Safety tools for decision support recommendations made to users of continuous glucose monitoring systems

Abstract

To provide systems and methods for determining if a decision support recommendation is to be presented to a user for treatment of a diabetic state.SOLUTION: The methods include receiving a plurality of input data items impacting a diabetic state of a user of a continuous glucose monitor, the input data items serving as input data to a process for determining a decision support recommendation; assigning a reliability level to each of the input data items; calculating a reliability metric based on the reliability levels assigned to the respective input data items; and determining a decision support recommendation based on the process and the input data and presenting the decision support recommendation to the user on a user interface only if the reliability metric exceeds a threshold.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Patent Application No. 62 / 786,089, filed December 28, 2018. The aforementioned application is incorporated by reference in its entirety and expressly made a part hereof.

[0002] The present embodiments relate to continuous analyte monitoring, and in particular to signal analysis and result presentation of continuous analyte monitoring systems. [Background technology]

[0003] Diabetes mellitus is a disease in which the pancreas cannot make enough insulin (Type I or insulin-dependent) and / or insulin is not effective (Type II or non-insulin-dependent). In the diabetic state, the victim suffers from hyperglycemia, which can cause many physiological disorders associated with the deterioration of small blood vessels, such as kidney failure, skin ulcers, or bleeding into the vitreous of the eye. Hypoglycemic reactions (low blood sugar) can be precipitated by inadvertent overdosing of insulin, or when normal administration of insulin or glucose-lowering drugs is accompanied by extreme exercise or poor food intake.

[0004] Traditionally, people with diabetes carry a self-monitoring blood glucose (SMBG) monitor, which typically requires an uncomfortable finger-prick method. Due to lack of comfort and convenience, people with diabetes usually measure their glucose levels only two to four times per day. Unfortunately, such time intervals are so far apart that people with diabetes may detect a hyperglycemic or hypoglycemic condition too late, sometimes resulting in dangerous side effects. With traditional methods, not only is it unlikely that people with diabetes will detect a dangerous condition in time to prevent it, but it is also likely that they will not know if their blood glucose level is rising (higher) or falling (lower) based on traditional methods. Thus, people with diabetes may not be able to make empirical insulin therapy decisions.

[0005] Another device that some diabetics use to monitor their blood glucose is a continuous analyte sensor, e.g., a continuous glucose monitor (CGM). A CGM typically includes a sensor that is placed invasively, minimally invasively, or non-invasively. The sensor measures the concentration of a given analyte, e.g., glucose, in the body and generates a raw signal that is generated by electronics associated with the sensor. The raw signal is converted to an output value that is displayed on a display. The output value resulting from the conversion of the raw signal is typically expressed in a format that provides meaningful information to the user, and users become accustomed to analyzing in formats such as blood glucose expressed in mg / dL. Summary of the Invention [Means for solving the problem]

[0006] In a first aspect, there is provided a method for determining whether a decision support recommendation should be presented to a user for treatment of a diabetic condition, the method comprising: identifying a user's current diabetic condition based at least in part on a glucose concentration level; receiving additional data that serves as input data along with the current diabetic condition to a process for determining a decision support recommendation; assigning a confidence level to the additional data; calculating a decision support recommendation using the process and the input data; and presenting the decision support recommendation to the user on a user interface only if the assigned confidence level of the additional data exceeds a threshold.

[0007] In an embodiment of the first aspect or any other embodiment, assigning the trust level includes assigning the trust level based at least in part on the identity of a source from which the additional data was obtained.

[0008] In an embodiment of the first aspect or any other embodiment, the additional data includes an indication of food and beverages ingested or to be ingested within a defined period of time.

[0009] In an embodiment of the first aspect or any other embodiment, the indicator comprises data manually entered by a user.

[0010] In an embodiment of the first aspect or any other embodiment, the index includes nutritional data about the food or beverage obtained from a third party.

[0011] In an embodiment of the first aspect or any other embodiment, the nutritional data is entered automatically by the application.

[0012] In an embodiment of the first aspect or any other embodiment, the nutritional data is automatically entered into the application by scanning the nutritional information.

[0013] In an embodiment of the first aspect or any other embodiment, the additional data includes information describing the user's past physiological response patterns.

[0014] In an embodiment of the first aspect or any other embodiment thereof, the additional data includes an indication of the level of user activity that has occurred or should occur within a defined period of time.

[0015] In an embodiment of the first aspect or any other embodiment, assigning the trust level includes assigning a higher trust to the additional information obtained through machine-to-machine communication than through manual input-to-machine communication.

[0016] In an embodiment of the first aspect or any other embodiment, the decision support recommendations include recommendations for consuming food or beverages.

[0017] In an embodiment of the first aspect or any other embodiment, the decision support recommendation comprises a recommendation to inject a calculated amount of insulin.

[0018] In an embodiment of the first aspect or any other embodiment, presenting the decision support recommendation includes determining when to present the decision support recommendation based at least in part on the confidence level.

[0019] In an embodiment of the first aspect or any other embodiment, receiving the additional data includes receiving a plurality of additional data items of different types, and assigning the trust level to the additional data includes independently assigning the trust level to each of the additional data items.

[0020] In an embodiment of the first aspect or any other embodiment, presenting the decision support recommendation includes presenting the decision support recommendation only if a weighted average of the individual confidence levels exceeds a first threshold and if none of the individual confidence levels is below a second threshold.

[0021] In an embodiment of the first aspect or any other embodiment, the first threshold and / or the second threshold are selected based at least in part on the type of decision support recommendation being calculated.

[0022] In an embodiment of the first aspect or any other embodiment, the threshold value is selected based at least in part on the type of decision support recommendation being determined.

[0023] In an embodiment of the first aspect or any other embodiment, presenting a decision support recommendation includes presenting a more conservative support recommendation that promotes patient safety when the assigned confidence level is within a range specified by the upper and lower thresholds.

[0024] In an embodiment of the first aspect or any other embodiment, the calculated amount of insulin is calculated as part of a process for determining a decision support recommendation.

[0025] In an embodiment of the first aspect or any other embodiment, assigning a confidence level to the additional data includes using machine learning techniques to assign a confidence level to the additional data.

[0026] In an embodiment of the first aspect or any other embodiment, the glucose concentration level is obtained from a glucose sensor in a continuous glucose monitor.

[0027] In an embodiment of the first aspect or any other embodiment, determining the user's current diabetic condition includes receiving sensor data from a glucose sensor associated with a continuous glucose monitor.

[0028] In an embodiment of the first aspect or any other embodiment, the user interface is integrated into a display device that receives sensor data from the glucose sensor.

[0029] In an embodiment of the first aspect or any other embodiment, the display device receives the sensor data from the glucose sensor via a wireless communication link.

[0030] In an embodiment of the first aspect or any other embodiment, presenting the decision support recommendations to the user on a user interface includes presenting the decision support recommendations to the user on a user interface of a display device.

[0031] In an embodiment of the first aspect or any other embodiment, a continuous glucose sensor application resident on a display device receives the sensor data and causes decision support recommendations to be presented to the user on a user interface.

[0032] In a second aspect, a method is provided for determining whether to calculate a decision support recommendation, the method comprising: identifying a user's current diabetes state based at least in part on a glucose concentration level; receiving additional data together with the current diabetes state that serves as input data to a process for determining a decision support recommendation; assigning a confidence level to the additional data; and calculating the decision support recommendation using the process and the input data only if the assigned confidence level of the additional data exceeds a threshold.

[0033] In an embodiment of the second aspect or any other embodiment, assigning the trust level includes assigning the trust level based at least in part on the identity of a source from which the additional data was obtained.

[0034] In an embodiment of the second aspect or any other embodiment, the additional data includes an indication of food and beverages ingested or to be ingested within a defined period of time.

[0035] In an embodiment of the second aspect or any other embodiment, the indicator comprises data manually entered by a user.

[0036] In an embodiment of the second aspect or any other embodiment, the index includes nutritional data about the food or beverage obtained from a third party.

[0037] In an embodiment of the second aspect or any other embodiment, the nutritional data is entered automatically by the application.

[0038] In an embodiment of the second aspect or any other embodiment, the nutritional data is automatically entered into the application by scanning the nutritional information.

[0039] In an embodiment of the second aspect or any other embodiment, the additional data includes information describing the user's past physiological response patterns.

[0040] In an embodiment of the second aspect or any other embodiment, the additional data includes an indication of the level of user activity that has occurred or should occur within a specified period of time.

[0041] In an embodiment of the second aspect or any other embodiment, assigning the trust level includes assigning a higher trust to the additional information obtained through machine-to-machine communication than through manual input-to-machine communication.

[0042] In an embodiment of the second aspect or any other embodiment, the decision support recommendations include recommendations for consuming food or beverages.

[0043] In an embodiment of the second aspect or any other embodiment, the decision support recommendation comprises a recommendation to inject a calculated amount of insulin.

[0044] In an embodiment of the second aspect or any other embodiment, receiving the additional data includes receiving a plurality of additional data items of different types, and assigning the trust level to the additional data includes independently assigning a trust level to each of the additional data items.

[0045] In an embodiment of the second aspect or any other embodiment, calculating the decision support recommendation includes calculating the decision support recommendation only if a weighted average of the individual confidence levels exceeds a first threshold and if none of the individual confidence levels is below a second threshold.

[0046] In an embodiment of the second aspect or any other embodiment, the threshold value is selected based at least in part on the type of decision support recommendation being determined.

[0047] In an embodiment of the second aspect or any other embodiment, the calculated amount of insulin is calculated as part of a process for determining a decision support recommendation.

[0048] In a third aspect, there is provided a method for determining whether a decision support recommendation should be presented to a user for treatment of a diabetic condition, the method comprising: receiving a plurality of input data items affecting the diabetic condition of a user of a continuous glucose monitor, the input data items serving as input data to a process for determining a decision support recommendation; assigning a confidence level to each of the input data items; calculating a confidence metric based on the confidence level assigned to each of the input data items; determining a decision support recommendation based on the process and the input data; and presenting the decision support recommendation to the user on a user interface only if the confidence metric exceeds a threshold.

[0049] In an embodiment of the third aspect or any other embodiment, calculating the confidence metric comprises calculating the confidence metric based on an average of the confidence levels of each of the input data items.

[0050] In an embodiment of the third aspect or any other embodiment, the average of the confidence levels is a weighted average.

[0051] In an embodiment of the third aspect or any other embodiment, the plurality of input data items includes sensor data from a glucose sensor in a continuous glucose monitor.

[0052] In an embodiment of the third aspect or any other embodiment, assigning the confidence level includes assigning the confidence level to the sensor data based on another factor selected from the group including signal quality, sensor calibration, connectivity, age of the glucose sensor, and expected useful life of the glucose sensor.

[0053] In an embodiment of the third aspect or any other embodiment, determining and presenting the decision support recommendation includes determining a decision support recommendation even if the reliability metric does not exceed a threshold, while presenting the decision support recommendation to a user on a user interface only if the reliability metric exceeds a threshold.

[0054] In an embodiment of the third aspect or any other embodiment, determining and presenting the decision support recommendation includes determining the decision support recommendation only if the reliability metric exceeds a threshold and presenting the decision support recommendation to the user only if the reliability metric exceeds a threshold.

[0055] In an embodiment of the third aspect or any other embodiment, assigning a confidence level to each of the input data items comprises assigning a confidence level to at least one of the input data items based on an uncertainty in the accuracy of a value of the at least one input data item.

[0056] In an embodiment of the third aspect or any other embodiment, assigning a confidence level to each of the input data items includes assigning a confidence level to at least one of the input data items based on a degree of uncertainty of an impact of the at least one input data item on the user's diabetic condition.

[0057] In an embodiment of the third aspect or any other embodiment, assigning the trust level includes assigning the trust level based at least in part on the identity of a source from which the additional data was obtained.

[0058] In an embodiment of the third aspect or any other embodiment, the additional data includes an indication of food and beverages consumed or to be consumed within a defined period of time.

[0059] In an embodiment of the third aspect or any other embodiment, the indicator comprises data manually entered by a user.

[0060] In an embodiment of the third aspect or any other embodiment, the index includes nutritional data about the food or beverage obtained from a third party.

[0061] In an embodiment of the third aspect or any other embodiment, the nutritional data is entered automatically by the application.

[0062] In an embodiment of the third aspect or any other embodiment, the nutritional data is automatically entered into the application by scanning the nutritional information.

[0063] In an embodiment of the third aspect or any other embodiment, the additional data includes information describing the user's past physiological response patterns.

[0064] In an embodiment of the third aspect or any other embodiment, the additional data includes an indication of the level of user activity that has occurred or should occur within a specified period of time.

[0065] In an embodiment of the third aspect or any other embodiment, assigning the trust level includes assigning a higher trust to the additional information obtained through machine-to-machine communication than through manual input-to-machine communication.

[0066] In an embodiment of the third aspect or any other embodiment, the decision support recommendations include recommendations for consuming food or beverages.

[0067] In an embodiment of the third aspect or any other embodiment, the decision support recommendation comprises a recommendation to inject a calculated amount of insulin.

[0068] In an embodiment of the third aspect or any other embodiment, presenting a decision support recommendation only if the assigned reliability of the additional data exceeds a threshold further includes determining when to present the decision support recommendation based at least in part on the reliability level.

[0069] In an embodiment of the third aspect or any other embodiment, presenting the decision support recommendation includes presenting the decision support recommendation only if a weighted average of the individual confidence levels exceeds a first threshold and if none of the individual confidence levels is below a second threshold.

[0070] In an embodiment of the third aspect or any other embodiment, the threshold value is selected based at least in part on the type of decision support recommendation being calculated.

[0071] In an embodiment of the third aspect, or any other embodiment, a higher threshold is assigned to decision support recommendations that are therapeutic than to decision support recommendations that are supportive.

[0072] In an embodiment of the third aspect or any other embodiment, the threshold value is selected based at least in part on the type of decision support recommendation being determined.

[0073] In an embodiment of the third aspect or any other embodiment, presenting a decision support recommendation includes presenting a more conservative support recommendation that promotes patient safety when the assigned confidence level is within a range specified by the upper and lower thresholds.

[0074] In an embodiment of the third aspect or any other embodiment, the plurality of additional data items includes an indication of food and beverages ingested or to be ingested within a defined period of time.

[0075] In an embodiment of the third aspect or any other embodiment, assigning the confidence levels includes assigning a lower confidence level to the input data items obtained from the food and beverage indicator when the indicator includes data manually entered by a user into the food and beverage indicator, and assigning a higher confidence level to the input data items obtained from the food and beverage indicator when the indicator includes nutritional data about the food and beverage obtained from a third party or nutritional data automatically entered by the application.

[0076] In an embodiment of the third aspect or any other embodiment, assigning confidence levels to the input data items comprises using machine learning techniques to assign confidence levels to the input data items.

[0077] In an embodiment of the third aspect or any other embodiment, the glucose concentration level is obtained from a glucose sensor in a continuous glucose monitor.

[0078] In an embodiment of the third aspect or any other embodiment, the user interface is integrated into a display device that receives sensor data from the glucose sensor.

[0079] In an embodiment of the third aspect or any other embodiment, the display device receives the sensor data from the glucose sensor via a wireless communication link.

[0080] In an embodiment of the third aspect or any other embodiment, presenting the decision support recommendations to the user on a user interface includes presenting the decision support recommendations to the user on a user interface of a display device.

[0081] In an embodiment of the third aspect or any other embodiment, a continuous glucose sensor application resident on the display device receives the sensor data and causes decision support recommendations to be presented to the user on a user interface.

[0082] In a fourth aspect, a system for providing decision support recommendations regarding a user's diabetic condition, the system comprising: a continuous glucose monitor (CGM) application running on a mobile device, the CGM application configured to receive data from a glucose sensor at least periodically or irregularly and display glucose calibration data; and a decision support recommendation application running as a subroutine within the CGM application or running as a parallel process within the CGM application on the mobile device and receiving data from the CGM application, the decision support recommendation application configured to perform a method described in the embodiments of the first, second, or third aspects or any other embodiment. [Brief description of the drawings]

[0083] [Figure 1] FIG. 1 is a block diagram of a preferred embodiment integrated system including a continuous glucose sensor and a drug delivery device. [Diagram 2] FIG. 1 is a front view of an electronic device configured for use with the present systems and methods. [Diagram 3]FIG. 2 is a functional block diagram of the electronic device of FIG. 1. [Figure 4] 2 illustrates a logical diagram of certain components of the system of FIG. 1. [Diagram 5] Indicates the categories of parameters or variables that may be used in the calculation of decision support recommendations (DSRs). [Figure 6A] 1 is an exemplary user interface in which decision support recommendation prompts are presented. [Figure 6B] 1 is an exemplary user interface in which decision support recommendation prompts are presented. [Figure 7] 1 is a flow chart illustrating a method according to present principles. [Figure 8] 1 is a graph of various input data items as a function of time that can be used to determine decision support recommendations. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0084] Continuous glucose monitoring (CGM) systems that display current glucose levels provide users with continuous information regarding their current metabolic state, and therefore, further provide continuous indicator information regarding trends in glucose level changes. This information facilitates the myriad therapeutic decisions that users with diabetes must make regarding food intake, exercise, and insulin delivery.

[0085] Even for users whose treatment does not include the use of insulin, monitoring can help indicate the effectiveness of food choices, physical activity, medications, and stress management efforts. Monitoring at different times on different days (as opposed to monitoring at the same time every day) can provide valuable information about blood glucose levels throughout the day. It can also allow for the observation of patterns in blood glucose levels, such as highs or lows at particular times of the day and / or associated with particular events.

[0086] The user can achieve improved metabolic control and reduced risk of acute metabolic deterioration by adapting the treatment in a personalized and flexible manner according to the current glucose level and the trend of the change of the glucose level. With this information, the user can determine, for example, whether he is at risk of a hypoglycemic event, whether the selected insulin dose was appropriate for the current diet, and whether physical activity needs to be supplemented, for example, with the intake of additional carbohydrates. In other words, the CGM system provides real-time, continuous feedback that the user can apply to reach the desired goal.

[0087] CGM systems permit retrospective analysis of large amounts of data, allowing the user and their medical team to generate statistics and detect patterns / trends that facilitate adjustments to treatment. This analysis can generate a variety of graphs and statistics, including, for example, glucose averages, standard deviations, percentage of time spent above, below and within target ranges, number of excursions above and below target ranges, detailed daily reports, and modal day reports that overlay trend data from multiple days onto a single chart. Notably, this analysis can generate alarms to alert the user to impending high or low glucose levels.

[0088] In addition to blood glucose level measurements, retrospective analysis will generally also require external information not obtained directly from the analyte sensors of the CGM system. The external data may correspond to past or present user indicators, such as, for example, how the user is feeling, what the user has eaten, the amount of user activity, etc. The external data, which may be user-entered data, derived data, or data obtained from other applications or sensors or other sources, may also constitute many other parameters, as described below.

[0089] Analysis of the large amount of historical information acquired by a CGM system may provide an understanding of patterns that allow for even greater metabolic control than is possible from simply recognizing current glucose levels and trends in changing glucose levels. For example, looking at CGM data after a meal, especially the first 1-2 hours, may reveal both the timing and magnitude of postprandial spikes. Users and clinicians may evaluate the postprandial effects of different food types for meal planning purposes and determine optimal timing of mealtime boluses, and / or the need for fast-acting insulin at mealtimes. As another example, analyzing glucose levels 3-4 hours after a bolus may provide useful insight regarding bolus dosing. Glucose levels that are consistently above or below target at this time may indicate a need to adjust the meal dose (insulin to carbohydrate ratio) or correction dose (insulin sensitivity). As yet another example, for those who use incretins to regulate glucose levels, glucose levels should remain fairly stable after a meal, so a review of postprandial patterns should indicate whether an adequate dose is being taken. In those taking fast-acting insulin along with incretins, the timing of insulin dosing can be evaluated. A drop in glucose immediately after eating, followed by a rise over a few hours, may indicate a need to delay or extend the delivery of insulin at mealtimes.

[0090] Analysis of historical information can also be used to better understand the extent to which exercise, and different forms of exercise, contribute to short-term and long-term glucose patterns. Similarly, the effects of stress and illness on glucose control can be analyzed.

[0091] Evaluating all available analytical data, including both glucose levels and trends, as well as external data, and using it to provide guidance to the user regarding the user's daily routine to improve disease management and identify the need for therapeutic intervention in a timely manner can be a challenging task for the user as well as for the clinician. To address this issue, the CGM system may include decision support functionality that uses analyte sensor data and / or other data, e.g., external data, as inputs to algorithms to provide user prompts on a display and / or provide commands to mechanical devices such as drug (e.g., insulin) pumps. The user prompts may include treatment recommendations for therapeutic interventions such as eating a snack, exercising, or taking appropriate medications. It is noted that the decision support recommendations may include recommendations provided for non-adjunctive use of the CGM system for treatment decisions that do not require self-monitoring of blood glucose (SMBG).

[0092] Instead of presenting specific treatment recommendations, in some cases, the decision support recommendations may include general guidelines to facilitate and encourage healthy behaviors on the part of the user. For example, some decision support recommendations may encourage the user to check their glucose levels frequently or to learn what factors cause the user's glucose levels to go up or down. Similarly, other decision support recommendations may encourage the user to eat a healthy diet, engage in physical activity such as taking a walk after meals, or take medications as directed. Some types of decision support recommendations may act as reminders, such as reminding the user that carbohydrates raise glucose levels. Still other types of decision support recommendations may be informational in nature, such as providing the user with information that allows the user to quickly and easily determine the serving size of a food. For example, such a decision support recommendation may indicate that a serving of meat is approximately equal to a deck of cards, and that a serving of fruit or vegetables is approximately the size of a baseball. As another example, such informational decision support recommendations may identify certain foods that are high or low carbohydrate foods, or high or low glycemic index foods.

[0093] As decision support algorithms become more commonly used in CGM systems, safety concerns become an important issue. It is essential that any treatment recommendations provided to a user by a decision support algorithm only promote improved health and safety, and that treatment recommendations that may cause medical complications be avoided. In some cases, more conservative or protective recommendations may need to be provided to ensure the user's safety than would typically be provided by a clinician given the same analytical data.

[0094] One factor that may be used by a CGM system to determine whether a particular decision support recommendation is safe enough to be presented to a user is the reliability of the input data (e.g., analyte sensor data and external data) used to generate the decision support recommendation. As described in more detail below, a reliability rating or level may be assigned to each individual input data item used to generate the recommendation. The reliability rating or level may be determined based on a variety of factors, such as the source of the input data item (e.g., data manually entered by the user versus sensor data), the accuracy of the data (e.g., whether nutritional data regarding diet or levels of physical activity are provided in qualitative or quantitative terms), etc.

[0095] Illustrative Continuous Glucose Monitoring System Exemplary embodiments disclosed herein relate to the use of glucose sensors to measure the concentration of glucose or a substance indicative of the concentration or presence of another analyte. In some embodiments, the glucose sensor is a continuous device, e.g., subcutaneous, transdermal, transcutaneous, non-invasive intraocular and / or intravascular (e.g., intravenous) device. In some embodiments, the device can analyze multiple intermittent blood samples. The glucose sensor can use any method of glucose measurement, including enzymatic, chemical, physical, electrochemical, optical, photochemical, fluorescence-based, spectrophotometric, spectroscopy (e.g., optical absorption spectroscopy, Raman spectroscopy, etc.), polarimetry, calorimetry, iontophoresis, radiometric, etc.

[0096] The glucose sensor can provide a data stream indicative of the concentration of the analyte in the host using any known detection method, including invasive, minimally invasive, and non-invasive sensing techniques. The data stream is typically a raw data signal that is used to provide a useful value of the analyte to a user, such as a patient or a medical professional (e.g., a physician) who may be using the sensor.

[0097] Although many of the descriptions and examples are directed to glucose sensors capable of measuring the concentration of glucose in a host, the systems and methods of the embodiments may be applied to any measurable analyte. Some exemplary embodiments described below utilize an implantable glucose sensor. However, it should be understood that the apparatus and methods described herein may be applied to any device capable of detecting the concentration of an analyte and providing an output signal representative of the concentration of the analyte.

[0098] As previously mentioned, in some embodiments, the analyte sensor is an implantable glucose sensor as described with reference to U.S. Patent No. 6,001,067 and U.S. Patent Publication No. 2011-0027127-A1. In some embodiments, the analyte sensor is a transcutaneous glucose sensor as described with reference to U.S. Patent Publication No. 2006-0020187-A1. In still other embodiments, the analyte sensor is a dual electrode analyte sensor as described with reference to U.S. Patent Publication No. 2009-0137887-A1. In still other embodiments, the sensor is configured to be implanted within a host blood vessel or outside the body as described in U.S. Patent Publication No. 2007-0027385-A1. These patents and publications are incorporated herein by reference in their entirety.

[0099] The following description and examples describe the present embodiments with reference to the drawings, in which reference numbers label elements of the present embodiments, and which are reproduced below in connection with a discussion of the corresponding drawing features.

[0100] FIG. 1 is a block diagram of a preferred embodiment integrated system including a continuous glucose sensor and a drug delivery device. This is an exemplary environment in which some embodiments described herein may be implemented. Here, an analyte monitoring system 100 includes a continuous analyte sensor system 8. The continuous analyte sensor system 8 includes a sensor electronics module 12 and a continuous analyte sensor 10. The system 100 may also include other devices and / or sensors, such as a drug delivery pump 2 and a reference analyte meter 4. The continuous analyte sensor 10 may be physically connected to the sensor electronics module 12 and may be integral (e.g., non-removably attached) or removably attached to the continuous analyte sensor 10. Alternatively, the continuous analyte sensor 10 may be physically separate from the sensor electronics module 12 and electronically coupled, such as via inductive coupling. Additionally, the sensor electronics module 12, the drug delivery pump 2 and / or the analyte reference meter 4 may communicate with one or more additional devices, such as any or all of the display devices 14, 16, 18, and / or 20. Display devices 14, 16, 18, and 20 generally include a processor, memory, storage, and other components sufficient to run applications including decision support modules.

[0101] In some implementations, the system 100 of FIG. 1 may also include a cloud-based processor 22 configured to analyze analyte data, drug delivery data, and / or other user-related data provided directly or indirectly over the network 24 from one or more of the sensor system 8, drug delivery pump 2, reference analyte meter 4, and display devices 14, 16, 18, 20. Based on the received data, the processor 22 may further process the data, generate reports providing statistics based on the processed data, trigger notifications to electronic devices associated with the host or the host's caretaker, or provide the processed information to any of the other devices of FIG. 1. In some exemplary implementations, the cloud-based processor 22 includes one or more servers. When the cloud-based processor 22 includes multiple servers, the servers may be geographically local or remote from one another. The network 24 may include any wired and wireless communication medium for transmitting data, including WiFi networks, cellular networks, the Internet, and any combination thereof.

[0102] In some example implementations, sensor electronics module 12 may include electronic circuitry associated with measuring and processing the data generated by continuous analyte sensor 10. This generated continuous analyte sensor data may also include algorithms that may be used to process and calibrate the continuous analyte sensor data, although these algorithms may be provided in other manners as well, such as by devices 14, 16, 18, and / or 20. Sensor electronics module 12 may include hardware, firmware, software, or a combination thereof, to provide measurements of the level of an analyte via a continuous analyte sensor, such as a continuous glucose sensor.

[0103] As mentioned above, sensor electronics module 12 may be coupled (e.g., wirelessly, etc.) with one or more devices, such as any or all of display devices 14, 16, 18, and 20. Display devices 14, 16, 18, and / or 20 may be configured to process and present information, such as sensor information, transmitted by sensor electronics module 12 for display on the display device. Display devices 14, 16, 18, and 20 may also trigger alarms and / or provide decision support recommendations based on the analyte sensor data.

[0104] In FIG. 1 , display device 14 is a key fob-like display device, display device 16 is a handheld application-specific computing device 16 (e.g., a DexCom G4™ Platinum receiver available from DexCom, Inc.), display device 18 is a general-purpose smartphone or tablet computing device 20 (e.g., a phone running the Android™ OS, Apple™ iPhone™, iPad™, or iPod™ Touch™ available from Apple, Inc.), and display device 20 is a computer workstation 20. In some exemplary implementations, the relatively small key fob-like display device 14 may be a computing device embodied in a watch, belt, necklace, pendant, jewelry, adhesive patch, pager, key fob, plastic card (e.g., credit card), identification (ID) card, or the like. This small display device 14 may include a relatively small display (e.g., smaller than display device 18) and may be configured to display a limited set of displayable sensor information, such as numeric values ​​26 and arrows 28. Some systems may also include a wearable device 21, such as described in U.S. Provisional Patent Application No. 61 / 904,341, entitled "Devices and Methods for Continuous Analyte Monitoring," filed November 14, 2013, and U.S. Patent No. 10,004,435, which claims priority thereto, the entire disclosures of which are expressly incorporated herein by reference in their entireties. Wearable device 21 may include any device(s) that is worn on or integrated into a user's optics, clothing, and / or body.Examples of devices include wearable devices, anklets, eyeglasses, rings, necklaces, armbands, pendants, belt clips, hair clips / hair ties, pins, cufflinks, tattoos, stickers, socks, sleeves, gloves, clothing (e.g., shirts, pants, underwear, bras, etc.), "clothing jewelry" such as zipper tabs, buttons, watches, shoes, contact lenses, subdermal implants, eyeglass lenses, cochlear implants, shoe inserts, orthodontic appliances (mouth), braces (body), medical wrappings, sports bands (wristbands, headbands), hats, bandages, hair pieces, nail polish, artificial joints / body parts, orthopedic pins / devices, implantable cardiac or nervous system devices, etc. Small display device 14 and / or wearable device 21 may include a relatively small display (e.g., smaller than display device 18) and may be configured to display graphical and / or numerical representations of sensor information such as numbers 26 and / or arrows 28. In contrast, display devices 16, 18, and 20 may be larger display devices that may be capable of displaying a larger set of displayable information, such as trend graphs 30 depicted on handheld receiver 16, in addition to other information such as numbers and arrows.

[0105] It will be understood that any other user equipment (e.g., computing device) configured to present at least information (e.g., drug delivery information, individual self-monitoring analyte measurements, heart rate monitor, caloric intake monitor, etc.) may be used in addition to or instead of those described with reference to FIG. 1.

[0106] 1, the continuous analyte sensor 10 includes a sensor for detecting and / or measuring an analyte, and the continuous analyte sensor 10 may be configured to continuously detect and / or measure an analyte as a non-invasive, subcutaneous, transcutaneous, and / or intravascular device. In some exemplary implementations, the continuous analyte sensor 10 may analyze multiple intermittent blood samples, although other analytes may be used as well.

[0107] In some exemplary implementations of FIG. 1, the continuous analyte sensor 10 may include a glucose sensor configured to measure glucose in blood using one or more measurement techniques, such as enzymatic, chemical, physical, electrochemical, fluorescent, spectrophotometric, polarimetric, calorimetric, iontophoretic, radiometric, immunochemical, etc. In implementations in which the continuous analyte sensor 10 includes a glucose sensor, the glucose sensor may include any device capable of measuring the concentration of glucose and may measure glucose using a variety of techniques, including invasive, minimally invasive, and non-invasive sensing techniques (e.g., fluorescence monitoring), to provide data, such as a data stream indicative of the concentration of glucose in a host. The data stream may be a raw data signal, converted into a calibrated and / or filtered data stream used to provide glucose values ​​to a host, such as a user, patient, or caregiver (e.g., a parent, relative, guardian, teacher, doctor, nurse, or any other individual interested in the health of the host). Additionally, the continuous analyte sensor 10 may be implanted as at least one of the following types of sensors: an implantable glucose sensor, a transcutaneous glucose sensor implanted in a host vasculature or external to the body, a subcutaneous sensor, a refillable subcutaneous sensor, an intraocular sensor, or an intravascular sensor.

[0108] In some implementations of FIG. 1, the continuous analyte sensor system 8 includes a Dexcom G4™ Platinum glucose sensor and transmitter commercially available from Dexcom, Inc. or in such sensor systems for continuously monitoring the glucose level of a host.

[0109] FIG. 2 illustrates one embodiment of an electronic device 200 configured for use with the present systems and methods. The electronic device 200 includes a display 202 and one or more input / output (I / O) devices, such as one or more buttons 204 and / or switches 206, that perform one or more functions when activated or clicked. In some embodiments, the electronic device 200 may be a mobile communication device. For example, in the illustrated embodiment, the electronic device 200 is a smartphone, and the display 202 includes a touch screen that also functions as an I / O device. In other embodiments, the electronic device 200 may include one or more devices other than a smartphone, such as a receiver for a CGM system, a smart watch, a tablet computer, a mini tablet computer, a handheld personal digital assistant (PDA), a game console, a multimedia player, a wearable device as described above, a screen in an automobile or other vehicle, etc. Although the electronic device 200 is illustrated as a smartphone, the electronic device 200 may be any of the other electronic devices mentioned herein and / or incorporate any or all of the functionality of the other electronic devices (including cases where some or all of the functionality is embodied on a remote server).

[0110] 3 is a block diagram of the electronic device 200 shown in FIG. 2 and illustrates its functional components according to some embodiments. The electronic device 200 includes a display 202 and one or more input / output ("I / O") device(s) 204, 206, as described above with respect to FIG. 2. The display 202 can be any device capable of displaying output, such as an LCD or LED screen. The input / output (I / O) devices 202, 204, 206 can include, for example, a keyboard (not shown), one or more buttons 204, one or more switches 206, etc. In embodiments including a touch screen, the display 202 also functions as an I / O device.

[0111] The electronic device 200 further includes a processor 208 (also referred to as a central processing unit (CPU)), a memory 210, a storage device 212, a transceiver 214, and may also include other components or devices (not shown). The memory 210 is coupled to the processor 208 via a system bus or a local memory bus 216. The processor 208 may be or include one or more programmable general-purpose or special-purpose microprocessors, digital signal processors (DSPs), programmable controllers, application specific integrated circuits (ASICs), programmable logic devices (PLDs), or the like, or a combination of such hardware-based devices.

[0112] The memory 210, during execution, provides the processor 208 with access to data and program information stored in the memory 210. Typically, the memory 210 includes random access memory (RAM) circuitry, read only memory (ROM), flash memory, or the like, or a combination of such devices.

[0113] Storage device 212 may include one or more internal and / or external mass storage devices, which may be or include any conventional medium for storing large amounts of data in a non-volatile manner. For example, storage device 212 may include conventional magnetic disks, optical disks, magneto-optical (MO) storage, flash-based storage devices, or any other type of non-volatile storage device suitable for storing structured or unstructured data. Storage device 212 may also include storage in the "cloud" using so-called cloud computing. Cloud computing relates to computing power that provides an abstraction between computing resources and their underlying technology architecture (e.g., servers, storage, networks) and enables convenient, on-demand network access to a shared pool of configurable computing resources that can be rapidly provisioned and released with minimal management effort or interaction with a service provider.

[0114] The electronic device 200 may perform various processes, such as, for example, data correlation, pattern analysis, and other processes. In some embodiments, the electronic device 200 may perform such processes by itself. Alternatively, such processes may be performed by one or more other devices, such as the one or more cloud-based processors 22 described above. In yet other embodiments, these processes may be performed in part by the electronic device 200 and in part by other devices. Various exemplary processes are described herein with reference to the electronic device 200. It should be understood that these exemplary processes are not limited to being performed solely by the electronic device 200. Furthermore, as used herein, the term "electronic device" should be construed to include other devices with which the electronic device 200 interacts, such as one or more cloud-based processors, servers, and the like.

[0115] The electronic device 200 may also include other devices / interfaces for performing various functions. For example, the electronic device 200 may include a camera (not shown).

[0116] The transceiver 214 enables the electronic device 200 to communicate with other computing systems, storage devices, and other devices over a network. Although the illustrated embodiment includes the transceiver 214, in alternative embodiments, a separate transmitter and a separate receiver may be used in place of the transceiver 214.

[0117] In some embodiments, the processor 208 can execute various applications, e.g., a CGM application, that may be downloaded to the electronic device 200, such as via the Internet and / or a cellular network. Data for the various applications may be shared between the electronic device 200 and one or more other devices / systems, and stored by the storage 212 and / or in one or more other devices / systems. The CGM application may include a decision support module and / or may include sufficient processing to operate the decision support evaluation functions and methods as described below.

[0118] In some of the present embodiments, the sensor 10 of the continuous analyte sensor system 8 of FIG. 1 is inserted into the skin of the host. A new sensor session is then initiated with the sensor 10, the sensor electronics 12, and the electronic device 200. Many techniques can be used to initialize the sensor 10. For example, initialization can be triggered when the sensor electronics 12 engages the sensor 10. In another example, initialization can be triggered by a mechanical switch, such as a switch (not shown) on a snap-fit ​​base that receives the sensor electronics 12. The switch automatically activates when the sensor electronics 12 snaps into the base. In another example, initialization can be menu-driven, and the user can be prompted by a user interface on the display 202 of the electronic device 200 to initiate initialization by making a selection on the user interface, such as by pressing a button or touching a designated area on the display 202 (which may include a touch screen). In another example involving a non-invasive sensor applied to the wearer's skin, the sensor 10 can sense contact with the skin and automatically start. Additionally, the analyte sensor system 8 can detect the use of a new sensor 10 using any of the techniques described above, automatically prompt the user to confirm the new sensor session via a prompt on the user interface of the system 8, and begin initialization in response to user confirmation in response to the prompt. Additional examples of initializing the sensor 10 can be found in U.S. Patent Application Serial No. 13 / 796,185, filed March 12, 2013, which corresponds to U.S. Patent No. 9,433,376, the entire disclosures of which are incorporated herein by reference.

[0119] FIG. 4 illustrates an exemplary logic diagram of a continuous analyte monitoring system 100, particularly illustrating components involved in depicting the determination and calculation of sensor results, and the determination of decision support recommendations based on the results and other factors. In particular, measurements from the sensor 10 are processed by sensor electronics 12 and sent to a mobile device 18, which is generally a smartphone. Although a smartphone is described here, it will be understood that any of the various electronic devices described above may be used to receive and display sensor data or other data and output results, as well as alerts and alarms based thereon. Additionally, a smartphone (or a device having similar smartphone capabilities) can transmit displayed notifications, results, recommendations, alerts, and alarms to various devices coupled thereto, for example, via Bluetooth™. Such devices include head-mounted displays such as Google Glass™, watches, and the like.

[0120] The mobile device 18 executes a CGM application 209 in which various monitoring and display functions are provided based on signals received from the sensor electronics 12. As part of this CGM application, a decision support module 211, which may also be a processor module, is provided for performing the decision support functions described herein. Although a decision support module is described, it will be understood that it may be replaced by any suitable functionality for performing the methods described herein. In some cases, the functionality of the decision support recommendation module 211 may be implemented as an application that runs as a subroutine within the CGM application or as an application that runs as a parallel process within the CGM application.

[0121] The mobile device 18 includes a display 202 for displaying notifications, results, recommendations, and alerts / alarms. The display 202 is depicted as a display screen and thus generally renders results visually, but it will be understood that notifications, outputs, results, and in more urgent cases alerts / alarms may also be communicated using other means, such as audibility. They may be communicated as an audible version of the displayed text or numbers. Alternatively, a tone or other sound, even a song or ring tone, may be rendered to the user as a personalized indication of the blood glucose level.

[0122] The mobile device 18 may further include memory 210 or storage 212 for retrieval and use of historical data, including user-entered data, as described in more detail below. Because the mobile device 18 is in network communication with various servers, historical data may also be retrieved from a network server 222. In addition to historical data, the server (or other network source) may further provide other external data that may be input into the determination that leads to the notification presented on the display 202.

[0123] The display 202 itself may provide an interface for a user to input data, for example using a touch screen interface, which may also be entered via buttons and switches 204 and 206, respectively. Some smartphones and many other computing devices may use a separate keyboard for the same purpose.

[0124] Signal processing may occur using the sensor electronics 12, using the mobile device 18, or using a combination of the two. Signal processing may also be performed in the cloud, for example, on a server 222 or other network source. In many cases, however, initial processing of the raw sensor signals, such as calibration, smoothing, or filtering, is performed by the sensor electronics 12, and an application on the mobile device 18 converts the signals received from the sensor electronics 12 into values ​​that are later shown on the display 202.

[0125] Input to the decision support module 5 illustrates how the measured blood glucose level is combined with other parameters or variables to result in a calculated or otherwise determined decision support recommendation value 252 that is then presented on a display screen of the mobile device. The calculation or determination is generally performed by the decision support module 211 on the mobile device 18, but may be determined in whole or in part by the server 222, or possibly by the sensor electronics 12. It will be understood that not all parameters and variables are input into all implementations of the determination of the decision support recommendation value 252.

[0126] Various parameters and variables are described, followed by examples of how they may be combined to provide decision support recommendations. Without intending to limit the scope of the configurations in any way, it is believed that particularly useful combinations include combinations of the current glucose value or the glucose value with the first derivative of the glucose value with respect to time. However, as will be appreciated by those of skill in the art given this teaching, numerous combinations are useful, and thus the scope of the invention should not be limited by the specific examples. In FIG. 5, decision support recommendations 252 are shown as being based at least on data 254 corresponding to a current measurement of glucose, and / or data 256 corresponding to historical measurements of glucose, and / or data 258 that is not directly related to glucose measurements and thus is referred to as "external data." Data 254 is generally a current measurement of glucose, measured, for example, in mg / dL. Data 256 corresponds to historical measurements of glucose, and may be divided into data 262 referred to as "recent history" measured glucose data and data 264 referred to as "old history" measured glucose data. Recent historical data 262 may have been measured over the minutes or hours prior to the currently measured glucose data 254, and therefore may be particularly useful for current trend analysis. Older historical data 264 may have been measured over the days, weeks, months, or years prior to the current measurement, and therefore may be particularly useful in calculating or determining overall patterns or trends (in which determination data 262 may be used).

[0127] The current measured glucose data 254 and the recent historical measured glucose data 262 can be used to calculate other types of data 266 based on current trends, such as data 268 corresponding to the rate of change of the glucose data over time, e.g., first derivative with respect to time, second derivative with respect to time, etc.

[0128] Data 258 may correspond to past or present indicators of the user, such as, for example, how the user is feeling, what the user has eaten, etc. Thus, while data 258 may have an indirect correlation with glucose levels, the data is not, in a functional sense, directly based on measured glucose values. Data 258 may also constitute many other variables, as described below.

[0129] Various parameters and variables based on the above types of data are described below. Again, it should be noted that the determination of a decision support recommendation in a particular implementation need not include all of the various types of data described, and in many cases will include only two or three types of data. Furthermore, the following description is merely exemplary, and other data types than those described below can be used. In particular, the calculation of the decision support recommendation can be performed by an algorithm on the mobile device, for example, as described above, which may take into account several or many variables in its determination of the decision support recommendation. These variables are evaluated algorithmically at or near the same time, but the following description will discuss in part the sequential effects of the variables on each other and on the determined decision support recommendation. In the context of a user interface of an electronic device such as a smartphone, the calculated decision support recommendation can lead to a notification presented on the user interface of the mobile device, and possibly even to an "actionable alert" (or alarm) that is displayed to the user and suggests one or more actions to be taken. In some implementations, the notification the user sees can simply be an indication of the user's condition, for example, that the user is in a normal state and no action needs to be taken. In other cases, the presentation may be of an alert or alarm condition. By unlocking the mobile device, performing a "swipe" action, or otherwise "drilling down" into the data underlying the existence of the alert or alarm condition, the user can see clear actions to be taken.

[0130] The first type of data that may be used, and the data involved in most implementations, is a glucose measurement. This first type of data may be in mg / dL or another numerical format, or may be processed or converted to derive another type of data that generally correlates with a glucose value. In some cases, the first type of data may also be used in a raw form received from the sensor electronics without significant processing, and / or may be processed by the sensor electronics. Processing may then be performed on a mobile device (or other device) running a decision support module or associated application to determine decision support recommendations, as needed. The first type of data may be further received from an intermediate module or conversion, for example, from another application running on a smartphone. Generally, this first type of data may be processed, for example, to calibrate, smooth, filter, or otherwise "clean up" the signal representing the data.

[0131] Generally, a current measurement of glucose is used, but it will be understood that the first type of data may also include one or more past values ​​of glucose, or even future values ​​of glucose as determined by a predictive algorithm, further details of which are described below.

[0132] In determining the decision support recommendation, all other factors being equal, a high glucose level tends to indicate a hyperglycemic state. In this case, the decision support recommendation may present the user with any of a variety of recommendations, such as suggesting that the patient hydrate, exercise, or take appropriate medication. Conversely, a low glucose level indicates a hypoglycemic state. In this case, the decision support recommendation may suggest, for example, that the patient eat a snack. Of course, many other schemes are understood and may be used in light of the present teachings.

[0133] Other types of data can be based on this first type of data, for example, the first derivative of the glucose value with respect to time can be used to determine the time rate of change of the glucose value, i.e., the glucose "velocity", i.e., whether the glucose value is increasing or decreasing and how quickly such changes are occurring. Thus, data values ​​representing the first derivative can be used in initial estimates of predictions of future glucose values ​​and also in determining decision support recommendations. Additionally, the direction and amplitude of the first derivative can be used to determine the weight of the same information in determining decision support recommendations.

[0134] 6A and 6B show examples of decision support recommendations that may be presented on a mobile device user interface 730 using glucose levels and their first derivatives as input data. In these examples, a current status 666 of the user's condition is presented. Additionally, for each status, one or more recommendations 668 are presented. In the example of FIG. 6A and 6B, three recommendations 668 are presented for each status 666. In some embodiments, the recommendations may be ordered based on safest to most aggressive.

[0135] The first derivatives of glucose values, and especially the higher order derivatives with respect to time, require that a certain amount of historical data be stored and used in the calculations. Such data is generally based on recent historical data, although it will be appreciated that older historical data may also be used and, as explained below, may provide useful information regarding user patterns, which may be analyzed abstractly or with respect to time periods, for example.

[0136] Another type of data that may be based on the glucose value, or more specifically, the first derivative, is the second derivative of the glucose value with respect to time, i.e., acceleration, which provides information about the rate at which changes in glucose levels are occurring and can often be used advantageously to determine how stable the changes in glucose levels are or whether they lead to excursions from the desired value.

[0137] In some implementations, the determination of the decision support recommendation may be based at least in part on the measured glucose value, the first or second derivative with respect to time of the measured glucose value, or both. In some implementations, the calculation of the decision support recommendation may be based on the above factors in combination with other factors described below. For example, another type of data based on the glucose value, e.g., the glucose value combined with a time derivative, may be used in combination with duration data (discussed below) to determine that a state has been reached where a DSR should be provided to the user. Similarly, the glucose value and / or the time derivative may be used in combination with food intake data to determine whether a decision support recommendation should be made.

[0138] Returning to the type of data based on glucose values, this may further include higher order derivatives with respect to time, glucose trace graphs over a period of time, the level and duration of the last significant glucose value excursion, e.g., the level of the last glucose peak, etc.

[0139] The duration that a measured glucose level remains within a predefined range is yet another type of data that may be used in the calculation and may be determined by analysis of glucose values, particularly values ​​over time. The predefined range may be arbitrarily defined, but may generally be indicative of a particular condition, e.g., high or low hyperglycemia, high or low hypoglycemia, or euglycemia.

[0140] More specifically, the time that a user spends in a range corresponding to a particular condition may provide important input to the decision support module, as this may indicate the need for therapeutic intervention, particularly in the case of hypoglycemic or hyperglycemic conditions. For example, if a user's glucose level is low, but the duration of a relatively low hypoglycemic event is long, a decision support recommendation to eat a snack may be presented based on its duration, since the likelihood of an excursion further downward is much higher. When using duration as a factor, the decision support module may use as input the duration itself, or the time that a particular condition exceeded a threshold duration, or other relevant parameters. Such data is generally made available via analysis of stored glucose values ​​over time.

[0141] Another type of data that may be used in the calculation or determination of a decision support recommendation corresponds to recent or historical events, particularly large excursions from expected or baseline values ​​of glucose levels. In particular, a user who has recently experienced a significant excursion is generally more likely to experience a significant current or future excursion. To address this issue, the decision support module may take such previous historical events into account in determining the decision support recommendation. For example, the level of the last glucose peak, or its duration (measured as a period above or within a threshold level), etc., may be used in such a determination. The level and / or duration of the last significant excursion or deviation of glucose values ​​from the baseline (or otherwise expected value) may be used in the determination, as it is often indicative of the user's current risk of a glycemic excursion, and in particular is an indicator of a higher likelihood of future excursions and deviations. As a subset of this type of data, the "last hypoglycemic / hyperglycemic event," including its level and duration, may be used in the determination. In any case, such data is generally made available via analysis of stored glucose values.

[0142] A further type of data that may be used in determining decision support recommendations uses older historical measured glucose values. In one case, and as described above, a pattern of glucose values ​​may be determined and used to inform a baseline from which excursions that may, for example, constitute a significant deviation from the baseline are measured. The pattern data may be, but is not necessarily, partly time-based or time-of-day based. In particular, users often follow very regular patterns based on eating, sleeping, exercise, or other activities that may affect glucose levels (such activities occurring at specific times of day). These may be advantageously used in determining whether excursions are expected outside of normal levels. If patterns are confirmed by time of day, the determined decision support recommendations may be more predictive and reliable and may provide more useful feedback. Such pattern data is generally made available through analysis of stored glucose values.

[0143] For example, a user may generally experience lower glucose values ​​in the morning than in the afternoon. The decision support module may adapt to this pattern to expect lower readings in the morning and higher readings in the afternoon. Similarly, a user may typically have an oatmeal meal in the morning, thus causing a spike in glucose values. The decision support module may determine that such a meal at approximately the same time each morning constitutes a pattern and may simply be taken into account and thus not require a decision support recommendation. Thus, taking into account the baseline patterned values, an analysis of a spike would not label it as a spike at all. Of course, other factors may affect the calculation of a decision support recommendation, which in combination may determine that a decision support recommendation should be made.

[0144] If the user's sleep patterns are known or can be determined by the decision support module, recommendations can be presented to the user, for example, suggesting the types of food that can be eaten before sleep or between dinner and sleep. Although eating and sleeping are disclosed elsewhere herein, it will be understood that patterns can be recognized or generated and used in determining decision support recommendations for other events, such as meetings, work, exercise, etc. Time of day information can be captured from any clock circuit or application, such as from a server, or from a mobile device or sensor electronics. Patterns can be based on detected events that occur in any kind of periodicity, such as during a daily, weekly, or monthly cycle. Such data is generally made available through analysis of stored glucose values, and various pattern recognition software applications can be advantageously used. In some cases, patterns can be detected and the user can be prompted to determine whether there is a particular cause for the pattern, such as usual meal times, exercise classes that occur at regular times, etc. Such prompts can be particularly used when the decision support module uses machine learning to determine daily or other periodic patterns or behaviors of a given user.

[0145] Along these lines, deviations outside of a recognized pattern may trigger similar user prompts, such as causing the decision support module to ask the user, "What did you change?" This may allow, for example, analysis and disambiguation of failed versus insufficient boluses.

[0146] Such pattern data may even provide anticipatory recommendations. For example, the pattern data may be used to suggest to the user where glucose levels are headed based on past historical data, which may result in decision support recommendations. For example, the decision support module may send an alert such as "It's almost 2pm. I often see lows at 2pm. Check X and take possible action Y," where X is a user-understandable variable, such as glucose level, and Y is the appropriate action to take given the currently determined decision support recommendation.

[0147] It will also be appreciated that other types of data related to deviations from normal glucose patterns may be used, but are not necessarily time-based. This may include cases where exercise (e.g., detected by movement or heart rate) is typically associated with reduced glucose levels. A "normal glucose pattern" may be learned for a particular user using known pattern recognition algorithms. Such deviations from normal patterns may then be defined and used as inputs into the determination of decision support recommendations. In some cases, a non-normative glycemic event may be a predictor of a higher risk state, at least in part due to the unexpected nature of the event, which may require a different type of output to the user, i.e., a different type of recommendation rendered on the display of the mobile device, or output to the insulin delivery device, i.e., pump. In this manner, the problem of dealing with non-time-based patterns may be effectively addressed.

[0148] Other types of data based on glucose values, or glucose values ​​measured over time, will be appreciated. For example, a glucose trace over a recent period, e.g., 6 hours, may be used to inform the calculation or determination of a current decision support recommendation.

[0149] Other types of data not based on glucose values ​​may be used in determining the decision support recommendation. A first category of such types of data is based on data from other sensors or sources or is input by the user. For example, the data may be of a type that includes anthropometric data corresponding to body measurements such as BMI or weight. Anthropometric data is particularly important for type II diabetes patients, but may have some influence for type I as well. In particular, for type II diabetes patients, changes in anthropometric measurements may have a significant influence on the determination of the decision support recommendation. For example, an improvement in the BMI of type II patients should lead to better decision support recommendations on average, all other aspects being equal. The measurements of anthropometric data may be captured semi-automatically, for example via a connected weight and height scale, or the values ​​for such BMI calculations may be manually entered by the user, for example at the user interface of the mobile device. Measurements may also be imported from other systems, including the cloud. In this way, the problem of treating all users the same, regardless of their anthropometric data, can be effectively addressed and solved.

[0150] Another type of data that may be used in determining decision support recommendations is data regarding the user's activity level, particularly the amount of activity, the type of activity, and the duration of activity (or a combination of these). In particular, quantification of the user's activity level can generally provide a better understanding of trends in glucose values. Activity information can feed into the determination of decision support recommendations and can also be useful for presenting to users who wish to receive additional information regarding why a particular decision support recommendation was made. This can also be used to determine what types of questions can be asked to help the user manage their diabetes.

[0151] Activity level measurements may be made via accelerometers, GPS data, or WiFi data indicative of location. Many mobile devices can, for example, count steps, or more generally determine whether a user of the mobile device is stationary, walking, running, driving, etc. It will be appreciated that such data, for example distances run, walked, or biked, may also be entered manually.

[0152] A related type of data is information about exercise, which is generally beneficial for diabetics and can help prevent hyperglycemia and hypoglycemia as well as assist in managing insulin delivery. However, exercise can have long-term effects on diabetes and can cause severe hypoglycemia after several hours in certain users. Therefore, due to this long time lag, it can be difficult to identify exercise as the cause of hypoglycemia. If exercise can be accurately detected, such as using the measurement devices described above, predictive analytics can be used to predict when it may start to affect glucose values ​​and therefore associated risk conditions that may require therapeutic intervention that can be recommended by the decision support module. For example, the decision support module may recommend that the user eat a light meal before or immediately after an exercise session, especially if the exertion level during the session was particularly high. Alternatively, or in addition, a recommendation may be presented to reduce insulin intake over the next few hours, as exercise may change insulin sensitivity. Exercise may generally be monitored using the same types of devices used to monitor activity and may include parameters such as duration of exercise, type of exercise, amount of calories burned, etc. It will be appreciated that such data may also be entered manually.

[0153] A further related type of data corresponds to sleep information. In particular, it is known that diabetic users are more likely to experience undetected hypoglycemic events during sleep. Movement, or lack thereof, and other factors can be used to detect sleep and assess risk accordingly. Other factors may include, for example, heart rate, user input, etc. A monitoring device, for example, a mobile device running a decision support module, may include a "night mode" feature or module that is instantiable by the user that may be used to assist in detecting sleep. Motion detection for such purposes may be performed as described above, for example, by using an accelerometer worn on the body. For example, a CGM sensor or transmitter may incorporate such an accelerometer or other motion detection circuitry. A phone or other motion detector placed adjacent to the user may detect how often the user moves, again indicative of sleep. In some cases, a motion detector in an alarm system may be used to provide such information and data. A heart rate monitor may measure changes in the user's heart rate. A user interface on the mobile device may also be used to assist in detecting sleep states. For example, if the user is not interacting with the mobile device at all, as determined by button presses, swipes, or other similar interactions, that may be associated with or consistent with a sleep state, or this may be learned by the decision support module as being associated with such a state. Conversely, if the user is interacting with the mobile device, the user may be considered not to be asleep. If the mobile device running the decision support module is equipped with a "sleep mode" feature, the user may activate it, in which case no assumptions about sleep or sleep detection are necessary.

[0154] Such a "sleep mode", "night mode", or sleep detection function can provide many advantages in certain implementations. In particular, by assigning a higher risk state to nighttime glycemic events than to daytime events, the system understands that the user is more likely to be unaware of their diabetes risk state and therefore glycemic events should be treated differently. In this way, the problem of user inattention during sleep, i.e., glucose excursion values ​​encountered during sleep, can be effectively addressed.

[0155] Another category of types of data that can be used in determining decision support recommendations corresponds to physiological data. One such physiological data type includes hydration information. In particular, dehydration is often associated with high blood glucose levels and can therefore be used to further inform the determination of decision support recommendations. Hydration information can be received, for example, from a Tanita BC-1000 body composition monitor in combination with a Garmin™ Connect system. It will be appreciated that such data can also be entered manually, at least at a qualitative level.

[0156] Another such physiological data type includes heart rate information. Heart rate can be indicative of exercise or other factors such as stress. If the heart rate or changes are due to exercise or other activity, it can be quantified using the activity monitor described above. Alternatively, the heart rate may be wirelessly communicated from a heart rate monitor or other application. In another implementation, it can be manually entered by the user using indicators such as "high heart rate," "normal heart rate," or a quantitative value if the user can measure it.

[0157] Another such physiological data type includes blood pressure information. In particular, the vascular effects of diabetes tend to increase the risk of hypertension. Therefore, monitoring it may be useful and can be factored into the determination of decision support recommendations. A variety of body-worn blood pressure monitors are available and can communicate blood pressure data in a wired or wireless manner to the device running the decision support module. Alternatively, the user may measure their own blood pressure and manually enter it into the device.

[0158] Further physiological data types include body temperature. Body temperature is often an indicator of disease, which in turn may affect diabetes risk status and therefore decision support recommendations. For example, body temperature and / or underlying disease may cause glycemic responses to various inputs or treatments to differ from those expected in other users or from those expected historically in the same user.

[0159] The body temperature data type may be captured by incorporating a temperature sensor in a sensor patch or using other such thermometers. This and other types of body temperature monitors may be found in U.S. Patent Application Serial No. 13 / 747,746, filed January 23, 2013, published as US2014 / 0005508A1, entitled "DEVICES, SYSTEMS, AND METHODS TO COMPENSATE FOR EFFECTS OF TEMPERATURE ON IMPLANTABLE SENSOR," owned by the assignee of the present application and incorporated herein by reference in its entirety. Temperature information may also be entered manually, either qualitatively or quantitatively.

[0160] Another type of data that may be used in determining decision support recommendations is context and behavioral information. In particular, such information may correspond to how the patient uses the mobile device, thus providing context to the particular data determined by the device. Behavioral input information may be obtained through the system and may include amount of interaction, glucose alert / alarm status, sensor data, number of screen hits, alarm analysis, events (e.g., characteristics associated with the user's response, time to response, glycemic control associated with the response, user feedback associated with the alarm, not acknowledging the alert / alarm within X minutes, time to acknowledge the alert / alarm, time in alert state, etc.), diabetes management data (e.g., CGM data, insulin pump data insulin sensitivity, patterns, activity data, calorie data), data regarding fatty acids, heart rate during exercise, IgG-antigliadin, stress levels (sweat / perspiration) from skin patch sensors, free amino acids, troponin, ketones, adipanectin, sweat, body temperature, etc. Input may be provided by sensors in data communication with the monitoring device. In some implementations, the information may be obtained through an intermediary, such as remote data storage.

[0161] Contextual information that may be provided as input for determining decision support recommendations includes a person's biology, location, sensing environment (e.g., light, sound levels), environmental data (e.g., weather, temperature, humidity, air pressure). Inputs may be received over a mesh network via peer-to-peer or machine-to-machine communications. Contextual information may include daily routine information obtained from a calendar application, which may change, especially between weekdays and weekends. Contextual information may include how often a person touches or grabs a device, even when not interacting with the device, based on sensed movements of the monitoring device.

[0162] The photo can provide context information. For example, a photo of one or more of a glucose meter reading, an insulin pen or pump JOB, a location (e.g., a gym, a park, a house, an Italian restaurant), or a meal can be used to provide context information. The photo can be processed, for example, to identify the caloric intake of the meal shown in the photo. The type of insulin used can also be provided to the monitoring system as a useful input to determining decision support recommendations. Context can also be provided by basal or bolus settings provided to or determined by the monitoring device.

[0163] Other inputs to the determination of decision support recommendations that constitute context / behavioral data may include data types referenced elsewhere for non-context / behavioral inputs, such as exercise information from a stationary bike or the like, glucose sensor information from a blood glucose (BG) meter or CGM, insulin delivery from an insulin delivery device, insulin on board calculations of the device, and other device provided or calculated information. Other context / behavioral data inputs to the determination of decision support recommendations may include hydration level, heart rate, target heart rate, internal temperature, outside temperature, outside humidity, analytes in the body, hydration input, power output (cycling), sweat rate, cadence, and adrenaline level, stress, illness / disease, metabolic rate / calorie burn rate, lipolysis rate, current weight, BMI, desired weight, target calories per day (expended), target calories per day (expanded), location, favorite foods, and level of exertion.

[0164] For any of the above-referenced behavioral or contextual inputs, the system may be configured to receive and / or generate analytical metrics based on the inputs. For example, a composite value may be generated based on the glucose level, temperature, and time of data that generated the index value for the user. The composite value may then be considered in determining decision support recommendations.

[0165] This information may be collected from a variety of sensors on and off the device, such as accelerometers, GPS, camera data, and third party tracking applications, including sleep cycle applications. For example, such tracking applications may use geolocation to determine context and behavior. Additionally, context and behavior may also be determined using social networking information available about the user, where a social networking feed associated with the user is configured to provide a source of data to the decision support module and / or provide an output thereto.

[0166] Other types of data that may be used in determining decision support recommendations include information regarding foods and beverages ingested, and insulin. Variables or parameters associated with these data types may include information regarding their amounts, their types, and the time and duration they were received.

[0167] In the case of foods and beverages consumed at a meal, these may be captured in a number of ways, for example, by the user manually entering food and beverage information into the device, for example, on a spreadsheet, using the mobile device's camera to capture a photo of the meal, or by entering data from a third party food application that may allow, for example, food items from a given restaurant (whose data has already been entered into the application) to be "checked off" as they are consumed and entered into the determination. In some cases, the user may be prompted for such information, for example, if the device detects a spike in glucose levels. The meal data may also be hypothesized using GPS or social networking data that indicates the user is near or has "checked in" to a known favorite restaurant (subject to confirmation by the user). The user may be prompted to confirm that they have ordered a "regular meal," and then the food data is automatically populated with the parameters of that meal, or alternatively, the prompt may offer an opportunity to enter other food choices if the user deviates from the usual options. In general, the meal data may be provided with details such as amount consumed, time of consumption, and other meal data that allows for a clinically relevant determination of the decision support recommendation. Such information can be used to effectively address the problems currently encountered in diabetes treatment based on its (and other) lack.

[0168] In some cases, inputs to the determination of decision support recommendations, such as food and beverages consumed, can be obtained by automatically aggregating data from third party applications without the user having to manually enter the input data. For example, when a user uses a third party application or a web browser on a display device to, for example, check in to a restaurant, order food for pickup or delivery, or click a photo of an Instagram food post, this information can be provided to a CGM application on the display device. This input data can be used by the decision support recommendation algorithm to provide relevant contextual information and insights to the user. For example, when a user checks in to a restaurant they have previously visited and this data is automatically provided to the CGM application, this data can be correlated with other available data to advise the user, for example, that their average glucose level was 189 mg / dl the last time they ate at this restaurant. In this way, information can be obtained from the user's existing application usage without additional burden on the user by requiring them to enter the same information into the CGM application. Such inter-application communication can also occur between a CGM application and a third party application associated with another device, such as a smartwatch or fitness tracker.

[0169] Another variable or parameter that may be factored into the determination of decision support recommendations is insulin levels. Data may be provided directly from an integrated insulin pump or from an EMR in the cloud. Such data may include information about the amount of insulin on board, insulin sensitivity, and past, current, and future planned basal and bolus levels. Data may be obtained by sensor data or other electronically communicated data, or may be provided by user input. One type of information that may be obtained from this data includes the time between the insulin bolus and the peak of the meal, which may be determined using insulin and glucose information.

[0170] A further type of data that may be used in determining decision support recommendations corresponds to stress levels. In particular, stress is known to affect diabetes and therefore the user's risk status. Such data may be provided via sensors, but may also be captured by asking the user to select from a variety of emoticons or other indicators of emotion. It may also be inferred from other sources, such as by analysis of events on the user's calendar or other regularly scheduled activities, e.g., work, exercise, family time, etc. Stress data may include information regarding the amount of stress, the type of stress, and the time the stress lasted.

[0171] A relevant type of data that may be used to determine decision support recommendations corresponds to a current health condition, which may overlap with a current emotional state. Such measurements may be captured manually via the device, including using the same types of emoticons as described above for stress, or may be imported from information in the cloud. Current health conditions and emotions, as well as anthropometric data, are known to have a significant impact, particularly on type II glucose control and insulin resistance. Health condition data may include information about current illnesses, the severity of the illness, how long the user has had the illness, etc.

[0172] Demographic data such as age or gender may also be used. In particular, demographic data may be collected from online stores, networks, or crowdsources, or manually entered into the device, and may provide information useful in determining decision support recommendations. For example, pediatric users are known to be prone to faster and higher blood sugar swings. As another example, it is believed that the risk status of older users, and especially type II diabetes users, may be higher with a particular blood sugar excursion compared to younger users with the same blood sugar excursion.

[0173] Another factor that may be used in determining decision support recommendations is the location of the sensor site. In particular, in some cases, the site or location of the CGM sensor may lead to sustained distinctions in blood glucose levels with respect to such location. These distinctions may be factored into the determination of decision support recommendations. Such data is generally manually entered by the user, but historical data may be used to avoid such user input if it is periodic and therefore a clear determination can be made. Additional details regarding the use of sensor site location may be found in U.S. Patent Application No. 61 / 904,396, filed November 14, 2013, entitled "INDICATOR AND ANALYTICS FOR SENSOR INSERTION IN A CONTINUOUS ANALYTE MONITORING SYSTEM AND RELATED METHODS," and U.S. Patent No. 9,480,401, which claims priority thereto, is owned by the assignee of the present application, and is incorporated herein by reference in its entirety.

[0174] Another factor that may affect the determination of the decision support recommendation is the cause of the increase or decrease in blood glucose level, if known. In this regard, it is noted that some changes in glucose levels are caused by stress and other changes are caused by food intake. Such data may be pre-processed or pre-associated prior to input into the decision support module, or may be associated during input. For example, food data may be processed in combination with glucose levels to determine whether the increase in glucose is caused by food or another cause, such as stress. The use of such data may effectively address problems previously seen by not considering such causes and effects.

[0175] The glucose values ​​(and derivative data) may be weighted by the decision support module based on signal quality, confidence level, etc. This is generally performed automatically by the electronic device based on analysis of signal data from the sensor electronics. However, any of the above variables or parameters may be input into the calculation of the decision support recommendation in a weighted manner, where the weighting is performed automatically, e.g., by signal analysis from the underlying sensors such as accelerometers, weight scales, etc., or by using manually entered data, e.g., from a physician or patient.

[0176] A summary of the types of data described is provided below in Table I. Note that certain parameters and variables occur in more than one data category.

[0177] [Table 1-1] [Table 1-2]

[0178] FIG. 7 shows a flow chart 40 illustrating a general use of the parameters and variables discussed above. In a first step, a number of inputs associated with a disease, such as diabetes, are received, the inputs corresponding to variables or parameters that may be measured, input by a user, or otherwise obtained, for example, via the cloud or other sources (step 272). Next, a decision support recommendation is calculated based on the received inputs (step 274). The decision support recommendation may be determined or calculated in a number of ways, as described below. The next step is to provide an indication of the decision support recommendation, such as an output (step 276). The output may be presented, for example, on a display or other user interface of a mobile device running a CGM application. Alternatively, or in combination, various types of advanced output (additional processing or additional details regarding the decision support recommendation processing, such as information regarding the inputs) may also be provided (step 282). In some implementations, the determined decision support recommendation may function to drive an integrated pump for a medication (step 275).

[0179] 8 illustrates a graph of various input data items as a function of time that may be used to determine a decision support recommendation. For simplicity of illustration, only a single input parameter is presented, in this case the measured glucose level, represented by trace 283 along the x-axis and y-axis, where the x-axis indicates time and the y-axis indicates glucose level. As shown, additional inputs are shown along the z-axis to illustrate that in more complex examples, trace 283 may be a multi-dimensional trace that takes into account a series of different inputs that the decision support module may use to determine a decision support recommendation. Examples of such inputs are discussed above.

[0180] As seen in FIG. 8, in region I, the user is initially in a hyperglycemic state and has a GUI with a low to medium range. In region II, it can be seen that the glucose value occupies a range of hyperglycemic values ​​(e.g., 180-400 mg / dL) 293 over a period Δt1. If the time Δt1 exceeds a predefined threshold, as the case of FIG. 8 assumes, the decision support module may present the user with a recommendation at time t1 to inject a bolus of insulin to mitigate the hyperglycemic event. In region III, after a delay of period Δt2, the glucose concentration begins to decrease. Region IV indicates an area where the glucose value is noisy for some reason, and therefore a low confidence level may be associated with that section of the signal. Thus, the decision support module provides the user with a recommendation at time t2 suggesting potential actions that may be taken to reduce the noise, such as checking the placement of the sensor, the quality of the communication link between the sensor and the mobile device, etc. Continuing further in time, area VII indicates that the user is heading toward a mild hyperglycemic state, and therefore the decision support module makes a recommendation at time t3 based on the glucose level trends and knowledge of the user's recent activity levels and exercise patterns suggesting that the user should exercise.

[0181] A safety module based on the reliability of inputs to a decision support module As mentioned previously, it is very important that whatever decision support recommendations are made, especially if they involve therapeutic interventions, are as safe as possible while minimizing the risk of harm to the user. Even if the recommendations strictly adhere to clinically established guidelines, there is still a risk that a particular recommendation may in fact be unsafe. One potential reason for this is that one or more of the inputs to the decision support module are inaccurate or otherwise unreliable, thereby misinterpreting the user's actual or predicted state or situation, leading to an inappropriate recommendation. The reasons why a particular input may be unreliable depend on the nature and source of the input, as well as various other factors. Examples of the use of these factors for some of the types of inputs discussed previously are provided below.

[0182] In accordance with the present principles, a reliability level or rating may be assigned to each input used by the decision support module. The assignment may be pre-set at the factory or may be made by an algorithm employed by the decision support module itself and / or by a cloud-based processor in communication with the decision support module. As used herein, the term reliability may encompass any uncertainty of accuracy or precision that may be known to be associated with measurements from individual sensors and the like, in some cases. Additionally or alternatively, the term reliability may encompass any uncertainty of the relationship of any of the input data items used in determining the decision support recommendation to the user's diabetic condition.

[0183] The confidence level may be a numerical indicator, e.g., 1 represents a very unreliable input value and 5 represents a highly reliable input value. Alternatively, the confidence level may be qualitatively classified, e.g., by confidence buckets representing low, medium, or high levels of confidence. Of course, other quantitative or qualitative confidence indicators may be envisioned, as would be understood by one of ordinary skill in the art. The confidence level assigned may reflect an absolute or relative level of confidence. In some cases, the confidence level may be set at the factory. In other cases, various analytical frameworks, e.g., machine learning, may be used to establish the confidence level of a given input or to modify a previously established confidence level that may have initially been set at the factory.

[0184] One factor that can be used to determine the trust level is the source of the input data item. For example, all else being equal, input data received directly from a sensor, e.g., an analyte sensor, a temperature sensor, etc., can be assigned a higher trust level than input data manually entered by a user. Similarly, input data items received from other devices (external to the CGM system) or other applications can be treated as more trustworthy than manually entered input data items. That is, manually entered data can be treated as less trustworthy than data obtained through machine-to-machine communication.

[0185] For example, an amount of physical activity expended during an exercise session obtained from a motion sensor, accelerometer, etc. may be treated as more reliable than an amount of physical activity obtained by a user manually inputting a subjective perception of the amount of exertion used during the exercise session. As another example, nutritional information provided by a nutritional database via an application may be treated as more reliable than nutritional information obtained solely from a user's subjective impressions and estimates of food intake.

[0186] Although sensor data is often considered more reliable than other input data sources, the data may still be subject to some degree of uncertainty due to factors such as signal quality, calibration, connectivity, and sensor age. For example, for a glucose sensor, the assigned confidence level may be reduced if the sensor signal is noisy, if the sensor has not been calibrated recently, or if the sensor is still in a break-in period or has exceeded its expected useful life. The reliability of input data from other types of sensors may also be reduced for similar reasons.

[0187] Another factor that may affect the reliability of input data is whether the data is quantitative or qualitative. For example, when a user manually inputs nutritional information for a meal, all other things being equal, an indication of a food serving size as "small," "medium," or "large" may be treated as less reliable than, for example, a serving size provided by weight in ounces or grams. That is, in some cases, quantitative data may be treated as more reliable than qualitative data.

[0188] Once confidence levels are assigned to the individual input data items used to generate a particular decision support recommendation, an overall confidence metric can be assigned based on the individual confidence levels. For example, the overall confidence can be an average, weighted or unweighted, of the individual confidence levels. In some cases, if an individual confidence level is deemed too low, it can be excluded from use in generating the recommendation; that is, a weight of zero can be assigned. In some embodiments, the decision support recommendation is presented to the user only if the overall confidence metric exceeds a certain threshold. In some cases, even if the overall confidence metric exceeds a threshold, the decision support recommendation is not presented to the user if any of the particular individual input data items have a confidence level below a certain specified value. That is, presentation of the recommendation can be suppressed based on an analysis of the overall confidence metric and the individual confidence levels. Of course, many variations in the calculation of the overall confidence metric will be understood.

[0189] The threshold value used to determine whether a recommendation is based on sufficiently reliable input data does not need to be a fixed value. Rather, in some cases, it may vary, for example, depending on the nature of the decision support recommendation. For example, recommendations that are classified as more conservative and safe, with less potential risk, may be presented at a lower threshold than recommendations that are classified as potentially high risk. The classification of decision support recommendations that are somewhat conservative may be pre-determined using any suitable guidelines. Furthermore, certain types of recommendations may be treated as being inherently more conservative than other types of recommendations. For example, in some cases, therapeutic recommendations may be classified as less conservative than supportive recommendations.

[0190] As discussed above, in some embodiments, decision-support recommendations are presented to the user only if the overall reliability metric exceeds a certain threshold. However, in other embodiments, decision-support recommendations are not even calculated unless the overall reliability metric exceeds a certain threshold. In some cases, even if the overall reliability metric exceeds a threshold, decision-support recommendations are not calculated if any of the particular individual input data items have a reliability level below a particular specified value. That is, calculation of decision-support recommendations may be suppressed based on analysis of the overall reliability metric and the individual reliability levels.

[0191] Any of a variety of mathematical frameworks can be used to determine both the confidence level of the decision support recommendations and the input data used to generate these recommendations, the overall confidence index, and the thresholds used to determine if and when the decision support recommendations are calculated or presented to the user. For example, as previously mentioned, in some cases, machine learning techniques may be used. Such techniques can be used on population data collected from field use to periodically and iteratively improve the performance of the safety module. This can be particularly advantageous in identifying certain sources of confidence error and uncertainty that occur infrequently and are therefore difficult to identify in clinical trials.

[0192] Other mathematical frameworks that may be used include adaptive learning, which learns aspects of a user over time, Bayesian analysis, which quantifies the reliability of current input data based on previous input data, and decision fusion methods, which determine the likelihood of an event occurring. In addition, heuristic techniques may be applied to the analytical framework to develop recommendations based on past experience. The development of such heuristic solutions may employ related techniques such as MPC, if / then logic, expert systems, logistic regression, neural networks, fuzzy logic, weight functions, and the use of regression models.

[0193] In some embodiments, a continuous glucose application resident on the display device (e.g., continuous glucose application 209 of FIG. 4) may periodically prompt the user to ensure that the application and the display device (its hardware, software, and / or firmware) are operating correctly. This is important because the display device may be supplied by the user and generally does not meet medical device standards and specifications. For example, the components of the display device, such as the power supply, display, housing, and software, are generally not medical grade. Prompts provided to the user by the application may be used as feedback to ensure that the display device is operationally complete. For example, in some cases, certain aspects of the display device's operation may be tested by presenting the user with various screens, such as a sample default trend screen, a sample warning screen, a sample decision support recommendation screen, a sample calibration screen, and / or a sample patient record screen. The user may then be prompted to answer questions regarding the quality and accuracy of the screens and the data therein before the CGM application is enabled on the display device.

[0194] Disclosed are systems and methods for determining decision support recommendations that can be presented to a user monitoring a diabetic condition to reduce the risk of complications resulting from the diabetic condition and improve metabolic control. Also disclosed are systems and methods for determining the reliability of individual input data items used to determine a given decision support recommendation and the overall reliability of the input data items, and determining whether the given decision support recommendation should be presented to a user based on the reliability level. Various methods are disclosed for determining the decision support recommendation and the reliability of the input data items.

[0195] The connections between elements shown in the figures represent exemplary communication paths. Additional communication paths may be included, either directly or through intermediaries, to further facilitate information exchange between elements. The communication paths may be bidirectional communication paths that allow elements to exchange information.

[0196] As used herein, the term "determining" encompasses a variety of actions. For example, "determining" can include calculating, computing, processing, deriving, investigating, looking up (e.g., looking up in a table, database, or another data structure), ascertaining, and the like. Also, "determining" can include receiving (e.g., receiving information), accessing (e.g., accessing data in a memory), and the like. Also, "determining" can include resolving, selecting, choosing, establishing, and the like.

[0197] As used herein, the term "message" encompasses a variety of formats for transmitting information. A message may include a machine-readable collection of information, such as an XML document, a fixed-field message, a comma-separated message, etc. In some implementations, a message may include a signal utilized to transmit one or more representations of information. Although described in the singular, it will be understood that a message may be composed / sent / stored / received, etc., in multiple parts.

[0198] The various operations of the methods described above may be performed by any suitable means capable of performing the operations, such as various hardware and / or software component(s), circuits, and / or module(s). Generally, any operation illustrated in the figures may be performed by a corresponding functional means capable of performing the operation.

[0199] The various illustrative logic blocks, modules, and circuits described in connection with this disclosure may be implemented or performed with a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array signal (FPGA) or other programmable logic device (PLD), discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but alternatively, the processor may be any commercially available processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0200] In one or more aspects, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functions may be stored on or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media, including any medium that facilitates transfer of a computer program from one place to another. A storage medium may be any available medium that can be accessed by a computer. By way of example, and not limitation, such computer-readable media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to convey or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. As used herein, disk and disc include compact discs (CDs), laser discs, optical discs, digital versatile discs (DVDs), floppy disks, and Blu-ray™ discs, where disks typically reproduce data magnetically, while discs reproduce data optically with lasers. Thus, in some aspects, computer readable media may include non-transitory computer readable media (e.g., tangible media). Additionally, in some aspects, computer readable media may include transitory computer readable media (e.g., signals). Combinations of the above should also be included within the scope of computer readable media.

[0201] The methods disclosed herein include one or more steps or actions for achieving the described method. The steps and / or actions of the methods may be interchanged with one another without departing from the scope of the claims. In other words, unless a specific order of steps or actions is specified, the order and / or use of specific steps and / or actions may be modified without departing from the scope of the claims.

[0202] Certain aspects may include a computer program product for performing the operations presented herein. For example, such a computer program product may include a computer-readable medium having instructions stored (and / or encoded) thereon, the instructions being executable by one or more processors to perform the operations described herein. In certain aspects, the computer program product may include packaging materials.

[0203] Software or instructions may be transmitted over a transmission medium. For example, if the software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of transmission media.

[0204] Furthermore, it should be understood that the modules and / or other suitable means for performing the methods and techniques described herein may be downloaded and / or otherwise obtained by a user terminal and / or base station as appropriate. For example, such devices may be coupled to a server to facilitate the transfer of means for performing the methods described herein. Alternatively, the various methods described herein may be provided via a storage means (e.g., RAM, ROM, a physical storage medium such as a compact disk (CD) or floppy disk, etc.), and the user terminal and / or base station may obtain the various methods upon coupling or providing the storage means to the device. Furthermore, any other suitable technique for providing the methods and techniques described herein to a device may be utilized.

[0205] It should be understood that the claims are not limited to the precise configuration and components illustrated above. Various modifications, changes and variations can be made in the arrangement, operation and details of the methods and apparatus described above without departing from the scope of the claims.

[0206] Unless otherwise defined, all terms (including technical and scientific terms) should be given their ordinary and customary meaning to those skilled in the art and should not be limited to a special or customized meaning unless expressly defined as such herein. It should be noted that the use of a particular term when describing a particular feature or aspect of the present disclosure should not be interpreted as implying that the term is redefined herein to be limited to include any particular characteristic of the feature or aspect of the present disclosure to which the term relates. Particularly in the appended claims, the terms and phrases used in this application and variations thereof should be interpreted as open-ended as opposed to limiting unless otherwise specified. As examples of the above, the term "including" should be interpreted to mean "including without limitation," "including but not limited to," etc.; the term "comprising," as used herein, is synonymous with "including," "containing," or "featuring," is inclusive or open-ended, and does not exclude additional, unrecited elements or method steps; the term "having" should be interpreted as "having at least," the term "including" should be interpreted as "including, but not limited to," the term "example" is used to provide illustrative examples of the items under discussion rather than an exhaustive or exclusive list of them, and includes terms such as "known," "conventional," "standard," and similar meanings. Adjectives such as the term "preferably," "preferred," "desired," or "desirable," and words of similar import, should not be construed as limiting the described subject matter to that available in a given period or at a given time, but rather as embracing known, conventional, or standard technology that may be available or known at any time now or in the future, and the use of terms such as "preferably," "preferred," "desired," or "desirable," and words of similar import, should not be understood to imply that a particular feature is critical, essential, or even important to the structure or function of the invention, but rather are merely intended to highlight alternative or additional features that may or may not be utilized in a particular embodiment of the invention.Similarly, a group of items joined by the conjunction "and" should not be construed as requiring the presence of every single one of those items in the group, but rather, unless otherwise stated, should be construed as "and / or." Similarly, a group of items joined by the conjunction "or" should not be construed as requiring mutual exclusivity between the groups, but rather, unless otherwise stated, should be construed as "and / or."

[0207] When a range of values ​​is provided, it is understood that the upper and lower limits, and each intervening value between the upper and lower limits of that range, are included within an embodiment.

[0208] With respect to the use of virtually any plural and / or singular term in this specification, a person skilled in the art can convert from plural to singular and / or from singular to plural as appropriate to the context and / or application. Various singular / plural permutations may be expressly set forth herein for clarity. The indefinite article "a" or "an" does not exclude plurals. A single processor or other unit may accomplish the functions of several items recited in the claims. The mere fact that certain means are recited in mutually different dependent claims does not indicate that a combination of these means cannot be used to advantage. Reference signs in the claims should not be construed as limiting the scope.

[0209] It will be further understood by those skilled in the art that, when a specific number is intended in an introduced claim recitation, such intent shall be clearly stated in the claim, and that in the absence of such a statement, no such intent exists. For example, to aid in understanding, the following appended claims may include the use of the introductory phrases "at least one" and "one or more" to introduce the claim recitation. However, the use of such phrases should not be construed as implying that the introduction of a claim recitation with the indefinite article "a" or "an" limits a particular claim that includes such an introduced claim recitation to an embodiment that includes only one such recitation, even when the same claim includes the introductory phrases "one or more" or "at least one" and an indefinite article such as "a" or "an" (e.g., "a" and / or "an" should typically be construed to mean "at least one" or "one or more"), and the same applies to the use of definite articles used to introduce claim recitations. In addition, even if a specific number of recitations of an introduced claim is explicitly recited, a person skilled in the art will recognize that such recitation should typically be interpreted to mean at least the recited number (e.g., the mere recitation of "two recitations" without other qualifiers typically means at least two recitations, or more than two recitations). Furthermore, when a conventional expression similar to "at least one of A, B, and C, etc." is used, such a structure is generally intended in the sense that a person skilled in the art would understand the conventional expression as including, for example, any combination of the recited items, including single elements (e.g., "a system having at least one of A, B, and C" includes, but is not limited to, a system having A alone, B alone, C alone, A and B together, A and C together, B and C together, and / or A, B, and C together, etc.).When a conventional expression similar to "at least one of A, B, or C, etc." is used, such a structure is generally intended in the sense that one of ordinary skill in the art would understand the conventional expression (e.g., "a system having at least one of A, B, or C" includes, but is not limited to, a system having A alone, B alone, C alone, A and B together, A and C together, B and C together, and / or A, B, and C together, etc.). It will be further understood by one of ordinary skill in the art that virtually any disjunction and / or phrase indicating two or more alternative terms, whether in the specification, claims, or drawings, should be understood to contemplate the possibility of including one of the terms, either of the terms, or both terms. For example, the phrase "A or B" is understood to include the possibilities of "A" or "B" or "A and B."

[0210] All numbers expressing quantities of ingredients, reaction conditions, and the like used herein should be understood to be modified in all instances by the term "about." Accordingly, unless otherwise indicated, the numerical parameters set forth herein are approximations that may vary depending on the desired properties sought to be obtained. At the very least, and not as an attempt to limit the application of the doctrine of equivalents to the scope of any claims in any application claiming priority to this application, each numerical parameter should be construed in light of the number of significant digits and ordinary rounding approaches.

[0211] All references cited herein are incorporated herein by reference in their entirety. To the extent that publications and patents or patent applications incorporated by reference conflict with the disclosure contained herein, the present specification is intended to supersede and / or take precedence over any such conflicting material.

[0212] This specification includes headings for reference and to aid in locating various sections. These headings are not intended to limit the scope of the concepts described therein. Such concepts may have applicability throughout the entire specification.

[0213] Moreover, while the foregoing has been described in some detail by way of illustration and example for purposes of clarity and understanding, it will be apparent to those skilled in the art that certain changes and modifications may be practiced. Thus, the description and examples should not be construed as limiting the scope of the invention to the specific embodiments and examples described herein, but rather as embracing all modifications and alternatives that fall within the true scope and spirit of the invention. [Explanation of symbols]

[0214] 2. Drug delivery pumps 4 Reference Meter 10 Continuous Analyte Sensors 12 Sensor Electronics 22 Cloud-Based Processors 24 Network 202 Display 202, 204, 206 I / O devices 208 processors 209 CGM Applications 210 Memory 211 Decision Support Module 212 Storage 214 Transceiver 222 Server with historical user data and / or external data

Claims

1. 1. A method for determining whether a user should be presented with decision support recommendations for treatment of a diabetic condition, comprising: determining a current diabetic state of the user based at least in part on the glucose concentration level; receiving additional data that serves, along with the current diabetes status, as input to a process for determining decision support recommendations; assigning a confidence level to the additional data; and calculating said decision support recommendations using said process and said input data; and presenting the decision support recommendation to the user on a user interface only if the assigned confidence level of the additional data exceeds a threshold.

2. The method of claim 1 , wherein assigning the trust level comprises assigning a trust level based at least in part on an identity of a source from which the additional data was obtained.

3. The method of claim 1 or 2, wherein the additional data includes an indication of food and beverages consumed or to be consumed within a defined period of time.

4. The method of claim 3 , wherein the indicator comprises data manually entered by the user.

5. The method of claim 3 or 4, wherein the indicators include nutritional data about foods or beverages obtained from a third party.

6. The method of claim 5 , wherein the nutritional data is entered automatically by an application.

7. The method of claim 5 , wherein the nutritional data is automatically entered into the application by scanning nutritional information.

8. The method of any one of claims 1 to 7, wherein the additional data comprises information describing the user's past physiological response patterns.

9. A method according to any preceding claim, wherein the additional data comprises an indication of the level of user activity which has occurred or should occur within a defined period of time.

10. The method of any one of claims 1 to 9, wherein assigning a confidence level comprises assigning a higher confidence to additional information obtained through machine-to-machine communication than through manual input-to-machine communication.

11. The method of any one of claims 1 to 10, wherein the decision support recommendations comprise recommendations for consuming food or beverages.

12. The method of any one of claims 1 to 10, wherein the decision support recommendations include a recommendation to inject a calculated amount of insulin.

13. The method of any one of claims 1 to 12, wherein presenting the decision support recommendation only if the assigned reliability of the additional data exceeds a threshold further comprises determining when to present the decision support recommendation based at least in part on the reliability level.

14. 14. The method of claim 1, wherein receiving additional data comprises receiving a plurality of additional data items of different types, and wherein assigning a trust level to the additional data comprises independently assigning a trust level to each of the additional data items.

15. 15. The method of claim 14, wherein presenting the decision support recommendation comprises presenting the decision support recommendation only if a weighted average of the individual confidence levels exceeds a first threshold and if none of the individual confidence levels is below a second threshold.

16. The method of any one of claims 1 to 15, wherein the threshold is selected based at least in part on the type of decision support recommendation being calculated.

17. 17. The method of claim 16, further comprising assigning a higher threshold to decision support recommendations that are therapeutic than to decision support recommendations that are supportive.

18. The method of any one of claims 1 to 16, wherein the threshold is selected based at least in part on the type of decision support recommendation being determined.

19. 19. The method of any one of claims 1 to 18, wherein presenting the decision support recommendation comprises presenting a more conservative support recommendation that promotes patient safety when the assigned confidence level is within a range specified by upper and lower thresholds.

20. The method of any one of claims 12 to 19, wherein the calculated amount of insulin is calculated as part of the process for determining the decision support recommendation.

21. A method according to any one of claims 2, 4 to 20, wherein the additional data comprises an indication of food and drink ingested or to be ingested within a defined period of time.

22. 22. The method according to any one of claims 3 to 21, wherein assigning a confidence level comprises: assigning a lower confidence level to the additional data obtained from the index of foods and beverages when the index comprises data manually entered by the user; and assigning a higher confidence level to the additional data obtained from the index of foods and beverages when the index comprises nutritional data regarding the foods and beverages obtained from a third party or nutritional data automatically entered by an application.

23. A method according to any preceding claim, wherein assigning a confidence level to the additional data comprises using machine learning techniques to assign a confidence level to the additional data.

24. The method of any one of claims 1 to 23, wherein the glucose concentration level is obtained from a glucose sensor in a continuous glucose monitor.

25. The method of any preceding claim, wherein determining the current diabetic condition of the user comprises receiving sensor data from a glucose sensor associated with a continuous glucose monitor.

26. 26. The method of claim 25, wherein the user interface is incorporated into a display device that receives the sensor data from the glucose sensor.

27. 27. The method of claim 26, wherein the display device receives the sensor data from the glucose sensor via a wireless communication link.

28. 30. The method of claim 27, wherein presenting the decision support recommendations to the user on a user interface comprises presenting the decision support recommendations to the user on the user interface of the display device.

29. 30. The method of claim 28, wherein a continuous glucose sensor application resident on the display device receives the sensor data and causes the decision support recommendations to be presented to the user on the user interface.

30. 1. A method for determining whether to calculate a decision support recommendation, comprising: determining a current diabetic state of the user based at least in part on the glucose concentration level; receiving additional data that serves, along with the current diabetes status, as input to a process for determining decision support recommendations; assigning a confidence level to the additional data; and and calculating the decision support recommendation using the process and the input data only if the assigned confidence level of the additional data exceeds a threshold.

31. 31. The method of claim 30, wherein assigning the trust level comprises assigning a trust level based at least in part on an identity of a source from which the additional data was obtained.

32. 32. The method of claim 30 or 31, wherein the additional data includes an indication of food and drink ingested or to be ingested within a defined period of time.

33. The method of claim 32 , wherein the indicator comprises data manually entered by the user.

34. 34. The method of claim 32 or 33, wherein the indicators include nutritional data about foods or beverages obtained from a third party.

35. 35. The method of claim 34, wherein the nutritional data is entered automatically by an application.

36. 35. The method of claim 34, wherein the nutritional data is automatically entered into the application by scanning nutritional information.

37. A method according to any one of claims 30 to 36, wherein the additional data comprises information describing the user's past physiological response patterns.

38. A method according to any one of claims 30 to 37, wherein the additional data comprises an indication of the level of user activity which has occurred or should occur within a defined period of time.

39. A method according to any one of claims 30 to 38, wherein assigning a confidence level comprises assigning a higher confidence to additional information obtained through machine-to-machine communication than through manual input-to-machine communication.

40. The method of any one of claims 30 to 39, wherein the decision support recommendations comprise recommendations for consuming food or beverages.

41. The method of any one of claims 30 to 40, wherein the decision support recommendations include a recommendation to inject a calculated amount of insulin.

42. 41. A method according to any one of claims 30 to 40, wherein receiving additional data comprises receiving a plurality of additional data items of different types, and wherein assigning a trust level to the additional data comprises independently assigning a trust level to each of the additional data items.

43. 43. The method of claim 42, wherein calculating the decision support recommendation comprises calculating the decision support recommendation only if a weighted average of the individual confidence levels exceeds a first threshold and if none of the individual confidence levels is below a second threshold.

44. The method of any one of claims 30 to 43, wherein the threshold is selected based at least in part on the type of decision support recommendation being determined.

45. The method of any one of claims 41 to 44, wherein the calculated amount of insulin is calculated as part of the process for determining the decision support recommendation.

46. 1. A method for determining whether a user should be presented with decision support recommendations for treatment of a diabetic condition, comprising: receiving a plurality of input data items affecting a diabetic condition of a user of a continuous glucose monitor, said input data items serving as inputs to a process for determining decision support recommendations; assigning a confidence level to each of said input data items; calculating a confidence metric based on the confidence levels assigned to each of the input data items; determining a decision support recommendation based on the process and the input data, and presenting the decision support recommendation to the user on a user interface only if the confidence metric exceeds a threshold.

47. 47. The method of claim 46, wherein calculating the confidence metric comprises calculating the confidence metric based on an average of the confidence levels of each of the input data items.

48. 48. The method of claim 46 or 47, wherein the average of confidence levels is a weighted average.

49. A method according to any one of claims 46 to 48, wherein the plurality of input data items comprises sensor data from a glucose sensor in the continuous glucose monitor.

50. 50. The method of claim 49, wherein assigning a confidence level includes assigning a confidence level to the sensor data based on another factor selected from the group including signal quality, sensor calibration, connectivity, age of the glucose sensor, and expected useful life of the glucose sensor.

51. A method according to any one of claims 46 to 50, wherein determining and presenting the decision support recommendation comprises determining the decision support recommendation even if the reliability metric does not exceed the threshold, while presenting the decision support recommendation to the user on the user interface only if the reliability metric exceeds the threshold.

52. A method according to any one of claims 46 to 50, wherein determining and presenting the decision support recommendation comprises determining the decision support recommendation only if the reliability metric exceeds the threshold, and presenting the decision support recommendation to the user only if the reliability metric exceeds the threshold.

53. 53. A method according to any one of claims 46 to 52, wherein assigning a confidence level to each of the input data items comprises assigning a confidence level to at least one of the input data items based on uncertainty in the accuracy of a value of the at least one input data item.

54. 54. A method according to any one of claims 46 to 53, wherein assigning a confidence level to each of the input data items comprises assigning a confidence level to at least one of the input data items based on a degree of uncertainty of the impact of the at least one input data item on the diabetic condition of the user.

55. A method according to any one of claims 46 to 54, wherein assigning the trust level comprises assigning a trust level based at least in part on the identity of a source from which the additional data was obtained.

56. A method according to any one of claims 46 to 55, wherein the plurality of input data items comprises an indication of food and drink ingested or to be ingested within a defined period of time.

57. 57. The method of claim 56, wherein the indicator comprises data manually entered by the user.

58. 57. The method of claim 56, wherein the indicators include nutritional data regarding foods or beverages obtained from a third party.

59. 59. The method of claim 58, wherein the nutritional data is entered automatically by an application.

60. 59. The method of claim 58, wherein the nutritional data is automatically entered into the application by scanning nutritional information.

61. A method according to any one of claims 46 to 60, wherein the input data items include information describing the user's past physiological response patterns.

62. A method according to any one of claims 46 to 61, wherein the input data items include an indication of the level of user activity which has occurred or should occur within a defined period of time.

63. A method according to any one of claims 46 to 62, wherein assigning a confidence level comprises assigning a higher confidence to additional information obtained through machine-to-machine communication than through manual input-to-machine communication.

64. 64. The method of any one of claims 46 to 63, wherein the decision support recommendations comprise recommendations for consuming food or beverages.

65. 65. The method of any one of claims 46 to 64, wherein the decision support recommendations include a recommendation to inject a calculated amount of insulin.

66. 66. The method of any one of claims 46 to 65, wherein presenting the decision support recommendation only if the assigned reliability of the additional data exceeds a threshold further comprises determining when to present the decision support recommendation based at least in part on the reliability level.

67. 66. A method according to any one of claims 46 to 65, wherein presenting the decision support recommendation comprises presenting the decision support recommendation only if a weighted average of the individual confidence levels exceeds a first threshold and if none of the individual confidence levels is below a second threshold.

68. The method of any one of claims 46 to 67, wherein the threshold is selected based at least in part on the type of decision support recommendation being calculated.

69. 70. The method of claim 68, further comprising assigning a higher threshold to decision support recommendations that are therapeutic than to decision support recommendations that are supportive.

70. The method of any one of claims 46 to 69, wherein the threshold is selected based at least in part on the type of decision support recommendation being determined.

71. 71. The method of any one of claims 46 to 70, wherein presenting the decision support recommendation comprises presenting a more conservative support recommendation that promotes patient safety when the assigned confidence level is within a range specified by upper and lower thresholds.

72. A method according to any one of claims 61 to 71, wherein the plurality of input data items comprises an indication of food and drink ingested or to be ingested within a defined period of time.

73. 73. The method of any one of claims 46 to 72, wherein assigning confidence levels comprises: assigning a lower confidence level to the input data items obtained from the index of foods and beverages when the index comprises data manually entered by the user; and assigning a higher confidence level to the input data items obtained from the index of foods and beverages when the index comprises nutritional data regarding the foods and beverages obtained from a third party or nutritional data entered automatically by an application.

74. A method according to any one of claims 46 to 73, wherein assigning confidence levels to the input data items comprises using machine learning techniques to assign confidence levels to the input data items.

75. The method of any one of claims 46 to 74, wherein the glucose concentration level is obtained from a glucose sensor in a continuous glucose monitor.

76. 76. The method of claim 75, wherein the user interface is incorporated into a display device that receives sensor data from the glucose sensor.

77. 77. The method of claim 76, wherein the display device receives the sensor data from the glucose sensor via a wireless communication link.

78. 78. The method of claim 77, wherein presenting the decision support recommendations to the user on a user interface comprises presenting the decision support recommendations to the user on the user interface of the display device.

79. 80. The method of claim 78, wherein a continuous glucose sensor application resident on the display device receives the sensor data and causes the decision support recommendations to be presented to the user on the user interface.

80. 1. A system for providing decision support recommendations regarding a diabetic condition of a user, comprising: a continuous glucose monitor (CGM) application running on the mobile device, the CGM application configured to at least periodically or irregularly receive data from a glucose sensor and display glucose calibration data; and a decision support recommendation application that runs as a subroutine within the CGM application or runs on the mobile device as a parallel process within the CGM application and receives data from the CGM application, the decision support recommendation application being configured to perform the method of claim 30.

Images