Computer program, information processing device, and information processing method

Abstract

This invention provides a computer program, an information processing device, and an information processing method that can record nursing records and ward logs in an analyzable format. [Solution] The computer program acquires patient information, accepts input of record information about the patient, generates a patient database based on the acquired patient information and the input record information, and by referring the patient information and record information recorded in the patient database with the acquired patient information and the accepted record information, it determines whether there is any omission in the record information to be recorded in the patient database. If it is determined that there is a omission in the record information, it accepts input of the omission in the record information and regenerates the patient database, and generates nursing records and ward logs based on the regenerated patient database.

Description

Technical Field

[0001] The present invention relates to a computer program, an information processing apparatus, and an information processing method.

Background Art

[0002] The main duties of a nurse include various medical acts and patient care, such as regular measurement of a patient's vital signs, and medical treatments such as medication and injection. Also, creating a nursing record that describes the care content and the content of medical acts performed on the patient when performing nursing duties, and creating documents such as a ward log are also essential duties of a nurse.

[0003] Patent Document 1 discloses a nursing record creation support apparatus that can recognize nursing information input by a nurse as line drawing data, edit it as character information, and create a nursing record.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, when a nurse inputs the content of medical acts and care performed on a patient, the items of patient information that need to be input can vary depending on various situations such as the patient's profile, symptoms on that day, and environment. As a result, there may be omissions in the patient information for creating a nursing record or a ward log, and it cannot be correctly analyzed as medical information. Therefore, it is desired that the nursing record and the ward log be recorded in an analyzable form.

[0006] The present invention has been made in view of such circumstances, and an object thereof is to provide a computer program, an information processing apparatus, and an information processing method capable of recording a nursing record and a ward log in an analyzable form. [Means for solving the problem]

[0007] (1) The computer program according to the present invention obtains patient information of a patient, accepts input of record information relating to the patient, generates a patient database based on the obtained patient information and the input record information, determines whether there is any omission in the record information to be recorded in the patient database by referring to the patient information and record information recorded in the patient database with the obtained patient information and the accepted record information, and if it is determined that there is a omission in the record information, accepts input of the omission in the record information to regenerate the patient database, and generates nursing records and ward logs based on the regenerated patient database, causing the computer to execute these processes.

[0008] Herein, embodiments of the present invention are as follows: (2) The computer program described in (1) above causes the computer to display a selection of candidate records to be recorded in the patient database and to obtain the selected candidate from the displayed options as the patient information.

[0009] (3) The computer program described in (2) above causes the computer to perform a process that generates a selection of candidate record information to be displayed in the patient database based on the difference between the record information recorded in the patient database and the input record information.

[0010] (4) If any of the computer programs described in (1) to (3) above is missing any recorded information, the computer program will cause the computer to perform the following actions: display a list of candidates for the missing recorded information, accept the selection of a displayed candidate, and accept the selected candidate as input for the missing recorded information.

[0011] (5) Any one of the computer programs described in (1) to (4) above will input the acquired patient information into a language model, obtain the candidate for missing record information generated by the language model, and have the computer perform the following processing: input the missing record information from the acquired candidates.

[0012] (6) Any one of the computer programs described in (1) to (5) above causes the computer to perform a process to acquire at least one of the patient's symptoms, vital information, medication information, and treatment information as the patient information.

[0013] (7) Any one of the computer programs described in (1) to (6) above inputs the symptoms obtained as patient information of the patient into a language model, and if the language model determines that there is a gap in the recorded information, it obtains the missing recorded information relating to subjective information generated by the language model, and causes the computer to perform the following processing: accept input of the obtained missing recorded information relating to subjective information.

[0014] (8) Any one of the computer programs described in (1) to (7) above inputs at least one of the vital information, medication information, and treatment information obtained as patient information of the patient into a language model, and if the language model determines that there is a gap in the recorded information, it obtains the missing recorded information relating to objective information generated by the language model, and causes the computer to perform a process that accepts input of the missing recorded information relating to the obtained objective information.

[0015] (9) Any one of the computer programs described in (1) to (8) above inputs the recorded information, which includes at least one of the patient's vital information, medication information, and treatment information, recorded in the generated patient DB, into a language model, obtains the evaluation information included in the nursing record generated by the language model, and causes the computer to perform the following processes: generate a nursing record using the obtained evaluation information.

[0016] (10) Any one of the computer programs described in (1) to (9) above causes the computer to perform a process to acquire at least one of the following as patient information: patient movement, severity classification, patient condition, specific medical procedure, use of specific medical device, and the condition of the patient requiring care.

[0017] (11) Any one of the computer programs described in (1) to (10) above inputs patient information for each medical facility into a language model, obtains candidate records of the record information generated by the language model, and causes the computer to perform a process of updating at least one of the record items in the medical facility's patient database and the record information recorded in the record items based on the obtained candidates.

[0018] (12) The information processing device according to the present invention includes a control unit, which acquires patient information of a patient, accepts input of record information relating to the patient, generates a patient database based on the acquired patient information and the input record information, determines whether there is any omission in the record information to be recorded in the patient database by referring to the patient information and record information recorded in the patient database and the acquired patient information and the accepted record information, and if it is determined that there is a omission in the record information, accepts input of the omission in the record information to regenerate the patient database, and generates nursing records and ward logs based on the regenerated patient database.

[0019] (13) The information processing method according to the present invention acquires patient information of a patient, accepts input of record information relating to the patient, generates a patient database based on the acquired patient information and the input record information, determines whether there is any omission in the record information to be recorded in the patient database by referring to the patient information and record information recorded in the patient database and the acquired patient information and the accepted record information, if it is determined that there is a omission in the record information, accepts input of the omission in the record information to regenerate the patient database, and generates nursing records and ward logs based on the regenerated patient database. [Effects of the Invention]

[0020] According to the present invention, nursing records and ward logs can be recorded in an analyzable format. [Brief explanation of the drawing]

[0021] [Figure 1] This figure shows an example of the configuration of the information processing system of this embodiment. [Figure 2]This is a diagram showing an example of information recorded in the medical information DB. [Figure 3] This is a diagram showing an example of the recorded information in the patient DB. [Figure 4] This is a diagram showing an example of the configuration of the patient DB. [Figure 5] This is a diagram showing an example of the method of recording the recorded information in the patient DB. [Figure 6] This is a diagram showing a first example of determining the presence or absence of leakage of the recorded information in the patient DB. [Figure 7] This is a diagram showing a second example of determining the presence or absence of leakage of the recorded information in the patient DB. [Figure 8] This is a diagram showing an example of the recorded information regarding the subjective information recorded in the patient DB. [Figure 9] This is a diagram showing an example of the recorded information regarding the objective information recorded in the patient DB. [Figure 10] This is a diagram showing an example of the configuration of the nursing record DB. [Figure 11] This is a diagram showing an example of the method of generating the subjective information of the nursing information. [Figure 12] This is a diagram showing an example of the subjective information of the nursing record. [Figure 13] This is a diagram showing an example of the method of generating the objective information of the nursing information. [Figure 14] This is a diagram showing an example of the objective information of the nursing record. [Figure 15] This is a diagram showing an example of the method of identifying a disease. [Figure 16] This is a diagram showing an example of the recorded information serving as the basis for a disease. [Figure 17] This is a diagram showing a specific example of the recorded information serving as the basis for a disease. [Figure 18] This is a diagram showing an example of the method of generating evaluation information. [Figure 19] This is a diagram showing an example of evaluation information. [Figure 20] This is a diagram showing an example of the method of generating nursing plan information. [Figure 21] This is a diagram showing an example of nursing plan information. [Figure 22] This figure shows an example of the record information that is recorded in the patient database when generating ward logs. [Figure 23] This figure shows an example of how patient database information is recorded. [Figure 24] This figure shows an example of the structure of a ward log database. [Figure 25] This figure shows an example of a ward log. [Figure 26] This figure shows an example of a processing procedure performed by an information processing device.

[0022] Embodiments of the present invention will be described below with reference to the drawings. Figure 1 is a diagram showing an example of the configuration of the information processing system of this embodiment. The information processing system includes an information processing device 50. The information processing device 50 is connected to a large language model (LLM) 100 and a data server 120 via a communication network 1.

[0023] The large-scale language model (also called the "language model") 100 is a machine learning model that utilizes natural language processing techniques to learn language patterns from existing text data and generate and understand text such as sentences and dialogues. The large-scale language model 100 divides text data into small chunks and performs embedding on each chunk to generate a vector representation (embending) of each chunk. The embedding process includes numerical representations of the chunk type and numerical representations of the chunk's positional relationship. As a result, each chunk is represented as a point in a high-dimensional space, where chunks with similar meanings are placed close to each other, and chunks with different meanings are placed far apart. The large-scale language model 100 can numerically capture the semantic relationships and similarities between chunks and understand the semantics, i.e., meaning, of text data. Furthermore, the large-scale language model 100 is equipped with an attention mechanism, and by appropriately adjusting the magnitude of the linear combination coefficient when linearly combining the numerical representations of each chunk calculated by the embedding process, it can obtain a numerical representation that expresses the complex structure of a sentence.

[0024] The large-scale language model 100 is composed of a deep neural network and can use models such as GPT-4, GPT-3.5, BERT, LaMDA, PaLM, LLaMA, or new models that will be developed and put into use in the future. The large-scale language model 100 of this embodiment can provide appropriate dialogue content for the patient in response to the content of the dialogue with the patient.

[0025] The data server 120 is equipped with a medical information database 121.

[0026] Figure 2 shows an example of the information recorded in the medical information DB121. The medical information DB121 contains information such as disease-specific response manuals, disease-specific treatment protocols (procedures), disease-specific clinical guidelines, nursing records, and ward logs.

[0027] Disease-specific response manuals include information such as key points for early detection and early intervention (key initial symptoms, peak onset periods, responses by healthcare professionals, etc.), an overview of side effects, criteria for identifying side effects (methods of identification), diseases requiring identification and methods of identification, treatment methods (main treatment methods when side effects occur), typical cases, and reference materials.

[0028] Disease-specific treatment protocols are, for example, predetermined standard treatment methods for specific diseases, including regulations, procedures, tests, and treatment plans. A protocol is a detailed set of instructions outlining its purpose, rationale, and methods.

[0029] Disease-specific treatment guidelines are documents that, for example, support decision-making by healthcare users and providers regarding important health issues. They evaluate the overall evidence (scientific basis) through systematic reviews and present recommendations that are considered optimal, taking into account the balance of benefits and harms. They are documents that present the most optimal treatments, etc., for various health-related issues based on evidence.

[0030] Nursing records are a collection of nursing records created in the past at various medical institutions.

[0031] The ward logs are a collection of ward logs created in the past at various medical institutions.

[0032] The large-scale language model 100 may be a language model that has learned from information recorded in the medical information DB 121 and is tuned to understand the contents of disease-specific response manuals, disease-specific treatment protocols (procedures), disease-specific clinical guidelines, nursing records, and ward logs.

[0033] The information processing device 50 includes a control unit 51 that controls the entire device, a communication unit 52, a memory 53, a display unit 54, an operation unit 55, a storage unit 56, an interface unit 58, and a recording medium reading unit 59, among others.

[0034] The storage unit 56 can be made up of semiconductor memory or a hard disk, and stores a computer program 57 (program product) and necessary information.

[0035] The control unit 51 may be configured by incorporating a required number of CPUs (Central Processing Units), MPUs (Micro-Processing Units), GPUs (Graphics Processing Units), etc. Alternatively, the control unit 51 may be configured by combining DSPs (Digital Signal Processors), FPGAs (Field-Programmable Gate Arrays), etc.

[0036] The communication unit 52 is equipped with a communication module and has the function of communicating with the large-scale language model 100, the data server 120, and other devices (not shown) via the communication network 1.

[0037] The display unit 54 is composed of a liquid crystal display or an organic EL display, and provides a user interface (UI) to the user by displaying the required information. Alternatively, an external display device may be provided instead of the display unit 54.

[0038] The operation unit 55 is, for example, a touch panel and can perform operations such as operating icons displayed on the display unit 54, moving and manipulating the cursor, and inputting text. The operation unit 55 may be composed of buttons and switches, or it may be composed of a keyboard and mouse. The operation unit 55 provides a user interface (UI) to the user by accepting user input. An external input device may be provided instead of the operation unit 55.

[0039] The interface unit 58 provides interface functions for accessing the patient database 70, nursing record database 80, and ward log database 90. Through the interface unit 58, information can be written to the patient database 70, nursing record database 80, and ward log database 90, and information can be read from the patient database 70, nursing record database 80, and ward log database 90.

[0040] The computer program 57 can be read by the recording medium reading unit 59 from a recording medium (e.g., an optically readable disc storage medium such as a CD-ROM) M and stored in the storage unit 56. The computer program 57 may also be read from a recording medium such as a storage device (semiconductor memory such as an SSD (Solid State Drive)) connected by a standard for connecting to a computer (e.g., USB (Universal Serial Bus) or other standards) and stored in the storage unit 56. Alternatively, the computer program 57 may be downloaded from an external device via the communication unit 52 and stored in the storage unit 56.

[0041] The memory 53 can be composed of semiconductor memory such as SRAM (Static Random Access Memory), DRAM (Dynamic Random Access Memory), or flash memory. The computer program 57 is loaded into the memory 53, and the control unit 51 can execute the computer program 57. The control unit 51 can execute the processing defined in the computer program 57. In other words, the processing performed by the control unit 51 is also the processing performed by the computer program 57.

[0042] Figure 3 shows an example of the recorded information stored in the patient database 70. The patient database 70 stores recorded information such as the patient's symptoms, vital signs, medication information, and treatment information, associated with each patient. Patient symptoms include, but are not limited to, chest symptoms (such as discomfort, pain, palpitations), abdominal symptoms (such as pain), and head symptoms (such as dizziness, pain, and lightheadedness).

[0043] Vital information includes, but is not limited to, blood pressure, pulse, heart rate, SpO2 (oxygen saturation), respiration, and body temperature. Medication information includes, but is not limited to, information on medications prescribed to the patient, including, but is not limited to, nitroglycerin, myrisrol (vasodilator), Inovan, Doptrex, norepinephrine (blood pressure raiser), and heparin (blood thinner). Procedure information includes, but is not limited to, blood sampling, electrocardiogram, puncture site confirmation, TR band adjustment, and oxygen administration.

[0044] As shown in Figure 3, for patient ID 01, symptoms can be represented by SyA, vital information by ViA, medication information by DoA, and treatment information by TrA. SyA represents a set of multiple symptoms grouped together as the code SyA for convenience, ViA represents a set of multiple vital information grouped together as the code ViA for convenience, DoA represents a set of multiple medication information grouped together as the code DoA for convenience, and TrA represents a set of multiple treatment information grouped together as the code TrA for convenience. Similarly, for patient ID 02, symptoms can be represented by SyB, vital information by ViB, medication information by DoB, and treatment information by TrB. For patient ID 03, symptoms can be represented by SyC, vital information by ViC, medication information by DoC, and treatment information by TrC. The same applies to other patients.

[0045] Figure 4 shows an example of the configuration of the patient database 70. The data items in the patient database include symptoms, vital signs, medication information, and treatment information. Symptoms include Sy1, Sy2, Sy3, etc. Symptoms Sy1, Sy2, Sy3, etc. include, but are not limited to, cough, fever, chest pain, shortness of breath, palpitations, etc.

[0046] Vital information includes vital signs Vi1, Vi2, Vi3, etc. These vital signs include, but are not limited to, blood pressure, pulse, heart rate, SpO2, respiration, and body temperature.

[0047] Medication information includes medication information Do1, Do2, Do3, etc. Medication information Do1, Do2, Do3, etc. include, but are not limited to, nitroglycerin, myrisrol, inovan, doptrex, norepinephrine, heparin, etc.

[0048] The procedure information includes procedure information Tr1, Tr2, Tr3, etc. Procedure information Tr1, Tr2, Tr3, etc. may include, but are not limited to, blood sampling, electrocardiogram, puncture site confirmation, TR band adjustment, oxygen administration, etc.

[0049] The information recorded in the patient database 70 can be recorded chronologically. For example, it can be recorded chronologically as follows: day of treatment (DAY0), day 1 of hospitalization (DAY1), day 2 of hospitalization (DAY2), and so on.

[0050] As shown in Figure 4, the recorded information for symptom Sy1 in the data items can be recorded chronologically in the order of treatment day (DAY0), hospitalization day 1 (DAY1), hospitalization day 2 (DAY2), etc., as recorded information ReS10, ReS11, ReS12, ... Similarly, the recorded information for symptom Sy2 in the data items can be recorded chronologically in the order of treatment day (DAY0), hospitalization day 1 (DAY1), hospitalization day 2 (DAY2), etc., as recorded information ReS20, ReS21, ReS22, ... And the recorded information for symptom Sy3 in the data items can be recorded chronologically in the order of treatment day (DAY0), hospitalization day 1 (DAY1), hospitalization day 2 (DAY2), etc., as recorded information ReS30, ReS31, ReS32, ...

[0051] The same applies to vital signs information (Vi1, Vi2, Vi3, ...), medication information (Do1, Do2, Do3, ...), and treatment information (Tr1, Tr2, Tr3, ...) as to symptoms (Sy1, Sy2, Sy3, ...).

[0052] Records regarding symptoms Sy1, Sy2, Sy3, ... are subjective information, while records regarding vital signs Vi1, Vi2, Vi3, ..., medication information Do1, Do2, Do3, ..., and treatment information Tr1, Tr2, Tr3, ... are objective information.

[0053] Figure 5 shows an example of how patient database information is recorded. When entering the record information ReS10 for the day of treatment (DAY0) for symptom Sy1, as shown in Figure 5, three options for record information ReS10 (candidates 1, 2, and 3) are displayed, allowing the user (nurse) to select one. In this way, instead of manually entering the record information directly, the user can select the required record information from the displayed options and record it in the patient database. Similarly, options for other symptoms, vital signs, medication information, and treatment information can be displayed and selected by the user. In this way, record information can be recorded in the patient database.

[0054] As described above, the control unit 51 can display a selection of candidate record information to be recorded in the patient database 70, and can acquire the selected candidate from the displayed options as patient information. Patient information can include at least one of the patient's symptoms, vital signs, medication information, and treatment information. This allows patient information to be entered into the patient database 70 table, eliminating the need to enter all patient information in text form. Record information in the patient database 70 is recorded simply by selecting the required option from the displayed choices. Patient information may also be entered manually.

[0055] Furthermore, the control unit 51 can generate a selection of candidate record information to be recorded in the displayed patient database based on the difference between the record information recorded in the patient database and the input record information.

[0056] Furthermore, since the system uses a selection-based input format, text-based input can be reduced. Duplicate input of the same information is eliminated, reducing the effort required for data entry. Additionally, the standardization of the recorded information allows for the creation of an analyzable patient database, enabling the acquisition of future clinical insights (business improvement). Moreover, the elimination of ambiguity in recorded information facilitates accurate information sharing, leading to more appropriate treatment.

[0057] Furthermore, nursing records and ward logs may vary in content and expression due to differences in the work experience and writing skills of the nurses who actually created them. These variations in content and expression can lead to errors when analyzing the information as medical data, preventing accurate analysis. By using a format where users select the appropriate item from a set of options, errors in analyzing the information as medical data can be reduced.

[0058] Next, we will explain how to determine whether or not there are any omissions in the recorded information stored in the patient database 70.

[0059] Figure 6 shows a first example of determining whether there are any omissions in the recorded information stored in the patient database 70. In this first example, the determination is made whether there are any omissions in the patient's symptoms among the patient information as recorded information. As illustrated in Figures 4 and 5, the information processing device 50 obtains patient information (patient's symptoms) by inputting the recorded information into the table of the patient database 70.

[0060] The information processing device 50 inputs the acquired patient information (symptoms) and prompt P1 to the large-scale language model 100. Prompt P1 can be, for example, "Based on the input patient information (symptoms), determine whether there is any missing record information regarding subjective information. If there is missing record information, generate and present candidates for the missing record information." but is not limited to this. The large-scale language model 100 generates candidates for missing record information regarding subjective information based on the input patient information (symptoms) and outputs the generated candidates for missing record information regarding subjective information to the information processing device 50. The record information regarding subjective information is record information regarding the patient's symptoms.

[0061] As described above, the control unit 51 determines whether there is any missing recorded information regarding the patient's symptoms and subjective information acquired as patient information, and if there is any missing recorded information, it inputs a prompt to the language model instructing it to generate the missing recorded information. If the language model determines that there is missing recorded information, the control unit 51 can acquire the missing recorded information regarding subjective information generated by the language model and accept input of the acquired missing recorded information regarding subjective information.

[0062] The information processing device 50 displays candidates for missing subjective information on the display unit 54. This allows the user to know, through the candidates for missing information, that there was a omission in the record information that should have been recorded in the patient DB 70, and what kind of record information was missing.

[0063] When a user inputs missing record information related to subjective information, the information processing device 50 can generate the patient DB 70 by writing the acquired patient information (symptoms) and the record information related to subjective information, which includes the input missing record information, to the patient DB 70.

[0064] As described above, if there is any missing recorded information, the control unit 51 can display a selection of candidates for the missing recorded information, accept the selection of a displayed candidate, and accept the selected candidate as input for the missing recorded information.

[0065] Furthermore, the control unit 51 inputs the symptoms acquired as patient information of the patient into a language model (large-scale language model 100), and if the language model determines that there is a gap in recorded information, it can acquire the missing recorded information related to subjective information generated by the language model and accept input of the acquired missing recorded information related to subjective information.

[0066] Figure 7 shows a second example of determining whether there are any omissions in the recorded information stored in the patient database 70. In this second example, the system determines whether there are any omissions in the patient's vital signs, medication information, and treatment information among the patient information. As illustrated in Figures 4 and 5, the information processing device 50 acquires patient information (patient's vital signs, medication information, and treatment information) by inputting the recorded information into the table of the patient database 70.

[0067] The information processing device 50 inputs the acquired patient information (vital signs information, medication information, treatment information) and prompt P2 to the large-scale language model 100. Prompt P2 can be, for example, "Based on the input patient information (vital signs information, medication information, treatment information), determine whether there is any missing record information regarding objective information. If there is missing record information, generate and present candidates for the missing record information." but is not limited to this. The large-scale language model 100 generates candidates for missing record information regarding objective information based on the input patient information (vital signs information, medication information, treatment information) and outputs the generated candidates for missing record information regarding objective information to the information processing device 50. The record information regarding objective information is record information regarding the patient's vital signs information, medication information, and treatment information.

[0068] As described above, the control unit 51 determines whether there is any missing record information regarding at least one of the patient information obtained as patient information, such as vital information, medication information, and treatment information, and objective information. If there is missing record information, it inputs a prompt to the language model instructing it to generate the missing record information. If the language model determines that there is missing record information, the control unit 51 can obtain the missing record information regarding objective information generated by the language model and accept input of the obtained missing record information regarding objective information.

[0069] The information processing device 50 displays candidates for missing record information related to objective information on the display unit 54. This allows the user to know, by displaying the candidates for record information, that there was a omission in the record information that should have been recorded in the patient DB 70, and what kind of record information was missing.

[0070] When a user inputs missing record information regarding objective information, the information processing device 50 can generate the patient DB 70 by writing the acquired patient information (vital information, medication information, treatment information) and the record information regarding objective information, which includes the input missing record information, to the patient DB 70.

[0071] As described above, if there is any missing recorded information, the control unit 51 can display a selection of candidates for the missing recorded information, accept the selection of a displayed candidate, and accept the selected candidate as input for the missing recorded information.

[0072] Furthermore, the control unit 51 inputs at least one of the patient's vital information, medication information, and treatment information acquired as patient information into the language model (large-scale language model 100), and if the language model determines that there is a gap in the recorded information, it can acquire the missing recorded information regarding objective information generated by the language model and accept input of the acquired missing recorded information regarding objective information.

[0073] Figure 8 shows an example of subjective information recorded in the patient database 70. As shown in Figure 8, patient symptoms can be scored to standardize the level of symptoms, thereby reducing variability in symptom judgment criteria among users. For example, the score can be expressed on a 5-point scale from 1 to 5, with score 1 being "none," score 2 being "slightly present," score 3 being "somewhat present," score 4 being "quite present," and score 5 being "very present."

[0074] As shown in Figure 8, for the symptom "cough" as a data item, scores of 3, 1, 1, 1, and 4 were recorded chronologically at DAY0, DAY1, DAY2, DAY3, and DAY4. For the symptom "fever," scores of 3, 1, 1, 1, and 4 were recorded chronologically at DAY0, DAY1, DAY2, DAY3, and DAY4. For the symptom "chest pain," scores of 3, 5, 5, 5, and 5 were recorded chronologically at DAY0, DAY1, DAY2, DAY3, and DAY4. For the symptom "shortness of breath," scores of 1, 5, 5, 5, and 5 were recorded chronologically at DAY0, DAY1, DAY2, DAY3, and DAY4.

[0075] By scoring symptoms, ambiguity in symptom descriptions can be eliminated, enabling the generation of a high-quality patient database.

[0076] Figure 9 shows an example of objective information recorded in the patient database 70. Data items may include, for example, "respiration," "chest auscultation," "body temperature," and "cough severity." Note that subjective information, such as "cough severity," where the degree of the condition is specified, may also be included in the objective information records. The data items are not limited to the example in Figure 9.

[0077] As shown in Figure 9, the recorded information regarding objective information can include details such as "shallow" for breathing, "right lower lung field rales attenuated, dullness 2 (partial)" for chest auscultation, "38.5°C" for body temperature, and "4" for the degree of cough.

[0078] Next, I will explain nursing records. Nursing records are documents that record the care provided and the patient's condition during nursing practice. Their purpose is to serve as proof of nursing practice, and to evaluate and improve the quality of nursing services. Nursing records consist of a basic structure that includes, for example, basic information including the condition and circumstances of the person requiring nursing care, a nursing plan to achieve the health problems and expected outcomes of the person requiring nursing care, progress records including the course of treatment, and a summary of the progress records.

[0079] Nursing records can be recorded in, for example, SOAP format. S (Subjective Information) includes subjective information based on what the patient or their family felt or said, mainly complaints and subjective symptoms. O (Objective Information) includes objective information obtained through tests, examinations, and observations. A (Assessment) includes the results of analyzing the data obtained from subjective and objective information. P (Nursing Plan) includes items that concretize the policy decided in the assessment, and the treatment methods and nursing plans to be implemented in the future. This specification describes an example of recording nursing records in SOAP format, but nursing records are not limited to SOAP format and may be in other formats such as DAR format or POS format.

[0080] Figure 10 shows an example of the structure of the Nursing Record DB80. The Nursing Record DB80 consists of subjective information, objective information, evaluation information, and nursing plan information for each patient. Subjective information mainly includes the patient's symptoms. Objective information includes information such as the patient's vital signs, medication information, and treatment information. Evaluation information includes evaluation results based on subjective and objective information. Nursing plan information includes future nursing plans based on evaluation information.

[0081] As shown in Figure 10, for patient ID 01, the nursing record consists of subjective information S1, objective information O1, evaluation information A1, and nursing plan information P1. Here, the codes S1, O1, A1, and P1 are used for convenience to represent the set (collection of information) of subjective information, objective information, evaluation information, and nursing plan information, respectively. The same applies to other patients.

[0082] Next, we will explain how nursing information is generated.

[0083] Figure 11 shows an example of a method for generating subjective information in nursing records. The information processing device 50 reads record information regarding the subjective information of a patient from the patient DB 70 for the required patient, and inputs the read subjective information record information and prompt P3 into the large-scale language model 100. Prompt P3 can be, for example, "Based on the entered subjective information record information, generate and present the subjective information for the nursing record," but is not limited to this. The subjective information record information includes, for example, the patient's symptoms, as illustrated in Figure 8.

[0084] As described above, the control unit 51 can input prompts to the language model instructing it to generate record information regarding the patient's subjective information and subjective information for nursing records, and can acquire the subjective information generated by the language model.

[0085] The large-scale language model 100 generates subjective information for nursing records based on recorded information about the input subjective information, and outputs the generated subjective information to the information processing device 50.

[0086] The information processing device 50 can acquire subjective information generated by the large-scale language model 100 and generate the nursing record DB 80 by writing the acquired subjective information to the subjective information column of the nursing record DB 80.

[0087] Figure 12 shows an example of subjective information in nursing records. Based on the patient's symptoms illustrated in Figure 8, for example, it could be written as follows: "The patient presented with a cough, fever, and chest pain upon arrival. The symptoms include chest pain and shortness of breath during breathing, and the chest pain and shortness of breath began three days ago."

[0088] Based on the subjective information mentioned above, it is clear that chest pain and shortness of breath began three days ago, and by eliminating ambiguity in subjective information, it is possible to generate high-quality nursing records.

[0089] Figure 13 shows an example of a method for generating objective information for nursing. The information processing device 50 reads record information regarding the objective information of a patient from the patient DB 70 for the required patient, and inputs the read record information regarding the objective information and prompt P4 into the large-scale language model 100. Prompt P4 can be, for example, "Based on the record information regarding the objective information that has been entered, generate and present the objective information for the nursing record," but is not limited to this. The record information regarding the objective information includes, for example, information such as the patient's vital signs, medication information, and treatment information, as illustrated in Figure 9.

[0090] As described above, the control unit 51 can input prompts to the language model instructing it to generate record information regarding objective patient information and objective information for nursing records, and can acquire the objective information generated by the language model.

[0091] The large-scale language model 100 generates objective information for nursing records based on recorded information about the input objective information, and outputs the generated objective information to the information processing device 50.

[0092] The information processing device 50 can generate the nursing record DB 80 by acquiring objective information generated by the large-scale language model 100 and writing the acquired objective information to the objective information column of the nursing record DB 80.

[0093] Figure 14 shows an example of objective information in nursing records. Based on the record information regarding the objective information of the patient illustrated in Figure 9, it could be written as follows: "Physical examination revealed that the patient has shallow breathing and a frequent cough. Auscultation of the chest revealed attenuated rales in the right lower lung field and partial dullness in the right lower lung field. Body temperature is elevated at 38.5°C, and other vital signs are normal."

[0094] Although not shown in the diagram, the control unit 51 inputs patient information for each medical facility and prompt P5 to the large-scale language model 100. Prompt P5 may be, for example, "Please generate and present candidate record information based on the entered patient information," but is not limited to this. The system may also acquire the candidate record information generated by the large-scale language model 100 and update at least one of the record items and the record information recorded in the patient DB 70 of the medical facility based on the acquired candidate.

[0095] Next, we will explain how to generate evaluation information. To generate evaluation information, it is first necessary to identify the patient's disease.

[0096] Figure 15 shows an example of a method for identifying a disease. The procedure manual is a standard treatment method predetermined for each disease, and is a disease-specific treatment protocol. As shown in Figure 15, the disease can be identified by comparing the patient's subjective and objective information with the documents described in the disease-specific procedure manual. For example, if the patient's subjective and objective information can be compared with the text in Procedure Manual I1, the disease can be identified as D1, which corresponds to Procedure Manual I1. If the patient's subjective and objective information can be compared with the text in Procedure Manual I2, the disease can be identified as D2, which corresponds to Procedure Manual I2. Furthermore, if the patient's subjective and objective information can be compared with the text in Procedure Manual I3, the disease can be identified as D3, which corresponds to Procedure Manual I3. Furthermore, if the patient's subjective and objective information can be compared with the text in Procedure Manual I4, the disease can be identified as D4, which corresponds to Procedure Manual I4.

[0097] Whether a patient's subjective and objective information can be compared with the text in the procedure manual can be determined, for example, by the number of common keywords between the subjective and objective information and the text described in the procedure manual.

[0098] Figure 16 shows an example of record information that serves as the basis for a disease diagnosis. For convenience, only objective record information is shown in the example in Figure 16 (subjective record information is omitted). As shown in Figure 16, the record information is represented as Re1, Re2, Re3, Re4, Re5, ... For example, if record information Re1, Re3, and Re4 apply to the patient, the suspected disease of the patient can be identified as disease D1. In other words, record information Re1, Re3, and Re4 are used as the basis for diagnosing that the patient's disease is disease D1.

[0099] Furthermore, if record information Re2, Re4, and Re5, which are objective information records concerning the patient, apply to the patient, the suspected disease of the patient can be identified as disease D2. In other words, record information Re2, Re4, and Re5 are used as the basis for diagnosing that the patient's disease is disease D2.

[0100] Furthermore, if record information Re1 and Re5, which are objective information records concerning the patient, apply to the patient in question, the suspected disease of the patient can be identified as disease D3. In other words, record information Re1 and Re5 are used as the basis for diagnosing that the patient's disease is disease D3.

[0101] As described above, the patient's disease can be identified based on the record information applicable to the patient.

[0102] Figure 17 shows specific examples of record information that serve as evidence for a disease. As shown in Figure 17, based on the patient's symptoms, test results, and examination results, for example, if the patient exhibits symptoms such as "cough," "sputum," "fever," "difficulty breathing," and "chest pain," the patient's disease can be identified as "pneumonia." Also, if the patient exhibits symptoms such as "difficulty breathing," "shortness of breath," "edema," and "fatigue," the patient's disease can be identified as "heart failure." Furthermore, if the patient exhibits symptoms such as "cough," "fever," "difficulty breathing," and "fatigue," the patient's disease can be identified as "infection." Also, if the patient exhibits symptoms such as "shortness of breath," "fatigue," "palpitations," and "pallor," the patient's disease can be identified as "anemia."

[0103] Figure 18 shows an example of a method for generating evaluation information. Once a patient's disease is identified based on subjective and objective information, evaluation information can be generated based on the procedure manual for the identified disease. The procedure manual contains detailed information on the diagnostic procedure, rationale for diagnosis, and diagnostic methods for the disease. Therefore, evaluation information can be generated by extracting the necessary documents from the documents describing the diagnostic procedure, rationale for diagnosis, and diagnostic methods described in the procedure manual.

[0104] To generate evaluation information, one can extract the necessary text by applying general language processing to the procedure manual. Alternatively, to generate evaluation information, one can create standardized documents in advance from the procedure manual and use these standardized documents as evaluation information.

[0105] Figure 19 shows an example of evaluation information. Based on the subjective information exemplified in Figure 12 and the objective information exemplified in Figure 14, evaluation information can be generated, for example, "The patient's symptoms strongly suggest acute pneumonia. Findings on chest auscultation, fever, cough, and chest pain support this."

[0106] Although not shown in the diagram, the control unit 51 inputs the recorded information, which includes at least one of the patient's vital signs, medication information, and treatment information, and the prompt P6, recorded in the generated patient DB 70, into the language model (large-scale language model 100). The prompt P6 may be, for example, "Based on the entered recorded information, generate and present evaluation information for the nursing record," but is not limited to this. The evaluation information included in the nursing record generated by the language model may be acquired, and the nursing record may be generated using the acquired evaluation information.

[0107] Next, we will explain how nursing care plan information is generated.

[0108] Figure 20 shows an example of a method for generating nursing care plan information. As shown in Figure 20, if the patient's disease is identified as disease D1, nursing care plan information NP1 can be generated based on procedure manual I11 corresponding to disease D1. Procedure manual I11 defines the treatment procedures and treatment plans for disease D1. Therefore, nursing care plan information can be generated by extracting the text about the treatment procedures and treatment plans described in procedure manual I11.

[0109] To generate nursing care plan information, one can extract the necessary text by applying general language processing to the procedure manual. Alternatively, to generate nursing care plan information, one can create standardized documents in advance from the procedure manual documents and use these standardized documents as nursing care plan information.

[0110] Furthermore, if the patient's disease is identified as disease D2, nursing care plan information NP2 can be generated based on procedure I12 corresponding to disease D2, and if the patient's disease is identified as disease D3, nursing care plan information NP3 can be generated based on procedure I13 corresponding to disease D3.

[0111] Figure 21 shows an example of nursing care plan information. Based on the assessment information exemplified in Figure 19, nursing care plan information can be generated as follows: "1. Schedule a chest X-ray for the patient to confirm the diagnosis of pneumonia. 2. Currently, prescribe antibiotics (e.g., amoxicillin) to alleviate symptoms. 3. Encourage rest and adequate fluid intake. 4. Observe the progress of symptoms and re-evaluate in a few days. 5. Consider hospitalization or additional tests as needed."

[0112] As described above, the control unit 51 acquires patient information of a patient, accepts input of record information concerning the patient, generates a patient database based on the acquired patient information and the input record information, and by referring the patient information and record information recorded in the patient database with the acquired patient information and the accepted record information, it determines whether there is any omission in the record information to be recorded in the patient database. If it is determined that there is a omission in the record information, it accepts input of the omission in the record information to regenerate the patient database, and can generate nursing records based on the regenerated patient database.

[0113] Even if there are omissions in the record information used to create nursing records, the system can prevent further omissions by accepting input of the missing record information, thereby enabling the generation of the patient database. This prevents situations where medical information cannot be analyzed due to omissions in the record information. Nursing records can be recorded in an analyzable format.

[0114] Next, we will explain the record information used to generate the ward log. Additional record information for generating the ward log can be added to the patient DB70 for each patient. When generating the ward log, record information from multiple patients is aggregated.

[0115] Figure 22 shows an example of the recorded information that is recorded in the patient database 70 when generating a ward log. Additional data items in the patient database that are recorded to generate the ward log include, but are not limited to, movement information, severity classification, patient status, specific medical procedures, usage of specific medical devices, and the status of patients requiring care.

[0116] Movement information refers to information that shows the movement of patients, such as admission, discharge, transfer out, and transfer in.

[0117] Severity classification is information that indicates classifications such as severe, moderate, and mild.

[0118] The patient's condition indicates whether they can walk independently without assistance, require one assistant for transport, or require two assistants for transport.

[0119] Specific medical procedures are information that indicates important medical procedures such as emergency surgery or specialized treatments.

[0120] The usage status of specific medical devices is information that indicates the usage status of critical equipment such as ventilators.

[0121] The condition of patients requiring care refers to the state of patients with special problems, and is information that indicates the condition of critically ill patients or patients requiring special care.

[0122] The information recorded in the patient database 70 can be recorded chronologically. For example, it can be recorded in chronological order as year / month / day 1, year / month / day 2, year / month / day 3, and so on.

[0123] As shown in Figure 22, the recorded information for movement information within the data items can be recorded chronologically in the order of year / month / day 1, year / month / day 2, year / month / day 3, ... as recorded information ReM1, ReM2, ReM3, ... Similarly, the recorded information for severity classification within the data items can be recorded chronologically in the order of year / month / day 1, year / month / day 2, year / month / day 3, ... as recorded information ReSv1, ReSv2, ReSv3, ... Similarly, the recorded information for patient status within the data items can be recorded chronologically in the order of year / month / day 1, year / month / day 2, year / month / day 3, ... as recorded information ReSt1, ReSt2, ReSt3, ... Similarly, the recorded information for specific medical procedures within the data items can be recorded chronologically in the order of year / month / day 1, year / month / day 2, year / month / day 3, ... as recorded information RePr1, RePr2, RePr3, ... Furthermore, the record information regarding the usage status of specific medical devices within the data items can be recorded chronologically in the order of year / month / day 1, year / month / day 2, year / month / day 3, ... as record information ReDe1, ReDe2, ReDe3, ... Similarly, the record information regarding the status of patients requiring care within the data items can be recorded chronologically in the order of year / month / day 1, year / month / day 2, year / month / day 3, ... as record information ReCa1, ReCa2, ReCa3, ...

[0124] Figure 23 shows an example of how record information is recorded in the patient database 70. When entering the record information ReM1 for the date 1 of the movement information data items, as shown in Figure 23, three options for record information ReM1 (candidates 1, 2, and 3) are displayed, and the user can select one. In this way, instead of manually entering the record information directly, the user can select the required record information from the displayed options and record it in the patient database 70. Similarly, options for record information can be displayed for other severity classifications, patient conditions, specific medical procedures, usage of specific medical devices, and the condition of patients requiring care, allowing the user to select one. In this way, record information can be recorded in the patient database 70.

[0125] The control unit 51 can acquire at least one of the following as patient information: patient movement (movement information), severity classification, patient condition, specific medical procedures, usage status of specific medical devices, and the condition of patients requiring care.

[0126] Figure 24 shows an example of the structure of the ward log DB90. The ward log DB90 can be generated by aggregating information on multiple patients hospitalized in the ward. The ward log DB90 includes information such as severity classification, movement information, the status of patients requiring care, specific medical procedures, nursing staff work status, handover items at shift changes, incidents and accidents, bed occupancy status, usage status of specific medical equipment, and patient status (transportation by stretcher, escort, walking independently).

[0127] Figure 25 shows an example of a ward log. As shown in Figure 25, the ward log includes information on work schedules such as day shifts, extended day shifts, and night shifts. The ward log also records information on patient movement, such as admission, discharge, transfer out, and transfer in, as well as severity classifications, patient status (transportation by stretcher, escort, walking independently), usage of specific medical devices such as ventilators, specific medical procedures such as dialysis, and the status of patients requiring care (list of patients requiring attention).

[0128] The control unit 51 can identify the record information to be recorded in the ward log by aggregating the record information recorded in the patient DB 70 for each of the multiple patients hospitalized in the ward, and generate the ward log DB 90 based on the identified record information.

[0129] As described above, the control unit 51 acquires patient information, accepts input of record information related to the patient, generates a patient database based on the acquired patient information and the input record information, and by referring the patient information and record information recorded in the patient database with the acquired patient information and the accepted record information, it determines whether there is any omission in the record information to be recorded in the patient database. If it is determined that there is a omission in the record information, it accepts input of the omission in the record information to regenerate the patient database, and can generate a ward log based on the regenerated patient database.

[0130] Even if there are omissions in the record information used to create the ward log, the system can prevent further omissions by accepting input of the missing record information, thereby enabling the generation of the patient database. This prevents situations where medical information cannot be analyzed due to omissions in the record information. Ward logs can be recorded in an analyzable format.

[0131] Figure 26 shows an example of the processing procedure by the information processing device 50. The control unit 51 displays a selection of candidate options for the recorded information to be recorded in the patient DB 70 (S11), and accepts the selection of a candidate option to acquire patient information (S12). The control unit 51 determines whether or not it has accepted the selection of all candidate options (S13), and if it has not accepted the selection of all candidate options (NO in S13), it continues the processing from step S11 onwards.

[0132] If all candidate options are selected (YES in S13), the control unit 51 inputs the acquired patient information into the language model (large-scale language model 100) and obtains candidate missing record information generated by the language model (S14).

[0133] The control unit 51 displays the acquired candidates and accepts input of the missing record information (S15), and generates the patient DB 70 based on the acquired patient information and the input record information (missing record information) (S16).

[0134] The control unit 51 generates nursing records using the recorded information stored in the generated patient DB 70 (S17). The control unit 51 updates the ward log by aggregating the recorded information stored in the patient DB 70 for each of the multiple patients hospitalized in the ward (S18), and then terminates the process. The updated ward log can be output at predetermined times (for example, in the morning or evening). [Explanation of symbols]

[0135] 1. Communication Network 50 Information Processing Devices 51 Control Unit 52 Communications Department 53 memory 54 Display section 55 Operation section 56 Memory section 57 Computer Programs 58 Interface section 59 Recording medium reading unit 70 Patient DB 80 Nursing Record Database 90 Ward Diary Database 100 Large-Scale Language Models 120 data servers 121 Medical Information Database

Claims

1. Obtain patient information from the patient, The system accepts input of record information regarding the aforementioned patient. A patient database is generated based on the acquired patient information and the entered record information. By referring to the patient information and record information recorded in the patient database and the acquired patient information and received record information, it is determined whether or not there are any omissions in the record information recorded in the patient database. If it is determined that there is a gap in the aforementioned record information, the system will accept input of the missing record information and regenerate the patient database. Based on the regenerated patient database, nursing records and ward logs are generated. A computer program that instructs a computer to perform a process.

2. The system displays a selection of candidate record information to be recorded in the patient database. The candidate selected from the displayed options is obtained as the patient information. A computer program according to claim 1 that causes a computer to perform a process.

3. The options for the record information to be displayed in the patient database are generated based on the difference between the record information recorded in the patient database and the input record information. A computer program according to claim 2 that causes a computer to perform a process.

4. If there is any missing information in the aforementioned recordings, the system will display a selection of candidates for the missing recordings. Accept the selection of the displayed candidates. The selected candidates will be accepted as input for the missing record information. A computer program according to claim 1 that causes a computer to perform a process.

5. The acquired patient information is input into a language model, and the candidate for missing record information generated by the language model is obtained. It accepts input of record information that was not selected from the acquired candidates. A computer program according to any one of claims 1 to 4, which causes a computer to perform a process.

6. At least one of the patient's symptoms, vital signs, medication information, and treatment information is obtained as patient information. A computer program according to any one of claims 1 to 4, which causes a computer to perform a process.

7. The symptoms obtained as patient information of the aforementioned patient are input into a language model, and if the language model determines that there is a gap in the recorded information, the missing recorded information relating to subjective information generated by the language model is obtained. We accept input of missing record information regarding the subjective information we have obtained. A computer program according to any one of claims 1 to 4, which causes a computer to perform a process.

8. At least one of the patient information obtained from the patient, including vital signs, medication information, and treatment information, is input into a language model, and if the language model determines that there is a gap in the recorded information, the missing recorded information relating to the objective information generated by the language model is obtained. We accept input of leaked record information regarding objective information obtained. A computer program according to any one of claims 1 to 4, which causes a computer to perform a process.

9. Record information, including at least one of the patient's vital signs, medication information, and treatment information, recorded in the generated patient database is input into the language model, and evaluation information included in the nursing record generated by the language model is obtained. Nursing records are generated using the acquired evaluation information. A computer program according to any one of claims 1 to 4, which causes a computer to perform a process.

10. The patient information includes at least one of the following: patient movement, severity classification, patient condition, specific medical procedures, use of specific medical devices, and the patient's condition requiring care. A computer program according to any one of claims 1 to 4, which causes a computer to perform a process.

11. Patient information for each medical facility is input into a language model, and candidate record information generated by the language model is obtained. Based on the acquired candidates, update at least one of the record items in the patient database of the medical facility and the record information recorded in the record items. A computer program according to any one of claims 1 to 4, which causes a computer to perform a process.

12. Equipped with a control unit, The control unit, Obtain patient information from the patient, The system accepts input of record information regarding the aforementioned patient. A patient database is generated based on the acquired patient information and the entered record information. By referring to the patient information and record information recorded in the patient database and the acquired patient information and received record information, it is determined whether or not there are any omissions in the record information recorded in the patient database. If it is determined that there is a gap in the aforementioned record information, the system will accept input of the missing record information and regenerate the patient database. Based on the regenerated patient database, nursing records and ward logs are generated. Information processing device.

13. Obtain patient information from the patient, The system accepts input of record information regarding the aforementioned patient. A patient database is generated based on the acquired patient information and the entered record information. By referring to the patient information and record information recorded in the patient database and the acquired patient information and received record information, it is determined whether or not there are any omissions in the record information recorded in the patient database. If it is determined that there is a gap in the aforementioned record information, the system will accept input of the missing record information and regenerate the patient database. Based on the regenerated patient database, nursing records and ward logs are generated. Information processing methods.

Images