WO2024037796A1 - Digital system for prioritizing admission and/or treatment of a patient - Google Patents

Digital system for prioritizing admission and/or treatment of a patient Download PDF

Info

Publication number
WO2024037796A1
WO2024037796A1 PCT/EP2023/069311 EP2023069311W WO2024037796A1 WO 2024037796 A1 WO2024037796 A1 WO 2024037796A1 EP 2023069311 W EP2023069311 W EP 2023069311W WO 2024037796 A1 WO2024037796 A1 WO 2024037796A1
Authority
WO
WIPO (PCT)
Prior art keywords
patient
treatment
user equipment
health condition
health
Prior art date
Application number
PCT/EP2023/069311
Other languages
French (fr)
Inventor
Azadeh MEHRABI
Matthias Wesselmann
Original Assignee
Biotronik Ag
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Biotronik Ag filed Critical Biotronik Ag
Publication of WO2024037796A1 publication Critical patent/WO2024037796A1/en

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/60—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
    • G16H40/67—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for remote operation
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/50—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for simulation or modelling of medical disorders

Definitions

  • the present invention relates to a digital system for optimizing the process of prioritizing admission/treatment at hospitals based on an urgency rating of a treatment of a patient.
  • the invention relates to machine learning based methods for predicting such an urgency rating as well as to a method of training a corresponding model.
  • a re-admission of a patient is highly unsatisfying.
  • the reason for this is that not only the patient consumes part of the already limited resources, but the fact that the patient is a re-admission indicates that the consumption of these resources could have been avoided if the patient would have been treated correctly at his earlier stay in the hospital. Therefore, the medical staff of a hospital is eager to keep and to reduce the readmission rate of patients as low as possible.
  • the re-admission rates of hospitals are even taken as a quality metric which can result in the hospital being penalized if the rate is too high.
  • Efforts have been made to optimize the process of prioritizing admission/treatment like providing a web-based system designed to automatically check-in and prioritize patients upon arrival to an emergency department.
  • discharge toolkits for educating patients have been developed. These toolkits try to explain patients about likely future pains or events after a surgery. For example, if the patient knows that light fever is most likely to occur within 1 week after a surgery, the patient can prepare and will probably not re-enter the emergency room because of the fever.
  • a first embodiment of the present invention is a computer-implemented method for training an artificial intelligence model for predicting an urgency rating of a treatment, the method comprising: inputting training data to the artificial intelligence model to train the model, the training data including a plurality of data samples, wherein each data sample comprises: data describing a health condition of a patient; and an urgency rating of a treatment associated with the data describing the health condition of the patient.
  • the urgency rating may be defined as either low, medium or high. Additionally, or alternatively, the urgency rating may be a value within a predefined interval/scale (e.g., on a scale from 1 to 10 where 1 is the lowest urgency rating and 10 is the highest urgency rating).
  • urgency rating of a treatment in the context of the present specification particularly indicates the need of the treatment and the priority of the treatment particularly compared to other patients. For example, a predicted urgency rating of a treatment having the value “low” or “zero” may be indicative for non-urgent need or no need for a treatment at all.
  • the artificial intelligence model trained according to this method may allow for accurate prediction of an urgency of a treatment of a patient based on his current health condition.
  • this may enable a system of triage based on data describing the health condition of the patients, without the patients needing to enter the hospital.
  • the training data may additionally or alternatively comprise: time-series data of a patient consisting of a plurality of data describing a health condition of the patient at different points in time each associated with an urgency rating of the treatment.
  • Training the artificial intelligence model based on time-series data of the patient may allow the model to predict a future health condition and/or a future urgency rating of the patient. This may allow for an increased accuracy of the urgency prediction as the model trained accordingly can analyze and predict the trend of a patient’s health condition. Therefore, a negative/positive trend can be identified way in advance improving the quality of treatment of the patient (e.g., valuable early treatment time can be won this way).
  • a third embodiment of the present invention is a computer-implemented method for predicting an urgency rating of a treatment of a patient, the method comprising: predicting the urgency rating of the treatment of the patient based on data describing a health condition of the patient using an artificial intelligence model, wherein the artificial intelligence model is trained according to the method of the 1 st embodiment.
  • Providing this method may allow for accurate prediction of an urgency of a treatment of a patient based on his current health condition. For example, a patient with a non-specific chest pain after a PCI may get a higher urgency rating (rate) compared to a patient with symptoms of a cold. Accordingly, if a patient’s current health condition results in the prediction of for example a low urgency rating, the patient may not be required to visit an emergency room avoiding the inefficient usage of resources. Accordingly, the triage is already made from outside the hospital as it can be identified whether a patient requires sooner or later treatment. This allows to improve and facilitate the flow of information, scheduling of treatments and patient transfers to the hospital. As a result, the new triage system using the provided method reduces the fluctuating workloads of hospitals resulting in a better-balanced workload and avoiding unnecessary triage situations in an emergency room.
  • the data describing a health condition of a patient comprise or are constituted by sensor data.
  • the method may additionally or alternatively comprise: predicting data describing a future health condition of the patient and, optionally, the urgency rating of the treatment of the patient based on the data describing the future health condition of the patient using an artificial intelligence model, wherein the artificial intelligence model is trained according to the method of the second embodiment.
  • the method may predict the future urgency rating of the treatment of the patient without the step of predicting the future health condition.
  • Providing this method may allow for an accurate prediction considering the trend of a patient’s health condition. Therefore, a negative/positive trend can be identified way in advance improving the quality of treatment of the patient (e.g., valuable early treatment time can be won this way).
  • the individual values of the data describing the health condition of the patient may be associated with a set of adjustable weights. These weights are adjusted according to a selected prediction metric of a set of prediction metrics before predicting the urgency rating of the treatment of the patient. Examples for such metrics may be - inter alia - a quality of life metric, a pain related metric, a lethality metric or an effort of treatment metric.
  • Adjusting the data describing the health condition before predicting the urgency rating may allow for increased efficiency in certain situations. For example, a rehabilitation hospital department may decide to choose the quality of life metric to determine whether a certain patient has a high demand of a corresponding treatment after a surgery to improve the healing process and to end up with a high quality of life.
  • a fifth embodiment of the present invention is a method for remote and/or wireless communication at a portable user equipment of a patient, the method comprising: collecting data describing a health condition of the patient from at least one data source; transmitting the collected data describing the health condition of the patient to a server; and receiving, from the server, an urgency rating of a treatment based on the health condition of the patient.
  • the method may only be performed when executed by the patient or it may be performed automatically. In case the method is performed automatically, the method may be performed on a regular basis (e.g., once a day, once every hour, once a week). The frequency may be configurable by the patient or may depend on the current and/or past health conditions and/or urgency ratings of the patient. The method may also be performed each time a change in the data describing the health condition of the patient is detected.
  • the method further comprises: receiving from the server or a user equipment of a health care provider, a treatment plan associated with the urgency rating. For example, if the urgency rating of a patient is low (e.g., 1 to 3) the associated treatment plan may contain a health advice (e.g., a less fat diet). Should the urgency rating be medium (e.g., 3 to 6) the associated treatment plan may contain for example a (recommendation for a) medical prescription or a scheduled appointment at a corresponding physician.
  • a treatment plan associated with the urgency rating For example, if the urgency rating of a patient is low (e.g., 1 to 3) the associated treatment plan may contain a health advice (e.g., a less fat diet). Should the urgency rating be medium (e.g., 3 to 6) the associated treatment plan may contain for example a (recommendation for a) medical prescription or a scheduled appointment at a corresponding physician.
  • the physician or the (recommendation for the) medical prescription may be selected based on the health condition of the patient (e.g., if the patient reports headache he may get a (recommendation for a) medical prescription for a headache medicine, if the patient shows light signs of heart issues the selected physician may be a cardiologist). However, should the urgency rating be high (e.g., 7 to 10), an ambulance call for transferring the patient to the hospital may be initiated and the treatment plan may contain an information for the patient about the emergency and that an ambulance is on its way.
  • health care provider in the context of the present specification particularly refers to a person, a group, a facility, or an organization which are involved in and contribute to health care in general or emergency management in particular.
  • Non limiting example for health care provider include medical staffs, hospital facilities, ambulances, rescue helicopters, personal physicians and the like.
  • Providing this method decreases the risk of the patient entering a critical health condition as it can be monitored frequently. Moreover, the threshold of a patient to report his symptoms and to get in contact with a physician is lowered. Therefore, the overall prioritizing admission/treatment process is improved as fewer emergency situations occur and the patient will most likely not enter the emergency room unnecessarily.
  • the health care provider e.g. medical staff of a hospital
  • the health care provider may already be informed about health condition of the patient and his expected time of arrival improving the quality of treatment (e.g. since the server may alert the corresponding hospital accordingly).
  • a sixth embodiment of the present invention is a method for remote and/or wireless communication at a server, the method comprising: receiving, from a portable user equipment of a patient, data describing a health condition of the patient; determining an urgency rating of a treatment of the patient based on the data describing the health condition of the patient; the method further comprising at least one of the following steps: transmitting the urgency rating of the treatment to the portable user equipment; transmitting the urgency rating of the treatment and the data describing the health condition of the patient to a user equipment of a health care provider for evaluating the urgency rating.
  • Providing this method may improve the overall prioritizing admission/treatment process as the urgency rating of the treatment of a patient can be determined without the patient being physically present in the emergency room.
  • determining the urgency rating of the treatment comprises: predicting the urgency rating of the treatment based on the data describing the health condition of the patient according to the method of the 3 rd embodiment; additionally or alternatively, it may comprise; predicting the urgency rating of the treatment according to the method of the 4 th embodiment; adding the data describing the health condition of the patient and the predicted urgency rating to time-series data of the patient indicating the health condition and associated urgency ratings of the patient over time; and retraining the artificial intelligence model according to the method of the 2 nd embodiment.
  • Determining the urgency rating using a specifically trained artificial intelligence model increases the accuracy of the urgency rating. Therefore, the assessment of the patient’s health condition is improved.
  • the method may further comprise: receiving, from the user equipment of the health care provider; an evaluated urgency rating.
  • the method may further comprise: retraining the artificial intelligence model according to the training method of the 1 st , the 2 nd or the subsequent embodiments using the data describing the health condition of the patient and/or the evaluated urgency rating as an additional data sample.
  • the server may evaluate the urgency rating additionally or alternatively to the evaluation of the health care provider, by determining data describing a health condition out of a plurality of data describing health conditions (each associated with an urgency rating) of previous treatments (from the same or other patients) similar to the data describing the health condition of the patient. The server may then compare the urgency rating associated with that similar data to the urgency rating predicted for the data describing the health condition of the patient. If the difference between both urgency ratings is acceptable (i.e., the difference is for example below a predefined threshold) the predicted urgency rating is accepted.
  • Providing these additional steps may increase the overall accuracy of the urgency rating prediction resulting in an improved assessment of the patient’s health condition.
  • the method may further comprise: selecting the user equipment of the health care provider from a plurality of user equipment of health care providers prior to transmitting the urgency rating of the treatment and the data describing the health condition of the patient to the user equipment of the health care provider; wherein selecting the user equipment of the health care provider is based at least on one of: a distance between the portable user equipment of the patient and the user equipment of the health care provider, a workload of the health care provider, a treatment specialization of the health care provider and/or a treatment success rate of the health care provider for the health condition described by the data.
  • the aforementioned treatment specialization and treatment success of the health care provider may also include or may be supplemented by or previous treatment history by the health care provider. For example, if the patient was treated at the one specific health care provider (e.g. a hospital) before, then the patient is most likely to be sent to the same health care provider by the system.
  • the treatment as well as the assessment/evaluation of a patient’s urgency level and health condition are improved. For example, in case a high urgency level is predicted (i.e., a transfer to an emergency department is required) selecting the health care provider (e.g. a corresponding hospital) closest to the patient can increase the patient’s change of survival.
  • the several criteria may also be combined. For example, if two different user equipment of two different health care providers have the same distance to a patient, the treatment specialization of each health care provider can be taken as an additional criteria for deciding which health care provider to choose.
  • the corresponding user equipment of the cardiologist is selected.
  • the distance may be determined based on position data, e.g. GPS data, received from the various user equipment.
  • a ninth embodiment of the present invention is a method for wireless communication at a user equipment of a health care provider, the method comprising: receiving, from a server, an urgency rating of a treatment and data describing a health condition of a patient; determining an evaluated urgency rating of the treatment based at least on the data describing the health condition of the patient; and transmitting the evaluated urgency rating of the treatment to the server.
  • the health care provider may also transmit a treatment plan based on the evaluated urgency rating to the server and/or directly to the patient. Additionally or alternatively, also the evaluated urgency rating may directly be transmitted to the patient.
  • the health care provider can evaluate the health condition of a patient more efficiently. As a result, fewer patients in critical condition are to be transported into the emergency room as these patients can efficiently be treated preventively by transmitting a corresponding treatment plan to the patient (directly or via the server). Furthermore, the situations in which a patient enters an emergency department without his health condition requiring it, can be reduced due to the improved treatment quality.
  • the data describing the health condition of the patient includes at least one of: a first set of health parameters and a second set of health parameters depending on the first set of health parameters, and/or captured vital signals of the patient.
  • the first set of health parameters may include or may come from at least one of: fever (e.g., temperature in C/F, duration of fever), pain (e.g., intensity on a scale from 1-10, duration, frequency, location), cough (dry/wet, pain, difficulty to breathe), nausea/vomiting, dizziness, blindness, numbness, deafness, rash, injury, coloration, blood pressure, pulse, flatulence, diarrhea and/or photos.
  • fever e.g., temperature in C/F, duration of fever
  • pain e.g., intensity on a scale from 1-10, duration, frequency, location
  • cough dry/wet, pain, difficulty to breathe
  • nausea/vomiting dizziness, blindness, numbness, deafness, rash, injury, coloration, blood pressure, pulse, flatulence, diarrhea and/or photos.
  • the first set of health parameters or a part of these parameters may be collected based on a general questionary as for example conducted by a general practitioner (e.
  • the first set of health parameters may be collected directly from the portable user equipment of the patient by means of sensors and/or may be requested from a database (e.g., a database from the general practitioner who has examined the patient).
  • the first set of health parameters may also be collected by means of other suitable sensors (e.g., a fever monitor connected to the portable user equipment of the patient).
  • the second set of health parameters depending on the first set of parameters may include or may come from at least one of: blood diagnostic (inflammation), tissue diagnostics (e.g., from an eternal lab), stool diagnostics, ultrasound examination, endoscopic images, X-Ray images, magnetic resonance imaging, long term blood pressure monitoring, pulse, fever (e.g., temperature in C/F, duration), pain (e.g., intensity on a scale from 1 to 10, duration, frequency, location), cough (dry/wet, pain, difficulty to breathe), nausea/vomiting, dizziness, blindness, numbness, deadness, rash, injury and/or coloration.
  • the second set of heath parameters may be collected based on a specific questionary as for example conducted by specialized practitioner.
  • the set of health parameters may depend on the first set of health parameters in a way that based on the values/result of the first health parameters the relevant parameters of the second health parameters are selected.
  • the second set of health parameters may be defined by a specialized practitioner after having received (e.g., via communication means) the first set of health parameters.
  • the second set of health parameters may be collected directly from the portable user equipment of the patient, e.g. by means of a user interface, one or more sensors and/or may be requested from a database (e.g., a database from external labs or specialized practitioners who for example have collected X-Ray images of the patient previously).
  • the second set of health parameters may also be collected by means of other suitable sensors (e.g., a fever monitor connected to the portable user equipment of the patient).
  • the captured vital signs may be collected via sensors (e.g. sensor data). These sensors may either be incorporated into the portable user equipment of the patient or may be connected or connectable with the portable user equipment. Examples for such sensors may be wearables like a smart watch, a medical device, a medical implant such as a pacemaker, and/or a step counter.
  • the captured vital signs may correspond to ECG information measured by the smart watch, a number of steps walked measured by the step counter or beats per minute measured by the pacemaker.
  • the data describing the health condition of the patient may also include a GPS information about the current location of the patient collected via a GPS sensor of the portable user equipment of the patient.
  • Providing the GPS information about the location of the patient may increase the patient’s chance of surviving as not only the nearest health care provider (e.g. medical staff or hospital) can be identified but also the corresponding ambulance has accurate location information resulting in a reduced time of transfer. Additionally, the health care provider is informed about the patient’s emergency in advance and can make appropriate preparations.
  • the nearest health care provider e.g. medical staff or hospital
  • the GPS position of the patient may be used to extract information on physical activity.
  • the velocity derived from the GPS characterizes the times where the patient was immobile, walking, cycling or using vehicles. Shifts in a patient’s behavior monitored over longer time period allow to detect if a patient health has improved or is deteriorated.
  • an Al may be trained to stabilize and improve physical activity and/or to detect suspicious deterioration of the patient in time to act before the heart is permanently damaged.
  • An eleventh embodiment of the present invention is a server comprising means for performing the method according to any of the embodiments 6 to 8 or 10 when referred back to any of the embodiments 6 to 8.
  • a twelfth embodiment of the present invention is a portable user equipment of a patient comprising means for performing the method according to the 5 th embodiment or 10 th embodiment when referred back to the 5 th embodiment.
  • a thirteenth embodiment of the present invention is a user equipment of a health care provider comprising means for performing the method according to the 9 th embodiment or the 10 th embodiment when referred back to the 9 th embodiment.
  • the user equipment of the health care provider may design as an app, a web-based application, an application specific device, a computer program and the like.
  • a fourteenth embodiment of the present invention is a digital triage system for remote and/or wireless communication, wherein the system comprises: a server according to the 11 th embodiment; and at least one of: a portable user equipment of a patient according to the 12 th embodiment; and/or a user equipment of a health care provider according to the 13 th embodiment.
  • a fifteenth embodiment of the present invention is a computer program comprising instructions, which when executed by a computer cause the computer to perform a method according to any of the embodiments 1 to 10.
  • FIG. 1 Illustration of a digital triage system in accordance with the embodiments of the present invention
  • Fig. 1 shows a system for prioritizing admission/treatment triage 100 (i.e. digital prioritizing admission/treatment system) in accordance with embodiments of the present invention.
  • the digital prioritizing admission/treatment system 100 comprises a server 120 and at least one portable user equipment of a patient 110 and/or at least one user equipment of a health care provider 130 in accordance with embodiments of the present invention.
  • the digital prioritizing admission/treatment system 100 and its components may perform the methods described herein with respect to the corresponding embodiments.
  • a corresponding computer program is installed on the components which comprises instructions which when executed by the computer (e.g., the processor of the component) causes the component to execute the corresponding method according to the embodiments of the present invention.
  • the patient or the care giver of the patient may register via his portable user equipment (e.g., a smartphone, tablet, or a smartwatch capable of wireless communication) 110 to the digital prioritizing admission/treatment system 100.
  • the registration may be done via an app installed on the portable user equipment of the user 110 or a webapp (e.g., hosted on the server 120).
  • the patient may provide data describing his health condition to the digital triage system 100.
  • the data may be transmitted to the server and stored into a corresponding database.
  • the data and/or the patient identity may be anonymized.
  • the portable user equipment of the patient 110 may further be connected to one or more data sources (e.g., external data bases, sensors like sensors incorporated in a wearable like a smartwatch, step counter, pacemaker or any other implant and/or medical device) from which the data describing the health condition of the patient may be collected and/or updated according to a configured modus (i.e., only collect data for the urgency rating prediction when executed by the patient explicitly or on a regular basis (e.g., daily) or whenever new data is available from one of the sources).
  • a configured modus i.e., only collect data for the urgency rating prediction when executed by the patient explicitly or on a regular basis (e.g., daily) or whenever new data is available from one of the sources.
  • a configured modus i.e., only collect data for the urgency rating prediction when executed by the patient explicitly or on a regular basis (e.g., daily) or whenever new data is available from one of the sources.
  • a first set of health parameters i.e., only collect
  • the data describing the health condition of the patient may be transmitted to the server 120 for determining the urgency rating of a treatment based on the health condition of the patient.
  • the server 120 may use an artificial intelligence model trained to predict an urgency rating based on such data according to embodiments of the present invention.
  • the artificial intelligence model may be deployed on the server 120 or at another location with the server 120 having access thereto.
  • the determined urgency rating may then be transmitted to the portable user equipment of the patient 110. This way, the patient gets an accurate assessment of his current health condition without making a physical presence at a hospital necessary or without a need of overcoming the patient’s threshold of contacting a health care provider.
  • the server is able to determine based on data of previous patients having a similar health condition (potentially evaluated by a health care provider) and who have been treated according to a certain treatment plan (potentially created by the health care provider) a treatment plan for the patient. Additionally, or alternatively, the server 120 may transmit the urgency rating and the corresponding data describing the health condition of the patient to a user equipment of a health care provider 130. The server may determine the user equipment of the health care provider 130 to which the data is transmitted out of a plurality of available user equipment of health care provider 130 according to several criteria like distance, treatment specialization of the health care provider etc. Accordingly, the server 120 may sort the plurality of available user equipment of health care provider 130 based on the several criteria and can thus determine the ideal health care provider (e.g. medical staff or hospital) for the patient.
  • ideal health care provider e.g. medical staff or hospital
  • the server may determine the urgency ratings from all of the patients and transmit a list ranked in descending order based on the urgency ratings to the one or more connected user equipment of health care providers .
  • the health care provider 130 may evaluate the (correctness of the) urgency rating.
  • the health care provider will analyze the data describing the health condition and check whether the health condition is as urgent as indicated by the rating.
  • the evaluated urgency rating may then be transmitted back to the server 120, which may allow the server 120 to retrain the artificial intelligence model (the urgency rating may generally refer to the same treatment, but it may also be possible that an additional urgency level relating to a different treatment is transmitted back to the server 120). In some examples, such back-transmission may be omitted.
  • the evaluated urgency rating may also be transmitted from the user equipment of the health care provider 130 to the portable user equipment of the patient 110 either via the server 120 or directly.
  • a treatment plan may be transmitted to the patient.
  • the treatment plan may include an ambulance transfer of the patient to the corresponding hospital or a physician directly contacting the patient.
  • the treatment plan may include a scheduled appointment for the patient at a physician.
  • the treatment plan may only contain an advise like a healthier diet.
  • One additional benefit of the present invention is that by monitoring and assessing the health condition of a patient, conclusions about the success of certain treatment plans can be drawn. Assuming a portable user equipment of a patient 110 transmits data describing the health condition of the patient to the server 120 which predicts a high urgency rating and the patient is transferred to the emergency department because the patient for example needs heart surgery. The system allows monitoring of the development of the health condition of the patient even after leaving the hospital or the rehabilitation after the surgery.
  • the corresponding health care provider would notice this and could call the patient or schedule an additional appointment.
  • an insurance company or the management of the health care provide would participate in the system 100 for prioritizing admission/treatment according to the invention, e.g. by getting access to aggregated information which can be obtain form the system.
  • the system 100 according to the invention may detect patients with susceptible symptoms earlier and would help to detect critical patients that do not consult a physician early enough to avoid grave consequences. That will avoid a number of emergency cases by preventive treatment.
  • the management in the hospital could have an overview about the cost, admission or readmission rate and staff could plan the treatment.

Abstract

A digital prioritizing admission/treatment system and corresponding devices and methods are disclosed. The prioritizing admission/treatment system, devices and methods are based on an urgency rating of a treatment of a patient. The urgency rating may be determined using an artificial intelligence model. A corresponding training method of the artificial intelligence is disclosed as well.

Description

Digital System for Prioritizing Admission and/or Treatment of a Patient
The present invention relates to a digital system for optimizing the process of prioritizing admission/treatment at hospitals based on an urgency rating of a treatment of a patient. In addition, the invention relates to machine learning based methods for predicting such an urgency rating as well as to a method of training a corresponding model.
Fast and correct assessing an urgency of a medical condition remains a huge challenge for all participating units in the health care systems, e.g. for first responders, emergency units in the hospital, etc, particularly during a global pandemic situation or in the aftermath of a catastrophic event. In such situations, resources are generally insufficient for all patients to be treated at once. Therefore, patients are ranked depending on the urgency of their treatment. However, also in ordinary times, an efficient system for prioritizing the admission of a patient, e.g. in a hospital is desired.
In situations in which patients are not physically present in the emergency room of a hospital, an optimal health care may not be provided. This is because the corresponding medical staff is not aware of the potential critical condition of the patient at home. Reasons for such situations may be a pandemic and a corresponding lockdown, but also bad infrastructure making the trip to the hospital difficult for the patient.
In general, a re-admission of a patient is highly unsatisfying. The reason for this is that not only the patient consumes part of the already limited resources, but the fact that the patient is a re-admission indicates that the consumption of these resources could have been avoided if the patient would have been treated correctly at his earlier stay in the hospital. Therefore, the medical staff of a hospital is eager to keep and to reduce the readmission rate of patients as low as possible. In some countries, the re-admission rates of hospitals are even taken as a quality metric which can result in the hospital being penalized if the rate is too high.
Efforts have been made to optimize the process of prioritizing admission/treatment like providing a web-based system designed to automatically check-in and prioritize patients upon arrival to an emergency department. In addition, to avoid situations requiring admission/treatment due to re-admissions of patients, discharge toolkits for educating patients have been developed. These toolkits try to explain patients about likely future pains or events after a surgery. For example, if the patient knows that light fever is most likely to occur within 1 week after a surgery, the patient can prepare and will probably not re-enter the emergency room because of the fever.
However, known systems for optimizing the prioritizing admission/treatment have a main drawback. These systems only start with the admission/treatment of the patient upon arrival of the patient to the emergency department. This renders the process highly inefficient due to several reasons. First, emergency patients entering the emergency department lose valuable time upon arrival because the registration and prioritizing process must be finished first. Second, even if the process upon arrival is optimized, the patient still loses highly valuable early treatment time as he/she only arrives when the emergency has already happened. In other cases, a patient enters the emergency room even if he/she could have been treated by a local physician as the symptoms are not critical. On the other hand, there may be patients having a high threshold of actually contacting an ambulance or a physician in general. That is why they usually wait out until their health conditions become critical resulting in a real emergency. Furthermore, prioritizing admission/treatment is usually done at an emergency department level which not only increases the workload of the corresponding medical staff but also prevents effective proactive treatment of symptoms.
In view of the plurality of drawbacks, there is a need for improved prioritizing admission/treatment.
This need is at least partially met by the system, devices and corresponding methods of the independent claims of the present invention. A first embodiment of the present invention is a computer-implemented method for training an artificial intelligence model for predicting an urgency rating of a treatment, the method comprising: inputting training data to the artificial intelligence model to train the model, the training data including a plurality of data samples, wherein each data sample comprises: data describing a health condition of a patient; and an urgency rating of a treatment associated with the data describing the health condition of the patient.
In some embodiments, the urgency rating may be defined as either low, medium or high. Additionally, or alternatively, the urgency rating may be a value within a predefined interval/scale (e.g., on a scale from 1 to 10 where 1 is the lowest urgency rating and 10 is the highest urgency rating).
The term “urgency rating of a treatment” in the context of the present specification particularly indicates the need of the treatment and the priority of the treatment particularly compared to other patients. For example, a predicted urgency rating of a treatment having the value “low” or “zero” may be indicative for non-urgent need or no need for a treatment at all.
The artificial intelligence model trained according to this method may allow for accurate prediction of an urgency of a treatment of a patient based on his current health condition.
In particular, this may enable a system of triage based on data describing the health condition of the patients, without the patients needing to enter the hospital.
According to a second embodiment, the training data may additionally or alternatively comprise: time-series data of a patient consisting of a plurality of data describing a health condition of the patient at different points in time each associated with an urgency rating of the treatment.
Training the artificial intelligence model based on time-series data of the patient, may allow the model to predict a future health condition and/or a future urgency rating of the patient. This may allow for an increased accuracy of the urgency prediction as the model trained accordingly can analyze and predict the trend of a patient’s health condition. Therefore, a negative/positive trend can be identified way in advance improving the quality of treatment of the patient (e.g., valuable early treatment time can be won this way).
A third embodiment of the present invention is a computer-implemented method for predicting an urgency rating of a treatment of a patient, the method comprising: predicting the urgency rating of the treatment of the patient based on data describing a health condition of the patient using an artificial intelligence model, wherein the artificial intelligence model is trained according to the method of the 1st embodiment.
Providing this method may allow for accurate prediction of an urgency of a treatment of a patient based on his current health condition. For example, a patient with a non-specific chest pain after a PCI may get a higher urgency rating (rate) compared to a patient with symptoms of a cold. Accordingly, if a patient’s current health condition results in the prediction of for example a low urgency rating, the patient may not be required to visit an emergency room avoiding the inefficient usage of resources. Accordingly, the triage is already made from outside the hospital as it can be identified whether a patient requires sooner or later treatment. This allows to improve and facilitate the flow of information, scheduling of treatments and patient transfers to the hospital. As a result, the new triage system using the provided method reduces the fluctuating workloads of hospitals resulting in a better-balanced workload and avoiding unnecessary triage situations in an emergency room.
In some examples, the data describing a health condition of a patient comprise or are constituted by sensor data.
According to a fourth embodiment, the method may additionally or alternatively comprise: predicting data describing a future health condition of the patient and, optionally, the urgency rating of the treatment of the patient based on the data describing the future health condition of the patient using an artificial intelligence model, wherein the artificial intelligence model is trained according to the method of the second embodiment. In some embodiments, when the artificial intelligence model has been trained additionally or alternatively on time-series data of the patient consisting of a plurality of urgency ratings of the treatments of the patient over time, the method may predict the future urgency rating of the treatment of the patient without the step of predicting the future health condition.
Providing this method may allow for an accurate prediction considering the trend of a patient’s health condition. Therefore, a negative/positive trend can be identified way in advance improving the quality of treatment of the patient (e.g., valuable early treatment time can be won this way).
In some embodiments, the individual values of the data describing the health condition of the patient may be associated with a set of adjustable weights. These weights are adjusted according to a selected prediction metric of a set of prediction metrics before predicting the urgency rating of the treatment of the patient. Examples for such metrics may be - inter alia - a quality of life metric, a pain related metric, a lethality metric or an effort of treatment metric.
Adjusting the data describing the health condition before predicting the urgency rating may allow for increased efficiency in certain situations. For example, a rehabilitation hospital department may decide to choose the quality of life metric to determine whether a certain patient has a high demand of a corresponding treatment after a surgery to improve the healing process and to end up with a high quality of life.
A fifth embodiment of the present invention is a method for remote and/or wireless communication at a portable user equipment of a patient, the method comprising: collecting data describing a health condition of the patient from at least one data source; transmitting the collected data describing the health condition of the patient to a server; and receiving, from the server, an urgency rating of a treatment based on the health condition of the patient.
In some embodiments, the method may only be performed when executed by the patient or it may be performed automatically. In case the method is performed automatically, the method may be performed on a regular basis (e.g., once a day, once every hour, once a week). The frequency may be configurable by the patient or may depend on the current and/or past health conditions and/or urgency ratings of the patient. The method may also be performed each time a change in the data describing the health condition of the patient is detected.
In some embodiments, the method further comprises: receiving from the server or a user equipment of a health care provider, a treatment plan associated with the urgency rating. For example, if the urgency rating of a patient is low (e.g., 1 to 3) the associated treatment plan may contain a health advice (e.g., a less fat diet). Should the urgency rating be medium (e.g., 3 to 6) the associated treatment plan may contain for example a (recommendation for a) medical prescription or a scheduled appointment at a corresponding physician. The physician or the (recommendation for the) medical prescription may be selected based on the health condition of the patient (e.g., if the patient reports headache he may get a (recommendation for a) medical prescription for a headache medicine, if the patient shows light signs of heart issues the selected physician may be a cardiologist). However, should the urgency rating be high (e.g., 7 to 10), an ambulance call for transferring the patient to the hospital may be initiated and the treatment plan may contain an information for the patient about the emergency and that an ambulance is on its way.
The term “health care provider” in the context of the present specification particularly refers to a person, a group, a facility, or an organization which are involved in and contribute to health care in general or emergency management in particular. Non limiting example for health care provider include medical staffs, hospital facilities, ambulances, rescue helicopters, personal physicians and the like.
Providing this method decreases the risk of the patient entering a critical health condition as it can be monitored frequently. Moreover, the threshold of a patient to report his symptoms and to get in contact with a physician is lowered. Therefore, the overall prioritizing admission/treatment process is improved as fewer emergency situations occur and the patient will most likely not enter the emergency room unnecessarily. However, should an unforeseen accident occur, which makes a transfer to the emergency room necessary, less time is lost as the health care provider (e.g. medical staff of a hospital) is already informed about the patient’s arrival (e.g. since the server may alert the corresponding hospital accordingly). Furthermore, the health care provider may already be informed about health condition of the patient and his expected time of arrival improving the quality of treatment (e.g. since the server may alert the corresponding hospital accordingly).
A sixth embodiment of the present invention is a method for remote and/or wireless communication at a server, the method comprising: receiving, from a portable user equipment of a patient, data describing a health condition of the patient; determining an urgency rating of a treatment of the patient based on the data describing the health condition of the patient; the method further comprising at least one of the following steps: transmitting the urgency rating of the treatment to the portable user equipment; transmitting the urgency rating of the treatment and the data describing the health condition of the patient to a user equipment of a health care provider for evaluating the urgency rating.
Providing this method may improve the overall prioritizing admission/treatment process as the urgency rating of the treatment of a patient can be determined without the patient being physically present in the emergency room.
According to a seventh embodiment, determining the urgency rating of the treatment comprises: predicting the urgency rating of the treatment based on the data describing the health condition of the patient according to the method of the 3rd embodiment; additionally or alternatively, it may comprise; predicting the urgency rating of the treatment according to the method of the 4th embodiment; adding the data describing the health condition of the patient and the predicted urgency rating to time-series data of the patient indicating the health condition and associated urgency ratings of the patient over time; and retraining the artificial intelligence model according to the method of the 2nd embodiment.
Determining the urgency rating using a specifically trained artificial intelligence model increases the accuracy of the urgency rating. Therefore, the assessment of the patient’s health condition is improved.
In some embodiments, when the server has transmitted the urgency rating to the user equipment of the health care provider for evaluating the urgency rating, the method may further comprise: receiving, from the user equipment of the health care provider; an evaluated urgency rating.
In some embodiments, when the server has received the evaluated urgency rating, the method may further comprise: retraining the artificial intelligence model according to the training method of the 1st, the 2nd or the subsequent embodiments using the data describing the health condition of the patient and/or the evaluated urgency rating as an additional data sample.
In some embodiments, the server may evaluate the urgency rating additionally or alternatively to the evaluation of the health care provider, by determining data describing a health condition out of a plurality of data describing health conditions (each associated with an urgency rating) of previous treatments (from the same or other patients) similar to the data describing the health condition of the patient. The server may then compare the urgency rating associated with that similar data to the urgency rating predicted for the data describing the health condition of the patient. If the difference between both urgency ratings is acceptable (i.e., the difference is for example below a predefined threshold) the predicted urgency rating is accepted.
Providing these additional steps may increase the overall accuracy of the urgency rating prediction resulting in an improved assessment of the patient’s health condition.
According to a eighth embodiment, the method may further comprise: selecting the user equipment of the health care provider from a plurality of user equipment of health care providers prior to transmitting the urgency rating of the treatment and the data describing the health condition of the patient to the user equipment of the health care provider; wherein selecting the user equipment of the health care provider is based at least on one of: a distance between the portable user equipment of the patient and the user equipment of the health care provider, a workload of the health care provider, a treatment specialization of the health care provider and/or a treatment success rate of the health care provider for the health condition described by the data. The aforementioned treatment specialization and treatment success of the health care provider may also include or may be supplemented by or previous treatment history by the health care provider. For example, if the patient was treated at the one specific health care provider (e.g. a hospital) before, then the patient is most likely to be sent to the same health care provider by the system.
Selecting the user equipment of the health care provider based on at least one of the mentioned criteria, the treatment as well as the assessment/evaluation of a patient’s urgency level and health condition are improved. For example, in case a high urgency level is predicted (i.e., a transfer to an emergency department is required) selecting the health care provider (e.g. a corresponding hospital) closest to the patient can increase the patient’s change of survival. The several criteria may also be combined. For example, if two different user equipment of two different health care providers have the same distance to a patient, the treatment specialization of each health care provider can be taken as an additional criteria for deciding which health care provider to choose. If the health condition of the patient for example indicates a heart issue and one of the two health care providers is cardiologist, the corresponding user equipment of the cardiologist is selected. The distance may be determined based on position data, e.g. GPS data, received from the various user equipment.
A ninth embodiment of the present invention is a method for wireless communication at a user equipment of a health care provider, the method comprising: receiving, from a server, an urgency rating of a treatment and data describing a health condition of a patient; determining an evaluated urgency rating of the treatment based at least on the data describing the health condition of the patient; and transmitting the evaluated urgency rating of the treatment to the server.
In some embodiments, the health care provider may also transmit a treatment plan based on the evaluated urgency rating to the server and/or directly to the patient. Additionally or alternatively, also the evaluated urgency rating may directly be transmitted to the patient.
Providing the user equipment of the health care provider with the information about the urgency rating of the treatment of the patient as well as the corresponding data describing the health condition improves the overall treatment quality. Due to the improved information flow (i.e., the medical staff has facilitated access to the health record of the patient) the health care provider can evaluate the health condition of a patient more efficiently. As a result, fewer patients in critical condition are to be transported into the emergency room as these patients can efficiently be treated preventively by transmitting a corresponding treatment plan to the patient (directly or via the server). Furthermore, the situations in which a patient enters an emergency department without his health condition requiring it, can be reduced due to the improved treatment quality.
According to a tenth embodiment, the data describing the health condition of the patient includes at least one of: a first set of health parameters and a second set of health parameters depending on the first set of health parameters, and/or captured vital signals of the patient.
In some embodiments, the first set of health parameters may include or may come from at least one of: fever (e.g., temperature in C/F, duration of fever), pain (e.g., intensity on a scale from 1-10, duration, frequency, location), cough (dry/wet, pain, difficulty to breathe), nausea/vomiting, dizziness, blindness, numbness, deafness, rash, injury, coloration, blood pressure, pulse, flatulence, diarrhea and/or photos. The first set of health parameters or a part of these parameters may be collected based on a general questionary as for example conducted by a general practitioner (e.g. it may be shown to the patient via a user interface of the portable user equipment, upon which the patient enters the data via the user interface, e.g. a touch- screen). The first set of health parameters may be collected directly from the portable user equipment of the patient by means of sensors and/or may be requested from a database (e.g., a database from the general practitioner who has examined the patient). The first set of health parameters may also be collected by means of other suitable sensors (e.g., a fever monitor connected to the portable user equipment of the patient).
In some embodiments, the second set of health parameters depending on the first set of parameters may include or may come from at least one of: blood diagnostic (inflammation), tissue diagnostics (e.g., from an eternal lab), stool diagnostics, ultrasound examination, endoscopic images, X-Ray images, magnetic resonance imaging, long term blood pressure monitoring, pulse, fever (e.g., temperature in C/F, duration), pain (e.g., intensity on a scale from 1 to 10, duration, frequency, location), cough (dry/wet, pain, difficulty to breathe), nausea/vomiting, dizziness, blindness, numbness, deadness, rash, injury and/or coloration. The second set of heath parameters may be collected based on a specific questionary as for example conducted by specialized practitioner. The set of health parameters may depend on the first set of health parameters in a way that based on the values/result of the first health parameters the relevant parameters of the second health parameters are selected. The second set of health parameters may be defined by a specialized practitioner after having received (e.g., via communication means) the first set of health parameters. The second set of health parameters may be collected directly from the portable user equipment of the patient, e.g. by means of a user interface, one or more sensors and/or may be requested from a database (e.g., a database from external labs or specialized practitioners who for example have collected X-Ray images of the patient previously). The second set of health parameters may also be collected by means of other suitable sensors (e.g., a fever monitor connected to the portable user equipment of the patient).
In some embodiments, the captured vital signs may be collected via sensors (e.g. sensor data). These sensors may either be incorporated into the portable user equipment of the patient or may be connected or connectable with the portable user equipment. Examples for such sensors may be wearables like a smart watch, a medical device, a medical implant such as a pacemaker, and/or a step counter. For example, the captured vital signs may correspond to ECG information measured by the smart watch, a number of steps walked measured by the step counter or beats per minute measured by the pacemaker.
Specifying the data describing the health condition of the patient allows for a highly accurate presentation of the patient’s actual health condition. Therefore, the urgency rating can be predicted/assessed more precisely.
In some embodiments, the data describing the health condition of the patient may also include a GPS information about the current location of the patient collected via a GPS sensor of the portable user equipment of the patient.
Providing the GPS information about the location of the patient may increase the patient’s chance of surviving as not only the nearest health care provider (e.g. medical staff or hospital) can be identified but also the corresponding ambulance has accurate location information resulting in a reduced time of transfer. Additionally, the health care provider is informed about the patient’s emergency in advance and can make appropriate preparations.
Additionally, the GPS position of the patient may be used to extract information on physical activity. The velocity derived from the GPS characterizes the times where the patient was immobile, walking, cycling or using vehicles. Shifts in a patient’s behavior monitored over longer time period allow to detect if a patient health has improved or is deteriorated.
Which such a simple to obtain activity tracker, an Al may be trained to stabilize and improve physical activity and/or to detect suspicious deterioration of the patient in time to act before the heart is permanently damaged.
Any successful Al training resides on choosing easily accessible and measurable output parameters that sufficiently correlate with the treatment success,
An eleventh embodiment of the present invention is a server comprising means for performing the method according to any of the embodiments 6 to 8 or 10 when referred back to any of the embodiments 6 to 8.
A twelfth embodiment of the present invention is a portable user equipment of a patient comprising means for performing the method according to the 5th embodiment or 10th embodiment when referred back to the 5th embodiment.
A thirteenth embodiment of the present invention is a user equipment of a health care provider comprising means for performing the method according to the 9th embodiment or the 10th embodiment when referred back to the 9th embodiment.
The user equipment of the health care provider may design as an app, a web-based application, an application specific device, a computer program and the like.
A fourteenth embodiment of the present invention is a digital triage system for remote and/or wireless communication, wherein the system comprises: a server according to the 11th embodiment; and at least one of: a portable user equipment of a patient according to the 12th embodiment; and/or a user equipment of a health care provider according to the 13th embodiment.
A fifteenth embodiment of the present invention is a computer program comprising instructions, which when executed by a computer cause the computer to perform a method according to any of the embodiments 1 to 10.
The following figure is provided to support the understanding of the present invention:
Fig. 1 Illustration of a digital triage system in accordance with the embodiments of the present invention
The following detailed description outlines the possible exemplary embodiments of the present invention.
Fig. 1 shows a system for prioritizing admission/treatment triage 100 (i.e. digital prioritizing admission/treatment system) in accordance with embodiments of the present invention. The digital prioritizing admission/treatment system 100 comprises a server 120 and at least one portable user equipment of a patient 110 and/or at least one user equipment of a health care provider 130 in accordance with embodiments of the present invention. The digital prioritizing admission/treatment system 100 and its components (server 120, portable user equipment of a patient 110, user equipment of a health care provider 130) may perform the methods described herein with respect to the corresponding embodiments. It may also be possible that a corresponding computer program is installed on the components which comprises instructions which when executed by the computer (e.g., the processor of the component) causes the component to execute the corresponding method according to the embodiments of the present invention.
In a first step, the patient or the care giver of the patient may register via his portable user equipment (e.g., a smartphone, tablet, or a smartwatch capable of wireless communication) 110 to the digital prioritizing admission/treatment system 100. The registration may be done via an app installed on the portable user equipment of the user 110 or a webapp (e.g., hosted on the server 120). When registering to the digital prioritizing admission/treatment system 100, the patient may provide data describing his health condition to the digital triage system 100. The data may be transmitted to the server and stored into a corresponding database. The data and/or the patient identity may be anonymized. The portable user equipment of the patient 110 may further be connected to one or more data sources (e.g., external data bases, sensors like sensors incorporated in a wearable like a smartwatch, step counter, pacemaker or any other implant and/or medical device) from which the data describing the health condition of the patient may be collected and/or updated according to a configured modus (i.e., only collect data for the urgency rating prediction when executed by the patient explicitly or on a regular basis (e.g., daily) or whenever new data is available from one of the sources). For collecting the data describing the health condition of the patient, they may be asked to provide a first set of health parameters according to a more general questionary. Based on the first set of health parameters, the patient may be asked to provide a second set of health parameters. These may be more specific as for example asked by a specialized physician. Furthermore, the patient may provide information about his current medication situation.
Once registered, the data describing the health condition of the patient may be transmitted to the server 120 for determining the urgency rating of a treatment based on the health condition of the patient. The server 120 may use an artificial intelligence model trained to predict an urgency rating based on such data according to embodiments of the present invention. The artificial intelligence model may be deployed on the server 120 or at another location with the server 120 having access thereto. The determined urgency rating may then be transmitted to the portable user equipment of the patient 110. This way, the patient gets an accurate assessment of his current health condition without making a physical presence at a hospital necessary or without a need of overcoming the patient’s threshold of contacting a health care provider. It may also be possible that the server is able to determine based on data of previous patients having a similar health condition (potentially evaluated by a health care provider) and who have been treated according to a certain treatment plan (potentially created by the health care provider) a treatment plan for the patient. Additionally, or alternatively, the server 120 may transmit the urgency rating and the corresponding data describing the health condition of the patient to a user equipment of a health care provider 130. The server may determine the user equipment of the health care provider 130 to which the data is transmitted out of a plurality of available user equipment of health care provider 130 according to several criteria like distance, treatment specialization of the health care provider etc. Accordingly, the server 120 may sort the plurality of available user equipment of health care provider 130 based on the several criteria and can thus determine the ideal health care provider (e.g. medical staff or hospital) for the patient.
In cases where a plurality of portable user equipment of different patients 110 are connected to the server, the server may determine the urgency ratings from all of the patients and transmit a list ranked in descending order based on the urgency ratings to the one or more connected user equipment of health care providers .
When the user equipment of the health care provider 130 (e.g., laptop, tablet, smart phone, PC) receives the urgency rating and the data describing the health condition of the patient, the health care provider may evaluate the (correctness of the) urgency rating. The health care provider will analyze the data describing the health condition and check whether the health condition is as urgent as indicated by the rating. The evaluated urgency rating may then be transmitted back to the server 120, which may allow the server 120 to retrain the artificial intelligence model (the urgency rating may generally refer to the same treatment, but it may also be possible that an additional urgency level relating to a different treatment is transmitted back to the server 120). In some examples, such back-transmission may be omitted. When retraining the artificial intelligence model on the evaluated urgency ratings, this may allow the model to identify correlations between certain health parameters and urgent treatments. This may result in a more accurate determining of the health condition of a patient and a corresponding urgency rating. The evaluated urgency rating may also be transmitted from the user equipment of the health care provider 130 to the portable user equipment of the patient 110 either via the server 120 or directly. In addition to the rating, a treatment plan may be transmitted to the patient. In case of a high urgency rating, the treatment plan may include an ambulance transfer of the patient to the corresponding hospital or a physician directly contacting the patient. In case of a medium urgency rating, the treatment plan may include a scheduled appointment for the patient at a physician. In case of a low urgency rating, the treatment plan may only contain an advise like a healthier diet. One additional benefit of the present invention is that by monitoring and assessing the health condition of a patient, conclusions about the success of certain treatment plans can be drawn. Assuming a portable user equipment of a patient 110 transmits data describing the health condition of the patient to the server 120 which predicts a high urgency rating and the patient is transferred to the emergency department because the patient for example needs heart surgery. The system allows monitoring of the development of the health condition of the patient even after leaving the hospital or the rehabilitation after the surgery. Should the health condition described via the corresponding data transmitted to and evaluated from the server 120 indicate that the patient is feeling better (e.g., indicated by a low pain indication or other vital signs like a high number of counted steps), one can assume that the surgery has been effective. However, should the data indicate that the health condition is not increasing (as supposed to), the corresponding health care provider would notice this and could call the patient or schedule an additional appointment.
Another benefit would arise if an insurance company or the management of the health care provide would participate in the system 100 for prioritizing admission/treatment according to the invention, e.g. by getting access to aggregated information which can be obtain form the system. For example, in low load situations, the system 100 according to the invention may detect patients with susceptible symptoms earlier and would help to detect critical patients that do not consult a physician early enough to avoid grave consequences. That will avoid a number of emergency cases by preventive treatment. The management in the hospital could have an overview about the cost, admission or readmission rate and staff could plan the treatment.

Claims

Claims
1. A computer-implemented method for training an artificial intelligence model for predicting an urgency rating of a treatment, the method comprising: inputting training data to the artificial intelligence model to train the model, the training data including a plurality of data samples, wherein each data sample comprises: data describing a health condition of a patient; and an urgency rating of a treatment associated with the data describing the health condition of the patient.
2. The method of the preceding claim, wherein the training data further comprises: time-series data of a patient consisting of a plurality of data describing a health condition of the patient at different points in time each associated with an urgency rating of the treatment.
3. A computer-implemented method for predicting an urgency rating of a treatment of a patient, comprising: predicting the urgency rating of the treatment of the patient based on data describing a health condition of the patient using an artificial intelligence model, wherein the artificial intelligence model is trained according to the method of claim 1 or 2.
4. The method of the preceding claim, the method further comprising: predicting data describing a future health condition of the patient and the urgency rating of the treatment of the patient based on the data describing the future health condition of the patient using an artificial intelligence model, wherein the artificial intelligence model is trained according to the method of claim 2.
5. A method for remote and/or wireless communication at a portable user equipment of a patient (110), the method comprising: collecting data describing a health condition of the patient from at least one data source; transmitting the collected data describing the health condition of the patient to a server (120); and receiving, from the server (120), an urgency rating of a treatment based on the health condition of the patient. A method for remote and/or wireless communication at a server (120), the method comprising: receiving, from a portable user equipment of a patient (110), data describing a health condition of the patient; determining an urgency rating of a treatment of the patient based on the data describing the health condition of the patient; the method further comprising at least one of the following steps: transmitting the urgency rating of the treatment to the portable user equipment of the patient (110); transmitting the urgency rating of the treatment and the data describing the health condition of the patient to a user equipment of a health care provider (130) for evaluating the urgency rating. The method of the preceding claim, wherein determining the urgency rating of the treatment comprises: predicting the urgency rating of the treatment based on the data describing the health condition of the patient according to the method of claim 3; and/or predicting the urgency rating of the treatment according to the method of claim 4; adding the data describing the health condition of the patient and the predicted urgency rating to time-series data of the patient indicating the health condition and associated urgency ratings of the patient over time; and retraining the artificial intelligence model according to the method of claim 2. The method of any of the preceding claims 6 to 7, wherein the method comprises: selecting the user equipment of the health care provider (130) from a plurality of user equipment of health care providers (130) prior to transmitting the urgency rating of the treatment and the data describing the health condition of the patient to the user equipment of the health care provider (130); wherein selecting the user equipment of the health care provider (130) is based at least on one of: a distance between the portable user equipment of the patient (110) and the user equipment of the health care provider (130), a workload of the health care provider, a treatment specialization of the health care provider and/or a treatment success rate of the health care provider for the health condition described by the data. A method for wireless communication at a user equipment of a health care provider (130), the method comprising: receiving, from a sever (120), an urgency rating of a treatment and data describing a health condition of a patient; determining an evaluated urgency rating of the treatment based at least on the data describing the health condition of the patient; and transmitting the evaluated urgency rating of the treatment to the server (120). The method of any of the preceding claims, wherein the data describing the health condition of the patient includes at least one of: a first set of health parameters, a second set of health parameters depending on the first set of health parameters, captured vital signals of the patient. A server (120) comprising means for performing the method according to any of the claims 6 to 8 or claim 10 when referred back to any of claims 6 to 8. A portable user equipment of a patient (110) comprising means for performing the method according to claim 5 or claim 10 when referred back to claim 5. A user equipment of a health care provider (130) comprising means for performing the method according to claim 9 or claim 10 when referred back to claim 9. A digital prioritizing admission/treatment system (100) for remote and/or wireless communication, wherein the system (100) comprises: a server (120) according to claim 11; and at least one of: a portable user equipment of a patient (110) according to claim 12; a user equipment of a health care provider (130) according to claim 13. A computer program comprising instructions, which when executed by a computer cause the computer to perform a method according to any of the claims 1 to 10.
PCT/EP2023/069311 2022-08-15 2023-07-12 Digital system for prioritizing admission and/or treatment of a patient WO2024037796A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
EP22190389 2022-08-15
EP22190389.1 2022-08-15

Publications (1)

Publication Number Publication Date
WO2024037796A1 true WO2024037796A1 (en) 2024-02-22

Family

ID=82940052

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2023/069311 WO2024037796A1 (en) 2022-08-15 2023-07-12 Digital system for prioritizing admission and/or treatment of a patient

Country Status (1)

Country Link
WO (1) WO2024037796A1 (en)

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120004925A1 (en) * 2010-06-30 2012-01-05 Microsoft Corporation Health care policy development and execution
US8380631B2 (en) * 2006-07-19 2013-02-19 Mvisum, Inc. Communication of emergency medical data over a vulnerable system
US20150213225A1 (en) * 2012-09-13 2015-07-30 Parkland Center For Clinical Innovation Holistic hospital patient care and management system and method for enhanced risk stratification
CA2945131A1 (en) * 2014-04-10 2015-10-15 Parkland Center For Clinical Innovation Holistic hospital patient care and management system and method for automated resource management
US20160217260A1 (en) * 2015-01-22 2016-07-28 Koninklijke Philips N.V. System, method and computer program product for patient triage
US20170017767A1 (en) * 2014-03-13 2017-01-19 Koninklijke Philips N.V. Patient watch-dog and intervention/event timeline
US20190180868A1 (en) * 2017-12-11 2019-06-13 UBQ, Inc. Optimizing emergency department resource management
US10517479B2 (en) * 2006-06-30 2019-12-31 Koninklijke Philips N.V. Mesh network personal emergency response appliance
EP3796328A1 (en) * 2011-11-23 2021-03-24 WhenMed VC LLC. Remotely-executed medical diagnosis and therapy including emergency automation

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10517479B2 (en) * 2006-06-30 2019-12-31 Koninklijke Philips N.V. Mesh network personal emergency response appliance
US8380631B2 (en) * 2006-07-19 2013-02-19 Mvisum, Inc. Communication of emergency medical data over a vulnerable system
US20120004925A1 (en) * 2010-06-30 2012-01-05 Microsoft Corporation Health care policy development and execution
EP3796328A1 (en) * 2011-11-23 2021-03-24 WhenMed VC LLC. Remotely-executed medical diagnosis and therapy including emergency automation
US20150213225A1 (en) * 2012-09-13 2015-07-30 Parkland Center For Clinical Innovation Holistic hospital patient care and management system and method for enhanced risk stratification
US20170017767A1 (en) * 2014-03-13 2017-01-19 Koninklijke Philips N.V. Patient watch-dog and intervention/event timeline
CA2945131A1 (en) * 2014-04-10 2015-10-15 Parkland Center For Clinical Innovation Holistic hospital patient care and management system and method for automated resource management
US20160217260A1 (en) * 2015-01-22 2016-07-28 Koninklijke Philips N.V. System, method and computer program product for patient triage
US20190180868A1 (en) * 2017-12-11 2019-06-13 UBQ, Inc. Optimizing emergency department resource management

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
NAPI N M ET AL: "Medical emergency triage and patient prioritisation in a telemedicine environment: a systematic review", HEALTH AND TECHNOLOGY, SPRINGER BERLIN HEIDELBERG, BERLIN/HEIDELBERG, vol. 9, no. 5, 11 August 2019 (2019-08-11), pages 679 - 700, XP036936192, ISSN: 2190-7188, [retrieved on 20190811], DOI: 10.1007/S12553-019-00357-W *
SALMAN OMAR H ET AL: "A review on utilizing machine learning technology in the fields of electronic emergency triage and patient priority systems in telemedicine: Coherent taxonomy, motivations, open research challenges and recommendations for intelligent future work", COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE, ELSEVIER, AMSTERDAM, NL, vol. 209, 16 August 2021 (2021-08-16), XP086768187, ISSN: 0169-2607, [retrieved on 20210816], DOI: 10.1016/J.CMPB.2021.106357 *

Similar Documents

Publication Publication Date Title
US11810657B2 (en) System and method for tracking patient recovery following orthopedic procedure
Sacco et al. Precise formulation and evidence‐based application of resource‐constrained triage
DE112006003199T5 (en) Non-obstructive, substantially continuous acquisition of the daily activities of a patient to indicate a change of state of the patient for the access of a remote operator
KR102543793B1 (en) Method and system for improving a physiological response
KR20150025629A (en) Apparatus for providing customized medical data to patients
Maningas et al. The Soterion Rapid Triage System: evaluation of inter-rater reliability and validity
Orwat et al. Adopting pervasive computing for routine use in healthcare
Lee et al. Developing risk‐adjusted quality indicators for pressure ulcers in long‐term care hospitals in the Republic of Korea
Visvesvaran et al. Quarcare-IoT based patient health monitoring system
US8428965B2 (en) System for clinical research and clinical management of cardiovascular risk using ambulatory blood pressure monitoring and actigraphy
CN117219243A (en) Cardiovascular disease assessment and management system
CN109979570B (en) First-aid handheld diagnosis equipment with X-ray detection function and diagnosis method
WO2024037796A1 (en) Digital system for prioritizing admission and/or treatment of a patient
Williams et al. Retrospective validation of a protocol to limit unnecessary transport of assisted-living residents who fall
Panicacci et al. Satnav e@ sycare telemedicine platform in the management of covid-19 patients: Field trial results
Berlowitz et al. Ambulatory care casemix measures
US20200381118A1 (en) Artificial intelligence (ai) system for individual health maintenance recommendations
CN104699932A (en) Remote health care system and method thereof
White Remote patient monitoring: Prevention of DFUs
KR102594948B1 (en) Metabolic syndrome ai primary doctor system
US20220199235A1 (en) Multi-Sensor Platform for Health Monitoring
Huft et al. Care of the pregnant offender
WO2023042971A1 (en) Method for predicting and managing menstrual disorder, and digital therapeutic agent
RU2011120069A (en) METHOD FOR PRESERVING A HEALTHY, LIFE-STATE OF A HUMAN AND EXTENDING ITS LIFE AND A DIAGNOSTIC SYSTEM IN ITS STAFF
EP3981322A1 (en) Automatically identifying pressure injuries

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23738804

Country of ref document: EP

Kind code of ref document: A1