Selection assistance device, data management device, selection assistance method, data management method, and recording medium

Abstract

This selection assistance device comprises a keyword acquisition unit, an estimation unit, a generation unit, a condition output unit, and an information acquisition unit. The keyword acquisition unit acquires a condition keyword, which is a keyword representing a condition for patients to be considered eligible for a clinical trial. The estimation unit estimates, from the condition keyword, a related keyword, which is a keyword related to the condition keyword in extraction of patients eligible for the clinical trial. The generation unit generates an index including the condition keyword and the estimated related keyword as extraction conditions for extracting patients eligible for the clinical trial. The condition output unit outputs the index to a data management device for managing patient-related information. The information acquisition unit acquires, from the data management device, information related to patients eligible for the clinical trial extracted on the basis of the index. The selection assistance device can assist in decision making in selection of clinical trial patients by acquiring information related to patients eligible for the clinical trial on the basis of the index.

Description

Selection support device, data management device, selection support method, data management method, and recording medium

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

[0002] In clinical trials of pharmaceuticals, the person in charge of selecting candidate clinical trial patients extracts patients who meet the selection criteria for clinical trial patients, for example, by checking the descriptions in medical records. For example, the person in charge extracts patients who may be suitable as clinical trial patients. Then, the person in charge extracts patients who meet the selection criteria for clinical trial patients, for example, by checking the details of the descriptions in the medical records. In addition, a system that supports the extraction of candidate clinical trial patients may be used to extract candidate clinical trial patients.

[0003] The clinical trial eligible patient selection device of Patent Document 1 compares anonymized prescription data with the criteria of the clinical trial to extract eligible patients for the clinical trial, and then constructs a database using the extracted results of eligible patients for the clinical trial.

[0004] JP 2016-24663 A

[0005] With the technology described in Patent Document 1, it may be difficult to set conditions for extracting information about 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 easily extract information about 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 a keyword acquisition means for acquiring condition keywords, which are keywords that represent the conditions of patients who will be subject to clinical trials; an estimation means for estimating related keywords, which are keywords related to the condition keywords, from the condition keywords when extracting patients who will be subject to clinical trials; a generation means for generating an index that includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who will be subject to clinical trials; a condition output means for outputting the index to a data management device that manages information about patients; and an information acquisition means for acquiring information about the patients who will be subject to clinical trials extracted based on the index from the data management device.

[0008] A data management device according to one aspect of the present disclosure includes an index acquisition means for acquiring an index that includes condition keywords, which are keywords that represent the conditions of patients who are subject to clinical trials, as extraction conditions for extracting patients who are subject to clinical trials, an extraction means for extracting information about the patients who are subject to clinical trials based on the index, and an extracted information output means for outputting the extracted information about the patients who are subject to clinical trials to a selection support device that outputs the condition keywords.

[0009] A selection support method according to one aspect of the present disclosure acquires condition keywords, which are keywords that represent the conditions of patients who will be subject to clinical trials; estimates related keywords, which are keywords related to the condition keywords, from the condition keywords when extracting patients who will be subject to clinical trials; generates an index that includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who will be subject to clinical trials; outputs the index to a data management device that manages information about patients; and acquires information about the patients who will be subject to clinical trials extracted based on the index from the data management device.

[0010] A data management method according to one aspect of the present disclosure obtains an index that includes condition keywords, which are keywords that represent the conditions of patients who are the subject of a clinical trial, as extraction conditions for extracting patients who are the subject of a clinical trial, extracts information about the patients who are the subject of a clinical trial based on the index, and outputs the extracted information about the patients who are the subject of a clinical trial to a selection support device that outputs the condition keywords.

[0011] 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 condition keywords, which are keywords that represent the conditions of patients who will be subject to clinical trials; a process of estimating related keywords, which are keywords related to the condition keywords, from the condition keywords when extracting patients who will be subject to clinical trials; a process of generating an index that includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who will be subject to clinical trials; and a process of outputting the index to a data management device that manages information about patients, and acquiring, from the data management device, information about the patients who will be subject to clinical trials that has been extracted based on the index.

[0012] A recording medium according to one aspect of the present disclosure non-temporarily records a data management program that causes a computer to execute the following processes: obtain an index that includes condition keywords, which are keywords that represent the conditions of patients who are subject to clinical trials, as extraction conditions for extracting patients who are subject to clinical trials; extract information about the patients who are subject to clinical trials based on the index; and output the extracted information about the patients who are subject to clinical trials to a selection support device that outputs the condition keywords.

[0013] According to the present disclosure, information about patients suitable for clinical trials can be easily extracted.

[0014] FIG. 1 is a diagram illustrating an example of the configuration of a clinical trial support 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 patients who are clinical trial subjects according to the present disclosure. FIG. 4 is a diagram illustrating an example of the configuration of a data management device according to the present disclosure. FIG. 5 is a diagram illustrating an example of the operation flow of the selection support device according to the present disclosure. FIG. 6 is a diagram illustrating an example of the operation flow of the data management device according to the present disclosure. FIG. 7 is a diagram illustrating an example of the hardware configuration of a selection support device and a data management device according to the present disclosure.

[0015] 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 support system. The clinical trial support system includes a selection support device 10, a data management device 20, a terminal device 30, and a medical data storage 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 connected to the terminal device 30, for example, via a network. The data management device 20 is connected to the medical data storage device 40, for example, via a network. Furthermore, there may be multiple data management devices 20, multiple terminal devices 30, and multiple medical data storage devices 40. For example, the data management device 20 may be connected to another data management device 20 via a network. The number of data management devices 20, multiple terminal devices 30, and multiple medical data storage devices 40 may be set as appropriate.

[0016] A clinical trial support system is, for example, a system that extracts information about patients who are clinical trial subjects. For example, the clinical trial support system extracts information about patients who are clinical trial subjects from among patients for whom information about treatment is stored. The clinical trial support system saves the information about patients who are clinical trial subjects extracted based on condition keywords as a database. For example, condition keywords are keywords that indicate the characteristics of patients who will be clinical trial subjects. For example, condition keywords are names of diseases to which the drug being tested is applied. Details of condition keywords will be explained later. Furthermore, the clinical trial support system searches the database for information about patients who are clinical trial subjects based on, for example, a search query that indicates detailed conditions for patients who are clinical trial subjects. In this way, the clinical trial support system uses, for example, condition keys to widely extract information about patients who could be clinical trial subjects and saves it as a database. Then, for example, the clinical trial support system uses a search query to search the database for information about patients who are likely to be selected as actual clinical trial patients.

[0017] A clinical trial subject is, for example, a patient who is the subject of a clinical trial. A clinical trial is, for example, a clinical trial conducted to obtain legal approval for the manufacture and sale of a pharmaceutical or medical device. A clinical trial support system is used, for example, when a clinical trial is planned, to extract patients who can be clinical trial subjects. A patient who can be a clinical trial subject is, for example, a patient who will be the subject of a detailed investigation to determine whether or not the patient is suitable for the clinical trial. The determination of whether or not the patient is suitable for the clinical trial is made, for example, by comparing information about the patient's treatment with the selection criteria for clinical trial patients. Information about the treatment is, for example, a record of one or more medical procedures including diagnosis, testing, medication, rehabilitation, counseling, and follow-up observation. Information about the patient's treatment is recorded, for example, in a medical record.

[0018] The information about patients who are clinical trial subjects is, for example, information about the attributes of patients who are candidate patients for the clinical trial and information about treatment. The information about patients who are clinical trial subjects may include information indicating whether or not the patient has consented to the clinical trial. The patient attributes are, for example, patient-specific information that does not change due to treatment. The patient attributes are, for example, one or more of the patient's age, gender, medical history, family medical history, work history, height, weight, and race. The patient attributes are not limited to the above. The information about treatment is, for example, the condition and progression of the disease. The information about treatment may include, for example, a treatment history. The treatment history is, for example, a history of surgery and medication. The treatment history is not limited to the above. The information about patients who are clinical trial subjects is, for example, pseudonymized information. The pseudonymized information is information from which information that identifies an individual can be restored when compared with other information. The information about patients who are clinical trial subjects may be anonymously processed information. The 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 that has the authority to view data, the information about the patient who is the subject of the clinical trial may include the patient's real name.

[0019] Information about patients who are the subject of a clinical trial is used, for example, to estimate the number of patients to be included in the clinical trial when creating a clinical trial protocol. Information about patients who are the subject of a clinical trial is also used, for example, to negotiate the implementation of the clinical trial with a person in charge at the medical institution treating the patient. The person in charge at the medical institution is, for example, a doctor treating the patient. The uses of information about treatment are not limited to the above.

[0020] Information about the treatment is entered into the electronic medical record, for example, via a terminal device used by the person in charge of treating the patient. The person in charge of treating the patient may be, for example, a doctor. The person in charge of treating the patient may also be a nurse, pharmacist, laboratory technician, physical therapist, or counselor. The person in charge of treating the patient is not limited to the above. Information about the treatment may also be output from, for example, an examination device to the medical data storage device 40. For example, an X-ray image of the patient and patient identification information may be output from an X-ray device to the medical data storage device 40. The patient identification information is, for example, a patient number. The patient identification information is not limited to the above.

[0021] The data management device 20, for example, extracts clinical trial patients from data stored in the medical data storage device 40. The data management device 20 extracts information about the clinical trial patients based on, for example, condition keywords and related keywords included in an index. The related keywords are, for example, keywords that indicate the same symptoms as the condition keywords when determining whether a patient is a clinical trial subject. The term "similar" can include "similar." The index is information that indicates the target to be extracted when extracting information from information related to treatment. The data management device 20 outputs information about the clinical trial patients to the selection support device 10 at a set timing. The selection support device 10, for example, stores information about the clinical trial patients acquired from the data management device 20 as a database. The selection support device 10 also acquires a search query and searches for information about the clinical trial patients corresponding to the search query from the stored information about the clinical trial patients. The selection support device 10 then outputs information about the clinical trial patients corresponding to the search query. The search query will be described later.

[0022] In the example of the clinical trial support system shown in FIG. 1 , the data management device 20 at the core hospital is further connected to the data management device 20 at the general hospital. In the clinical trial support system shown in FIG. 1 , the selection support device 10 acquires, for example, information about clinical trial subject patients extracted by the data management device 20 at the general hospital via the data management device 20 at the core hospital. The core hospital is, for example, a medical institution that accepts patients from general hospitals. Patients from general hospitals are accepted, for example, when advanced medical technology and equipment are required for patient treatment. The core hospital is, for example, a university hospital, a specialized medical institution, or a regional base hospital. The general hospital is, for example, a medical institution that works in cooperation with the core hospital to treat patients with conditions that require advanced medical technology and equipment. The general hospital is a medical institution that examines patients using conventional medical technology, refers patients to the core hospital, and provides further treatment for patients who have returned from the core hospital. The data management device 20 at the general hospital may also acquire information about clinical trial subject patients from other general hospitals.

[0023] Each data management device 20 adds information about the clinical trial subject patients extracted by itself to information about the clinical trial subject patients acquired from a downstream data management device 20, and outputs the combined information to an upstream device. Here, the upstream device refers to, for example, one of the data management devices 20 that is closer to the selection support device 10. For example, a core hospital may be referred to as a first hospital, a general hospital directly affiliated with the core hospital as a second hospital, and a general hospital affiliated with the second hospital as a third hospital. In this case, the selection support device 10 acquires, for example, information about the clinical trial subject patients extracted by the data management device 20 of the second hospital via the data management device 20 of the first hospital. Furthermore, the selection support device 10 acquires, for example, information about the clinical trial subject patients extracted by the data management device 20 of the third hospital via the data management devices 20 of the second hospital and the first hospital. Extracting information about the clinical trial subject patients in this manner makes it possible to widely extract patients suitable for clinical trials from, for example, patients receiving treatment at hospitals in a certain region.

[0024] 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 includes a keyword acquisition unit 11, an estimation unit 12, a generation unit 13, a condition output unit 14, and an information acquisition unit 15. The selection support device 10 may further include, for example, a storage unit 16, a search query acquisition unit 17, a search unit 18, and a search result output unit 19.

[0025] The keyword acquisition unit 11 acquires condition keywords, which are keywords that represent conditions for patients who are to be the subject of a clinical trial. The condition keywords are, for example, keywords that represent conditions for extracting patients who are to be the subject of a clinical trial.

[0026] The condition keyword may include, for example, the disease name of a patient to be extracted as a clinical trial subject. For example, if it is desired to extract patients with an infection caused by the novel coronavirus, the condition keyword may be set to "novel coronavirus." The condition keyword may be a combination of multiple keywords. Furthermore, when the condition keyword is a combination of multiple keywords, it may also include a keyword indicating the symptoms of the clinical trial subject. For example, when it is desired to extract patients with a novel coronavirus infection and pneumonia symptoms as candidate clinical trial patients, the condition keyword may be a combination of "novel coronavirus," indicating the disease name, and "pneumonia," indicating the symptoms. Furthermore, when the condition keyword is a combination of multiple keywords, it may also include a keyword indicating the stage of progression of the disease. For example, when it is desired to extract patients with lung cancer as clinical trial subjects, the condition keyword may be a combination of "lung cancer," indicating the disease name, and "stage 3," indicating the stage of progression of the disease. Furthermore, information indicating that the condition keyword should not be converted into a related keyword may be added to the condition keyword. For example, when conversion into another keyword is unnecessary, such as in the case of a keyword indicating an age condition, information indicating that the condition keyword should not be converted into a related keyword is added to the condition keyword. The information included in the condition keyword is not limited to the above.

[0027] The estimation unit 12 estimates related keywords, which are keywords related to the condition keywords, from the condition keywords when extracting patients who are subjects of a clinical trial. The related keywords are, for example, keywords that extract patients who are subjects of the same clinical trial as the patients extracted by the condition keywords. For example, the related keywords are keywords that can extract patients who have the same medical condition as the patients extracted using the condition keywords. The medical condition is, for example, one or more pieces of information from the name of the disease, symptoms of the disease, and the stage of progression of the disease. Furthermore, "same" can also include "similar." Furthermore, when the condition keyword is a combination of multiple keywords, the estimation unit 12 estimates related keywords for each of the multiple keywords.

[0028] For example, the estimation unit 12 estimates different disease names with the same symptoms as related keywords from the disease name indicated by the condition keyword. For example, suppose that patient A is extracted when the condition keyword is "novel coronavirus." In this case, patient B, whose electronic medical record lists "COVID-19," is not extracted as a candidate clinical trial patient despite being infected with the same virus. In such a case, the estimation unit 12 estimates the related keyword "COVID-19" from the condition keyword "novel coronavirus." By estimating the related keyword from the condition keyword, it becomes possible to extract patients with the same condition using the condition keyword and the related keyword.

[0029] The estimation unit 12 may estimate, as the related keywords, keywords that indicate symptoms estimated from the disease name indicated by the condition keyword. For example, when the condition keyword is "COVID-19," the estimation unit 12 may estimate, as the related keywords, keywords such as "abnormal breathing noises" and "pneumonia symptoms on CT scan," which indicate symptoms. Furthermore, when information indicating that a condition keyword should not be converted into a related keyword is added to the condition keyword, the estimation unit 12 may exclude the condition keyword to which information indicating that the keyword should not be converted into a related keyword is added from the targets for estimating related keywords.

[0030] The estimation unit 12 estimates related keywords from the condition keywords using, for example, an estimation model. The estimation model is, for example, a machine learning model that uses the condition keywords as input and estimates related keywords. The estimation model estimates, as related keywords, terms that can extract patients with the same condition as the condition keywords from terms included in a dictionary. The dictionary includes, for example, terms in the medical field.

[0031] The estimation model converts each of the condition keywords and dictionary terms into embedding vectors. The estimation model, for example, calculates the Euclidean distance or cosine similarity between the embedding vectors converted from the condition keywords and the embedding vectors converted from the dictionary terms. The estimation model, for example, estimates, as related keywords, terms in the dictionary whose Euclidean distance or cosine similarity between the embedding vectors and the condition keywords satisfies a criterion. The criterion for Euclidean distance or cosine similarity is set so that patients with the same medical condition can be extracted.

[0032] The estimation model is generated, for example, by learning the relationship between condition keywords and related keywords. The estimation model is generated, for example, by learning the relationship between condition keywords included in the training data and terms included in a dictionary, and information indicating the suitability of the terms as related keywords. For example, Word2Vec can be used as the estimation model. A language model other than Word2Vec may be used as the estimation model. For example, GPT-2 (Generative Pre-trained Transformer-2), GPT-3, GPT-3.5, or GPT-4 may be used as the estimation model. Furthermore, 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 as the estimation model. The language model used for the estimation model is not limited to the above.

[0033] The estimation unit 12 may estimate related keywords corresponding to the condition keywords using a list associating the condition keywords with related keywords. The list associating the condition keywords with related keywords is generated, for example, by an operator with medical knowledge associating terms that can be extracted for patients with the same medical condition. The estimation unit 12 estimates related keywords corresponding to the condition keywords by, for example, referring to data associating the condition keywords with related keywords. When multiple related keywords are associated with one condition keyword, the estimation unit 12 estimates the multiple associated related keywords as related keywords corresponding to the condition keyword.

[0034] The generation unit 13 generates an index that includes the condition keyword and the estimated related keywords as extraction conditions for extracting patients who are subject to clinical trials. The index is, for example, an index related to the extraction of patients who are subject to clinical trials. The index includes, for example, the condition keyword and the related keywords as indexes related to the extraction of patients who are subject to clinical trials. In this case, the index is, for example, a condition that is satisfied by patients who are subject to clinical trials. For example, if the condition keyword is "novel coronavirus" and the related keyword is "COVID-19," the index includes "novel coronavirus" and "COVID-19." The index may include multiple condition keywords and the related keywords for each of the condition keywords.

[0035] The generating unit 13 may generate an index to which a priority is added when extracting information about clinical trial patients. That is, the index includes, for example, a priority when extracting information about clinical trial patients as an index related to the extraction of clinical trial patients. In this case, the priority is an index indicating the priority when extracting clinical trial patients. The priority is used, for example, to determine the interval at which the data management device 20 extracts information about clinical trial patients. The priority is set, for example, so that the higher the priority, the shorter the interval at which the data management device 20 extracts information about clinical trial patients. Furthermore, the priority may be an index indicating an index for which extraction processing is prioritized when the processing load on the data management device 20 is high, for example.

[0036] Priorities are set, for example, based on the rate at which the condition progresses. For example, if the disease being studied is a viral infection, the condition changes more quickly than a chronic disease such as hypertension. For this reason, in the case of a viral infection, it is desirable that the information about the patient being studied be extracted from information about the most recent treatment possible. In this case, the priority of the viral infection is set higher than the priority of hypertension.

[0037] The priority may be set based on the frequency of use of at least one of the condition keyword and the related keyword. For example, the priority of an index including a keyword whose frequency of use is rapidly increasing is set to be higher than that of other indexes. The priority may also be acquired from, for example, the terminal device 30 together with the condition keyword. In this case, the priority is set by, for example, the user of the terminal device 30.

[0038] The condition output unit 14 outputs an index to the data management device 20 that manages information related to the patient's treatment. The condition output unit 14 outputs the index to the index acquisition unit 21 of the data management device 20, for example, at the timing when the index is generated. The condition output unit 14 may output the index at a timing set as the timing for outputting the index. In this case, the timing for outputting the index is set, for example, using a time or a time interval. The timing for outputting the index is set, for example, so that the results of extraction based on the index are acquired in batch processing performed by the information acquisition unit 15. For example, if batch processing is performed at midnight and the extracted information is reflected in the batch processing if the index is output by 9:00 PM, the timing for outputting the index is set, for example, to 9:00 PM. The timing for outputting the index by the condition output unit 14 is not limited to the above. The timing for outputting the index by the condition output unit 14 can be set as appropriate.

[0039] The information acquisition unit 15 acquires information about the clinical trial subject patients extracted based on the index from the extracted information output unit 25 of the data management device 20. The information acquisition unit 15 acquires information about the clinical trial subject patients extracted based on the index from the extracted information output unit 25 of the data management device 20, for example, by batch processing. The batch processing is performed, for example, at a predetermined interval. The predetermined interval is set so as to enable obtaining candidate clinical trial patients based on new information while preventing excessive load on the system. The predetermined interval may be set based on the priority included in the index. For example, the higher the priority of the index, the shorter the predetermined interval is set. The predetermined interval may also be set based on the disease name or the stage of progression of the disease included in the index. For example, the faster the disease progresses, the shorter the predetermined interval is set. For example, the higher the stage of progression of the disease, the shorter the predetermined interval is set. The batch processing may be performed at a set date and time. The batch processing may also be performed irregularly. The timing of the batch processing may be set as appropriate.

[0040] The information about the patients who are the subject of the clinical trial may be the number of patients extracted for each index. The information about the patients who are the subject of the clinical trial may also include information indicating whether or not the patient has consented to the clinical trial.

[0041] The storage unit 16 stores information about patients who are subject to clinical trials acquired by the information acquisition unit 15. The storage unit 16 stores the information about patients who are subject to clinical trials, for example, as a database in which indexes are associated with information about patients who are subject to clinical trials. The storage unit 16 updates the database so that the latest data for each index is accessible. The storage unit 16 may also store a history of data before the update.

[0042] When the estimation unit 12 estimates related keywords using an estimation model, the storage unit 16 stores, for example, the estimation model. When the search unit 18 searches for information on clinical trial subjects corresponding to a search query using a search model, the storage unit 16 stores, for example, the search model. The estimation model and the search model may each be stored in a storage means other than the storage unit 16.

[0043] FIG. 3 shows an example of information about patients who are clinical trial subjects. The example in FIG. 3 is the extraction result for a patient with index "1." In the example in FIG. 3, the condition keyword for the index numbered "1" is "corona" and the related keyword is "COVID-19." In this case, the example in FIG. 3 is information about a patient whose treatment information included the related keyword "COVID-19." In addition, information about the patients who are clinical trial subjects in the example in FIG. 3 has been extracted, including information about "consent," "symptoms," and "treatment." In the example in FIG. 3, patients who have contracted COVID-19, have symptoms of pneumonia, are receiving oxygen inhalation treatment, and have consented to the clinical trial are extracted. In addition, the information about the patients who are clinical trial subjects in the example in FIG. 3 has been processed to hide the patient's name. Information that has been processed to hide the patient's name is pseudonymized information or anonymously processed information.

[0044] The search query acquisition unit 17 acquires, for example, a search query for searching information about patients who are subject to a clinical trial. The search query acquisition unit 17 acquires, for example, a search query for searching information about patients who are subject to a clinical trial from the terminal device 30 that outputs the condition keywords. The search query indicates, for example, conditions for searching information about patients who are subject to a clinical trial from stored data, which is necessary for selecting patients who are subject to a clinical trial. The search query may also include conditions for excluding patients from the search. The search query indicates, for example, conditions more detailed than the condition keywords. For example, a condition keyword, such as "lung cancer," is a condition for gathering broad information, while a search query is a condition for narrowing down patients to be selected as patients who are subject to a clinical trial, such as "lung cancer," "age 45 or older," and "no complications." The search query may also be a selection criterion for patients who are subject to a clinical trial. The selection criterion includes, for example, a patient inclusion criterion for a clinical trial and a patient exclusion criterion for a clinical trial. The selection criterion used as the search query may be, for example, either the patient inclusion criterion for a clinical trial or the patient exclusion criterion for a clinical trial. The search query may also be part of the selection criterion for patients who are subject to a clinical trial.

[0045] The search query acquisition unit 17 may accept a search query when the user of the selection support device 10 or the terminal device 30 has been authenticated. The authentication of the user of the selection support device 10 or the terminal device 30 is performed using, for example, an ID and a password. The authentication of the user of the selection support device 10 or the terminal device 30 is not limited to authentication using an ID and a password.

[0046] The search unit 18 searches, for example, for information indicated by a search query acquired by the search query acquisition unit 17 from information about candidate clinical trial patients stored in the storage unit 16. The search query acquisition unit 17 searches, for example, for patients whose information about patients who are the subject of a clinical trial matches the conditions indicated by the search query. Matching can include similarity.

[0047] The search unit 18 may use a search model to search for information about patients who are clinical trial subjects that correspond to the search query. The search model is, for example, a machine learning model that receives the search query as input and extracts information about patients who are clinical trial subjects that are the search target.

[0048] The search model is generated, for example, by learning the relationship between the search query and information about the patient who is the subject of the clinical trial. The search model is generated, for example, by converting the search query and information about the patient who is the subject of the clinical trial into feature vectors and learning the relationship between the feature vectors. The search model is generated, for example, by deep learning using a neural network. The learning algorithm for generating the search model is not limited to the above. The search model is generated, for example, in a system external to the selection support device 10.

[0049] The search unit 18 may use a language model to search for information about patients who are clinical trial subjects that corresponds to the search query. For example, GPT-2, GPT-3, GPT-3.5, or GPT-4 may be used as the language model. Furthermore, T5, BERT, RoBERTa, or ELECTRA may also be used as the search model. The language model used for the search model is not limited to the above.

[0050] In addition, when the user or terminal device 30 of the selection support device 10 that outputs the search query and the condition keyword is the same, the search unit 18 may search for information about patients who are clinical trial subjects corresponding to the index that includes the condition keyword.

[0051] The search result output unit 19 outputs, for example, information about the patient who is the subject of the clinical trial searched by the search unit 18. The search result output unit 19 outputs, for example, the information about the patient who is the subject of the clinical trial searched by the search unit 18 to the terminal device 30. When the information about the patient who is the subject of the clinical trial is anonymized information or pseudonymized information, the search result output unit 19 outputs, for example, the information about the patient who is the subject of the clinical trial that is anonymized information or pseudonymized information.

[0052] The search result output unit 19 may output, as the search result, the number of patients whose information about the patients being the subject of the clinical trial matches the search query. The search result output unit 19 may also output the number of patients whose information about the patients being the subject of the clinical trial matches the index. The number of matching patients may include the number of similar patients.

[0053] 4 is a diagram showing an example of a search result for information on patients who are clinical trial subjects corresponding to a search query. The example in FIG. 4 shows the search results when the search query is "disease name: coronavirus," "symptoms: pneumonia," and "consent to clinical trial." The search results in FIG. 4 show that 10 patients who have contracted coronavirus, have symptoms of pneumonia, and have consented to the clinical trial have been extracted.

[0054] A specific example of the configuration of the data management device 20 will be described. Fig. 5 is a diagram showing an example of the configuration of the data management device 20. The data management device 20 basically includes an index acquisition unit 21, an extraction unit 24, and an extracted information output unit 25. The data management device 20 may further include, for example, a related word estimation unit 22, an update unit 23, an extracted information acquisition unit 26, and an extracted data storage unit 27.

[0055] The index acquisition unit 21 acquires an index that includes condition keywords, which are keywords that represent the conditions of patients who are subjects of a clinical trial, as extraction conditions for extracting patients who are subjects of a clinical trial. The index acquisition unit 21 may acquire an index that includes condition keywords and related keywords as extraction conditions for extracting patients who are subjects of a clinical trial. The index acquisition unit 21 acquires the index from, for example, the condition output unit 14 of the selection support device 10. When multiple data management devices 20 are connected via a network, the index acquisition unit 21 may acquire indexes from the other data management devices 20.

[0056] The related word estimation unit 22 estimates related keywords, which are keywords related to condition keywords, for example, when extracting patients who are clinical trial subjects. The related word estimation unit 22 estimates related keywords from condition keywords, for example, using an estimation model. The related word estimation unit 22 estimates related keywords from condition keywords, for example, using an estimation model generated using training data different from the estimation model used by the estimation unit 12 of the selection support device 10. That is, the related word estimation unit 22 estimates related keywords from condition keywords using an estimation model generated using training data different from the estimation model used by the estimation unit 12 of the selection support device 10 and the same algorithm as the estimation model used by the estimation unit 12 of the selection support device 10. For example, the related word estimation unit 22 estimates related keywords using an estimation model generated based on training data that uses terms unique to or frequently used terms as keywords in hospitals from which the data management device 20 extracts data. The related word estimation unit 22 may also estimate related keywords using related keywords included in an index as input to the estimation model. In this way, it is possible to extract a wider range of information about patients who are clinical trial subjects by using an estimation model generated for each hospital from which data is to be extracted by the data management device 20. In addition, an estimation model may be generated for each department from which data is to be extracted by the data management device 20.

[0057] For example, the related word estimation unit 22 may estimate related keywords from condition keywords included in the index when related keywords are not estimated in the selection support device 10. In this case, the estimation model may be the same as the estimation model used by the estimation unit 12 of the selection support device 10.

[0058] The related word estimation unit 22 may estimate related keywords corresponding to the condition keywords using a list associating the condition keywords with the related keywords. Alternatively, the related word estimation unit 22 may estimate related keywords using related keywords included in an index acquired from the selection support device 10 as the condition keywords. The list associating the condition keywords with the related keywords may be generated, for example, by an operator with medical knowledge associating terms that can be extracted for patients with the same medical condition. The list associating the condition keywords with the related keywords may be a list using terms unique to or frequently used in a hospital from which the data management device 20 extracts data. The related word estimation unit 22 estimates related keywords corresponding to the condition keywords by, for example, referring to data associating the condition keywords with the related keywords. When multiple related keywords are associated with one condition keyword, the related word estimation unit 22 estimates the multiple associated related keywords as related keywords corresponding to the condition keyword.

[0059] The update unit 23 updates the index by, for example, adding related keywords to an index that includes a condition keyword. Furthermore, when the related word estimation unit 22 estimates related keywords using related keywords included in the index as condition keywords, the update unit 23 may update the index by adding the related keywords estimated by the related word estimation unit 22. For example, assume that the index includes "novel coronavirus" as a condition keyword and "COVID-19" as a related keyword. In this case, if the related word estimation unit 22 estimates the related keyword "novel pneumonia" from the condition keyword or the related keyword, the update unit 23 adds "novel pneumonia" to the index. That is, the update unit 23 generates an updated index that includes "novel coronavirus" as a condition keyword and "COVID-19" and "novel pneumonia" as related keywords.

[0060] The extraction unit 24 extracts information about the clinical trial patient based on the index. The extraction unit 24 extracts information about the clinical trial patient from information about treatment based on, for example, condition keywords included in the index. For example, the extraction unit 24 extracts information about the clinical trial patient from the electronic medical record based on condition keywords included in the index. If the index includes related keywords, the extraction unit 24 extracts information about the patient who is the subject of the clinical trial based on, for example, the condition keywords and the related keywords.

[0061] The extraction unit 24 searches for, for example, patients whose information on treatment matches the information indicated by the index. The match may include similarity. The extraction unit 24 may use an extraction model to search for information on treatment corresponding to the index. The extraction model is, for example, a machine learning model that uses condition keywords and related keywords included in the index as input and extracts information on treatment as information on patients who are clinical trial subjects.

[0062] The extraction model is generated, for example, by learning the relationship between keywords included in the index and information related to treatment. The extraction model is generated, for example, by converting the keywords included in the index and information related to treatment into feature vectors and learning the relationship between the feature vectors. The extraction model is generated, for example, by deep learning using a neural network. 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.

[0063] The extraction unit 24 may extract information about patients who are clinical trial subjects using a language model. For example, GPT-2, GPT-3, GPT-3.5, or GPT-4 may be used as the language model. Furthermore, T5, BERT, RoBERTa, or ELECTRA may be used as the estimation model. The language model used for the estimation model is not limited to the above. When the update unit 23 updates the index, the extraction unit 24 extracts information about the patient based on, for example, the updated index. That is, the extraction unit 24 extracts information about the patient based on the condition keywords and related keywords included in the updated index.

[0064] The extracted information output unit 25 outputs information about the extracted patients to the selection support device 10 that is the output source of the condition keywords. The extracted information output unit 25 outputs information about patients who are the subject of clinical trials, for example, by batch processing at a predetermined cycle. The predetermined cycle is set, for example, so as to enable obtaining candidate clinical trial patients based on new information while preventing excessive load on the system. The predetermined cycle may be set based on the priority included in the index. For example, the higher the priority of the index, the shorter the predetermined cycle is set.

[0065] The predetermined period may be set based on the name of the disease or the stage of the disease included in the index. The extracted information output unit 25 outputs information about the patient who is the subject of the clinical trial, for example, at a period based on the name of the disease or the stage of the disease included in the index. For example, the predetermined period is set to be shorter as the disease progresses more rapidly. Also, for example, the predetermined period is set to be shorter as the disease progresses more rapidly.

[0066] When the other data management device acquires information about the clinical trial subject patient extracted based on the index, the extracted information output unit 25 outputs, for example, to the selection support device 10, the information about the clinical trial subject patient acquired by the extracted information acquisition unit 26 from the other data management device and the information about the clinical trial subject patient extracted by the extraction unit 24. The other data management device is, for example, an information processing device having a configuration similar to that of the data management device 20 and connected to the data management device 20 via a network. In the example of FIG. 1, the other data management device corresponds to, for example, a data management device in a general hospital. That is, the other data management device is an information processing device that extracts information about the clinical trial subject patient from information about the treatment of a patient different from the data management device 20. For example, if the side closer to the selection support device 10 on the network is considered upstream and the side farther from the selection support device 10 is considered downstream, the other data management device is an information processing device downstream of the data management device 20. For example, the extracted information output unit 25 outputs the information about the clinical trial subject patient at the same cycle as the cycle at which the extracted information acquisition unit 26 acquires information about the clinical trial subject patient from the other data management device. In this case, the extracted information output unit 25 outputs the information about the patient who is the subject of the clinical trial, for example, after the timing at which the extracted information acquisition unit 26 acquires the information about the patient who is the subject of the clinical trial from the other data management device. Furthermore, the extracted information output unit 25 may output the information about the patient who is the subject of the clinical trial at a cycle shorter than the cycle at which the extracted information acquisition unit 26 acquires the information about the patient who is the subject of the clinical trial from the other data management device. Furthermore, the extracted information output unit 25 may output the information about the patient who is the subject of the clinical trial at a cycle longer than the cycle at which the extracted information acquisition unit 26 acquires the information about the patient who is the subject of the clinical trial from the other data management device.

[0067] The extracted information acquisition unit 26 acquires, for example, information about patients who are subject to clinical trials that has been extracted by another data management device based on an index. The extracted information acquisition unit 26 acquires the information about patients who are subject to clinical trials, for example, by batch processing at a predetermined cycle. The predetermined cycle is, for example, the same cycle as the cycle at which the extracted information output unit 25 outputs information about patients who are subject to clinical trials.

[0068] The predetermined period for acquiring information about the patients who are the subject of the clinical trial from the other data management devices may be different from the period during which the extracted information output unit 25 outputs information about the patients who are the subject of the clinical trial. For example, the period is set to be longer than the period during which the extracted information output unit 25 outputs information about the patients who are the subject of the clinical trial. In other words, the synchronization period is set so that synchronization with the other data management devices occurs less frequently than the period during which synchronization with the selection support device 10 occurs. This is because, for example, medical institutions using other data management devices have a small number of patients, and are unlikely to be frequently selected as patients who are the subject of the clinical trial from those medical institutions. Furthermore, medical institutions using other data management devices may, for example, have insufficient human resources and information processing system resources.

[0069] The extracted data storage unit 27 stores, for example, information about the patients who are the subject of the clinical trial extracted by the extraction unit 24. The extracted data storage unit 27 stores, for example, the information about the patients who are the subject of the clinical trial as a database in which indexes are associated with information about the patients who are the subject of the clinical trial. The extracted data storage unit 27 updates the database so that, for example, the information about the patients who are the subject of the clinical trial corresponding to each index becomes the latest information. The extracted data storage unit 27 may also store a history of data before the update.

[0070] When the extracted information acquiring unit 26 acquires information about patients who are subject to clinical trials from another data management device, the extracted data storage unit 27 stores, for example, the information about patients who are subject to clinical trials acquired by the extracted information acquiring unit 26. The extracted data storage unit 27 integrates and stores, for example, the information about patients who are subject to clinical trials from another data management device that the extracted information acquiring unit 26 acquires and the information about patients who are subject to clinical trials extracted by the extraction unit 24, for each index.

[0071] When the extraction unit 24 extracts information about patients who are clinical trial subjects using an extraction model, the extraction data storage unit 27 stores, for example, the extraction model. The extraction model may be stored in a storage means other than the extraction data storage unit 27.

[0072] The terminal device 30 is, for example, a terminal device used for processing to access the selection support device 10 and acquire information about patients who are subject to clinical trials. The terminal device 30, for example, outputs condition keywords to the keyword acquisition unit 11 of the selection support device 10. The terminal device 30 also outputs a search query to the search query acquisition unit 17 of the selection support device 10. The terminal device 30 then acquires information about patients who are subject to clinical trials from the search result output unit 19 of the selection support device 10. The terminal device 30 then outputs the information about patients who are subject to clinical trials to a display device (not shown).

[0073] The terminal device 30 is a terminal device used, for example, by a person in charge of a pharmaceutical company or a person in charge of an institution entrusted with clinical trial work by a pharmaceutical company. An institution entrusted with clinical trials by a pharmaceutical company is, for example, a CRO (Contract Research Organization). Also, a person in charge of an institution entrusted with clinical trials by a pharmaceutical company is, for example, a CRA (Clinical Research Associate). The CRA may be an employee of the pharmaceutical company.

[0074] The terminal device 30 may be used by personnel belonging to a medical institution and personnel of an institution entrusted with work by the medical institution. Personnel belonging to a medical institution include, for example, doctors, nurses, pharmacists, or clinical laboratory technicians. Personnel belonging to a medical institution are not limited to the above. Furthermore, an institution entrusted by a medical institution to handle clinical trials is, for example, an SMO (Site Management Organization). Furthermore, a person in charge of an institution entrusted by a hospital to handle clinical trials is, for example, a CRC (Clinical Research Coordinator).

[0075] The medical data storage device 40 stores, for example, information regarding patient treatment. The information regarding patient treatment is, for example, electronic medical record data entered by a doctor. The information in the electronic medical record may also be data entered by a nurse or a laboratory technician. The information regarding patient treatment may 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 medical data storage device 40 may store the information regarding patient treatment as anonymously processed information or pseudonymized information. The medical data storage device 40 outputs, for example, information regarding the treatment of the patient to be extracted to the extraction unit 24 of the data management device 20 as information regarding the patient who is the subject of the clinical trial.

[0076] The following describes the operation of the selection support device 10. Fig. 6 is a diagram showing an example of the flow of processing in the selection support device 10 to acquire information about patients who are the subject of a clinical trial based on condition keywords.

[0077] The keyword acquiring unit 11 acquires condition keywords (step S101). The condition keywords are keywords that represent the conditions of patients who are the subject of a clinical trial. The keyword acquiring unit 11 acquires the condition keywords from, for example, the terminal device 30.

[0078] When the condition keywords are acquired, the estimation unit 12 estimates related keywords from the condition keywords (step S102). The related keywords are keywords related to the condition keywords when extracting patients who are the subject of a clinical trial.

[0079] When related keywords have been estimated for all the acquired condition keywords, i.e., when related keywords have been estimated for all the acquired condition keywords (Yes in step S103), the generation unit 13 generates an index that includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who are the subject of the clinical trial (step S104).

[0080] When the index is generated, the condition output unit 14 outputs the index to the index acquisition unit 21 of the data management device 20 that manages information about patients (step S105).

[0081] The index acquisition unit 21 of the data management device 20 acquires an index, for example, from the condition output unit 14 of the selection support device 10. The extraction unit 24 of the data management device 20 extracts information about the clinical trial patient, for example, based on the index. Then, the extracted information output unit 25 of the data management device 20 outputs the information about the clinical trial patient to the information acquisition unit 15 of the selection support device 10, for example.

[0082] The information acquisition unit 15 acquires information about the patients who are the subjects of the clinical trial, extracted based on the index, from the extracted information output unit 25 of the data management device 20 (step S106).

[0083] When the information about the patient who is the subject of the clinical trial is acquired, the information acquisition unit 15 stores the information about the patient who is the subject of the clinical trial extracted based on the index, for example, in the storage unit 16 (step S107).

[0084] The search query acquisition unit 17 acquires, for example, a search query for searching for information about patients who are candidates for clinical trials (step S108). The search query acquisition unit 17 acquires the search query for searching for information about patients who are candidates for clinical trials, for example, from the terminal device 30 that has output the conditional keywords. Once the search query is acquired, the search unit 18 searches, for example, for information indicated by the search query from information about candidate patients for clinical trials stored in the storage unit 16 (step S109).

[0085] When the information on the candidate clinical trial patients is searched, the search result output unit 19 outputs, for example, the search result of the information on the patients who are the subject of the clinical trial (step S110). The search result output unit 19 outputs, for example, the search result of the information on the patients who are the subject of the clinical trial to the terminal device 30.

[0086] In step S103, if there are any conditional keywords for which related keywords have not been estimated, that is, if there are any conditional keywords for which related keywords have not been estimated among the acquired conditional keywords (No in step S103), the process returns to step S102, and the estimation unit 12 estimates related keywords for the conditional keywords for which related keywords have not been estimated.

[0087] The following describes the operation of the data management device 20. Fig. 7 is a diagram showing an example of the processing flow for extracting information about patients who are the subject of a clinical trial based on an index in the data management device 20.

[0088] The index acquisition unit 21 acquires an index including condition keywords as extraction conditions for extracting patients who are the subject of a clinical trial (step S121). The condition keywords are keywords that represent the conditions for patients who are the subject of a clinical trial. The index acquisition unit 21 acquires the index from, for example, the condition output unit 14 of the selection support device 10.

[0089] Once the index is acquired, the extraction unit 24 extracts information about the patient who is the subject of the clinical trial based on the index (step S122).

[0090] After extracting the information about the patient who is the subject of the clinical trial, the extraction unit 24 stores the extracted information about the patient who is the subject of the clinical trial in the extracted data storage unit 27, for example, in association with an index.

[0091] If information about the clinical trial patient has not been acquired from another data management device (No in step S123), the extracted information output unit 25 outputs the extracted information about the clinical trial patient (step S124). The extracted information output unit 25 outputs the extracted information about the clinical trial patient to, for example, the information acquisition unit 15 of the selection support device 10 that is the output source of the condition keyword.

[0092] Furthermore, in step S23, if information about the clinical trial patient has been acquired from another data management device (Yes in step S123), the extracted information output unit 25 outputs, for example, the extracted information about the clinical trial patient and the information about the clinical trial patient acquired from the other data management device (step S125). The extracted information output unit 25 outputs, for example, the extracted information about the clinical trial patient and the information about the clinical trial patient acquired from the other data management device to the information acquisition unit 15 of the selection support device 10 that is the output source of the condition keyword.

[0093] The selection support device 10 estimates related keywords from the acquired condition keywords. The selection support device 10 also generates an index including the condition keywords and the estimated related keywords as extraction conditions for extracting patients who are candidates for clinical trials. The selection support device 10 outputs the index to the data management device 20. The selection support device 10 then acquires information about the patients who are candidates for clinical trials extracted using the index. By using an index including the condition keywords and the related keywords estimated based on the condition keywords, it is possible to extract potential patients who are candidates for clinical trials from among patients whose patient-related information does not match the condition keywords. Therefore, the selection support device 10 can increase the amount of information extracted about patients who are candidates for clinical trials. Furthermore, by estimating related keywords from the condition keywords, users can obtain information about patients who are candidates for clinical trials without being aware of variations in terms included in the patient-related information. Therefore, the selection support device 10 can easily extract information about patients who are suitable for clinical trials. Furthermore, by extracting information about patients who are candidates for clinical trials, the selection support device 10 can support decision-making in selecting patients for clinical trials.

[0094] When the index includes information indicating the extraction priority, the selection support device 10 can acquire information about patients who are subject to clinical trials that have been extracted based on the extraction priority. For example, in a clinical trial of a new drug for a disease with a rapidly progressing condition, if patients who are subject to clinical trials are not selected based on new information, the patient's condition may have changed since the time of extraction, making the patient's condition unsuitable for the clinical trial. In such a case, by increasing the priority of disease names with a rapidly progressing condition and increasing the frequency of extracting information about patients who are subject to clinical trials, the selection support device 10 can acquire information suitable for selecting patients who are subject to clinical trials when patients who are subject to clinical trials must be selected based on new information.

[0095] Furthermore, the selection support device 10 can output the results of a search based on a search query from information about patients who are candidates for clinical trials extracted based on keyword conditions, thereby outputting information about patients that matches the search query from a wide range of patient information. For example, a person in charge of selecting patients for clinical trials can select patients for clinical trials by specifying detailed conditions from a wide range of patient information. Therefore, by using the selection support device 10, patients for clinical trials can be efficiently selected.

[0096] In addition, by storing information about patients who are clinical trial subjects extracted as pseudonymized information or anonymously processed information in the memory unit 16 and outputting it based on a search query, the selection support device 10 can output information about patients who are clinical trial subjects while, for example, protecting the patients' personal information.

[0097] The data management device 20 extracts information about patients who are the subject of a clinical trial based on the index. After extracting the information about the patients who are the subject of a clinical trial, the data management device 20 stores the extracted information about the patients who are the subject of a clinical trial in association with the index. The data management device 20 then outputs the information about the patients who are the subject of a clinical trial to the selection support device 10. By extracting and outputting the information about the patients who are the subject of a clinical trial based on the index in this way, the data management device 20 can easily extract information about patients who are suitable for the clinical trial.

[0098] Furthermore, by extracting and outputting information about patients who are the subject of a clinical trial using an index that includes related keywords, the data management device 20 can improve the possibility of extracting information about a sufficient number of patients who are the subject of a clinical trial. Furthermore, when synchronizing with the selection support device 10 based on the index, only necessary information can be synchronized, which can speed up processing and reduce the required computer resources.

[0099] Furthermore, when outputting information about patients who are clinical trial subjects extracted by other data management devices and information about patients who are clinical trial subjects extracted by its own device, the data management device 20 can, for example, output information about patients who are clinical trial subjects in a state in which information extracted by two or more devices is integrated to the selection support device 10. For example, when the amount of information extracted by other data management devices is small, the data management device 20 can output information about patients who are clinical trial subjects extracted by its own device and other data management devices to the selection support device 10 while suppressing the load.

[0100] Each process in the selection support device 10 may be distributed and executed among multiple information processing devices connected via a network. For example, the processes in the keyword acquisition unit 11, the estimation unit 12, the generation unit 13, the condition output unit 14, and the information acquisition unit 15 may be executed in different information processing devices, and the processes in the search query acquisition unit 17, the search unit 18, and the search result output unit 19 may be executed in different information processing devices. It may be set as appropriate which information processing device executes each process in the selection support device 10.

[0101] Furthermore, each process in the data management device 20 may be distributed and executed among a plurality of information processing devices connected via a network. For example, the processes in the index acquisition unit 21, extraction unit 24, extracted information output unit 25, and extracted information acquisition unit 26, and the processes in the related word estimation unit 22 and update unit 23 may be executed in different information processing devices. It can be set as appropriate which information processing device executes each process in the data management device 20.

[0102] Each process in the selection support device 10 and the data management device 20 can be realized by executing a computer program on a computer. Fig. 8 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 and the data management device 20. 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.

[0103] The CPU 101 reads and executes computer programs for 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 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 terminal device 30 and the medical data storage device 40 may also have a configuration similar to that of the computer 100.

[0104] 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.

[0105] Some or all of the above-described embodiments can be described as, but are not limited to, the following supplementary notes.

[0106] [Supplementary Note 1] A selection support device comprising: a keyword acquisition means for acquiring condition keywords which are keywords that represent the conditions of patients who are subjects of a clinical trial; an estimation means for estimating related keywords which are keywords related to the condition keywords from the condition keywords when extracting patients who are subjects of the clinical trial; a generation means for generating an index which includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who are subjects of the clinical trial; a condition output means for outputting the index to a data management device which manages information about patients; and an information acquisition means for acquiring information about the patients who are subjects of the clinical trial extracted based on the index from the data management device.

[0107] [Supplementary Note 2] The selection support device according to Supplementary Note 1, wherein the related keywords are keywords that extract patients who are subjects of the same clinical trial as the clinical trial in which the patient extracted by the conditional keywords is the subject.

[0108] [Supplementary Note 3] The selection support device according to Supplementary Note 1 or 2, wherein the related keyword is a keyword indicating at least one of a disease name similar to a disease name indicated by the condition keyword and a symptom that appears in the disease name indicated by the condition keyword.

[0109] [Supplementary Note 4] The selection support device according to any one of Supplementary Notes 1 to 3, wherein the acquisition means acquires information about the patients who are the subject of the clinical trial by batch processing at a predetermined interval.

[0110] [Supplementary Note 5] The selection support device according to Supplementary Note 4, wherein the information acquisition means acquires information about the patient who is the subject of the clinical trial at a period based on the name of the disease or the stage of progression of the disease included in the index.

[0111] [Supplementary Note 6] The selection support device according to any one of Supplementary Notes 1 to 5, wherein the information acquisition means acquires information about the patient who is the subject of the clinical trial in a pseudonymized form.

[0112] [Supplementary Note 7] The selection support device described in any one of Supplementary Notes 1 to 6 further comprises: a query acquisition means for acquiring a search query for searching information about the patient who is the subject of the clinical trial; a search means for searching information about the patient who is the subject of the clinical trial for information indicated by the search query; and a search result output means for outputting the information searched by the search means to a terminal device that has output the search query.

[0113] [Supplementary Note 8] A data management device comprising: an index acquisition means for acquiring an index including condition keywords, which are keywords that represent the conditions of patients who are subject to clinical trials, as extraction conditions for extracting patients who are subject to clinical trials; an extraction means for extracting information about patients who are subject to clinical trials based on the index; and an extracted information output means for outputting the extracted information about the patients who are subject to clinical trials to a selection support device that is an output source of the condition keywords.

[0114] [Supplementary Note 9] The data management device according to Supplementary Note 8, further comprising: a related word estimation means for estimating related keywords, which are keywords related to the condition keywords, from the condition keywords when extracting patients who are the subject of the clinical trial; and an update means for updating the index including the condition keywords by adding the related keywords to the index, wherein the extraction means extracts information about the patients based on the updated index.

[0115] [Supplementary Note 10] The data management device according to Supplementary Note 9, wherein the related keywords are keywords that extract patients who are subjects of the same clinical trial as the patient extracted by the conditional keywords.

[0116] [Supplementary Note 11] The data management device according to Supplementary Note 9 or 10, wherein the related keyword is a keyword indicating at least one of a disease name similar to a disease name indicated by the condition keyword and a symptom that appears in the disease name indicated by the condition keyword.

[0117] [Supplementary Note 12] The data management device according to any one of Supplementary Notes 9 to 11, wherein the extracted information output means outputs information about the patients who are the subject of the clinical trial by batch processing at a predetermined cycle.

[0118] [Supplementary Note 13] The data management device according to Supplementary Note 12, wherein the extracted information output means outputs information about the patient who is the subject of the clinical trial at a period based on the name of the disease or the stage of progression of the disease included in the index.

[0119] [Supplementary Note 14] The data management device according to any one of Supplementary Notes 9 to 13, wherein the extracted information output means outputs information about the patient who is the subject of the clinical trial in a pseudonymized form.

[0120] [Supplementary Note 15] The data management device described in any one of Supplementary Notes 9 to 14, further comprising an extracted information acquisition means for acquiring information about patients extracted based on an index, and the extracted information output means outputs the information about patients acquired by the extracted information acquisition means and the information about patients extracted by the extraction means.

[0121] [Supplementary Note 16] The data management device according to Supplementary Note 15, wherein the extracted information acquisition means acquires the information about the patient extracted based on the index at the same cycle as the cycle at which the extracted information output means outputs the information about the patient.

[0122] [Supplementary Note 17] The data management device according to Supplementary Note 15 or 16, wherein the extracted information output means outputs the information about the patient acquired by the extracted information acquisition means and the information about the patient extracted by the extraction means after the timing at which the extracted information acquisition means acquires the information about the patient extracted based on the index.

[0123] [Supplementary Note 18] The data management device according to Supplementary Note 15, wherein the extracted information acquisition means acquires information about the patient extracted based on the index at a period longer than a period at which the extracted information output means outputs information about the patient extracted based on the index.

[0124] [Supplementary Note 19] A selection support method comprising: acquiring condition keywords which are keywords that represent the conditions of patients who will be subject to a clinical trial; estimating, from the condition keywords, related keywords which are keywords related to the condition keywords when extracting patients who will be subject to the clinical trial; generating an index which includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who will be subject to the clinical trial; outputting the index to a data management device which manages information about patients; and acquiring, from the data management device, information about the patients who will be subject to the clinical trial extracted based on the index.

[0125] [Supplementary Note 20] A data management method comprising: acquiring an index including condition keywords, which are keywords that represent the conditions of patients who are subject to clinical trials, as extraction conditions for extracting patients who are subject to clinical trials; extracting information about the patients who are subject to clinical trials based on the index; and outputting the extracted information about the patients who are subject to clinical trials to a selection support device that outputs the condition keywords.

[0126] [Supplementary Note 21] A recording medium that non-temporarily records a selection support program that causes a computer to execute the following processes: a process of acquiring condition keywords, which are keywords that represent the conditions of patients who will be subject to clinical trials; a process of estimating related keywords, which are keywords related to the condition keywords, from the condition keywords when extracting patients who will be subject to clinical trials; a process of generating an index that includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who will be subject to clinical trials; a process of outputting the index to a data management device that manages information about patients; and a process of acquiring information about the patients who will be subject to clinical trials extracted based on the index from the data management device.

[0127] [Supplementary Note 22] A recording medium that non-temporarily records a data management program that causes a computer to execute the following processes: a process of acquiring an index that includes condition keywords, which are keywords that represent the conditions of patients who are subject to clinical trials, as extraction conditions for extracting patients who are subject to clinical trials; a process of extracting information about patients who are subject to clinical trials based on the index; and a process of outputting the extracted information about patients who are subject to clinical trials to a selection support device that outputs the condition keywords.

[0128] Furthermore, some or all of the configurations described in Supplements 2 to 7 that are dependent on Supplement 1 above may also be dependent on Supplements 19 and 21 in the same dependent relationship as Supplements 2 to 7. Furthermore, not limited to Supplement 1, Supplement 19, and Supplement 21, 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.

[0129] Furthermore, some or all of the configurations described in Supplementary Notes 9 to 18, which are dependent on Supplementary Note 8, may also be dependent on Supplementary Notes 20 and 22 in the same dependent relationship as Supplementary Notes 9 to 18. Furthermore, not limited to Supplementary Notes 8, 20, and 22, 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.

[0130] 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.

[0131] REFERENCE SIGNS LIST 10 Selection support device 11 Keyword acquisition unit 12 Estimation unit 13 Generation unit 14 Condition output unit 15 Information acquisition unit 16 Storage unit 17 Search query acquisition unit 18 Search unit 19 Search result output unit 20 Data management device 21 Index acquisition unit 22 Related word estimation unit 23 Update unit 24 Extraction unit 25 Extracted information output unit 26 Extracted information acquisition unit 27 Extracted data storage unit 30 Terminal device 40 Medical data storage 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: a keyword acquisition means for acquiring condition keywords which are keywords that represent the conditions of patients who are subjects of a clinical trial; an estimation means for estimating related keywords which are keywords related to the condition keywords when extracting patients who are subjects of the clinical trial from the condition keywords; a generation means for generating an index which includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who are subjects of the clinical trial; a condition output means for outputting the index to a data management device which manages information about patients; and an information acquisition means for acquiring information about the patients who are subjects of the clinical trial extracted based on the index from the data management device.

2. The selection support device according to claim 1, wherein the related keywords are keywords that extract patients who are subjects of the same clinical trial as the clinical trial in which the patients extracted by the condition keywords are subjects.

3. The selection support device according to claim 1 or 2, wherein the related keywords are keywords indicating at least one of a disease name similar to the disease name indicated by the condition keyword and a symptom that appears in the disease name indicated by the condition keyword.

4. The selection support device according to any one of claims 1 to 3, wherein the information acquisition means acquires information about the patients who are the subject of the clinical trial by batch processing at a predetermined interval.

5. The selection support device according to claim 4, wherein the information acquisition means acquires information about the patient who is the subject of the clinical trial at a period based on the name of the disease or the stage of progression of the disease included in the index.

6. The selection support device according to any one of claims 1 to 5, wherein the information acquisition means acquires information about the patient who is the subject of the clinical trial in a pseudonymized form.

7. A selection support device as described in any one of claims 1 to 6, further comprising: a query acquisition means for acquiring a search query for searching information about the patient who is the subject of the clinical trial; a search means for searching information about the patient who is the subject of the clinical trial for information indicated by the search query; and a search result output means for outputting the information searched by the search means to a terminal device that outputs the search query.

8. A data management device comprising: an index acquisition means for acquiring an index containing condition keywords, which are keywords that represent the conditions of patients who are subject to clinical trials, as extraction conditions for extracting patients who are subject to clinical trials; an extraction means for extracting information about patients who are subject to clinical trials based on the index; and an extracted information output means for outputting the extracted information about the patients who are subject to clinical trials to a selection support device that outputs the condition keywords.

9. The data management device according to claim 8, further comprising: a related word estimation means for estimating related keywords, which are keywords related to the condition keywords, from the condition keywords when extracting patients who are the subject of the clinical trial; and an update means for updating the index including the condition keywords by adding the related keywords to the index, wherein the extraction means extracts information about patients based on the updated index.

10. The data management device according to claim 9, wherein the related keywords are keywords that extract patients who are subjects of the same clinical trial as the patient extracted by the condition keywords.

11. The data management device according to claim 9 or 10, wherein the related keywords are keywords indicating at least one of a disease name similar to the disease name indicated by the condition keyword and a symptom that appears in the disease name indicated by the condition keyword.

12. A data management device according to any one of claims 9 to 11, wherein the extracted information output means outputs information about the patients who are the subject of the clinical trial by batch processing at a predetermined interval.

13. The data management device according to claim 12, wherein the extracted information output means outputs information about the patient who is the subject of the clinical trial at a period based on the name of the disease or the stage of progression of the disease included in the index.

14. A data management device as described in any one of claims 9 to 13, further comprising an extracted information acquisition means for acquiring information about the patients who are the subject of the clinical trial extracted based on the index, and wherein the extracted information output means outputs the information about the patients who are the subject of the clinical trial acquired by the extracted information acquisition means and the information about the patients who are the subject of the clinical trial extracted by the extraction means.

15. The data management device according to claim 14, wherein the extracted information output means outputs the information about the patient acquired by the extracted information acquisition means and the information about the patient extracted by the extraction means after the timing at which the extracted information acquisition means acquires the information about the patient extracted based on the index.

16. The data management device according to claim 14, wherein the extracted information acquisition means acquires information about patients extracted based on the index at a period longer than the period at which the extracted information output means outputs information about patients extracted based on the index.

17. A selection support method comprising: acquiring condition keywords that are keywords that represent the conditions of patients who will be subject to a clinical trial; estimating, from the condition keywords, related keywords that are keywords related to the condition keywords when extracting patients who will be subject to the clinical trial; generating an index that includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who will be subject to the clinical trial; outputting the index to a data management device that manages information about patients; and acquiring, from the data management device, information about the patients who will be subject to the clinical trial that has been extracted based on the index.

18. A data management method comprising: obtaining an index containing condition keywords, which are keywords that represent the conditions of patients who are subject to clinical trials, as extraction conditions for extracting patients who are subject to clinical trials; extracting information about the patients who are subject to clinical trials based on said index; and outputting the extracted information about the patients who are subject to clinical trials to a selection support device that outputs the condition keywords.

19. A recording medium that non-temporarily records a selection support program that causes a computer to execute the following processes: a process of acquiring condition keywords, which are keywords that represent the conditions of patients who will be subject to clinical trials; a process of estimating related keywords, which are keywords related to the condition keywords, from the condition keywords when extracting patients who will be subject to clinical trials; a process of generating an index that includes the condition keywords and the estimated related keywords as extraction conditions for extracting patients who will be subject to clinical trials; a process of outputting the index to a data management device that manages information about patients; and a process of acquiring information about the patients who will be subject to clinical trials extracted based on the index from the data management device.

20. A recording medium that non-temporarily records a data management program that causes a computer to execute the following processes: obtaining an index that includes condition keywords, which are keywords that represent the conditions of patients who are subject to clinical trials, as extraction conditions for extracting patients who are subject to clinical trials; extracting information about patients based on the index; and outputting the extracted information about the patients to a selection support device that outputs the condition keywords.

Images