Systems and methods for providing titration dose guidance

Abstract

A method for providing a titration insulin dose guidance recommendation (DGR) for a diabetic patient is provided, wherein a plurality of dose guidance algorithms (DGAs) adapted to calculate a dose recommendation for the patient is provided. Each DGA is characterized by a level of aggressiveness (LA) and a corresponding level of required data quality for a blood glucose history (BGH). When a DGR is requested by the patient one of the DGAs is selected based on a determined BGH data quality level for a BGH data set comprising BG values for a given period of lapsed time. The determination is based one or more of the following BGH properties: (i) number of measured BG values, (ii) variability of measured BG values, and (iii) qualifier type for the BG measurements. The qualifier type may comprise fasting BG measurements, non-fasting BG measurements, and CGM measurements.

Description

[0001] SYSTEMS AND METHODS FOR PROVIDING TITRATION DOSE GUIDANCE

[0002] The present disclosure generally relates to systems and methods for assisting patients and health care practitioners in managing insulin treatment to diabetics. In a specific aspect the present invention relates to systems and methods suitable for use in a diabetes management system in which a patient is titrated to reach a blood glucose target range for a determined insulin target dose.

[0003] BACKGROUND

[0004] People suffering from diabetes have been treated with insulin for more than a century. Diabetics are depending on insulin products to prolong their lives and increase their life quality. Injections of insulin are by far the most used method of diabetics to take their insulin. Different types of diabetes, different stages of each type and different lifestyles and individual physiological differences, all contribute to complicate optimisation of treatment regimens and as a result modern treatment is based on highly personalised treatment regimens based on extensive research and development as well as carefully supervised introduction of treatment by titration.

[0005] Development of treatment is not only focussed on increasing the efficiency and prolonging the life of diabetics, but also on postponing and ultimately avoiding the serious adverse long-term effects of the condition, as well as making the treatment itself easier and less of a daily burden for diabetics.

[0006] It is known that the better the insulin management fit with the actual need, the better the efficiency and the longer adverse effects of the condition can be postponed. Furthermore, severe over- or under-dosing can be immediately life threatening and must be avoided. Thus, large fluctuations of insulin levels in the body should be limited to occur only when needed, which is most often related to ingestion of meals and exercising. The better the natural regulation of insulin in the body of a non-diabetic can be mimicked, the better, obviously. An obvious technical solution would be to measure the blood glucose level, determine the need for insulin, and inject the required amount at intervals of a few minutes. However, this would be extremely expensive and not improve the life quality of the diabetics, as it would become a full-time occupation to regulate blood glucose / insulin levels.

[0007] To decrease the number of required injections and thereby the interference with the diabetic’s life and routines, different types of insulin has been developed to allow treatment of different types of diabetes and to allow combination of products to optimise treatment regimens on an individual basis. In particular, products of long-lasting effect of slowly releasing insulin in the user’s body has been developed to provide a basic level of insulin, called basal treatment. This ensures insulin levels do not get too low during sleep and outside meal-situations.

[0008] As a new user (patient) is introduced to treatment with insulin, the user has to go through a period of titration. For insulins for use once weekly or other fixed intervals, the often-preferred titration method is the 3-FASTING SMBG method, according to which the FASTING SMBG (Fasting Blood Glucose, being a measurement of the blood glucose concentration after 8 hours of fasting) is measured on titration day and the prior two days. The measurements are done as SMBG (Self Measured Blood Glucose) values and should preferably be in the range between 4.4 mmol / L - 7.2mmol / L. A mean value of these three measurements is used to avoid inefficient titration due to bad measurements and random variations in plasma Glucose variations caused by unusual activities of the user.

[0009] Thus, if Sunday is titration day for a new user of a once weekly insulin product, the user is required to obtain SMBG values Friday, Saturday and Sunday and calculate a mean value. If the value is above the target area, the dose should be increased and if it is below, the dose should be decreased. This process continues until the weekly measurements becomes stabile and within target range.

[0010] However, such measurements are uncomfortable for the users, as they are not only required to sting a finger to obtain blood on which to measure, but also required to be fasting 8 hours each Friday, Saturday and Sunday.

[0011] Furthermore, by requiring 3 measurements to reduce the impact of a bad measurement, the likelihood of a bad measurement disturbing the outcome is also increased, as more measurements are performed.

[0012] The current titration method and target range is a best compromise for a population, based on a “one rule-fits all”-approach. As the BG values varies during the day, depending on the activities of the individual and the individual itself, it would be expected that user would increase stability of BG levels and provide better treatment outcome, if it was possible and feasible to provide a more personalised titration procedure and treatment. Thus, different solutions based on modelling of the expected effect over time from an injection, have been introduced. These models may be more or less complex, but all need some input data from the specific user.

[0013] Baseline data (fasting SMBG measured with no insulin given) and fasting SMBG’s each week at different doses administered can be used as input in mathematical modelling, and the effect to be expected from changes of the dose given estimated. This will allow the system to predict the result of next week’s measurement from a given change in dose, whereby the system can be used to optimise suggestion of a new dose setting. As the fasting SMBG is then measured the following week and thus added to the historical data, the algorithm can improve the modelling accuracy and gradually adapt the parameters of the model to fit the specific individual’s response to the drug better and better.

[0014] As the model quickly becomes able to predict the following weeks fasting SMBG value within reasonable tolerances, the system will be able to detect bad measurements and outliers and request a measurement to be redone, accompanied with suggestions of how to avoid errors, such as washing hands just prior to measuring to avoid contamination of the sample and such. By becoming able to detect bad measurements “earlier” in the sense that they may be detected even when being just outside limited tolerances and not having to be obviously off to an extend where the user would become suspicious, fewer measurements are needed, as one of the purposes of requiring three measurements where to reduce the impact of a bad measurement.

[0015] WO 2023 / 144364 discloses a data-driven basal insulin titration algorithm that provides personalized decision support to guide dose adjustments to find the optimal insulin dose is provided. A for the subject representative prediction algorithm is adapted to calculate the mean and the variance of the resulting fasting blood glucose value (FBG) for the subject as a function of the size of an injected dose of insulin, the prediction algorithm being adaptive in response to obtained FBG and corresponding insulin dose size data for the subject. Based on predicted mean and variance FBG values a policy algorithm is adapted to calculate the probability for a corresponding policy target as a function of insulin dose size. The dose size recommendation is determined as the insulin dose size having the highest calculated probability for meeting the policy target.

[0016] When using the current fasting SMBG-3, the measurements obtained will be a few measurements by the end of the week. Since the BG level varies significantly during the day (with meals and activity) and during the week (as the effect of the basal insulin first increases and then slowly decreases), the user may become both hypo-glycaemic and hyper-glycaemic in short intervals during the week, without it being detected during the weekly SMBG measurements.

[0017] To minimise the occurrence of such events, an even more personalized titration method may be used based on CGM (Continuous Glucose Measuring) to obtain a large number of BG values through a number of weeks. This would provide input data for a much more detailed modelling and allow for an indexing of different dose settings, by introducing penalty functions with different weighing for hypo-glycaemic and hyper-glycaemic occurrences. Although a high level of titration safety can be provided using CGM data, for many patients it is not deemed attractive to have to bother with and take care of a skin-mounted CGM sensor.

[0018] Correspondingly, it is an object of the present invention to provide a titration set-up for a long- acting type of insulin which is both safe and effective to reach a titration target, which is considered not too burdensome to the patient, and which can reduce the involvement and / or the burden on the HCP (Health Care Practitioner, e.g. a doctor or a nurse).

[0019] DISCLOSURE OF THE INVENTION

[0020] In the disclosure of the present invention, embodiments and aspects will be described which will address one or more of the above objects or which will address objects apparent from the below disclosure as well as from the description of exemplary embodiments. In the context of the present invention, when not used in combination with specific values, the terms “fasting blood glucose” (FBG) and “fasting plasma glucose” (FPG) respectively BG and PG are used interchangeable. Correspondingly, in the disclosure of the invention the terms FBG / BG represents FBG / BG as well as FPG / PG.

[0021] In a first aspect of the invention a computing system for providing a titration dose guidance recommendation for a long-acting insulin for a query subject (patient) to treat diabetes mellitus is provided, wherein the system comprises one or more processors and a memory. The memory comprises instructions for executing by the one or more processors the steps of: obtaining a plurality of dose guidance algorithms (DGAs) adapted to calculate a long-acting insulin dose recommendation for the query subject based on a blood glucose target range (BGTR), and a blood glucose history (BGH), each DGA being characterized by a level of aggressiveness (LA) and a corresponding level of required BGH data quality, obtaining a BGTR for the query subject, obtaining a BGH data set, comprising at least one blood glucose (BG) measurement for the query subject taken over a time course, each respective BG measurement comprising (i) a BG value, (ii) a corresponding blood glucose timestamp representing when in the time course the respective glucose measurement was made, and (iii) a qualifier type indicating the type of BG measurement, and determining a BGH data quality level for the BGH data set, the determination being based one or more of the BGH properties (i) number of measured BG values, (ii) variability of measured BG values, and (iii) qualifier type for the BG measurements. The memory comprises instructions for executing by the one or more processors the further steps of: receiving a dose guidance request (DGR) for the query subject, based on the determined BGH data quality level for the BGH data set, selecting a DGA from the plurality of DGAs, and calculating a dose recommendation utilizing the selected DGA.

[0022] In this way a data-driven insulin titration system is provided that to a high degree assures personalized and safe decision support to guide dose adjustments to find an optimal insulin dose and which can reduce the involvement and / or the burden on the HCP as the resulting dose recommendation is influenced by the actual quality of the provided BG data and not only the assumed or expected quality.

[0023] The term “level of aggressiveness” characterizes how much a given DGA for a given BG input and BGTR may recommend adjusting a dose, i.e. a more aggressive DGA will recommend a higher dose adjustment and a less aggressive DGA will recommend a lower dose adjustment. A least aggressive DGA will not recommend any adjustment. According to the invention a more aggressive DGA requires a corresponding higher level of BGH data quality than a less aggressive DGA.

[0024] In the context of the present invention the term “measurement” also covers BG values derived from a number of measured BG values, e.g. a Titration Glucose Level (TGL) value based on a number of measured FBG values or a TGL value based on a large number of CGM measurements.

[0025] The qualifier type may comprise fasting SMBG measurement, non-fasting SMBG measurement, and CGM measurement. The distinction between fasting and non-fasting SMBG measurement may e.g. be based on user input or be estimated from the time the measurement is made.

[0026] The level of BGH data quality is determined based on e.g. the number of measured BG values, the variability of measured BG values, and the type of the BG measurements, e.g. a high number of BG values with low variability provides a high level of BGH data quality. Correspondingly, BG values based on CGM measurements may be deemed to provide a higher level of BGH data quality than fasting SMBG measurement which then can be deemed to provide a higher level of BGH data quality than non-fasting SMBG measurement.

[0027] The time course over which the BGH data is collected will typically depend on the kind of long- acting insulin for which a titration dose is to be recommended, e.g. for a once-weekly insulin such as icodec the time course may be the week before the dose guidance request is made.

[0028] Correspondingly, for a once-weekly insulin a BGH data set comprising a single non-fasting SMBG measurement will result in a low level of BGH data quality whereas a BGH data set based on a complete week of CGM measurements will result in a high level of BGH data quality.

[0029] Selecting a DGA from the plurality of DGAs may be based on a look-up table in which a required minimum level of BGH data quality for each DGA is listed, e.g. a most-aggressive DGA may require BG values based solely on CGM measurement, an intermediate-aggressive DGA may require a given number of fasting SMBG measurements with a given variability, and a least-aggressive DGA may require only a single fasting SMBG measurements within a given range or a number of non-fasting SMBG measurements. If the level of BGH data quality is too low or no BGH data set is provided a “no adjustment” DGA may be selected. Disregarding the general set-up, low BG values of even low quality may result in a lowering dose adjustment.

[0030] As input to the DGA one or more TGL values may be determined based on the BGH data set, e.g. all BG values for a given period, e.g. a day or a week, may be condensed to a single TGL value. For a given period of lapsed time an average or minimum TGL value may be determined based on one or more TGL values. The selection of the current DGA may be based on at least one of the following TGL properties: (i) number of measured BG values upon which an average or minimum TGL value is determined, (ii) variability of TGL values, and (iii) qualifier type for the BG measurements upon which the one or more TGL value / s is / are determined.

[0031] At least one of the DGAs may be in the form of a look-up table. At least one DGA may be adapted to calculate a dose recommendation for the query subject based additionally on an Insulin Dose Event History (IH), the method comprising the further step of: obtaining an IH data set, comprising at least one dose event during all or a portion of the time course, each dose event of the at least one dose event comprising (i) an insulin dose amount, and (ii) a corresponding dose event timestamp representing when in the time course the respective dose event occurred. The plurality of DGAs may be based on a single DGA in which a factor is applied to a calculated recommendation based on the level of BGH data quality.

[0032] When a given DGA has been selected as a current DGA, either by the HCP or by a previously by the system, the selection of a new DGA based on the determined data quality level may comprise the options of (i) maintaining the current DGA as the selected DGA if the determined BGH data quality level corresponds to or exceeds the required BGH data quality level, or (ii) selecting a new DGA with a lower LA if the determined BGH data quality level does not correspond to the required BGH data quality level for the current DGA. The current DGA may remain as the

[0033] Alternatively, when a given DGA has been selected as a current DGA, the selection of a new DGA based on the determined data quality level may comprise the options of (i) maintaining the current DGA as the selected DGA if the determined BGH data quality level corresponds to or exceeds the required BGH data quality level, (ii) selecting a new DGA with a higher LA if the determined BGH data quality level exceeds the required BGH data quality level for the current DGA, or selecting a new DGA with a lower LA if the determined BGH data quality level does not correspond to the required BGH data quality level for the current DGA.

[0034] Based on the BGH data set, the system may determine whether the BG values for the query subject are within the BGTR at a given level of acceptance, e.g. a maximum of 15% above the BGTR and 0% below the BGTR. If the BG values for the query subject are not within the BGTR at the given level of acceptance, then the system may suggest a BG measurement regime corresponding to an increased level of BGH data quality. An increased level of BGH data quality may be suggested depending on one or more recent determined BGH data quality levels having met the BGH data quality level for a current DGA.

[0035] In an exemplary embodiment the memory comprises instructions for executing by the one or more processors the further step of obtaining a time schedule setting out when a DGR is generated (e.g. a daily DGR or a weekly DGR), i.e. the patient will not have to request a DGR manually, but it will be generated automatically allowing the recommended dose to be displayed, e.g. on the patient’s smartphone.

[0036] In an exemplary embodiment one or more of the DGAs is / are adapted to calculate an estimated risk of hypoglycaemia and adjust the BGTR accordingly. The memory may further comprise instructions for executing by the one or more processors the further steps of based on BGH and BGTR calculate a deviation value as an indicator for the performance of the titration process and based on the deviation value provide a recommendation in respect of the number of BG measurements to be provided for a given period of time.

[0037] In a further aspect of the invention a method for providing a titration dose guidance recommendation for a query subject to treat diabetes mellitus is provided, the method comprising the steps of obtaining a plurality of dose guidance algorithms (DGAs) adapted to calculate a long- acting insulin dose recommendation for the query subject based on a blood glucose target range (BGTR) and a blood glucose history (BGH), each DGA being characterized by a level of aggressiveness (LA) and a corresponding level of required BGH data quality, obtaining a BGTR for the query subject, and obtaining a BGH data set, comprising at least one blood glucose (BG) measurement for the query subject taken over a time course, each respective BG measurement comprising (i) a BG value, (ii) a corresponding blood glucose timestamp representing when in the time course the respective glucose measurement was made, and (iii) a qualifier type indicating the type of BG measurement. The method comprises the further steps of: determining a BGH data quality level for the BGH data set, the determination being based one or more of the following BGH properties: (i) number of measured BG values, (ii) variability of measured BG values, and (iii) qualifier type for the BG measurements, and receiving a dose guidance request (DGR) for the query subject. Based on the determined BGH data quality level for the BGH data set, the method comprises the further steps of selecting a DGA from the plurality of DGAs, and calculating a dose recommendation utilizing the selected DGA.

[0038] The method may comprise additional steps or have properties as described above for a corresponding system.

[0039] BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In the following embodiments of the invention will be described with reference to the drawings, wherein fig. 1 shows a graphical representation of a titration process running over four 16 weeks periods, and figs. 2A and 2B illustrate how a TGL value can be determined based on CGM data.

[0041] DESCRIPTION OF EXEMPLARY EMBODIMENTS The present invention relates to a system and algorithm adapted to provide titration dose guidance recommendations for user to treat diabetes mellitus.

[0042] In such a system a given algorithm is used to generate recommended insulin doses and treatment advice for diabetes patients based on BG data, insulin dosing history and, in more advanced applications, other factors like meals, physical activity, stress, illness etc. may be taken into consideration.

[0043] Essentially such a system comprises a back-end engine (“the engine”) used in combination with an interacting system in the form of a client and an operating system. The client from the engine’s perspective is the software component that requests dose guidance. The client gathers the necessary data (e.g. FPG values) and requests dose guidance from the engine. The client then receives the response from the engine.

[0044] On a small local scale, the engine may run directly as an app on a given user’s smartphone or on a blood glucose meter (BGM) and thus be a self-contained application comprising 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 cloud-based large-scale diabetes management system. Such a cloud-based system would allow the engine to always be up to date (in contrast to app-based systems running entirely on e.g. the patient’s smartphone), would allow advanced methods such as machine learning and artificial intelligence to be implemented, and would allow data to be used in combination with other services in a greater “digital health” setup. Such a cloud-based system ideally would manage a large amount of patient requests for dose recommendations.

[0045] Although a “complete” engine may be designed to manage all computing aspects, it may be desirable to divide the engine into a local and a cloud version to allow the patient-near day-to- day part of the dose guidance system to run independently of any reliance upon cloud computing. For example, when the user via the client app makes a request for dose guidance the request is transmitted to the engine which will return a dose recommendation. In case cloud access is not available the client app would run a dose-recommendation calculation using local data. Dependent upon the user’s app-settings the user may or may not be informed.

[0046] With reference to fig. 1 an illustrative scenario for a user (patient) being titrated for treatment with a long-acting type of insulin is shown. Typically starting from a high BG level, the patient is “aggressively” titrated during a first period to reach the target BG range (using an “aggressive” titration algorithm) after which the patient will continue with the determined target dose which ideally will keep the BG values within the target range. During the titration process the BG level is found often, e.g. based on daily FBG values determined using SMBG measurements, or, as shown, using CGM data. After having determined the titration target dose the patient is shifted to a scheme where fewer FBG values is used merely to assure that the patient continues to be within the BG target range taking the determined target dose, e.g. using a BGM as shown.

[0047] However, as also shown if fig. 1 , after a second period with measured BG values within target range a number of measured BG values are outside the target range, this resulting in the patient's HCP starting a “corrective” titration process during a third period based on a higher frequency of FBGs and using an appropriate titration scheme. In the shown example this measure does not result in acceptable BG values within the target range for which reason the HCP decides to start a second more intensive titration process during a fourth period based on CGM measurements and using a corresponding titration algorithm which may or may not be the same as the initially used titration algorithm. As appears, after a second titration period the patient has (again) been treated to target using the newly determined target dose.

[0048] Turning to the specific titration example shown in fig. 1 a patient has been titrated with a once- weekly long-acting insulin analog (e.g. icodec from Novo Nordisk A / S, Denmark) during four 16 weeks periods. During the first 16 weeks period the patient is titrated to target using CGM BG data and an “aggressive” titration algorithm taking advantage of the high data quality provided by the CGM. As most titration algorithms are not designed to receive BG data corresponding to the sampling rate of a typical CGM, e.g. BG values measured every 5 minutes), the CGM data is used to determine a “titration glucose level” TGL based on 24 hours of CGM values (see below). The daily TGL values may then be used to determine a TGL value for an entire week (e.g. as an average or as a minimum TGL value) which is then used as an input to the titration algorithm.

[0049] The initial CGM-based titration algorithm may e.g. be obtained as a personalized physiological model derived from BG and insulin history data obtained from the patient during a modeldetermining period prior the titration process per se. Alternatively, a titration algorithm may be selected based on demographic properties for the patient, e.g. age, sex, BMI etc. Titration may also be based on a simple look-up table, see below. Having reached the titration BG target at the end of the first 16 weeks period, the patient is shifted to a steady observation period in which an FBG value is determined once weekly as an SMBG value using a traditional BG meter. As appears, at the end of the second period a number of FBG values outside the BG target range is determined, this indicating that the correct target dose was not determined initially. However, having only a single FBG value per week the HCP will be cautious to (aggressively) titrate the patient up based solely on such scarce BG data.

[0050] Corresponding, the patient is shifted to a 3 FBG determinations per week, this allowing the HCP to titrate the patient more aggressively. The input to the titration algorithm will typically be a single TGL value determined as an average or a minimum for the week’s FBG values. However, as appears, at the end of the third period the FBG values are generally outside the BG target range, this indicating that the correct target dose has not yet been determined.

[0051] Consequently, the patient is shifted back to a more aggressive CGM-based titration scheme in which the same titration algorithm used initially may be used. As appears, at the end of the fourth 16 weeks period the patient’s BG values are back in the target range based on a new target dose after which the patient can be shifted back to a once-weekly FBG scheme.

[0052] As illustrated, a successful titration process will often need to be shifted between different schemes relying on different amounts of BG data, this shifting typically being authorized and supervised by the HCP which will actively have to decide how to adjust the titration scheme to reach a target dose resulting in measured FBG values within the target range. During an initial aggressive titration process the patient may be well motivated to follow the instructions and provide the BG measurements required by the chosen titration algorithm, however, for a subsequent shift back from a less burdensome BG scheme to a more burdensome BG scheme the patient may be less motivated. For example, the patient may not provide the required number of fasting SMBGs or may not carry a CGM as required over a longer period. Such considerations may make the HCP reluctant to intensify the titration process.

[0053] Correspondingly, in an aspect of the present invention a titration set-up is provided which can reduce the involvement and / or the burden on the HCP and which ideally may provide a set-up which would allow the patent to perform an “automated” titration process which is adaptive to the BG input data received. More specifically, a method and a system for implementing the method are provided in which a plurality of Dose Guidance Algorithms (DGAs) adapted to calculate a dose recommendation for a patient based on a Blood Glucose Target (BGTR) and a Blood Glucose History (BGH) is obtained. The DGAs may be “advanced” based on physiological models and adapted to a given patient during an initial modelling period in which also insulin dose event history (IH) is utilized. Alternatively, the DGAs may be in the form of simple look-up tables or flow-charts. The different DGAs may be variations of the same DGA, e.g. implemented more or less aggressively based on available BG titration data.

[0054] In addition to an initial or updated BGTR the dose recommendation is based on an obtained BGH data set comprising at least one Blood Glucose (BG) measurement taken over a time course, e.g. a week for titration with a once-weekly insulin analogue. Each respective BG measurement comprising a BG value, a corresponding blood glucose timestamp representing when in the time course the respective glucose measurement was made, and a qualifier type indicating the type of BG measurement.

[0055] In the present context the term “measurement” also covers BG values derived from a number of measured BG values, e.g. a TGL value based on a number of measured FBG values or a TGL value based on a large number of CGM measurements.

[0056] Depending on the available DGAs also an IH data set may be obtained.

[0057] When the patient makes a Dose Guidance Request (DGR) one of the pluralities of DGAs is selected as the current DGA. The selection is performed for a given period of lapsed time, e.g. a week, and is based on at least one of the following BGH properties: number of measured BG values, variability of measured BG values, and qualifier type for the BG measurements, e.g. based on SMBG values or CGM values. Utilizing the selected DGA a dose recommendation is determined.

[0058] In the below scheme a set-up is shown in which the DGAs all are in the form of simple look-up tables:

[0059] As appears, the number of available FBG measurements determine the allowed increments. In the case where few measurements are available, the allowed increments of the insulin dose are restricted to provide a safe titration with limited FBG data available. Conversely, if more FBG measurements are available, larger steps are allowed.

[0060] In case of no available FBG data during the dosing period, no titration will take place. This is recognised by the algorithm, and subsequently a reminder can be provided to the patient to measure the FBG.

[0061] If a prediction model is implemented and running in the background a given FBG value can be categorized as an outlier when outside the defined confidence interval defined by the prediction model. In such a case the FBG value may be ignored or a more conservative and safer DGA may be used.

[0062] When dose recommendations are based on CGM data TGL values will typically have to be estimated. This could be achieved by applying e.g. a 3 hour “sliding window” across CGM readings one data point at the time, see figs. 2A and 2B, calculating the mean of each 3 hours interval from the current time up until the time of the last dose adjustment. The lowest 3 hour mean for each day will be recorded as a TGL value and used for titration. An alternative method for estimating an FBG / TGL value from CGM data is disclosed in applicant’s WO 2024 / 133412.

[0063] An example of a personalized advanced DGA is disclosed in applicant’s WO 2023 / 144364, which is hereby incorporated by reference. Such an advanced model may be developed during an initial training period, i.e. prior to the initial aggressive CGM-based titration period and may be refined further during the titration period. Such a model may be used as the basis for additional DGAs in which the recommendations is “scaled back” based on the amount and quality of the BG data input, e.g. a given calculated recommendation may be multiplied by an appropriate factor, e.g. 0.5 or 0.25.

[0064] The system may be provided with algorithms adapted to calculate an estimated risk of hypoglycaemia and adjust the BGTR accordingly. For example, if a single FBG is provided (not an outlier), and the dose recommendation for the week results in a hypoglycaemia risk, the dose recommended for the subsequent week, will be decreased (i.e., titrated down). Furthermore, at least 3 FBG measurements are requested to be provided by the patient for the subsequent week.

[0065] To further assist the HCP (or to “fully automate” the titration process), the quantity and quality of the requested BG input may be generated by the system and controlled by the achieved results from the titration process. To determine the required level of BG input a BG prediction model or actual measured BG values may be used as input.

[0066] When “more data” are requested, at least 3 FBG (day -2, -1 and 0) or more are expected / in- tended depending on the situation. This is described in the following cases below specified as levels 1 , 2 or 3, depending on the accuracy of the model and / or the current mean FBG.

[0067] In other words, the level is increased if the model accuracy is poor (to fit the model with more data) or if the mean FBG is too high or too low (to allow for larger dose changes). If the model accuracy remains poor while in level 3, HCP involvement may be required to assess the causes of the poor accuracy. If the model obtains a satisfactory accuracy (i.e. FBG values inside the 95% confidence interval), fewer FBG measurements are needed for titration, thus the level is decreased.

[0068] In the above description of exemplary embodiments, the different structures and means provid- ing the described functionality for the different components have been described to a degree to which the concept of the present invention will be apparent to the skilled reader. The detailed design and specification for the different components are considered the object of a normal design procedure performed by the skilled person along the lines set out in the present specification.

[0069] *****

Claims

CLAIMS1. A computing system for providing a titration dose guidance recommendation for a long-acting insulin for a query subject to treat diabetes mellitus, wherein the system comprises one or more processors and a memory, the memory comprising: instructions for executing by the one or more processors the steps:(a) obtaining a plurality of dose guidance algorithms (DGAs) adapted to calculate a long- acting insulin dose recommendation for the query subject based on a blood glucose target range (BGTR) and a blood glucose history (BGH), each DGA being characterized by a level of aggressiveness (LA) and a corresponding level of required BGH data quality,(b) obtaining a BGTR for the query subject,(c) obtaining a BGH data set, comprising at least one blood glucose (BG) measurement for the query subject taken over a time course, each respective BG measurement comprising:(i) a BG value,(ii) a corresponding blood glucose timestamp representing when in the time course the respective glucose measurement was made, and(iii) a qualifier type indicating the type of BG measurement,(d) determining a BGH data quality level for the BGH data set, the determination being based one or more of the following BGH properties:(i) number of measured BG values,(ii) variability of measured BG values, and(iii) qualifier type for the BG measurements,(e) receiving a dose guidance request (DGR) for the query subject,(f) based on the determined BGH data quality level for the BGH data set, selecting a DGA from the plurality of DGAs, and(g) calculating a dose recommendation utilizing the selected DGA.

2. A computing system as in claim 1 , wherein a given DGA has been selected as a current DGA, the selection of a DGA in step (g) comprising the options:maintaining the current DGA as the selected DGA if the determined BGH data quality level corresponds to or exceeds the required BGH data quality level, or selecting a new DGA with a lower LA if the determined BGH data quality level does not correspond to the required BGH data quality level for the current DGA.

3. A computing system as in claim 1 , wherein a given DGA has been selected as a current DGA, the selection of a DGA in step (g) comprising the options: maintaining the current DGA as the selected DGA if the determined BGH data quality level corresponds to or exceeds the required BGH data quality level, selecting a new DGA with a higher LA if the determined BGH data quality level exceeds the required BGH data quality level for the current DGA, or selecting a new DGA with a lower LA if the determined BGH data quality level does not correspond to the required BGH data quality level for the current DGA.

4. A computing system as in any of claims 1-3, wherein: based on the BGH data set, determining whether the BG values for the query subject are within the BGTR at a given level of acceptance, and if the BG values for the query subject are not within the BGTR at the given level of acceptance, then(h) suggesting to the query subject a BG measurement regime corresponding to an increased level of BGH data quality.

5. A computing system as in claim 4, wherein an increased level of BGH data quality is suggested only if one or more recent determined BGH data quality levels have met the BGH data quality level for a current DGA.

6. A computing system as in any of claims 1-5, wherein the qualifier type comprises: fasting SMBG measurement, non-fasting SMBG measurement, and CGM measurement.

7. A computing system as in any of claims 1-6, wherein at least one of the DGAs is in the form of a look-up table.

8. A computing system as in any of claims 1-7, wherein one or more titration glucose level (TGL) values is / are determined based on the BGH data set.

9. A computing system as in claim 8, wherein for the given time course an average or minimum TGL value is determined based on the one or more TGL values.

10. A computing system as in claim 9, wherein the selection of the current DGA is based on at least one of the following TGL properties:(i) number of measured BG values upon which an average or minimum TGL value is determined,(ii) variability of TGL values, and(iii) qualifier type for the BG measurements upon which the one or more TGL value / s is / are determined.

11. A computing system as in any of claims 1-10, wherein at least one DGA is adapted to calculate a dose recommendation for the query subject additionally based on an insulin dose event history (I H) , the method comprising the further step of: obtaining an IH data set, comprising at least one dose event during all or a portion of the time course, each dose event of the at least one dose event comprising:(i) an insulin dose amount, and(ii) a corresponding dose event timestamp representing when in the time course the respective dose event occurred.

12. A computing system as in any of claims 1-11 , the memory comprising instructions for executing by the one or more processors the further step: obtaining a time schedule setting out when a DGR is generated.

13. A computing system as in claim 12, wherein the time schedule generates:(i) a daily DGR, or(ii) a weekly DGR.

14. A computing system as in any of claims 1-13, wherein one or more of the DGAs is / are adapted to calculate an estimated risk of hypoglycaemia and adjust the BGTR accordingly.

15. A computing system as in any of claims 1-14, the memory comprising instructions for executing by the one or more processors the further steps: based on BGH and BGTR, calculating a deviation value as an indicator for the performance of the titration process, and based on the deviation value, providing a recommendation in respect of the number of BG measurements to be provided for a given period of time.*****

Images