System and method for compliance evaluation of regimens

Abstract

A diabetes management system adapted to determine compliance for a subject being treated according to a basal insulin regimen, the system being adapted to receive regimen data, BG data, and insulin injection data. If one or more insulin injections have not been received according to a prescribed regimen and are thus missing, the system is adapted to calculate an expected dose response BG value for each missing injection. By comparing the received BG data corresponding to the missing insulin injection, it can be determined whether the subject is compliant with the basal insulin regimen for a given confidence interval.

Description

Technical Field

[0001] The present disclosure generally relates to systems and methods for assisting patients and healthcare providers in the management of insulin therapy for diabetes. In certain aspects, the invention relates to systems and methods suitable for use in a diabetes management system that provides an optimized individualized basal insulin titration regimen.

Background Art

[0002] True diabetes mellitus (DM) is an insulin secretion disorder leading to hyperglycemia and varying degrees of peripheral insulin resistance. Type 2 true diabetes mellitus is characterized by a progressive impairment of normal physiological insulin secretion. In healthy individuals, basal insulin secretion by pancreatic β-cells occurs continuously, maintaining a steady blood glucose level over a long period between meals. Even in healthy individuals, there is meal-induced secretion where insulin is rapidly released with an initial phase 1 spike in response to a meal, followed by a prolonged insulin secretion that returns to the basal level after 2 - 3 hours. Persistent poorly controlled hyperglycemia over several years can cause multiple health complications. True diabetes mellitus is one of the main causes of early morbidity and mortality throughout the world.

[0003] Effective control of blood / plasma glucose can prevent or delay many of these complications, but once established, they may not be reversible. Therefore, achieving good glycemic control in the effort to prevent diabetes complications is a major goal in the treatment of type 1 and type 2 diabetes. To administer insulin drug treatment regimens, adjustable step sizes, as well as smart titrators using physiological parameter estimation and a predetermined fasting blood glucose target value, have been developed.

[0004] There are many non-insulin treatment options for diabetes, but as the disease progresses, the most reliable response is usually with insulin. In particular, since diabetes is associated with progressive β-cell loss, many patients, especially those with long-term disease, need to ultimately transition to insulin because, depending on the degree of hyperglycemia (e.g., HbA1c ≥ 8.5%), other medications are unlikely to provide sufficient benefit.

[0005] The ideal insulin regimen aims to mimic the physiological profile of insulin secretion as closely as possible. The insulin profile has two main components: continuous basal secretion and a post-meal spike due to food intake. Basal secretion controls glucose overnight and during fasting, while the post-meal spike controls post-meal hyperglycemia.

[0006] Based on onset and duration of action, injectable formulations are broadly classified into basal (long-acting analogs [e.g., insulin detemir and insulin glargine] and ultra-long-acting analogs [e.g., insulin degludec (for once-daily dosing) and insulin icodec (intended for once-weekly dosing)], and intermediate-acting insulin [e.g., isophane insulin, etc.], as well as prandial (rapid-acting analogs [e.g., insulin aspart, insulin glulisine, and insulin lispro]). Premixed insulin formulations incorporate both a basal insulin component and a prandial insulin component.

[0007] Basal insulin is typically the only (initial) insulin treatment for type 2 diabetes, while for type 1 diabetes, basal insulin can be used in combination with rapid-acting insulin before meals.

[0008] Generally, to determine the optimal basal insulin dose for a given patient, the patient starts at an initial safe recommended low dose of basal insulin (typically, 10 U / day) and is titrated, and then increased until the fasting plasma glucose (FPG) is within the target range, generally 80 - 130 mg / dL at a dose typically of 40 - 70 U / day for type 2 diabetes. Alternatively, fasting blood glucose (FBG) values can be used. Dose adjustments should be more moderate and less frequent as the target is approached, and if any hypoglycemia occurs, a dose reduction adjustment is recommended.

[0009] However, all of this is a major barrier not only to starting insulin therapy, which is necessary to adjust treatment to individual needs and maintain glycemic control, but also to optimizing the dose and intensifying the regimen.

[0010] One challenge for effective titration is treatment adherence. Failure to initiate, optimize, and intensify basal insulin therapy is caused by clinical inertia leading to inadequate treatment adherence. This is typically related to forgetfulness, recognition of the need for the medication, fear of hypoglycemia, and lack of confidence or uncertainty regarding insulin titration.

[0011] Connected injection devices provide patients and clinicians with insights regarding treatment adherence and changes that can be implemented immediately for treatment. Some decision support tools, such as insulin titration support applications, use input from connected devices to provide reliable and safe guidance. Since algorithms use data to calculate safe and efficient recommended doses for patients, it is important that the data from connected devices be complete. Accordingly, it is important to know whether the recommended dose has been administered.

[0012] The nature of the injection data is that they are sparse, meaning that they represent whether the patient adheres to the treatment regimen. Thus, "missing" data points represent non - adherence and thereby represent an important input to the algorithm.

[0013] In the context of bolus calculation, the problem of missing dose data has been addressed, for example, in Patent Document 1, which discloses that the evaluation of BG values may indicate whether a given recommended dose was actually taken or not taken because, for example, the expected meal was skipped. The evaluation of BG values is based on general considerations regarding the set BG target range.

[0014] Patent Document 2 discloses a data collection device used in combination with an insulin infusion pump. The device is adapted to replace missing data based on BG data and using predictive learning. The missing data may relate to non - insulin drugs such as analgesics, allergy medications, or cold medications.

[0015] Patent Document 3 discloses a diabetes management system adapted to determine adherence for a subject being treated according to a basal insulin regimen. The system is adapted to receive regimen data that sets the current dose size and prescribed injection periodicity, and a plurality of fasting blood glucose (FBG) measurements of the subject taken over time, thereby establishing an FBG history (FBGH), and a plurality of insulin injection data sets during all or a portion of the time elapsed, thereby establishing an insulin injection history (IH). The system includes one or more processors and a memory that, when executed by one or more processors, includes instructions for performing data optimization based on FBGH and IH by processing missing data in which one or more temporal gaps in the subject data are interpolated by resampling the subject data at a predetermined time interval.

[0016] In view of the problem of compliance with treatment considered above, and thus the lack of dosage data during basal insulin titration, it is an object of the present invention to provide a method and system that enable more efficient and safe insulin titration despite the lack of dosage data.

Prior Art Documents

Patent Documents

[0017]

Patent Document 1

Patent Document 2

Patent Document 3

Summary of the Invention

[0018] In the disclosure of the present invention, embodiments and aspects are described that address one or more of the above objectives, or objectives that are apparent not only from the following disclosure but also from the description of the exemplary embodiments.

[0019] In summary, the present invention is based on the recognition that "missing" insulin dosage data points may occur due to technical problems such as loss of communication to the injection device, or due to the patient not properly installing an add-on device connected to their injection device. Accordingly, if it is possible to distinguish non-compliance from missing data points, this would be beneficial for a given titration regimen.

[0020] Accordingly, in a first aspect of the present invention, there is provided a diabetes management system adapted to determine compliance for a subject being treated according to a basal insulin regimen. The system receives regimen data that sets a current dose size and a prescribed injection periodicity, and a plurality of fasting blood glucose (FBG) measurements of the subject that are obtained over time and thereby establish a Fasting Blood Glucose History (FBGH), each respective glucose measurement among the plurality of glucose measurements including (i) an FBG value, and (ii) a corresponding FBG timestamp, and a plurality of insulin injection data sets during all or a portion of the time elapsed, thereby establishing an Insulin Injection History (IH), each injection data set including (i) an injection amount, and (ii) an injection timestamp representing at which point in time during the time elapsed the injection occurred. The system comprises one or more processors and a memory, the memory including instructions that, when executed by the one or more processors, implement a method including determining whether, for a period of time, one or more insulin injection data sets have been received (logged) according to a prescribed injection regimen and are thus not missing. If an insulin injection data set(s) is missing, the method includes calculating, for each missing injection, an expected dose-response FBG value based on the received FBGH and IH and using a dose-response algorithm, and determining, for a given confidence interval, whether the received FBG value corresponding to the missing insulin injection corresponds to the calculated dose-response FBG value. If the received FBG value corresponds to the calculated dose-response FBG value, the system determines that the subject is compliant with the basal insulin regimen, or if the received FBG value does not correspond to the calculated dose-response FBG value, the system determines that the subject is not compliant with the basal insulin regimen. The FBG value may be derived from received CGM data.

[0021] In this way, using the knowledge about FBGH and IH, it is possible to determine, according to the titration regimen, whether an unlogged (i.e., missing) injection has actually been performed by the patient.

[0022] In an exemplary embodiment, the predicted dose response FBG value is based on the received FBGH and IH as well as regimen dose data, i.e., for unlogged injections, the calculation is based on the regimen data that sets the time and dose size of the unlogged injection.

[0023] In an exemplary embodiment, the diabetes management system is adapted to further provide insulin dose recommendations for a subject, and the method includes an additional step of receiving a dose guidance request (DGR), and a step of determining whether the subject has adhered to the regimen regarding insulin injection for a predetermined length of time prior to receiving the DGR. If the subject has adhered (with or without missing dose data), the system provides an updated dose recommendation based on the received FBGH and IH, or if the subject has not adhered, the system maintains the current dose recommendation. The predetermined length of time prior to receiving the DGR is the time since the last previous DGR was made.

[0024] In a second aspect of the present invention, a method is provided for determining compliance for a subject being treated according to a basal insulin regimen. The method includes obtaining regimen data that sets a current dose size and a prescribed injection periodicity, obtaining a plurality of fasting blood glucose (FBG) measurements of the subject taken over time thereby establishing a FBG history (FBGH), wherein each respective glucose measurement among the plurality of glucose measurements includes (i) an FBG value and (ii) a corresponding FBG timestamp, obtaining a plurality of insulin injection data sets during all or a portion of the time elapsed between the FBG measurements thereby establishing an insulin injection history (IH), wherein each injection data set includes (i) an injection amount and (ii) an injection timestamp representing at what point in time during the time elapsed the injection occurred. The method includes a further step of determining during a period of time whether one or more insulin injection data sets have not been received and are thus missing according to the prescribed injection regimen. If an insulin injection data set(s) is missing, the method includes calculating for each missing injection an expected dose response FBG value based on the received FBGH and IH and using a dose response algorithm, and for a given confidence interval, determining whether the received FBG values corresponding to the missing insulin injections correspond to the calculated dose response FBG values. If the received FBG values correspond to the calculated dose response FBG values, the method determines that the subject is compliant with the basal insulin regimen, or if the received FBG values do not correspond to the calculated dose response FBG values, the method determines that the subject is not compliant with the basal insulin regimen. The FBG values may be derived from received CGM data.

[0025] The expected dose response FBG values may be based on the received FBGH and IH as well as regimen dose data.

[0026] The method may be further adapted to provide insulin dosage recommendations for a subject, the method comprising an additional step of receiving a dosage guidance request (DGR), an additional step of determining whether the subject has adhered to the regimen with respect to insulin injection for a predetermined length of time prior to receiving the DGR, an additional step of providing an updated dosage recommendation based on the received FBGH and IH if the subject has adhered, or an additional step of maintaining the current dosage recommendation if the subject has not adhered. The predetermined length of time prior to receiving the DGR is the time since the last previous DGR was generated.

[0027] Embodiments of the present invention are further described below with reference to the drawings.

Brief Description of the Drawings

[0028]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Modes for Carrying Out the Invention

[0029] Overall, a diabetes dosage guidance system is provided that serves people with diabetes by generating recommended insulin dosages. In such systems, a given algorithm is used to generate recommended insulin dosages and treatment advice for diabetic patients based on BG and insulin administration history.

[0030] Essentially, such systems include a back-end engine ("engine"), which is a primary aspect of the present invention, used in combination with systems that interact in the form of clients and operating systems.

[0031] From the perspective of the engine, a client is a software component that requests dosage guidance. The client collects the necessary data (e.g., CGM data, insulin dosage data, patient parameters) and requests dosage guidance from the engine. The client then receives a response from the engine.

[0032] On a small local scale, the engine may be directly executed as an app on a given user's smartphone, and thus may be a self - contained application that includes both the client and the engine. Alternatively, the system setup may be designed to be implemented as a back - end engine adapted to be used as part of a large - scale cloud - based diabetes management system. Such cloud - based systems allow the engine to always be up - to - date (e.g., in contrast to app - based systems that fully operate on a patient's smartphone, etc.), enable the implementation of advanced methods such as machine learning and artificial intelligence, and allow data to be used in combination with other services in a larger "digital health" setting. Such cloud - based systems will ideally handle the demands of a large number of patients for dosage recommendations.

[0033] A "complete" engine may be designed to accommodate all computing modalities, but it may be desirable to split the engine into local and cloud versions to enable a patient-proximate, daily dosing guidance system to operate independently of any reliance on cloud computing. For example, when a user makes a dosing guidance request via a client app, the request is sent to the cloud engine and dosing recommendations are returned. If cloud access is unavailable, the client app will perform dosing recommendation calculations using the current local algorithm. Depending on the user's app settings, the user may or may not be notified.

[0034] To provide safe and efficient dosing guidance recommendations, it is important that the data on which the recommendation calculations are based be as complete as possible.

[0035] Accordingly, in a first aspect, the present invention provides a diabetes management system adapted to determine treatment compliance for a subject being treated according to a basal insulin regimen.

[0036] More specifically, the present invention aims to distinguish non-compliance periods and missing data periods in a time series of paired injection and glucose data. This is done by using periods of known compliance (i.e., periods where data points are available) to build a model of the dosing response and looking at the corresponding glucose data to calculate the probability of new missing injection data points due to non-compliance or due to true missing injection data points.

[0037] The advantage of this solution is that it enables more stable and accurate dosing guidance in settings where insulin pen connectivity is used. By calculating the probability of compliance during a missing data period from a connected pen, dosing guidance that would otherwise be delayed can be provided as soon as the connection is restored.

[0038] More specifically, an exemplary diabetes management system is provided that is adapted to determine compliance for a subject being treated according to a basal insulin regimen. The system is adapted to receive (a) regimen data that sets a current dose size and a prescribed injection periodicity, (b) a plurality of fasting blood glucose (FBG) measurements of the subject taken over time thereby establishing a FBG history (FBGH), each respective glucose measurement of the plurality of glucose measurements including (i) an FBG value and (ii) a corresponding FBG timestamp, and (c) a plurality of insulin injection data sets during all or a portion of the time elapsed thereby establishing an insulin injection history (IH), each injection data set including (i) an injection amount and (ii) an injection timestamp representing at what point in time during the time elapsed the injection occurred.

[0039] The system comprises one or more processors and a memory storing instructions that, when executed by the one or more processors, implement a method including (A) determining that, for a period of time, one or more insulin injection data sets have not been received according to a prescribed injection regimen and are thus missing, and (B) calculating, for each missing injection, a predicted dose response FBG value using a dose response algorithm based on the received FBGH and IH.

[0040] The method further includes the step of (C) determining, for a given confidence interval, whether a received FBG value corresponding to a missing insulin injection corresponds to the calculated dose response FBG value, and (i) determining that the subject is compliant with the basal insulin regimen if the received FBG value corresponds to the calculated dose response FBG value, or (ii) determining that the subject is not compliant with the basal insulin regimen if the received FBG value does not correspond to the calculated dose response FBG value.

[0041] In an exemplary embodiment, the patient uses a continuous glucose monitor (CGM) that provides BG data based on the determination of fasting BG (FBG) values, as well as a connected drug delivery device, e.g., a pen device with an add-on device that logs injections with a smartphone app that provides daily insulin titration guidance. The app increments by up to 4 units if fasting glucose is above range, decrements by up to 4 units if FBG is below range, and no change is recommended otherwise. This is done every three days.

[0042] Example 1: Compliance The patient has been compliant with the treatment for the first 29 days of the period (as can be seen from the logged data) and is currently taking 40 units (IU) of insulin. - From days -33 to -37, injections were not logged (see Figure 1) - On day -38, the patient requests a dose recommendation. - The dose response algorithm (see below) creates a dose response for the missing days from days 1 to -37, assuming a dose of 44 units was taken from days -33 to -37 (see Figure 2). - The evaluation algorithm (see below) identifies that the mean of the fasting glucose measurements from days -33 to -37 falls within the 95% confidence interval of the response. - Based on the above, the evaluation algorithm determines that the patient has been compliant with taking 44 units of insulin daily and continues the titration from day -32 by suggesting a 4-unit dose increase.

[0043] The above is summarized in Table 1 below.

Table 1

[0044] Example 2: Non-compliance The patient has been compliant with the treatment for the first 29 days of the period (as can be seen from the logged data) and is currently taking 40 units of insulin. - From days -33 to -37, injections were not logged (see Figure 3) - On the 38th day, the patient requests a dosage recommendation. - The dose - response algorithm calculates the dose - response from the data on days 1 - 37, assuming that a dose of 44 units was administered on days 33 - 37 (see Figure 2). - The evaluation algorithm identifies that the mean of the fasting glucose measurements on days 33 - 37 falls outside the 95% confidence interval of the response. - Based on the above, the evaluation algorithm determines that the patient was non - compliant and suggests no dose change, continuing the titration from day 32.

Table 2

[0045] In Examples 1 and 2, during the period of 5 injections according to the prescribed regimen, all 5 injections are missing. The period may be shorter or longer since one or some of the injection dose sizes may be known.

[0046] In the above examples, the dose - response algorithm is based on a global linear regression model that assumes a linear relationship between glucose measurements. That is, TIFF2025520645000004.tif7170, where TIFF2025520645000005.tif5170 is TIFF2025520645000006.tif6170 and also TIFF2025520645000007.tif5170 is the day corresponding to the insulin injection performed.

[0047] Below, the usage method in model identification is described. The method is used to investigate whether the predicted dose - response is detectable from the data and whether outliers in the data have a significant effect on model identification. Then, the parameters from three model structures are identified, and which model best fits the individual dose - response is identified.

[0048] To estimate α and β, the ordinary least squares (LSQ) method is used. It can be written in matrix form for all n points for each individual. TIFF2025520645000008.tif7170 where TIFF2025520645000009.tif9170 and in the equation TIFF2025520645000010.tif9170 and TIFF2025520645000011.tif8170 the ordinary LSQ estimate is that the residual is TIFF2025520645000012.tif7170 assumed to be. Here, TIFF2025520645000013.tif8170 is usually the average response TIFF2025520645000014.tif7170 is the estimated parameter and is also found by minimizing the sum of the squared residuals. TIFF2025520645000015.tif9170

[0049] The solution is a vector, TIFF2025520645000016.tif5170 which is the estimated value of the unknown parameter θ. By substituting Equation (3) into two norms, TIFF2025520645000017.tif8170 is obtained.

[0050] This equation for θ is minimized and the derivative with respect to the parameter is taken, thereby obtaining the normal equations. TIFF2025520645000018.tif7170 and TIFF2025520645000019.tif5170 is solved for to obtain the estimate of the unknown parameter. TIFF2025520645000020.tif7170

[0051] The distribution of the parameter estimates is In TIFF2025520645000021.tif7170, TIFF2025520645000022.tif6170 is the estimated noise covariance, and is TIFF2025520645000023.tif8170. And n and TIFF2025520645000024.tif4170 are the number of data points and parameters, respectively.

[0052] The estimated noise covariance is used to calculate the confidence interval, for example, and is TIFF2025520645000025.tif10170, where TIFF2025520645000026.tif5170 is the predicted glucose value, TIFF2025520645000027.tif4170 indicates the confidence interval (e.g., 90% vs. 95%), and TIFF2025520645000028.tif3170 is the number of data points.

[0053] In the example shown, all data points are weighted the same as in the normal LSQ method, but these points can be weighted by a different weighting, e.g., a robust LSQ method.

[0054] In the normal LSQ method, all data points are assumed to be of equal quality. However, this may not be the case when considering the level of variability in SMBG data. To minimize the sensitivity to outliers and errors, weighted LSW can be used, and instead of minimizing the term in (4), minimize TIFF2025520645000029.tif9170. Where TIFF2025520645000030.tif4170 is the residual It is the weight of TIFF2025520645000031.tif4170. The weights can be selected in different ways using knowledge about the data. The normal equation is, TIFF2025520645000032.tif becomes 7170. In the formula, TIFF2025520645000033.tif is 7170, and When solving for TIFF2025520645000034.tif5170, an estimate of the unknown parameter TIFF2025520645000035.tif7170 is obtained.

[0055] Next, the distribution of the parameter estimates is, TIFF2025520645000036.tif8170.

[0056] Here, the bisquare-type weighting is used for robust LSQ that minimizes the influence of outliers on the fit. The method is iterative, giving full weight to small residuals and zero weight to residuals larger than expected by chance. The weights are TIFF2025520645000037.tif10170 are calculated iteratively. In the formula, TIFF2025520645000038.tif4170 is the adjusted and normalized residual, TIFF2025520645000039.tif4170, which is for weighted LSQ. TIFF2025520645000040.tif12170

[0057] Here TIFF2025520645000041.tif5170 is the leverage of the residual TIFF2025520645000042.tif4170, that is, the degree to which the i-th residual affects the fit, K is the tuning constant, s is the robust variance, It is TIFF2025520645000043.tif8170.

[0058] The strength of the vice-square type weighting lies in the ability to fit the data in a manner similar to the normal LSQ method while excluding the influence of outliers.

[0059] As another method, as shown in Figure 5, an asymmetric weighting function considering time or user error can be used. This shows the asymmetric weighting function of the residuals of (13) in the left figure. The right figure shows forgetting the weights in (14) for the valid memory level range of 10 days. It should be noted that only 3 out of 7 points are shown. This is due to the structure of the SMBG data where only the last 3 values before dose adjustment are available.

[0060] If a person measures SMBG incorrectly at other times, the glucose concentration is expected to be equal to or higher than the actual pre-breakfast SMBG. This generally results in an error in the positive direction of the SMBG measurement value. Therefore, it can be predicted that outliers caused by user error tend to increase the measured glucose concentration.

[0061] Considering user error and the severity of low glucose, the weighting function is designed such that low SMBG values are weighted higher than high SMBG values. However, it should be noted that insulin affects the glucose level and it is not desirable to exclude information about the dose response. Therefore, the SMBG reading is weighted compared to other SMBG readings given the same insulin dose. Therefore, the weighting function is It is TIFF2025520645000044.tif9170, and where TIFF2025520645000045.tif5170 is the i-th SMBG measurement value, TIFF2025520645000046.tif5170 is the corresponding insulin injection, and TIFF2025520645000047.tif7170 is TIFF2025520645000048.tif7170 all SMBG measurement values TIFF2025520645000049.tif7170 where TIFF2025520645000050.tif7170 is. If there is only one SMBG measurement for a corresponding insulin dose, TIFF2025520645000051.tif7170 and TIFF2025520645000052.tif7170 is. The weighting function is illustrated on the left side of Figure 5.

[0062] As shown above, the compliance or non - compliance determination can be used to provide insulin administration recommendations to a subject in an efficient and safe manner. More specifically, in the system described above, the method implemented can include an additional step of receiving a dose guidance requirement (DGR), and an additional step of determining whether the subject has been compliant with a regimen regarding insulin injection for a predetermined length of time before the DGR is received. If the subject has been compliant, the system will provide an updated dose recommendation based on the received FBGH and IH (e.g., 48 units as described above), or if the subject has not been compliant, the system will maintain the current dose recommendation (e.g., 44 units as described above).

[0063] In the above description of the exemplary embodiments, different structures and means for providing the functionality described for different components have been described to the extent that would be apparent to a reader skilled in the concepts of the present invention. The detailed construction and specifications for different components are considered the subject of normal design procedures by those skilled in the art along the lines set forth herein. *****

Claims

1. A diabetes management system adapted to determine compliance for a subject being treated according to a basal insulin regimen, the system comprising: - regimen data that sets a current dose size and a prescribed injection periodicity; and - a plurality of fasting blood glucose (FBG) measurements of the subject obtained over a period of time, thereby establishing an FBG history (FBGH), each respective glucose measurement within the plurality of glucose measurements comprising: (i) an FBG value; and (ii) a corresponding FBG timestamp, the FBG measurements; and - a plurality of insulin injection data sets during all or a portion of the period of time, thereby establishing an insulin injection history (IH), each injection data set comprising: (i) an injection amount; and (ii) an injection timestamp indicating when the injection occurred within the period of time, the insulin injection data sets, adapted to receive; The system comprises one or more processors and a memory, the memory when executed by one or more processors: - determining, for a period, whether one or more insulin injection data sets have not been received and are thus missing according to the prescribed injection regimen; and if an insulin injection data set(s) is missing: - calculating, for each missing injection, a predicted dose response FBG value based on the received FBGH and IH and using a dose response algorithm; - determining, for a given confidence interval, whether the received FBG value corresponding to the missing insulin injection corresponds to the calculated dose response FBG value; - if the received FBG value corresponds to the calculated dose response FBG value, determining that the subject is compliant with the basal insulin regimen; or - if the received FBG value does not correspond to the calculated dose response FBG value, determining that the subject is not compliant with the basal insulin regimen, the diabetes management system comprising instructions for implementing a method comprising.

2. The diabetes management system according to claim 1, wherein the predicted dose response FBG value is based on the received FBGH, IH, and regimen dose data.

3. Further adapted to provide insulin administration recommendations to the subject, the method comprising: - An additional step of receiving a dosage guidance request (DGR), and - An additional step of determining whether the subject adhered to the regimen regarding insulin injection for a predetermined length of time before receiving the DGR, and - An additional step of providing updated dosage recommendations based on the received FBGH and IH if the subject adhered, or - An additional step of maintaining the current dosage recommendation if the subject did not adhere, the diabetes management system according to claim 1 or 2.

4. The diabetes management system according to any one of claims 1 to 3, wherein the predetermined length of time before the DGR is received is the time since the last previous DGR was created.

5. The diabetes management system according to any one of claims 1 to 4, wherein the FBG value is derived from the received CGM data.

6. A method for determining adherence for a subject being treated according to a basal insulin regimen, comprising: - Obtaining regimen data that sets the current dosage size and the prescribed injection periodicity, and - Obtaining a plurality of fasting blood glucose (FBG) measurements of the subject taken over time, thereby establishing an FBG history (FBGH), wherein each respective glucose measurement within the plurality of glucose measurements is (i) an FBG value, and (ii) a corresponding FBG timestamp, and - Obtaining a plurality of insulin injection data sets during all or a portion of the time elapsed, thereby establishing an insulin injection history (IH), wherein each injection data set includes (i) the injection amount, and (ii) an injection timestamp representing when the injection occurred within the time elapsed, and - Determining for a period whether one or more insulin injection data sets have not been received and are thus missing according to the prescribed injection regimen; and if an insulin injection data set(s) is missing, - Calculating an expected dosage response FBG value for each missing injection based on the received FBGH and IH and using a dosage response algorithm, and - Determining for a given confidence interval whether the received FBG value corresponding to the missing insulin injection corresponds to the calculated dosage response FBG value. - When the received FBG value corresponds to the calculated dose-response FBG value, determining that the subject is compliant with the basal insulin regimen, or - When the received FBG value does not correspond to the calculated dose-response FBG value, determining that the subject is not compliant with the basal insulin regimen, a method comprising.

7. The method according to claim 6, wherein the predicted dose-response FBG value is based on received FBGH and IH and regimen dose data.

8. Further adapted to provide insulin dose recommendations for the subject, the method comprising - An additional step of receiving a dose guidance request (DGR); - An additional step of determining whether the subject was compliant with the regimen with respect to insulin injection for a predetermined length of time prior to receiving the DGR; - An additional step of providing an updated dose recommendation based on received FBGH and IH if the subject was compliant, or - An additional step of maintaining the current dose recommendation if the subject was not compliant, the method according to claim 6 or 7.

9. The method according to any one of claims 6 to 8, wherein the predetermined length of time prior to the DGR being received is the time since the last previous DGR was created.

10. The method according to any one of claims 6 to 9, wherein the FBG value is derived from received CGM data.

Images