Monitoring system for predicting risk of occurrence of adverse event, and method for predicting risk of occurrence of adverse event

Abstract

The present disclosure addresses the problem of providing a monitoring system for predicting the risk of occurrence of an adverse event in a patient being treated with a fluid therapeutic agent.　The monitoring system predicts the risk of occurrence of an adverse event in a patient being treated with a fluid therapeutic agent, and comprises: a pre-treatment patient information input unit for inputting individual pre-treatment patient information of the patient acquired in advance from the patient before treatment with the therapeutic agent; a determination criterion setting unit for setting a determination criterion for the risk of occurrence of the adverse event on the basis of the pre-treatment patient information; an intra-treatment biological information acquisition unit for acquiring intra-treatment biological information in the patient being treated with the therapeutic agent; an adverse event occurrence risk prediction unit for predicting the risk of occurrence of an adverse event in the patient being treated with the fluid therapeutic agent, on the basis of the determination criterion and the intra-treatment biological information; and a storage unit for storing the pre-treatment patient information, the intra-treatment biological information, and the determination criterion.

Description

Monitoring system for predicting the risk of occurrence of adverse events and method for predicting the risk of occurrence of adverse events 【0001】 The present disclosure relates to a monitoring system for predicting the risk of occurrence of adverse events in a patient being treated with a fluid therapeutic agent, a method for predicting the risk of occurrence of adverse events in a patient being treated with a fluid therapeutic agent, and a program for predicting the risk of occurrence of adverse events in a patient being treated with a fluid therapeutic agent. 【0002】 In hospital treatment and diagnosis, gases or liquids may be injected into a patient's body to treat a predetermined disease. For example, in the medical field, patients are treated with therapeutic agents for various diseases such as antibiotics, immunosuppressants, diuretics, and anticancer agents by intravenous drip or the like in the body. 【0003】 In addition, dialysis therapy can be cited as one of the methods of injecting a liquid into the body. Since dialysis is a treatment that performs extracorporeal circulation, adverse events due to dialysis have been reported both at home and abroad. Therefore, measures to prevent adverse events are also essential. In dialysis, the hospital staff in charge of treatment checks the patient's medical record in advance and then starts dialysis. When the patient's blood pressure reaches a predetermined value set in advance during dialysis, a warning lamp on the dialysis monitoring device is displayed. Then, the staff in charge of treatment checks the patient's condition and determines whether to interrupt dialysis or to observe the patient closely. The above-mentioned predetermined set value of blood pressure is not different for each patient, and it is generally set manually on the dialysis monitoring device uniformly for all patients. 【0004】 Here, even if the staff in charge of treatment rushes to the patient's bed during dialysis after seeing the display of the warning lamp on the dialysis monitoring device, it may be difficult to determine whether to interrupt dialysis or not. In addition, in order to avoid the risk of occurrence of adverse events, even in a situation where dialysis can originally continue, dialysis is sometimes interrupted just to be on the safe side. 【0005】Incidentally, a system has been disclosed (see Patent Document 1) that evaluates the risk of adverse events when injecting fluids into a patient. This system is programmed to acquire patient-related data, determine an initial risk prediction for the patient related to the fluid injection based on the patient data, provide the initial risk prediction to a user device before the fluid injection is performed on the patient, acquire sensor data related to the patient, and determine the current risk prediction for the patient related to the fluid injection. However, this technology relates to the risks associated with the administration of contrast agents and does not determine the risk prediction for patients undergoing treatment with therapeutic agents. 【0006】 Furthermore, the inventors have been conducting research on the medical applications of hydrogen gas and hydrogen-containing liquids. For example, they have disclosed a pharmaceutical composition containing hydrogen gas to improve the prognosis after the return of spontaneous circulation (see Patent Document 2), a pharmaceutical composition containing hydrogen gas to improve and / or stabilize circulatory dynamics after hemorrhagic shock (see Patent Document 3), and a hydrogen-containing organ preservation solution containing dissolved hydrogen at a predetermined concentration or higher, used to wash and / or preserve organs for transplantation (see Patent Document 4). However, these have not yet been put into practical use. 【0007】 Japanese Patent Publication No. 2024-123227, Japanese Patent Publication No. 2017-119653, International Publication No. 2018 / 021175, International Publication No. 2020-217575 【0008】 Kazuhisa Sugai et al., Scientific Reports volume 10, Article number: 20173 (2020) 【0009】 The objective of this disclosure is to provide a monitoring system that predicts the risk of adverse events occurring in patients being treated with fluid therapeutic agents on a patient-by-patient basis. 【0010】As a result of diligent research to solve the above problems, the inventors have found a system that sets criteria for determining the risk of adverse events based on pre-treatment patient information obtained in advance from patients before treatment with a fluid therapeutic agent, and predicts the risk of adverse events in patients being treated with a fluid therapeutic agent based on these criteria and in-treatment biological information of patients being treated with the fluid therapeutic agent. 【0011】In other words, the present disclosure is as follows: [1] A monitoring system for predicting the risk of adverse events occurring in a patient being treated with a fluid therapeutic agent, comprising: a pre-treatment patient information input unit for inputting individual pre-treatment patient information of a patient obtained in advance from the patient before treatment with the therapeutic agent; a judgment criterion setting unit for setting judgment criteria for the risk of adverse events occurring based on the pre-treatment patient information; a treatment biological information acquisition unit for acquiring treatment biological information of the patient being treated with the therapeutic agent; an adverse event risk prediction unit for predicting the risk of adverse events occurring in the patient being treated with the fluid therapeutic agent based on the judgment criteria and the treatment biological information; and a storage unit for storing the pre-treatment patient information, the treatment biological information and the judgment criteria. [2] The monitoring system according to [1], characterized in that the treatment biological information acquisition unit acquires treatment biological information of the patient being treated with the therapeutic agent in real time. [3] The monitoring system according to [1] or [2] above, characterized in that the judgment criterion setting unit obtains a corrected judgment value by correcting the basic judgment value based on the patient's individual pre-treatment patient information and uses that as the judgment criterion. [4] The monitoring system according to any one of [1] to [3] above, characterized in that the patient's individual pre-treatment patient information includes information regarding the occurrence of adverse events during treatment during similar treatments in the past. [5] The monitoring system according to any one of [1] to [4] above, comprising a display unit that displays the pre-treatment patient information and the biological information during treatment. [6] The monitoring system according to any one of [1] to [5] above, wherein the pre-treatment patient information is at least one selected from the group consisting of blood pressure, dry weight, body weight, weight gain rate between dialysis, weight gain between dialysis, heart rate, oxygen saturation, age, hematocrit value, body temperature, circulating blood volume, SOFA score, respiratory cycle parameters, electrocardiogram, cardiothoracic ratio, parameters obtained by echocardiography, indicators showing the degree of edema, medical information, medical history of a specified disease, family history, treatment history, medications currently being taken, smoking habits, blood glucose level, biochemical test results, hematological test results, patient information during treatment from previous similar treatments, and the concentration of the active ingredient of the therapeutic agent in the fluid in the body.[7] The monitoring system according to any one of [1] to [6] above, wherein the biological information during treatment is at least one selected from the group consisting of blood pressure, heart rate, oxygen saturation, hematocrit value, weight change, body temperature, respiratory cycle parameters, circulating blood volume, rate of change of circulating blood volume, peripheral venous pressure, central venous pressure, pulmonary artery pressure, pulmonary artery wedge pressure, mixed venous blood oxygen saturation, blood concentration of the active ingredient of the therapeutic agent in the fluid within the body, and water removal rate. [8] The monitoring system according to any one of [1] to [7] above, characterized in that the fluid is a gas containing hydrogen gas or a liquid in which hydrogen gas is dissolved. [9] The monitoring system according to any one of [1] to [8] above, further comprising a risk notification unit that notifies the patient to stop or adjust treatment with the therapeutic agent in the fluid when the risk prediction unit predicts a high risk of adverse event occurrence.

[10] The monitoring system according to any one of [1] to [9] above, further comprising a treatment control unit that controls treatment with the fluid therapeutic agent when the adverse event risk prediction unit predicts a high risk of adverse events.

[11] The monitoring system according to any one of [1] to

[10] above, wherein the patient is a dialysis patient and the pre-treatment patient information includes blood pressure, inter-dialysis weight gain rate, or dry weight.

[12] The monitoring system according to

[11] above, wherein the pre-treatment patient information includes treatment patient information from past dialysis sessions.

[13] The monitoring system according to any one of [1] to

[12] above, characterized in that the judgment criterion setting unit creates a normal model which is the probability distribution of the treatment biological information in a normal state in which no adverse events occur based on the pre-treatment patient information and uses it as the judgment criterion, and the adverse event risk prediction unit calculates the probability that the acquired treatment biological information arises from the normal model based on the normal model and the treatment biological information, and predicts the risk of adverse events based on that probability.

[14] A method for predicting the risk of adverse events in a patient being treated with a fluid therapeutic agent, which is performed by a monitoring system for predicting the risk of adverse events in a patient being treated with a fluid therapeutic agent, the monitoring system comprising: a pre-treatment patient information input unit for inputting pre-treatment patient information obtained in advance from the patient before treatment with the therapeutic agent; a judgment criterion setting unit for setting judgment criteria for the risk of adverse events based on the pre-treatment patient information; a treatment biological information acquisition unit for acquiring treatment biological information of the patient being treated with the therapeutic agent; an adverse event risk prediction unit for predicting the risk of adverse events in the patient being treated with the fluid therapeutic agent based on the judgment criteria and the treatment biological information; and a storage unit for storing the pre-treatment patient information, the treatment biological information and the judgment criteria, wherein the monitoring system comprises: a pre-treatment patient information input step for inputting pre-treatment patient information obtained in advance from the patient before treatment with the therapeutic agent; a judgment criterion setting step for setting judgment criteria for the risk of adverse events based on the pre-treatment patient information; and a treatment biological information acquisition step for acquiring treatment biological information of the patient being treated with the therapeutic agent. A method for predicting the risk of adverse events in a patient being treated with a fluid therapeutic agent, comprising: an adverse event risk prediction step that predicts the risk of adverse events occurring in the patient being treated with the fluid therapeutic agent based on the above judgment criteria and the biological information during treatment; and a storage step that stores the pre-treatment patient information, the biological information during treatment and the judgment criteria.

[15] A program capable of predicting the risk of adverse events occurring in a patient being treated with a fluid therapeutic agent, characterized in that a computer is made to function as a monitoring system as described in any of [1] to

[13] above. 【0012】 This disclosure makes it possible to assist in predicting the risk of adverse events in patients being treated with fluid therapeutic agents. 【0013】This is a functional block diagram of a monitoring system for predicting the risk of adverse events according to an embodiment of this disclosure. This figure shows a screen for inputting pre-treatment patient information obtained in advance from the patient before treatment with the therapeutic agent in the monitoring system for predicting the risk of adverse events according to an embodiment of this disclosure. This figure shows a screen displaying in-treatment biological information of the patient being treated with the acquired therapeutic agent in the monitoring system for predicting the risk of adverse events according to an embodiment of this disclosure. This is a flowchart showing the operation flow of the monitoring system for predicting the risk of adverse events according to an embodiment of this disclosure. 【0014】 Next, embodiments of this disclosure will be described with reference to the attached drawings to facilitate understanding of this disclosure. Note that parts of the drawings that are not relevant to the explanation may be omitted. Furthermore, all content described in the patent and non-patent documents referenced herein is incorporated herein by reference as a whole. 【0015】 [Configuration of a monitoring system for predicting the risk of adverse events] Figure 1 is a functional block diagram showing an example of an embodiment of the monitoring system 1 for predicting the risk of adverse events related to this disclosure (hereinafter referred to as "this monitoring system"). 【0016】 This monitoring system 1 is implemented by an information processing device such as a personal computer. This monitoring system 1 executes a program (hereinafter referred to as "this program") that can predict the risk of adverse events related to this disclosure occurring. This program, which operates on this monitoring system 1, works in cooperation with the hardware resources of this monitoring system 1 to implement a method (hereinafter referred to as "this method") for predicting the risk of adverse events related to this disclosure occurring. 【0017】 The hardware resources of this monitoring system 1 include, for example, processors such as a CPU (Central Processing Unit), an MPU (Micro Processing Unit), and a DSP (Digital Signal Processor). The processors execute the instructions written in this program to realize the various means of this monitoring system 1 described below. 【0018】Furthermore, when an information processing device (not shown) different from the monitoring system 1 executes this program, the information processing device functions in the same way as the monitoring system 1, thereby realizing this method. 【0019】 This monitoring system 1 comprises a pre-treatment patient information input unit 11, a judgment criterion setting unit 12, a treatment-in-treatment biological information acquisition unit 13, an adverse event risk prediction unit 14, a storage unit 20, and a display unit 15. 【0020】 [Fluid / Patient] Fluid refers to gas or liquid. Examples of gases include hydrogen, carbon dioxide, nitric oxide, oxygen, nitrogen, and other noble gases such as helium, argon, and xenon, or mixtures thereof. The gas is administered to the patient by inhalation through the nose or mouth. A gas inhalation mask may be used during inhalation. Possible concentrations of the gas contained in the gaseous therapeutic agent include 0.01-100%, 0.1-95%, 1-90%, 5-80%, 10-70%, and 20-60%. For example, in the case of hydrogen gas, possible concentrations include 0.01-100%, 0.05-50%, and 0.1-30%. In the case of oxygen gas, possible concentrations include 0.01-100%, 1-90%, 5-85%, and 10-80%. Furthermore, as a gaseous therapeutic agent, for example, a mixed gas consisting of 66.7% hydrogen and 33.3% oxygen gas can be cited. 【0021】 Examples of liquids include, in addition to dialysate, hydrogen, oxygen, carbon dioxide, nitric oxide, carbon monoxide, and hydrogen sulfide (H2). ２ S) is a liquid containing a dissolved gas such as helium, or a liquid containing various therapeutic agents for diseases such as catecholamines, antibiotics, immunosuppressants, diuretics, and anticancer drugs. The above liquid may also contain a solid that generates the above gas. Examples of liquids containing dissolved hydrogen gas include hydrogen-dissolved water and hydrogen-dissolved dialysis fluid. There are no particular restrictions on the amount of gas contained in the liquid, but for example, in the case of hydrogen, the concentrations are 0.1 to 1.6 mg / L, 0.2 to 1.2 mg / L, and 0.3 to 1.0 mg / L under conditions of 20°C and 1 atm. 【0022】The hydrogen-dissolved water described above can be produced, for example, by pressurizing and sealing gas into water, or by electrolysis of water. Specifically, hydrogen produced by electrolysis of purified tap water can be reverse-osmotically used to create a dialysis solution rich in hydrogen. 【0023】 Examples of patients include those with kidney disease, cardiovascular diseases such as myocardial infarction, post-cardiac arrest encephalopathy (including hypoxic-ischemic encephalopathy associated with neonatal asphyxia), and post-cardiac arrest syndrome, lifestyle-related diseases or metabolic syndrome such as hypertension, diabetes, arteriosclerosis, fatty liver, lipid metabolism disorders, and sleep disorders, ischemic-reperfusion injury such as cerebral infarction, neurodegenerative diseases such as dementia, Parkinson's disease, or depression, cancer, inflammation such as sepsis, rheumatoid arthritis, asthma, chronic obstructive pulmonary disease (COPD), and periodontal disease, skin diseases such as atopic dermatitis, and eye diseases such as age-related macular degeneration. Among the above patients, kidney disease patients receiving dialysis treatment are particularly suitable. 【0024】 Patients receiving treatment with fluid therapeutic agents are not limited to those receiving treatment involving the continuous administration of liquid or gaseous therapeutic agents into the body, but also include patients whose blood drawn out of the body is continuously treated with therapeutic agents such as dialysis fluid. "Continuous administration of liquid or gaseous therapeutic agents into the body" refers, for example, to continuous administration of liquid therapeutic agents via intravenous drip or continuous inhalation of gaseous therapeutic agents. 【0025】 [Pre-treatment patient information input unit] The pre-treatment patient information input unit 11 receives pre-treatment patient information obtained in advance from patients before treatment with therapeutic agents. Input can be done manually, for example, using a touch panel, keyboard, or mouse; wirelessly by transmitting data via Wi-Fi or Bluetooth®; or by connecting to the device that acquired the pre-treatment patient information via a wired connection and transmitting the data via a wired connection. The input data is stored in the storage unit 20 for each data entry. 【0026】The pre-treatment patient information input unit 11 may include one, at least two, at least three, at least four, at least five, or at least six items selected from the group consisting of blood pressure, dry weight, body weight, weight gain rate between dialysis, weight gain between dialysis, heart rate, oxygen saturation, age, hematocrit value, body temperature, circulating blood volume, SOFA score, respiratory cycle parameters, electrocardiogram, cardiothoracic ratio, parameters obtained from echocardiography, an index indicating the degree of edema, medical information, medical history of specified diseases, family history, treatment history, medications currently being taken, smoking habits, blood glucose level, biochemical test results, hematological test results, patient information during treatment from previous similar treatments, and the concentration of the active ingredient of the therapeutic agent in the fluid in the body. The concentration of the active ingredient of the therapeutic agent in the fluid in the body may include the concentration of the active ingredient in the blood, exhaled breath, or skin. Furthermore, blood pressure data includes not only diastolic blood pressure, systolic blood pressure, or mean blood pressure, but also information on blood pressure trends, such as whether the range of variation in diastolic, systolic, or mean blood pressure tends to be higher or lower than the average value for healthy individuals. The parameters related to the respiratory cycle mentioned above include respiratory rate, inspiratory time, expiratory time, inspiratory time to expiratory time (IE ratio), work of breathing, start of inspiration, end of inspiration, start of expiration, end of expiration, and breathing depth. These parameters related to the respiratory cycle can be acquired by this monitoring system 1, which is equipped with sensors such as pressure sensors, temperature sensors, position sensors, acoustic sensors, and cameras, using these sensors. 【0027】 Here, "past treatments of the same type" refers to treatments previously performed on the same patient as the treatment being performed now, using the same method. Specifically, in the case of a patient receiving dialysis, this refers to previous dialysis treatments the patient has received, i.e., the last dialysis treatment or any dialysis treatment prior to that. Such information on past treatments of the same type is particularly useful for understanding the patient's unique biological response trends in treatments that are performed continuously and repeatedly on the same patient, such as dialysis. 【0028】In cases where the patient is a dialysis patient, it is preferable that the pre-treatment patient information includes one, two, or three items selected from the group consisting of blood pressure, interdialysis weight gain rate, or dry weight. 【0029】 The above pre-treatment patient information can be obtained by measuring or calculating each piece of information in advance using a device capable of measuring that information, or by using existing calculation methods. For example, blood pressure can be measured with a blood pressure monitor, heart rate with a heart rate monitor, weight with a scale, oxygen saturation with a pulse oximeter, hematocrit with a hematocrit meter, and electrocardiogram with an electrocardiograph. Dry weight can be obtained, for example, by measuring the weight of a dialysis patient after excess fluid has been removed by dialysis. The weight gain rate between dialysis sessions and the amount of weight gain between dialysis sessions can be calculated using the weight after the previous dialysis session and the patient's weight before treatment. Biochemical test results include total protein levels (albumin and globulin), albumin level, albumin-to-globulin ratio, total bilirubin level, direct bilirubin level, AST level, ALT level, LDH level, alkaline phosphotase level, γ-GTP level, cholinesterase level, CPK level, amylase level, glucose level (blood glucose level), HbA1c level, sodium level, potassium level, chloride level, CRP level, triglyceride level, total cholesterol level, HDL-C level, urea nitrogen level, creatinine level, uric acid level, and iron level. Hematological test results include white blood cell count, red blood cell count, hemoglobin count, hematocrit count, platelet count, neutrophil count, lymphocyte count, monocyte count, eosinophil count, basophil count, prothrombin time, activated partial thromboplastin time, fibrinogen count, and D-dimer count. 【0030】Furthermore, some of the pre-treatment patient information described above is not limited to information about the patient being treated with the therapeutic agent, but may also include corresponding patient information obtained from other patients receiving similar treatment with the therapeutic agent. In addition, in cases where the same type of treatment is repeatedly performed, such as in dialysis, the pre-treatment patient information described above may also include patient information from previous similar treatments. Patient information from previous similar treatments includes not only the patient's biological information during treatment in previous similar treatments, but also information related to the occurrence of adverse events during treatment in previous similar treatments, such as whether or not adverse events occurred during treatment in previous treatments, and if so, the type of adverse event and the patient's biological information before and after the adverse event occurred. Specifically, as pre-treatment patient information when the same type of treatment is repeatedly performed, taking the case of a patient undergoing dialysis for the second time using hydrogen-dissolved dialysis as an example, the patient's blood pressure and hematocrit value during the first dialysis treatment, the presence or absence of seizures during dialysis treatment, and the blood pressure and hematocrit value before and after the seizure occurred can be cited. 【0031】 The above pre-treatment patient information refers to the patient's information before treatment with the fluid therapeutic agent. The term "before treatment with the fluid therapeutic agent" can be adjusted as appropriate depending on the type of patient information, but for example, it can be any time before 1 second before treatment with the therapeutic agent, such as 1 minute, 5 minutes, 30 minutes, 1 hour, 3 hours, 12 hours, 24 hours, 48 ​​hours, 72 hours, 1 week, 1 year ago, 5 years ago, 10 years ago, or the time elapsed since similar treatment was performed in the past. Furthermore, the above pre-treatment patient information may also be information linked to the time it was measured, such as weight 48 hours before drug administration. 【0032】When manually entering pre-treatment patient information, the display unit 15 displays the items of pre-treatment patient information that can be entered. All items may be displayed, items may be grouped into predetermined categories and the categories may be displayed first, or a search function may be used to allow the user to search for and display items of pre-treatment patient information. Alternatively, items of pre-treatment patient information related to adverse events that are expected to occur for each patient's disease may be defined in advance, and when the patient's disease is entered, the items of pre-treatment patient information corresponding to that disease may be extracted, displayed, and made available for input. In the case of dialysis patients, for example, as shown in Figure 2, pre-treatment patient information related to dialysis may be extracted and displayed on the display unit 15, and pre-treatment patient information may be made available for input. 【0033】 [Intra-treatment biological information acquisition unit] The intra-treatment biological information acquisition unit 13 acquires intra-treatment biological information of patients being treated with a therapeutic agent. Preferably, the intra-treatment biological information acquisition unit 13 acquires intra-treatment biological information of patients being treated with a therapeutic agent in real time, and the acquired data is stored in the storage unit 20. 【0034】 The biological information acquired by the biological information acquisition unit 13 during treatment includes one, at least two, at least three, at least four, at least five, or at least six selected from the group consisting of blood pressure, heart rate, oxygen saturation, hematocrit value, weight change, body temperature, respiratory cycle parameters, circulating blood volume, rate of change in circulating blood volume, peripheral venous pressure, central venous pressure, pulmonary artery pressure, pulmonary artery wedge pressure, mixed venous blood oxygen saturation, blood concentration of the active ingredient of the therapeutic agent in the fluid within the body, and ultrafiltration rate. 【0035】 In cases where the patient is undergoing dialysis, it is preferable to include blood pressure, respiratory cycle parameters, and / or hematocrit values ​​as the above-mentioned biological information during treatment. Furthermore, at least two of these biological information may be combined. 【0036】Blood pressure can be measured with a blood pressure monitor, and circulating blood flow, rate of change of circulating blood flow, and hematocrit can be measured by irradiating the blood circuit with a near-infrared laser. Peripheral venous pressure, central venous pressure, pulmonary artery pressure, pulmonary artery wedge pressure, mixed venous blood oxygen saturation, etc., can be obtained from sensors installed in circuits such as catheters and extracorporeal circulation circuits. Parameters related to the respiratory cycle may be obtained as respiratory waveforms, for example, by using pressure sensors or flow sensors connected to a transnasal cannula inserted into the nasal cavity. 【0037】 If the fluid is a gas containing a predetermined gas or a liquid in which a predetermined gas is dissolved, the biological information during treatment may be the concentration of the predetermined gas in the body, such as in the blood, breath, or skin, of a patient undergoing treatment with the drug. Methods for measuring the gas concentration in the body include, for example, measuring blood, breath, or skin samples taken from the patient by gas chromatography, measuring using electrode methods, polarography, etc., or measuring using a commercially available dissolved gas concentration meter. In the case of dialysis patients, the blood sample taken from the patient may be blood taken before and / or after passing through the dialyzer of the dialysis machine, and can be collected from blood tubes or ports. Alternatively, electrodes for measuring hydrogen may be placed inside the blood tube to continuously measure the blood hydrogen concentration of the patient during dialysis. Specifically, in the case of a patient undergoing treatment with hydrogen inhalation or hydrogen-dissolved dialysis fluid, it is preferable to include the blood hydrogen concentration as the biological information during treatment. 【0038】 The above-mentioned biological information during treatment is input to the biological information acquisition unit 13 via wired or network connection from each biological information measuring device connected to the patient undergoing treatment. 【0039】In the treatment-dependent biological information acquisition unit 13, the treatment-dependent biological information may be acquired in real time by a device provided in the monitoring system 1, or it may be acquired by receiving measured data in real time from a device that measures the treatment-dependent biological information. For example, if the patient is a dialysis patient, the blood pressure, heart rate, and hematocrit values ​​in the treatment-dependent biological information can be obtained by the monitoring system 1 receiving data in real time from the respective measuring devices mounted on the dialysis monitoring device. 【0040】 It is preferable to continuously measure and acquire the biological information of the patient undergoing treatment with the therapeutic agent. 【0041】 In this disclosure, "real-time" means the continuous and timely acquisition or reception of biometric information, and is not limited to cases where such acquisition or reception occurs immediately without any time delay. For example, even if information is acquired intermittently or periodically at predetermined time intervals, such as every few seconds to a few minutes, or if there are delays due to the communication environment or information processing, it is included in "real-time" in this disclosure as long as it is possible to grasp changes in the biometric information of the patient during treatment. Furthermore, even if the acquisition or reception of biometric information is temporarily interrupted during treatment due to procedures on the patient, equipment adjustments, or communication malfunctions, it is included in "real-time" acquisition in this disclosure as long as monitoring can be considered to have continued throughout the entire treatment period. 【0042】 The biological information during treatment may be displayed on the display unit 15. All items of the acquired biological information during treatment may be displayed, or items may be grouped into predetermined categories and the categories may be displayed first, or a search function may be used to search for and display items of patient information during treatment. In addition, the biological information during treatment required for each patient's disease may be predetermined, and the biological information during treatment corresponding to that disease may be extracted and displayed. In the case of dialysis patients, for example, the biological information during treatment is displayed on the display unit 15 as shown in Figure 3, and is displayed in real time. 【0043】[Criterion Setting Unit] Based on the pre-treatment patient information, the criterion setting unit 12 sets the criteria for determining the risk of occurrence of adverse events for each piece of in-treatment biological information. The set criteria are stored in the storage unit 20. 【0044】 The method for setting the criteria is as follows. First, based on guidelines regarding the disease to be treated, the instructions for the therapeutic agent, etc., a general criterion determination value for determining the risk of occurrence of adverse events in that patient when treated with the fluid therapeutic agent is calculated as the "basic determination value" for each piece of in-treatment biological information. Then, a corrected determination value obtained by correcting the basic determination value based on the pre-treatment patient information may be set as the criterion for determining the risk of occurrence of adverse events. 【0045】 The corrected determination value may be a value corrected to stricter conditions than the general criterion determination value, a value corrected to looser conditions than the general criterion determination value, or a value under the same conditions as the general criterion determination value, and can be appropriately set based on the pre-treatment patient information. When a mathematical model is used for setting the criteria, a linear model, etc. derived from statistical data in a past group of patients of the same type is used to estimate the initial value of the "latent state" indicating the stability of the circulatory dynamics, etc. with the pre-treatment patient information as an explanatory variable, and based on the estimated value, adjust the parameters of the prior probability distribution, or adjust the "intercept of the linear predictor (bias term)" and "weight coefficient" in the calculation formula for the probability of occurrence of adverse events. This enables determination that reflects the physiological reserve ability, the strength of the compensatory function, etc. of the individual patient, rather than a statistical general theory. 【0046】Furthermore, in other embodiments of this disclosure, the setting of the judgment criteria may be performed using a statistical probability model, specifically a Bayesian Anomaly Detection method, or a Kalman filter, particle filter, etc. In addition to conventional anomaly detection using thresholding or methods that score only the distance from normal data, by calculating the deviation from normal data as a "probability" and further combining it with a prior probability based on pre-treatment patient information, more accurate predictions become possible. In this case, the judgment criterion setting unit 12 defines the biological information acquired during treatment in a predetermined period at the start of treatment (for example, 15 to 30 minutes from the start of dialysis) as the "normal state" for that patient. Then, based on the distribution of data during this period, a probability model (predictive distribution) of normal biological information is constructed and set as the judgment criterion (normal model). That is, by inputting the pre-treatment patient information and the biological information at the start of treatment of the target patient into a general model (population model) constructed from a vast amount of past patient data, a "personalized normal model" specific to that patient is constructed. In this probability model, it is preferable to perform Bayesian estimation that introduces uncertainty into parameters such as the mean and variance, as well as the latent state itself. 【0047】The correction determination value may be set to a plurality of values that are corrected from the basic determination value to conditions that are stepwise stricter or looser than a predetermined value or ratio. When using a mathematical model, similar to the setting of this correction determination value, as a prior probability in Bayesian estimation, a distribution that places emphasis on a state where the risk of a harmful event is likely to occur (the side where the latent state is unstable) may be set. That is, depending on the type and degree of patient information before treatment, the correction determination value may not be limited to one step with respect to the basic determination value, but may be a value corrected to conditions that are two or three steps stricter or looser. For example, if a decrease in systolic blood pressure of 20 mmHg as biological information during treatment is the basic determination value, the correction determination value may be set to a stricter condition in two steps, such as a decrease in systolic blood pressure of 15 mmHg and a decrease in systolic blood pressure of 12 mmHg, depending on the patient information before treatment, or the correction determination value may be set to a looser condition in two steps, such as a decrease in systolic blood pressure of 25 mmHg and a decrease in systolic blood pressure of 30 mmHg, depending on the patient information before treatment. Also, the correction determination value may be calculated individually according to the type of harmful event to be predicted. Specifically, when predicting the risk of spasm and when predicting the risk of heart failure, a difference may be provided in the strict conditions and loose conditions when calculating the correction determination value. Alternatively, in a setting where an alarm is issued when the risk probability of blood pressure decrease exceeds a predetermined ratio, for example, 50%, the ease of increase (sensitivity) of the probability may be set high. 【0048】The above-mentioned corrected judgment value may be calculated by performing a statistical analysis on the presence or absence of adverse events based on the basic judgment value, pre-treatment patient information, and intra-treatment biological information, creating a predictive model formula to predict the risk of adverse events, and then applying the pre-treatment patient information or intra-treatment biological information to that predictive model formula. Alternatively, a trained model may be generated based on the basic judgment value, pre-treatment patient information, and intra-treatment biological information, and the risk of adverse events may be calculated using that trained model. In this case, the pre-treatment patient information functions as an "initial risk bias" or "offset" in the predictive model. For example, in patients with a history of heart failure or elderly patients, the autonomic nervous system's compensatory function for maintaining hemodynamics (maintaining blood pressure through sympathetic nerve tension) is either weak or already in a state where compensatory function is at work when treatment begins. Therefore, the judgment criterion setting unit 12 optimizes the judgment criteria for these patients by underestimating the initial value of the "latent state" indicating hemodynamic stability (setting a higher starting point for the probability of risk occurrence), or by setting a larger rate coefficient for deterioration of the condition, so that risk can be detected earlier and with minor changes in biological information than in healthy individuals. 【0049】 The calculation of the corrected threshold value will be explained using dialysis patients as an example. 1. For first-time dialysis patients, based on literature (e.g., Jennifer E. Flythe et al., KDIGO executive conclusions, Kidney International (2020) 97, 861-876), a 20 mmHg decrease in systolic blood pressure is set as the basic threshold for seizures or headaches. Next, for patients with a tendency towards low systolic blood pressure as indicated by pre-treatment patient information, or for patients with heart failure or ischemic heart disease, a stricter corrected threshold value of 15 mmHg is set. This is to reflect in the model the limited compensatory capacity for the decrease in circulating blood volume in such patients. On the other hand, for patients with a tendency towards high systolic blood pressure as indicated by pre-treatment patient information, a more lenient corrected threshold value of 25 mmHg is set, as the acceptable range of blood pressure reduction is larger. Note that the above method for determining the corrected threshold value can also be applied, for example, when hydrogen gas is inhaled during dialysis treatment. 【0050】2. For dialysis patients undergoing their second or subsequent dialysis treatments, the biological information obtained during past treatments functions as training data for learning the patient's "individuality (parameters)." Specifically, the probability distribution (posterior distribution) updated at the end of the previous treatment is stored in the memory unit 20 and set as the prior distribution (initial risk prediction) for the current treatment. As a result, the patient's unique normal model is updated and accumulated with each repeated treatment, and the judgment criteria are automatically optimized so that an alarm is activated at the optimal timing for the patient to notify them of the risk. Similarly, a decrease in systolic blood pressure of 20 mmHg is set as the judgment value for the occurrence of seizures or headaches. Next, if seizures occurred during the previous dialysis as pre-treatment patient information, the corrected judgment value is set to a stricter condition of 15 mmHg, and if arrhythmia occurred during the previous dialysis as pre-treatment patient information, the corrected judgment value is set to an even stricter condition of 12 mmHg. As a result, for patients who have experienced an adverse event once, the model is automatically updated to ensure a safety margin. On the other hand, if no seizures occurred during the previous dialysis session with a systolic blood pressure drop of 20 mmHg, the corrected threshold is set to a more lenient 25 mmHg. Furthermore, if no other adverse events, including seizures, occurred during the first dialysis session, the corrected threshold is set to an even more lenient 30 mmHg. The method for determining the above corrected threshold can also be applied, for example, when using hydrogen-dissolved dialysate during dialysis. 【0051】 The calculation of the corrected judgment value may be set by combining multiple pre-treatment patient information sets. 【0052】 [Prediction of the Risk of Adverse Events] Adverse events during treatment with therapeutic agents include, for example, cardiovascular adverse events such as hypotension, hypertension, arrhythmia, and myocardial ischemia; hematological adverse events such as hemolysis, thrombosis, and anemia; neurological adverse events such as dialysis ataxia syndrome, headache, dizziness, numbness, and convulsions; gastrointestinal adverse events such as nausea, vomiting, constipation, and liver damage; skin and allergic adverse events such as rash and anaphylaxis; and renal and urinary tract adverse events such as renal impairment, frequent urination, and urinary retention. Adverse events in dialysis patients undergoing dialysis treatment include adverse blood pressure fluctuations such as hypotension, arrhythmia, organ ischemia (angina pectoris, abdominal pain, leg pain), convulsions, and muscle cramps. 【0053】 The adverse event risk prediction unit 14 predicts the risk of adverse events in patients being treated with a fluid therapeutic agent based on the judgment criteria and the biological information during treatment. Specifically, for example, if the biological information obtained from a patient being treated with a fluid therapeutic agent is above the judgment criteria, a high risk of adverse events can be predicted, and if the biological information obtained from a patient being treated with a fluid therapeutic agent is below the judgment criteria, a low risk of adverse events can be predicted. Alternatively, the probability that the biological information obtained from a patient being treated with a fluid therapeutic agent deviates from a normal model can be calculated, and if this probability is low, i.e., if the probability of deviation from the normal model is high, a high risk of adverse events can be predicted, and if the probability that the biological information obtained from a patient being treated with a fluid therapeutic agent deviates from a normal model is high, i.e., if the probability of deviation from the normal model is low, a low risk of adverse events can be predicted. Furthermore, the prediction of the risk of adverse events may include not only threshold processing to determine whether the numerical value of the acquired biological information exceeds a standard value, but also statistical inference processing using time-series data. Conventionally, it is widely practiced to predict the risk of adverse events based on biological information obtained from patients being treated with a therapeutic agent. However, the determination of whether or not adverse events occur varies from patient to patient, and treatment staff decide whether or not to discontinue the administration of therapeutic drugs based on their own judgment. This monitoring system 1 uses not only biological information during treatment but also patient information before treatment, thereby improving the accuracy of the determination of whether or not adverse events will occur for each patient, and suppressing fluctuations in the judgment of individual treatment staff. 【0054】 The risk of adverse events includes not only the risk of adverse events occurring that are not currently present, but also the risk of mild adverse events currently occurring worsening. This monitoring system 1 evaluates the risk of adverse events in real time using the posterior probability of abnormal occurrence calculated based on biological information during treatment. This enables judgments based on probabilistic confidence rather than threshold judgments, reducing unnecessary alarms and achieving early detection of adverse events. 【0055】 [Storage Unit] The storage unit 20 stores information used by the monitoring system 1 to implement this method, such as pre-treatment patient information acquired by the pre-treatment patient information input unit 11, intra-treatment biological information acquired by the intra-treatment biological information acquisition unit 13, and judgment criteria set by the judgment criteria setting unit 12. The storage unit 20 is composed of semiconductor memory elements such as an HDD (Hard Disk Drive), SSD (Solid State Drive), RAM (Random Access Memory), and flash memory. 【0056】 [Display Unit] The display unit 15 displays pre-treatment patient information and in-treatment biological information, and may also display judgment criteria and predicted results of the risk of adverse events. The items to be displayed may be selectable by the treatment staff operating this monitoring system 1. A monitor or tablet can be used as the display unit 15. 【0057】 [Risk Notification Unit] The risk notification unit notifies the risk of adverse events in the adverse event risk prediction unit 14. Notifications can be made by displaying them on warning lamps or the display unit 15. For example, a yellow lamp can be lit when the risk of adverse events is high, and a green lamp can be lit when the risk of adverse events is low. Notifications may also be sent to the outside of this monitoring system 1 by wireless transmission methods such as Wi-Fi or Bluetooth®, or by wired transmission by connecting to a device that acquired pre-treatment patient information via a wired connection. 【0058】Traditionally, even when warnings were displayed, treatment was frequently continued without adverse events occurring. This was because warnings were displayed uniformly when the values ​​of predetermined items in the patient's biological information fell below a certain standard, without considering individual patient-specific warning levels. In such situations, intervention was required even for patients who did not require it, increasing the burden on treatment staff. This monitoring system 1 can predict the risk of adverse events based on individual patient information, enabling more accurate prediction of the risk of adverse events for patients and reducing the burden on treatment staff. 【0059】 [Treatment Control Unit] The treatment control unit controls the treatment with the fluid therapeutic agent when the risk prediction unit 14 predicts a high risk of adverse events. Specifically, in the case of dialysis, the treatment control unit can reduce the ultrafiltration rate from 15 mL / kg / hr to 13 mL / kg / hr or 10 mL / kg / hr, reduce the ultrafiltration rate from the initial setting to 1 / 2 or 1 / 4, or stop ultrafiltration altogether, or conversely, increase the ultrafiltration rate from the initial setting to 3 / 2, 4 / 2 or 5 / 2. It can also control the dialysis monitoring device to interrupt dialysis itself. 【0060】 Furthermore, when using a hydrogen gas supply device to administer hydrogen to patients during dialysis, if a high risk of adverse events is predicted, the hydrogen gas supply device can be controlled to slow down the hydrogen inhalation rate or interrupt the hydrogen inhalation. In addition, the ultrafiltration rate and the hydrogen inhalation rate may be adjusted. 【0061】 [Adverse Event Recording Unit] This monitoring system 1 may be equipped with an adverse event recording unit. When an adverse event occurs, it can record biological information during treatment at the time of the event and before and after it. The recorded data will be used as pre-treatment patient information for the next treatment of the same type. 【0062】[Operation] Next, the operation of this monitoring system 1 will be explained. This monitoring system 1 is used according to steps S1 to S5 shown in Figure 4 to predict the risk of adverse events occurring in patients being treated with a fluid therapeutic agent. 【0063】 (Step S1) Pre-treatment patient information obtained in advance from the patient by measurement or interview before treatment with the therapeutic agent is input into the monitoring system 1. Input may be performed via each item of pre-treatment patient information displayed on a touch panel, or it may be performed in real time by wireless transmission from the device that measured the pre-treatment patient information. (Step S2) Based on the pre-treatment patient information, criteria for determining the risk of adverse events for each biological information during treatment are set. (Step S3) Treatment with the fluid therapeutic agent is started, and biological information during treatment is acquired. (Step S4) Based on the criteria and the patient's biological information during treatment, the risk of adverse events is predicted. (Step S5) If the prediction is that the risk of adverse events is high, the treatment with the therapeutic agent is controlled (stopped or adjusted, etc.). If treatment is stopped, the series of processes is terminated. If treatment is adjusted and continued, the process returns to Step S3. On the other hand, if the prediction is that the risk of adverse events is not high (low risk), the treatment completion conditions, such as whether a predetermined treatment time has elapsed, are determined. If the treatment completion conditions are not met, the process returns to Step S3, and the acquisition of biological information during treatment and risk prediction continue. If the conditions for completing treatment are met, the series of procedures will be terminated. 【0064】The following describes a specific example of the blood pressure reduction prediction process using this monitoring system 1. In this example, time-series parameters based on respiratory waveforms obtained from a nasal cannula of a dialysis patient are used. (1) Model initialization (setting of prior distribution) In the pre-treatment patient information input step, information on patient A (in their 70s, with diabetic nephropathy and a history of heart failure) is entered. The judgment criterion setting unit 12 uses this pre-treatment patient information as an explanatory variable and sets a prior probability distribution for the initial value of a hemodynamic stability index called the "latent state" of patient A, using coefficients of a linear model estimated from statistical data of past similar patient groups. These coefficients of the linear model were calculated by estimating the initial value of each patient's latent state from the initial respiratory and blood pressure fluctuations using a Kalman filter for past dialysis patient data, and then performing a regression analysis with the estimated value as the dependent variable and the pre-treatment patient information as the explanatory variable. Risk factors such as a history of heart failure and advanced age have been confirmed to be significantly associated with the "unstable side" of the latent state in previous statistical analyses, and this system refers to the corresponding coefficients to set the prior probability for patient A's initial state. (2) Calibration Phase (Construction of a Patient A-Specific Normal Model) For a predetermined period after the start of dialysis (e.g., 15 minutes), the system processes respiratory waveforms acquired from a nasal cannula at approximately 50-100 Hz per second in real time and calculates features such as respiratory cycle, inspiratory time, expiratory time, IE ratio, and respiratory variability. Simultaneously, samples are acquired from a blood pressure monitor every minute or from a continuous blood pressure monitor. A state-space model is applied to the time-series data during the calibration period, and the posterior distribution of the latent state is updated sequentially. In this process, the statistical features of the respiratory-circulatory response specific to patient A (mean, variance, autocorrelation, cross-correlation structure of respiration and blood pressure fluctuations, etc.) are estimated. By integrating these features and the distribution of the latent state, a "normal model" specific to patient A (a model that expresses the normal range as a probability distribution) is formed. (3) Detection of warning signs (estimation of hidden conditions) In the middle of dialysis, as fluid removal progresses, even if there is no obvious change in the blood pressure value itself, changes in respiratory parameters such as an increase in respiratory rate, a decrease in inspiratory time, and a sudden change in the IE ratio may be observed in patient A. The prediction unit 14 calculates the degree of abnormality by comparing such multivariate changes with a normal model.Specifically, the posterior probability distribution of the latent state is recalculated using the observed respiratory parameters to determine "how much the latest state deviates from the probability distribution of the normal model." Statistically, this degree of deviation corresponds to the Mahalanobis distance based on the innovation (prediction error) and its covariance in the Kalman filter, or the likelihood ratio using a particle filter. This estimation indicates that if the change in respiratory data suggests "the limit of the sympathetic nervous system's compensatory response," the probability of deterioration of the latent state increases, and this is judged to be a sign of a drop in blood pressure. (4) Prediction and Control The prediction unit 14 predicts future biological information from the current latent state and calculates, for example, the probability that systolic blood pressure will fall below a predetermined threshold in a few tens of minutes. This prediction uses a method that simulates future blood pressure fluctuations using a Bayesian predictive distribution, with the patient A's specific normal model and the latest latent state as input. If the predicted probability of a drop in blood pressure exceeds a predetermined standard, the treatment control unit can be configured to automatically reduce the fluid removal rate from 15 mL / kg / hr to 10 mL / kg / hr to prevent a drop in blood pressure and simultaneously issue an alarm to the therapist. This allows for treatment adjustments tailored to patient A's condition before a drop in blood pressure actually occurs. 【0065】 Furthermore, this disclosure can be implemented not only as the monitoring system 1 described above, but also as a method for predicting the risk of adverse events, including the processing performed by each component of the monitoring system 1. In addition, this disclosure can be implemented as a program for causing a computer to execute each step included in the method, or as a computer-readable recording medium on which the program is recorded. That is, each step in the flowchart or sequence diagram relating to the embodiment of this disclosure also represents the processing procedure of the program, and the function of executing each step is realized through the cooperation of hardware resources (processor, memory, etc.) and software (program). 【0066】 1 Monitoring System 11 Pre-treatment patient information input unit 12 Judgment criterion setting unit 13 In-treatment biological information acquisition unit 14 Adverse event risk prediction unit 15 Display unit 20 Storage unit

Claims

1. A monitoring system for predicting the risk of adverse events in a patient being treated with a fluid therapeutic agent, comprising: a pre-treatment patient information input unit for inputting individual pre-treatment patient information of a patient obtained in advance from the patient before treatment with the therapeutic agent; a judgment criterion setting unit for setting judgment criteria for the risk of adverse events based on the pre-treatment patient information; a treatment biological information acquisition unit for acquiring treatment biological information of the patient being treated with the therapeutic agent; an adverse event risk prediction unit for predicting the risk of adverse events in the patient being treated with the fluid therapeutic agent based on the judgment criteria and the treatment biological information; and a storage unit for storing the pre-treatment patient information, the treatment biological information and the judgment criteria.

2. The monitoring system according to claim 1, characterized in that the treatment-in-treatment biological information acquisition unit acquires in real time the treatment-in-treatment biological information of the patient being treated with the therapeutic agent.

3. The monitoring system according to claim 1 or 2, characterized in that the judgment criterion setting unit obtains a corrected judgment value by correcting the predetermined basic judgment value based on the patient's individual pre-treatment patient information and uses that as the judgment criterion.

4. The monitoring system according to any one of claims 1 to 3, characterized in that the individual pre-treatment patient information of the patient includes information regarding the occurrence of adverse events during treatment of the same type in the past.

5. The monitoring system according to any one of claims 1 to 4, further comprising a display unit that displays the pre-treatment patient information and the in-treatment biological information.

6. The monitoring system according to any one of claims 1 to 5, wherein the pre-treatment patient information is at least one selected from the group consisting of blood pressure, dry weight, body weight, weight gain rate between dialysis, weight gain between dialysis, heart rate, oxygen saturation, age, hematocrit value, body temperature, circulating blood volume, SOFA score, parameters related to the respiratory cycle, electrocardiogram, cardiothoracic ratio, parameters obtained by echocardiography, an index indicating the degree of edema, medical information, medical history of a specified disease, family history, treatment history, medications currently being taken, smoking habits, blood glucose level, biochemical test results, hematological test results, patient information during treatment from previous similar treatments, and the concentration of the active ingredient of the therapeutic agent in the fluid in the body.

7. The monitoring system according to any one of claims 1 to 6, wherein the biological information during treatment is at least one selected from the group consisting of blood pressure, heart rate, oxygen saturation, hematocrit value, weight change, body temperature, respiratory cycle parameters, circulating blood volume, rate of change in circulating blood volume, peripheral venous pressure, central venous pressure, pulmonary artery pressure, pulmonary artery wedge pressure, mixed venous blood oxygen saturation, blood concentration of the active ingredient of the therapeutic agent in the fluid in the body, and ultrafiltration rate.

8. The monitoring system according to any one of claims 1 to 7, characterized in that the fluid is a gas containing hydrogen gas or a liquid in which hydrogen gas is dissolved.

9. The monitoring system according to any one of claims 1 to 8, further comprising a risk notification unit that notifies the user to stop or adjust treatment with the fluid therapeutic agent when the risk prediction unit predicts a high risk of an adverse event occurring.

10. The monitoring system according to any one of claims 1 to 9, further comprising a treatment control unit that controls treatment with the fluid therapeutic agent when the risk of occurrence of an adverse event is predicted to be high in the adverse event risk prediction unit.

11. The monitoring system according to any one of claims 1 to 10, wherein the patient is a dialysis patient, and the pre-treatment patient information includes blood pressure, inter-dialysis weight gain rate, or dry weight.

12. The monitoring system according to claim 11, wherein the pre-treatment patient information includes patient information during past dialysis treatments.

13. The monitoring system according to any one of claims 1 to 12, characterized in that the judgment criterion setting unit creates a normal model which is the probability distribution of the biological information during treatment in a normal state in which no adverse events occur, based on the pre-treatment patient information, and uses it as the judgment criterion; and the adverse event occurrence risk prediction unit calculates the probability that the acquired biological information during treatment arises from the normal model based on the normal model and the biological information during treatment, and predicts the risk of adverse events occurring based on that probability.

14. A method for predicting the risk of adverse events in a patient being treated with a fluid therapeutic agent, performed by a monitoring system for predicting the risk of adverse events in a patient being treated with a fluid therapeutic agent, the monitoring system comprising: a pre-treatment patient information input unit for inputting individual pre-treatment patient information of a patient obtained in advance from the patient before treatment with the therapeutic agent; a judgment criterion setting unit for setting judgment criteria for the risk of adverse events based on the pre-treatment patient information; a treatment biological information acquisition unit for acquiring treatment biological information of the patient being treated with the therapeutic agent; an adverse event risk prediction unit for predicting the risk of adverse events in the patient being treated with the fluid therapeutic agent based on the judgment criteria and the treatment biological information; and a storage unit for storing the pre-treatment patient information, the treatment biological information and the judgment criteria, wherein the monitoring system comprises: a pre-treatment patient information input step for inputting pre-treatment patient information obtained in advance from the patient before treatment with the therapeutic agent; a judgment criterion setting step for setting judgment criteria for the risk of adverse events based on the pre-treatment patient information; and a treatment biological information acquisition step for acquiring treatment biological information of the patient being treated with the therapeutic agent. A method for predicting the risk of adverse events in a patient being treated with a fluid therapeutic agent, comprising: an adverse event risk prediction step that predicts the risk of adverse events occurring in the patient being treated with the fluid therapeutic agent based on the above-mentioned judgment criteria and the biological information during treatment; and a storage step that stores the pre-treatment patient information, the biological information during treatment, and the judgment criteria.

15. A program capable of predicting the risk of adverse events occurring in patients being treated with a fluid therapeutic agent, characterized in that a computer is made to function as a monitoring system according to any one of claims 1 to 13.

Images