Information processing apparatus, information processing method, and program

Abstract

To provide information useful for selecting a patient to be retroactively introduced.SOLUTION: The information processing apparatus generates a prediction model by machine learning using information on a patient as an explanatory variable and information on a profit from medical treatment of the patient as an objective variable. For example, the information processing apparatus generates a prediction model using the medical examination information and the accounting information of the patient as explanatory variables and the patient ROI score of the patient as an objective variable. The information processor uses the prediction model to predict information related to the profit by the medical examination of the patient to be processed in the future from the information of the patient to be processed (step S13). The information processor displays the predicted result on a display part (step S14).SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] In Japan, patients can generally visit any hospital for their first visit without a referral letter. As a result, compared to other countries, Japan has a higher number of outpatients at large and medium-sized hospitals. In particular, at medium-sized hospitals, first-time patients often visit the hospital in the same way as they would at a clinic. Furthermore, outpatients who do not have a referring hospital find it difficult to refer them back to a clinic even after their condition has stabilized. Patients whose condition has already improved have a low unit price for medical treatment. If a large number of such patients remain at large or medium-sized hospitals, it puts pressure on the hospital's management.

[0003] The following are techniques for analyzing the financial status of hospitals: Patent Document 1 discloses a management support system that collects performance information from multiple medical institutions and performs benchmark analysis. This management support system creates comparative information between the performance information of a specific medical institution and the performance information of other medical institutions based on comparison conditions set by the specific medical institution.

[0004] Patent Document 2 discloses a technology that provides each medical institution with a platform-type medical system and performs business analysis using information from each medical institution. The reservation sales calculation system described in Patent Document 2 predicts daily sales for patients with appointments by adding up the medical fee points corresponding to the scheduled consultation contents on a target day. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Laid-Open No. 2001-216379 [Patent Document 2] Japanese Patent Publication No. 2023-10513 Summary of the Invention [Problem to be solved by the invention]

[0006] Currently, doctors select patients to refer to clinics based on their own judgment. Ideally, doctors should select patients to refer based not only on the patient's condition but also on the medical fees they will likely receive in the future. However, systems such as electronic medical records do not have the functionality to analyze or predict revenue from patient care. Furthermore, it is difficult for mid-sized hospitals to secure personnel who can analyze and predict revenue from patient care.

[0007] Furthermore, even with the techniques described in Patent Document 1 or Patent Document 2, it was difficult to predict future revenues from medical treatment of patients.

[0008] The present invention has been made in consideration of the above-mentioned problems in the prior art, and an object of the present invention is to provide information useful for selecting patients to be referred back. [Means for solving the problem]

[0009] In order to solve the above problem, the invention described in claim 1 is an information processing device comprising: a first prediction means for predicting information regarding future revenue from medical treatment for a patient to be processed from information about the patient to be processed using a first model generated by machine learning with patient information as an explanatory variable and information regarding revenue from medical treatment for the patient as a target variable; and a display control means for displaying the results predicted by the first prediction means on a display means.

[0010] The invention described in claim 2 is an information processing device described in claim 1, in which the patient information used as an explanatory variable in generating the first model includes at least one of gender, age, disease name, and medical history, and the first prediction means predicts information regarding revenue from medical treatment for the patient being processed at one or more future times.

[0011] The invention described in claim 3 is an information processing device described in claim 1 or 2, further comprising a second prediction means for predicting an index indicating the ease of reverse referral of a patient to be processed from the information of the patient to be processed using a second model generated by machine learning with patient information as an explanatory variable and an index indicating the ease of reverse referral of the patient as a target variable, and the display control means displays the result predicted by the second prediction means on the display means.

[0012] The invention described in claim 4 is an information processing device described in claim 3, in which the patient information used as an explanatory variable in generating the second model includes at least one of information regarding reverse referral, gender, age, disease name, and medical history, and the second prediction means predicts an index indicating the likelihood of reverse referral of the patient to be processed at the time of prediction by the second prediction means.

[0013] The invention described in claim 5 is an information processing device described in claim 3, further comprising a calculation means for calculating a score indicating the degree to which reverse referral of the patient to be processed is recommended based on information regarding revenue from medical treatment for the patient to be processed and an index indicating the ease of reverse referral for the patient to be processed, and the display control means displays the score calculated by the calculation means on the display means.

[0014] The invention described in claim 6 is an information processing device described in claim 5, wherein the calculation means calculates the score so that the higher the revenue from medical treatment of the patient to be processed, the lower the score, and the higher the ease of reverse referral of the patient to be processed.

[0015] The invention described in claim 7 is an information processing method including a prediction step of predicting information regarding future revenue from medical treatment for a patient to be processed from information about the patient to be processed using a model generated by machine learning with patient information as an explanatory variable and information regarding revenue from medical treatment for the patient as a target variable, and a display control step of displaying the results predicted in the prediction step on a display means.

[0016] The invention described in claim 8 is a program for causing a computer to function as a prediction means for predicting information regarding future revenue from medical treatment for a patient to be processed from information about the patient to be processed, using a model generated by machine learning with patient information as an explanatory variable and information regarding revenue from medical treatment for the patient as a target variable, and a display control means for displaying the results predicted by the prediction means on a display means. [Effects of the Invention]

[0017] According to the present invention, it is possible to provide information that is useful for selecting patients to be referred back. [Brief explanation of the drawings]

[0018] [Figure 1] 1 is a system configuration diagram of a reverse introduction support system according to a first embodiment of the present invention. [Figure 2] FIG. 2 is a diagram for explaining a method for generating a first model. [Figure 3] 1 is an example of medical information used to generate a first model. [Figure 4] 10 is a flowchart showing a first machine learning process. [Figure 5] 10 is a flowchart showing a first score calculation process. [Figure 6] 10 is an example of a first score display screen. [Figure 7] FIG. 10 is a diagram for explaining a method for generating a second model in the second embodiment. [Figure 8] 10 is a flowchart showing a second machine learning process. [Figure 9] 10 is a flowchart showing a second score calculation process. [Figure 10] 10 is an example of a second score display screen. [Figure 11] 11 is a flowchart showing a third score calculation process in the third embodiment. [Figure 12] 10 is an example of a third score display screen. DETAILED DESCRIPTION OF THE INVENTION

[0019] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Advantages and features provided by the embodiments will be understood from the following detailed description and drawings. However, the scope of the present invention is not limited to the embodiments disclosed below or the examples shown in the drawings.

[0020] [First embodiment] First, a first embodiment of an information processing device according to the present invention will be described. FIG. 1 is a system configuration diagram of a reverse referral support system 100. The reverse referral support system 100 includes an information processing device 10, a data server 20, an electronic medical record system 30, a medical receipt system 40, an accounting system 50, and a financial system 60. The reverse referral support system 100 is used in a medical institution. Here, the medical institution is assumed to be a medical institution larger than a predetermined size. For example, the medical institution is a medium-sized hospital or a large-sized hospital.

[0021] The information processing device 10 acquires various information from the data server 20. The information processing device 10 provides the user with information to support reverse referral of patients from medical institutions to other medical institutions. Reverse referral refers to the referral of a patient whose condition has stabilized to the referring doctor or a local clinic (physical office). Here, the target is primarily patients receiving outpatient treatment at medical institutions.

[0022] The data server 20, electronic medical record system 30, receipt system 40, accounting system 50, financial system 60, etc. are existing systems within the medical institution (hospital). The data server 20 collects and stores various information from the electronic medical record system 30, receipt system 40, accounting system 50, financial system 60, etc. within the medical institution. The data server 20 is configured as a data warehouse. The data server 20 may periodically acquire various information from each system, or may acquire various information when information is updated in each system.

[0023] The electronic medical record system 30 manages information related to the electronic medical records of patients in a medical institution. The information related to the electronic medical records includes medical information, etc. The medical information includes the patient's symptoms, test results, treatment details, etc. The medical receipt system 40 is a system for creating medical receipts (medical fee statements). Medical receipts are documents used by medical institutions to bill insurers for medical expenses. The medical receipt system 40 manages medical receipt information for patients within the medical institution. The medical receipt information includes the name of the disease, test details, test costs, treatment details, treatment costs, etc. The accounting system 50 manages accounting information within a medical institution. The accounting information includes income, expenditure, profit and loss, etc. of the medical institution. The financial system 60 manages financial information within the medical institution. The financial system 60 may be a common system with the receipt system 40 or the accounting system 50.

[0024] The information processing device 10 includes a control unit 11, an operation unit 12, a display unit 13, a communication unit 14, a storage unit 15, a database 16, and the like. The control unit 11 includes a CPU (Central Processing Unit), RAM (Random Access Memory), etc. The control unit 11 comprehensively controls the processing operations of each unit of the information processing device 10. Specifically, the CPU reads out various programs stored in the storage unit 15. The CPU expands the read programs into the RAM and performs various processes in cooperation with the programs.

[0025] The operation unit 12 includes a keyboard, a mouse, etc. The keyboard includes cursor keys, character input keys, various function keys, etc. The operation unit 12 outputs operation signals input by key operations on the keyboard or mouse operations to the control unit 11. The operation unit 12 may include a touch screen configured integrally with the display unit 13. In this case, the operation unit 12 outputs operation signals to the control unit 11 according to the position of the touch operation on the touch screen.

[0026] The display unit 13 (display means) includes an LCD (Liquid Crystal Display), etc. The display unit 13 displays various screens according to instructions of a display signal input from the control unit 11.

[0027] The communication unit 14 is configured by a network interface, etc. The communication unit 14 transmits and receives data to and from an external device connected via a communication network N, such as a LAN (Local Area Network).

[0028] The storage unit 15 is a storage device configured with an HDD (Hard Disk Drive), an SSD (Solid State Drive), a non-volatile semiconductor memory, etc. The storage unit 15 stores various programs, and parameters and data required to execute the various programs. The storage unit 15 also has a trained model storage unit 151. The trained model storage unit 151 stores a prediction model generated by machine learning.

[0029] The database 16 is a database for managing various information obtained from the data server 20 .

[0030] The control unit 11 acquires various pieces of patient information from the data server 20 via the communication unit 14. Note that the control unit 11 does not have to acquire the information managed in the data server 20 via the communication network N. For example, the control unit 11 may import the information managed in the data server 20 as a CSV file.

[0031] The control unit 11 generates a first model by machine learning using patient information as an explanatory variable and information on revenue from medical treatment of the patient as a target variable. The patient information used as an explanatory variable in generating the first model includes at least one of gender, age, disease name, and hospital history.

[0032] The control unit 11 (first prediction means) uses the first model to predict information about future profits from medical treatment for the patient to be processed, based on information about the patient to be processed. The control unit 11 predicts information about profits from medical treatment for the patient to be processed at one or more future times.

[0033] The control unit 11 (display control means) causes the result predicted by the first prediction means to be displayed on the display unit 13. In other words, the control unit 11 causes the display unit 13 to display information regarding future profits from medical treatment of the patient to be processed.

[0034] Next, a method for generating the first model will be described with reference to Fig. 2. Here, a patient ROI (Return On Investment) score is used as information relating to revenue from medical treatment for the patient. The control unit 11 generates a first model by machine learning using the patient's medical information and accounting information as explanatory variables and the patient ROI score of the patient as a response variable.

[0035] The medical information is information related to the medical treatment of a patient. The medical information may include the patient's attributes (gender, age, etc.). The medical information used to generate the first model includes gender, age, medical department category, medical department, disease name, outpatient / hospitalization history, test status, condition, etc. The medical information is mainly information managed in the electronic medical record system 30. FIG. 3 shows an example of the clinical information used to generate the first model.

[0036] The accounting information used to generate the first model includes income, expenses, etc. The accounting information is mainly information managed in the receipt system 40, the accounting system 50, or the financial system 60. Income includes outpatient income, inpatient income, and surgery income. Outpatient income includes medical fees, tests, medication, rehabilitation, etc. Inpatient income includes hospital fees, tests, medication, rehabilitation, etc. Surgery income includes surgery fees, tests, medication, etc. Costs include fixed costs and variable costs. Fixed costs include salary costs, depreciation costs, rent, land costs, and utility costs. Variable costs include pharmaceutical costs, medical material costs, patient meal material costs, and testing outsourcing costs. When generating the first model, only information that can change depending on the presence of each patient needs to be used as costs. Alternatively, costs that do not correspond to individual patients can be converted to values ​​per patient and the converted values ​​can be used.

[0037] The control unit 11 acquires learning data (past data) used for machine learning from the data server 20. For data whose content may change over time, the date and time or period corresponding to the data is added to the data.

[0038] The control unit 11 uses data from a reference date and time when calculating the patient ROI score as the medical information and accounting information used for machine learning. Hereinafter, the "reference date and time" will be referred to as the "reference time." The reference time when calculating the patient ROI score may be, for example, the time of the first visit. For income, expenses, etc., information from a predetermined period prior to the reference time may be used.

[0039] The control unit 11 uses the patient ROI score one year after the baseline time and the patient ROI score three years after the baseline time as the objective variables used in the machine learning. The control unit 11 calculates the patient ROI score one year after the baseline time according to the following formula (1). Patient ROI score from baseline to 1 year later = (profit from medical treatment for the patient in question for one year from the baseline) / (costs for medical treatment of the relevant patient in one year from the base point) × 100 ... Formula (1)

[0040] The control unit 11 calculates the patient ROI score three years after the baseline according to the following formula (2). Patient ROI score from baseline to 3 years later = (profit from medical treatment for the patient in question over a three-year period from the baseline) / (costs for medical treatment of the relevant patient over the three years from the baseline) x 100 ...Equation (2)

[0041] Next, the operation of the reverse introduction support system 100 will be described. 4 is a flowchart showing a first machine learning process executed by the information processing device 10. The first machine learning process is a process for generating a first model that predicts a patient ROI score.

[0042] First, the control unit 11 acquires various patient information from the data server 20 via the communication unit 14 (step S1). The various information includes electronic medical record information, medical receipt information, accounting information, financial information, etc. Specifically, the control unit 11 acquires medical information and accounting information to be used as explanatory variables. The control unit 11 also acquires information for calculating a patient ROI score to be used as a target variable. The information for calculating the patient ROI score includes income from the patient's medical care, expenses for the patient's medical care, etc. The control unit 11 stores the acquired various information for each patient in the database 16.

[0043] Next, the control unit 11 calculates a patient ROI score for each patient based on the income and expenses of each patient for a predetermined period (step S2). Specifically, the control unit 11 subtracts expenses from the income of each patient for the predetermined period to calculate the profit of each patient for the predetermined period. The predetermined periods are one year from the reference time and three years from the reference time. The control unit 11 calculates the patient ROI score one year after the reference time according to the above formula (1). The control unit 11 also calculates the patient ROI score three years after the reference time according to the above formula (2).

[0044] Next, the control unit 11 constructs a first model by machine learning using the patient's medical information and accounting information as explanatory variables and the patient ROI score of the patient as a target variable (step S3). The first model is generated so that it can output the patient ROI score one year after the baseline time and / or the patient ROI score three years after the baseline time. In the first model, the user may be able to arbitrarily change the stage at which the patient ROI score is output. For example, in step S2, the predetermined period for calculating the patient ROI score may be arbitrarily changeable.

[0045] Next, the control unit 11 evaluates the first model (step S4). The control unit 11 evaluates the validity of the first model using input data and output data for which the correct answers are known. For example, the control unit 11 may set aside a portion of the data acquired in step S1 as data for evaluation and use this portion of the data for evaluation. This data for evaluation is not used to generate the first model. If the evaluation result of the first model does not reach a usable level, the control unit 11 may add data for learning and regenerate the first model.

[0046] Thereafter, the control unit 11 stores the first model in the trained model storage unit 151 of the storage unit 15 (step S5). This completes the first machine learning process.

[0047] 5 is a flowchart showing the first score calculation process executed by the information processing device 10. The first score calculation process is a process of predicting a patient ROI score using a first model and displaying the prediction result.

[0048] First, the control unit 11 acquires the medical information and accounting information of the target patient from the data server 20 via the communication unit 14 (step S11). Specifically, data similar to the explanatory variables shown in FIG. 2 is used as the medical information and accounting information of the target patient. However, the control unit 11 does not need to acquire all of the explanatory variables shown in FIG. 2 for the target patient, and may acquire only some of the explanatory variables. Here, all patients treated at a medical institution may be considered target patients. Alternatively, patients who meet certain conditions may be considered target patients.

[0049] Next, the control unit 11 calls the first model from the trained model storage unit 151 of the memory unit 15 (step S12).

[0050] Next, the control unit 11 calculates the patient ROI score for the target patient using the first model (step S13). The control unit 11 inputs the medical information and accounting information of the target patient into the first model and obtains the patient ROI score for the target patient. The control unit 11 uses the first model to predict the patient ROI score for the target patient one year from the present time and the patient ROI score for the target patient three years from the present time, based on the medical information and accounting information of the target patient. Here, the present time is used as the reference time. More precisely, the date and time of the data used as the medical information and accounting information of the target patient is used as the reference time.

[0051] Next, the control unit 11 causes the display unit 13 to display the patient ROI score for each patient (step S14). This completes the first score calculation process.

[0052] FIG. 6 shows an example of the first score display screen 131 displayed on the display unit 13. The first score display screen 131 displays, for each patient, the name, medical department, disease name, progress, outpatient history, patient ROI score one year from the present time, and patient ROI score three years from the present time. A doctor considers patients with relatively low patient ROI scores as "patients who should be referred back." When displaying the first score display screen 131, the control unit 11 may sort the patients in descending order of patient ROI score. Alternatively, the control unit 11 may display a predetermined number of patients on the first score display screen 131, starting with the lowest patient ROI score.

[0053] As described above, according to the first embodiment, the control unit 11 of the information processing device 10 generates a first model through machine learning using patient information as an explanatory variable and information on revenue from medical treatment for the patient as a target variable. Using the first model, the control unit 11 predicts information on future revenue from medical treatment for the target patient based on the information on the target patient. The control unit 11 displays the predicted information on revenue from medical treatment for the patient on the display unit 13. This allows the control unit 11 to provide information useful for selecting patients to be referred. The control unit 11 contributes to the management of medical institutions more than in the past, when doctors select patients based on their own intuition. Furthermore, by using the first model generated by machine learning, the control unit 11 can reduce the amount of work required to select patients to be referred. For example, doctors can select patients to be referred based on each patient's patient ROI score one year from the present and three years from the present. Doctors can refer patients with relatively low patient ROI scores to other medical institutions, thereby improving the management efficiency of medical institutions using the reverse referral support system 100. This will alleviate congestion at the medical institution and reduce the workload of medical staff.

[0054] The patient information used as an explanatory variable in generating the first model includes at least one of gender, age, disease name, and medical history. This allows the control unit 11 to use the first model to predict information related to the profits from medical treatment of the patient to be processed based on the gender, age, disease name, medical history, etc. of the patient to be processed.

[0055] Furthermore, when the control unit 11 predicts information relating to revenue from medical treatment of the patient to be processed at multiple future times, it can provide information on trends or fluctuations in the information relating to the revenue.

[0056] [Second embodiment] Next, a second embodiment to which the present invention is applied will be described. The reverse introduction support system of the second embodiment has the same configuration as the reverse introduction support system 100 shown in the first embodiment. Therefore, the same components as those in the first embodiment are designated by the same reference numerals, and illustrations and explanations thereof are omitted. The configuration and processing characteristic of the second embodiment will be explained below.

[0057] The control unit 11 of the information processing device 10 generates a second model by machine learning using the patient information as an explanatory variable and an index indicating the ease of reverse referral of the patient as a response variable. The patient information used as an explanatory variable in generating the second model includes at least one of information about reverse referral, gender, age, disease name, and medical history. Hereinafter, "information about reverse referral" will be referred to as "reverse referral information."

[0058] The control unit 11 (second prediction means) predicts an index indicating the likelihood of reverse referral of the patient to be treated from the information of the patient to be treated using the second model. The control unit 11 predicts an index indicating the likelihood of reverse referral of the patient to be treated at the time of prediction by the second prediction means.

[0059] The control unit 11 (display control means) causes the result predicted by the second prediction means to be displayed on the display unit 13. That is, the control unit 11 causes the display unit 13 to display an index indicating the ease of reverse referral of the patient to be processed.

[0060] Next, a method for generating the second model will be described with reference to Fig. 7. Here, a reverse referral acceptance score (reverse referral acceptance rate) is used as an index indicating the ease of reverse referral of a patient. The control unit 11 generates a second model by machine learning using the medical information and reverse referral information of the patient as explanatory variables and the reverse referral acceptance score of the patient as a response variable.

[0061] The medical information used to generate the second model includes gender, age, medical department category, medical department, disease name, outpatient / hospitalization history, examination status, condition, etc. The medical information is mainly information managed in the electronic medical record system 30.

[0062] The reverse referral information used to generate the second model includes the presence or absence of a referral source, the number of reverse referral inquiries, and the number of reverse referral acceptances. The presence or absence of a referral source is information indicating the presence or absence of a referral source medical institution (clinic, etc.) that referred a patient to a medical institution using the reverse referral support system 100. The number of reverse referral inquiries is the number of times a doctor at a medical institution using the reverse referral support system 100 has approached a patient about a reverse referral. The number of reverse referral acceptances is the number of times a patient has accepted a reverse referral in response to an inquiry from a doctor. The reverse referral information is information that is mainly managed in the electronic medical record system 30. The period for counting the number of reverse referral inquiries and the number of reverse referral acceptances may be the entire period that the patient is treated at this medical institution, or a predetermined period going back from the present time.

[0063] The control unit 11 acquires, from the data server 20, learning data (past data) to be used for machine learning.

[0064] The control unit 11 uses data at a reference time when calculating the reverse referral acceptance score as the medical information and reverse referral information used for machine learning. The reference time when calculating the reverse referral acceptance score is usually the current time (prediction time).

[0065] The control unit 11 uses the reverse introduction acceptance score as a target variable used in machine learning. The control unit 11 calculates the reverse introduction acceptance score according to the following formula (3). Reverse referral acceptance score = (number of reverse referrals accepted) / (number of reverse referral inquiries) × 100 ... Equation (3)

[0066] Next, the operation of the reverse introduction support system 100 of the second embodiment will be described. The first machine learning process (see FIG. 4) is the same as in the first embodiment, and therefore description thereof will be omitted. In this way, in the second embodiment as well, the control unit 11 generates a first model.

[0067] 8 is a flowchart showing the second machine learning process executed by the information processing device 10 of the second embodiment. The second machine learning process is a process for generating a second model that predicts the reverse referral acceptance score.

[0068] First, the control unit 11 acquires various types of patient information from the data server 20 via the communication unit 14 (step S21). The various types of information include electronic medical record information, medical receipt information, accounting information, financial information, etc. Specifically, the control unit 11 acquires medical information and reverse referral information, which are used as explanatory variables. The control unit 11 also acquires information for calculating a reverse referral acceptance score, which is used as a target variable. The information for calculating the reverse referral acceptance score includes the number of reverse referral inquiries, the number of reverse referral acceptances, etc. The control unit 11 stores the acquired various types of information in the database 16 for each patient.

[0069] Next, the control unit 11 calculates the reverse referral acceptance score (reverse referral acceptance rate) of each patient based on the number of reverse referral acceptances and the number of reverse referral inquiries for each patient (step S22). Specifically, the control unit 11 calculates the reverse referral acceptance score according to the above formula (3).

[0070] Next, the control unit 11 constructs a second model by machine learning using the patient's medical information and reverse referral information as explanatory variables and the patient's reverse referral acceptance score as a response variable (step S23).

[0071] Next, the control unit 11 evaluates the second model (step S24). The control unit 11 evaluates the validity of the second model using input data and output data for which the correct answers are known. For example, the control unit 11 may set aside a portion of the data acquired in step S21 as data for evaluation and use this portion of the data for evaluation. This data for evaluation is not used to generate the second model.

[0072] Thereafter, the control unit 11 stores the second model in the trained model storage unit 151 of the storage unit 15 (step S25). This completes the second machine learning process.

[0073] 9 is a flowchart showing the second score calculation process executed by the information processing device 10 of the second embodiment. The second score calculation process is a process for predicting the patient ROI score and the reverse referral acceptance score and displaying the prediction results.

[0074] First, the control unit 11 acquires the medical information, accounting information, and reverse referral information of the target patient from the data server 20 via the communication unit 14 (step S31). Specifically, data similar to the explanatory variables shown in FIGS. 2 and 7 is used as the medical information, accounting information, and reverse referral information of the target patient. However, the control unit 11 does not need to acquire all of the explanatory variables shown in FIGS. 2 and 7 for the target patient, and may acquire only some of the explanatory variables. Here, all patients treated at a medical institution may be considered target patients. Alternatively, patients who meet certain conditions may be considered target patients.

[0075] Next, the control unit 11 calls the first model from the trained model storage unit 151 of the storage unit 15 (step S32).

[0076] Next, the control unit 11 calculates the patient ROI score for the target patient using the first model (step S33). The control unit 11 inputs the medical information and accounting information of the target patient into the first model and obtains the patient ROI score for the target patient. The control unit 11 uses the first model to predict the patient ROI score for the target patient one year from the present time and the patient ROI score for the target patient three years from the present time, based on the medical information and accounting information of the target patient.

[0077] Next, the control unit 11 calls the second model from the trained model storage unit 151 of the storage unit 15 (step S34).

[0078] Next, the control unit 11 calculates a reverse referral acceptance score for the target patient using the second model (step S35). The control unit 11 inputs the medical information and reverse referral information of the target patient into the second model and acquires a reverse referral acceptance score for the target patient. The control unit 11 predicts the reverse referral acceptance score of the target patient at the prediction time (current time) from the medical information and reverse referral information of the target patient using the second model.

[0079] Next, the control unit 11 causes the display unit 13 to display the patient ROI score and the reverse referral acceptance score for each patient (step S36). This completes the second score calculation process.

[0080] FIG. 10 shows an example of the second score display screen 132 displayed on the display unit 13. The second score display screen 132 displays, for each patient, the name, medical department, disease name, progress, outpatient history, reverse referral acceptance score, patient ROI score one year from the present time, and patient ROI score three years from the present time. A doctor considers patients with relatively low patient ROI scores and relatively high reverse referral acceptance scores as "patients who should be reverse referred." When displaying the second score display screen 132, the control unit 11 may sort the patients in descending order of patient ROI scores. Alternatively, when displaying the second score display screen 132, the control unit 11 may sort the patients in descending order of reverse referral acceptance scores. Furthermore, the control unit 11 may display a predetermined number of patients on the second score display screen 132, starting with the lowest patient ROI score. Furthermore, the control unit 11 may display a predetermined number of patients on the second score display screen 132, starting with the highest reverse referral acceptance score.

[0081] As described above, according to the second embodiment, the control unit 11 of the information processing device 10 can provide information useful for selecting patients to be reversely referred, similarly to the first embodiment.

[0082] Furthermore, the control unit 11 generates a second model through machine learning using patient information as an explanatory variable and an index indicating the ease of reverse referral for the patient as a target variable. The control unit 11 uses the second model to predict an index indicating the ease of reverse referral for the patient to be processed from the information of the patient to be processed. The control unit 11 displays the predicted index indicating the ease of reverse referral for the patient to be processed on the display unit 13. This allows the control unit 11 to provide information useful for selecting patients to be reverse referred. For example, a doctor can select patients to be reverse referred by referring to each patient's reverse referral acceptance score. By a doctor reverse referring patients with a relatively high reverse referral acceptance score to another medical institution, the number of patients at the medical institution using the reverse referral support system 100 can be smoothly adjusted.

[0083] The patient information used as an explanatory variable in generating the second model includes at least one of reverse referral information, gender, age, disease name, and hospital history. This allows the control unit 11 to use the second model to predict an index showing the likelihood of reverse referral for the patient to be processed, based on the reverse referral information, gender, age, disease name, hospital history, etc. of the patient to be processed.

[0084] Furthermore, the control unit 11 predicts an index indicating the ease of reverse referral of the patient to be processed at the time of prediction, thereby enabling the doctor to select the patient to be reverse referred, taking into consideration the ease of reverse referral at the time of prediction.

[0085] [Third embodiment] Next, a third embodiment to which the present invention is applied will be described. The reverse introduction support system of the third embodiment has the same configuration as the reverse introduction support system 100 shown in the first embodiment. Therefore, the same components as those in the first embodiment are designated by the same reference numerals, and illustrations and explanations thereof are omitted. Below, the configuration and processing characteristic of the third embodiment will be explained.

[0086] The control unit 11 (calculation means) of the information processing device 10 calculates a score indicating the degree to which reverse referral is recommended for the patient to be processed, based on information regarding revenue from medical treatment of the patient to be processed and an index indicating the ease of reverse referral for the patient to be processed. Hereinafter, the "score indicating the degree to which reverse referral is recommended" will be referred to as the "reverse referral recommendation score." The control unit 11 calculates the reverse referral recommendation score so that the higher the revenue from medical treatment of the patient to be processed, the lower the reverse referral recommendation score. The control unit 11 calculates the reverse referral recommendation score so that the higher the ease of reverse referral for the patient to be processed, the higher the reverse referral recommendation score.

[0087] The control unit 11 (display control means) causes the display unit 13 to display the reverse introduction recommendation score calculated by the calculation means.

[0088] The control unit 11 calculates the reverse introduction recommendation score according to the following formula (4). Reverse referral recommendation score = {(100 - patient ROI score) + reverse referral acceptance score} / 2 ...Formula (4)

[0089] As shown in formula (4), the higher the patient ROI score, the lower the reverse referral recommendation score. Also, the higher the reverse referral acceptance score, the higher the reverse referral recommendation score.

[0090] Next, the operation of the reverse introduction support system 100 of the third embodiment will be described. The first machine learning process (see FIG. 4) is the same as that in the first embodiment, and therefore description thereof will be omitted. In this way, in the third embodiment as well, the control unit 11 generates a first model. The second machine learning process (see FIG. 8) is the same as that in the second embodiment, and therefore description thereof will be omitted. In this way, in the third embodiment as well, the control unit 11 generates a second model.

[0091] 11 is a flowchart showing the third score calculation process executed by the information processing device 10 of the third embodiment. The third score calculation process is a process for calculating a reverse introduction recommendation score and displaying the calculation result. The processing of steps S41 to S45 is similar to the processing of steps S31 to S35 of the second score calculation processing (see FIG. 9), and therefore a description thereof will be omitted.

[0092] Next, the control unit 11 calculates a reverse referral recommendation score for the target patient from the patient ROI score and the reverse referral acceptance score of the target patient according to the above formula (4) (step S46).

[0093] Next, the control unit 11 causes the display unit 13 to display the patient ROI score, the reverse referral acceptance score, and the reverse referral recommendation score for each patient (step S47). This completes the third score calculation process.

[0094] FIG. 12 shows an example of the third score display screen 133 displayed on the display unit 13. The third score display screen 133 displays, for each patient, the name, reverse referral recommendation score, reverse referral acceptance score, patient ROI score one year from the present, and patient ROI score three years from the present. A doctor considers patients with relatively high reverse referral recommendation scores as "patients who should be reverse referred." The reverse referral recommendation score displayed on the third score display screen 133 is a value calculated from the reverse referral acceptance score and the patient ROI score one year from the present. Alternatively, or in addition, the control unit 11 may display a reverse referral recommendation score calculated from the reverse referral acceptance score and the patient ROI score three years from the present. When displaying the third score display screen 133, the control unit 11 may sort patients in descending order of reverse referral recommendation score. Furthermore, the control unit 11 may display a predetermined number of patients on the third score display screen 133, starting with the highest reverse referral recommendation score.

[0095] As described above, according to the third embodiment, the control unit 11 of the information processing device 10 can provide information useful for selecting patients to be reversely referred, similarly to the first and second embodiments.

[0096] The control unit 11 calculates a reverse referral recommendation score for the patient to be processed based on information regarding revenue from medical treatment for the patient to be processed and an index indicating the ease of reverse referral for the patient to be processed. The control unit 11 displays the calculated reverse referral recommendation score on the display unit 13. This allows the control unit 11 to provide information useful for selecting patients to be reverse referred. Doctors can select patients to be reverse referred by referring to the reverse referral recommendation score for each patient. By doctors reverse referring patients with relatively high reverse referral recommendation scores to other medical institutions, the management efficiency of medical institutions using the reverse referral support system 100 is improved and the number of patients at the medical institution can be smoothly adjusted.

[0097] The control unit 11 lowers the reverse referral recommendation score as the profit from medical treatment of the patient to be processed increases, and the control unit 11 raises the reverse referral recommendation score as the ease of reverse referral of the patient to be processed increases. In this way, the control unit 11 can provide the degree of recommendation for reverse referral of the patient.

[0098] The descriptions in the above embodiments are examples of the information processing device, information processing method, and program according to the present invention, and are not limited to these. The detailed configuration and detailed operation of each part constituting the device can also be changed as appropriate within the scope of the present invention. For example, characteristic processes in each embodiment may be combined and executed.

[0099] In the above embodiments, the case where the patient ROI score (ratio of profit to cost) is used as information on the profit from the patient's medical treatment has been described. Instead, the profit from the patient's medical treatment, income from the patient's medical treatment, etc. may be used as information on the profit from the patient's medical treatment.

[0100] In the above second and third embodiments, a case has been described in which a reverse referral acceptance score (reverse referral acceptance rate) is used as an index indicating the ease of reverse referral of a patient. Alternatively, the distance from the patient's home to the medical institution to which the reverse referral is made, the severity of the patient's illness, etc. may be used as an index indicating the ease of reverse referral. For example, generally, the closer the distance from the patient's home to the medical institution to which the reverse referral is made, the more likely a doctor will reverse refer the patient. Furthermore, the less severe the patient's illness, the more likely a doctor will reverse refer the patient.

[0101] In the third embodiment, the calculation method of the reverse introduction recommendation score using the above formula (4) has been described, but the present invention is not limited to this. A reverse introduction recommendation score calculated by another method may also be used.

[0102] In the above embodiments, the control unit 11 of the information processing device 10 has been described as using the first model to predict the patient ROI score one year after the baseline time and the patient ROI score three years after the baseline time. That is, the control unit 11 has been described as predicting information related to the profits from treating patients at two future timings. Alternatively, the control unit 11 may predict information related to the profits from treating patients at one future timing, or may predict information related to the profits from treating patients at three or more future timings.

[0103] In each of the above embodiments, the processing executed by the information processing device 10 may be performed by a plurality of devices in cooperation with each other. Furthermore, instead of the operation unit 12 and the display unit 13 of the information processing device 10, an operation unit and a display unit of an external device accessible to the information processing device 10 may be used. For example, the control unit 11 of the information processing device 10 may predict information regarding revenue from medical treatment of a patient to be processed in response to an operation by a doctor from the operation unit of the external device, and display the predicted result on the display unit of the external device. The control unit 11 of the information processing device 10 may predict an index indicating the ease of reverse referral for a patient to be processed in response to an operation by a doctor from the operation unit of the external device, and display the predicted result on the display unit of the external device. The control unit 11 of the information processing device 10 may calculate a reverse referral recommendation score for a patient to be processed in response to an operation by a doctor from the operation unit of the external device, and display the calculated reverse referral recommendation score on the display unit of the external device.

[0104] In the above-described embodiments, the systems cited as sources of various information that the information processing device 10 acquires from the data server 20 are merely examples and do not limit the present invention.

[0105] The computer-readable medium for storing the program for executing each process is not limited to the above examples. A carrier wave may also be used as a medium for providing program data via a communication line.

[0106] The above-disclosed embodiments are for the purpose of explanation and example only, and are not intended to be limiting. The scope of the present invention should be interpreted by the claims. [Explanation of symbols]

[0107] 10. Information processing equipment 11 Control section 12 Control section 13 Display section 14 Communications Department 15 Storage section 16 Databases 20 Data Server 30 Electronic Medical Record System 40 Receipt System 50 Accounting System 60 Financial Systems 100 Reverse introduction support system 151 Trained model storage N Communication Network

Claims

1. a first prediction means for predicting information regarding future profits from medical treatment of a patient to be treated from the information of the patient to be treated, using a first model generated by machine learning with patient information as an explanatory variable and information regarding profits from medical treatment of the patient as a target variable; a display control means for displaying the result predicted by the first prediction means on a display means; An information processing device comprising:

2. The patient information used as an explanatory variable in generating the first model includes at least one of gender, age, disease name, and hospital history, the first prediction means predicts information regarding profits from medical treatment of the patient to be processed at one or more future timings; The information processing device according to claim 1 .

3. The method further comprises a second prediction means for predicting an index indicating the likelihood of reverse referral of a patient to be treated from the information of the patient to be treated, using a second model generated by machine learning with patient information as an explanatory variable and an index indicating the likelihood of reverse referral of the patient as a target variable; the display control means causes the display means to display the result predicted by the second prediction means; 3. The information processing device according to claim 1 or 2.

4. The patient information used as an explanatory variable in generating the second model includes at least one of information on reverse referral, gender, age, disease name, and hospital history, The second prediction means predicts an index indicating the likelihood of reverse referral of the patient to be treated at the time of prediction by the second prediction means. The information processing device according to claim 3 .

5. The system further includes a calculation means for calculating a score indicating the degree to which a reverse referral of the patient to be treated is recommended based on information regarding revenue from medical treatment of the patient to be treated and an index indicating the ease of reverse referral of the patient to be treated, the display control means causes the display means to display the score calculated by the calculation means; The information processing device according to claim 3 .

6. The calculation means calculates the score so that the higher the profit from medical treatment of the patient to be processed, the lower the score, and so that the higher the likelihood of reverse referral of the patient to be processed, the higher the score. The information processing device according to claim 5 .

7. a prediction step of predicting information regarding future revenue from medical treatment for a patient to be treated from the information of the patient to be treated using a model generated by machine learning with patient information as an explanatory variable and information regarding revenue from medical treatment for the patient as a target variable; a display control step of displaying the result predicted in the prediction step on a display means; An information processing method including:

8. Computer, a prediction means for predicting information about future revenues from medical treatment for a patient to be treated, based on the information about the patient to be treated, using a model generated by machine learning with patient information as an explanatory variable and information about revenues from medical treatment for the patient as a target variable; a display control means for displaying the result predicted by the prediction means on a display means; A program to function as a

Images