Medical examination time prediction system and medical examination time prediction method

Abstract

Provided is a medical examination time prediction system for predicting a medical examination call time of a subject patient, the medical examination time prediction system comprising: a storage device that stores and accumulates, for each medical examination, progress of a plurality of processes from reception to call and results data relating to a specimen; a processing device that calculates a predicted execution time for each process of medical examination of the subject patient on the basis of the results data and predicts a medical examination call time; and a database in which the predicted medical examination call time calculated by the processing device is recorded and which can be accessed from a terminal carried by the subject patient, the processing device recalculating the predicted execution time for an uncompleted process on the basis of the results data after a partial process relating to medical examination of the subject patient is completed, and updating the predicted medical examination call time of the subject patient recorded in the database.

Description

Examination time prediction system and examination time prediction method

[0001] The present invention relates to a consultation time prediction system and a consultation time prediction method.

[0002] Systems are known that predict the waiting time until a patient is called in for sample collection (e.g., blood collection) after receiving a request for sample collection (Patent Documents 1 and 2).

[0003] International Publication No. 2022 / 185810 Patent No. 4156813

[0004] Even after sample collection, patients may have to wait in the hospital until they are called in for their examination, and for inexperienced patients, such as first-time patients, it can be unfamiliar to the hospital and they may have to wait without knowing how long it will take before their examination, which can be a psychological burden. Therefore, it is conceivable to apply Patent Document 1, for example, to predict the time until the examination at the time of reception.

[0005] However, the circumstances between reception and being called for sample collection are significantly different from those between reception and being called for a medical examination. For example, while simple clerical or mechanical tasks may be involved between reception and being called for blood collection, there are no time-consuming or highly complex manual processes that require human judgment. Therefore, the time from reception to being called for blood collection can be predicted in a consistent manner for each patient. In contrast, the time from reception to being called for a medical examination involves time-consuming processes such as sample analysis and pre-processing, such as centrifugation, as well as the highly complex manual process of determining whether a retest is required based on the sample analysis results. The number of such manual processes and whether a retest is required varies from patient to patient, making it difficult to predict the time of the medical examination for each patient at the reception stage in the same way as the time of blood collection. For patients who wait based on the indicated time, if the time passes without being called and further time passes without knowing the situation, this can lead to anxiety, such as wondering whether there is some mistake and they will not be called, or whether they are waiting in the wrong place.

[0006] An object of the present invention is to provide a consultation time prediction system and a consultation time prediction method that can indicate the time until a patient is called in for a consultation whenever necessary.

[0007] In order to achieve the above object, the present invention provides an examination time prediction system for predicting a call time for a target patient, comprising: a storage device that stores and accumulates, for each examination, the progress of multiple processes from reception to call, and actual data related to the specimen; a processing device that calculates a predicted implementation time for each process of the examination of the target patient based on the actual data, and predicts the call time for the examination; and a database in which the predicted call times for the examination calculated by the processing device are recorded and which is accessible from a terminal carried by the target patient, and in which, after some of the processes related to the examination of the target patient have been completed, the processing device recalculates the predicted implementation times for the incomplete processes based on the actual data, and updates the predicted call times for the target patient recorded in the database.

[0008] According to the present invention, the time until the patient is called in for medical examination can be displayed each time according to the patient's needs.

[0009] FIG. 1 is a flow chart showing an example of the processes within a hospital leading up to an examination. FIG. 1 is a schematic diagram of an examination time prediction system according to one embodiment of the present invention. FIG. 2 is a conceptual diagram of a process for predicting an examination time in an examination time prediction system according to one embodiment of the present invention. FIG. 3 is a conceptual diagram of a calculation of a predicted examination time performed at the time of blood collection reception in an examination time prediction system according to one embodiment of the present invention. FIG. 4 is a conceptual diagram of a calculation of a predicted examination time performed at the time of completion of blood collection in an examination time prediction system according to one embodiment of the present invention. FIG. 5 is a conceptual diagram of a calculation of a predicted examination time performed at the time of laboratory sample reception in an examination time prediction system according to one embodiment of the present invention. FIG. 6 is a conceptual diagram of a calculation of a predicted examination time performed at the time of testing completion in an examination time prediction system according to one embodiment of the present invention. FIG. 7 is a diagram showing an example of a reception slip that notifies a target patient of the predicted examination time at reception. FIG. 8 is a diagram showing an example of a screen of a mobile terminal that displays the predicted time.

[0010] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0011] The consultation time prediction system and consultation time prediction method of this embodiment predict the consultation call time for each patient who comes to the hospital and presents each patient who has completed check-in with an estimate of the waiting time until they are called for the consultation. The estimated waiting time presented to each patient is updated as the hospital processes leading up to the consultation progress, thereby improving accuracy. For ease of explanation, the following description focuses on one patient to explain the system and processes, and the subject patient is referred to as the "target patient" and other patients as "other patients." Furthermore, in this embodiment, blood is used as an example of the target patient's sample.

[0012] 1 is a flow chart showing an example of the process in a hospital leading up to a consultation. At a hospital, the process from when a patient checks in for blood collection (specimen collection) to when they are called in to be examined by a doctor broadly includes blood collection and specimen testing, and each of these blood collection and specimen testing involves multiple steps.

[0013] To explain an example using Figure 1, a patient who visits a hospital first checks in (hospital reception) and follows instructions to move to the blood collection room where blood will be collected (movement to blood collection room). After completing the blood collection application (blood collection application) in the blood collection room, the patient waits to be called and has their blood collected (blood collection performed), then moves to the front of the examination room (an example of a waiting area) (movement to front of examination room) and waits until they are called to be examined by a doctor. The patient then waits to be called and enters the examination room (movement inside examination room) and is examined by a doctor.

[0014] Meanwhile, in the blood collection room, a blood collection reception is held in response to the patient's request (blood collection reception).The patient is then called in to have their blood collected (blood collection call / blood collection completion).The blood collected from the patient is transported to the testing room along with the blood of other patients.

[0015] While the patient waits for a doctor's examination after blood collection, the patient's sample is received in the examination room (laboratory sample reception) and placed in a pre-treatment device (sample pre-treatment). This pre-treatment involves centrifugation, which may take time depending on the test items being measured. The pre-treated sample is then placed in an automated analyzer, which measures the test items being measured (sample measurement / test completion). The measurement results are verified as needed (result validation), registered in a database, and made available for viewing by the doctor on a terminal in the examination room (result viewing). If deemed necessary as a result of result validation or result viewing, or if an error occurs in the automated analyzer during sample measurement, the sample measurement may be repeated. In the examination room, the patient is called (examination call), and the doctor examines the patient based on the sample measurement results (examination completion).

[0016] The examination time prediction system of this embodiment predicts the time from when the target patient checks in for blood collection to when they are called in for an examination by a doctor, which in the example of Figure 1 is the approximate time it will take for them to move around the examination room, and updates the prediction each time according to the progress of the process so that the target patient can check the timing when they will be called in for an examination as appropriate.

[0017] -Consultation time prediction system- Figure 2 is a schematic diagram of an examination time prediction system according to one embodiment of the present invention, and Figure 3 is a conceptual diagram of the consultation time prediction process in the examination time prediction system according to one embodiment of the present invention. The examination time prediction system 1 shown in Figure 2 is configured to include, for example, a computer 100 installed on a local network of a hospital and a database 200 installed on the cloud. In a narrow sense, the consultation time prediction system of the present invention corresponds to the consultation time prediction system 1, but in a broad sense, it refers to a system including a host system X, described below, and the consultation time prediction system 1.

[0018] The examination time is predicted by the processing device 120 of the computer 200, and the predicted examination call time calculated by the computer 100 is recorded in the database 200. The database 200 is stored in a server (not shown) installed on the cloud, and unlike when installed on a local network of a hospital, it can be accessed from a mobile terminal M such as a smartphone carried by the patient.

[0019] 2 is merely an example, and the system may be configured such that database 200 is installed on the hospital's local network and patients are allowed to access database 200. It is also possible to provide the functions of computer 100 to a server installed outside the hospital's local network.

[0020] Next, we will explain the functions of computer 100. Computer 100 is configured to include a storage device 110 that stores various databases necessary for predicting examination times, and a processing device 120 that executes processing to predict examination times based on the data stored in storage device 110.

[0021] The storage device 110 stores and accumulates the progress of multiple processes from reception (e.g., blood collection reception in FIG. 1 ) to calling (e.g., movement within the examination room in FIG. 1 ) and actual data related to samples for each examination. For example, the storage device 110 stores a test information DB 111, a timestamp information DB 112, and a learning model storage unit 113.

[0022] Specifically, patient sample information (FIG. 3) such as measurement test items, sample type, whether or not retesting is required, etc. is recorded as performance data for each patient sample reception number in the test information DB 111. This data is input via the external system communication processor 121 from a host system X that manages multiple systems in the hospital, or from a so-called blood collection operation support system Y that streamlines blood collection operations in blood collection rooms.

[0023] The timestamp information DB 112 records, as actual data for each patient sample reception number, test progress information ( FIG. 3 ) related to the current examination, such as the number of people waiting at the time of blood collection reception, the number of people waiting at the time of specimen reception in the laboratory, the number of people waiting at the time of completion of the examination, the time of blood collection reception, the time of blood collection call (time of blood collection), the time of blood collection completion (time of specimen collection completion), the time of specimen reception in the laboratory, the time of specimen pretreatment input, the time of specimen input into the analyzer, the time of specimen test completion when specimen test results are reported, and the time of examination call. The actual data related to the progress of the process, i.e., the test progress information, includes data input from the host system X and the blood collection operation support system Y, such as the blood collection call time, as well as the current number of people waiting (e.g., the number of people waiting for examination at the time of blood collection reception) calculated by the processing device 120 (prediction processing unit 123) based on the progress of the examination process for other patients. This number of people waiting serves as basic information when the prediction processing unit 123 calculates the predicted examination call time for the target patient.

[0024] The learning model storage unit 113 stores a learning model 130 (FIG. 3) for predicting examination times, which is constructed by the processing device 120.

[0025] Processing device 120 The processing device 120 includes an external system communication processing unit 121, a learning model construction unit 122, a prediction processing unit 123, and a screen display processing unit 124. The functions of these processing units can be realized as hardware elements such as circuits that execute each process, or software elements such as programs.

[0026] The external system communication processing unit 121 receives patient specimen information and test progress information from the host system X and the blood collection service support system Y, and stores the patient specimen information in the test information DB 111 and the test progress information in the timestamp information DB 112, respectively.

[0027] The learning model construction unit 122 constructs a learning model 130 ( FIG. 3 ) by machine learning the patient sample information stored in the test information DB 111 and the test progress information stored in the timestamp information DB 112, and stores the constructed learning model 130 in the learning model storage unit 113. The learning model 130 is a prediction means that inputs data related to the target patient, such as measurement test items, and outputs the time of a doctor's call for consultation. The learning model construction unit 122 constructs and updates the learning model 130 periodically (e.g., once a day) based on performance data up to the previous day (patient sample information and test progress information). The patient sample information used to construct the learning model 130 includes the reception number, measurement test items, sample type, whether or not retesting is required, and whether or not centrifugation is required. The test progress information also includes the number of people waiting at the time of blood collection reception, the number of people waiting at the time of laboratory specimen reception, the number of people waiting at the time of completion of the test, the time of blood collection reception, the time of blood collection call, the time of blood collection completion, the time of laboratory specimen reception, the time of pre-processing input, the time of analysis device input, the time of specimen test completion, and the time of consultation call.

[0028] The prediction processing unit 123 calculates a predicted implementation time for each step of the examination for the target patient based on the performance data recorded in the storage device 110, and predicts the examination call time. For example, for the target patient, the prediction processing unit 123 calculates the number of people waiting based on the test progress information of other patients recorded in the timestamp information DB 112 at each time point, such as when blood collection is accepted, when laboratory samples are accepted, and when the test is completed, and stores this information in the timestamp information DB 112 as the test progress information of the target patient. Specifically, in this embodiment, the prediction processing unit 123 reads the learning model 130 that has learned the performance data from the learning model storage unit 113, uses the patient sample information and test progress information of the target patient stored in the test information DB 111 and the timestamp information DB 112 as input, and calculates a predicted implementation time for the uncompleted step as the output of the learning model 130. After calculating each predicted time, the prediction processing unit 123 stores the calculated predicted time in the timestamp information DB 112, transmits it to the higher-level system X via the external system communication processing unit 121, and records it in the database 200 on the cloud.

[0029] In this case, the calculation of the predicted consultation call time may include uncertain factors that significantly affect the calculation result. For example, the time required until the consultation call varies depending on whether the sample needs to be retested or whether centrifugation is required in the pretreatment process. Therefore, the prediction processing unit 123 calculates the predicted consultation call time for the target patient based on the actual data for each case, including whether the target patient's sample was centrifuged and whether or not a retest was performed. That is, when classifying by whether centrifugation is required and whether or not a retest is required, predicted times are calculated for four cases: when neither centrifugation nor retesting is required, when both centrifugation and retesting are required, when centrifugation is required but retesting is not required, and when centrifugation is not required but a retest is required. The predicted consultation call times registered in the database 200, i.e., the predicted consultation call times that the target patient can view, can be configured to allow the time for each case to be confirmed on the target patient's mobile device M, or the latest or earliest time can be displayed.

[0030] In this embodiment, the processing device 120 is configured to calculate the predicted consultation call time for the target patient using machine learning, but the method for calculating the consultation call time is not limited to machine learning. For example, an algorithm may be constructed in advance to calculate the predicted implementation time of each step, including the consultation call time, based on performance data, and the consultation call time may be calculated based on the performance data in accordance with this algorithm.

[0031] Here, after some of the steps related to the examination of the target patient are completed, the processing device 120 (prediction processing unit 123) recalculates the predicted implementation times (including the predicted consultation call times) for the incomplete steps based on the performance data, and updates the predicted consultation call times for the target patient recorded in the database 200. The predicted times for the incomplete steps calculated by the prediction processing unit 123 vary depending on the progress of the steps. Furthermore, when a step whose implementation time was predicted for the target patient is completed, the actual implementation time is sequentially recorded in the timestamp information DB 112 as performance data. For example, at the time of blood collection acceptance, in this embodiment, the prediction processing unit 123 calculates the predicted blood collection call time, the predicted specimen test completion time, and the predicted consultation call time. When blood is subsequently collected from the patient, the time when the target patient is actually called for blood collection is recorded in the timestamp information DB 112 and is used as basic information for subsequent construction of the learning model 130. When the process progresses further and blood collection is completed, the predicted time for the completed blood collection call is excluded from the calculation, and the predicted time for the incomplete process, such as the predicted time for completion of the sample test and the predicted time for the consultation call, is calculated.

[0032] In addition, actual data regarding the implementation time, including the blood collection call time, is sent to the host system X via the related system by manual operation or automatic processing, and is input from the host system X to the examination time prediction system 1 via the external system communication processing unit 121.

[0033] The screen display processing unit 124 outputs the data of the examination information DB 111 and the timestamp information DB 112 to an output device Z such as a monitor or a printer in response to the operation of a terminal in the examination room, for example. By outputting the data of the examination information DB 111 and the timestamp information DB 112 to the output device Z, it is possible to check the examination status for each day or each time period. The examination call time for the target patient calculated by the prediction processing unit 123 can be output to the output device Z and can also be printed on paper, for example, and provided to the target patient.

[0034] - Operation - When accepting blood collection When accepting blood collection, the processing device 120 calculates the blood collection call time, sample test completion time, and consultation call time from the measurement test items for the target patient, the sample type, the number of people waiting at the time of sample collection acceptance as the current number of people waiting, and the time of sample collection acceptance.

[0035] Specifically, when a patient accepts a blood collection order, the order number, test items to be measured, and sample type are input as patient sample information for the target patient from the host system X and the blood collection operation support system Y via the external system communication processor 121 into the examination time prediction system 1, and the time of blood collection is also input as test progress information for the target patient. The external system communication processor 121 stores the received patient sample information for the target patient in the test information DB 111, and stores the test progress information in the timestamp information DB 112. After saving the patient sample information and test progress information input in conjunction with the blood collection order for the target patient, the examination time prediction system 1 proceeds to processing by the prediction processor 123.

[0036] FIG. 4 is a conceptual diagram illustrating the calculation of the predicted consultation time performed during blood collection reception. The prediction processing unit 123 calculates the number of people waiting at the time of blood collection reception for the target patient based on the test progress information of other patients in the timestamp information DB 112, and stores the calculated number in the timestamp information DB 112. As shown in FIG. 4, the prediction processing unit 123 also reads the reception number, test item, specimen type, number of people waiting at the time of blood collection reception, and blood collection reception time for the target patient from the test information DB 111 and the timestamp information DB 112, inputs the information into the learning model 130 stored in the learning model storage unit 113, and calculates a predetermined predicted time. The predicted times calculated in this example are the predicted blood collection call time, the predicted specimen test completion time, and the predicted consultation call time. The prediction processing unit 123 saves these calculated predicted times in the timestamp information DB 112, transmits them to the host system X via the external system communication processing unit 121, and records the target patient's examination in the cloud database 200.

[0037] At the time of blood collection reception, the host system X outputs and provides to the target patient a reception slip 300 such as the one shown in Fig. 8. The information printed on the reception slip 300 includes the reception time, patient name, patient number, reception number, predicted blood collection call time, probability of being called around the predicted blood collection call time, predicted consultation call time, probability of being called around the predicted consultation time, and a two-dimensional code 400 of a URL for viewing the predicted time. The two-dimensional code 400 is a URL issued to access the database 200.

[0038] FIG. 9 shows an example of the display when the two-dimensional code 400 of the URL for viewing the predicted time is read and accessed on the mobile device M. The information displayed on the mobile device M includes the reception number, the predicted blood collection call time, the probability of being called around the predicted blood collection call time, the predicted consultation call time, the probability of being called around the predicted consultation time, the time from the current time until the next call, and the number of people who will be called before the target patient. Unlike the printed reception slip 300 of FIG. 8, the display in FIG. 9 is updated to the latest content depending on the progress of the process at the time the target patient accesses. Note that while this embodiment illustrates an example in which the two-dimensional code 400 is printed on the paper reception slip 300, the URL can also be provided electromagnetically rather than via paper.

[0039] At the time of blood collection reception, there are many incomplete steps before the patient is called for a consultation, and it is unknown whether a re-test will be required, so it is difficult to predict the time when the patient will be called for a consultation. However, it is important to inform the patient at the time of reception of an estimate of the time until the patient is called for a consultation.

[0040] - When blood collection is completed When blood collection is completed (when sample collection is completed), the processing device 120 again calculates the sample test completion time and the consultation call time from the measurement test items, sample type, blood collection reception time, blood collection call time, and blood collection completion time for the target patient.

[0041] When blood collection for the target patient is completed, the examination time prediction system 1 receives a reception number as patient sample information for the target patient from the blood collection operation support system Y via the external system communication processing unit 121, and also receives new inputs of the blood collection call time and blood collection completion time as test progress information for the target patient. Based on the reception number, the external system communication processing unit 121 stores the newly received test progress information for the target patient, i.e., the blood collection call time and blood collection completion time, in the timestamp information DB 112. After storing the blood collection call time and blood collection completion time input upon completion of blood collection for the target patient, the examination time prediction system 1 proceeds to processing by the prediction processing unit 123.

[0042] FIG. 5 is a conceptual diagram illustrating the calculation of the predicted consultation time when blood collection is completed. As shown in FIG. 5 , the prediction processing unit 123 reads the patient's reception number, measurement test item, specimen type, number of people waiting at blood collection reception (the number of people waiting at blood collection completion may be recalculated), blood collection reception time, blood collection call time, and blood collection completion time from the test information DB 111 and the time stamp information DB 112, and inputs these information into the learning model 130 stored in the learning model storage unit 113 to calculate a predetermined predicted time. Since the blood collection call and blood collection completion that were incomplete or subject to time prediction at the time of blood collection reception are now completed and actual results, the blood collection call time and blood collection completion time are added as input items to the learning model 130 when blood collection is completed relative to the time of blood collection reception. The predicted times calculated at blood collection completion are the predicted sample test completion time and the predicted consultation call time, and the time prediction for the completed blood collection call is excluded. The prediction processing unit 123 stores these calculated predicted times in the timestamp information DB 112, transmits them to the higher-level system X via the external system communication processing unit 121, and records the examination of the target patient in the database 200 on the cloud.

[0043] This updates the predicted time of specimen test completion and the predicted time of consultation call in the database 200, and the target patient can check the latest predicted time updated upon completion of blood collection via the mobile terminal M.

[0044] -When the specimen is received in the laboratory- When the specimen is received in the laboratory, the processing device 120 again calculates the specimen test completion time and the consultation call time from the measurement test items for the target patient, the specimen type, the number of people waiting at the time of specimen reception in the laboratory as the current number of people waiting, the blood collection reception time, the blood collection call time, the blood collection completion time, and the specimen reception time in the laboratory.

[0045] When a sample for a target patient is received in a laboratory, a reception number is input as patient sample information for the target patient from the host system X via the external system communication processing unit 121 to the examination time prediction system 1, and a laboratory sample reception time is also newly input as examination progress information for the target patient. Based on the reception number, the external system communication processing unit 121 stores the newly received examination progress information for the target patient, i.e., the laboratory sample reception time, in the timestamp information DB 112. After storing the laboratory sample reception time, the examination time prediction system 1 proceeds to processing by the prediction processing unit 123.

[0046] FIG. 6 is a conceptual diagram illustrating the calculation of the predicted examination time when a specimen is received from a laboratory. The prediction processing unit 123 calculates the number of people waiting at the time of receiving the specimen from the laboratory based on the test progress information of other patients stored in the timestamp information DB 112, and stores the calculated number of people waiting in the timestamp information DB 112. As shown in FIG. 6, the prediction processing unit 123 also reads the reception number, measurement test item, specimen type, number of people waiting at the time of receiving the specimen from the test information DB 111 and the timestamp information DB 112, the reception time of the blood collection, the call time for blood collection, the time of completion of blood collection, and the time of receiving the specimen from the laboratory for the target patient, and inputs the information into the learning model 130 stored in the learning model storage unit 113 to calculate a predetermined predicted time. Since the laboratory specimen reception that was incomplete at the time of blood collection completion is now completed and considered actual, the laboratory specimen reception time is added as an input item to the learning model 130 when the specimen is received from the laboratory relative to the time of blood collection completion, and the number of people waiting at the time of receiving the specimen from the laboratory is used instead of the number of people waiting at the time of receiving the blood collection. The predicted times calculated when the specimen is received from the laboratory are the predicted specimen testing completion time and the predicted consultation call time. The prediction processing unit 123 stores these calculated predicted times in the timestamp information DB 112, transmits them to the host system X via the external system communication processing unit 121, and records the consultation of the target patient in the database 200 on the cloud.

[0047] This updates the predicted time of sample test completion and the predicted time of consultation call in the database 200, and the target patient can check the latest predicted time updated as the sample is received in the laboratory via the mobile terminal M. At the time of sample reception in the laboratory, there are fewer uncompleted processes compared to when blood collection is completed, and the number of people waiting, which is included in the basic information for prediction calculation, is updated to the current number of people waiting at the time of reception in the laboratory, so the predicted time becomes a more accurate value.

[0048] Furthermore, in large hospitals, collected samples are not transported one by one from the blood collection room to the testing room, but rather a certain number of samples are transported as a lot, and the time from completion of blood collection to transport varies depending on the timing of completion of blood collection. Therefore, it is very important to recalculate the number of people waiting and the predicted time of consultation call when the test room receives samples.

[0049] -When sample testing is completed- When sample testing is completed, the processing device 120 again calculates the time when the doctor will be called for an examination from the measurement test items for the target patient, the sample type, the number of people waiting at the time the test is completed as the current number of people waiting, the time when blood collection is accepted, the time when blood collection is called, the time when blood collection is completed, the time when the sample is accepted by the laboratory, the time when sample pre-processing is started (the time when sample pre-processing is started in Figure 1), the time when the sample is inserted into the automated analyzer (the time when sample measurement is started in Figure 1), and the time when sample testing is completed.

[0050] When the specimen testing is completed, the consultation time prediction system 1 receives a reception number as patient specimen information for the target patient from the host system X via the external system communication processor 121, and also receives new test progress information for the target patient, including the pre-processing input time, analyzer input time, and specimen testing completion time. Based on the reception number, the external system communication processor 121 stores the newly received test progress information for the target patient, i.e., the pre-processing input time, analyzer input time, and specimen testing completion time, in the timestamp information DB 112. After storing the test progress information input upon completion of the specimen testing for the target patient, the consultation time prediction system 1 proceeds to processing by the prediction processor 123.

[0051] 7 is a conceptual diagram of the calculation of the predicted examination time performed when testing is completed. The prediction processing unit 123 calculates the number of people waiting at the completion of testing for the target patient's sample from the test progress information of other patients in the timestamp information DB 112, and stores the calculated number in the timestamp information DB 112. Furthermore, as shown in FIG. 7, the prediction processing unit 123 reads the reception number, measurement test item, sample type, number of people waiting at the completion of testing, blood collection reception time, blood collection call time, blood collection completion time, laboratory sample reception time, pretreatment input time, analyzer input time, and sample testing completion time for the target patient from the test information DB 111 and the timestamp information DB 112, and inputs these information into the learning model 130 stored in the learning model storage unit 113 to calculate a predetermined predicted time. Since the testing and pre-processing of samples that were incomplete at the time of sample reception in the laboratory have now been completed and are considered actual results, the pre-processing input time, analyzer input time, and sample testing completion time are added as input items to the learning model 130 when the sample is completed compared to when the sample was received in the laboratory, and the number of people waiting at the time of testing completion is used instead of the number of people waiting at the time of sample reception in the laboratory. The predicted time calculated at the time of testing completion is the predicted consultation call time. The prediction processing unit 123 saves the calculated predicted time in the timestamp information DB 112 and sends it to the host system X via the external system communication processing unit 121, and also records the consultation of the target patient in the cloud database 200.

[0052] This updates the predicted consultation call time in the database 200, and the patient can check the latest predicted consultation call time updated as the specimen testing is completed via the mobile terminal M. At the time of completion of testing, there are even fewer uncompleted processes compared to when the specimen was received in the laboratory, and the number of people waiting, which is included in the basic information for prediction calculation, is updated to the current number of people waiting at the time of completion of testing, so the predicted consultation call time becomes a more accurate value.

[0053] In addition, retesting may be performed depending on the results of the specimen test, and time-consuming processes such as centrifugation may be added depending on the measurement items. Therefore, the specimen test completion time may deviate from the previous prediction. Therefore, it is important to recalculate the number of people waiting and the predicted consultation call time when the specimen test is completed in order to present a more accurate predicted consultation call time to the target patient.

[0054] The examination information DB 111 and the timestamp information DB 112 can be output to the output device Z at any time via the screen display processing unit 124, allowing doctors and other personnel to check the examination status for each day or each time period. The predicted waiting time for each patient can also be output to the output device Z and checked.

[0055] -Effects- (1) According to this embodiment, after some steps related to the examination of a target patient are completed, the predicted implementation times for the incomplete steps are recalculated based on actual data, and the predicted examination call time for the target patient recorded in the database 200 is gradually updated. This makes it possible to provide the patient with the latest information on the time until they are called for examination whenever necessary, depending on the progress of the step. Furthermore, because the implementation times of the steps, which change to actual results as the step progresses, are added to the basic information for the prediction calculation, the accuracy of the prediction of the examination call time improves as the step progresses.

[0056] For example, when blood collection is accepted, the basic information for time prediction does not include the time of blood collection completion, but when blood collection is completed, the actual blood collection completion time can be added to the basic information for calculating the consultation call time.When samples are accepted by the laboratory, the current time when the samples are accepted by the laboratory and the newly calculated number of people waiting can be added to the basic information.Furthermore, when sample testing is completed, the basic information can include the time of sample pretreatment performed during the sample testing, the time of input into the automated analyzer, the time the sample testing is completed, and the newly calculated number of people waiting at the current time.By using such sequential calculations, the calculation accuracy of the consultation call time can be improved according to the progress of the process.

[0057] (2) When calculating the predicted time of a patient's appointment call using machine learning as in this embodiment, the reliability of the time prediction by the learning model 130 can be expected to improve as actual data accumulates.

[0058] (3) When calculating the predicted consultation call time depending on whether or not the patient's sample has been centrifuged and whether or not retesting is required, the consultation call time can be presented in a flexible manner depending on the measurement test items and whether or not retesting is required.

[0059] (4) Even if access from the target patient's general mobile terminal M is restricted on the hospital's local network, by recording the consultation call time in the database 200 installed on the cloud as in this embodiment, the target patient can access the database 20 at any time from his or her own general mobile terminal M and check the latest predicted consultation call time.

[0060] (5) By outputting the reception slip 300 on which the URL for accessing the database 200 is printed as a two-dimensional code 400, the target patient can easily access the database 200 and check the latest predicted time of call for consultation.

[0061] -Modifications- The present invention is not limited to the above-described embodiments and may include various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those including all of the described configurations. For example, it is possible to replace some of the configurations with other configurations. It is also possible to delete some of the configurations of the embodiments or add other configurations.

[0062] For example, the examination time prediction system 1 can be incorporated into a computer provided in an automatic analyzer. The storage device 110 can also be configured with multiple storage devices, and the test information DB 111, the timestamp information DB 112, and the learning model storage unit 113 can be shared and stored among multiple storage devices. In this case, the storage devices only need to be connected via a network, and do not necessarily have to be provided in or connected to the same computer. The same applies to the processing device 120, and the functions of the learning model construction unit 122 and the prediction processing unit 123 can be shared among multiple processing devices.

[0063] Furthermore, the basic information for calculating the predicted consultation call time is not limited to the aspects described in the above embodiment and can be modified as appropriate. For example, when predicting the consultation call time upon completion of blood collection, the blood collection completion time is known as the latest progress time, so it is possible to reduce the calculation load by excluding earlier blood collection call times from the basic information. Similarly, when predicting the consultation call time upon receipt of specimens from the laboratory, the blood collection reception time, blood collection call time, and blood collection completion time may be excluded from the basic information. When predicting the consultation call time upon completion of specimen testing, the blood collection reception time, blood collection call time, blood collection completion time, laboratory specimen reception time, pretreatment start time, and analyzer loading time may be excluded from the basic information.

[0064] 1...examination time prediction system, 110...storage device, 111...examination information DB (actual data), 112...timestamp information DB (actual data), 113...learning model storage unit, 120...processing device, 121...external system communication processing unit, 122...learning model construction unit, 123...prediction processing unit, 124...screen display processing unit, 130...learning model, 200...database, 300...reception slip, 400...two-dimensional code (URL)

Claims

1. An examination time prediction system for predicting a call time for a target patient, comprising: a storage device which stores and accumulates, for each examination, the progress of a plurality of steps from reception to calling, and actual data related to the specimen; a processing device which calculates a predicted implementation time for each step of the examination of the target patient based on the actual data, and predicts the examination call time; and a database in which the predicted examination call times calculated by the processing device are recorded and which is accessible from a terminal carried by the target patient, wherein, after some of the steps related to the examination of the target patient are completed, the processing device recalculates the predicted implementation times for uncompleted steps based on the actual data, and updates the predicted examination call time for the target patient recorded in the database.

2. The examination time prediction system of claim 1, wherein the performance data on the progress of the process includes a current number of patients waiting calculated based on the progress of other patients' examination processes, and the processing device calculates a predicted time of the patient to be called for examination based on the number of patients waiting.

3. The examination time prediction system according to claim 1, wherein the performance data on the specimen includes measurement test items and specimen types.

4. The consultation time prediction system according to claim 1, wherein the processing device calculates the predicted consultation call time for the target patient using machine learning.

5. The examination time prediction system of claim 1, wherein the processing device calculates a predicted examination call time for the target patient based on the performance data for each case of whether or not the sample of the target patient has been centrifuged and for each case of whether or not retesting has been performed.

6. The consultation time prediction system according to claim 1, wherein the database is installed on a cloud.

7. The examination time prediction system of claim 4, wherein the process includes sample collection and sample testing, and the processing device calculates the blood collection call time, the sample testing completion time, and the examination call time from the time of acceptance of the sample collection, the measurement test item related to the target patient, the sample type, the number of people waiting at the time of acceptance of the sample collection as the current number of people waiting, and the time of acceptance of the sample collection.

8. The examination time prediction system of claim 7, wherein the processing device recalculates the specimen test completion time and the examination call time from the measurement test items, specimen type, and specimen collection completion time related to the target patient when the specimen collection is completed.

9. The examination time prediction system of claim 8, wherein the processing device recalculates the specimen test completion time and the examination call time from the specimen reception time at the laboratory when the specimen is received in the laboratory, the measurement test item related to the target patient, the specimen type, the number of people waiting at the time of specimen reception at the laboratory as the current number of people waiting, and the specimen reception time at the laboratory.

10. The consultation time prediction system of claim 9, wherein the processing device recalculates the consultation call time from the time when the specimen testing is completed in the examination room, the measurement test item related to the target patient, the specimen type, the number of people waiting at the time of completion of the test as the current number of people waiting, and the time when the specimen testing is completed.

11. The examination time prediction system according to claim 1, further comprising a URL for accessing said database, which is issued.

12. The examination time prediction system according to claim 11, further comprising a reception slip on which the URL is printed, said reception slip being outputted.

13. A method for predicting examination times, which comprises storing and accumulating, for each examination, the progress of a plurality of steps from reception to calling, and actual data relating to samples, calculating a predicted time of performance for each step of the examination of a target patient based on the actual data, and predicting a call time for the target patient, characterized in that, after some of the steps relating to the examination of the target patient are completed, the predicted times of performance for uncompleted steps are recalculated based on the actual data, and the predicted call time for the target patient recorded in a database accessible from a terminal carried by the target patient is updated.

Images