Selection assistance device, selection assistance method, and recording medium

Abstract

A selection assistance device according to the present invention is provided with an acquisition unit, an extraction unit, an estimation unit, and an output unit. The acquisition unit acquires a patient condition that is a medical condition required in a clinical trial patient and a selection condition that is a condition for selecting a clinical trial patient. The extraction unit extracts, as selection candidates from among patients receiving medical treatment, patients whose information regarding the medical treatment of the patient matches the patient condition. The estimation unit estimates, from among the selection candidates, a candidate for a clinical trial patient on the basis of the selection condition. The output unit outputs information regarding the estimated candidate for a clinical trial patient. The selection assistance device can provide assistance in the decision-making process for selecting a clinical trial patient by estimating a clinical trial patient candidate from among patients extracted on the basis of the patient condition and acquiring information regarding the patient to be subjected to the clinical trial.

Description

Selection support device, selection support method, and recording medium

[0001] The present disclosure relates to a selection support device and the like.

[0002] In clinical trials of pharmaceuticals, patients who meet the selection criteria for clinical trial patients are selected based on, for example, the descriptions in medical records. In selecting clinical trial patients, the person in charge, for example, extracts patients who may be suitable as clinical trial patients from the descriptions in the medical records. Then, the person in charge, for example, carefully checks the contents of the medical records to select patients who are presumed to be suitable for the clinical trial as candidate clinical trial patients. In addition, a system that supports the selection of candidate clinical trial patients may be used to select candidate clinical trial patients.

[0003] The clinical trial matching device of Patent Document 1 stores data that associates the medication conditions with the clinical trial results for each of a plurality of clinical trials, and searches for a clinical trial that matches a patient from among the plurality of clinical trials based on the patient's information.

[0004] International Publication No. 2021 / 145439

[0005] The technology described in Patent Document 1 may not be accurate enough in estimating patients suitable for clinical trials.

[0006] In order to solve the above-mentioned problems, the present disclosure aims to provide a selection support device that can improve the accuracy of estimating patients suitable for clinical trials.

[0007] In order to solve the above problems, a selection support device according to one aspect of the present disclosure includes an acquisition means for acquiring patient conditions, which are medical conditions required of clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients; an extraction means for extracting, from among patients receiving treatment, patients whose information regarding the patient's clinical trial matches the patient conditions as selection candidates; an estimation means for inferring candidate clinical trial patients from among the selection candidates based on the selection conditions; and an output means for outputting information regarding the inferred candidate clinical trial patients.

[0008] A selection support method according to one aspect of the present disclosure acquires patient conditions, which are medical conditions required of clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients; extracts, from among patients receiving treatment, patients whose information about the patient's clinical trial matches the patient conditions as selection candidates; predicts, from among the selected candidates, potential clinical trial patients based on the selection conditions; and outputs information about the predicted potential clinical trial patients.

[0009] A recording medium according to one embodiment of the present disclosure non-temporarily records a selection support program that causes a computer to execute the following processes: a process of acquiring patient conditions, which are medical conditions required of clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients; a process of extracting, from among patients receiving treatment, patients whose information about the patient's clinical trial matches the patient conditions as selection candidates; a process of inferring candidate clinical trial patients from among the selected candidates based on the selection conditions; and a process of outputting information about the inferred candidate clinical trial patients.

[0010] According to the present disclosure, it is possible to improve the accuracy of estimating patients suitable for clinical trials.

[0011] FIG. 1 is a diagram illustrating an example of the configuration of a clinical trial matching system according to the present disclosure. FIG. 2 is a diagram illustrating an example of the configuration of a selection support device according to the present disclosure. FIG. 3 is a diagram illustrating an example of information about candidate clinical trial patients according to the present disclosure. FIG. 4 is a diagram illustrating an example of information about candidate clinical trial patients according to the present disclosure. FIG. 5 is a diagram illustrating an example of the operation flow of a selection support device according to the present disclosure. FIG. 6 is a diagram illustrating an example of the hardware configuration of a selection support device according to the present disclosure.

[0012] Embodiments of the present disclosure will be described in detail with reference to the drawings. FIG. 1 is a diagram illustrating an example of the configuration of a clinical trial matching system. The clinical trial matching system includes, for example, a selection support device 10, a data management device 20, a terminal device 30, and a medical terminal device 40. The selection support device 10 is connected to the data management device 20, for example, via a network. The selection support device 10 is also connected to the terminal device 30, for example, via a network. The data management device 20 is also connected to the medical terminal device 40, for example, via a network. There may be multiple data management devices 20, multiple terminal devices 30, and multiple medical terminal devices 40. The number of data management devices 20, multiple terminal devices 30, and multiple medical terminal devices 40 can be set as appropriate.

[0013] A clinical trial matching system, for example, is a system that predicts candidate clinical trial patients. Clinical trial matching, for example, is a system that predicts candidate clinical trial patients and matches institutions conducting clinical trials with patients who are likely to be suitable for the clinical trial. The clinical trial matching system predicts candidate clinical trial patients based, for example, on selection conditions for clinical trial patients and information about the treatment of patients receiving treatment at medical institutions. The selection conditions are, for example, conditions that clinical trial patients must meet when conducting a clinical trial. The selection conditions may also be, for example, conditions that clinical trial patients are desirable to meet when conducting a clinical trial. For example, the selection conditions may be conditions that allow the clinical trial to be conducted efficiently when the clinical trial patient meets the selection conditions. For example, a clinical trial is more likely to be conducted if the patient's condition meets medical conditions. On the other hand, for example, the more clinical trial patients are receiving treatment at the same hospital, the more efficient the clinical trial can be. In such cases, the selection conditions are set, for example, as a condition that the patients receive treatment at the same hospital. The selection conditions may also specify the name of the hospital.

[0014] A candidate clinical trial patient is, for example, a patient who is a candidate for a clinical trial. That is, a candidate clinical trial patient is, for example, a patient who is highly likely to be suitable for a clinical trial. A candidate clinical trial patient is, for example, a patient who is used to estimate the number of patients to be included in a clinical trial when a clinical trial protocol is being prepared. Furthermore, a candidate clinical trial patient is a patient who is selected when selecting patients who will actually be included in the clinical trial when the clinical trial is being conducted.

[0015] A clinical trial is, for example, a clinical trial conducted to obtain legal approval for the manufacture and sale of pharmaceuticals or medical devices. A clinical trial patient is, for example, a patient who is in a suitable condition to be a subject of a clinical trial. A suitable subject of a clinical trial is, for example, a patient who is likely to be able to obtain significant data in a clinical trial to verify the effectiveness and safety of a new drug or a new medical device. A new drug is, for example, a newly developed pharmaceutical product with no clinical track record. An institution that conducts a clinical trial is, for example, a pharmaceutical company. The institution that conducts a clinical trial may be an institution that has been entrusted with clinical trial work by a pharmaceutical company. An institution that has been entrusted with clinical trials by a pharmaceutical company is, for example, a CRO (Contract Research Organization).

[0016] Information regarding patient treatment is, for example, information indicating the patient's condition that may affect the outcome of the clinical trial. Information regarding patient treatment is the attributes and medical records of candidate clinical trial patients. Patient attributes are, for example, patient-specific information that is largely unchanged by treatment. Patient attributes are, for example, one or more of the patient's age, gender, medical history, family medical history, employment history, height, weight, and race. Patient attributes are not limited to the above. Medical records are, for example, information recorded in a medical chart. Medical records are records of one or more items of diagnosis, tests, medication, surgery, follow-up, and patient condition. Medical records may include the condition and progression of the disease. Medical records may include test data. Medical records may include one or more items of information other than those recorded in the medical chart, such as patient disease information, complication information, biomarkers, disease status, guideline score, medical history, effects, and test results. Medical records may also include information recorded on a medical prescription. Medical records are not limited to the above.

[0017] The information regarding the patient's treatment may include information indicating whether the patient has consented to the clinical trial. The information regarding the patient's treatment may include a desired amount of remuneration to be paid for the clinical trial. The remuneration to be paid for the clinical trial is, for example, money paid by the institution conducting the clinical trial to at least one of the doctor, medical institution, and patient when the clinical trial is conducted. The information regarding the patient's treatment may also include the environment in which the patient is receiving treatment. The environment in which the patient is receiving treatment is, for example, the region, hospital, ward, or floor in which the patient is hospitalized. The floor is, for example, the floor number or medical department within the ward. The location in which the treatment is being received is not limited to the above. The environment in which the patient is receiving treatment may also include information about the doctor treating the patient. In this case, the selection conditions include at least one of the region, hospital, ward, and floor in which the patient is hospitalized.

[0018] The information regarding the patient's treatment may be, for example, pseudonymized information. Pseudonymized information is, for example, called "pseudonymized information." Pseudonymized information is information from which information that identifies an individual can be restored when compared with other information. The information regarding the patient who is the subject of the clinical trial may be anonymized information. Pseudonymized information is, for example, called "anonymized information." Anonymously processed information is information from which information that identifies an individual cannot be restored even when compared with other information. For example, when the selection support device 10 is used by an entity authorized to view the information, the information regarding the patient's treatment may include the patient's real name. The entity authorized to view the data may be, for example, a person in charge of a medical institution or a person in charge of an institution entrusted by the medical institution with the clinical trial. The person in charge of the medical institution may be, for example, a doctor. The person in charge of the medical institution may be a nurse, pharmacist, clinical laboratory technician, physical therapist, or counselor. The person in charge of the medical institution is not limited to the above.

[0019] Here, an example of the configuration of the selection support device 10 will be described. Fig. 2 is a diagram showing an example of the configuration of the selection support device 10. The selection support device 10 basically includes an acquisition unit 11, an extraction unit 12, an estimation unit 13, and an output unit 15. The selection support device 10 may further include, for example, a generation unit 14 and a storage unit 16.

[0020] The acquisition unit 11 acquires patient conditions, which are medical conditions required for clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients. The patient conditions are, for example, selection criteria for clinical trial patients. The selection criteria include, for example, criteria for inclusion of patients in the clinical trial and criteria for exclusion of patients from the clinical trial. The inclusion criteria are, for example, criteria for patients who are suitable for the clinical trial. That is, the inclusion criteria are criteria for patients who can be subjects of the clinical trial if they meet the conditions indicated by the inclusion criteria. The exclusion criteria are criteria for patients who are not suitable for the clinical trial. That is, the exclusion criteria are criteria for patients who are not suitable for the clinical trial if they meet the conditions indicated by the exclusion criteria. The inclusion criteria and the exclusion criteria may each include multiple criteria. The patient conditions may be some of the criteria included in the selection criteria. The patient conditions may also be criteria that relax the conditions indicated by some of the criteria included in the selection criteria. For example, if the selection criteria stipulate that the patient must be 40 years of age or older, the patient conditions may be set as being 35 years of age or older. By using some of the selection criteria or criteria that relax some of the selection criteria as patient conditions, it is possible to extract a wide range of patients who could be candidates for clinical trial patients in the extraction unit 12. Furthermore, the patient conditions are not limited to the selection criteria for clinical trial patients.

[0021] The selection conditions are, for example, conditions desired by the entity conducting the clinical trial for the clinical trial patients. The entity conducting the clinical trial is, for example, a pharmaceutical company. The selection conditions are, for example, conditions that must be met by the clinical trial patients other than the selection criteria. The selection conditions may include part of the selection criteria. The selection conditions may also be conditions that further limit the conditions indicated by the criteria included in the selection criteria. For example, if the selection criteria state that the age must be 40 years or older, the selection conditions may also be set to state that the age must be 50 years or older and younger than 55 years old.

[0022] When there are multiple entities conducting the clinical trial, the acquisition unit 11 may acquire selection conditions from each of the entities conducting the clinical trial. The acquisition unit 11 may acquire patient conditions and selection conditions from each of the entities conducting the clinical trial. Each of the entities conducting the clinical trial is, for example, a pharmaceutical company. For example, if pharmaceutical company A and pharmaceutical company B are seeking candidates for clinical trial patients, the acquisition unit 11 acquires the selection conditions of each pharmaceutical company from pharmaceutical company A and pharmaceutical company B. The entity conducting the clinical trial may also be a medical institution. The entity conducting the clinical trial is not limited to the above.

[0023] The acquisition unit 11 may acquire a priority for each condition included in the selection conditions. The priority is, for example, an index indicating how much importance is attached to each condition included in the selection conditions. In other words, the priority is, for example, the weight of each condition included in the selection conditions in selecting candidate clinical trial patients. For example, if the selection conditions include age, hospital where patient is hospitalized, and pre-existing conditions, the priority is the weight of age, hospital where patient is hospitalized, and pre-existing conditions in selecting candidate clinical trial patients.

[0024] The acquisition unit 11 may acquire the patient conditions and the selection conditions at different times. For example, the acquisition unit 11 may acquire the selection conditions after the extraction unit 12 acquires patients who meet the patient conditions as selection candidates based on the patient conditions. Furthermore, the acquisition unit 11 may acquire selection conditions in which some of the conditions have been changed after acquiring the selection conditions. The acquisition unit 11 acquires the patient conditions and the selection conditions from, for example, the terminal device 30.

[0025] The extraction unit 12 extracts, as selection candidates, patients whose information about their treatments satisfies the patient conditions from among patients receiving treatment. As a process for extracting selection candidates, the extraction unit 12 extracts, for example, identifiers of the selection candidates and information about the treatments of the selection candidates. Patients receiving treatment are, for example, patients whose information about their treatments is stored in the data management device 20. In other words, the extraction unit 12 extracts, as selection candidates, patients who satisfy the patient conditions from among patients whose information about their treatments is stored in the data management device 20.

[0026] As a process of extracting selection candidates, the extraction unit 12 outputs a query including patient conditions to the data management device 20, for example. Then, the extraction unit 12 extracts selection candidates by obtaining information about the selection candidates extracted based on the query from the data management device 20. Furthermore, in the process of extracting selection candidates, the extraction unit 12 obtains, for example, information about the selection candidates along with information about the selection candidates. Furthermore, the extraction unit 12 may access a storage area of ​​the data management device 20 and extract patients who meet the patient conditions as selection candidates from among patients whose information about treatments is stored in the storage area.

[0027] Furthermore, for each of the extracted selection candidates, the extraction unit 12 extracts information on items from the information on treatment that the estimation unit 13 uses to estimate candidate clinical trial patients. For example, the extraction unit 12 extracts information on items set as extraction items from the information on patient treatment stored in the data management device 20. The extraction items indicate, for example, the content of information to be extracted from the information on patient treatment. For example, when information on the disease name, symptoms, age, ward where treatment is being received, and medication history is extracted from the information on treatment, the disease name, symptoms, age, ward where treatment is being received, and medication history are set as extraction items. The extraction items may, for example, be included in the patient conditions.

[0028] Extraction items are set based on, for example, the use of the new drug that is the subject of the clinical trial. For example, the information that should be referenced when selecting candidate patients for a clinical trial is significantly different between lung cancer and cerebral infarction. Therefore, extraction items are set according to the disease that the new drug that is the subject of the clinical trial is intended to treat.

[0029] The extracted items may include information indicating whether or not the patient has consented to the clinical trial. In this case, the extraction unit 12 acquires, for example, information about the treatment of the selected candidate, including whether or not the patient has consented to the clinical trial. The extracted items may also include the amount of remuneration for the clinical trial desired by at least one of the doctor, medical institution, and patient.

[0030] When the extraction unit 12 accesses the storage area of ​​the data management device 20 to extract selection candidates, the extraction unit 12 may use an extraction model to extract patients who meet the patient conditions. For example, the extraction unit 12 searches for patients whose information on treatment matches the patient conditions. Matching may include similarity. The extraction model is, for example, a machine learning model that extracts selection candidates using the patient conditions as input.

[0031] The extraction model is generated, for example, by learning the relationship between information about patient conditions and patient treatment and suitability as a clinical trial candidate. The extraction model is generated, for example, by converting information about patient conditions and patient treatment into feature vectors and learning the relationship between the two feature vectors and suitability as a clinical trial candidate. For example, Word2Vec can be used for the conversion to feature vectors. A language model other than Word2Vec may also be used for the conversion to feature vectors. The extraction model is generated, for example, by deep learning using a neural network. Alternatively, the extraction model may convert information about patient conditions and patient treatment into feature vectors and extract selection candidates based on the Euclidean distance or cosine similarity between the two vectors. The learning algorithm for generating the extraction model is not limited to the above. The extraction model may also be generated, for example, in a system external to the data management device 20. The extraction model may also be generated by the generation unit 14.

[0032] The extraction unit 12 may use a language model to extract information about patients who are clinical trial subjects. For example, the extraction unit 12 outputs patient conditions and a prompt to the language model requesting the language model to extract patients who meet the patient conditions. The language model then extracts selection candidates from patients whose treatment-related information is stored based on the patient conditions and the prompt. For example, the language model may use Generative Pre-trained Transformer-2 (GPT-2), GPT-3, GPT-3.5, or GPT-4. Furthermore, for the estimation model, T5 (Text-to-Text Transfer Transformer), BERT (Bidirectional Encoder Representations from Transformers), RoBERTa (Robustly optimized BERT approach), or ELECTRA (Efficiently Learning an Encoder that Classifies Token Replacements Accurately) may be used. The language model used for the extraction model is not limited to the above.

[0033] The extraction unit 12 may extract selection candidates by comparing information about the patient's treatment extracted using the language model with patient conditions. For example, the extraction unit 12 outputs a prompt to the language model to extract information about the patient's age, disease name, and symptoms from the information about the patient's treatment. The extraction unit 12, for example, determines the degree of match between the information extracted by the language model and the patient conditions. Then, the extraction unit 12 extracts patients whose degree of match is equal to or exceeds a standard as selection candidates. The standard for the degree of match is set, for example, so that the extracted patients are patients who may be suitable for the clinical trial.

[0034] When the information related to treatment is pseudonymized information or anonymously processed information, the extraction unit 12 extracts, for example, information related to the selected candidate in a pseudonymized or anonymous state. Pseudonymized information is, for example, information from which information that identifies an individual can be restored when compared with other information. Anonymously processed information is, for example, information from which information that identifies an individual cannot be restored even when compared with other information.

[0035] The estimation unit 13 estimates candidate clinical trial patients from among the selected candidates based on the selection conditions. When selection criteria have been acquired from the entities conducting the clinical trials, the estimation unit 13 estimates candidate clinical trial patients suitable for each of the entities conducting the clinical trials, for example, based on the selection conditions. In other words, the estimation unit 13 estimates candidate clinical trial patients to be assigned to each of the entities conducting the clinical trials, for example, based on the selection conditions. The estimation unit 13 estimates candidate clinical trial patients suitable for each of the entities conducting the clinical trials, for each of the entities conducting the clinical trials, based on the selection conditions created by each of the entities conducting the clinical trials.

[0036] The estimation unit 13 estimates candidate clinical trial patients from selected candidates using an estimation model. The candidate clinical trial patients are, for example, patients who meet selection conditions. The estimation model is, for example, a machine learning model that uses the selection conditions as input and estimates candidate clinical trial patients from selected candidates.

[0037] The estimation model, for example, converts the selection conditions and information about the treatment of the selection candidate into feature vectors. Then, the estimation model estimates the compatibility of the information about the treatment with the selection conditions based on the feature vectors converted from the selection conditions and the feature vectors converted from the information about the treatment. Then, patients whose compatibility meets or exceeds a standard are predicted as candidates for clinical trial patients. The compatibility standard is set so that meaningful data can be obtained in the clinical trial when the patients who are clinical trial patients are selected as clinical trial subjects.

[0038] When assigning clinical trial candidates to entities conducting clinical trials, the estimation model may, for example, assign each selected candidate to the pharmaceutical company corresponding to the selection criteria with the highest degree of fit. The estimation model may randomly assign each selected candidate whose degree of fit meets a standard to an entity conducting clinical trials. The estimation model may assign each selected candidate whose degree of fit meets a standard to an entity conducting clinical trials so that the number of people assigned to the entities conducting clinical trials is uniform.

[0039] The estimation model may also estimate the degree of fit for each condition included in the selection conditions. When estimating the degree of fit for each condition included in the selection conditions, the estimation model may, for example, estimate, as a candidate clinical trial patient, a patient whose information on treatment satisfies a criterion for all conditions included in the selection conditions. When estimating the degree of fit for each condition included in the selection conditions, the estimation model may, for example, estimate, as a candidate clinical trial patient, a patient whose average degree of fit satisfies a criterion for all conditions included in the selection conditions.

[0040] The estimation unit 13 may estimate candidate clinical trial patients based on the priority of each condition included in the selection conditions. For example, the estimation unit 13 estimates candidate clinical trial patients by using the priority of each condition included in the selection conditions as a weight when the estimation model makes an estimation. In this case, for example, parameters in the estimation model are set so that a greater weight is assigned to a variable corresponding to a condition with a higher priority.

[0041] The estimation unit 13 may estimate candidate clinical trial patients using an estimation model according to the disease for which the new drug being tested is to be used. For example, the estimation unit 13 estimates candidate clinical trial patients using an estimation model according to the site where the disease occurs or the name of the disease. The name of the disease is, for example, the name of the disease. For example, the estimation unit 13 estimates candidate clinical trial patients using different estimation models for a cancer treatment drug and a viral infection treatment drug. The different estimation models are, for example, learning models with different learning data and / or learning algorithms. The estimation unit 13 may estimate candidate clinical trial patients using an estimation model according to the medical department in which the new drug is expected to be used. For example, the estimation unit 13 estimates candidate clinical trial patients using different estimation models for a new drug expected to be used in neurosurgery and a new drug expected to be used in gastroenterology.

[0042] The estimation unit 13 may estimate candidate clinical trial patients using, for example, a language model. For example, the estimation unit 13 outputs selection conditions and a prompt to the language model requesting extraction of patients who meet the selection conditions. For example, the language model used may be a language model that differs from the language model used by the extraction unit 12 in at least one of training data and algorithm. For example, the language model used may be a language model that differs from the language model used by the extraction unit 12 in at least one of training data and algorithm. For example, the language model may be GPT-2, GPT-3, GPT-3.5, or GPT-4. Furthermore, the estimation model may be T5, BERT, RoBERTa, or ELECTRA. The language model used to estimate candidate clinical trial patients is not limited to the above.

[0043] The estimation unit 13 may also calculate the degree of compatibility based on the Euclidean distance or cosine similarity between feature vectors. When calculating the degree of compatibility based on the Euclidean distance or cosine similarity between feature vectors, the estimation unit 13 estimates, for example, the degree of compatibility of information related to treatment with respect to the selection conditions. Then, patients whose degree of compatibility is equal to or exceeds a standard are estimated as candidates for clinical trial patients.

[0044] The estimation unit 13 may estimate candidate clinical trial patients based on the patient conditions and the selection conditions. For example, when patient conditions and selection conditions are acquired from each entity conducting a clinical trial, the estimation unit 13 estimates candidate clinical trial patients that meet the conditions of each entity conducting a clinical trial using an estimation model based on the patient conditions and selection conditions. In this case, the estimation model estimates candidate clinical trial patients that meet the conditions of each entity conducting a clinical trial based on, for example, the suitability of the patient's treatment for the patient conditions and selection conditions. The estimation unit 13 may assign the selected candidate as a candidate clinical trial patient to the entity conducting the clinical trial that presents the patient conditions and selection conditions with the highest suitability based on the suitability estimated by the estimation model. In this case, the estimation model estimates, for example, the suitability of the patient's treatment for the patient conditions and selection conditions.

[0045] The estimation unit 13 may calculate the degree of conformance between the patient conditions and selection conditions and the information on the patient's treatment, and assign the selection candidate as a candidate clinical trial patient to the entity with the highest degree of conformance. The estimation unit 13, for example, converts the patient conditions and selection conditions and the information on the patient's treatment into feature vectors, respectively. Then, the estimation unit 13 estimates the degree of conformance of the information on the treatment to the patient conditions and selection conditions, for example, based on the feature vectors converted from the patient conditions and selection conditions and the feature vectors converted from the information on the treatment. Then, the estimation unit 13 predicts patients whose degree of conformance is equal to or exceeds a standard as candidate clinical trial patients. The standard of conformance is set so that meaningful data can be obtained in the clinical trial when the patient as a clinical trial patient is selected as a clinical trial subject.

[0046] The estimation unit 13 may assign the selected candidates to entities conducting the clinical trial based on the amount of remuneration to be paid for conducting the clinical trial. In this case, the selection conditions include, for example, information on the amount of remuneration to be paid for conducting the clinical trial. For example, when the same patient meets the selection conditions for entities conducting the clinical trial, the estimation unit 13 assigns the selected candidate as a candidate clinical trial patient to the entity that will pay the highest amount of remuneration for conducting the clinical trial. The estimation unit 13 may also assign the selected candidates to entities conducting the clinical trial based on the amount of remuneration desired by at least one of the doctor, medical institution, and patient. For example, the estimation unit 13 assigns the selected candidates to entities conducting the clinical trial in order of the lowest desired remuneration among patients who meet the selection conditions.

[0047] The estimation unit 13 may select clinical trial patients to be assigned to each entity conducting the clinical trial based on the environment in which the selected candidates are receiving treatment. The environment in which the selected candidates are receiving treatment is, for example, the location where the treatment is being received. The location where the treatment is being received is, for example, the region, hospital, ward, or floor where the patient is hospitalized. The floor is, for example, the floor number or medical department within the ward. The location where the treatment is being received is not limited to the above. In this case, the selection conditions include at least one of the region, hospital, ward, and floor where the patient is hospitalized.

[0048] The generation unit 14 generates the estimation model by, for example, machine learning the relationship between the selection conditions and information on treatment and suitability as a clinical trial patient. The information on treatment is, for example, information corresponding to the extracted items among the information on treatment. Suitability as a clinical trial patient is determined, for example, using performance data indicating whether the patient was actually selected as a clinical trial patient. That is, the generation model is generated so that the degree of conformance to the selection criteria is high when, for example, a patient whose treatment information is used for learning has actually been selected as a clinical trial patient. The estimation model is generated, for example, by converting the selection conditions and information on treatment into feature vectors and learning the relationship between the feature vectors and suitability as a clinical trial patient. The generation unit 14 may generate the estimation model by, for example, machine learning the relationship between the patient conditions, selection conditions, information on treatment, and suitability as a clinical trial patient. The estimation model is generated, for example, by deep learning using a neural network. The learning algorithm for generating the estimation model is not limited to the above. The estimation model may be generated, for example, in a system external to the selection support device 10.

[0049] The generation unit 14 may retrain the estimation model. For example, the generation unit 14 retrains the estimation model using information indicating whether the estimated clinical trial patient candidate is selected as an actual clinical trial patient as a label. For example, by retraining based on performance data corresponding to the disease to which the new drug undergoing clinical trials is applied, it is possible to estimate the clinical trial patient candidate using an estimation model optimized according to the disease.

[0050] The output unit 15 outputs information about the candidate clinical trial patients estimated by the estimation unit 13. The output unit 15 outputs, for example, the number of candidate clinical trial patients estimated by the estimation unit 13. The output unit 15 may output, for example, the candidate clinical trial patients estimated by the estimation unit 13. The output unit 15 may output information about the treatment of each candidate clinical trial patient estimated by the estimation unit 13. Items included in the information about the treatment of each patient are set according to the purpose of the information. For example, an item indicates the classification of information included in the item. For example, an item is information for understanding the content of the information included in the item. For example, an item corresponding to information indicating the name of a hospital where a candidate clinical trial patient is receiving treatment is "hospital." The purpose of the information is, for example, a task performed using the information about the candidate clinical trial patients. For example, the purpose of the information is the creation of a clinical trial protocol. The purpose of the information may be the selection of patients at the stage of conducting a clinical trial. For example, at the stage of creating a clinical trial protocol, it is sufficient to know the approximate number of patients to be included in the clinical trial, whereas detailed information about the patient's condition may be required when selecting patients at the stage of conducting the clinical trial. For this reason, when used in patient selection at the stage of conducting a clinical trial, more items may be output than when used in creating a clinical trial protocol.

[0051] The output unit 15 may output candidate clinical trial patients assigned to each entity conducting the clinical trial. Each entity conducting the clinical trial is, for example, a pharmaceutical company conducting the clinical trial. The output unit 15 may output candidate clinical trial patients assigned to each entity conducting the clinical trial and information regarding the treatment of each patient.

[0052] The output unit 15 may output information indicating whether or not each of the candidate clinical trial patients estimated by the estimation unit 13 has consented to the clinical trial. The output unit 15 may also output information indicating the remuneration desired for the clinical trial by at least one of the medical institution and the patient, for each of the candidate clinical trial patients estimated by the estimation unit 13.

[0053] When candidate clinical trial patients are estimated based on multiple selection conditions, the output unit 15 may output the number of candidate clinical trial patients for each selection condition. Furthermore, when candidate clinical trial patients are estimated based on multiple selection conditions, the output unit 15 may output information regarding the treatment of each candidate clinical trial patient for each selection condition.

[0054] When the information regarding the patient's treatment is pseudonymized or anonymized, the output unit 15 may output the information regarding the treatment of the candidate clinical trial patient in a pseudonymized or anonymized state.

[0055] The output unit 15 may output information about the candidate clinical trial patients when some of the patient conditions are relaxed. The output unit 15 may also output information about the candidate clinical trial patients when some of the selection conditions are relaxed. The output unit 15 also outputs information about the candidate clinical trial patients estimated by the estimation unit 13 to, for example, the terminal device 30.

[0056] FIG. 3 is an example of information about candidate clinical trial patients output by the output unit 15. In the example of FIG. 3, information corresponding to "disease name," "symptoms," and "hospital" is output for each candidate clinical trial patient. The "disease name" in the example of FIG. 3 is, for example, the name of the disease from which each patient is suffering. The "symptoms" in the example of FIG. 3 are symptoms that the patient is experiencing. Furthermore, the "hospital" in the example of FIG. 3 is, for example, the name of the hospital where the patient is receiving treatment. Furthermore, in the example of FIG. 3, patients are identified using "patient numbers." Therefore, the information about candidate clinical trial patients shown in the example of FIG. 3 is pseudonymized information.

[0057] FIG. 4 shows an example of information about candidate clinical trial patients when the candidate clinical trial patients are assigned to multiple pharmaceutical companies. In the example of FIG. 4 , similar to the example of FIG. 3 , information corresponding to "disease name," "symptoms," and "hospital" is output for each candidate clinical trial patient. Also, in the example of FIG. 4 , the entity to which each candidate clinical trial patient is assigned is output as "pharmaceutical company." In the example of FIG. 4 , "pharmaceutical company" is, for example, the name of the pharmaceutical company to which each candidate clinical trial patient is assigned. Also, in the example of FIG. 4 , the output unit 15 may output information about candidate clinical trial patients so that pharmaceutical companies to which the candidate clinical trial patients are assigned can view only the candidate clinical trial patients assigned to their company. Also, in the example of FIG. 4 , the output unit 15 may output information about candidate clinical trial patients by sorting by pharmaceutical company.

[0058] FIG. 5 shows an example of outputting the number of clinical trial patient candidates who meet the selection criteria as information regarding clinical trial patient candidates. In the example of FIG. 5, "consent," "hospital," "disease name," "symptoms," and "number of patients" are output as "items." In the example of FIG. 5, "item" is the name of the item. The name of the item indicates the type of information included. In the example of FIG. 5, "disease name" is, for example, the name of the disease each patient suffers from. In the example of FIG. 5, "symptoms" are symptoms that the patients are experiencing. In addition, in the example of FIG. 5, "hospital" is, for example, the name of the hospital where the patient is receiving treatment. In addition, in the example of FIG. 5, "number of patients" is, for example, the number of clinical trial patient candidates who meet the selection criteria. For example, a pharmaceutical company representative uses the "number of patients" in the example of FIG. 5 as a reference when creating a clinical trial protocol.

[0059] The memory unit 16 stores information related to the process of estimating candidate clinical trial patients. The memory unit 16 stores, for example, information related to the selected candidates extracted by the extraction unit 12. The memory unit 16 stores, for example, the selected candidates extracted by the extraction unit 12 and information related to the treatment of each of the selected candidates. The memory unit 16 may also store candidate clinical trial patients estimated by the estimation unit 13 in association with information related to the treatment of each patient. The memory unit 16 stores, for example, an extraction model. The memory unit 16 also stores, for example, an estimation model. The extraction model and the estimation model may be stored in a storage means other than the memory unit 16.

[0060] The data management device 20 stores, for example, information related to the patient's treatment. The information related to the patient's treatment is, for example, data in an electronic medical record entered by a doctor. The information in the electronic medical record may also be data entered by a nurse or a medical technician. The information related to the patient's treatment may also be information on one or more items of the patient's disease information, complication information, biomarkers, disease status, guideline score, medical history, effects, and test results other than those described in the electronic medical record. The data management device 20 may store the information related to the patient's treatment as anonymously processed information or pseudonymized information. The data management device 20 outputs the information related to the patient's treatment to, for example, the extraction unit 12 of the selection support device 10.

[0061] The terminal device 30 is, for example, a terminal device used by a person in charge of conducting a clinical trial. The person in charge of conducting the clinical trial is a person in charge of a pharmaceutical company or a person in charge of an institution that has been entrusted with work related to the clinical trial by the pharmaceutical company. The person in charge of an institution that has been entrusted with the clinical trial by the pharmaceutical company is, for example, a CRA (Clinical Research Associate). The CRA may be an employee of the pharmaceutical company.

[0062] The terminal device 30 may be used by a person in charge of a medical institution that is requested by a pharmaceutical company to conduct a clinical trial. For example, the person in charge of the medical institution receives an inquiry from the pharmaceutical company regarding candidate clinical trial patients. The person in charge of the medical institution then operates the terminal device 30 to obtain information regarding the candidate clinical trial patients and respond to the pharmaceutical company. The person in charge of the medical institution is, for example, a doctor or a nurse. The person in charge between medical areas may be a staff member of the hospital in charge of the clinical trial. The person in charge of the medical institution is not limited to the above. The person in charge of the medical institution may also be a person in charge of an institution that has been entrusted by the medical institution to handle the clinical trial. The institution that has been entrusted by the medical institution to handle the clinical trial is, for example, an SMO (Site Management Organization). The person in charge of an institution that has been entrusted by the medical institution to handle the clinical trial is, for example, a CRC (Clinical Research Coordinator).

[0063] The terminal device 30 acquires, for example, patient conditions and selection conditions input by the person in charge of conducting the clinical trial. Then, the pharmaceutical company's terminal device 30 outputs, for example, the patient conditions and selection conditions to the acquisition unit 11 of the selection support device 10. The terminal device 30 also acquires information about candidate clinical trial patients from, for example, the output unit 15 of the selection support device 10. Then, the terminal device 30 outputs, for example, the information about candidate clinical trial patients to a display device (not shown).

[0064] The medical terminal device 40 is, for example, a terminal device that inputs information about a treatment. The information about a patient's treatment is, for example, data from an electronic medical record entered by a doctor. The information about a patient's treatment may also be data entered by a nurse or a laboratory technician. The information about a patient's treatment may also be information about one or more items of the patient's disease information, complication information, biomarkers, disease status, guideline score, medical history, effects, and test results, other than those listed in the electronic medical record. The information about treatment may also be information about prescribed medications. Information about clinical trials among the information about treatment may be input by a person in charge of an institution entrusted with the work by a medical institution. An institution entrusted with handling clinical trials by a medical institution may be, for example, a site management organization (SMO). An institution entrusted with handling clinical trials by a hospital may be, for example, a clinical research coordinator (CRC). The medical terminal device 40 outputs information about treatment to, for example, the data management device 20.

[0065] The following describes the operation of the selection support device 10 to estimate candidate clinical trial patients. Fig. 6 is a diagram showing an example of the flow of processing in the selection support device 10 to estimate patients who will be subject to clinical trials.

[0066] The acquisition unit 11 acquires patient conditions, which are medical conditions required for clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients (step S11). The acquisition unit 11 acquires the patient conditions and selection conditions from, for example, the terminal device 30.

[0067] After acquiring the patient conditions and the selection conditions, the extraction unit 12 extracts, as selection candidates, patients whose treatment-related information matches the patient conditions (step S12). The extraction unit 12 extracts selection candidates from patients whose treatment-related information is stored in the data management device 20, for example.

[0068] When the selection candidates are extracted, the estimation unit 13 estimates candidates for clinical trial patients from among the selection candidates based on the selection conditions (step S13). That is, the estimation unit 13 estimates, for example, the entities to which the selected candidates will be assigned. If the selected candidate does not satisfy any of the entities' selection conditions, the estimation unit 13, for example, excludes the selected candidate from the subjects to be assigned.

[0069] When the allocation destination estimation for the selected candidate is completed (Yes in step S14), the output unit 15 outputs information about the selected clinical trial patient candidate (step S15).

[0070] If there are candidates for which allocation estimation has not been completed (No in step S14), the process returns to step S13, and the estimation unit 13 estimates candidates for clinical trial patients from among the selected candidates, for example, based on the selection conditions.

[0071] The selection support device 10 acquires patient conditions, which are medical conditions required for clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients. The selection support device 10 also extracts patients whose treatment information meets the patient conditions as selection candidates. The selection support device 10 then predicts candidate clinical trial patients from the selection candidates based on the selection conditions. In this way, by extracting candidate clinical trial patients based on medical conditions and predicting patients who meet the conditions for selecting clinical trial patients from the extracted patients, the selection support device 10 can improve the accuracy of predicting patients who are suitable for clinical trials.

[0072] Furthermore, by estimating candidate clinical trial patients suitable for each entity conducting the clinical trial, the selection support device 10 can accurately match patients with, for example, each pharmaceutical company that is the entity conducting the clinical trial.

[0073] Furthermore, by predicting candidate clinical trial patients based on selection criteria including at least one of the region, hospital, ward, and floor where the patient is hospitalized, the selection support device 10 can predict, for example, candidate clinical trial patients for whom a clinical trial can be efficiently conducted. Furthermore, by predicting patients who will be clinical trial subjects from among the selected candidates based on the selection criteria, the selection support device 10 can support, for example, decision-making in selecting clinical trial patients.

[0074] Each process in the selection support device 10 may be distributed and executed in a plurality of information processing devices connected via a network. For example, the processes in the extraction unit 12 and the estimation unit 13 and the generation unit 14 may be performed in different information processing devices. Also, for example, the processes in the extraction unit 12 and the estimation unit 13 may be performed in different information processing devices. It can be set as appropriate which information processing device performs each process in the selection support device 10.

[0075] Each process in the selection support device 10 can be realized by executing a computer program on a computer. Fig. 7 shows an example of the configuration of a computer 100 that executes a computer program that performs each process in the selection support device 10. The computer 100 includes a CPU (Central Processing Unit) 101, a memory 102, a storage device 103, an input / output I / F (Interface) 104, and a communication I / F 105.

[0076] The CPU 101 reads and executes computer programs for performing each process from the storage device 103. The CPU 101 may be configured with a combination of multiple CPUs. The CPU 101 may also be configured with a combination of a CPU and another type of processor. For example, the CPU 101 may be configured with a combination of a CPU and a graphics processing unit (GPU). The memory 102 is configured with a dynamic random access memory (DRAM) or the like, and temporarily stores the computer programs executed by the CPU 101 and data being processed. The storage device 103 stores the computer programs executed by the CPU 101. The storage device 103 is configured with, for example, a non-volatile semiconductor storage device. Other storage devices such as a hard disk drive may also be used for the storage device 103. The input / output I / F 104 is an interface that accepts input from an operator and outputs display data, etc. The communication I / F 105 is an interface that transmits and receives data to and from other information processing devices. The data management device 20, the terminal device 30, and the medical terminal device 40 may also have a configuration similar to that of the computer 100.

[0077] The computer program used to execute each process can also be stored and distributed on a computer-readable recording medium that non-temporarily stores data. Examples of recording media that can be used include magnetic tapes for recording data and magnetic disks such as hard disks. Optical disks such as CD-ROMs (Compact Disc Read Only Memory) can also be used as recording media. Non-volatile semiconductor storage devices can also be used as recording media.

[0078] Some or all of the above-described embodiments can be described as, but are not limited to, the following supplementary notes.

[0079] [Supplementary Note 1] A selection support device comprising: an acquisition means for acquiring patient conditions, which are medical conditions required of clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients; an extraction means for extracting, from among patients receiving treatment, patients whose information on the treatment of said patients matches said patient conditions, as selection candidates; an estimation means for inferring candidate clinical trial patients from among said selection candidates based on said selection conditions; and an output means for outputting information on said inferred candidate clinical trial patients.

[0080] [Supplementary Note 2] The selection support device described in Supplementary Note 1, wherein the acquisition means acquires the selection conditions from each entity conducting the clinical trial, and the estimation means estimates candidate clinical trial patients suitable for the clinical trial for each entity conducting the clinical trial based on the selection conditions.

[0081] [Supplementary Note 3] The selection support device according to Supplementary Note 1 or 2, wherein the estimation means estimates the candidate clinical trial patients based on the priority of each condition included in the selection conditions.

[0082] [Supplementary Note 4] The selection support device according to any one of Supplementary Notes 1 to 3, wherein the estimation means estimates the candidate clinical trial patients based on the amount of remuneration paid for conducting the clinical trial.

[0083] [Supplementary Note 5] The selection support device described in any one of Supplementary Notes 1 to 4, wherein the estimation means estimates the candidate clinical trial patient based on at least one of the region, hospital, ward, and floor where the candidate patient is hospitalized.

[0084] [Supplementary Note 6] The selection support device according to any one of Supplementary Notes 1 to 5, wherein the extraction means extracts information about the selected candidates in a pseudonymized or anonymized state.

[0085] [Supplementary Note 7] The selection support device according to any one of Supplementary Notes 1 to 6, wherein the estimation means estimates candidates for clinical trial patients based on the patient conditions and the selection conditions.

[0086] [Supplementary Note 8] The selection support device according to any one of Supplementary Notes 1 to 7, wherein the extraction means extracts information about the selected candidate, including whether or not the candidate has agreed to participate in the clinical trial.

[0087] [Supplementary Note 9] The selection support device according to any one of Supplementary Notes 1 to 8, wherein the estimation means estimates the candidate clinical trial patients from the selected candidates using an estimation model for estimating the candidate clinical trial patients.

[0088] [Supplementary Note 10] The selection support device according to Supplementary Note 9, further comprising a generation unit that generates the estimation model by machine learning the relationship between the selection conditions and whether or not a patient is selected as a clinical trial patient.

[0089] [Supplementary Note 11] A selection support method that takes patient conditions, which are medical conditions required of clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients; extracts, from patients receiving treatment, patients whose information about the patient's treatment matches the patient conditions, as selection candidates; predicts, from the selection candidates, candidates for clinical trial patients based on the selection conditions; and outputs information about the predicted candidates for clinical trial patients.

[0090] [Supplementary Note 12] A recording medium that non-temporarily stores a selection support program that causes a computer to execute the following processes: a process of acquiring patient conditions, which are medical conditions required of clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients; a process of extracting, from among patients receiving treatment, patients whose information on the treatment of said patients matches said patient conditions, as selection candidates; a process of selecting, from said selection candidates, candidate clinical trial patients based on said selection conditions; and a process of outputting information on said selected candidate clinical trial patients.

[0091] Furthermore, some or all of the configurations described in Supplementary Notes 2 to 10, which are dependent on Supplementary Note 1, may also be dependent on Supplementary Notes 11 and 12 in the same dependent relationship as Supplementary Notes 2 to 10. Furthermore, not limited to Supplementary Notes 1, 11, and 12, some or all of the configurations described as Supplements may be made dependent on various hardware, software, various recording means for recording software, or systems, within the scope of each of the above-mentioned embodiments.

[0092] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0093] REFERENCE SIGNS LIST 10 Selection support device 11 Acquisition unit 12 Extraction unit 13 Estimation unit 14 Generation unit 15 Output unit 16 Storage unit 20 Data management device 30 Terminal device 40 Medical terminal device 100 Computer 101 CPU 102 Memory 103 Storage device 104 Input / output I / F 105 Communication I / F

Claims

1. A selection support device comprising: an acquisition means for acquiring patient conditions, which are medical conditions required of clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients; an extraction means for extracting, from among patients receiving treatment, patients whose information on the treatment of said patients matches said patient conditions as selection candidates; an estimation means for inferring candidate clinical trial patients from among said selection candidates based on said selection conditions; and an output means for outputting information on said inferred candidate clinical trial patients.

2. The selection support device according to claim 1, wherein the acquisition means acquires the selection conditions from each entity conducting the clinical trial, and the estimation means estimates candidate clinical trial patients suitable for the clinical trial for each entity conducting the clinical trial based on the selection conditions.

3. The selection support device according to claim 1 or 2, wherein the estimation means estimates the candidate clinical trial patients based on the priority of each condition included in the selection conditions.

4. A selection support device according to any one of claims 1 to 3, wherein the estimation means estimates the candidate clinical trial patients based on the amount of remuneration paid for conducting the clinical trial.

5. A selection support device according to any one of claims 1 to 4, wherein the estimation means estimates the candidate clinical trial patient based on at least one of the region, hospital, ward, and floor where the candidate patient is hospitalized.

6. The selection support device according to any one of claims 1 to 5, wherein the extraction means extracts information about the selected candidates in a pseudonymized or anonymized state.

7. The selection support device according to any one of claims 1 to 6, wherein the estimation means estimates candidates for clinical trial patients based on the patient conditions and the selection conditions.

8. The selection support device according to any one of claims 1 to 7, wherein the extraction means extracts information about the selected candidate, including whether or not the candidate has agreed to participate in the clinical trial.

9. The selection support device according to any one of claims 1 to 8, wherein the estimation means estimates the candidate clinical trial patients from the selected candidates using an estimation model for estimating the candidate clinical trial patients.

10. The selection support device according to claim 9, further comprising a generation means for generating the estimation model by machine learning the relationship between the selection conditions and whether or not a patient has been selected as a clinical trial patient.

11. A selection support method comprising: acquiring patient conditions, which are medical conditions required of clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients; extracting, from among patients receiving treatment, patients whose information on the patient's treatment matches the patient conditions as selection candidates; predicting, from among the selection candidates, potential clinical trial patients based on the selection conditions; and outputting information on the predicted potential clinical trial patients.

12. The selection support method according to claim 11, further comprising: obtaining the selection conditions from each entity conducting the clinical trial; and predicting the candidate clinical trial patients to be assigned to each entity conducting the clinical trial based on the selection conditions.

13. The selection support method according to claim 11 or 12, wherein the candidate clinical trial patients are estimated based on the priority of each of the conditions included in the selection conditions.

14. A selection support method according to any one of claims 11 to 13, wherein the candidate clinical trial patients are estimated based on the amount of remuneration paid for conducting the clinical trial.

15. A selection support method according to any one of claims 11 to 14, wherein the candidate clinical trial patients are estimated based on at least one of the region, hospital, ward, and floor where the candidate patients are hospitalized.

16. A selection support method according to any one of claims 11 to 15, wherein information about the selected candidates is extracted in a pseudonymized or anonymized state.

17. A selection support method according to any one of claims 11 to 16, wherein candidate clinical trial patients are estimated based on the patient conditions and the selection conditions.

18. A selection support method according to any one of claims 11 to 17, wherein information about the selected candidates, including whether or not they have agreed to participate in a clinical trial, is extracted.

19. A selection support method according to any one of claims 11 to 18, wherein the candidate clinical trial patients are estimated using an estimation model that selects candidate clinical trial patients from the selected candidates.

20. A recording medium that non-temporarily stores a selection support program that causes a computer to execute the following processes: a process of acquiring patient conditions, which are medical conditions required of clinical trial patients, and selection conditions, which are conditions for selecting clinical trial patients; a process of extracting, from among patients receiving treatment, patients whose information regarding the patient's treatment meets the patient conditions, as selection candidates; a process of selecting, from among the selection candidates, candidate clinical trial patients based on the selection conditions; and a process of outputting information regarding the selected candidate clinical trial patients.

Images