Analyte monitoring system and method

[0001] This application claims priority under 35 USC §119 to Provisional Application No. 60/804,170 filed Jun. 7, 2006 entitled “Analyte Monitoring”, and to Provisional Application No. 60/804,169 filed Jun. 7, 2006 entitled “Analyte Monitoring System” the disclosure of each of which are incorporated in their entirety by reference for all purposes

[0002] The association of chronic hyperglycemia and the devastating long-term complications of diabetes was clearly established by the Diabetes Control and Complication Trial (DCCT) (The Diabetes Control and Complications Trial Research Group. “The effect of intensive treatment of diabetes on the development and progression of long-term complications of insulin-dependent diabetes mellitus” N Engl J Med 329: 978-986, 1993; Santiago J V “Lessons from the Diabetes Control and Complications Trial” Diabetes 1993, 42: 1549-1554).

[0003] The DCCT found that in patients receiving intensive insulin therapy, there was a reduced risk of 76% for diabetic retinopathy, 50% for diabetic nephropathy and 60% for diabetic neuropathy. The long-term benefits of tight glycemic control have been further substantiated by the Epidemiology of Diabetes Interventions and Complications study which found over a 50% reduced risk of macrovascular disease as a result of intensive insulin therapy (The Diabetes Control and Complications Trial/Epidemiology of Diabetes Intervention and Complication (DCCT/EDIC) Study Group, “Intensive diabetes treatment and cardiovascular disease in patients with type 1 diabetes”, 353, 2643-2653, 2005).

[0004] However, the DCCT found that patients receiving intensive insulin therapy were at a threefold increased risk of severe hypoglycemia. Patients adhering to intensive insulin therapy regimens were found to have lowered thresholds for activation of neurogenic warning systems and consequently were at increased risk for more severe hypoglycemic events. (Amiel S A, Tamborlane W V, Simonson D C, Sherwin R S., “Defective glucose counterregulation after strict glycemic control of insulin-dependent diabetes mellitus.” N Engl J. Med. 1987 28; 316(22):1376-83).

[0005] The increased risk of hypoglycemia and the fear associated with patients' perception of that risk has been cited as the leading obstacle for patients to achieve the targeted glycemic levels (Cryer P E. “Hypoglycaemia: The limiting factor in the glycemic management of type I and type II diabetes” Diabetologia, 2002, 45: 937-948). In addition to the problem of chronic hyperglycemia contributing to long-term complications and the problem of acute iatrogenic hypoglycemia contributing to short-term complications, recent research suggests that transient episodes of hyperglycemia can lead to a wide range of serious medical problems besides previously identified microvascular complications as well as macrovascular complications such as increased risk for heart disease. (Haffner S “The importance of postprandial hyperglycemia in development of cardiovascular disease in people with diabetes” International Journal of Clinical Practice, 2001, Supplement 123: 24-26; Hanefeld M: “Postprandial hyperglycemia: noxious effects on the vessel wall” International Journal of Clinical Practice, 2002, Supplement 129: 45-50).

[0006] Additional research has found that glycemic variation and the associated oxidative stress may be implicated in the pathogenesis of diabetic complications (Hirsh I B, Brownlee M “Should minimal blood glucose variability become the gold standard of glycemic control?” J of Diabetes and Its Complications, 2005, 19: 178-181; Monnier, L., Mas, E., Ginet, C., Michel, F., Villon L, Cristol J-P, and Collette C, “Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia in patients with type 2 diabetes”. JAMA 2006, 295, 1681-1687). Glycemic variation has also been identified as a possible explanation for the increased prevalence of depression in both type 1 and type 2 diabetes (Van der Does F E. De Neeling J N, Snoek F J, Kostense P J, Grootenhuis P A, Bouter L M, and R J Heine: Symptoms and well-being in relation to glycemic control in type II diabetes Diabetes Care, 1996, 19: 204-210; De Sonnaville J J. Snoek F J. Colly L P. Deville W. Wijkel D. Heine R J: “Well-being and symptoms in relation to insulin therapy in type 2 diabetes” Diabetes Care, 1998, 21:919-24; Cox D J, Gonder-Frederick L A, McCall A, et al. “The effects of glucose fluctuation on cognitive function and QOL: the functional costs of hypoglycaemia and hyperglycaemia among adults with type 1 or type 2 diabetes” International Journal of Clinical Practice, 2002, Supplement 129: 20-26).

[0007] The potential benefits of continuous glucose monitoring have been recognized by numerous researchers in the field (Skyler J S “The economic burden of diabetes and the benefits of improved glycemic control: the potential role of a continuous glucose monitoring system” Diabetes Technol Ther 2 (Suppl 1): S7-S12, 2000; Tansey M J, Beck R W, Buckingham B A, Mauras N, Fiallo-Scharer R, Xing D, Kollman C, Tamborlane W V, Ruedy K J, “Accuracy of the modified Continuous Glucose Monitoring System (CGMS) sensor in an outpatient setting: results from a diabetes research in children network (DirecNet) study.” Diab. Tech. Ther. 7(1):109-14, 2005; Klonoff, D C: “Continuous glucose monitoring: Roadmap for 21st century diabetes therapy” Diabetes Care, 2005, 28: 1231:1239). Accurate and reliable real-time continuous glucose monitoring devices have the ability to alert patients of high or low blood sugars that might otherwise be undetected by episodic capillary blood glucose measurements.

[0008] Continuous glucose monitors have the potential to permit more successful adherence to intensive insulin therapy regimens and also to enable patients to reduce the frequency and extent of glycemic fluctuations. However, the development of this technology has proceeded more slowly than anticipated. For example, two recent comprehensive reviews of decades of research in the field cited the lack of accuracy and reliability as the major factor limiting the acceptance of this new technology as well as the development of an artificial pancreas (Chia, C. W. and Saudek, C. D., “Glucose sensors: toward closed loop insulin delivery” Endocrinol. Metab. Clin. N. Am., 33, 174-195, 2004; Hovorka, R. “Continuous glucose monitoring and closed-loop systems” Diabet. Med. 23, 1-12, 2006).

[0009] As continuous analyte monitoring becomes more prevalent, of use are continuous analyte sensors and systems that are accurate to such a high degree that confirmatory analyte measurement are not needed to verify the continuous sensing measurements, e.g., prior to a user relying on the continuous measurements. Also of interest are such sensors that work in concert with a drug delivery device.

[0010] Generally, the present disclosure relates to methods and devices for monitoring of the level of an analyte using a continuous and/or automatic in vivo monitoring analyte sensor. Embodiments include sensors in which at least a portion of the sensor is adapted to be positioned beneath the skin of a user and which are adapted for providing clinically accurate analyte data, i.e., data with accuracy sufficient so that a user may confidently rely on the sensor results, e.g., to manage a disease condition and/or make a healthcare decision based thereon. Accordingly, sensors capable of providing clinically accurate (i.e., clinically relevant) analyte information to a user are provided.

[0011] Embodiments include continuous analyte monitoring systems that do not require additional analyte information obtained by a second system and/or sensor to confirm the results reported by the continuous sensing system.

[0012] Embodiments also include high accuracy continuous analyte sensors and systems with drug delivery systems e.g., insulin pumps, or the like. A communication link (e.g., by cable or wirelessly such as by infrared (IR) or RF link or the like) may be provided for transfer of data from the sensor to the drug delivery device. The drug delivery device may include a processor to determine the amount of drug to be delivered using sensor data, and may deliver such drug automatically or after user direction to do so.

[0013] Also provided are methods of analyte monitoring using highly accurate continuous analyte sensors.

[0014] These and other objects, features and advantages of the present disclosure will become more fully apparent from the following detailed description of the embodiments, the appended claims and the accompanying drawings.

[0024] Before the various embodiments of the present disclosure is described, it is to be understood that this disclosure is not limited to particular embodiments described, as such may, of course, vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting, since the scope of the present disclosure will be limited only by the appended claims.

[0025] Where a range of values is provided, it is understood that each intervening value, to the tenth of the unit of the lower limit unless the context clearly dictates otherwise, between the upper and lower limit of that range and any other stated or intervening value in that stated range, is encompassed within the present disclosure. The upper and lower limits of these smaller ranges may independently be included in the smaller ranges is also encompassed within the present disclosure, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the present disclosure.

[0026] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this present disclosure belongs. Although any methods and materials similar or equivalent to those described herein can also be used in the practice or testing of various embodiments of the present disclosure, exemplary methods and materials are now described. All publications mentioned herein are incorporated herein by reference to disclose and describe the methods and/or materials in connection with which the publications are cited.

[0027] It must be noted that as used herein and in the appended claims, the singular forms “a”, “an”, and “the” include plural referents unless the context clearly dictates otherwise.

[0028] The publications discussed herein are provided solely for their disclosure prior to the filing date of the present application. Nothing herein is to be construed as an admission that the present disclosure is not entitled to antedate such publication by virtue of prior invention. Further, the dates of publication provided may be different from the actual publication dates which may need to be independently confirmed.

[0029] As will be apparent to those of skill in the art upon reading this disclosure, each of the individual embodiments described and illustrated herein has discrete components and features which may be readily separated from or combined with the features of any of the other several embodiments without departing from the scope or spirit of the present disclosure.

[0030] The present disclosure is applicable to analyte monitoring systems using a sensor—at least a portion of which is positioned beneath the skin of the user, for the in vivo determination of a concentration of an analyte, such as glucose, lactate, and the like, in a body fluid. The sensor may be, for example, subcutaneously positioned in a patient for the continuous or periodic monitoring an analyte in a patient's interstitial fluid. This may be used to infer the glucose level in the patient's bloodstream. The sensors of the subject disclosure also include in vivo analyte sensors for insertion into a vein, artery, or other portion of the body containing fluid. A sensor of the subject disclosure may be configured for monitoring the level of the analyte over a time period which may range from hours, days, weeks, or longer, as described in greater detail below.

[0031] More specifically, FIG. 1 illustrates a data monitoring and management system such as, for example, analyte (e.g., glucose) monitoring system 100, in accordance with one embodiment of the present disclosure. The subject disclosure is further described primarily with respect to a glucose monitoring system for convenience and such description is in no way intended to limit the scope of the present disclosure. It is to be understood that the analyte monitoring system may be configured to monitor a variety of analytes. Analytes that may be monitored include, for example, acetyl choline, amylase, bilirubin, cholesterol, chorionic gonadotropin, creatine kinase (e.g., CK-MB), creatine, DNA, fructosamine, glucose, glutamine, growth hormones, hormones, ketones, lactate, peroxide, prostate-specific antigen, prothrombin, RNA, thyroid stimulating hormone, and troponin, and the like. The concentration of drugs, such as, for example, antibiotics (e.g., gentamicin, vancomycin, and the like), digitoxin, digoxin, drugs of abuse, theophylline, and warfarin, and the like, may also be monitored.

[0032] The analyte monitoring system 100 includes a highly accurate sensor 101, a transmitter unit 102 coupled to the sensor 101, and a receiver unit 104 which is configured to communicate with the transmitter unit 102 via a communication link 103. The receiver unit 104 may be further configured to transmit data to a data processing terminal 105 for evaluating the data received by the receiver unit 104. Moreover, the data processing terminal in one embodiment may be configured to receive data directly from the transmitter unit 102 via a communication link 106 which may optionally be configured for bi-directional communication. Some or all of the various components may be separate components, or some or all may be integrated into a single unit.

[0033] Only one sensor 101, transmitter unit 102, receiver unit 104, communication link 103, and data processing terminal 105 are shown in the embodiment of the analyte monitoring system 100 illustrated in FIG. 1. However, it will be appreciated by one of ordinary skill in the art that the analyte monitoring system 100 may include one or more sensor 101, transmitter unit 102, receiver unit 104, communication link 103, and data processing terminal 105. Moreover, within the scope of the present disclosure, the analyte monitoring system 100 may be a continuous monitoring system, or semi-continuous, or a discrete monitoring system. In a multi-component environment, each device is configured to be uniquely identified by each of the other devices in the system so that communication conflict is readily resolved between the various components within the analyte monitoring system 100.

[0034] In one embodiment of the present disclosure, the sensor 101 is physically positioned in or on the body of a user whose analyte level is being monitored. The sensor 101 may be configured to continuously sample the analyte level of the user and convert the sampled analyte level into a corresponding data signal for transmission by the transmitter unit 102. In one embodiment, the transmitter unit 102 is coupled to, e.g., mounted on, the sensor 101 so that both devices are positioned on the user's body. The transmitter unit 102 performs data processing such as filtering and encoding on data signals, each of which corresponds to a sampled analyte level of the user, for transmission to the receiver unit 104 via the communication link 103.

[0035] In one embodiment, the analyte monitoring system 100 is configured as a one-way RF communication path from the transmitter unit 102 to the receiver unit 104. In such embodiment, the transmitter unit 102 transmits the sampled data signals received from the sensor 101 without acknowledgement from the receiver unit 104 that the transmitted sampled data signals have been received. For example, the transmitter unit 102 may be configured to transmit the encoded sampled data signals at a fixed rate (e.g., at one minute intervals) after the completion of the initial power on procedure. Likewise, the receiver unit 104 may be configured to detect such transmitted encoded sampled data signals at predetermined time intervals. Alternatively, the analyte monitoring system 100 may be configured with a bi-directional RF (or otherwise) communication between the transmitter unit 102 and the receiver unit 104.

[0036] Additionally, in one aspect, the receiver unit 104 may include two sections. The first section is an analog interface section that is configured to communicate with the transmitter unit 102 via the communication link 103. In one embodiment, the analog interface section may include an RF receiver and an antenna for receiving and amplifying the data signals from the transmitter unit 102, which are thereafter, demodulated with a local oscillator and filtered through a band-pass filter. The second section of the receiver unit 104 is a data processing section which is configured to process the data signals received from the transmitter unit 102 such as by performing data decoding, error detection and correction, data clock generation, and data bit recovery.

[0037] In certain embodiments, in operation, the receiver unit 104 is configured to detect the presence of the transmitter unit 102 within its range based on, for example, the strength of the detected data signals received from the transmitter unit 102 or a predetermined transmitter identification information. Upon successful synchronization with the corresponding transmitter unit 102, the receiver unit 104 is configured to begin receiving from the transmitter unit 102 data signals corresponding to the user's detected analyte level. More specifically, the receiver unit 104 in one embodiment is configured to perform synchronized time hopping with the corresponding synchronized transmitter unit 102 via the communication link 103 to obtain the user's detected analyte level.

[0038] Referring again to FIG. 1, the data processing terminal 105 may include a personal computer, a portable computer such as a laptop or a handheld device (e.g., personal digital assistants (PDAs)), and the like, each of which may be configured for data communication with the receiver via a wired or a wireless connection. Additionally, the data processing terminal 105 may further be connected to a data network (not shown) for storing, retrieving and updating data corresponding to the detected analyte level of the user.

[0039] Within the scope of the present disclosure, the data processing terminal 105 may include an infusion device such as an insulin infusion pump or the like, which may be configured to administer insulin to patients, and which may be configured to communicate with the receiver unit 104 for receiving, among others, the measured analyte level. Alternatively, the receiver unit 104 may be configured to integrate an infusion device therein so that the receiver unit 104 is configured to administer insulin therapy to patients, for example, for administering and modifying basal profiles, as well as for determining appropriate boluses for administration based on, among others, the detected analyte levels received from the transmitter unit 102.

[0040] Additionally, the transmitter unit 102, the receiver unit 104 and the data processing terminal 105 may each be configured for bi-directional wireless communication such that each of the transmitter unit 102, the receiver unit 104 and the data processing terminal 105 may be configured to communicate (that is, transmit data to and receive data from) with each other via the wireless communication link 103. More specifically, the data processing terminal 105 may in one embodiment be configured to receive data directly from the transmitter unit 102 via the communication link 106, where the communication link 106, as described above, may be configured for bi-directional communication.

[0041] In this embodiment, the data processing terminal 105 which may include an insulin pump, may be configured to receive the analyte signals from the transmitter unit 102, and thus, incorporate the functions of the receiver 103 including data processing for managing the patient's insulin therapy and analyte monitoring. In one embodiment, the communication link 103 may include one or more of an RF communication protocol, an infrared communication protocol, a Bluetooth enabled communication protocol, an 802.11x wireless communication protocol, or an equivalent wireless communication protocol which would allow secure, wireless communication of several units (for example, per HIPPA requirements) while avoiding potential data collision and interference.

[0042] Continuous Glucose Monitoring Sensors and Systems

[0043] As described above, the various embodiments of the present disclosure relate to continuous analyte sensors and systems having a high degree of accuracy, e.g., as demonstrated by a Clark Error Grid, Parks Error Grid, Continuous Glucose Error Grid, MARD analysis, and the like. The high degree of accuracy permits a user to rely on the results of the sensor without the need to confirm sensor results. In certain embodiments, the sensors have at least about 80% of its paired data points within zone A of one or more of the Clark Error Grid, the Consensus Error Grid, or the Continuous Glucose Error Grid Analysis, e.g., at least about 85% of its paired data points within zone A of one or more of the Clark Error Grid, the Consensus Error Grid, or the Continuous Glucose Error Grid Analysis, e.g., at least about 90% of its paired data points within zone A of one or more of the Clark Error Grid, the Consensus Error Grid, or the Continuous Glucose Error Grid Analysis, e.g., at least about 95% of its paired data points within zone A of one or more of the Clark Error Grid, the Consensus Error Grid, or the Continuous Glucose Error Grid Analysis.

[0044] In certain embodiments, a sensor may have about 80% or greater, e.g., 85% or greater, e.g., 90% or greater of its paired data points within zone A of the Clark Error Grid, and 80% or greater, e.g., 85% or greater, e.g., 90% or greater, of its paired data points within zone A of the Consensus Error Grid.

[0045] The sensors are continuous analyte monitoring sensors. The sensors are adapted to continuously or periodically monitor analyte levels for a period of time, e.g., usually at least about 24 hours, e.g., about 1 day to about 30 days, e.g., about 3 days to about 7 days, e.g., a 5 day sensor or 7 day sensor.

[0046] Embodiments of the clinically accurate continuous glucose monitoring systems of the present disclosure include four components: a small, miniaturized analyte sensor element (which may be an electrochemical or optical sensor) for placement in the subcutaneous adipose tissue in the arm or abdomen (or elsewhere); a disposable sensor delivery unit containing a spring-loaded sharp for mechanical insertion of the sensor into the tissue and a sensor support mount; a transmitter (e.g., wireless transmitter) which connects to the sensor support mount on the skin surface and to the inserted electrochemical sensor; and a hand-held receiver device for communication (e.g., wireless) with the transmitter and for the communication (e.g., audio and/or visual display) of the continuous glucose values to the user. The system may also include a data management system in which information from the receiver (and/or transmitter) is forwarded (e.g., wirelessly or otherwise) to a data management system such as a personal computer (“PC”), personal digital assistant (“PDA”), telephone, facsimile machine, drug delivery device (e.g., internal or external insulin pump) or the like.

[0047] Embodiments of the sensors of the present disclosure vary, but in all embodiments have a high degree of accuracy. In other words, the sensors' accuracy enables a user of the system to solely and confidently rely on the sensors' results that are reportable to the user, e.g., to manage a disease state such as diabetes or the like, make healthcare decisions (e.g., insulin delivery, meals, exercise, etc.). In this manner, adjunctive measurements are not required to confirm the readings of the highly accurate sensors of the present disclosure, thereby eliminating burdensome and painful fingersticks required for testing analyte using conventional blood analyte monitoring systems such as blood glucose test strips and the like, used for such confirmations.

[0048] In certain embodiments a sensor is adapted to be wholly or partially positioned beneath the skin surface of a user. A sensor may be a transcutaneous sensor in which a portion of the sensor is configured to be positioned beneath a skin surface and portion is configured to be positioned above the skin surface. In many embodiments at least a portion of the sensor is configured to be inserted into the subcutaneous adipose tissue. Sensors may vary in size, where in certain embodiments a sensor may be about 5.5 mm long, about 600 microns wide and about 250 microns thick. Sensors having different lengths and/or widths and/or thicknesses are also encompassed by the present disclosure. The sensors are configured to accurately measure an analyte, e.g., glucose concentration in the interstitial fluid, which has correlates with blood glucose. The sensor is typically provided to a user as a sterile, single-use disposable element.

[0049] The sensors may be configured to continuously monitor analyte levels of a user for a period of time. In certain embodiments, the period of time ranges from about 1 day to about 30 days, e.g., from about 3 days to about 7 days, where in certain embodiments a sensor may configured for up to about five days of continuous use. A system may include two or more sensors, which may be temporally overlapped for a certain period of usage time, thereby extending the amount of time of continuous sensing and/or doing away with any time gaps that may result from removing a first sensor and inserting a second. Furthermore, a sensor may be calibrated from a previous sensor in certain embodiments.

[0050] The glucose measurement is made using sensing chemistry. Sensing chemistry may include an enzyme and may include a mediator. In certain embodiments, the sensing chemistry is a modified glucose oxidase polymeric matrix with an osmium dopant in the supporting polymer matrix. The sensing chemistry (also referred to as the “transduction chemistry”) used in the sensors of the present disclosure permits detection of signal, e.g., a nanoampere electrical current from the reaction with an applied potential, such as of only about 40 mV.

[0051] More specifically, in one embodiment, the sensor includes at least one working electrode formed on a substrate. The sensor may also include at least one counter electrode (or counter/reference electrode) and/or at least one reference electrode. The counter electrode and/or reference electrode may be formed on the substrate or may be separate units. For example, the counter electrode and/or reference electrode may be formed on a second substrate which is also implanted in the patient or, for some embodiments of the implantable sensors, the counter electrode and/or reference electrode may be placed on the skin of the patient with the working electrode or electrodes being implanted into the patient.

[0052] The working electrode or electrodes are formed using conductive traces disposed on the substrate. The counter electrode and/or reference electrode, as well as other optional portions of the sensor, such as a temperature probe, may also be formed using conductive traces disposed on the substrate. These conductive traces may be formed over a smooth surface of the substrate or within channels formed by, for example, embossing, indenting or otherwise creating a depression in the substrate.

[0053] A sensing layer is often formed proximate to or on at least one of the working electrodes to facilitate the electrochemical detection of the analyte and the determination of its level in the sample fluid, particularly if the analyte can not be electrolyzed at a desired rate and/or with a desired specificity on a bare electrode. The sensing layer may include an electron transfer agent to transfer electrons directly or indirectly between the analyte and the working electrode. The sensing layer may also contain a catalyst to catalyze a reaction of the analyte. The components of the sensing layer may be in a fluid or gel that is proximate to or in contact with the working electrode. Alternatively, the components of the sensing layer may be disposed in a polymeric or sol-gel matrix that is proximate to or on the working electrode. In one aspect, the components of the sensing layer are non-leachably disposed within the sensor. Further, the components of the sensor are immobilized within the sensor.

[0054] In addition to the electrodes and the sensing layer, the sensor may also include a temperature probe, a mass transport limiting layer, a biocompatible layer, and/or other optional components, as described below. Each of these items enhances the functioning of and/or results from the sensor, as discussed below.

The Substrate

[0055] The substrate may be formed using a variety of non-conducting materials, including, for example, polymeric or plastic materials and ceramic materials. Suitable materials for a particular sensor may be determined, at least in part, based on the desired use of the sensor and properties of the materials.

[0056] In some embodiments, the substrate is flexible. In other embodiments, the sensors are made using a relatively rigid substrate to, for example, provide structural support against bending or breaking.

Conductive Traces

[0057] At least one conductive trace is formed on the substrate for use in constructing a working electrode. In addition, other conductive traces may be formed on the substrate for use as electrodes (e.g., additional working electrodes, as well as counter, counter/reference, and/or reference electrodes) and other components, such as a temperature probe. The conductive traces may be formed on the substrate by a variety of techniques, including, for example, photolithography, screen printing, or other impact or non-impact printing techniques. The conductive traces may also be formed by carbonizing conductive traces in an organic (e.g., polymeric or plastic) substrate using a laser.

[0058] The conductive traces are typically formed using a conductive material 56 such as carbon (e.g., graphite), a conductive polymer, a metal or alloy (e.g., gold or gold alloy), or a metallic compound (e.g., ruthenium dioxide or titanium dioxide). The formation of films of carbon, conductive polymer, metal, alloy, or metallic compound are well-known and include, for example, chemical vapor deposition (CVD), physical vapor deposition, sputtering, reactive sputtering, printing, coating, and painting.

[0059] In addition to the particles of carbon, metal, alloy, or metallic compound, the conductive ink may also contain a binder. The binder may optionally be cured to further bind the conductive material within the channel and/or on the substrate.

[0060] Suitable redox couples for binding to the conductive material of the reference electrode include, for example, redox polymers (e.g., polymers having multiple redox centers.). In one aspect, the reference electrode surface may be non-corroding so that an erroneous potential is not measured. Examples of conductive materials include less corrosive metals, such as gold and palladium, and may include non-corrosive materials including non-metallic conductors, such as carbon and conducting polymers. A redox polymer can be adsorbed on or covalently bound to the conductive material of the reference electrode, such as a carbon surface of a conductive trace. Non-polymeric redox couples can be similarly bound to carbon or gold surfaces.

[0061] A variety of methods may be used to immobilize a redox polymer on an electrode surface. One method is adsorptive immobilization. This method is particularly useful for redox polymers with relatively high molecular weights. The molecular weight of a polymer may be increased, for example, by cross-linking.

[0062] Another method for immobilizing the redox polymer includes the functionalization of the electrode surface and then the chemical bonding, often covalently, of the redox polymer to the functional groups on the electrode surface.

Sensing Layer

[0063] Some analytes, such as oxygen, can be directly electrooxidized or electroreduced on the working electrode. Other analytes, such as glucose and lactate, require the presence of at least one electron transfer agent and/or at least one catalyst to facilitate the electrooxidation or electroreduction of the analyte. Catalysts may also be used for those analyte, such as oxygen, that can be directly electrooxidized or electroreduced on the working electrode. For these analytes, each working electrode has a sensing layer formed proximate to or on a working surface of the working electrode. Typically, the sensing layer is formed near or on only a small portion of the working electrode, often near a tip of the sensor. This limits the amount of material needed to form the sensor and places the sensing layer 64 in the best position for contact with the analyte-containing fluid (e.g., a body fluid, sample fluid, or carrier fluid).

Electron Transfer Agent

[0064] In many embodiments, the sensing layer contains one or more electron transfer agents in contact with the conductive material of the working electrode. In some embodiments of the present disclosure, there is little or no leaching of the electron transfer agent away from the working electrode during the period in which the sensor is implanted in the patient. A diffusing or leachable (i.e., releasable) electron transfer agent often diffuses into the analyte-containing fluid, thereby reducing the effectiveness of the electrode by reducing the sensitivity of the sensor over time.

[0065] In some embodiments of the present disclosure, to prevent leaching, the electron transfer agents are bound or otherwise immobilized on the working electrode or between or within one or more membranes or films disposed over the working electrode. The electron transfer agent may be immobilized on the working electrode using, for example, a polymeric or sol-gel immobilization technique. Alternatively, the electron transfer agent may be chemically (e.g., ionically, covalently, or coordinatively) bound to the working electrode, either directly or indirectly through another molecule, such as a polymer, that is in turn bound to the working electrode.

[0066] In general, electron transfer agents may be electroreducible and electrooxidizable ions or molecules having redox potentials that are a few hundred millivolts above or below the redox potential of the standard calomel electrode (SCE). Further, the electron transfer agents are not more reducing than about −150 mV and not more oxidizing than about +400 mV versus SCE.

Catalyst

[0067] The sensing layer may also include a catalyst which is capable of catalyzing a reaction of the analyte. The catalyst may also, in some embodiments, act as an electron transfer agent. One example of a suitable catalyst is an enzyme which catalyzes a reaction of the analyte. In one aspect, the catalyst is non-leachably disposed on the sensor, whether the catalyst is part of a solid sensing layer in the sensor or solvated in a fluid within the sensing layer. In a further aspect, the catalyst is immobilized within the sensor (e.g., on the electrode and/or within or between a membrane or film) to prevent unwanted leaching of the catalyst away from the working electrode and into the patient. This may be accomplished, for example, by attaching the catalyst to a polymer, cross linking the catalyst with another electron transfer agent (which can be polymeric), and/or providing one or more barrier membranes or films with pore sizes smaller than the catalyst.

Biocompatible Layer

[0068] An optional film layer is formed over at least that portion of the sensor which is subcutaneously inserted into the patient. This optional film layer may serve one or more functions. The film layer prevents the penetration of large biomolecules into the electrodes. This is accomplished by using a film layer having a pore size that is smaller than the biomolecules that are to be excluded. Such biomolecules may foul the electrodes and/or the sensing layer thereby reducing the effectiveness of the sensor and altering the expected signal amplitude for a given analyte concentration. The fouling of the working electrodes may also decrease the effective life of the sensor. The biocompatible layer may also prevent protein adhesion to the sensor, formation of blood clots, and other undesirable interactions between the sensor and body.

Interferent-Eliminating Layer

[0069] An interferent-eliminating layer may be included in the sensor. The interferent-eliminating layer may be incorporated in the biocompatible layer or in the mass transport limiting layer (described below) or may be a separate layer. Interferents are molecules or other species that are electroreduced or electrooxidized at the electrode, either directly or via an electron transfer agent, to produce a false signal. In one embodiment, a film or membrane prevents the penetration of one or more interferents into the region around the working electrodes. In one aspect, this type of interferent-eliminating layer is much less permeable to one or more of the interferents than to the analyte.

Mass Transport Limiting Layer

[0070] A mass transport limiting layer may be included with the sensor to act as a diffusion-limiting barrier to reduce the rate of mass transport of the analyte, for example, glucose or lactate, into the region around the working electrodes. By limiting the diffusion of the analyte, the steady state concentration of the analyte in the proximity of the working electrode (which is proportional to the concentration of the analyte in the body or sample fluid) can be reduced. This extends the upper range of analyte concentrations that can still be accurately measured and may also expand the range in which the current increases approximately linearly with the level of the analyte. Particularly useful materials for the film layer are membranes that do not swell in the analyte-containing fluid that the sensor tests.

[0071] Suitable membranes include 3 to 20,000 nm diameter pores. Membranes having 5 to 500 nm diameter pores with well-defined, uniform pore sizes and high aspect ratios may be used. In one embodiment, the aspect ratio of the pores may be two or greater, or in one aspect five or greater.

[0072] Embodiments of the system include a receiver that includes both the signal processing algorithms and the user interface system for operation of the system and display of the results—although one or both may be incorporated wholly or partially into the transmitter of the system. In operation, the glucose display on the main screen of the receiver is updated during a predetermined time period, e.g., about once a minute or the like, and gives the instantaneous continuous glucose value. Also provided may be the direction and/or rate of change averaged over a predetermined period of time, e.g., the preceding fifteen minutes, or the like. The direction may be communicated using any suitable audio and/or visual indicator(s). For example, direction may be displayed with trend arrows that give quantitative ranges of the rate of change in units of about 1 mg/dL/min from about −2 mg/dL/min to about +2 mg/dL/min. The receiver may also include threshold and/or projected warnings—audible and/or visual warnings. These may be settable at the factory and/or by the user to different glucose levels to provide warnings of actual and impending hypo- or hyperglycemia. Other warnings may also be included, e.g., battery level, and the like. Time-to-calibrate indicators may also be included.

[0073] The system may also include a blood glucose (“BG”) meter for use with glucose test strips which may be used for calibration of the continuous glucose sensor, but as noted above, is not needed to confirm the continuous sensor results. The BG meter may be a separate, though connectable component, or may be integrated into the receiver as a single unitary device. For example, the receiver may include a test strip port and a processor to process a reading from the test strip. The built-in blood glucose meter eliminates the possibility of transcription errors during sensor calibration and also provides the user with a backup glucose meter system.

[0074] The continuous glucose systems of the present disclosure may be calibrated according to a predetermined calibration schedule. In certain embodiments, this schedule may be limited to factory-only calibration. However in certain embodiments, the calibration schedule may include calibrations by the user. For example, over the period of use of the system, it may be calibrated from about 0 to about 10 times, e.g., from about 1 to about 5 times, e.g., about 4 times. An exemplary calibration schedule may include calibration 4 times over a 5 day period, e.g., at 10, 12, 24 and 72 hours after sensor insertion. In certain embodiments, the system may be configured for single point calibration, e.g., as described in U.S. Pat. No. 6,121,009 and elsewhere. In other embodiments, exemplary calibration schedule may include calibration 1-2 times over a 5-7 day period. The system may be configured to accept calibration values that fall within a certain range or are at least meet a threshold value. For example, calibration values may be accepted for blood glucose input between about 60 and about 300 mg/dL and when the absolute rate of change of glucose is estimated to be less than about 2 mg/dL/min. These constraints on the acceptance of calibration input values are designed to limit the potential adverse effects of the intrinsic physiological lag between interstitial fluid glucose and blood glucose.

[0075] In the embodiments in which at least one calibration by the user is required, the system may be configured so that it does not display (i.e., does not report to the user) real-time glucose values from the continuous monitor until the first calibration, e.g., at about ten hours after sensor insertion in certain instances. This delay after insertion is designed so that the initial system calibration is performed after the sensor has reached a stable equilibrium with the surrounding tissue.

[0076] Moreover, in one embodiment, the use of fingerstick calibration in response to the Freestyle Navigator® system hypoglycemic alarm may increase the overall system accuracy.

[0077] An exemplary, analyte sensor and sensing system having the high accuracy described herein is the Freestyle Navigator® continuous glucose monitoring system from Abbott Diabetes Care, Inc., of Alameda, Calif.

[0078] Kits

[0079] Finally, kits are also provided. Embodiments of the subject kits may include one or more highly accurate sensors as described herein. Embodiments may also include a sensor insertion device and/or transmitter and/or receiver. Embodiments may also include a drug delivery device such as an insulin pump.

[0080] In certain embodiments, a kit may include a blood glucose meter to be used with the continuous sensing system, e.g., for calibration. The meter may be a separate component from continuous sensing components (in which case a communication link for transferring data from the meter to the sensing system (such as to the receiver) may be included) or may be integrated therein, e.g., the receiver may include a blood glucose meter.

[0081] The subject kits may also include written instructions for using a sensor. The instructions may be printed on a substrate, such as paper or plastic, etc. As such, the instructions may be present in the kits as a package insert, in the labeling of the container of the kit or components thereof (i.e., associated with the packaging or sub-packaging) etc. In other embodiments, the instructions are present as an electronic storage data file present on a suitable computer readable storage medium, e.g., CD-ROM, diskette, etc. In yet other embodiments, the actual instructions are not present in the kit, but means for obtaining the instructions from a remote source, e.g. via the Internet, are provided. An example of this embodiment is a kit that includes a web address where the instructions can be viewed and/or from which the instructions can be downloaded. As with the instructions, this means for obtaining the instructions is recorded on a suitable substrate.

[0082] In many embodiments of the subject kits, the components of the kit are packaged in a kit containment element to make a single, easily handled unit, where the kit containment element, e.g., box or analogous structure, may or may not be an airtight container, e.g., to further preserve the one or more sensors and additional reagents (e.g., control solutions), if present, until use.

[0083] The accuracy of a highly accurate continuous monitoring system such as the Freestyle Navigator® continuous glucose monitoring system measuring glucose in the interstitial fluid is studied, in comparison with a laboratory reference method over five days of sensor wear.

Study Design and Methods

[0084] Fifty-eight subjects with Type 1 diabetes ranging in age from 18-64 were enrolled in a multi-center, prospective, single-arm study. Each subject wore two sensors simultaneously—one on the arm and the other on the abdomen. All the FreeStyle Navigator® devices were calibrated with separate capillary fingerstick measurements at 10, 12, 24 and 72 hours after sensor insertion. Data from the continuous glucose monitor was collected at one-minute intervals for the entire study. Measurements from the FreeStyle Navigator® system were compared with reference venous sample measurements taken in an in-patient clinical research center once every fifteen minutes over a fifty hour time period covering a distribution over the entire 120 hour wear period for the Freestyle Navigator® sensor.

[0085] The subjects were admitted to a healthcare facility either in the evening or in the morning for sensor insertion. The sensors were inserted by a health care professional on both the lateral or posterior upper arm and the right or left lower abdominal quadrant using the disposable sensor delivery unit. The subjects returned to the clinic approximately nine hours later for the placement of the venous access line and for the calibration of the sensor using the built-in FreeStyle® blood glucose meter. Calibration of the FreeStyle Navigator® device in this study was deliberately scheduled to occur at different times of day as well as both pre- and post-prandially. During two separate periods in which the subjects were in the clinic and venous samples were being taken, each subject was administered intravenous insulin or a 75 gram fast-acting glucose drink, such as Glucola, in order to obtain data for evaluation of the sensor performance during deliberately-induced periods of rapidly-falling and rapidly-rising glucose. Data from the sensor and transmitter were stored in the receiver with a one minute frequency, but were not displayed to the subjects or the clinic staff. Throughout the study, all the subjects continued with their previously established diabetes management regimen. The high frequency and volume of the venous blood samples, 2.5 mL once every fifteen minutes, required a limitation of 50 hours of intensive testing in order to maintain the total volume of blood drawn from each subject within accepted safety limits. Subjects were assigned to different study schedules so as to provide an optimal distribution of the fifty hours of laboratory reference data over the total five day duration of the sensor life.

[0086] FIG. 2 illustrates five-day data from the Freestyle Navigator® continuous glucose monitor (arm and abdomen) and 50 hours of YSI venous sampling taken two separate in-patient admissions from one subject. The timing of the glucose and insulin challenges is also shown. The shaded blocks are night time. The black solid line is the Freestyle Navigator® sensor in the arm, the dashed line is the sensor in the abdomen. YSI measurements are shown in white triangles. The plus and cross symbols are the Freestyle Navigator® system blood glucose calibrations for the arm and abdominal sensors, respectively.

[0087] Referring to FIG. 2, a typical profile plot for the five-days of the study with one-minute data from the arm and abdominal sensors as well as the fifteen minute venous samples taken over three separate periods during the five days. The glucose concentration from the venous sample was measured using a YSI 2300 STAT Plus™ Glucose & Lactate Analyzer YSI analyzer (YSI Life Sciences, Yellow Springs, Ohio). All YSI measurements were made in duplicate from a single blood sample. YSI measurements were multiplied by 1.12 to obtain plasma equivalent value.

Results

[0088] A number of separate metrics were used to evaluate the accuracy and performance of the FreeStyle Navigator® system compared with the venous blood samples measured with the laboratory reference method. These metrics included the Clarke error grid, the Consensus error grid, the mean and median absolute relative difference as well as cross-correlation statistics for comparison of abdominal and upper arm sensors. The sensor performance was evaluated for the entire five days, for each day individually as well as diurnally and nocturnally. Characteristic physiological lag times were derived from analysis of the data. The data was also analyzed using the Continuous Glucose Error Grid Analysis (CG-EGA) (Kovatchev, 2004). Finally, the accuracy of the FreeStyle Navigator® system compared to the venous reference samples was analyzed as a function of the measured rates of change in the underlying blood glucose.

[0089] Comparison of the FreeStyle Navigator® continuous glucose measurements with the laboratory reference method (n=20,362) gave a mean absolute relative difference of 12.8% and a median absolute relative difference of 9.3%. The percentage in the clinically-accurate Clarke Error Grid zone A was 81.7% and 16.7% in the clinically-acceptable B zone. This included periods of high rates of change of blood glucose during intravenous glucose and insulin challenges. The precision of the matched Freestyle Navigator® sensors worn on the arm and abdomen had a coefficient of variation of 10% (n=312,953). The accuracy remained unchanged over the five days with the percent of data in the Clarke Error Grid Zone A equal to 82.5% on the first day and 80.9% on the fifth day.

Clinical Accuracy Overall

[0090] FIG. 3 shows the Clarke error grid for the study reported herein. More specifically, FIG. 3 illustrates an overall Clarke error grid showing 81.7% in the clinically-accurate A zone, 16.7% of the paired points in the clinically-acceptable or benign error zone B and only 1.7% outside of the A and B zones

[0091] The Clarke error grid was developed to assess the clinical implications of new glucose monitoring technology relative to accepted reference methods (Cox D J, Clarke W L, Gonder-Frederick L A, Pohl S, Hoover C, Snyder A, “Accuracy of perceiving blood glucose in IDDM”, Diabetes Care, 8(6):529-36, 1985; Clarke W L, Cox D, Gonder-Frederick L A, Carter W and Pohl S L. “Evaluating clinical accuracy of systems for self-monitoring of blood glucose” Diabetes Care, 10, 622-628, 1987). There were a total of 20,362 paired points for all 58 subjects with YSI venous measurements and Freestyle Navigator® system interstitial fluid glucose measurements. 81.7% of the paired points fell in the Clarke error grid zone A indicating a high level of clinical accuracy. There were 16.7% of the paired points in the clinically-acceptable (benign error) zone “B”, 0.1% in the overtreatment error zone “C”, 1.9% in the failure to detect error zone “D” and 0.01% in the clinically inaccurate and dangerous error zone “E”.

[0092] The Consensus error grid has been proposed as an alternative to the original error grid zone demarcations, specifically to eliminate the physical proximity of the clinically-unacceptable D zone with the clinically-accurate A zone in the lower left portion of the grid. The results of the Clarke error grid and the Consensus error grid are summarized in the Table (1) below. The Consensus error grid was also defined with five distinct risk levels, but the definitions were specified in terms of effect on clinical action by the patient. Zone A has no effect. Zone B has little or no effect. Zone C has altered clinical action. Zone D has altered clinical action with significant medical risk. Zone E has altered clinical action with potentially dangerous consequences.

[0093] On the Clarke error grid, there were 316 individual points in the D zone. Ninety-five percent of these points were in the lower left quadrant of the error grid.

[0094]

TABLE (1)
Summary statistics of Clarke and Consensus Error Grid
	Clarke		Consensus
	Error		Error
	Grid		Grid
Zone	%	N = 20362	%	N = 20362
A	81.7	16627	85.5	17419
B	16.7	3398	13.6	2776
C	0.1	19	0.8	161
D	1.6	316	0.03	6
E	0.0	2	0.0	0

[0095] On the Consensus error grid, by contrast, the number of points in the significant medical risk D zone is reduced to 6. In addition to the reduction in D zone points, the Consensus error grid shows a higher percentage in the clinically-accurate A zone, a slightly lower percentage in the clinically-acceptable B zone, a slightly higher percentage in the altered clinical action C zone and no points in the dangerous consequence E zone.

[0096] The performance of the Freestyle Navigator® system was also assessed using the mean and median absolute relative difference between the sensor interstitial glucose measurements and the YSI venous sample measurements. The mean absolute relative difference was 12.8% and the median absolute relative difference was 9.3%. A comparison of accuracy and performance by day shows that the system's performance on the fifth day is equivalent to the performance of the first or second day. Table (2) contains data with the error grid statistics as well as the mean and median absolute relative difference from the study separated by day.

[0097]

TABLE (2)
Clarke Error Grid, mean and median absolute relative difference by day
	Day 1	Day 2	Day 3	Day 4	Day 5
Zone	N	%	N	%	N	%	N	%	N	%
A	4354	82.5	3215	82.4	2903	79.4	1688	84.0	4467	80.9
B	865	16.4	646	16.6	668	18.3	285	14.2	934	16.9
C	12	0.2	4	0.1	1	0.0	0	0.0	2	0.0
D	47	0.9	34	0.9	82	2.2	37	1.8	116	2.1
E	0	0.0	2	0.1	0	0.0	0	0.0	0	0.0
Mean ARD		12.6		12.3		14.1		11.9		13.0
Median ARD		9.4		9.3		9.9		7.8		9.5
Total	5278	100.0	3901	100.0	3654	100.0	2010	100.0	5519	100.0

[0098] Additional analysis was done comparing the accuracy and performance of the Freestyle Navigator® system nocturnally and diurnally. The percentage of points in the Clarke error grid A zone was 87.1% at night and 80.6% during the day. The difference in accuracy during the day may be associated with the higher rates of change during the daytime, when all of the glucose and insulin challenges were conducted.

[0099] The data from the present study has also been analyzed using the Continuous Glucose Error Grid Analysis (CG-EGA), designed to incorporate the extra temporal dimension of data provided by continuous glucose monitoring systems (Kovatchev et al.). The rate analysis using the CG-EGA gave a 81.1% in the rate error grid A zone, 14.4% in the rate error grid B zone, 1.5% in the rate error grid C zone, 2.3% in the rate error grid D zone, and 0.7% in the rate error grid E zone. The point analysis using the CG-EGA gave a 83.6% in point error grid A zone, 15.0% in point error grid B zone, 0.1% in point error grid C zone, 1.3% in point error grid D zone, and 0% in point error grid E zone. The CG-EGA analysis combining rate and point information revealed that accuracy, measured as a percentage of accurate readings plus benign errors, was 97.5% (94.2% accurate, 3.4% benign). The CG-EGA accuracy stratified by glycemic state gave 60.4% in hypoglycemia (53.1% accurate, 7.3% benign), 99.3% in euglycemia (95.7% accurate, 3.6% benign) and 98.2% in hyperglycemia (95.4% accurate, 2.8% benign). The difference in accuracy between the hypoglycemic, euglycemic, and hyperglycemic ranges may be related to the high rate of change often associated with the descent into hypoglycemia. Standard egression analysis and Deming regression analysis both gave small, but significant offsets 24.9 and 14.3 mg/dL) that could contribute to the slight decrease in accuracy in hypoglycemia.

[0100] FIGS. 4A and 4B give an expanded view of the data from FIG. 2 on a four-hour time axis and centered about the glucose challenge and the insulin challenge, respectively. More specifically, FIG. 4A illustrates a zoomed in view (four hour duration) of Freestyle Navigator®™ sensor data and YSI measurements during the glucose challenge. Referring to FIG. 4A, the continuous glucose sensor data in one minute intervals are shown in the two solid curves (solid from the arm, dashed from the abdomen). The 15 minute YSI venous sample data are shown in the triangles. The time between the nadir of the YSI data and the Freestyle Navigator® system is approximately 24 minutes. The time between the peak of the YSI data and the Freestyle Navigator® system data is approximately 19 minutes.

[0101] Additionally, FIG. 4B shows data from two Freestyle Navigator® sensors, compared with fifteen minutes venous samples measured with the YSI from the insulin challenge in one patient in the study. Referring to FIG. 4B, the Freestyle Navigator® projected alarm, would have alerted the subject to an impending hypoglycemic event 26 minutes before the blood sugar crossed the 70 mg/dL hypoglycemic threshold. At the time of the alarm, the Freestyle Navigator® system glucose was approximately 175 mg/dL and the YSI reading was approximately 90 mg/dL and the rate of change was −3.5 mg/dL/min.

[0102] Both FIGS. 4A and 4B show the temporal tracking of the FreeStyle Navigator® system compared against the venous reference samples. The expanded temporal axis used in FIGS. 4A and 4B also permits more direct visualization of the time lag between the Freestyle Navigator® system interstitial fluid glucose measurement and the venous reference sample measurements. The temporal offset between the FreeStyle® Navigator system and the venous reference measurements was also analyzed by applying a time shift in order to minimize the mean absolute relative difference.

[0103] After correction for the calibration bias, this resulted in an average 12.8 minute lag between the glucose values measured in the interstitial fluid and in the venous samples. This is consistent with previously published studies on the physiological lag between interstitial fluid glucose and blood glucose (see for example: Rebrin K, Steil G M, van Antwerp W P, Mastrotoraro J J, “Subcutaneous glucose predicts plasma glucose independent of insulin: implications for continuous monitoring”, Am J Physiol., 277(3 Pt 1):E561-71, 1999; Steil G M, Rebrin K, Mastrototaro J, Bernaba B, Saad M F, “Determination of plasma glucose during rapid glucose excursions with a subcutaneous glucose sensor”, Diab. Tech. Ther, 5:27-31, 2003; Thennadil S N, Rennert J L, Wenzel B J, Hazen K H, Ruchti T L, Block M B, “Comparison of glucose concentration in interstitial fluid, and capillary and venous blood during rapid changes in blood glucose levels”, Diab. Tech. Ther., 3(3):357-65, 2001).

[0104] The performance of the arm and abdominal sensors was comparable with equivalent Clarke error grid statistics and mean absolute relative difference. The precision of the matched Freestyle Navigator® sensors worn on the arm and abdomen had a coefficient of variation of 10% (n=312, 953). There was no difference in performance of the sensor as a function of age, gender or ethnicity. However, there were small but measurable differences in the accuracy of the sensor depending on the subject's BMI and also on the years since diagnosis. Subjects with BMI less than 25.0 had 78.8% in the Clarke error grid A zone (N=4844), whereas subjects with BMI between 25.0 and 30.0 had 82.2% in the Clarke error grid A zone (N=7855) and subjects with BMI greater than 30.0 had 84.4% in the Clarke error grid A zone (N=3928). Similarly, there were small but measurable differences in accuracy depending on the years since diagnosis of type 1 diabetes. The highest accuracy, 88.5% in the Clarke error grid A zone, was found in subjects who had been diagnosed with diabetes for five years or less (N=2066) and 81.3% for subjects diagnosed between 5 and 25 years (N=9133). Subjects diagnosed with type 1 diabetes for over 25 years had 79.9% in the Clarke error grid A zone (N=5448).

Clinical Accuracy Under Special Circumstances

[0105] The evaluation of the overall accuracy and performance of the FreeStyle Navigator® continuous glucose monitor included periods of deliberately-induced rapidly rising and rapidly falling blood glucose, i.e. in response to the glucose and insulin challenges. There were significant differences in the accuracy compared with the laboratory reference measurements depending on the different rates of change of the underlying blood glucose. Table (3) gives the Clarke error grid statistics and the median absolute relative difference percentage as a function of the rate of change of blood glucose as determined by the YSI measurements. The effect of the physiological lag on the accuracy of the sensor values compared to venous reference samples is more pronounced at the high rates of change, particularly during when the absolute rate of change exceeds 2 mg/dL/min.

[0106]

TABLE (3)
Rate of change and Clarke error grid statistics and median ARD
Rate of
Change		Clarke Error Grid Region	Median
(mg/dL/min)	N	A	B	C	D	E	ARD %
<−2	601	54.6	42.3	1.3	1.8	0.0	17.4
−2 to −1	1728	71.7	26.2	0.3	1.8	0.0	11.8
−1 to 1	14653	84.9	13.5	0.0	1.5	0.0	8.5
1 to 2	1954	79.8	18.9	0.0	1.3	0.0	11.0
>2	691	63.5	34.7	0.0	1.7	0.0	16.9

[0107] FIG. 5 illustrates the rate of change histogram showing underlying rate of change at high resolution (in units of 0.25 mg/dL/min) and in units of the Navigator receiver trend arrows (1.0 mg/dL/min). The rate of change of glucose as measured by the sensor was between −1 and +1 mg/dL/min 74.6% of the time. Referring to FIG. 5, there is a slight difference in the measured occurrence of absolute rates of change less than 1 mg/dL/min due to the different sampling frequency and temporal extent of the Freestyle Navigator® system and YSI measurements.

[0108] The Freestyle Navigator® trend arrows would have been in the horizontal position indicating an absolute rate of change less than 1 mg/dL/min 74.1% of the time for which the YSI data revealed 71.9% of all readings in this range. Both values are consistent with previously reported results (see for example: Dunn T C, Eastman R C, Tamada J A, “Rates of glucose change measured by blood glucose meter and the GlucoWatch Biographer during day, night, and around mealtimes”, Diabetes Care 27: 2161-2165, 2004; Kovatchev, B. P., Clarke, W. L., Breton, M., Brayman, K. and McCall, A. “Quantifying Temporal Glucose Variability in Diabetes via Continuous Glucose Monitoring: Mathematical Methods and Clinical Application” Diab. Technol. Thera., 7, 849-862, 2005).

[0109] FIG. 6 illustrates Clarke error grid for YSI rates of change between −1 to 1 mg/dL/min showing increase in accuracy during modest rates of change. Referring to FIG. 6, whereas the overall percentage of paired points in the Clarke error grid A zone was 81.7%, the percentage in the A zone for rates of change between −1 mg/dL/min and +1 mg/dL/min was significantly higher at 84.9%. Similarly, the mean and median absolute relative differences at these times were 11.4% and 8.5% respectively.

[0110] The accuracy of the Freestyle Navigator® continuous glucose monitor was evaluated in comparison to a standard laboratory reference method using venous blood samples. The overall mean and median absolute relative difference of the sensor in the current study of 12.8% and 9.3% represent a significantly higher level of accuracy than previously published results from other continuous glucose monitoring systems (see for example, Diabetes Research in Children Network (DirecNet) Study Group: “The Accuracy of the CGMS in Children with Type 1 Diabetes: Results of the Diabetes Research in Children Network (DirecNet) Accuracy Study”. Diabetes Technol Ther 5(5):781-789, 2003; Diabetes Research in Children Network (DirecNet) Study Group: “The Accuracy of the GlucoWatch G2 Biographer in Children with Type 1 Diabetes: Results of the Diabetes Research in Children Network (DirecNet) Accuracy Study”. Diabetes Technol Ther 5(5):791-800, 2003; Tansey M J, Beck R W, Buckingham B A, Mauras N, Fiallo-Scharer R, Xing D, Kollman C, Tamborlane W V, Ruedy K J, “Accuracy of the modified Continuous Glucose Monitoring System (CGMS) sensor in an outpatient setting: results from a diabetes research in children network (DirecNet) study.” Diab. Tech. Ther. 7(1):109-14, 2005; Garg S., Zisser H., Schwartz S., Bailey T., Kaplan R., Ellis S. and Jovanovic L, “Improvement in glycemic excursions with a transcutaneous, real-time continuous glucose sensor”, Diabetes Care, 29, 44-50, 2006).

[0111] The high accuracy of the system as measured by the percentage in the Clarke error grid A zone and the mean and median absolute relative differences remained high over the entire five days. There was a small, but measurable improvement in the Clarke error grid statistics and the absolute relative difference values on the fourth day. This is due principally to the fact that there were no glucose challenges administered on the fourth day of the study resulting in fewer rates of change on that day less than 2 mg/dL/min than on other days. In addition, there may be a small increase in accuracy on the fourth day associated with the final system calibration at 72 hours after sensor insertion. Similarly, the slight decrease in accuracy observed on the third and fifth days of the sensor wear may be associated with the fact that these days had a greater number of glucose and insulin challenges than other days in the study, resulting in more absolute rates of change on those days in excess of 2 mg/dL/min.

[0112] A significant portion of the apparent discrepant points between the Freestyle Navigator® and the venous reference samples are likely due to the physiological lag alone. An example of the effect of physiological lag on accuracy is the point at the nadir of the curves in FIG. 4B, which is categorized in the Clarke error grid analysis as a clinically unacceptable D zone point. In this case, although the point-wise comparison of the Freestyle Navigator® sensor value and the venous reference sample value suggests a failure to detect a hypoglycemic event, it is clear from the data that the Freestyle Navigator® system is correctly tracking the fall of the subject's glucose level.

[0113] In the case shown in FIG. 4B, with the projected alarm capability enabled and the detection threshold set at 70 mg/dL, the device would have alerted the user to a predicted change in clinical state from euglycemia to hypoglycemia when the Freestyle Navigator® glucose value was approximately 175 mg/dL and the measured rate of glucose decrease was in excess of −3.5 mg/dL/min. At that moment, the trend arrow was in the downward vertical direction, indicating a rate of glucose decrease of greater than 2 mg/dL/min, and the device's alarm would have used predictive algorithm to identify that the subject would be hypoglycemic in thirty minutes.

[0114] At the time when the projected alarm would have alerted the subject to an impending hypoglycemic event, the YSI reading was approximately 90 mg/dL. An interpolation of the YSI data indicates that the subject's blood sugar crossed the 70 mg/dL threshold for hypoglycemia approximately twenty-six minutes later. Although the paired YSI and Freestyle Navigator® system points at the nadir of the curve result in a D zone point on the Clarke error grid, it is clear from a detailed analysis that the projected alarm would have alerted the subject to an impending hypoglycemic event in a timely manner.

[0115] Another important measure of the clinical accuracy, and ultimately the clinical utility, of the Freestyle Navigator®system is the percentage of points in the clinically-accurate Clarke error grid A zone. A recent numerical simulation study evaluated the effect of sensor inaccuracy on the statistics associated with glucose monitoring error grid analysis using data from a clinical trial of a continuous glucose monitoring system in type 1 children and adolescents (Kollman et al., 2005). In the numerical study, paired points from the actual continuous glucose monitoring system and a laboratory reference method were randomly “shuffled” to simulate a high degree of sensor inaccuracy. The study found that 78% of the randomly shuffled paired points were still in the combined A and B zones of the Clarke error grid. A more useful measure of the clinical accuracy and utility of new glucose monitoring technology may be the percentage of points in the clinically-accurate Clarke error grid A zone alone. (Kollman C, Wilson D M, Wysocki T, Tamborlane W V, Beck R W, “Limitations of the Statistical measures of Error in Assessing the Accuracy of continuous Glucose Sensors”, Diab. Tech. Ther., 7(5):665-672, 2005). An alternative to the more commonly-used metric of combined A and B zone percentage is to rely instead on the total percentage in the A zone alone. The results of the present study showing the Freestyle Navigator® system achieving 81.7% in the A zone alone represent a new level of performance for continuous glucose monitoring systems.

[0116] The high accuracy and performance of the Freestyle Navigator® system at night is also in contrast with previous reports of continuous glucose monitoring systems that exhibited sustained periods of anomalous nocturnal hypoglycemia (see for example: McGowan K. Thomas W, Moran A. “Spurious reporting of nocturnal hypoglycemia by CGMS in patients with tightly controlled type I diabetes” Diabetes Care 2002; 25: 1499-1503; Metzger My Leibowitz G, Wainstein J, Glaser B, Raz I. “Reproducibility of glucose measurements using the glucose sensor” Diabetes Care 2002; 25: 1185-1191; Mauras N, Beck R W, Ruedy K J, Koliman C, Tamborlane W V, Chase H P “Lack of accuracy of continuous glucose sensors in hialthy nondiabetic children: results of the Diabetes Research in Children Network (DirecNet) accuracy study” J Pediatr 2004; 144:770-775).

[0117] The difference in accuracy as a function of BMI may be related to the length of the Freestyle Navigator® sensor and the thickness of the subcutaneous adipose tissue layer in subjects with BMI less than 25. Anthropometric data strongly suggests that the insertion of the Freestyle Navigator® sensor in the upper arm or abdomen will result in the sensor being placed as intended in the subcutaneous adipose tissue layer in most individuals (Horejsi, R., Moller, R., Pieber, T R, Wallner, S., Sudi, K, Reibnegger, G. and Tafeit “Differences of subcutaneous adipose tissue topography between type 2 diabetic men and healthy controls” Exp. Biol. Med., 227, 794-798, 2002). However, in some individuals with low BMI, the data indicate that the subcutaneous adipose tissue layer thickness on the posterior arm upper arm or even the lower abdominal quadrant may be only slightly greater than the required 6 mm thickness needed to properly accommodate the sensor. Although the overall sensor performance in subjects with BMI less than 25 is still excellent (78.8% in the clinically-accurate Clarke error grid A zone), there is a small but measurable difference when compared with subjects with BMI greater than 30 (84.4% in the clinically-accurate Clarke error grid A zone). In the low BMI subjects with reduced subcutaneous adipose tissue layer thickness, the proximity of skeletal muscle tissue to the sensor in the adipose tissue could increase the effect reported by Moberg et al. in which tissue glucose nadirs were not only delayed relative to plasma, but also reduced especially during insulin-induced hypoglycemia (Moberg E, Hagstrom-Toft E, Amer P. and Bolinder J. “Protracted glucose fall in subcutaneous adipose tissue and skeletal muscle compared with blood during insulin-induced hypoglycaemia” Diabetologia 40, 1320-1326, 1997).

[0118] In the present study, the apparent difference in accuracy as a function of years since diagnosis is most likely also a result of the weak dependence of accuracy on BMI. The 6 subjects with a diagnosis of diabetes less than five years, for whom there was the highest percentage in the Clarke error grid A zone and the lowest median absolute relative difference, also by chance had the highest mean BMI (29.8). Similarly, the 18 subjects with lowest BMI (<24.9) in the study happened to also have the highest mean years since diagnosis of diabetes (30.1 years).

Insulin Adjustment Analysis

[0119] The Insulin Adjustment Analysis evaluates the difference between insulin dosing based on Freestyle Navigator® Continuous Glucose Monitoring System (CM) readings and that based on reference readings. The interpretation of the analysis is best understood considering a hypothetical patient with a glucose target level of 90-120 mg/dL and an insulin sensitivity of 30 mg/dL/unit. The glucose target level represents aggressive therapy where the therapeutic goal is to keep glucose squarely in the normal range. The analysis is targeted to meet the requirements of intensive insulin therapy. The choice of insulin sensitivity was made to simplify interpretation—the treatment differences between Navigator CM and YSI are calculated in whole number differences in the units of insulin. This seemingly arbitrary choice of the hypothetical patient has no influence on the results of the Insulin Adjustment Analysis—the choice was based on the goals of intensive insulin therapy and the ease of interpretation of the results.

[0120] The Insulin Adjustment Analysis data is reported as differences in units of insulin. (see Table 4). This is an intermediate result that allows a more detailed characterization of the data than the final summary (see Table 5). Decisions with Navigator CM were rated Correct 89.3% (1180/1322) of the time and Acceptable 7.6% (100/1322) of the time. Since the Acceptable rating translates to a glucose adjustment to within the normal glucose range, accurate adjustments are the sum of Correct and Acceptable categories, 96.8% (1280/1322).

[0121]

TABLE 4
Treatment Difference for the Hypothetical Patient with Insulin
Sensitivity = 30 mg/dL/unit and Glucose Target = 90-120 mg/dL
Navigator CM-
YSI Treatment
Difference	Glucose <200 mg/dL	Glucose ≧200 mg/dL
(Units of insulin)	N	%	Category	N	%	Category
−4	0	0	Hyperglycemia 2	4	0.6	Hyperglycemia 2
−3	1	0.1	Hyperglycemia 1	13	2.0	Hyperglycemia 1
−2	11	1.6	Acceptable	78	12.1	Acceptable
−1	120	17.7	Correct	215	33.4	Correct
0	353	52.0	Correct	240	37.3	Correct
1	173	25.5	Correct	79	12.3	Correct
2	18	2.7	Possible Error	11	1.7	Acceptable
3	2	0.3	Error	3	0.5	Possible Error
4	1	0.1	Error	0	0	Error
Total	679	100		643	100

[0122]

TABLE 5
Insulin Adjustment Analysis Summary
Category	Effect on Blood Glucose	N	%
Correct	Within ±30 mg/dL of target	1180	89.3
	glucose
Acceptable	Within normal glucose range	100	7.6
Possible Error (hypo)	60 mg/dL below target glucose	21	1.6
Error (hypo)	≧90 mg/dl below target glucose	3	0.2
Hyperglycemia 1	90 mg/dL above target glucose	14	1.1
Hyperglycemia 2	≧120 mg/dl above target glucose	4	0.3
Total		1322	100

[0123] In summary, this analysis describes 3 occurrences of “Error (hypo)” and 4 occurrences of “hyperglycemia 2” being potentially indicated from 1322 decision points analyzed.

Glucose Peak

[0124] Continuous glucose monitoring provides the ability to identify and quantify the maximum glucose excursions after meals and during the night. The quantification of glucose peaks was clinically accurate 88.1% of the time and clinically useful 97.6% of the time (see Table 6).

[0125]

TABLE 6
Glucose Peak Analysis
Difference	Clinical Assessment	N	%
±15 mg/dL	Accurate	263	41.5
±45 mg/dL	Accurate	295	46.6
±75 mg/dL	Useful	60	9.5
±105 mg/dL	Misclassification	14	2.2
±135 mg/dL	Misclassification	1	0.2
Total		633	100.0

Insulin Adjustment Analysis

[0126] The Insulin Adjustment Analysis evaluates the hypothetical difference between insulin dosing based on Navigator CM readings to that based on a blood glucose meter such as Freestyle Blood Glucose (BG) readings. The interpretation of the analysis is best understood considering a hypothetical patient with a glucose target level of 90-120 mg/dL and an insulin sensitivity of 30 mg/dL/unit. The glucose target level represents aggressive therapy where the therapeutic goal is to keep glucose squarely in the normal range. The analysis is targeted to meet the requirements of intensive insulin therapy. The choice of insulin sensitivity was made to simplify interpretation—the treatment differences between Navigator CM and Freestyle BG YSI (see Table 7) are calculated in whole number differences in the units of insulin. This seemingly arbitrary choice of the hypothetical patient has no influence on the results of the Insulin Adjustment Analysis—the choice was based the goals of intensive insulin therapy and the ease of interpretation of the results.

[0127] The Insulin Adjustment Analysis data is reported as differences in units of insulin (see Table 7). There were 6,040 paired (Navigator CM-Freestyle BG) glucose readings available at times of subject-reported insulin dosing or bedtime in the Home Use Study. The analysis is summarized in Table 8 with 86.5% (5226/6040) of the readings correct and 94.3% (5696/6040) accurate or acceptable. These results provide approximately 89.3% (1180/1322) correct and 96.8% (1280/1322) accurate or acceptable.

[0128]

TABLE 7
Treatment Difference for the Hypothetical Patient
with Insulin Sensitivity = 30 mg/dL/
unit and Glucose Target = 90-120 mg/dL
Difference
in Insulin	Glucose < 200 mg/dL	Glucose ≧ 200 mg/dL
Dose (Units)	N	(%)	N	(%)
4	0	0	1	0.0
3	11	0.3	2	0.1
2	84	2.1	14	0.7
1	810	20.1	162	8.1
0	2362	58.5	530	26.5
−1	675	16.7	687	34.3
−2	89	2.2	367	18.3
−3	8	0.2	163	8.1
−4	0	0	75	3.7
Total	4039	—	2001	—

[0129]

TABLE 8
Insulin Adjustment Analysis Summary
Category	Effect on Blood Glucose	N	%
Correct	Within ±30 mg/dL of target	5226	86.5
	glucose
Acceptable	Within normal glucose range	470	7.8
Possible Error	60 mg/dL below target glucose	86	1.4
(hypo)
Error (hypo)	≧90 mg/dl below target	12	0.2
	glucose
Hyperglycemia 1	90 mg/dL above target glucose	171	2.8
Hyperglycemia 2	≧120 mg/dl above target	75	1.2
	glucose
Total		6040	100

[0130] Insulin dosing or bedtime was not indicated for 5,447 of the 11,487 Freestyle BG duplicate points. The Insulin Adjustment Analysis was also conducted using the 5,447 Freestyle BG duplicate points for which there was no indication of insulin injection to determine if there was a substantive difference between the two populations. The Insulin Adjustment Analysis data is reported as differences in units of insulin (See Table 9). The results are slightly better for the points where insulin injections were not indicated (See Table 10) with 89.4% (4868/5447) correct and 95.5 (5203/5447) correct or acceptable.

[0131]

TABLE 9
Treatment Difference for the Hypothetical Patient with
Insulin Sensitivity = 30 mg/dL/unit and
Glucose Target = 90-120 mg/dL - Non-insulin Injection Points
Difference
in Insulin	Glucose < 200 mg/dL	Glucose ≧ 200 mg/dL
Dose (Units)	N	(%)	N	(%)
4	2	0.0	0	0
3	11	0.3	1	0.1
2	95	2.3	26	2.0
1	876	21.2	132	10.0
0	2473	59.9	388	29.5
−1	588	14.2	411	31.2
−2	81	2.0	228	17.3
−3	5	0.1	88	6.7
−4	0	0	42	3.2
Total	4131	—	1316	—

[0132]

TABLE 10
Insulin Adjustment Analysis Summary
Non-insulin Injection Points
Category	Effect on Blood Glucose	N	%
Correct	Within ±30 mg/dL of target	4868	89.4
	glucose
Acceptable	Within normal glucose range	335	6.2
Possible Error	60 mg/dL below target glucose	96	1.8
(hypo)
Error (hypo)	≧90 mg/dl below target glucose	13	0.2
Hyperglycemia 1	90 mg/dL above target glucose	93	1.7
Hyperglycemia 2	≧120 mg/dl above target glucose	42	0.8
Total		5447	100

[0133] When a patient adjusts an insulin dose using a blood glucose meter such as Freestyle Blood Glucose monitor, there is no indication if glucose is changing. If glucose is rising at the time of glucose dosing, there is insufficient insulin to stabilize blood glucose and the predicted insulin dose will be too small. Likewise, if glucose is descending, there is already insulin in the blood, and the predicted insulin dose will be too large. The rate of glucose change indicated by Navigator CM at the time of insulin dosing (see Table 11) indicates glucose changes >±2 mg/dL/minute 4.0% of the time, and >±1 mg/dL/minute 18.3% of the time. The agreement of static the blood glucose meter readings with static reference readings is excellent, but the interpretation of this agreement to suggest accurate insulin dosing with the blood glucose meter is not correct. When insulin is dosed with no knowledge of changing glucose levels, the dosing will be incorrect a significant fraction of the time. The determination of 94.3% Navigator CM dosing accuracy in this study and 96.8% Navigator CM dosing accuracy in a previous study provide realistic estimations when the rate of glucose change is also known.

[0134]

TABLE 11
Navigator CM Rate Indication at the Time of Insulin Dosing
Navigator CM Rate of Change (mg/dL/minute)	N	(%)
>2.0	330	3.3
1.0 to 2.0	897	9.0
−1.0 to 1.0	8140	81.7
−2.0 to −1.0	526	5.3
<−2.0	72	0.7

[0135] The description below details a further user study results from a highly accurate continuous glucose monitoring system such as, for example, Freestyle Navigator® system. Of the 137 subjects enrolled in the investigation, 123 completed the 40-day monitoring period. The other 14 subjects withdrew from the study due to non-compliance with protocol demands (n=8) or difficulties handling the device (n=6). None of the discontinued subjects participated in the unblinded portion of the study. The glucose data available for the discontinued subjects was included in the paired point analysis.

[0136] The performance of the Freestyle Navigator® was assessed using the absolute relative difference between the sensor interstitial glucose measurements and the blood glucose measurements. Data from 961 sensors with 11,487 paired FreeStyle BG reference values were evaluated. The mean absolute relative difference was 14.4% and the median absolute relative difference was 11.1%. The mean absolute relative difference indicates that, on average, the CM reading was 14.4% higher or lower than the corresponding BG reading. The median absolute relative difference indicates that the CM reading was equally as likely to be within 11.1% of the BG reading, either higher or lower, as it was to be outside of that range.

[0137] The equation for the Deming regression had a slope of 0.83, an intercept of 21.8 mg/dL and correlation coefficient of 0.92. These results demonstrate a strong correlation between CM and BG readings.

[0138] FIG. 7 shows the Clarke error grid for the study. There were a total of 11,487 paired points with averaged duplicate BG reference values and interpolated CM values, from 131 subjects. No paired points were available from six subjects. Of the 11,487 paired points, 77.2% fell in the Clarke error grid zone A, indicating a high level of correspondence between the reference blood glucose measurements and the CM results. There were 19.6% of the paired points in zone B and only 3.2% outside the A and B zones. Results for all the Clarke error grid zones are shown in Table 12 below. The results of the Consensus error grid are also included in Table 12.

[0139]

TABLE 12
Summary statistics of Clarke and Consensus Error Grid
	Clarke Error		Consensus
	Grid		Error Grid
Zone	N	(%)	N	(%)
A	8863	77.2	9180	79.9
B	2255	19.6	2194	19.1
C	1	0.0	109	0.9
D	365	3.2	4	0.0
E	3	0.0	0	0.0
N paired points	11487		11487

[0140] On the Clarke error grid, there were 365 individual points in the D zone. On the Consensus error grid, by contrast, the number of points in the D zone is reduced to four. In addition, the Consensus error grid shows 79.9% in the A zone, 99.0% in the A and B zones, less than 1% in the C and D zones and no points in the E zone.

[0141] A comparison of accuracy and performance by day of sensor wear shows that the system's performance on the fifth day is nearly equivalent to the performance on the first or second day. Table 13 contains data with the error grid statistics as well as the mean absolute relative difference from the study separated by day.

[0142]

TABLE 13
Clarke Error Grid, absolute relative difference by day
	Day 1	Day 2	Day 3	Day 4	Day 5
Zone	N/(%)	N/(%)	N/(%)	N/(%)	N/(%)
Clarke A	1061	(77.8)	2182	(77.4)	2110	(77.7)	1884	(79.3)	1626	(73.5)
Clarke B	266	(19.5)	551	(19.6)	516	(19.0)	427	(18.0)	495	(22.4)
Clarke C	0	(0.0)	0	(0.0)	0	(0.0)	0	(0.0)	1	(0.0)
Clarke D	36	(2.6)	84	(3.0)	91	(3.3)	63	(2.7)	91	(4.1)
Clarke E	1	(0.1)	1	(0.0)	0	(0.0)	1	(0.0)	0	(0.0)
N paired	1364		2818		2717		2375		2213
points
Consensus A
Consensus B
Consensus C
Consensus D
Consensus E
N paired	1364		2818		2717		2375		2213
points
Mean	14.8		14.3		14.0		13.9		15.3
ARD
Median
ARD

[0143] Table 14 shows that the CM readings are optimal when blood glucose is relatively stable (i.e., when the rate is within +/−1 mg/dL/min). As expected the bias increases somewhat as the magnitude of the rate of glucose change increases. However, the displayed rate arrow provides the necessary information to properly interpret the glucose result in these situations. The mean bias for glucose <100 mg/dL and the mean percent bias for glucose ≧100 mg/dL become increasingly positive as the rate decreases from +2 mg/dL/minute to −2 mg/dL/minute. Lag in the interstitial readings versus capillary blood glucose readings is the explanation for this result. When glucose levels were rising, the CM values were low, on average, versus BG with the difference versus BG lower for rising glucose (>1 mg/dL/minute) than for stable glucose (±1 mg/dL/minute). When glucose levels were falling CM was high, on average, versus BG with the difference versus BG higher for falling glucose (<1 mg/dL/minute) than for stable glucose (±1 mg/dL/minute).

[0144]

TABLE 14
Difference measures vs. glucose rate of change
	Navigator CM Rate of Change
	(mg/dL per minute)	Mean	Median	N
Difference (mg/dL) for glucose <100 mg/dL
	>2.0	3.7	−1.2	3
	1.0 to 2.0	4.7	5.5	33
	−1.0 to 1.0	7.6	7.1	2028
	−2.0 to −1.0	17.9	18.7	261
	<−2.0	26.5	24.4	50
Absolute difference (mg/dL) for glucose <100 mg/dL
	>2.0	11.0	9.8	3
	1.0 to 2.0	12.5	9.9	33
	−1.0 to 1.0	13.3	10.8	2028
	−2.0 to −1.0	21.5	19.4	261
	<−2.0	32.4	27.0	50
Percent difference % for glucose ≧=100 mg/dL
	>2.0	−13.7	−14.3	152
	1.0 to 2.0	−10.9	−10.7	581
	−1.0 to 1.0	−3.5	−3.7	7245
	−2.0 to −1.0	6.9	6.8	432
	<−2.0	7.5	9.1	69
Absolute % difference % for glucose ≧=100 mg/dL
	>2.0	17.0	16.1	152
	1.0 to 2.0	14.8	12.6	581
	−1.0 to 1.0	12.2	9.8	7274
	−2.0 to −1.0	15.9	12.6	432
	<−2.0	18.3	14.5	69

[0145] The Clarke EGA as a function of Navigator rate (Table 15) exhibits the expected behavior. When glucose is descending by at least −2 mg/dL/min, there is a higher likelihood that a reading would fall into the left Zone D than when the glucose is stable or rising. When glucose is rising, there is a higher likelihood that a reading would fall into the right Zone D. The rate arrow provides the valuable information to properly interpret the glucose result (i.e. when glucose is rapidly descending Navigator CM tends to be higher than Navigator BG and when glucose is rapidly ascending Navigator CM tends to be lower than Navigator BG).

[0146]

TABLE 15
Clarke EGA vs. glucose rate of change
Zone	<−2.0	%	−2.0 to −1.0	%	−1.0 to 1.0	%	1.0 to 2.0	%	>2.0	%
A	61	51.3	425	61.3	7372	79.3	455	74.1	101	65.2
B	45	37.8	194	28.0	1688	18.1	149	24.3	48	31.0
C	0	0.0	1	0.1	0	0.0	0	0.0	0	0.0
D	12	10.1	73	10.5	240	2.6	10	1.6	6	3.9
E	1	0.8	0	0.0	2	0.0	0	0.0	0	0.0
Total	119		692		9302		614		155

Sensor Success Measures

[0147] The rate of successful sensor insertions was evaluated from reported results of each sensor insertion attempt, as well as the electronic records stored by the Receiver. The electronic records were used to determine whether each sensor was detected by the Receiver, and whether the user followed the steps in the labeling. The percentage of insertions that were successful, when used as directed, was similar for the blinded (96.0%) and unblinded (96.8%) phases of the study (96.4% overall). The percentage of successful insertions was similar for the arm (95.7%) and abdomen (97.4%) insertion sites. Abdomen insertions may have been more successful because it is easier to see the entire insertion process at the abdomen site when inserting a sensor on oneself.

[0148] The success rate for the initial Sensor Calibration process was evaluated from results recorded in the receiver log data for each successful sensor insertion attempt. The time required to complete the first sensor calibration was evaluated in addition to the overall success or failure. The percentage of sensors that were successfully calibrated and produced glucose results within the first 12 hours was calculated. Sensor calibration is not allowed within the first 10 hours. Sensors that could not be calibrated because conditions were out of range were excluded, e.g., if the glucose was changing too rapidly for calibration. The percentage of sensors that were successfully calibrated within 12 hours, when used as directed, was similar for the blinded (90.5%) and unblinded (92.6%) phases of the study (91.5% overall).

[0149] Sensor duration was evaluated as the time duration from sensor insertion to the last CM glucose result reported for the sensor. Some sensors were removed early by user error or discretion, or because of protocol logistics such as the end of the trial. These sensors are excluded from analysis, unless the sensor reached the nominal 5-day sensor life (>108 hours). The median sensor life was similar for the blinded (119.9 hours) and unblinded (120.0 hours) phases of the study. The percentage of sensors, used as directed, that produced glucose results for 108 hours or more was similar for the blinded (83.5%) and unblinded (83.0%) phases of the study. Sensors on the arm tended to have slightly longer duration (86.2% for >108 hours) than those on the abdomen (79.4%), because there is somewhat less flexing and folding of the skin at the posterior arm insertion site than on the abdomen, improving the effectiveness of the skin adhesive that holds the sensor in place.

Glycemic Analysis

[0150] The change in glycemic status between the blinded and unblinded phases of the study was stratified by type 1 and type 2 diabetes. During the unblinded phase when alarms were set, subjects were instructed to perform a BG test when alarms were triggered. Some important differences in controlling glucose concentration with insulin administration between the two types of diabetes are the following:

• • [0151] Subjects with type 2 diabetes are less likely to induce hypoglycemia with insulin because they are insensitive to insulin. Type 1 subjects, with normal insulin sensitivity are much more likely to induce hypoglycemia.
  • [0152] Subjects with type 2 diabetes can reduce hyperglycemia by reducing carbohydrate ingestion and allowing endogenous insulin to reduce blood glucose. Patients with type 1 diabetes produce no endogenous insulin, so a reduction of carbohydrates is not a viable strategy for controlling glucose. Controlling glucose with injected insulin is much more difficult than control with endogenous insulin.

[0153] The time spent in hypoglycemic (<70 mg/dL), euglycemic (70-180 mg/dL) and hyperglycemic ranges is illustrated in FIG. 8 for type 1 and 2 subjects in the blinded and unblinded phases of the study.

[0154] The type 1 subjects improved in the unblinded phase by reducing time in hypoglycemia. The time spent below the 70 mg/dL threshold for hypoglycemia was reduced by 42% from 1.4 hours to 0.8 hours (p<0.0001). The time spent in hyperglycemia (>180 mg/dL) did not change.

[0155] For type 2 subjects, the duration of hyperglycemia improved in the unblinded phase. The time spent in the euglycemic range increased by 12% (p=0.0027) and the time spent >180 mg/dL decreased by 18% (p=0.0057). As anticipated, the measures of hypoglycemia for type 2 subjects, which were low in the blinded phase, were largely unchanged in the unblinded phase.

[0156] Accordingly, a continuous analyte monitoring system in one embodiment includes an analyte sensor having at least about 80% of its paired data points within zone A and at least about 95% of its paired data points within zone A and zone B of the Clarke Error Grid, a transmitter capable of receiving information from the sensor, and a receiver capable of receiving information from the transmitter.

[0157] In one aspect, analyte sensor has at least about 85% of its paired data points within zone A of the Clarke Error Grid.

[0158] In a further aspect, the analyte sensor has at least about 90% of its paired data points within zone A of the Clarke Error Grid.

[0159] In still a further aspect, the analyte sensor has more than approximately 90% of its paired data points within zone A of the Clarke Error Grid.

[0160] Additionally, in another aspect, the analyte sensor has at least about 85% of its paired data points within zone A of the Consensus Error Grid, and further, where the analyte sensor has at least about 90% of its paired data points within zone A of the Continuous Glucose Error Grid Analysis.

[0161] The analyte sensor may be a glucose sensor.

[0162] In yet another aspect, the system may not require confirmation of analyte data obtained by the system.

[0163] The system may include a drug delivery device, where one or more of the transmitter and the receiver may be adapted to transmit analyte information to the drug delivery device.

[0164] In another aspect, the analyte sensor may be calibrated using single point calibration.

[0165] A continuous analyte monitoring system in accordance with another embodiment includes an analyte sensor having at least about 85% of its paired data points within zone A and at least about 95% of its paired data points within zone A and zone B of the Consensus Error Grid, a transmitter capable of receiving information from the sensor, and a receiver capable of receiving information from the transmitter.

[0166] The analyte sensor may have at least about 85% of its paired data points within zone A of the Consensus Error Grid.

[0167] The analyte sensor may have at least about 90% of its paired data points within zone A of the Consensus Error Grid.

[0168] The analyte sensor may have more than approximately 90% of its paired data points within zone A of the Consensus Error Grid.

[0169] In another aspect, the system may not require confirmation of analyte data obtained by the system.

[0170] The system may include a drug delivery device, where one or more of the transmitter and the receiver may be adapted to transmit analyte information to the drug delivery device.

[0171] Also, the analyte sensor may be calibrated using single point calibration.

[0172] A method of monitoring glucose levels in accordance with still another embodiment includes determining glucose concentration using a first transcutaneously positioned analyte sensor, reporting glucose concentration to a user, where a second sensor is not used to confirm the accuracy of the first transcutaneously positioned analyte sensor.

[0173] In one aspect, determining may include over a period of time ranging from about 1 day to about 7 days.

[0174] The first transcutaneously positioned analyte sensor may have at least about 85% of its paired data points within zone A of the Clarke Error Grid.

[0175] The first transcutaneously positioned analyte sensor may have at least about 90% of its paired data points within zone A of the Clarke Error Grid.

[0176] The first transcutaneously positioned analyte sensor may have more than about 90% of its paired data points within zone A of the Clarke Error Grid.

[0177] The first transcutaneously positioned analyte sensor may be a glucose sensor.

[0178] The method in a further aspect may include determining health related information based on the reported glucose concentration, where the health related information may include a bolus amount, or one or more of a food intake, medication dosage level, or activity level.

[0179] Also, the medication dosage level may include insulin dosage level.

[0180] In a further aspect, the method may include transmitting the reported glucose concentration, and where transmitting may include one or more of a wired transmission or a wireless transmission.

[0181] In still another aspect, the method may include calibrating the first transcutaneously positioned analyte sensor using single point calibration.

[0182] The first transcutaneously positioned analyte sensor may have at least about 95% of its paired data points within zone A and zone B of the Clarke Error Grid.

[0183] The first sensor may have at least about 85% of its paired data points within zone A.

[0184] A method of monitoring glucose levels in accordance with yet another embodiment includes determining glucose concentration using a first transcutaneously positioned analyte sensor, reporting glucose concentration to a user, where accuracy of the first transcutaneously positioned analyte sensor is established other than with a second sensor.

[0185] In one aspect, the first transcutaneously positioned analyte sensor has at least about 85% of its paired data points within zone A of the Clarke Error Grid.

[0186] In another aspect, the first transcutaneously positioned analyte sensor has at least about 90% of its paired data points within zone A of the Clarke Error Grid.

[0187] In still another aspect, the first transcutaneously positioned analyte sensor has more than about 90% of its paired data points within zone A of the Clarke Error Grid.

[0188] The first transcutaneously positioned analyte sensor may be a glucose sensor.

[0189] A method of monitoring glucose levels using a single glucose sensor in accordance with still yet a further embodiment includes transcutaneously positioning a glucose sensor in a patient for a period of time, determining glucose concentration of the patient using the transcutaneously positioned glucose sensor, and using one or more additional devices during the period of time only to calibrate the glucose sensor but not to confirm the accuracy of the transcutaneously positioned glucose sensor.

[0190] The glucose sensor in one embodiment has at least about 85% of its paired data points within zone A and at least about 95% of its paired data points within zone A and zone B of the Clarke Error Grid.

[0191] The glucose concentration may be determined over a period of time ranging from about 1 day to about 7 days.

[0192] In a further aspect, the glucose sensor has at least about 85% of its paired data points within zone A of the Clarke Error Grid.

[0193] In yet another aspect, the glucose sensor has at least about 90% of its paired data points within zone A of the Clarke Error Grid.

[0194] The glucose sensor in still another aspect has more than approximately 90% of its paired data points within zone A of the Clarke Error Grid.

[0195] In still a further aspect, the method may include determining a health related information based on the determined glucose concentration, where the health related information includes one or more of a food intake, medication dosage level, or activity level, and further, where medication dosage level includes insulin dosage level.

[0196] The method may include transmitting data associated with the determined glucose concentration, where transmitting may include one or more of a wired transmission or a wireless transmission.

[0197] Also, calibration of the glucose sensor may include performing single point calibration.

[0198] An analyte monitoring system in accordance with still yet another embodiment includes an analyte sensor configured to detect one or more analyte levels of a patient, a transmitter unit operatively coupled to the analyte sensor, the transmitter unit configured to transmit one or more signals associated with the detected one or more analyte levels, and a receiver unit configured to receive the transmitted one or more signals associated with the detected one or more analyte levels, where the accuracy of the detected one or more analyte levels relied upon to make a clinically relevant decision is established without using a blood glucose measurement.

[0199] In one aspect, the clinically relevant decision may include healthcare decision.

[0200] The clinically relevant decision may include a bolus amount determination.

[0201] The blood glucose measurement may include a confirmatory blood glucose measurement.

[0202] The detected one or more analyte level may be calibrated, for example, using single point calibration.

[0203] The transmitter unit may be configured to wirelessly transmit the one or more signals to the receiver unit.

[0204] The analyte sensor in one embodiment has at least about 85% of its paired data points within zone A and at least about 95% of its paired data points within zone A and zone B of the Clarke Error Grid.

[0205] An analyte monitoring device in accordance with still yet a further embodiment includes a receiver unit for receiving one or more signals related to an analyte level detected by an electrochemical sensor, the receiver unit including a display to display an indication of the analyte level, where the electrochemical sensor has at least about 85% of its paired data points within zone A and at least about 95% of its paired data points within zone A and zone B of the Clarke Error Grid.

[0206] The electrochemical sensor may have at least about 85% of its paired data points within zone A of the Clarke Error Grid.

[0207] The electrochemical sensor may have at least about 90% of its paired data points within zone A of the Clarke Error Grid.

[0208] The electrochemical sensor may have more than approximately 90% of its paired data points within zone A of the Consensus Error Grid.

[0209] The receiver unit may be configured to calibrate the one or more signals related to the analyte level, and further, where the receiver unit may be configured to display the calibrated one or more signals related to the analyte level without a confirmatory blood glucose measurement.

[0210] In another aspect, the receiver unit may be configured to calibrate the one or more signals related to the analyte level using single point calibration.

[0211] The receiver unit may be configured to display the one or more signals related to the analyte level without a confirmatory blood glucose measurement.

[0212] The receiver unit in one embodiment may include one of an rf receiver or an rf transceiver.

[0213] The receiver unit in still a further aspect may be configured to calibrate the one or more signals related to the analyte level using a calibration value of less that about one microliter of body fluid, where the body fluid includes blood.

[0214] The receiver unit may include an alarm configured to indicate when the analyte level is at or near a threshold level.

[0215] The threshold level may include one of hypoglycemia, impending hypoglycemia, hyperglycemia, or impending hyperglycemia.

[0216] The alarm may include one or more of an audible signal, a visual display, or a vibratory signal.

[0217] The alarm may be configured to automatically deactivate after a predetermined time period.

[0218] The receiver unit in one aspect may be a portable handheld unit.

[0219] The receiver unit may be configured for wearing on or under an article of clothing.

[0220] The receiver unit may include an rf transceiver configured to receive or transmit the one or more signals related to an analyte level.

[0221] In still another aspect, the display may be configured to display one or more of analyte level trend information, rate of change information associated with the analyte level, basal profile information, bolus amount information, or therapy related information.

[0222] In a further aspect, the receiver may include a blood glucose meter.

[0223] The display may be configured to display the indication of the analyte level at least one or more of once per minute, once per five minutes, once per ten minutes, or over a predetermined time period, where the predetermined time period may include one or more of less than 24 hour period, one day, three days, seven days, fourteen days, twenty one days, twenty eight days, less than thirty days, or greater than thirty days.

[0224] A monitoring device in a further embodiment includes a portable housing, an rf receiver coupled to the portable housing, the rf receiver configured to wirelessly receive one or more signals related to an analyte level of a patient detected by an electrochemical sensor, a processing unit coupled to the portable housing and to the rf receiver, the processing unit configured to process the one or more signal received by the rf receiver, and a display unit coupled to the portable housing and the processing unit, the display unit configured to display an indication associated with the one or more signals related to the analyte level of the patient, where the electrochemical sensor has at least about 85% of its paired data points within zone A and at least about 95% of its paired data points within zone A and zone B of the Consensus Error Grid.

[0225] The electrochemical sensor may have at least about 85% of its paired data points within zone A of the Consensus Error Grid.

[0226] The electrochemical sensor may have at least about 90% of its paired data points within zone A of the Consensus Error Grid.

[0227] An analyte monitoring device in accordance with still another embodiment includes a receiver unit for receiving one or more signals related to an analyte level detected by an electrochemical sensor, the receiver unit including a display to display an indication of the analyte level, and the receiver unit further configured to process one or more signals related to analyte related therapy for communication with a drug administration system, where the electrochemical sensor has at least about 85% of its paired data points within zone A and at least about 95% of its paired data points within zone A and zone B of the Clarke Error Grid.

[0228] In one aspect, the electrochemical sensor has at least about 90% of its paired data points within zone A of the Clarke Error Grid.

[0229] Various other modifications and alterations in the structure and method of operation of this disclosure will be apparent to those skilled in the art without departing from the scope and spirit of the present disclosure. Although the present disclosure has been described in connection with specific embodiments, it should be understood that the embodiments of the present disclosure as claimed should not be unduly limited to such specific embodiments. It is intended that the following claims define the scope of the present disclosure and that structures and methods within the scope of these claims and their equivalents be covered thereby.