on a list of general medication instruction sentences. The predetermined order can take various forms, but as an example, it can be in descending order of reference frequency. Then, the support information generation unit 258 generates support information including data on problems related to the patient and the relationships between problems related to the patient, data on a list of recommended medication instructions, and data on a list of general medication instructions, and outputs the generated support information to the support information provision unit 259.

[0065] The support information providing unit 259 transmits the support information input from the support information generating unit 258 (information including data on problems related to the patient and the relationships between problems related to the patient, data on a list of recommended medication instructions, and data on a list of general medication instructions) to the terminal device 10 (output to the terminal device 10 via the network 40). The reconstruction executing unit 260 corrects the hearing contents (listening items) by performing reconstruction based on the data of the hearing contents in the consideration information received by the support request receiving unit 253 and the intermediate state (problem related to the patient) generated by the problem identifying unit 256. For example, the reconstruction executing unit 260 reconstructs the hearing contents (listening items) by using rule-based inference or machine learning-based inference.

[0066] [Operation] Next, the operation of the support information providing system 1 will be described.

[0067] [Initial support information display process] 18 and 19 are flowcharts showing the flow of the initial support information display process executed by the terminal device 10. FIG. The initial support information display process is executed on the terminal device 10 when the patient visits the pharmacy for the first time, and is started in response to an instruction to execute the initial support information display process being input via the input unit 815 of the terminal device 10. In step S1, the UI display control unit 51 receives UI information from the support information providing server 30 and displays a patient selection screen (see FIG. 6). In step S2, the UI display control unit 51 receives the selection of the patient to be interviewed, thereby specifying information for identifying the patient.

[0068] In step S3, the UI display control unit 51 receives input of prescription data issued to the selected patient. The patient identification information and prescription data are transmitted to the support information providing server 30. In step S4, the UI display control unit 51 receives UI information from the support information providing server 30 and displays a medical interview display screen for the first visit corresponding to the selected patient (see FIG. 8). When the first support information display process is executed, since the patient to be interviewed is visiting the clinic for the first time (first medical examination), the UI display control unit 51 receives UI information from the support information providing server 30 in step S4, displays a questionnaire display screen (see FIG. 7), and accepts input to the questionnaire display screen. Then, the UI display control unit 51 transmits the questionnaire results entered on the questionnaire display screen to the support information providing server 30, and then receives UI information for the medical interview display screen and displays a medical interview display screen corresponding to the patient.

[0069] In step S5, the UI display control unit 51 accepts input of answers to the medical interview display screen for the first visit. Data of the answers (medical interview results) to the medical interview display screen for the first visit is transmitted to the support information providing server 30. In step S6, the target information acquisition unit 52 acquires various pieces of information to be considered (information to be considered) from the information input to the UI screen regarding the patient for whom medication instruction is to be given. In step S7, the support request unit 53 transmits a request for providing support information to the support information providing server 30 together with the consideration target information acquired by the target information acquisition unit 52. In step S8, the support information acquisition unit 54 acquires UI information including support information transmitted from the support information providing server 30 in response to the support information request transmitted by the support request unit 53.

[0070] In step S9, the UI display control unit 51 displays a support information display screen (problem display screen) showing a problem related to the patient based on the acquired UI information (see FIG. 9). In step S10, the UI display control unit 51 determines whether or not an operation to select a problem related to the patient has been performed. If an operation to select a problem related to the patient has been performed, the result of determination in step S10 is YES, and the process proceeds to step S11. On the other hand, if an operation to select a problem related to the patient has not been performed, the result of step S10 is determined to be NO, and the process proceeds to step S20. In step S11, the UI display control unit 51 displays a transition screen corresponding to a problem related to the selected patient (see FIG. 10). In step S12, the UI display control unit 51 accepts an input (determination of medication instruction content or invocation of an inference function) for a transition screen corresponding to a problem related to the selected patient.

[0071] In step S13, the UI display control unit 51 determines whether an operation for determining medication instruction content or an operation for calling an inference function has been performed. If an operation for deciding on medication instruction content is performed, it is determined to be a "decision operation" in step S13, and the process proceeds to step S20. On the other hand, if an operation for calling the inference function has been performed, it is determined to be an "inference operation" in step S13, and the process proceeds to step S14. In step S14, the UI display control unit 51 displays a transition screen corresponding to the selected inference function (see FIG. 11). In step S15, the UI display control unit 51 accepts an operation for inputting an inference content (selection of an element of a side effect or input of a free word).

[0072] In step S16, the UI display control unit 51 determines whether or not an operation for executing inference (such as an operation of the "answer side effects" button) has been performed. If an operation for executing inference has not been performed, the determination in step S16 is NO, and the process proceeds to step S15. On the other hand, if an operation for executing inference has been performed, the determination in step S16 is YES, and the process proceeds to step S17. In step S17, the UI display control unit 51 displays a screen (inference result display screen) that displays the selected inference result. In step S18, the UI display control unit 51 accepts an operation for selecting medication instruction content.

[0073] In step S19, the UI display control unit 51 determines whether or not an operation for determining the medication instruction content has been performed. If an operation for determining the medication instruction content has been performed, the determination in step S19 is YES, and the process proceeds to step S20. On the other hand, if no operation for determining the medication instruction content has been performed, the result of step S19 is determined to be NO, and the process proceeds to step S18.

[0074] In step S20, the UI display control unit 51 determines whether or not an operation to display recommended medication instruction contents (such as an operation of the "Instruction" button) has been performed. If the operation to display the recommended medication instruction contents has not been performed, the result of step S20 is determined to be NO, and the process proceeds to step S9. On the other hand, if an operation to display recommended medication instruction contents has been performed, the result of determination in step S20 is YES, and the process proceeds to step S21.

[0075] In step S21, the UI display control unit 51 displays a support information display screen showing recommended medication instruction content. In step S22, the UI display control unit 51 determines whether or not an operation has been performed on the support information display screen showing the recommended medication instruction content. At this time, the UI display control unit 51 accepts an operation of checking the checkbox of the medication instruction sentence, an operation of inputting the next instruction content, and the like. If an operation is performed on the support information display screen showing the recommended medication instruction content, the result of determination in step S22 is YES, and the process proceeds to step S21. On the other hand, if no operation has been performed on the support information display screen, the result of step S22 is determined to be NO, and the process proceeds to step S23.

[0076] In step S23, the UI display control unit 51 determines whether or not an operation to confirm the medication instruction content has been performed. If the operation to confirm the medication instruction contents has not been performed, the result of step S23 is determined to be NO, and the process proceeds to step S21. On the other hand, if an operation to confirm the medication instruction contents has been performed, the result of determination in step S23 is YES, and the process proceeds to step S24. In step S24, the UI display control unit 51 transmits the confirmed medication instruction content to the medication history as an instruction history. After step S24, the initial support information display process ends.

[0077] [Revisit support information display processing] 20 and 21 are flowcharts showing the flow of the revisit support information display process executed by the terminal device 10. FIG. The revisit support information display process is executed by the terminal device 10 when the patient returns to the clinic, and is started in response to an instruction to execute the revisit support information display process being input via the input unit 815 of the terminal device 10.

[0078] Steps S31 to S33 of the revisit support information display process are the same as steps S1 to S3 of the first-time support information display process. In step S34, the UI display control unit 51 receives UI information from the support information providing server 30 and displays a medical interview display screen for a return visit according to the selected patient (see FIG. 8).

[0079] In step S35, the UI display control unit 51 accepts input of answers to the medical interview display screen for the return visit. Data of the answers (medical interview results) to the medical interview display screen for the return visit is transmitted to the support information providing server 30. In step S36, the target information acquisition unit 52 acquires data on the medication instruction content from the patient's previous visit. The acquired data on the medication instruction content from the previous visit is considered to be one of the information to be considered. In step S37, the target information acquisition unit 52 acquires various pieces of information to be considered (information to be considered) from the information input to the UI screen regarding the patient for whom medication instruction is to be given.

[0080] In step S38, the support request unit 53 transmits a request for providing support information to the support information providing server 30 together with the information to be considered acquired by the target information acquisition unit 52. The information to be considered that is transmitted at this time includes data on the contents of medication instructions given to the patient at their previous visit. The processing in steps S39 to S55 is the same as that in steps S8 to S24 of the revisit support information display processing.

[0081] [Information Management Processing] FIG. 22 is a flowchart showing the flow of the information management process executed by the pharmacy computer 20. The information management process is started in response to an instruction to execute the information management process being input via the input unit 815 of the pharmacy computer 20.

[0082] In step S61, the DB management unit 153 determines whether or not a prescription for a patient has been received. If the patient's prescription has not been received, the determination in step S61 is NO, and the process proceeds to step S63. On the other hand, if the patient's prescription has been accepted, the determination in step S61 is YES, and the process proceeds to step S62. In step S62, the DB management unit 153 updates the patient attribute DB 171 and the prescription DB 172 with the data of the currently accepted prescription.

[0083] In step S63, the DB management unit 153 determines whether or not a request to send the patient's medication history data has been received from the support information providing server 30. If a request to transmit the patient's medication history data has not been received from the support information providing server 30, the result of step S63 is determined to be NO, and the process proceeds to step S65. On the other hand, if a request to transmit the patient's medication history data has been received from the support information providing server 30, the result of the determination in step S63 is YES, and the process proceeds to step S64. In step S64, the DB management unit 153 transmits the requested data of the patient's medication history to the support information providing server 30.

[0084] In step S65, the DB management unit 153 determines whether or not the current medication history data (data on the medication instruction content explained to the patient, etc.) has been received from the support information providing server 30. If the current medication history data has not been received from the support information providing server 30, the determination in step S65 is NO, and the process proceeds to step S67. On the other hand, if the current medication history data has been received from the support information providing server 30, the determination in step S65 is YES, and the process proceeds to step S66. In step S66, the DB management unit 153 updates the medication history DB 173 with the received medication history data.

[0085] In step S67, the DB management unit 153 determines whether it is the predetermined timing for synchronizing the data of the various databases managed by the pharmacy computer 20 with the data of the various databases provided in the support information providing server 30. If the predetermined timing for synchronizing the data of the various databases managed by the pharmacy computer 20 with the data of the various databases provided on the support information providing server 30 has not arrived, the result of step S67 is determined to be NO, and the information management process is repeated. On the other hand, if the timing has arrived at the preset time for synchronizing the data in the various databases managed by the pharmacy computer 20 with the data in the various databases provided on the support information providing server 30, the answer in step S67 is YES, and processing proceeds to step S68.

[0086] In step S68, the DB management unit 153 updates each database by sending and receiving data with the support information providing server 30, and synchronizes the data in the various databases managed by the pharmacy computer 20 with the data in the various databases provided on the support information providing server 30. After step S68, the information management process is repeated.

[0087] [Support information provision processing] FIG. 23 is a flowchart showing the flow of the support information providing process executed by the support information providing server 30. The support information providing process is started in response to an instruction to execute the support information providing process being input via the input unit 815 of the support information providing server 30.

[0088] In step S71, the support request receiving unit 253 receives a request to provide support information together with consideration subject information from the terminal device 10. In step S72, the feature extraction unit 255 performs natural language processing on the received information to be considered (hearing results, impressions, etc.). In step S73, the feature extraction unit 255 extracts feature amounts from the consideration object information. In step S74, the problem identification unit 256 performs machine learning-based inference using the feature quantities extracted by the feature extraction unit 255 as input, and generates an intermediate state represented by an index of the problem related to the patient. At this time, the relevance of the problem related to the patient to the information under consideration is also generated.

[0089] In step S75, the presented information evaluation unit 257 performs rule-based inference based on the problems related to the patient identified by the problem identification unit 256 and the relevance of the problems related to the patient, thereby evaluating each medication instruction statement related to the problem related to the patient.

[0090] In step S76, the support information generation unit 258 acquires medication instruction sentences from the drug DB 176 as medication instruction content (support information) recommended for the patient, based on the evaluation value of the medication instruction sentence determined by the presented information evaluation unit 257. That is, the support information generation unit 258 acquires medication instruction sentences evaluated based on the relevance of the patient's problem, and generates data of a list of recommended medication instruction sentences by arranging medication instruction sentences that have been assigned an evaluation value equal to or greater than a preset threshold in descending order of evaluation value. Furthermore, the support information generation unit 258 acquires medication instruction sentences that have been assigned an evaluation value less than a preset threshold by the presented information evaluation unit 257, and arranges them in a predetermined order to generate data of a list of general medication instruction sentences. As a result, support information including data of the list of recommended medication instruction sentences and data of the list of general medication instruction sentences is generated.

[0091] In step S77, the support information providing unit 259 transmits the support information input from the support information generating unit 258 (information including data on problems related to the patient and the relationships between problems related to the patient, data on a list of recommended medication instructions, and data on a list of general medication instructions) to the terminal device 10 (output to the terminal device 10 via the network 40). After step S77, the support information providing process is repeated.

[0092] [Reconstruction process] FIG. 24 is a flowchart showing the flow of the reconstruction process executed by the support information providing server 30. The reconstruction process is started in response to an instruction to execute the reconstruction process being input via the input unit 815 of the support information providing server 30. Note that the reconstruction process may be executed every time the support information providing process is executed, or at a preset time (for example, 3:00 AM).

[0093] In step S81, the reconstruction executing unit 260 acquires data on the hearing contents in the consideration information acquired in providing support information for a plurality of patients, and the generated intermediate state (problem related to the patient). In step S82, the reconstruction execution unit 260 reconstructs the hearing content based on the data of the hearing content in the consideration information received by the support request receiving unit 253 and the intermediate state (problem related to the patient) generated by the problem identification unit 256. After step S82, the reconstruction process ends.

[0094] As described above, in the support information providing system 1 according to this embodiment, when a patient for whom medication instruction is to be provided visits the pharmacy for the first time, the following information is acquired as the consideration target information: data on the patient's attributes who brought the prescription, data on the medical interview results entered on the medical interview display screen, prescription data indicating the contents of the prescription brought by the patient, etc. Furthermore, when the patient is a returning visitor, the following information is acquired as the consideration target information: data on the patient's attributes who brought the prescription, data on the medical interview results entered on the medical interview display screen, prescription data indicating the contents of the prescription brought by the patient, data on the medication instruction content from the previous visit, etc. Then, features are extracted from the consideration target information, and machine learning-based inference is performed using the features as inputs to generate an intermediate state represented by an index of the patient's problem. At this time, the relevance of the patient's problem to the consideration target information is also generated. Furthermore, by performing rule-based inference based on the patient's problems and the relevance of the patient's problems, each medication instruction statement related to the patient's problems is evaluated, and support information including recommended medication instructions is provided based on the evaluation results.

[0095] Therefore, by performing a first inference (here, machine learning-based inference) from the information under consideration, an intermediate state (patient problem) can be generated that embodies as information one state of the thought process of the pharmacist providing medication instruction to the patient. Furthermore, by performing a second inference (here, rule-based inference) from the intermediate state (patient problem), support information including recommended medication instruction text is generated. Furthermore, for patients who return to the pharmacy, medication instruction content inferred from the previous medication instruction content and the results of this interview is recommended with a higher evaluation value. Therefore, it becomes possible to support medication instruction using computer-based inference, thereby more appropriately supporting the work of pharmacists. Furthermore, when a patient returns to the pharmacy, it is possible to recommend medication instructions that are appropriate for the patient's current condition, such as the prescribed medication, the medication instructions given to the patient in the past, and the patient's symptoms. Furthermore, when medication instruction is assisted by a computer, appropriate medication instruction statements can be inferred using a procedure that conforms to the thought process of a pharmacist. In addition, when inference is performed using deep learning or the like, it is generally difficult to grasp the inference process, making it difficult to determine the basis for the inference results. However, by performing two-stage inference via an intermediate state, as in this embodiment, it is possible to confirm the basis for the recommended medication instructions.

[0096] [Variation 1] In the above-described embodiment, a client-server type support information providing system 1 is constructed, and a request for support information is made from the terminal device 10 to the support information providing server 30 to obtain the support information. In contrast, by providing the support information provision function of the support information providing server 30 in a single device (e.g., a terminal device 10 or a pharmacy computer 20, etc.), the functions of the support information providing system 1 can be realized in a single information processing device 800 (i.e., realized as a stand-alone system).

[0097] FIG. 25 is a block diagram showing the functional configuration of a stand-alone information processing device 800 having a support information providing function. As shown in FIG. 25, when configured as a standalone type, in a single information processing device 800, the functions of the UI display control unit 51 of the terminal device 10, the UI information generation unit 252 of the support information providing server 30, the support request reception unit 253, the target information acquisition unit 254, the feature extraction unit 255, the problem identification unit 256, the presented information evaluation unit 257, the support information generation unit 258, the support information provision unit 259 and the reconstruction execution unit 260 are provided in a CPU 811, and each database managed by the support information providing server 30 (or the pharmacy computer 20) is provided in a memory unit 817.

[0098] Furthermore, when the support information providing system 1 is configured as a client-server system, the combination of information processing devices that configure the system is not limited to the examples shown in the above-described embodiments. For example, it is possible to distribute the functions of the pharmacy computer 20 or the support information providing server 30 across more servers, or to integrate the functions of the pharmacy computer 20 and the support information providing server 30 into a single server.

[0099] [Variation 2] In the above embodiment, an example has been described in which six types of problems shown in FIG. 9 are displayed as problems relating to patients. In contrast, it is possible to display problems relating to a wider variety of patients. It is also possible to sequentially set the number of problems related to the patient according to the patient attributes and prescription data, and display intermediate states. Furthermore, when generating a problem relating to a patient, it is possible to provide a range of content for the problem relating to the patient, such as making the content more specific or more abstract. This makes it possible to more appropriately represent the thought process that pharmacists use when providing medication instructions to patients.

[0100] [Variation 3] In the above-described embodiment, a first stage of inference is performed from the information under consideration to generate an intermediate state (a problem relating to the patient) that embodies as information one state of the thought process in which a pharmacist provides medication instructions to a patient, and a second stage of inference is performed from the intermediate state (the problem relating to the patient) to generate support information including recommended medication instructions. In contrast to this, by performing inference at more stages, it is possible to generate intermediate states at multiple stages and display them appropriately.

[0101] For example, one state of the thought process in which a drug provides medication instructions to a patient can be set to n stages (n is an integer of 2 or greater) according to the level of abstraction, and n intermediate stages that embody these can be generated. In this case, from the information to be considered, the first intermediate stage state is generated by a first inference, the second intermediate stage state is generated from the first intermediate stage state by a second inference, and the nth intermediate stage state is generated sequentially by an nth inference. Then, support information including recommended medication instructions can be generated from the nth intermediate stage state by an n+1th inference. As an example, when generating an intermediate state, machine learning-based inference can be performed, and when generating support information including recommended medication instructions, rule-based inference can be performed. This makes it possible to more specifically trace the thought process of pharmacists (especially experienced pharmacists) when providing medication instructions to patients, while dividing the pharmacist's thought process into more appropriate stages and showing intermediate states (problems related to the patient). Therefore, it becomes possible to more appropriately support the work of pharmacists.

[0102] As described above, the support information providing system 1 according to this embodiment includes the terminal device 10 and the support information providing server 30. The terminal device 10 and the support information providing server 30 are configured to be able to communicate with each other via a network 40 . The terminal device 10 includes a UI display control unit 51 and a support request unit 53, and the support information providing server 30 includes a target information acquisition unit 254, a problem identification unit 256, a support information generation unit 258, and a support information providing unit 259. In the terminal device 10, the support request unit 53 sends a request to the support information providing server 30 to generate support information for pharmaceutical guidance based on target information that is to be considered in providing pharmaceutical guidance to the patient and includes the content of the previous pharmaceutical guidance to the patient and the results of the current interview. The UI display control unit 51 displays the support information transmitted from the support information providing server 30 in response to a request from the support request unit 53 . In the support information providing server 30, the target information acquiring unit 254 acquires the consideration target information based on the request from the support request unit 53. The problem identification unit 256 generates support information including the content of recommended pharmaceutical guidance by making inferences based on the information to be considered. The support information providing unit 259 provides the support information generated by the problem identifying unit 256 to the terminal device 10 . As a result, when a patient who is the subject of pharmaceutical guidance is a returning patient, the content of the recommended pharmaceutical guidance is inferred based on the information to be considered, including the content of the previous pharmaceutical guidance given to the patient and the results of the current interview. Therefore, it becomes possible to provide medication instructions that are appropriate for the patient's current condition, thereby more appropriately supporting the work of pharmacists.

[0103] The problem identifying unit 256 generates, as support information, information representing an intermediate state expressed using pharmacological judgment factors as indicators when providing medication instruction, by performing a first inference based on the information to be considered. This makes it possible to generate an intermediate state (e.g., a problem related to the patient) that embodies as information one state of the thought process in which a pharmacist provides medication instructions to a patient by performing a first inference (e.g., machine learning-based inference) from the information under consideration.

[0104] The support information generating unit 258 generates the content of recommended pharmaceutical guidance as support information by performing a second inference based on the information representing the intermediate state. This makes it possible to generate support information including recommended medication instructions by performing a second inference (for example, rule-based inference) from an intermediate state (for example, a problem related to a patient).

[0105] The support information providing server 30 includes a feature extraction unit 255 . The feature extraction unit 255 extracts feature amounts from the consideration target information acquired by the target information acquisition unit 254 based on preset extraction conditions. The problem identifying unit 256 generates information representing an intermediate state by performing a first inference based on the feature amount extracted from the information under consideration. This makes it possible to generate an intermediate state that more accurately reflects the content expressed by the consideration target information.

[0106] The first inference involves machine learning-based inference that uses the information under consideration as input and uses as training data pharmaceutical decision-making factors that come to mind about the patient when a pharmacist recognizes the target information. This allows intermediate states to be generated that reflect the pharmacist's thinking.

[0107] The second inference includes rule-based inference based on association of preset support information with information representing an intermediate state. This allows support information to be generated from an intermediate state by simple and clear processing.

[0108] The terminal device 10 or the support information providing server 30 according to this embodiment also includes a problem identifying unit 256 , a support information generating unit 258 , and a support information providing unit 259 . The problem identification unit 256 or the support information generation unit 258 generates support information for pharmaceutical guidance including the content of recommended pharmaceutical guidance by making inferences based on target information that is taken into consideration when providing pharmaceutical guidance to the patient and includes the content of the previous pharmaceutical guidance to the patient and the results of the current interview. The support information providing unit 259 outputs the support information generated by the problem identifying unit 256 or the support information generating unit 258 . As a result, when a patient who is the subject of pharmaceutical guidance is a returning patient, the content of the recommended pharmaceutical guidance is inferred based on the information to be considered, including the content of the previous pharmaceutical guidance given to the patient and the results of the current interview. Therefore, it becomes possible to provide medication instructions that are appropriate for the patient's current condition, thereby more appropriately supporting the work of pharmacists.

[0109] The present invention is not limited to the above-described embodiment, and any modifications and improvements that can achieve the object of the present invention are included in the present invention. For example, in the above embodiment, a list of medication instructions is presented as support information, but the present invention is not limited to this. That is, the objects presented as support information in the present invention include various types of medical information.

[0110] Furthermore, in the above-described embodiment, after the patient has been interviewed, the patient waits while the pharmacist dispenses the medication. During this waiting time, simple lifestyle guidance information identified from the interview results may be presented to the patient (for example, displayed on the patient's smartphone, etc.). Furthermore, in the above-described embodiment, a display screen for conducting a medical interview with the patient and a display screen for listing medication instructions (a list of recommended medication instructions for patients and a list of general medication instructions) after the contents have been confirmed by the pharmacist may be displayed on a device carried by the patient (for example, the patient's smartphone, etc.). Furthermore, problems related to the patient after the contents have been confirmed by the pharmacist may also be displayed on a device carried by the patient (for example, the patient's smartphone, etc.).

[0111] In the above embodiment, the patient-related problems are shown as a circular graph as shown in Fig. 9, but the output format of the patient-related problems is not limited to this. For example, it is possible to output numerical values ​​representing the relevance of the patient-related problems as a radar chart. In this case, it becomes easy to display changes in the patient-related problems from past medication instructions.

[0112] The above-described series of processes can be executed by hardware or software. In other words, the functional configurations in the above-described embodiments are merely examples and are not particularly limited. That is, it is sufficient that any of the computers constituting the support information providing system 1 has a function capable of executing the above-described series of processes as a whole, and the functional blocks used to realize this function are not particularly limited to the examples shown. Furthermore, one functional block may be configured as a single piece of hardware, a single piece of software, or a combination thereof.

[0113] Furthermore, the recording medium containing the program for executing the above-mentioned series of processes may be configured not only as a removable medium distributed separately from the device main body in order to provide the program to the user, but also as a recording medium provided to the user in a state where it is pre-installed in the device main body.

[0114] Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments. Furthermore, the effects described in the present embodiments are merely a list of the most preferable effects resulting from the present invention, and the effects of the present invention are not limited to those described in the present embodiments. [Explanation of symbols]

[0115] 1 Support information providing system, 10 Terminal device, 20 Pharmacy computer, 30 Support information providing server, 40 Network, 51 User interface display control unit (UI display control unit), 52, 254 Target information acquisition unit, 53 Support request unit, 54 Support information acquisition unit, 151 Receipt management unit, 152 Medication history management unit, 153, 251 DB management unit, 171, 271 Patient attribute database (Patient attribute DB), 172, 272 Prescription database (Prescription DB), 173, 273 Medication history database (Medication history DB), 174, 274 Interview result database (Interview result DB), 175, 275 Hearing content database (Hearing content DB), 176, 276 Drug database (Drug DB), 252 User interface information generation unit (UI information generation unit), 253 Support request reception unit, 255 Feature extraction unit, 256 Problem identification unit, 257 presentation information evaluation unit, 258 support information generation unit, 259 support information provision unit, 260 reconstruction execution unit, 800 information processing device, 811 CPU, 812 ROM, 813 RAM, 814 bus, 815 input unit, 816 output unit, 817 storage unit, 818 communication unit, 819 drive, 820 imaging unit, 831 removable media

Claims

1. A support information providing system including a terminal device and a server configured to be able to communicate via a network, The terminal device a support request means for transmitting to the server a request for generating support information for pharmaceutical guidance based on target information that is to be considered in pharmaceutical guidance for the patient and includes the content of the previous pharmaceutical guidance for the patient and the result of the current interview; a support information display means for displaying the support information transmitted from the server in response to a request from the support request means, The server a target information acquisition means for acquiring the target information based on a request from the support request means; an inference means for generating the support information including the content of recommended pharmaceutical guidance by making an inference based on the target information; a support information providing means for providing the support information generated by the inference means to the terminal device; Equipped with The inference means performs a first inference based on the target information, thereby generating, as the support information, information representing an intermediate state represented using pharmaceutical judgment factors as indicators when providing medication instructions, and reconstructs the content of interviews to be conducted with the patient from the next time onwards based on the generated information representing the intermediate state.

2. The support information providing system according to claim 1, wherein the inference means generates the content of recommended pharmaceutical guidance as the support information by performing a second inference based on the information representing the intermediate state.

3. The server a feature extraction means for extracting feature amounts from the target information acquired by the target information acquisition means based on a preset extraction condition; 3. The support information providing system according to claim 1, wherein the inference means generates information representing the intermediate state by performing the first inference based on the feature extracted from the target information.

4. The support information providing system according to any one of claims 1 to 3, characterized in that the first inference includes a machine learning-based inference constructed using the target information as input and pharmaceutical judgment factors that come to mind regarding a patient when a pharmacist recognizes the target information as training data.

5. 3. The support information providing system according to claim 2, wherein the second inference includes rule-based inference based on a correspondence between the support information and the information representing the intermediate state that is set in advance.

6. an inference means for generating support information for pharmaceutical guidance including the content of recommended pharmaceutical guidance by making inferences based on target information including the content of the previous pharmaceutical guidance given to the patient and the results of the current interview, which are to be taken into consideration in pharmaceutical guidance for the patient; a support information output means for outputting the support information generated by the inference means; Equipped with The inference means performs a first inference based on the target information, thereby generating, as the support information, information representing an intermediate state represented using pharmaceutical judgment factors as indicators when providing medication instructions, and reconstructs the content of interviews to be conducted with the patient from the next time onwards based on the generated information representing the intermediate state.

7. an inference step of generating support information for pharmaceutical guidance including the content of recommended pharmaceutical guidance by making inferences based on target information that is to be taken into consideration in pharmaceutical guidance for the patient and includes the content of the previous pharmaceutical guidance for the patient and the results of the current interview; a support information output step of outputting the support information generated in the inference step; Including, In the inference step, a first inference is made based on the target information, thereby generating, as the support information, information representing an intermediate state represented using pharmaceutical judgment factors as indicators when providing medication instructions, and reconstructing the content of interviews to be conducted with the patient from the next time onwards based on the generated information representing the intermediate state.

8. On the computer, an inference function that generates support information for pharmaceutical guidance, including the content of recommended pharmaceutical guidance, by making inferences based on target information that is to be taken into consideration in pharmaceutical guidance for a patient and includes the content of the previous pharmaceutical guidance for the patient and the results of the current interview; and a support information output function that outputs the support information generated by the inference function; Realize this, The inference function is a program characterized in that it generates, as the support information, information representing an intermediate state represented using pharmaceutical judgment factors as indicators when providing medication instructions by performing a first inference based on the target information, and reconstructs the content of interviews to be conducted with the patient from the next time onwards based on the information representing the intermediate state generated.

Images