Consumables Availability Forecast for Point-of-Care Medical Analyzers

Abstract

The present application describes a device for generating predicted consumable status data for a point-of-care medical analyzer, comprising an input unit and a processing unit. The input unit is configured to receive initial consumable status data for the point-of-care medical analyzer, the initial consumable status data defining an initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer, and receive an analyzer identifier at a first time index. The input unit is configured to receive a second time index. The processing unit is configured to generate predicted consumable status data using a predicted consumable consumption model selected using at least the analyzer identifier and the initial consumable status data, the predicted consumable status data defining an amount of point-of-care medical analyzer consumables predicted to be present in the point-of-care medical analyzer at a second time index. The predicted consumable consumption model includes (i) a consumable model for the point-of-care medical analyzer and (ii) a point-of-care medical analyzer usage model.

Description

Technical Field

[0001] The present invention generally relates to an apparatus for generating predictive consumable status data for a point-of-care medical analyzer, a method for generating predictive consumable status data for a point-of-care medical analyzer, a system, a computer program element, and a computer-readable medium. Background Art

[0002] Automated point-of-care medical analyzers have almost replaced manual testing operations in hospitals, diagnostic testing centers, and point-of-care (POC). For example, such devices can be used for blood gas testing, urea testing, diabetes management, and hematology.

[0003] For example, a typical high-performance blood gas analyzer can test a variety of electrolytes, metabolites and biomarkers, such as pCO2, pO2, pH, Na+, K+, Ca2+, Cl-, glucose, lactate and hemoglobin, D-dimer, troponin, CKMB, myoglobin, NT-proBNP, CRP, phCG, PCT, etc. A typical automated blood gas analyzer can achieve a throughput of about 30 samples per hour, where multiple blood tests are performed in a single sample, providing a total of about 300 tests per hour.

[0004] Such automated point-of-care medical analyzers require a range of reagents. Typically, in a diagnostic testing center or large hospital, there are a large number and variety of automated point-of-care medical analyzers, making effective management of reagents complicated.

[0005] In some cases, one or more personnel may be assigned the task of keeping the point-of-care testing equipment up and running. These personnel are responsible for monitoring whether the equipment is always calibrated, whether the predetermined quality controls are performed, and whether there are enough consumables. These point-of-care coordinators do not use the analyzers frequently and therefore do not know the hour-to-hour status of the consumables. Therefore, they often rely on the input of the operator, who is usually a nurse. In order to ensure that the point-of-care medical analyzer is always up and running, the nurse may call the point-of-care coordinator in advance before the analyzer consumables run out to ensure that the consumables are replaced and the point-of-care testing equipment is always available. This often results in the replacement of consumables before they need to be replaced, thus resulting in waste.

[0006] US 2003 / 0172009A1 discusses a consumables supply management system in which clinical testing and analysis equipment in a medical institution and a computer of a consumables supplier are connected via a communication line. However, this system can be further improved.

[0007] Therefore, there is a need for an improved method of reagent management for an automated point-of-care medical analyzer. Summary of the invention

[0008] It would be advantageous to have an improved technique for point-of-care medical analyzer consumable availability prediction.The objects of the invention are solved by the subject matter of the independent claims, wherein further embodiments are contained in the dependent claims.

[0009] An apparatus for generating predictive consumable status data for a point-of-care medical analyzer, comprising:

[0010] - input unit; and

[0011] -processor.

[0012] The input unit is configured to receive initial consumable status data for the point-of-care medical analyzer, the initial consumable status data defining an initial amount of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer, and an analyzer identifier at a first time index.

[0013] The input unit is configured to receive a second time index.

[0014] The processing unit is configured to generate predicted consumable status data using a predicted consumable consumption model selected using at least the analyzer identifier and the initial consumable status data, the predicted consumable status data defining an amount of the point-of-care medical analyzer consumable predicted to be present in the point-of-care medical analyzer at a second time index.

[0015] The predictive consumables consumption model includes (i) a consumables model for a point-of-care medical analyzer, which defines the consumable depletion rate of point-of-care medical analyzer consumables used for testing in the point-of-care medical analyzer, and (ii) a point-of-care medical analyzer usage model, which defines the incidence rate of testing.

[0016] Predicted consumable status data is generated by a processing unit based on an initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer, an aggregation of at least an output of a consumable model and a point-of-care medical analyzer usage model between a first time index and a second time index, and a location of the point-of-care medical analyzer.

[0017] Advantageously, a predicted future consumable status of one or more consumables in a point-of-care medical analyzer may be quickly and automatically generated taking into account the complexity of an individual testing protocol and data regarding the frequency of use of the point-of-care medical analyzer.

[0018] Optionally, the apparatus is configured to receive historical usage data for a point-of-care medical analyzer defined by an analyzer identifier.

[0019] Advantageously, a point-of-care medical analyzer or a networked computer in communication with the point-of-care medical analyzer can provide predictive consumable status data with improved accuracy because consumable levels at a certain historical month, week, day, or even time can be used to improve the accuracy of the predictive consumable status data.

[0020] Optionally, the apparatus is configured to receive a consumable usage rule set defining point-of-care medical analyzer consumables consumption over time.

[0021] Advantageously, forward-looking prediction rule sets defining consumable usage can be used to improve the accuracy of predictive consumable status data. For example, a rule set can be provided for the preset of a specific blood gas analyzer, which specifies that a greater number of blood gas tests are performed every Friday according to the hospital's regular schedule.

[0022] Optionally, the apparatus is configured to receive a consumable capacity metric for a point-of-care medical analyzer consumable using the analyzer identifier.

[0023] Optionally, the device is configured to generate a consumable model based on at least one of historical usage data, a consumable usage rule set, a consumable handling model for a point-of-care medical analyzer, and a consumable capacity metric for a point-of-care medical analyzer identified by an analyzer identifier and run from initial consumable status data collected at a first time index.

[0024] Advantageously, the combination of several information sources enables a more accurate assessment of the amount of point-of-care medical analyzer consumable remaining in the point-of-care medical analyzer at the second time index.

[0025] Optionally, the consumables model for the point-of-care medical analyzer includes a linear model, a rule-based machine learning system, experimental fitting data, and / or a Bayesian model.

[0026] Advantageously, a wide range of appropriate data analysis models can be provided to fit the network of point-of-care medical analyzers employed. For example, a hospital department with a relatively constant caseload (in other words, with a well-defined schedule for testing) may give priority to rule-based models or linear models. For example, a large hospital may use experimentally fitted data and / or Bayesian models to define the incidence of medical tests caused by seasonally occurring diseases.

[0027] Optionally, the apparatus is configured to receive usage pattern data obtained prior to the first time index for a point-of-care medical analyzer defined by the analyzer identifier.

[0028] Advantageously, a historical model of consumable consumption may be generated from usage pattern data that accurately characterizes the historical usage of a particular point-of-care medical analyzer installed in a hospital or clinic.

[0029] Optionally, the apparatus is configured to receive usage location metadata of a point-of-care medical analyzer defined by an analyzer identifier.

[0030] Optionally, the apparatus is configured to receive a future usage plan of the point-of-care medical analyzer defined by the analyzer identifier between the first time index and the second time index.

[0031] Optionally, the apparatus is configured to generate a point-of-care medical analyzer usage model based on at least one of the usage pattern data, usage location metadata, and future usage plans.

[0032] Optionally, the point-of-care medical analyzer uses the model to output an expected number of tests between the first and second time indexes.

[0033] Optionally, the outputs of at least the consumables model and the point-of-care medical analyzer usage model aggregated between the first time index and the second time index include calculating the test consumption of point-of-care medical analyzer consumables used for testing in the point-of-care medical analyzer, and multiplying the test consumption by the expected number of tests to generate predicted consumables status data at the second time index, and / or as a variable between the first time index and the second time index.

[0034] Advantageously, a user can define a time window during which the consumables present in the point-of-care medical analyzer and the consumption of the consumables need to be predicted. For example, a point-of-care medical analyzer technician can define a time window corresponding to their vacation. The usage model will provide the expected number of tests between the first and second time indexes, while the point-of-care medical analyzer technician is away. Conversely, it can be predicted at what time index the point-of-care medical analyzer becomes completely depleted of important reagents under the care of the point-of-care medical analyzer technician. The point-of-care medical analyzer technician is able to predict when a given reagent needs to be refilled before going on vacation.

[0035] Optionally, the apparatus is configured to receive expiration date data for point-of-care medical analyzer consumables and factor the point-of-care medical analyzer consumables into the consumables model only between the first time index and an expiration date defined in the expiration date data.

[0036] Optionally, the device further comprises:

[0037] - A data output unit. The data output unit is configured to output the predicted consumable status data at the second time index, and / or as a variable between the first time index and the second time index.

[0038] Optionally, the device also includes an output display.

[0039] The output display is configured to display a GUI that includes a representation of the point-of-care medical analyzer and an initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer, to receive a second time index from a user via the GUI interface, and to display on the GUI an indication of how urgently replacement of the point-of-care medical analyzer consumables present in the point-of-care medical analyzer at the second time index is required.

[0040] Advantageously, a user can quickly and simply enter a duration during which, for example, one or more point-of-care medical analyzers are used without maintenance personnel. A rapidly generated GUI enables a manager to assess the remaining capacity of one or more point-of-care medical analyzers to provide a series of tests. In addition, the GUI can use the modeling described herein to provide predictions for various configuration arrangements of the medical analysis system. For example, a user can investigate whether using a first and second point-of-care medical analyzer offline can keep a certain point-of-care medical analyzer running longer with remaining reagent inventory.

[0041] Optionally, the device is configured to display multiple representations on the GUI, each representation corresponding to a unique point-of-care medical analyzer among a plurality of point-of-care medical analyzers, configured to receive a second time index from a user via the GUI interface, and configured to sort the multiple representations in the GUI based on a level of urgency for replacement of point-of-care medical analyzer consumables for each of the plurality of point-of-care medical analyzers.

[0042] Optionally, the apparatus is configured to generate and display a list of consumables and / or quantities of consumables for replenishing point-of-care medical analyzers in the plurality of point-of-care medical analyzers.

[0043] Optionally, the device is a blood gas analyzer.

[0044] According to a second aspect, there is provided a method for generating predicted consumable status data for a point-of-care medical analyzer, comprising:

[0045] receiving initial consumable status data for a point-of-care medical analyzer, the initial consumable status data defining an initial amount of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer, and receiving an analyzer identifier at a first time index;

[0046] receiving a second time index;

[0047] generating, using a predictive consumable consumption model selected using at least the analyzer identifier and the initial consumable status data, predictive consumable status data defining an amount of a point-of-care medical analyzer consumable predicted to be present in the point-of-care medical analyzer at a second time index;

[0048] wherein the predictive consumables consumption model comprises (i) a consumables model for a point-of-care medical analyzer, the consumables model defining a consumables consumption rate of point-of-care medical analyzer consumables used for testing in the point-of-care medical analyzer, and (ii) a point-of-care medical analyzer usage model, the point-of-care medical analyzer usage model defining an incidence rate of testing; and

[0049] Wherein the predicted consumable status data is generated based on an initial amount of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer and an aggregation of at least outputs of a consumable model and a point-of-care medical analyzer usage model between a first time index and a second time index.

[0050] Optionally, the method includes receiving historical usage data for a point-of-care medical analyzer defined by an analyzer identifier.

[0051] Optionally, the method includes receiving a consumable usage rule set that defines point-of-care medical analyzer consumable usage over time.

[0052] Optionally, the method includes receiving a consumable capacity metric for a point-of-care medical analyzer consumable using the analyzer identifier.

[0053] Optionally, the method includes generating a consumable model based on at least one of historical usage data, a consumable usage rule set, a consumable handling model for a point-of-care medical analyzer, and a consumable capacity metric for a point-of-care medical analyzer identified by an analyzer identifier and run from initial consumable status data collected at a first time index.

[0054] Optionally, in the method, the consumables model for the point-of-care medical analyzer comprises a linear model, a rule-based machine learning system, experimental fitting data, and / or a Bayesian model.

[0055] Optionally, the method includes receiving usage pattern data acquired prior to the first time index for a point-of-care medical analyzer defined by the analyzer identifier.

[0056] Optionally, the method includes receiving usage location metadata for a point-of-care medical analyzer defined by an analyzer identifier.

[0057] Optionally, the method includes receiving a future usage plan of the point-of-care medical analyzer defined by the analyzer identifier between the first time index and the second time index.

[0058] Optionally, the method includes generating a point-of-care medical analyzer usage model based on at least one of the usage pattern data, the usage location metadata, and a future usage plan.

[0059] Optionally, in the method, the point-of-care medical analyzer uses the model to output an expected number of tests between the first and second time indexes.

[0060] Optionally, in the method, aggregating outputs of at least the consumables model and the point-of-care medical analyzer usage model between the first time index and the second time index comprises:

[0061] Calculate test consumption of point-of-care medical analyzer consumables used for testing in a point-of-care medical analyzer and multiply the test consumption by the expected number of tests to generate predicted consumable status data at a second time index and / or as a variable between the first time index and the second time index.

[0062] Optionally, the method wherein receiving expiration date data for a point-of-care medical analyzer comprises factoring point-of-care medical analyzer consumables into the consumables model only between the first time index and an expiration date defined in the expiration date data.

[0063] Optionally, the method includes outputting predicted consumable status data at a second time index, and / or as a variable between the first time index and the second time index.

[0064] Optionally, the method includes displaying a GUI including a representation of the point-of-care medical analyzer and an initial amount of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer.

[0065] Optionally, the method includes displaying on the GUI an indication of how urgently replacement of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer at the second time index is required.

[0066] Optionally, the method includes displaying multiple representations on the GUI, each representation corresponding to a unique point-of-care medical analyzer among a plurality of point-of-care medical analyzers, receiving a second time index from a user via the GUI interface, and sorting the multiple representations in the GUI based on an urgency level for replacement of a point-of-care medical analyzer consumable for each of the plurality of point-of-care medical analyzers.

[0067] Optionally, the method includes generating a list of consumables and / or quantities of consumables for replenishing point-of-care medical analyzers in the plurality of point-of-care medical analyzers.

[0068] According to a third aspect, there is provided a system comprising:

[0069] -Point-of-care medical analyzers;

[0070] - an apparatus according to the first aspect or one of its embodiments; and

[0071] Data communication network.

[0072] The point of care medical analyzer is configured to transmit initial consumable status data of the point of care medical analyzer defining an initial amount of a point of care medical analyzer consumable present in the point of care medical analyzer and an analyzer identifier at a first time index to the device via the data communications network.

[0073] Optionally, the system comprises:

[0074] -Mobile data terminal.

[0075] The mobile computer is configured as:

[0076] displaying a GUI including a representation of a point-of-care medical analyzer and an initial amount of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer;

[0077] sending a second time index to the device via the GUI interface via the data network;

[0078] receiving response data from the device; and

[0079] An indication of how urgently replacement of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer at the second time index is displayed on the GUI of the mobile device.

[0080] Optionally, a system is provided wherein at least one point-of-care medical analyzer is configured to identify an installed point-of-care medical analyzer consumable using a point-of-care medical analyzer consumable identifier, and the at least one point-of-care medical analyzer is configured to transmit, via a data communications network, an intermediate quantity of a point-of-care medical analyzer consumable present in the installed point-of-care medical analyzer consumable in association with the point-of-care medical analyzer consumable identifier.

[0081] According to a fourth aspect, there is provided a computer program element comprising instructions which, when executed by a computer, enable the computer to perform the method according to the second aspect and embodiments thereof.

[0082] According to a fifth aspect, there is provided a computer program product tangibly embodied on a carrier medium, comprising software code for performing the method according to the second aspect.

[0083] Point-of-care testing (POCT or bedside testing) is defined as medical diagnostic testing performed at or near the point of care, at the time and place of patient care. This is in stark contrast to the historical model, in which testing was confined entirely or largely to a medical laboratory, requiring samples to be sent from the point of care and then waiting hours or days for results, during which time care had to continue without the needed information.

[0084] The driving concept behind POCT is to bring testing to patients quickly and easily. This increases the likelihood that patients, physicians, and care teams will receive results faster, allowing for better clinical management decisions at the point of care. POCT includes: blood glucose testing, blood gas and electrolyte analysis, rapid coagulation testing, rapid cardiac marker diagnostics, drug abuse screening, urine testing, pregnancy testing, fecal occult blood analysis, food pathogen screening, hemoglobin diagnostics, infectious disease testing, and cholesterol screening.

[0085] POCT is typically accomplished through the use of deliverable, portable, and handheld instruments. When handheld devices are not available, small benchtop analyzers or fixed equipment can also be used - the goal is to collect samples and obtain results in a very short time at or near the patient's location so that the treatment plan can be adjusted as needed before the patient leaves. Cheaper, faster, and smarter POCT devices have increased the use of POCT methods by making them cost-effective for many diseases. In addition, it is highly desirable to measure multiple analytes simultaneously in the same sample, thereby achieving rapid, low-cost, and reliable quantification. Therefore, multiplexed point-of-care testing (xPOCT) has become more important for medical diagnostics over the past decade.

[0086] In the following applications, the term "predicted consumable state data" is a data structure that can be stored in volatile and / or non-volatile storage media (such as computer RAM or hard drive, respectively), which defines, at a given time index, the predicted consumption level of one or more reagents remaining within a point-of-care medical analyzer, taking into account the number of medical tests performed within a time range prior to the given time index, and the amount of reagents used for a particular type of test.

[0087] For example, the data structure may include a numeric element as a fixed or floating point number providing the remaining fluid volume of each reagent in the analyzer in milliliters. Alternatively, the data structure may include a numeric element as a floating point number or integer representing the percentage of reagent remaining.

[0088] Furthermore, the predicted consumable status data may be a single value (providing a snapshot of reagent consumption at one time), or the predicted consumable status data may be an N-dimensional state vector defining the predicted remaining reagent quantities for each reagent type used in the analyzer at multiple time points in the future.

[0089] Optionally, one or more numerical elements of the predicted consumable status data may include a statistical measure, such as the likelihood of reaching a given volume level at a specific time index. Optionally, one or more numerical elements of the predicted consumable status data may include a predicted range of volumes that can be obtained with a given probability. For example, the predicted consumable status data additionally includes a volume range spanning an average predicted consumable level, wherein the volume range defines a 5%, 10%, 20%, 30%, 40%, or 50% likelihood of reaching the lower and / or upper portion of the volume range.

[0090] In the following application, the term "initial consumable status data" means automatically digitized fluid volume or reagent quantity data obtained by an electronic measuring device of a point-of-care medical analyzer, such as a photodiode or phototransistor level monitoring device, or an ultrasonic fluid level monitoring device. The initial consumable status data may be automatically transmitted by the point-of-care medical analyzer, or a device monitoring the point-of-care medical analyzer may "poll" the point-of-care medical analyzer for level information at will.

[0091] In the following applications, the term "analyzer identifier" refers to a code or serial number that uniquely identifies a point-of-care medical analyzer on a computer network that includes multiple point-of-care medical analyzers, such that initial consumable status data read from the uniquely identified point-of-care medical analyzer can be used to predict future reagent status for the uniquely identified point-of-care medical analyzer.

[0092] In the following application, the term "time index" refers to a data object that can define the status of a consumable, for example, in terms of time. Many different types of time formats can be used, with a resolution of seconds, minutes, hours, days, or weeks, depending on the application of the point-of-care medical analyzer.

[0093] In the following applications, the term "predicted consumable depletion model" refers to a computer-implemented algorithm for predicting the consumption of one or more reagents in a unique point-of-care medical analyzer over a time range. The "predicted consumable depletion model" includes at least two elements: a "consumable model" that defines the amount of one or more reagents consumed (expend) during the implementation of a defined test; and a "point-of-care medical analyzer usage model" that defines how many tests occur between a first and a second time index.

[0094] In the simplest case, the "consumables model" is parameterized by the number of tests of a particular type that are formed. For a given number of tests performed, the output of the "consumables model" defines the total consumption of one or more reagents (e.g., in milliliters), which are defined by the protocol required to implement a specific test (such as, for example, tests for pCO2, pO2, pH, Na+, K+, Ca2+, Cl-, glucose, lactate, or hemoglobin, etc.). The consumables model can optionally take into account changes in specific tests (in particular, it can take into account, for example, the volumes of different reagents required for different volumes of blood). Alternatively, the consumables model can statistically define the volume of the delivered reagent, in other words, by providing the expected volume of the reagent as the mean of a probability distribution (such as a normal distribution) and defining the change in the amount of the reagent delivered to the patient according to the standard deviation of the normal distribution. It should be understood that modern point-of-care medical analyzers have a certain degree of accuracy in their ability to monitor the volume of the delivered reagent. However, with frequent and prolonged use of a point-of-care medical analyzer, even small deviations from the expected average amount of reagent delivered per test can result in significant errors between the expected and actual amounts of reagent remaining in the point-of-care medical analyzer.

[0095] Optionally, the "consumables handling model for a point-of-care medical analyzer" will include a model of the volume of the fluid handling system of the point-of-care medical analyzer, including data from a liquid delivery system showing how many mL of liquid have been drawn from a bag. The liquid delivery system includes pumps, valves controlling which bag liquid is drawn from and thus which liquid is introduced into the measurement chamber.

[0096] To take into account the situation where the same point-of-care medical analyzer can perform multiple different tests, the term "consumable model" is optionally a combination of multiple "consumable sub-models", where each consumable sub-model represents one of multiple unique tests that can be performed on the same point-of-care medical analyzer.

[0097] A "point of care medical analyzer usage model" refers to a computer in a modified algorithm that defines the usage of a given type of point of care medical analyzer during the time between a first and a second time index. The point of care medical analyzer usage model provides an accurate representation of the number and type of tests applied using a given point of care medical analyzer. Of course, the usage of a single point of care medical analyzer can be defined by a wide range of factors, some of which can be listed as follows without limitation: whether a given point of care medical analyzer is located in a professional medical testing center, a large point of care location (such as a hospital) or a small point of care location (such as a doctor's surgery (point of care medical analyzer location)) will reflect the usage rate of the point of care medical analyzer. The department in which the point of care medical analyzer is placed, and whether it is placed with other point of care medical analyzers of the same type, will have an impact on the intensity of use. The use of a point of care medical analyzer may vary according to different parts of the day, week, month, or even year. For example, a medical testing center can be configured to analyze a large number of samples according to a set schedule (e.g., a given number of blood gas tests are performed on every repeated Friday). Optionally, these trends can be pre-programmed. Optionally, such trends may be observed using simple straight-line statistical models or more advanced Bayesian inference or machine learning. In summary, the point-of-care medical analyzer usage model optionally provides a statistical inference of the number and type of tests applied using a given point-of-care medical analyzer between the first and second time indexes. Optionally, the point-of-care medical analyzer usage model may apply historical point-of-care medical analyzer usage records under the assumption that future usage will be similar to past usage.

[0098] In the following application, the term "aggregation" refers to the combination of "initial consumable status data", and the respective outputs of the "consumable model" and the "point-of-care medical analyzer usage model" to provide an absolute or statistically qualified estimate of one or more reagents remaining in a given point-of-care medical analyzer at a second time index.

[0099] In a simple case, the user wants to know the amount of distilled water remaining in the point-of-care medical analyzer after one week of use. The user can enter a specific future date as the first and second time indexes into the "predicted consumables consumption model". The point-of-care medical analyzer will be asked to display the remaining 1000mL of distilled water (initial consumables status data). The point-of-care medical analyzer usage model will be defined between the first and second time indexes (such as one week), and according to, for example, historical usage data, 10 tests of a given protocol requiring distilled water are required. The "consumables model" will show that for a given protocol, each iteration of the protocol requires 20mL of distilled water. Therefore, in this case, the aggregation will be the total number of tests multiplied by the total amount of distilled water required for each test. Since the results of 10 tests will still leave enough distilled water for future tests, the predicted consumables status data output will reflect the final amount. Those skilled in the art will understand that the previous example is a basic example, and many analytical protocols require 2 to 10 reagents, all of which have different consumption rates, which are difficult to track. For example, in the case of a consumables model and a point-of-care medical analyzer usage model defined in terms of continuous random variables, the aggregation process is more complex than the multiplication discussed above, and Bayesian inference defining "predicted consumables status data" can be applied, for example, based on prior knowledge of the usage of point-of-care medical analyzers.

[0100] In the following applications, the term "historical usage data" refers to recorded reagent consumption trends that are related to time for a uniquely identified point-of-care medical analyzer, or to a type of point-of-care medical analyzer that is similar to the uniquely identified analyzer.

[0101] In the following application, the term "consumable usage rule set" defines the volume of one or more consumables (reagents) used for one or more tests. More complex rule sets may define statistical changes in consumable usage for a large number of tests.

[0102] In the following application, the term "usage location data" may define the department or hospital in which a point-of-care medical analyzer is used. The consumables model may optionally associate the "usage location data" with historically stored reagent consumables models so that when the analyzer is relocated within the point-of-care, the predicted consumables status data remains accurate.

[0103] Additionally, the location of use is important when the point-of-care coordinator should replace supplies or replenish local supplies storage.

[0104] Since point-of-care medical analyzers can be moved around, the location of the analyzer can be dynamically derived in the system based on well-known positioning methods. Alternatively, the location can be manually entered into the system as a specific department of the hospital or a specific floor of the hospital, which may not give the exact location, but a defined area to look for it.

[0105] In the following application, the term "future usage plan" is a data structure that defines the number and type of tests performed at least between the first and second time indexes. The data is preferably extracted from an automated hospital control system (e.g., running on a hospital server). In other words, when a healthcare professional requires an appointment for a large number of patients to perform various medical examinations (such as blood tests), the point-of-care medical analyzer usage model is updated based on future usage plan data. Optionally, in this case, the predictive consumables consumption model reduces the statistical significance of the predictive point-of-care medical analyzer usage model and emphasizes the future usage plan data in the model because the future usage plan data is a more reliable source of information about the use of the point-of-care medical analyzer. Optionally, the future usage plan can be continuously updated from a wide range of locations, such as multiple minor surgeries and / or multiple hospitals. Therefore, a more accurate estimate of the predictive consumables status data can be provided. Therefore, the basic idea of ​​the present invention is to aggregate information related to the predicted number and type of activities on a specific point-of-care medical analyzer using a model of how an activity on the analyzer is converted into consumables usage. This enables prediction of future reagent levels within the point-of-care medical analyzer and optionally predicts the probability of the prediction being correct. These predictions and probabilities can be presented to the user in a clear and user-friendly manner on a user interface (GUI). BRIEF DESCRIPTION OF THE DRAWINGS

[0106] Exemplary embodiments of the invention will be described in the following schematic drawings, which are not presented to scale.

[0107] Figure 1a ) schematically shows a medical diagnostic device.

[0108] Figure 1b ) schematically shows a functional diagram of a medical diagnostic device.

[0109] Figure 2 A system according to the third aspect is schematically shown.

[0110] Figure 3 An apparatus according to the first aspect is schematically shown.

[0111] Figure 4 A method according to the second aspect is schematically illustrated.

[0112] Figure 5 Examples of applicable data structures are schematically shown.

[0113] Figure 6 An algorithm for predicting the status of consumables is schematically shown.

[0114] Figure 7a) shows a variation of the user interface (GUI).

[0115] Figure 7b ) shows another variation of the GUI.

[0116] Figure 7c ) shows yet another variation of the GUI. DETAILED DESCRIPTION

[0117] Figure 1a ) shows a medical analysis device 10a. The medical analysis device 10a includes a housing 12 having an inlet 14, the inlet 14 being suitable for receiving a test container, such as a vacuum container or a capillary tube (not shown), containing a body fluid sample (such as a blood sample) obtained from a patient. Optionally, the inlet can allow the sampled blood to be injected directly from a test syringe. The medical analysis device 10a also includes a display screen 16a for showing, for example, the progress of a blood gas test. The various reagent fluids used in the test are provided by replaceable cartridges, sealed bags or bottles (not shown). Another example of consumables used by the medical analysis device 10a is a "sensor cassette" of a replaceable functional sensor, which is used to contact a sample from a patient. Typically, the sensor of the "sensor cassette" includes a sensor containing a rare metal alloy (such as vanadium bronze) and is functionalized with a membrane containing an active ingredient. Eventually, after thousands of uses, for example, such a "sensor cassette" may also need to be replaced.

[0118] In use, a user of the medical analysis device 10a inserts a test container containing a blood sample (such as a whole blood sample) into the inlet 14 and initiates an automatic test. During the test process, a certain proportion of the reagent fluid is exhausted from its replaceable cartridges, sealed bags and / or bottles. For example, after thousands of tests, the sensors in the sensor cartridge will be exhausted. Eventually, the medical analysis device will reach a point where it cannot perform further tests (or only a subset of the tests) before the reagent fluid is replenished or the sensor cartridge is exhausted.

[0119] Figure 1b ) schematically shows Figure 1a ) is a system diagram of a medical analysis device 10a shown in FIG. In particular, the medical analysis device 10b includes an embedded computer 18, a testing subsystem 20, an external communication connection 22, a system power supply 24, and a display 16b.

[0120] The analysis subsystem 20 includes a receptacle 26 in mechanical communication with the inlet 14, which is capable of receiving a test container of an appropriate shape containing a fluid patient sample (such as blood or plasma). The analysis subsystem 20 also includes a reagent fluid supply 28. Optionally, the reagent fluid supply 28 is one or more replaceable cartridges, or a sealed bag (solution bag) that can be engaged with the receptacle.

[0121] Optionally, the replaceable cartridge includes a data storage device that can be read by the processor 18 of the fluid analyzer 10. For example, the data storage device may contain production date data, expiration date data, and production volume of the liquid in the bag, and / or other metadata identifying the reagent. Optionally, the data storage device may include a cartridge identifier so that the fluid analyzer 10 can query a database via a data communication network to obtain information about the reagent contained in the cartridge. Optionally, the data storage device is an optically readable indicator, such as a QR code and / or a bar code. Optionally, the data storage device is an electrically readable indicator, such as an RFID tag or an electrically connectable programmable memory element.

[0122] The reagent fluid supply 28 in a replaceable cartridge or sealed bag may include one or more reservoirs pre-filled with process liquids of known composition (such as those known to those skilled in the art: QC1, QC2, QC3, CAL3 (S1940), CAL2 (S1930), RINSE / CAL1 (S1920)). Those skilled in the art will appreciate that other chemistries may be provided depending on the exact assay desired.

[0123] This replaceable cartridge (not shown) can be equipped with for example a silicone rubber nipple, so that when the bottom seal of the receptacle is engaged, the replaceable cartridge is fluidically communicated with the receptacle 26, so that the reagent fluid supply 28 included in the replaceable cartridge can flow into the receptacle 26. Or, a transmembrane needle can be used to approach the reagent. Alternatively, the reagent fluid supply 28 comprises a fixed container with a fluid aspiration assembly. Those skilled in the art will appreciate that the network of fluid conduits, pumps, etc. enables the reagent fluid supply 28 to be fluidically communicated with the receptacle.

[0124] The components of the analysis subsystem 20 are connected using a liquid transport system (not shown). In practice, a subsample of a patient sample having a volume of several mL and / or reagents from a reagent fluid supply 28 are transported around the conduits of the analysis subsystem 20 separated by air bubbles. Sensors contained in the liquid transport subsystem detect and control the sequence of movement of the subsample and reagents around the liquid transport subsystem to enable sample analysis to proceed.

[0125] The analysis subsystem 20 also includes a test detector 30 (such as vanadium bronze supplemented with functionalized enzymes). The purpose of this unit is to observe the results of the patient sample contacting the test detector 30, thereby presenting the results of the analysis. For example, the test detector may include a sensor cartridge - a mechanical array of multiple sensors (usually made of, for example, functionalized vanadium bronze) that are typically exposed to patient blood samples and reagents thousands of times before being exhausted.

[0126] After the subsample of the patient sample is analyzed, a liquid delivery system (not shown) delivers the subsample of the patient sample and / or the spent reagent to a waste collection container (not shown). Optionally, this collection vessel can be another container, such as a separate bag, which is a modular part of a reagent solution package.

[0127] The analysis subsystem 20 also includes a reagent fluid supply monitoring unit 32. The purpose of this unit is to monitor the level of reagent fluid contained within the reagent fluid supply 28 (whether a replaceable cartridge, bottle or other type of receptacle).

[0128] Alternatively, the reagent fluid supply monitoring unit 32 can use the fluid level monitoring of the reagent receiver to monitor the level drop of the reagent fluid supply 28 when the machine is used. For example, a photodiode or a phototransistor can be used. In addition, an ultrasonic radar can be used to monitor the level of the reagent liquid supply. The technician will understand that many different types of fluid level monitoring can be applied. The reagent fluid supply monitoring unit 32 digitizes the reagent fluid supply level and transmits it to the communication bus 34.

[0129] The medical analysis device 10a, 10b also includes an embedded computer 18. The purpose of the embedded computer is to coordinate the sequence of events to perform the test on the patient sample, transmit the results for further analysis, and perform various housekeeping duties associated with the medical analysis device 10a, 10b, including monitoring the fluid levels in the reagent fluid supply 28.

[0130] The embedded computer 18 includes an internal input / output unit 36 ​​for communicating with embedded hardware inside the medical analysis device 10a, 10b. For example, the internal input / output unit 36 ​​can be an "I2C" bus, a USB (TM) bus, etc. The embedded computer 18 also includes a non-volatile memory 38 (such as an embedded hard disk), a volatile memory 40 (such as RAM), a display driver 42, an external input / output unit 44 (such as an Ethernet port), an input unit 46 (such as a keyboard and / or a touch screen driver) and a processing unit 48 (such as an embedded Intel (TM) processor). These components are connected via a communication bus 50. The technician will understand the basic functions of these units, and there are many variations and implementations.

[0131] In operation, embedded computer 18 monitors inlet 14 for the presence of a sample via test detector 30. At the same time, embedded computer 18 receives a (optionally real-time) digital representation of the fluid level in reagent fluid supply 28 from reagent fluid supply monitoring unit 32.

[0132] Thus, embedded computer 18 may forward reagent fluid supply 28 level updates to external network 22 or device via external input / output unit 44. Optionally, embedded computer 18 may forward reagent fluid supply 28 level updates more frequently as the level of the reagent fluid drops to its lowest possible value.

[0133] Figure 2 A system 80 according to the second aspect is shown. The system 80 comprises a plurality of medical analysis devices 10x, 10y, 10z, each as described in conjunction with FIG. 1 . The system further comprises an optional data warehouse (server, cloud server) 88, an optional mobile data device 84 (e.g., a "smart phone" operating an iOS (TM) or Android operating system), an optional wireless communication base station 90, and an apparatus 60 according to the second aspect. All of these devices are communicatively coupled using a data communication network 82 (which may typically be a combination of a LAN, WAN, WiFi, GSM, LTE network, etc.). The operation of the system will be described later in this application.

[0134] It is often difficult for a point-of-care coordinator to monitor the amount of reagents remaining in each medical analysis device 10x, 10y, 10z. Different point-of-care medical analyzers require refilling at different times depending on their use. The rate at which reagents are consumed depends at least on the types of tests being run on the medical analysis device and the intensity of those tests. Such point-of-care medical analyzers are used at different intensities depending on whether they are installed in a busy town hospital or a local clinic, depending on the time of year, etc. Therefore, it has been essentially impossible for a point-of-care coordinator to accurately predict future supply requirements for reagents used in such medical tests, up to now.

[0135] In use, the device 60 polls (or receives updates from) the medical analysis devices 10x, 10y, 10z located at the point of care or testing facility 85. The medical analysis devices 10x, 10y, 10z transmit data to the device 60 via the data communication network 82. Specifically, the medical analysis devices 10x, 10y, 10z may transmit at least first, second, and third initial consumable status data reflecting the respective reagent fill levels of the medical analysis devices 10x, 10y, 10z to the device 60. Optionally, point of care medical analyzer usage information or measured consumable consumption rates may also be transmitted to the device 60 for calculation.

[0136] Those skilled in the art will appreciate that the medical analysis device 10x, 10y, 10z can be equipped with a sufficiently powerful data processor to generate the predicted consumable status data by itself. In this case, the predicted consumable status data (e.g., the fluid level in the reservoir) or the number of sensor elements used in the sensor box can be displayed on the display of the medical analysis device 10x, 10y, 10z, or forwarded to the device 60 or the mobile data device 84, so that the care point coordinator or technician can check the overall status of the medical analysis device 10x, 10y, 10z at the care point 85. Optionally, the medical analysis device 10x, 10y, 10z is configured to generate an order to the reagent supplier when the output of the predicted consumable consumption model indicates that reagent consumption is inevitable, or the sensor box is close to the end of its predicted life. Accordingly, the order can be transmitted to the reagent supplier via the data communication network 82.

[0137] Figure 3 An apparatus 60 according to the first aspect is shown.

[0138] A first aspect includes an apparatus 60 for generating predicted consumable status data for a point-of-care medical analyzer 10, comprising:

[0139] an input unit 62; and

[0140] Processing unit 64.

[0141] The input unit 62 is configured to receive initial consumable status data for the point-of-care medical analyzer, the initial consumable status data defining an initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer, and an analyzer identifier at a first time index.

[0142] The input unit 62 is configured to receive a second time index.

[0143] The processing unit 64 is configured to generate predicted consumable status data using a predicted consumable consumption model selected using at least the analyzer identifier and the initial consumable status data, the predicted consumable status data defining an amount of the point-of-care medical analyzer consumable predicted to be present in the point-of-care medical analyzer at a second time index.

[0144] The predictive consumables consumption model includes (i) a consumables model for a point-of-care medical analyzer, which defines the consumables consumption rate of point-of-care medical analyzer consumables used for testing in the point-of-care medical analyzer, and (ii) a point-of-care medical analyzer usage model, which defines the incidence of testing.

[0145] The predicted consumable status data is generated by the processing unit 64 based on an initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer and an aggregation of at least the outputs of the predicted consumable consumption model and the point-of-care medical analyzer usage model between the first time index and the second time index.

[0146] Therefore, it is recommended to receive the latest measurements of reagents from analyzers deployed at the point of care and predict for each analyzer when the various reagent consumables will run out.

[0147] The apparatus 60 may be implemented on a wide range of computing devices. For example, a personal computer, a server, a smart phone, a cloud server, and an embedded computer (optionally within a point-of-care medical analyzer) may all be configured to perform as an apparatus according to the first aspect.

[0148] According to an optional embodiment of the first aspect, the means 60 applies a device-specific average usage in order to give a more accurate estimate of when a device is available.

[0149] For example, the proposed framework is flexible and can optionally integrate domain knowledge and monitoring statistics of such a point-of-care medical analyzer system to determine whether the analyzer will run out of consumables on a public holiday or weekend. This enables the point-of-care coordinator to order and replace reagents in a timely manner before they run out.

[0150] Optionally, the status of all reagent consumables in point-of-care medical analyzers can be aggregated in one GUI view. This enables the operator to see which devices will experience consumable outages in the future. Optionally, the data can be visualized as a "calendar view" enabling the operator to simulate how consumable levels evolve over time, optionally based on the current levels of consumables.

[0151] In other words, for each analyzer and each type of consumable, one or more of the following aspects can be considered when predicting the status of the consumable: the current level of the consumable in the analyzer (measured using the reagent sensor 28); the expiration date of the consumable, the predicted number of measurements for a particular analyzer; and the average usage of the consumable per test.

[0152] Optionally, the information is presented on the GUI in a chronological order, enabling a point-of-care coordinator to monitor predictions of future events leading to consumable outages on the analyzer.

[0153] Optionally, information is displayed on the GUI, such as with a schematic cross-sectional view of a replaceable cartridge or sealed bag showing actual and / or predicted reagent levels remaining in the replaceable cartridge, sealed bag, or the estimated remaining life of the sensor cartridge.

[0154] Figure 4A method according to a second aspect is shown. Specifically, a method for generating predicted consumable status data for a point-of-care medical analyzer is provided, comprising:

[0155] receiving 100 initial consumable status data for a point-of-care medical analyzer, the initial consumable status data defining an initial amount of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer, and receiving an analyzer identifier at a first time index;

[0156] Receiving 102 a second time index;

[0157] generating 104 predicted consumable status data defining an amount of a point-of-care medical analyzer consumable predicted to be present in the point-of-care medical analyzer at a second time index using a predicted consumable consumption model selected using at least the analyzer identifier and the initial consumable status data;

[0158] wherein the predictive consumables consumption model comprises (i) a consumables model for a point-of-care medical analyzer, the consumables model defining a consumables consumption rate of point-of-care medical analyzer consumables used for testing in the point-of-care medical analyzer, and (ii) a point-of-care medical analyzer usage model, the point-of-care medical analyzer usage model defining an incidence rate of testing; and

[0159] Wherein the predicted consumable status data is generated based on an initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer and an aggregation of outputs of at least a predictive consumable consumption model and a point-of-care medical analyzer usage model between a first time index and a second time index.

[0160] Figure 5 One possible data structure 106 is shown, which may be transmitted between the point of care 85 and the device 60 by the point of care medical analyzer 10x, 10y, 10z, for example, as Figure 2 As shown. Specifically, the data structure 106 encapsulates initial consumable status data transmitted from the point-of-care medical analyzer, which defines the initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer. The data structure 106 includes a packet header 108 (which can be, for example, a TCP / IP header in addition to any specific application software header). In addition, the data structure 106 includes a plurality of consumable status fields 110b, 110c, 110d. Each consumable status field includes data measured by the reagent fluid supply monitoring unit 32, for example, defining the reagent supply or fill level of a sealed bag within the point-of-care medical analyzer, or the estimated remaining life of a sensor cartridge.

[0161] Figure 6A functional diagram highlighting the algorithm of the present invention is shown. A predictive consumable consumption model defining a consumable consumption model 120 and a point-of-care medical analyzer usage model 122 is shown.

[0162] The consumables consumption model 120 keeps track of how one activity of the analyzer is converted into consumables usage. In other words, for a particular point-of-care medical analyzer, it tracks how one use of the point-of-care medical analyzer is converted into consumables usage. This can be achieved by using some constants, data from actual consumables, and historical consumption information about the particular point-of-care medical analyzer.

[0163] The consumable consumption model 120 is optionally a continuously self-correcting predictive model that is optionally updated each time an input changes.

[0164] Optionally, the consumable consumption model 120 is a lookup table that defines a list of reagent volumes that are consumed for each type of test performed.

[0165] Optional inputs to the consumables consumption model 120 are, for example, historical usage parameters 124, values ​​derived from research and development 126 (such as various constants, coefficients, and rules that affect predictions, such as "when the analyzer has been running for less than 23 hours, test type III uses an additional 5.6% of consumables"). Another optional input to the consumables consumption model is consumables data reported from production 128. This provides detailed information, for example, about the content and capacity of each type of consumable.

[0166] The output 130 of the consumable consumption model 120 is a volume or volume set defining the consumable usage for each activity (test) type. Optionally, the output volume or volume set may, for example, form a mean or distribution of a statistical distribution, enabling the definition of an output probability of a given reagent volume per test.

[0167] Point-of-care medical analyzer usage models 122 predict the predicted amount and type of activity for a specific point-of-care medical analyzer. These models work by calculating how input parameters affect the ultimate amount of usage of the point-of-care medical analyzer. Parameters that have the greatest impact on usage are given a higher weight when predicting the ultimate activity. Using the inputs and historical usage patterns, the model will predict the future usage of the analyzer and optionally derive a probability of correct prediction.

[0168] The point-of-care medical analyzer usage model 122 is optionally a continuously self-correcting predictive model that is updated when its inputs change. Typical input parameters to the point-of-care medical analyzer usage model 122 are historical usage pattern data 132 for a given point-of-care medical analyzer, device metadata such as a model or configuration of a point-of-care medical analyzer 134, and / or point-of-care medical analyzer location information 136 (e.g., the department or hospital in which the point-of-care medical analyzer is located). The point-of-care medical analyzer usage model 122 outputs a predicted number of activities between a first and a second time index 138. An aggregator 140 combines the predicted number of consumables used for each activity of a single device between the first and second time indexes and the predicted number of different activities to produce a prediction of when a particular point-of-care medical analyzer will run out of a given reagent consumable or multiple consumables.

[0169] An initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer 148 , as measured by sensors within the point-of-care medical analyzer, is input into the aggregator 140 .

[0170] Optionally, the aggregator 140 may output a probability estimate of when a particular point-of-care medical analyzer will run out of a given reagent consumable between the first and second time indexes.

[0171] Optionally, the predicted consumable status data generated as described above can be continuously updated so that whenever one of the inputs to the model changes, the entire system or a branch of the system will quickly adapt to these changes. This provides a system that can quickly and accurately adapt to changes in the environment in which the analyzer is located. In addition, since many inputs change frequently, the accuracy of the system may be affected. However, according to the above system, the probability of predicted error data can be provided to the user.

[0172] FIG. 7 shows three different types of GUIs (Graphical User Interfaces) that may be displayed to a user of an apparatus system according to the present application.

[0173] It should be understood that a wide range of output options can be provided. For example, a graphical summary of the reagent status of one or more point-of-care medical analyzers can be provided in a tabular or numerical form on a screen of a personal computer. The graphical summary can be printed out using a printer. In addition, the graphical summary can be resized and rearranged to enable a series of smart phone applications to display the reagent status of one or more point-of-care medical analyzers to a user. It will be understood in the following discussion that fewer or more point-of-care medical analyzers can be provided than those used in the examples shown, and these examples are for illustration only.

[0174] Figure 7a) shows an example of a GUI format 150 that can be displayed on a data output unit of a device 60 and / or a point-of-care medical analyzer 10x as described above. The GUI is shown in a "tab" format, wherein tabs 150a, 150b, 150c, and 150d represent the reagent status of four different point-of-care medical analyzers. The first tab 150a is shown with a first reagent indicator 152 and a second reagent indicator 154. A drop-down menu 156 has been used to select a second time index. The method according to the second aspect has been used to calculate the reagent remaining amount at the second time index. Optionally, the first and second reagent indicators are bar graphs, and / or numerical time displays showing the remaining amount of the first and second reagents in a given point-of-care medical analyzer. Optionally, this embodiment of the GUI is configured to display a warning indicator 158 if the level of the first reagent is (for example) insufficient to meet the predicted workload of the point-of-care medical analyzer. Optionally, the numerical display 160 can define the remaining time during which the given first and second reagents are available in view of the predicted workload. This display technology may be better suited for simple (optionally monochrome) displays on point-of-care medical analyzers.

[0175] Figure 7b ) shows another version of a GUI format 162 that can be displayed on a data output unit of the device 60 and / or point-of-care medical analyzer 10x as described above. The GUI displays a tabular forecast 164 of reagent usage percentages for a given week. For example, this display option is more suitable for output as a printed report by printing to paper.

[0176] Figure 7c ) shows another version of a GUI format 168 that can be displayed on a data output unit of the device 60 and / or point-of-care medical analyzer 10x as described above. The GUI has three rows on the left, with the first row 170 representing a first point-of-care medical analyzer, the second row 172 representing a second point-of-care medical analyzer, and the third row 174 representing a third point-of-care medical analyzer. The width dimension of the illustrated GUI represents early and late stages in the manner of a "Gantt" chart. In the illustrated case, the point-of-care medical analyzers 170, 172, 174 are identical, and each requires reporting of two reagent quantities. However, the skilled person will understand that the GUI can report fewer or more reagent quantities. Time navigation arrows 176a and 176b enable the user to "scan" the time limit indicator 178 backward and forward in time, respectively.

[0177] For the first point-of-care medical analyzer 170, a first bar 180 defines the availability of a first reagent. A second bar 182 defines the availability of a second reagent. The bar 182 for the first point-of-care medical analyzer 172 is interrupted at 184, illustrating the replacement of the second reagent.

[0178] Optionally, a time limit indicator 178 is set by the user to a time point on the GUI corresponding to the second time index until which the point-of-care medical analyzer must not run out of consumables. Optionally, the rows of the first point-of-care medical analyzer 170, the second point-of-care medical analyzer 172, and the third point-of-care medical analyzer 174 may be sorted according to how urgently the consumables need to be replaced. Optionally, a "to-do list" of important point-of-care medical analyzer refilling tasks may be generated.

[0179] Optionally, the bars of the GUI may be configured to display as "hard" (single color) bars that are interrupted at the point in time when the availability of a particular reagent is guaranteed to be below a probability level calculated by the predictive consumable consumption model.

[0180] Optionally, the bar of the GUI can be configured to be displayed as a "variable intensity" bar, where the estimated probability of correctness of the predicted consumable consumption model deteriorates. For example, it is naturally expected that small errors in the predicted consumable consumption model will be magnified over a longer period of time in the future. Displaying a higher probability of error with a weaker bar intensity can emphasize to the user that at a given date displayed on the GUI, when the bar intensity becomes too weak, the reagent needs to be replaced.

[0181] According to a fourth aspect, there is provided a computer program element comprising instructions which, when executed by a computer, enable the computer to perform the method of the first aspect or an embodiment thereof.

[0182] According to a fifth aspect, there is provided a computer program product tangibly embodied on a carrier medium, comprising software code for performing the method of the first aspect.

[0183] The computer program elements may comprise, for example, a data structure containing computer instructions which, when executed by a processing unit of a computer, form the computational steps of the method of the second aspect.

[0184] The invention can be embodied on a computer which comprises the computer program elements from the outset or which has received the computer program elements in the form of an update, either by means of a disc or the Internet.

[0185] Alternatively, the computer readable medium may comprise an optical storage or distribution medium, such as a CD-ROM disc, a DVD, or a solid-state storage or distribution medium, such as a USB stick or a disk.

[0186] Optionally, a program for providing the computer program elements via downloading over the Internet is provided.

[0187] Although the present invention has been illustrated and described with reference to the drawings and this specification, these should be considered illustrative and exemplary rather than restrictive. The present invention is not limited to the disclosed embodiments. Reasonable changes to the embodiments discussed herein may be made by a skilled person studying the drawings, the disclosure, and the dependent claims. The reference signs should not be interpreted as limiting the scope.

Claims

1. An apparatus (60) for generating predictive consumable status data for a point-of-care medical analyzer, comprising: Input unit (62); and Processing unit (64); wherein the input unit (62) is configured to receive initial consumable status data of a point-of-care medical analyzer, the initial consumable status data defining an initial amount of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer, and to receive an analyzer identifier at a first time index; Wherein, the input unit (62) is configured to receive a second time index; wherein the apparatus (60) is configured to receive usage location metadata of the point-of-care medical analyzer defined by the analyzer identifier; wherein the processing unit (64) is configured to generate predicted consumable status data using a predicted consumable consumption model selected using at least the analyzer identifier and the initial consumable status data, the predicted consumable status data defining an amount of a point-of-care medical analyzer consumable predicted to be present in the point-of-care medical analyzer at the second time index; wherein the predictive consumables consumption model comprises (i) a consumables model for the point-of-care medical analyzer, the consumables model defining a consumables consumption rate of the point-of-care medical analyzer consumables used for testing in the point-of-care medical analyzer, and (ii) a point-of-care medical analyzer usage model, the point-of-care medical analyzer usage model defining an incidence rate of the test, wherein the consumables model for the point-of-care medical analyzer and / or the point-of-care medical analyzer usage model are continuous self-correcting predictive models that are updated when their input parameters change; and Wherein, the predicted consumable status data is generated by a processing unit (64) based on the initial consumable status data defining an initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer, and an aggregation of the outputs of at least the consumable model and the point-of-care medical analyzer usage model between the first time index and the second time index.

2. The device (60) according to claim 1, further configured to: Historical usage data for a point-of-care medical analyzer defined by the analyzer identifier is received.

3. The device (60) according to claim 2, further configured to: A consumable usage rule set is received, the consumable usage rule set defining the consumption of the point-of-care medical analyzer consumables over time.

4. The device (60) according to claim 3 is further configured to: A consumable capacity metric for a consumable of the point-of-care medical analyzer is received using the analyzer identifier.

5. The device (60) according to claim 4, further configured to: The consumable model is generated based on at least one of the historical usage data, a consumable handling model for the point-of-care medical analyzer, the consumable usage rule set, and a consumable capacity metric for the point-of-care medical analyzer identified by the analyzer identifier and run from the initial consumable status data collected at the first time index.

6. The device (60) according to claim 1 is further configured to: Usage pattern data obtained prior to the first time index for the point-of-care medical analyzer defined by the analyzer identifier is received.

7. The device (60) according to claim 6, further configured to: A future usage schedule of the point-of-care medical analyzer defined by the analyzer identifier between a first time index and a second time index is received.

8. The device (60) according to claim 7, further configured to: The point-of-care medical analyzer usage model is generated based on at least one of the usage pattern data, the usage location metadata, and the future usage plan.

9. The device (60) according to claim 8, in, The point-of-care medical analyzer outputs an expected number of tests between the first and second time indexes using a model.

10. The device (60) according to claim 9: in, Aggregating outputs of at least a predictive consumable consumption model and a point-of-care medical analyzer usage model between the first time index and the second time index comprises: Calculate the test consumption of the point-of-care medical analyzer consumables used for testing in the point-of-care medical analyzer, and multiply the test consumption by the expected number of tests to generate predicted consumables status data at the second time index, and / or as a variable between the first time index and the second time index.

11. The apparatus (60) according to claim 1, further configured to: receiving expiration date data for consumables of the point-of-care medical analyzer; The point-of-care medical analyzer consumable is factored into the consumable model only between the first time index and a due date defined in the due date data.

12. The apparatus (60) according to claim 1, further configured to: Generate predictive consumable status data for multiple point-of-care medical analyzers.

13. A method for generating predictive consumable status data for a point-of-care medical analyzer, comprising: receiving initial consumable status data for a point-of-care medical analyzer, the initial consumable status data defining an initial amount of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer and an analyzer identifier at a first time index; receiving a second time index; receiving usage location metadata of the point-of-care medical analyzer defined by the analyzer identifier; generating predicted consumable status data using a predicted consumable consumption model selected using at least the analyzer identifier and the initial consumable status data, the predicted consumable status data defining an amount of a point-of-care medical analyzer consumable predicted to be present in the point-of-care medical analyzer at the second time index; wherein the predictive consumables consumption model comprises (i) a consumables model for the point-of-care medical analyzer, the consumables model defining a consumables consumption rate of the point-of-care medical analyzer consumables used for testing in the point-of-care medical analyzer, and (ii) a point-of-care medical analyzer usage model, the point-of-care medical analyzer usage model defining an incidence rate of the test, wherein the consumables model for the point-of-care medical analyzer and / or the point-of-care medical analyzer usage model are continuous self-correcting predictive models that are updated when their input parameters change; and Wherein, the predicted consumable status data is generated based on the initial consumable status data defining an initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer, and an aggregation of the outputs of at least the consumable model and the point-of-care medical analyzer usage model between the first time index and the second time index.

14. A system (80) for generating predictive consumable status data for a point-of-care medical analyzer, comprising: at least one point-of-care medical analyzer (10x, 10y, 10z); The device (60) according to any one of claims 1 to 12; Data communication network (82); Wherein, the point-of-care medical analyzer (10x, 10y, 10z) is configured to transmit initial consumable status data of the point-of-care medical analyzer to the device via the data communication network, wherein the initial consumable status data defines an initial amount of point-of-care medical analyzer consumables present in the point-of-care medical analyzer, and transmits an analyzer identifier at a first time index.

15. The system (80) of claim 14, further comprising: Mobile data devices (84); Wherein, the mobile data device is configured as: displaying a GUI including a representation of the point-of-care medical analyzer and an initial amount of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer; sending, via the data network, a second time index to the device via a GUI interface; receiving response data from the device (60); and An indication of how urgently replacement of a point-of-care medical analyzer consumable present in the point-of-care medical analyzer at the second time index is displayed on the GUI of the mobile device.

16. The system (80) according to any one of claims 14 or 15, in, The at least one point-of-care medical analyzer (10x, 10y, 10z) is configured to identify an installed point-of-care medical analyzer consumable using a point-of-care medical analyzer consumable identifier; and Wherein, at least one of the point-of-care medical analyzers (10x, 10y, 10z) is configured to transmit an intermediate quantity of a point-of-care medical analyzer consumable present in the installed point-of-care medical analyzer consumable in association with the point-of-care medical analyzer consumable identifier via a data communications network (82).

Images