Compatible Asymmetric Drug Cost Elements in a Control System for Drug Supply

Abstract

An exemplary embodiment provides a drug delivery device that uses a cost function in its control system to determine drug dosage. The cost function can have a drug cost element and a performance cost element. An exemplary embodiment can use a cost function with a drug cost element that scales in an asymmetric manner for different ranges of inputs (i.e., different candidate drug dosages). The change in scaling for different input ranges provides added flexibility to tailor the drug cost element to the user and thus provide better management of drug delivery to the user and better compliance with performance targets. An exemplary embodiment can use a cost function with a drug cost element (such as an insulin cost element) of zero for the candidate dosage for a range of candidate dosages (e.g., below a reference dosage).

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims the benefit of U.S. Provisional Patent Application No. 63 / 165,252, filed on March 24, 2021, and U.S. Provisional Patent Application No. 63 / 158,918, filed on March 10, 2021, the entire contents of these patent documents are hereby incorporated by reference in their entirety.

Background Art

[0002] Some control systems seek to minimize a cost function. An example of such a control system is a control device for a drug delivery device such as an automated insulin delivery (AID) device. The cost function for an AID device typically weighs the risk of insulin under - delivery or over - delivery against the risk of glucose excursions below or above a control target. In some AID devices, the cost function sums a glucose cost element and an insulin cost element. The glucose element captures the magnitude of glucose excursions above and / or below a control target predicted with a candidate dosage, and the insulin element captures the magnitude of insulin above or below a standard dosage (such as a basal dosage) that would be delivered with the candidate dosage. The control system applies the cost function to each candidate dosage of insulin and selects the candidate dosage with the minimum cost.

Summary of the Invention

Problems to be Solved by the Invention

[0003] The AID system assumes that the basal dosage will maintain the blood glucose concentration at the target blood glucose concentration. Unfortunately, this assumption does not hold for many users. Formulations for standard basal dosages (e.g., basal dosages calculated from the total daily insulin (TDI)) do not match the true needs of many users. For example, let's say a user requires more insulin than the standard basal dosage. Candidate dosages above the standard basal dosage are penalized by the insulin cost element of the cost function. The insulin cost element is a quadratic expression in the formulation of the cost function. Therefore, the magnitude of the insulin cost element increases rapidly as the candidate dosage increases above the standard basal dosage. This makes it difficult to compensate for relatively small magnitudes of glucose deviation, because due to the rapidly increasing insulin cost element, the cost for an appropriate candidate dosage to correct relatively small magnitudes of glucose rapidly increases. Therefore, the system prefers relatively small changes rather than relatively large changes to the insulin dosage, and thus compensation for such relatively small magnitudes of glucose deviation can take a long time. Therefore, a user may have a persistent small magnitude of glucose deviation, which can be unhealthy for the patient. This result means that the user may have a blood glucose concentration that is persistently higher or lower than the target.

[0004] Another problem associated with the standard formulation of the cost function for the AID system is that there is a penalty for a decrease in insulin supply because there is a delta in relation to the basal dosage. This becomes a problem in cases where the user requires a rapid decrease in dosage to avoid the risk of progressing to hypoglycemia. **Means for Solving the Problem**

[0005] According to a first aspect of the invention, a drug delivery device includes a memory for storing data and computer programming instructions, and a pump for delivering a drug to a user. The drug delivery device further includes a processor for executing computer programming instructions to determine a value of a cost function for a candidate dosage for the user. The cost function has, for example, a performance cost element (for deviation of glucose from a desired target) and a drug cost element. The drug cost element is configured to be asymmetric about a threshold value, a standard basal amount, or a customized amount configured for the user. Further, the processor is configured to execute computer programming instructions to select, based on the value of the cost function for the candidate dosage, a dosage to be delivered to the user by the pump from among the candidate dosages.

[0006] The threshold amount may be an average basal dosage or a specific basal dosage for the user. The threshold amount may be a multiple of the basal dosage for the user, and the multiple is greater than 1. The multiple may be a ratio of the average blood glucose concentration over a predetermined time interval to the target blood glucose concentration. The multiple may be a ratio of the average basal dosage delivered to the user over that interval to an estimated value of the basal dosage over a predetermined interval derived from the total daily drug for the user. The step of selecting a dosage may include the step of selecting one of the candidate dosages having the minimum value of the cost function. The drug delivery device can deliver at least one of insulin, a glucagon-like peptide-1 (GLP-1) agonist, pramlintide, combinations thereof, or another type of drug. The drug cost element may be zero or substantially zero for any of the candidate dosages less than the threshold amount.

[0007] According to another aspect of the invention, a drug delivery device includes a memory for storing data and computer programming instructions, and a pump for delivering a drug to a user. The drug delivery device includes a processor for executing computer programming instructions to determine a value of a cost function for a candidate dosage of the drug for the user and to select, based on the value of the cost function for the candidate dosage, a dosage to be delivered to the user by the pump from among the candidate dosages. The cost function has a performance cost element and a drug cost element. The scaling of the drug cost element is quadratic above a first threshold and linear above a second threshold that is above the first threshold.

[0008] The scaling of the drug cost element may be linear below a first threshold. The drug cost element may have a fixed value in at least one case of a candidate dosage that is below the first threshold. The fixed value may be zero or substantially zero.

[0009] According to a further aspect of the invention, a drug delivery device includes a memory for storing data and computer programming instructions, and a pump for delivering a drug to a user. Also, the drug delivery device includes a processor for executing computer programming instructions to determine a value of a cost for a candidate dosage for the user and to select, based on the value of the cost for the candidate dosage, a dosage to be delivered to the user by the pump from among the candidate dosages for the user. The cost has a performance cost element and a drug cost element. The cost is calculated in different ways for different ranges of candidate dosages.

[0010] The cost can be calculated to be negligible for any of the candidate dosages in one of the ranges below the first threshold. The first threshold may be the average basal dosage or a specific basal dosage for the user. The first threshold may be a multiple of the average basal dosage or a specific basal dosage for the user, and the multiple may be greater than 1. The cost may include a drug cost element calculated by quadratic formulation in one of the ranges above the first threshold and above the second threshold. The cost may include a drug cost element calculated by linear formulation in one of the ranges above the second threshold and above the third threshold. The cost can be calculated by using a cost function, and the cost function can be scaled in different ways in at least two of the ranges. The cost function can include a performance cost element and a drug cost element, and the drug cost element can be different in at least two of the ranges so that the cost function is scaled in different ways in at least two of the ranges. The cost function can have a drug cost element, and the cost can be determined for the first of the ranges by using a formula for the drug cost that is different from the one used when determining the cost for the second of the ranges. The cost can be determined by different cost functions for each of the ranges. The drug may be one of insulin, a glucagon-like peptide-1 (GLP-1) agonist, pramlintide, their co-formulations, or another type of drug.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8A

Figure 8B

Figure 9

DETAILED DESCRIPTION OF THE INVENTION

[0012] The exemplary embodiment relates to a drug delivery device that uses a cost function in its control system to determine a drug dosage. The cost function can have a drug cost element and a performance cost element. The exemplary embodiment can use a cost function having a drug cost element that scales in an asymmetric manner for different ranges of inputs (i.e., different candidate drug dosages). The variation in scaling in different input ranges provides added flexibility for adapting the drug cost element to the user and thus for providing relatively good management of drug delivery to the user and relatively good compliance with performance targets.

[0013] The drug delivery device of the exemplary embodiment can supply any of a variety of drugs. The drugs supplied by the drug delivery device of the exemplary embodiment can include, without limitation, insulin, glucagon-like peptide-1 (GLP-1) agonists, pramlintide, co-formulations of two or more of these, glucagon, hormonal agents, pain management agents, chemotherapeutic agents, antibiotics, antiviral drugs, anticoagulants, blood coagulants, antidepressants, antiepileptic drugs, antipsychotics, antihypertensives, statins, therapeutic agents, and pharmaceuticals.

[0014] Exemplary embodiments can provide a change to the cost function of a drug delivery device in relation to a conventional drug delivery device. For example, an exemplary embodiment can use a cost function having a drug cost element of zero (such as an insulin cost element) for candidate dosages within a range of candidate dosages (e.g., less than a reference dosage). Alternatively, instead, the drug cost element can have a negligible value such as being substantially zero for a range of candidate dosages. In the case of an insulin or GLP-1 agonist delivery device, this reduces the drug cost element such that there is a reduced penalty for a decrease in drug dosage, and in some instances, makes it relatively easier to avoid hypoglycemia. This change to the drug cost element is configured to reflect the view that the consequences of hypoglycemia are generally more dangerous than hypoglycemia itself.

[0015] In other exemplary embodiments, the drug cost element can have a fixed positive value (i.e., a constant value) for candidate dosages within a range. In still other exemplary embodiments, the drug cost element can be a linear expression in the formulation of the cost function for candidate dosages within a range of candidate dosages. The net effect of these changes to the conventional drug cost element is to reduce the penalty resulting from the drug cost element at candidate dosages less than the standard dosage in relation to the conventional drug cost element.

[0016] The cost function does not need to use a singular or single - type representation for the drug cost elements over the range of all possible candidate dosages. The drug cost representation can be a constant, linear, quadratic, or non - quadratic exponential representation. The drug cost elements can have different representations over the range of candidate dosages. For example, a drug cost element can be a constant for a first range of candidate dosages, a non - constant linear representation for a second range of candidate dosages, and a quadratic representation for a third range of candidate dosages. In some embodiments, the drug cost element representation can be of a single type (e.g., constant, linear, quadratic, or exponential) over the range of candidate dosages, but can have different formulations. For example, a drug cost element can be expressed as x + 2 for a first range and 2x+3 for a different range, where x is a variable such as a delta in relation to a base dosage.

[0017] In some exemplary embodiments, different formulations of the cost function may be used for different ranges, or, in some cases, different cost functions may be used for different ranges. The differences do not necessarily have to be due only to changes to the drug cost elements. The scaling of the drug cost elements does not have to be the same over all possible candidate dosages. The cost function can be asymmetric across a threshold or reference dosage.

[0018] In addition, the exemplary embodiments can also change the reference dosage used when determining the drug cost element. First, the reference dosage can be set to be greater than the basal dosage. Thus, the reference dosage may be relatively well-suited for users having a greater than normal drug need, such as a relatively large insulin need. Second, the reference dosage can be customized to the user's actual average basal amount (such as the average in the most recent interval). As a result, the value of the reference dosage is relatively well-suited to the user. Third, the reference dosage can be customized based on the user's most recent actual discrepancy between the basal supply of the drug and the bolus supply of the drug. The reference dosage is not limited to be 50% of the TDI in the cost calculation.

[0019] Figure 1 depicts an exemplary drug delivery system 100 suitable for delivering drugs such as insulin, GLP-1 agonists, or other drugs such as those detailed above to user 108 according to the exemplary embodiments. The drug delivery system 100 includes a drug delivery device 102. The drug delivery device 102 may be a wearable device worn on the body of user 108. The drug delivery device 102 can be directly coupled to the user (e.g., directly attached to a body part and / or skin of user 108 via an adhesive or the like). In one example, the surface of the drug delivery device 102 may include an adhesive to facilitate attachment to user 108.

[0020] The drug supply device 102 can include a controller 110. The controller 110 can be implemented in hardware, software, or any combination thereof. The controller 110 can be, for example, a microprocessor, a logic circuit, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or a microcontroller coupled to a memory. The controller 110 can maintain not only the date and time but also other functions (such as calculations or the like). The controller 110 is operable to execute a control application 116 stored in a storage 114 that enables the controller 110 to implement a control system for controlling the operation of the drug supply device 102. The control application 116 can control the supply of drugs to the user 108 as described herein. The storage 114 can hold user histories 111 such as the history of automated drug supply, the history of bolus drug supply, meal event history, exercise event history, sensor data, and the like. In addition, the controller 110 can be operable to receive data or information. The storage 114 can include both primary memory and secondary memory. The storage 114 can include random access memory (RAM), read only memory (ROM), optical storage, magnetic storage, removable storage media, semiconductor storage, or the like.

[0021] The drug supply device 102 can, of course, include a reservoir 112 for storing the drug to be supplied to the user 108. It can provide a fluid path to the user 108, and the drug supply device 102 can discharge the drug from the reservoir 112 to supply the drug to the user 108 via the fluid path. The fluid path can include, for example, a tube that couples the drug supply device 102 to the user 108 (such as a tube that couples a cannula to the reservoir 112).

[0022] For example, there may be one or more communication links between one or more devices physically separated from the drug supply device 102, including the management device 104 and / or the sensor 106 of the user and / or the caregiver of the user. The communication link may include any known communication protocol or standard, such as Bluetooth®, Wi-Fi, short-range communication standards, cellular standards, or any other wireless protocol, and any wired or wireless communication link operating according to such protocol or standard. Also, the drug supply device 102 may include a user interface 117, such as an integrated display device for displaying information to the user 108 and, in some embodiments, for receiving information from the user 108. The user interface 117 may include a touch screen and / or one or more input devices such as buttons, knobs, or keyboards.

[0023] The drug supply device 102 can interface with a network 122. The network 122 may include a local area network (LAN), a wide area network (WAN), or a combination thereof. The computing device 126 can interface with the network and the computing device can communicate with the insulin supply device 102.

[0024] The drug supply system 100 can include one or more sensors 106 for detecting the level of one or more specimens. The one or more sensors 106 can be coupled to the user 108, for example, by an adhesive or the like, and can provide information or data regarding one or more medical conditions and / or physical attributes of the user 108. The one or more sensors 106, in some exemplary embodiments, can provide periodic blood glucose concentration measurements and can be another type of device or sensor that provides continuous glucose monitoring (CGM) or blood glucose measurements. The one or more sensors 106 may be physically separate from the drug delivery device 102 or may be an integrated element thereof. The one or more sensors 106 can provide data to the controller 110 that notifies the controller 110 of one or more measurements or detected specimen levels of the user 108. The information or data provided by the one or more sensors 106 can be used to regulate the drug delivery operation of the drug delivery device 102.

[0025] In addition, the drug supply system 100 can also include a management device 104. In some embodiments, the management device 104 is not required, rather, the drug supply device 102 can manage itself. The management device 104 can be a special-purpose device such as a dedicated personal diabetes manager (PDM) device. The management device 104 can be a programmed general-purpose device such as any portable electronic device including a dedicated controller such as, for example, a processor, a microcontroller, or the like. The management device 104 can be used to program or adjust the operation of the drug supply device 102 and / or the sensor 104. The management device 104 can be, for example, any portable electronic device including a dedicated device, a smartphone, a smartwatch, or a tablet. In the example depicted, the management device 104 can include a processor 119 and a storage 118. The processor 119 can execute a process for managing and controlling the supply of drugs to the user 108. Also, the processor 119 can be operable to execute programming code stored within the storage 118. For example, the storage can be operable to store one or more control applications 120 for execution by the processor 119. One or more control applications 120 (or, 116) can be responsible for controlling the drug supply device 102, such as, for example, the supply of insulin to the user 108. The storage 118 can store one or more control applications 120, a history 121 such as those described above for the drug supply device 102, and other data and / or programs.

[0026] The management device 104 may include a user interface (UI) 123 for communicating with the user 108. The user interface 123 can include a display such as a touch screen for displaying information. Also, the touch screen can be used to receive input when it is a touch screen. Further, the user interface 123 can also include input elements such as a keyboard, buttons, one or more knobs, or the like. The user interface 123 can be used to consider data or history or provide input for purposes such as generating a change in the basic drug dosage, supplying a drug bolus, or changing one or more parameters used by the control applications 116 / 120.

[0027] The management device 104 can interface with a network 124 such as a LAN or WAN, or a combination of such networks. The management device 104 can communicate with one or more servers or cloud services 128 over the network 124.

[0028] Other devices, such as smartwatch 130, fitness monitor 132, and / or another wearable device 134, may form part of the drug delivery system 100. These devices can communicate with the drug delivery device 102 to receive information from and / or issue commands to the drug delivery device 102. These devices 130, 132, and 134 can execute computer programming instructions to perform some of the control functions otherwise performed by the controller 110 or the processor 119. These devices 130, 132, and 134 can include a display for displaying information such as analyte levels, such as current blood glucose levels, the drug in the filled state, the drug delivery history, and the like. The display can display a user interface for providing input, for example, to effect a change in the basal drug dosage, the delivery of a drug bolus, or a change in one or more parameters used by the control application 116 / 120. Also, these devices 130, 132, and 134 can have a communication connection with the sensor 106 to directly receive analyte data.

[0029] The control system of the exemplary embodiment depends on the cost function as described above. The control system attempts to minimize the total penalty of the cost function over a wide range of possible candidate drug dosages. FIG. 2 depicts a flowchart 200 of exemplary steps that may be performed when selecting the next dosage to be supplied to the user in the exemplary embodiment. At 202, the control system selects candidate dosages from which a dosage can then be selected. At 204, by using the cost function, the cost is calculated for each candidate dosage. The cost function has the candidate dosage as an input and the cost for the candidate dosage as an output. At 206, the candidate dosage having the best cost is selected. Depending on the manner in which the cost function is configured, the best cost may be the minimum value or the maximum value. For the purposes of the following description, it is assumed that the candidate dosage with the minimum cost is the one with the best cost. At 208, the selected dosage is supplied to the user. A control signal can be generated and transmitted from the controller 110 operating the control application 116 to the pump 113 so that the pump 113 supplies the selected drug dosage to the user. Thereafter, the process 200 can be repeated for another subsequent cycle (e.g., every 5 minutes).

[0030] To understand the manner in which the exemplary embodiment modifies the cost function for the drug delivery device, it is useful to observe the conventional cost function for the drug device. An example of a cost function for a drug delivery device is the cost function for an insulin delivery device. The conventional cost function for an insulin delivery device is as follows:

Equation

[0031] Figure 3 depicts a flowchart 300 of exemplary steps that may be performed to calculate a cost function such as that detailed in Equation 1. At 302, a drug cost element is calculated. In the case of the conventional cost function of this cost example, the performance cost element is

Equation

Equation

[0032] Conventional cost functions for drug supply devices tend to define the drug cost function as a linear cost or a quadratic cost. In other words, the drug cost function is expressed as a linear expression or a quadratic expression. The above-mentioned conventional cost function has a quadratic drug cost element, and the reason is that

Number

[0033] As can be observed from FIG. 4, according to both the conventional linear drug cost element and the conventional secondary drug element, there is a penalty for the supply of insulin dosage below the standard basal dosage. The exemplary embodiments described hereinafter and in subsequent figures recognize that such a penalty may not be desirable. As described above, such a penalty suppresses the supply of insulin dosage below the standard basal dosage. Thus, according to some exemplary embodiments, the drug cost element is zero or substantially zero (i.e., slightly above zero but still negligible) for a required dosage less than the standard basal dosage (i.e., the reference dosage described above). This formulation of the drug cost element allows the control system to freely reduce the drug supply as needed to change the predicted performance metric to the set point. This control scheme can be relatively aggressive in reducing the drug supply while maintaining compensation for an increase in drug supply in relation to the standard basal dosage.

[0034] FIG. 5 depicts a flowchart 500 of exemplary steps that may be performed in an exemplary embodiment to achieve zero cost associated with a cost function. At 502, a check is performed to determine whether a candidate dosage is less than a threshold, such as less than the standard basal dosage. At 504, if the candidate dosage is less than the threshold, the drug cost element is set to zero or a value that is substantially zero. At 506, if the candidate dosage is greater than or equal to the threshold, the drug cost element is calculated by using the formulation defined in the cost function.

[0035] An exemplary formulation of a cost function that provides a zero-cost drug cost element value for an insulin supply threshold or less than the standard basal dosage is as follows:

Equation

[0036] The reference value used as a threshold at which the drug cost element is zero or substantially zero below it may be a value other than the standard basal dosage. Often, a user may have greater drug needs than the standard basal formulation. Thus, by fixing the supply at the standard basal dosage, persistent low-level performance deviations above the blood glucose target cannot be removed. Accordingly, some exemplary embodiments can set the threshold based on the average positive performance deviation above the blood glucose target. This allows for a greater supply of the drug than the standard basal dosage of the drug to reduce the positive performance deviation without penalty.

[0037] One use of an adaptable threshold greater than the standard basal dosage is for an insulin delivery device. In such a use, the threshold can be calculated as shown in flowchart 600 of FIG. 6. At 602, the ratio of the user's average blood glucose concentration over a predetermined period to the target blood glucose concentration is calculated. This ratio reflects the extent to which the user's average blood glucose concentration exceeds the target blood glucose concentration. At 604, the basal dosage is multiplied by the ratio to determine the threshold.

[0038] The cost function with a modified threshold based on the average glucose deviation can be expressed as the following formula,

Equation

[0039] Alternatively, the threshold value can also be set based on the reference value of the actual basal / TDI difference of the user. Conventionally, the basal amount for the user is set as half of the user's TDI. Unfortunately, this rule of thumb does not work well for some users. To address such users, exemplary embodiments can determine the threshold value based on the user's actual basal / TDI ratio. FIG. 7 depicts a flowchart 700 of steps that can be performed to determine such a threshold value. At 702, the ratio of the average basal dosage for the user over a predetermined period (e.g., one day) to half of the user's TDI is determined. If the user has insulin needs above normal (i.e., above basal), the ratio will be greater than 1. At 704, the standard basal dosage for the user (i.e., 0.5 TDI) is multiplied by the ratio to determine the threshold reference value. This allows the threshold to exceed the basal dosage, in which case the average actual basal dosage is above the standard basal dosage.

[0040] The formulation of the cost function may be as described above for the deviation of the average glucose, but [Number] may be formulated in a different manner as follows, [Number] Here, I btotalis the total basal insulin over a predetermined interval, and I bactual(i) is the actual basal insulin in cycle i.

[0041] It may be desirable for the drug cost element in the cost function to be asymmetric between different ranges to suit the user's drug needs. In other words, the drug cost element can be scaled in different ways in different ranges. FIG. 8A depicts a flowchart 800 of an example where different scalings are used for different ranges of drug candidate dosages. In this example, at 802, the drug cost element has zero cost below a first threshold. This example is shown in plot 820 of FIG. 8B. Plot 820 shows a curve 822 with linear scaling and a curve 824 with quadratic scaling, similar to what is depicted in FIG. 4. Curve 825 represents the drug cost element over the range of candidate dosages, and in this case, the scaling changes at different basal dosage rates. Region 826 of the curve before the required insulin supply of 0.1 U / h has a flat constant value of 0, as described above in relation to FIG. 5. At 804, the scaling of curve 825 changes to quadratic scaling in region 828 extending from 0.1 U / h to 0.3 U / h. This region 828 represents the range of required dosages when the user's blood glucose concentration is near the target and the required insulin dosage slightly exceeds the basal dosage. For this region 828, quadratic scaling is well-suited, and in this case, the insulin level is rising but not rising so much as to pose a threat of hypoglycemia. However, as the candidate dosage exceeds 0.3 U / h, the quadratic scaling can become overly aggressive and increase the risk of hypoglycemia. Thus, at 806, the scaling is reduced. Specifically, the insulin cost element has linear scaling when the required insulin dosage is above a second threshold (e.g., 0.3 U / h) within region 830. As a result, the risk of hypoglycemia due to the supply of excessive insulin dosages is reduced.

[0042] The above example shows that the drug cost element and the cost function can be changed over different ranges of the input (i.e., drug dosage). More generally, it should be noted that the exemplary embodiments may use a cost function having an asymmetry with respect to the drug cost element. FIG. 9 depicts a flowchart 900 of exemplary steps that can be performed to provide asymmetry over a range. At 902, a range of candidate dosages of the drug is identified. At 904, an expression is assigned for the drug cost element for each range, a cost formulation for each range is assigned, or a separate cost function is assigned for each range. This assignment provides an asymmetry over the range of candidate dosages. This asymmetry enables the control system to relatively well customize the drug cost element to the user's needs.

[0043] Although the description herein focuses on exemplary embodiments, it should be understood that various changes in form and detail may be made without departing from the intended scope of the claims appended hereto in relation to the exemplary embodiments.

Claims

1. A drug supply device, a memory for storing data and computer programming instructions, a pump for supplying a drug to a user, and a processor for executing the computer programming instructions, and has, the processor, determines a value of a cost function for a candidate dosage for the user, the cost function having a performance cost element and a drug cost element, the drug cost element is configured to be asymmetric about a threshold amount, the threshold amount being a multiplication value of a weighting factor greater than 1 and a standard basic dosage for the user, executes the computer programming instructions to select a dosage to be supplied to the user by the pump from among the candidate dosages based on the value of the cost function for the candidate dosage. A drug supply device.

2. The drug supply device according to claim 1, wherein the standard basic dosage is an average basic dosage for the user.

3. The drug supply device according to claim 1, wherein the weighting factor is a ratio of an average blood glucose concentration in a time interval to a target blood glucose concentration.

4. The drug supply device according to claim 1, wherein the weighting factor is a ratio of an average basic dosage supplied to the user in the interval to an estimated value of the basic dosage in the interval derived from the total daily drug for the user.

5. Selecting the dosage includes selecting one of the candidate dosages having a minimum value for the cost function, the drug supply device according to claim 1.

6. The drug supply device supplies at least one of insulin, a glucagon-like peptide-1 (GLP-1) agonist, or pramlintide, the drug supply device according to claim 1.

7. A drug supply device, a memory for storing data and computer programming instructions, a pump for supplying a drug to a user, and a processor for executing the computer programming instructions, and has, the processor, determines a value of a cost function for a candidate dosage for the user, the cost function having a performance cost element and a drug cost element, the drug cost element is configured to be asymmetric about a threshold amount, and the drug cost element is substantially zero for any of the candidate dosages less than the threshold amount, A drug supply device that executes the computer programming instructions to select, from among the candidate dosages, the dosage to be supplied to the user by the pump based on the value of the cost function for the candidate dosages. **Claim 8** A drug supply device, comprising: a memory for storing data and computer programming instructions; a pump for supplying a drug to a user; a processor for executing the computer programming instructions, wherein the processor determines a value of a cost function for candidate dosages of the drug for the user, the cost function having a performance cost element and a drug cost element; the scaling of the drug cost element is quadratic above a first threshold value and linear above a second threshold value that is above the first threshold value; A drug supply device that executes the computer programming instructions to select, for the user, from among the candidate dosages, the dosage to be supplied to the user by the pump based on the value of the cost function for the candidate dosages. **Claim 9** The drug supply device according to claim 8, wherein the scaling of the drug cost element is linear below the first threshold value. **Claim 10** The drug supply device according to claim 9, wherein the drug cost element has a fixed value for at least one of the candidate dosages that is below the first threshold value. **Claim 11** The drug supply device according to claim 10, wherein the fixed value is zero or substantially zero. **Claim 12** A drug supply device, comprising: a memory for storing data and computer programming instructions; a pump for supplying a drug to a user; a processor for executing the computer programming instructions, wherein the processor determines a value of a cost for candidate dosages for the user, the cost having a performance cost element and a drug cost element; the cost is calculated in different ways for different ranges of the candidate dosages, and the cost is calculated to be negligible for any of the candidate dosages in one of the ranges below a first threshold value; A drug supply device that executes the computer programming instructions to select, for the user, from among the candidate dosages, the dosage to be supplied to the user by the pump based on the value of the cost for the candidate dosages.

13. The drug supply device according to claim 12, wherein the first threshold value is the average basic dosage or a specific basic dosage for the user.

14. The drug supply device according to claim 13, wherein the first threshold value is a multiplication value of a multiple and the average basic dosage or the specific basic dosage for the user, and the multiple is greater than 1.

15. The drug supply device according to claim 12, wherein the cost includes the drug cost element calculated by quadratic formulation in one of the ranges exceeding the second threshold value that exceeds the first threshold value.

16. The drug supply device according to claim 15, wherein the cost includes the drug cost element calculated by linear formulation in one of the ranges exceeding the third threshold value that exceeds the second threshold value.

17. The drug supply device according to claim 12, wherein the cost is calculated using different cost functions and cost function scales in at least two of the ranges.

18. The drug supply device according to claim 17, wherein the cost function includes a performance cost element and a drug cost element, and the drug cost element is different in at least two of the ranges so that the drug cost element scales in different ways in at least two of the ranges.

19. The drug supply device according to claim 17, wherein the cost function has a drug cost element, and the cost is determined for the first of the ranges using a formula for drug cost that is different from the one used when determining the cost for the second of the ranges. 【Claim 】 The drug supply device according to claim 12, wherein the cost is determined by different cost functions for each range.

21. The drug supply device according to claim 12, wherein the drug is one of insulin, glucagon-like peptide-1 (GLP-1) agonist, or pramlintide.

Images