Method for improving a drug library for an infusion pump, infusion pump and server

Abstract

The present invention describes a method for improving a drug library for an infusion pump The method comprises the step of obtaining infusion settings for at least one drug from at least one infusion pump and the step of determining a typical usage pattern for the at least one drug based on the obtained infusion settings for the at least one drug. The infusion settings comprise at least one input parameter the at least one pump is programmed with for at least one infusion with the at least one drug and / or the infusion settings comprise information on at least one component connected to the at least one pump for at least one infusion with the at least one drug. A further object of the present invention is a computer-readable storage medium comprising an improved drug library.

Description

[0001] Method for improving a drug library for an infusion pump, Infusion pump and Server

[0002] Description

[0003] The invention relates to a method for improving a drug library for an infusion pump, an infusion pump according to the preamble of claim 13, a server according to the preamble of claim 14 and a computer-readable storage medium according to the preamble of claim 15.

[0004] Infusion pumps are used to administer medication to a patient in an automated way. Infusion pumps are typically programmed by a healthcare professional in order to start the infusion. During programming, the healthcare professional specifies infusion parameters such as a bolus or dose rates for the drug to be administered. Errors in programming, in particular in the specification of the bolus or the dose rate, can be critical and endanger the patient, either by overdosing or underdosing the patient to be infused.

[0005] Current infusion pumps are typically equipped with a drug library that provides a dose error reduction system (DERS). The drug library provides a list of drugs and, for each drug, hard and soft limits for e.g. the bolus, the dose rate or flow rate (ml / h) of each drug, wherein hard limits cannot be overridden during programming, whereas soft limits provide an information or alert, but can be overridden. Hard limits are typically set such that extremely wrong programming of a user is prevented. Soft limits provide a recommendation or guideline that can be ignored and overridden by the user defined within an authorized range. Setting meaningful and useful soft limits is a difficult task. If soft limits are set too tight, they are overridden too often, leading to an alert fatigue on the side of the healthcare professionals programming the pump. If the soft limits are set too loose, no additional security is provided through the soft limit as compared to the hard limit. Another challenge is that dosage or flow rates may differ depending on the setting, e.g. ICU or neonatal unit, and on the patient group. For example, different boluses, dosage rates or flow rates apply to infants and adults, for different patient weights or sex.

[0006] It is an object of the instant invention to overcome these disadvantages of the prior art, improving dose-error reduction on infusion pumps, reducing alert fatigue and hence, increasing patient safety. This object is achieved by means of a method for improving a drug library for an infusion pump comprising the features of claim 1.

[0007] Accordingly, the method comprises the step of obtaining infusion settings for at least one drug from at least one infusion pump. In this way, information on the infusion settings actually and in practice used for an infusion of the at least one drug at the at least one infusion pump is obtained. The infusion setting are e.g. retrieved or received from the infusion pump. The infusion settings comprise at least one input parameter the at least one pump is programmed with for at least one infusion with the at least one drug. Hence, information is obtained on at least one input parameter that is being programmed for an infusion with the at least one drug with the at least one infusion pump. Obtaining at least one input parameter comprises, for example, obtaining the input parameter and its programmed value. Alternatively or additionally, the infusion settings comprise information on at least one component connected to the at least one pump for at least one infusion with the at least one drug. In this way, information on at least one component connected to the at least one pump during an infusion with the at least one drug is obtained. For example, information on the at least one component comprises information on the type of tube used for infusion.

[0008] The method further comprises the step of determining a typical usage pattern for the at least one drug based on the obtained infusion settings for the at least one drug. In this way, a usage pattern is derived which is based on the actual usage of the infusion pump, in particular on the infusion settings used in practice for an infusion with the at least one drug. The typical usage pattern comprises for example infusion settings that are typically being used for the infusion of the at least one drug at an infusion pump. A typical usage pattern comprises e.g. one or more input parameters used for infusion with the at least one drug. A typical usage pattern can also comprise a combination of one or more input parameters used for infusion with the at least one drug and information on at least one component, e.g. a tube type, connected to an infusion pump during infusion with the at least one drug. A usage pattern for at least one drug, e.g. infusion settings used for infusion with the at least one drug, is, for example, a typical usage pattern if it is determined to be of particular relevance, e.g. being a frequent usage pattern.

[0009] According to one aspect, the steps of obtaining infusion settings for the at least one drug and determining a typical usage pattern for the at least one drug are repeated. With an increasing number of obtained infusion settings for the at least one drug, the determined typical usage pattern becomes more and more precise, leading to an update of the typical usage pattern with every obtained infusion settings. In one aspect, the step of determining a typical usage pattern for the at least one drug is carried out only after the step of obtaining infusion settings has been repeated a defined number of times Ndata. This leads to an update of the typical usage pattern only if the number of data points obtained on the infusion settings has increased by Ndata. According to one aspect, the method is carried out for each drug or a specified set of drugs the at least one infusion pump is programmed with for an infusion.

[0010] According to one aspect, the step of determining a typical usage pattern for the at least one drug is carried out only if a certain criterium is satisfied, e.g. if the number of obtained infusion settings for the at least one drug exceeds a certain threshold Nstart. This ensures that the determined typical usage pattern is built on a certain amount of data, hence, providing a certain level of reliability.

[0011] According to one aspect, obtaining infusion settings comprises obtaining at least one first infusion parameter the at least one pump is programmed with for at least one infusion with the at least one drug, wherein the at least one first infusion parameter determines the amount of the at least one drug to be infused. The first infusion parameter comprises, for example, a dose rate (e.g. in mg / h, mg / kg / h), a flow rate (e.g. in ml / h), a drug concentration (e.g. in mg / ml), an infusion rate (e.g. in ml / h), a bolus set (e.g. in mg) and / or a loading / induction dose (e.g. in mg / kg). Obtaining at least one first infusion parameter comprises obtaining information on the at least one first infusion parameter and its value. In this way, information on the amount of the at least one drug to be administered, e.g. its dose rate or its flow rate, is being obtained.

[0012] According to one aspect, determining a typical usage pattern for the at least one drug based on the obtained infusion settings comprises determining a typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug. A typical value for the at least one first infusion parameter comprises for example a value obtained for the at least one first infusion parameter that occurs with a certain frequency, e.g. the most commonly used value obtained for the at least one first infusion parameter or an averaged or mean value. A typical range comprises, for example, a range of values within which a certain percentage of the obtained values for the at least one first infusion parameter lies. An upper limit is, for example, a value above which all or a certain percentage of the obtained values for the at least one first infusion parameter lies. Correspondingly, a lower limit is, for example, a value below which all or a certain percentage of the obtained values for the at least one first infusion parameter lies. The typical value, the typical range, the lower limit and / or the upper limit comprise, in one aspect, a confidence interval. In this way, information on the typically programmed value, the typically programmed range of values, a lower limit of the values and / or an upper limit for the values programmed for the at least one drug is provided.

[0013] According to a further aspect, obtaining infusion settings comprises obtaining at least one second infusion parameter the at least one pump is programmed with for at least one infusion with the at least one drug, wherein the at least one second infusion parameter carries information on the patient to be infused or on the location of the infusion. The second infusion parameter comprises, for example, a ward, a patient weight, a patient body surface area (BSA), a patient age and / or a patient sex. The ward is in particular the ward the infusion pump is used at during the infusion with the at least one drug. The at least one second infusion parameter provides additional information on the drug application, e.g. its setting and / or the patient to whom the drug is being administered. This allows for a more specific determination of a typical usage pattern. In one aspect, several second infusion parameters are obtained, e.g. the ward and the patient weight and the patient age and the patient sex. The patient age is e.g. specified in years, years and months, years, months and days or in the form of a category (e.g. infant, child, adult), wherein the category is defined by certain age intervals, for example.

[0014] According to one aspect, determining a typical usage pattern for the at least one drug based on the obtained infusion settings comprises determining a typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug depending on the at least one second infusion parameter. In particular, a typical usage pattern is being determined based on the at least one first infusion parameter and the at least one second infusion parameter. In this way, more specific typical usage patterns can be derived. E.g. a typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug is determined for each value of the obtained at least one second infusion parameter. E.g. a typical dose rate, a typical range for a dose rate, an upper limit for a dose rate and / or a lower limit for a dose rate is or are being determined for different wards, different patient ages or patient age groups or patient age categories, different patient sex etc. In another example, a typical bolus, a typical range for a bolus, an upper limit for a bolus and / or a lower limit for a bolus is or are being determined for different wards, different patient ages or patient age groups or patient age categories, different patient sex etc. The more second infusion parameters are obtained, the more specific the determined typical usage pattern can be defined. E.g. a typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug can be defined depending on one second infusion parameter, e.g. patient weight or ward, or defined depending on two or more second infusion parameters, e.g. patient weight and ward, or others.

[0015] According to one aspect, artificial intelligence, e.g. a machine-learning model, is being used to determine the typical usage pattern for the at least one drug based on the obtained infusion settings. In this way, a typical usage pattern can be determined based on the actual usage of the at least one pump, e.g. on the actually used infusion settings used for the at least one drug. For example, the artificial intelligence, e.g. the machine-learning model, receives at least a first and at least a second infusion parameter obtained as infusion settings for the at least one drug as input. Based on this input, the artificial intelligence, e.g. the machine-learning model, learns to predict a value of the at least one first infusion parameter given a value for the at least one second infusion parameter, or given a set of values for a set of second infusion parameters. In one aspect, the artificial intelligence, e.g. the machine-learning model, determines a typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug, in one aspect depending on the at least one second infusion parameter, artificial intelligence, e.g. a machine-learning model, constitutes a particular efficient way for determining a typical usage pattern based on the obtained infusion settings.

[0016] According to one aspect, the method comprises the step of improving a drug library by adding information based on the determined typical usage pattern to the drug library. In this way, an improved drug library is being provided which takes into account the actual programming behaviour of users, in particular healthcare professionals, and thus the infusion settings used in real clinical practice for the at least one drug. In this way, an evolutive drug library is being provided that evolves with and adapts to clinical practice and standards.

[0017] According to one aspect, improving the drug library comprises adding an improved upper limit and / or an improved lower limit for the at least one first infusion parameter for the at least one drug to the drug library, wherein the improved upper limit and / or the improved lower limit are determined based on the typical usage pattern. The improved lower limit and / or the improved upper limit is added, for example, as a soft limit to the drug library. A soft limit is an overridable limit. The soft limit is e.g. displayed to a user, e.g. on a pump display. An alert is in one example being issued, e.g. on a pump display, if a soft limit is being crossed. The soft limit can be overridden and infusion with the infusion parameter beyond the soft limit be initiated or continued. The improved lower limit and / or the improved upper limit are, in one example, added as additional limits for the at least one drug to the drug library. Alternatively, the improved lower limit and / or the improved upper limit replace existing lower and / or upper limits in the drug library. E.g. the drug library is a basic drug library comprising soft limits and these soft limits are being replaced by the improved lower and / or upper limits for the at least one drug. In one aspect, adding an improved upper limit and / or an improved lower limit for the at least one first infusion parameter for the at least one drug to the drug library is carried out only if a certain criterium is satisfied, e.g. a certain confidence reached.

[0018] According to one aspect, improving the drug library comprises adding an improved upper limit and / or an improved lower limit for the at least one first infusion parameter for the at least one drug depending on the at least one second infusion parameter to the drug library, wherein the improved upper limit and / or the improved lower limit are determined based on the typical usage pattern. By making the improved upper limit and / or the improved lower limit dependent on the at least one second infusion parameter, more specific improved upper and / or lower limits obtained. E.g. an improved upper limit and / or an improved lower limit for the at least one drug can be added to be drug library for different values of the at least one second infusion parameter. In one aspect, an improved upper limit and / or an improved lower limit for the at least one drug for different wards, patient weights or weight intervals, patient ages or age intervals or age categories and / or for different patient sex are added to the drug library. In this way, a more precise setting of upper and / or lower limits can be provided. In particular, a drug library with more specific soft limits can be provided.

[0019] According to one aspect, the method is executed on the at least one pump and the infusion settings are obtained from infusions on the at least one pump. In this way, infusion settings are collected locally for a single pump depending on the usage of this specific pump. In one aspect, the infusion pump comprises a drug library, e.g. stored on a storage medium of the pump, and the drug library is improved by adding information based on the determined typical usage pattern to the drug library.

[0020] According to one aspect, the method is executed on a server, wherein the server is configured to be connected to a plurality of infusion pumps and infusion settings are obtained from infusions on the plurality of infusion pumps. In this way, infusion settings are collected from a plurality of pumps. A larger set of data can be obtained in a reduced amount of time and the reliability of the determined usage pattern for the at least one drug be increased. Moreover, different usage scenarios and usage habits are taken into account as data from more pumps is being collected and used in the determination of the typical usage pattern. In one aspect, infusion settings are obtained and typical usage patterns are determined for each of a plurality of drugs.

[0021] A further object is a computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method according to any of the claims 1 to 13. A further object is a computer-readable storage medium comprising instructions which, when executed by a computer, cause the computer to carry out a method according to any of the claims 1 to 13. The computer program and the computer- readable storage medium implement the technical effects and the advantages described regarding the aforementioned method for improving a drug library.

[0022] A further object of the present invention is an infusion pump comprising a processor wherein the processor is configured to carry out a method according to claim 12. The infusion pump implements the technical effects and the advantages described regarding the method according to claims 1 to 12. A further object of the present invention is a server configured to be connected to a plurality of infusion pumps, wherein the server is configured to carry out a method according to claim 13. The server implements the technical effects and the advantages described regarding the method according to claims 1 to 11 and 13.

[0023] A further object of the present invention is computer-readable storage medium, wherein the computer-readable storage medium comprises an improved drug library obtained through a method according to any of the claims 9 to 12. In this way, a computer-readable medium with an improved drug library is being provided which has been developed in line with actual clinical practice. The computer-readable storage medium comprising the improved drug library is e.g. part of an infusion pump. Alternatively or additionally, the computer-readable storage medium comprising the improved drug library is part of an infusion station, e.g. comprising a storage rack for multiple infusion pumps to be arranged on the bedside of a patient, to which multiple infusion pumps can be communicatively connected. The infusion pumps connected to the infusion station can hence share the same improved drug library. In another alternative, the computer-readable storage medium comprising the improved drug library is part of a server which can be communicatively connected, e.g. via a wireless or via a wired connection, to multiple infusion pumps. In particular, the improved drug library comprises information based on at least one determined typical usage pattern. With the information added to the drug library, the drug library can provide improved dose-error reduction mechanisms. In one aspect, the improved drug library comprises an improved upper limit and / or an improved lower limit for at least one first infusion parameter for at least one drug, wherein the improved upper limit and / or the improved lower limit is e.g. an overrideable limit. For example, the improved upper limit and / or the improved lower limit for the at least one first infusion parameter for at least one drug is provided in the drug library in addition to a soft limit, providing an additional limit reflecting actual clinical practice. In one aspect, the drug library comprises an improved upper limit and / or an improved lower limit for at least one first infusion parameter for at least one drug depending on at least one second infusion parameter, e.g. as an overrideable limit. In this way, a more specific upper limit and / or lower limit for the at least one first infusion parameter for the at least one drug is provided, reflecting the at least one second infusion parameter. The improved upper and / or improved lower limit for the at least one first infusion parameter is, in one example, provided in the drug library depending on a plurality of second infusion parameters. This allows for a further differentiation. In one example, the at least one second infusion parameter is the ward, i.e. the station of the care unit in which the at least one infusion pump is being used during the at least one infusion with the at least one drug. In this example, different drug libraries are being obtained for different values of the at least one second infusion parameter, i.e. the different wards. Each ward-specific drug library comprises information added to it based on the determined typical usage pattern for the at least one drug for that ward. E.g. each ward-specific drug library comprises ward-specific improved upper and / or lower limits for the at least one first infusion parameter for the at least one drug. In this way, a more specific drug library can be provided. Additionally to the ward as at least one second infusion parameter, other second infusion parameters can be obtained as part of obtaining the infusion settings for the at least one drug. In this way, the ward-specific drug libraries can be provided with information based on a typical usage pattern differentiating for different second infusion parameters, such as patient weight, patient age and / or patient sex.

[0024] Further, a method for using an improved drug library, in particular, a method for dose-error reduction using an improved drug library, is provided, wherein a programming pattern is being compared to information added to the drug library based on a determined typical programming pattern. In one aspect, a value programmed for the at least one first infusion parameter is compared to a typical value for that infusion parameter. The typical value has, e.g. been determined using artificial intelligence, e.g a machine-learning model. Alternatively or additionally, a modification of a programmed at least one first infusion parameter is compared to a typical value for a modification of that infusion parameter. In one aspect, a value programmed for the at least one first infusion parameter is compared to an improved upper and / or an improved lower limit for the at least one infusion parameter for the at least one drug. The improved upper and / or the improved lower limit have been determined by the aforementioned method. In particular, the improved upper and / or improved lower limit have been determined based on obtained infusion settings, in particular using artificial intelligence, e.g. using a machine-learning model. In one example, an alert is being issued if the programmed value for the at least one infusion parameter deviates from the typical value for that infusion parameter. Alternatively or additionally, an alert is being issued if the modification of a programmed value for the at least one infucation parameter deviated from the typical value for the modification of that infusion parameter. In one example, an alert is issued if the programmed value for the at least one infusion parameter is below the improved lower limit and / or above the improved upper limit. The alert is e.g. a graphical alert issued on a display of an infusion pump about to be programmed, on a display of an infusion station used for the programming on at least one connected infusion pump or on any other display connected to at least one infusion pump about to be programmed. In one example, the comparison of the value programmed for the at least one first infusion parameter to the improved upper and / or the improved lower limit consists of a display of the programmed value, the improved upper limit and / or the improved lower limit, e.g. on a display used for programming at least one infusion pump about to be programmed.

[0025] The idea underlying the invention shall subsequently be described in more detail with respect to the embodiment shown in the drawings. Herein:

[0026] Fig. 1 illustrates a method for improving a drug library for an infusion pump according to one aspect of the present invention;

[0027] Fig. 2 illustrates a method for improving a drug library for an infusion pump according to one aspect of the present invention; and

[0028] Fig. 3 illustrates an improved drug library according to one aspect of the present invention and a method for using an improved drug library according to one aspect of the present invention.

[0029] Fig. 1 illustrates a method 1 for improving a drug library according to an aspect of the present invention. The improved drug library is to be used on or with an infusion pump. According to the present invention, a drug library provides a set of drugs with upper and / or lower limits for their admissible dose rates, flow rates or other infusion parameters that specify the amount to be administered. The upper or lower limits can be defined as hard or as soft limits in the drug library. A hard limit cannot be overridden by a user programming the infusion pump. A soft limit is a recommended limit which can be overridden by a user, e.g. after additional confirmation by the user. The method 1 comprises the step of obtaining 101 infusion settings for at least one drug from at least one infusion pump. The infusion settings comprise at least one input parameter the at least one infusion pump is programmed with for at least one infusion with the at least one drug. Alternatively or additionally, obtaining 101 the infusion settings comprises obtaining information on at least one component connected to the at least one pump for the at least one infusion with the at least one drug. In one example, the step of obtaining 101 infusion settings for at least one drug from at least one infusion pump is being repeated, e.g. until a sufficient amount of data on the infusion settings for the at least one drug has been obtained. In one example, during every infusion on a specific infusion pump or a set of infusion pumps, infusion settings for the drug infused on that pump or the set of pumps are being obtained 101. In one alternative, infusion settings are obtained only for drugs infused on that pump or the set of pumps out of a set of specified drugs. In this way, a continuous monitoring and gathering of data on the infusion settings used on that pump or on the set of pumps is being obtained.

[0030] For example, obtaining 101 infusion settings comprises obtaining at least one first infusion parameter the at least one infusion pump is programmed with for at least one infusion with the at least one drug. E.g. at least one first infusion parameter and its programmed value are received or retrieved during programming of the at least one infusion pump. One or multiple first infusion parameters are, for example, obtained during the step 101. A first infusion parameter in the sense of the present invention is a parameter that determines the amount of a drug to be administered. Examples of first infusion parameters are a dose rate or a flow rate, but also, a drug concentration, an infusion rate, a bolus set and / or a loading / induction dose comprise first infusion parameters. In one example, obtaining 101 infusion settings comprises obtaining, in addition to the at least one first infusion parameter, at least one second infusion parameter the at least one pump is programmed with for at least one infusion with the at least one drug. A second infusion parameter is an infusion parameter that specifies information on the patient to be infused or on the location of the infusion. A second infusion parameter is, for example, a ward (of a healthcare unit, e.g. an ICU, a neonatal unit, a psychiatric unit, etc.), a patient weight, a patient body surface area (BSA), a patient age or a patient sex. Either one second infusion parameter or multiple second infusion parameters are being obtained during the step of obtaining 101 infusion settings. The more first and second infusion parameters are obtained, the more detailed knowledge on the clinical infusion practice is being gathered. The method further comprises the step 102 of determining a typical usage pattern for the at least one drug based on the obtained infusion settings for the at least one drug. A typical usage pattern comprises a typical programming pattern for the at least one drug, as it derived from the obtained infusion settings for the at least one drug. In one variant, the typical usage pattern for the at least one drug is determined based on the obtained infusion settings gathered during a single infusion with the at least one drug. Additionally or alternatively, the typical usage pattern for the at least one drug is determined based on the obtained infusion settings gathered during multiple infusions with the at least one drug. In this way, a larger data set is used to determine the typical usage pattern, making it statistically more relevant and reliable. E.g. the step of obtaining 101 infusion settings for the at least one drug is repeated multiple times, before the step of determining 102 a typical usage pattern is being carried out. In one variant, the step of obtaining 101 infusion settings for at least one drug and the step of determining 102 a typical usage pattern for at least one drug are repeated multiple times, e.g. for the same drug, for different drugs, on one infusion pump or across multiple infusion pumps. In one example, the steps of obtaining 101 infusion settings and determining 102 a typical usage pattern are carried out for every infusion on a pump or a set of pumps, for all infused drugs or for a specified set of drugs. In this way, reliable typical usage patterns for multiple drugs are obtained quickly and a continuous update of the determined 102 typical usage patterns for the different drugs is obtained, keeping the typical usage patterns aligned with clinical practice.

[0031] In one example, determining 102 a typical usage pattern for the at least one drug comprises determining a typical value, a typical range, an upper limit and / or a lower limit for at least one first infusion parameter for the at least one drug. E.g. during step 101 , at least one first infusion parameter is being obtained, and in step 102, a typical value, a typical range, an upper limit and / or a lower limit for that first infusion parameter is being determined based on the obtained first infusion parameter. If, during the step of obtaining 101 infusion settings, at least one second infusion parameter is obtained, determining 102 a typical usage pattern comprises determining a typical usage pattern for the at least one drug based on the obtained infusion settings comprises determining a typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug depending on the at least one second infusion parameter. E.g. one typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug is determined for each value or for each relevant value or for each available value of the at least one second infusion parameter. In one variant, one typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug is determined for multiple second infusion parameters, e.g. a ward and a patient weight, a ward, a patient age and a patient weight, or any other number or combination of second infusion parameters. In this way, e.g. a multi-dimensional array of typical values, typical ranges, upper limits and / or lower limit is being obtained for the at least one drug, wherein the dimension of the array depends on the number of second infusion parameters obtained. In the depicted example, the typical usage pattern is being determined 101 using artificial intelligence, in particular a machine-learning model, based on the obtained infusion settings. In this way, a particular efficient method for determining 102 a typical usage pattern is being provided.

[0032] Fig. 2 illustrates a method 1 for improving a drug library according to a further aspect of the present invention. The method 1 according to the aspect detailed with regard to Fig. 2 comprises all the features detailed with regard to the aspect illustrated in Fig. 1 . In addition, the method 1 comprises the additional step of improving 103 a drug library by adding information based on the determined typical usage pattern to the drug library. In this way, an improved drug library is provided which takes into account clinical practice. In particular, improving 103 the drug library comprises adding an improved upper limit and / or an improved lower limit for the at least one first infusion parameter for the at least one drug to the drug library. The improved upper limit and / or the improved lower limit for the at least one first infusion parameter for the at least one drug is / are determined based on the typical usage pattern. In this way, a drug library with improved limits based on clinical practice is provided, enabling improved dose-error reduction, e.g. by reducing programming errors and reducing alert fatigue. In one example, the improved upper limit and / or improved lower limit are identical to the upper limit and lower limit, respectively, as determined 102 based on the obtained 101 infusion settings, in particular the obtained at least one first infusion parameter. Alternatively, the improved upper limit and / or improved lower limit are identical to the upper limit and lower limit, respectively, as determined 102 based on the obtained 101 infusion settings, i.e. the obtained at least one first infusion parameter depending on at least one second infusion parameter. In this way, improved upper and / lower limits for at least one first infusion parameter for at least one drug are provided for each value or each relevant value or each available value of one or multiple second infusion parameter(s). Improved upper and / or lower limits are for the at least one first infusion parameter are, in one example, added to different drug libraries, depending on the value of the at least one second infusion parameter. In particular, for each ward, an improved upper and / or lower limit for the at least one infusion parameter for the at least one drug is added to a wardspecific drug library. In the depicted example, the improved upper limit and / or lower limit are added to the drug library as soft, i.e. overrideable, limits for the at least one first infusion parameter, in particular the dose rate or flow rate, of the at least one drug. In one example, the steps of obtaining 101 infusion settings and determining 102 a typical usage pattern for the at least one drug is being repeated multiple times on the same infusion pump or across multiple pumps, before the step of adding 103 information based on the determined typical usage pattern, in particular improved upper and / or lower limits for the at least one first infusion parameter for the at least one drug, to the drug library. In another example, the improved upper and / or lower limits are update at the drug library at regular intervals, e.g. every time or every Nth time a typical usage pattern has been determined 102, in particular when a lower and / or upper limit for the at least one first infusion parameter for the at least one drug has been determined.

[0033] Fig. 3 illustrates a display 2 for programming an infusion pump. The display 2 is part of an infusion pump, an infusion station to be connected to multiple infusion pumps, or a server to be connected to multiple infusion pumps. The display 2 is connected to a computer- readable storage medium comprising an improved drug library according to an aspect of the present invention. The computer-readable storage medium is e.g. part of an infusion pump, an infusion station to be connected to multiple infusion pumps, or a server to be connected to multiple infusion pumps. The improved drug library comprises a list of drugs in order to aid programming of infusion pumps. Upon programming of an infusion pump via the display 2, at least one first infusion parameter is input for the at least one drug to be administered. In the present example, the at least one first infusion parameter is input via display keys 2. As an example, a flow rate in ml / h is depicted as the at least one first infusion parameter. The at least one first infusion parameter can take on a value in a certain range 22 (here given in ml / h), ranging from a minimal to a maximal value. The improved drug library of the present example provides a hard lower limit 231 and a hard upper limit 232 for the at least one first infusion parameter for all drugs or for a subset of the drugs listed in the drug library. Values for the first infusion parameter below the hard lower limit 231 and above the hard upper limit 232 are not admissible. As the hard limits 231 , 232 are set such that severe implications for patient health through over- or underdosing are being prevented. Additionally, the improved drug library comprises a soft lower limit 241 and a soft upper limit 242 for the at least one first infusion parameter. These soft limits 241 , 242 are mere recommendations based on experience and known practice and as such, can be crossed. As depicted in the present example, the programmed value 26 is smaller than the soft lower limit 241 , but larger than the hard lower limit 231. Hence, programming and subsequent infusion with this value of the at least first parameter can be continued after issuing of an alert 27 and additional confirmation by a user. In addition to these basic hard and soft limits 231 , 232, 241 , 242, the improved drug library comprises an improved lower limit 251 and an improved upper limit 252. These are determined based on the typical usage pattern, in particular, they are identical to the lower limit and upper limit, respectively, determined based on the obtained infusion settings, in particular, based on the obtained at least one first infusion parameter and, optionally, one or more second infusion parameters. Thus, the improved upper and lower limits 251 , 252 are determined based on clinical practice and are more specific as the provided hard and soft limits 231 , 232, 241 , 242, in so far as they also take into account one or more second infusion parameters. The improved upper and lower limits 251 , 252 hence enable dose-error reduction based on actual clinical practice in a more specific way, thus also reducing alert fatigue and increasing patient safety.

[0034] List of Reference Numerals

[0035] 1 method for improving a drug library

[0036] 101 obtaining infusion settings 102 determining typical usage pattern

[0037] 103 adding information to drug library

[0038] 2 display

[0039] 21 input keys

[0040] 22 input range for first infusion parameter 231 hard lower limit for first infusion parameter

[0041] 232 hard upper limit for first infusion parameter

[0042] 241 soft lower limit for first infusion parameter

[0043] 242 soft upper limit for first infusion parameter

[0044] 251 improved lower limit for first infusion parameter 252 improved upper limit for first infusion parameter

[0045] 26 programmed value of first infusion parameter

[0046] 27 alert

Claims

Claims1 . A method (1) for improving a drug library for an infusion pump, wherein the method (1) comprises: obtaining (101) infusion settings for at least one drug from at least one infusion pump, wherein the infusion settings comprise at least one input parameter the at least one pump is programmed with for at least one infusion with the at least one drug and / or wherein the infusion settings comprise information on at least one component connected to the at least one pump for at least one infusion with the at least one drug; and determining (102) a typical usage pattern for the at least one drug based on the obtained infusion settings for the at least one drug.

2. The method (1) according to claim 1 , characterized in that obtaining (101) infusion settings comprises obtaining at least one first infusion parameter the at least one pump is programmed with for at least one infusion with the at least one drug, wherein the at least one first infusion parameter determines the amount of the at least one drug to be infused.

3. The method (1) according to claim 2, characterized in that the first infusion parameter comprises a bolus, a dose rate and / or a flow rate.

4. The method (1) according to one of the preceding claims, characterized in that determining (102) a typical usage pattern for the at least one drug based on the obtained infusion settings comprises determining a typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug.

5. The method (1) according to one of the preceding claims, characterized in that obtaining (101) infusion settings comprises obtaining at least one second infusion parameter the at least one pump is programmed with for at least one infusion with the at least one drug, wherein the at least one second infusion parameter carries information on the patient to be infused or on the location of the infusion.

6. The method (1) according to claim 2, characterized in that the second infusion parameter comprises a ward, a patient weight, a patient body surface area, a patient age and / or a patient sex.

7. The method (1) according to claim 6, characterized in that determining (102) a typical usage pattern for the at least one drug based on the obtained infusion settings comprises determining a typical value, a typical range, an upper limit and / or a lower limit for the at least one first infusion parameter for the at least one drug depending on the at least one second infusion parameter.

8. The method (1) according to any of the preceding claims, characterized in that a machine-learning model is being used to determine (102) the typical usage pattern for the at least one drug based on the obtained infusion settings.

9. The method (1) according to any of the preceding claims, characterized in that the method comprises the step of improving (103) a drug library by adding information based on the determined typical usage pattern to the drug library.

10. The method (1) according to claim 9, characterized in that improving (103) the drug library comprises adding an improved upper limit (252) and / or an improved lower limit (251) for the at least one first infusion parameter for the at least one drug to the drug library, wherein the improved upper limit (252) and / or the improved lower limit (251) are determined based on the typical usage pattern; and / or wherein improving the drug library comprises adding an improved upper limit (252) and / or an improved lower limit (251) for the at least one first infusion parameter for the at least one drug depending on the at least one second infusion parameter to the drug library, wherein the improved upper limit (252) and / or the improved lower limit (251) are determined based on the typical usage pattern.

11. The method (1) according to any of the preceding claims, characterized in that the method (1) is executed on the at least one pump and the infusion settings are obtained (101) from infusions on the at least one pump.

12. The method (1) according to any of the claims 1 to 11 , characterized in that the method is executed on a server, wherein the server is configured to be connected to a plurality of infusion pumps and infusion settings are obtained (101) from infusions on the plurality of infusion pumps.

13. An infusion pump comprising a processor, characterized in that the processor is configured to carry out the method according to claim 12.

14. A server configured to be connected to a plurality of infusion pumps, characterized in that the server is configured to carry out the method according to claim 13.

15. A computer-readable storage medium, characterized in that the computer-readable storage medium comprises an improved drug library obtained through the method according to any of the claims 9 to 12.

Images