HbA₁c check method basing on patient glucose concentration in blood, hypoglycemia evaluation method and diabetes mellitus patient check instrument

Description of the invention

The present invention relates to a method of controlling HbA_1c based on the patient's blood glucose data, how the patient is assessed for hypoglycemia, and the patient's diabetes monitoring device.

This invention was created with the support of the US Government in the form of grant no. NIH / NIDDK: RO1 DK 28288 and NIH / NIDDK: RO1 DK 51562, both from the National Institutes of Health. The US government has certain rights to the invention.

The system in question relates generally to glycemic control in diabetic patients, and in particular to a computerized system and method for assessing the prediction of glycosylated hemoglobin (HbA1_C.and HbAO and the risk of impending hypoglycaemia.

Extensive research including studies on diabetes control and complications (DCCT) (see DCCT Research Group: The Effect Of Intensive Treatment Of Diabets On The Development And Progression Of Long-Term Complications Of Insulin-Dependent Diabetes Mellitus. New England Journal of Medicine, 329 ; 978-986, 1993), Stockholm Diabetes Intervention Research (see Reichard P. Phil M; Mortality and Treatment Side Effects During Long-term Intensified Conventional Insulin Treatment in the Stockholm Diabetes Intervention Study, Diabetes, 43: 313-317, 1994) and research into the future of diabetes in the United Kingdom (see UK Prospective Diabetes Study Group: Effect of Intensive Blood Glucose Control With Metformin On Complications in Patients With Type 2 Diabetes (UKPDS 34). Lancet, 352: 837-853, 1998) have repeatedly demonstrated that the most effective way to prevent long-term diabetic complications is to keep blood glucose (BG) levels within the normal range during intensive insulin therapy.

However, the same studies also documented some detrimental effects of intensive insulin therapy, the most serious of which is the increased risk of frequent severe hypoglycemia (SH), a condition known as an episode of neuroglycopenia that prevents self-treatment and requires external help to recover (see DCCT Research Group: Epidemiology of Severe Hypoglycemia In The Diabetes Control and Complications Trial. American Journal of Medicine, 90: 450-459, 1991, and DCCT Research Group: Hypoglycemia in the Diabetes Control and Complications Trial. Diabetes, 46: 271-286,1997). Since severe hypoglycaemia can cause accidents, coma and even death, patients and healthcare professionals are reluctant to undergo intensive care. Consequently, hypoglycemia is considered a major barrier to improved glycemic control (Cryer PE: Hypoglycemia is the Limiting Factor in the Management of Diabetes. Diabetes Metab Res Rev. 15: 42-46,1999).

People with diabetes face the lifelong problem of optimizing the maintenance of tight glycemic control without increasing the risk of hypoglycaemia. The main challenge with this problem is to create simple and reliable methods that allow patients to both control their glycaemia and risk hypoglycemia, but which can be applied to their everyday environment.

It has been known for over twenty years that glycosylated hemoglobin is a marker for glycemic control in patients with diabetes (type I or type II). Many researchers have looked into this relationship and have found that glycosylated hemoglobin essentially reflects the average blood glucose levels of a patient over the previous two months. Since blood glucose levels fluctuate widely over time in most diabetics, it has been proposed to see a real association between integrated glucose control and HbA1c only in patients who are known to have stable glucose control over an extended period of time.

Early studies with such patients showed an almost deterministic relationship between mean blood glucose over the previous 5 weeks and HbA1c, with this curvilinear relationship having a correlation coefficient of 0.98 (see Aaby Svendsen P, Lauritzen T. Soegard U, Nerup J (1982) Glycosylated Hemoglobin and Steady-State Mean Blood Glucose Concentration in Type 1 (Insulin-Dependent) Diabetes, Diabetologia 23, 403-405). In 1993, the DCCT group concluded that HbA1c is a "logical candidate" for glycosylated hemoglobin research. The DCCT group determined a linear relationship between the preceding mean blood glucose value and HbA1c (see Santiago JV (1993). Lessons from the Diabetes Control and Complications Trial, Diabetes, 42, 1549-1554).

Guidelines have been developed that HbA1c 7% corresponds to a mean blood glucose value of 8.3 mM (150 mg / dL), HbA1c 9% corresponds to a mean blood glucose value of 11.7 mM (210 mg / dL), and 1% an increase in HbA1c corresponds to an increase in mean blood glucose of 1.7 mM (30 mg / dL, 2). The DCCT group also suggested that since it is impractical to measure blood glucose mean value directly, you can help control your blood glucose with one simple test, namely HbA1c. However, studies clearly show that HbA1c is not sensitive to hypoglycemia.

In fact, there is no reliable factor before the immediate threat of a patient with severe hypoglycemia. The DCCT group concluded that only about 8% of future severe hypoglycemic events can be predicted from known variables such as history of hypoglycemia, low HbA1 value_C. and unawareness of hypoglycemia. One final review details the current clinical status of the problem in detail, along with the severe hypoglycemia prevention options available to patients and healthcare professionals (see Bolli, GB: How To Ameliorate The Problem of Hypoglycemia In Intensive As Well As Nonintensive Treatment Of Type I Diabetes. Diabetes Care. , 22, Appendix 2: B43-B52,1999).

Contemporary home blood glucose monitoring devices provide a means of measuring blood glucose frequently by self-monitoring blood glucose. However, the problem with self-monitoring of blood glucose is that there is no link between the data collected by blood glucose monitoring devices and HbA1c and hypoglycaemia. In other words, there are currently no reliable ways to assess HbA1c and identify impending hypoglycaemia based on blood glucose self-monitoring readings (see Bremer T and Gough DA: Is blood glucose predictable from previous values? A solicitation for data. Diabetes 48: 445-451) , 1999).

A blood glucose meter is known in the industry and is essentially any device that can function as a blood glucose determination mechanism. A blood glucose meter or mechanism, device, tool or system therefor, includes various conventional methods of collecting a blood sample (e.g. finger prick) for each test and for the determination of glucose with an instrument that reads the glucose concentration by electromechanical or colorimetric methods. Recently, various methods of determining the concentration of blood analytes without taking a blood sample have been developed. For example, US Patent No. 5,267,152 (Yang et al.) Describes a non-invasive method of measuring blood glucose using near-infrared, scattered and reflected laser radiation spectroscopy. Similar near infrared spectroscopic devices are also described in US Patent Nos. 5,086,229 (Rosenthal et al.) And 4,975,581 (Robinson et al.).

US Patent No. 5,139,023 (Stanley) describes a transdermal blood glucose monitoring device that relies on the use of a permeation enhancer (e.g., bile salt) to facilitate transdermal glucose movement with a concentration gradient determined between the interstitial fluid and the receiving medium. US Patent No. 5,036,861 (Sembrowich) describes a passive glucose monitor that picks up sweat through a piece of skin, wherein a cholinergic agent is used to stimulate sweat secretion from the sweat exocrine gland. Similar sweat collection devices are described in US Patent No. 5,076,273 (Schoendorfer) and US Patent No. 5,140,985 (Schroeder).

In addition, US Patent No. 5,279,543 (Glikfeld) describes the use of iontophoresis to non-invasively sample substances through the skin into a container on the surface of the skin. It is stated there that this sampling procedure can be combined with a glucose-specific biosensor or with glucose-specific electrodes to monitor blood glucose levels. In addition, International Publication No. WO 96/00110 (Tamada) describes an iontophoretic device for transdermal monitoring of a target substance, where an iontophoretic electrode is used to move the analyte into the container and the biosensor is used to detect the target analyte present in the container. Finally, US Patent No. 6,144,869 (Berner) describes a sampling system designed to measure the concentration of an analyte present.

In addition, the blood glucose meter may include indwelling catheters and subcutaneous sampling of tissue fluid.

The object of the present invention is to develop the missing link by proposing three different but compatible algorithms for the assessment of HbA1c and the risk of hypoglycemia from blood glucose self-monitoring data to predict the short-term and long-term risk of hypoglycemia and the long-term risk of hyperglycemia.

Inventors previously reported that one reason for the missing link between the routinely available blood glucose self-monitoring data and the assessment of HbA1c and the risk of hypoglycemia

The fact that the sophisticated data collection and clinical evaluation methods used in diabetes research are rarely supported by diabetes specific and mathematically complex statistical procedures.

In response to the need for statistical analyzes that take into account the specific distribution of blood glucose data, the inventors developed a symmetrical transformation of the blood glucose measurement scale (see Kovatchev BP, Cox DJ, Gonder-Frederick LA, and WL Clarke (1997). Symmetization of the Blood Glucose Measurement Scale and Its Applications, Diabetes Care, 20, 1655-1658), which works as follows. Blood glucose levels are measured in mg / dL in the US and mmol / L (or mM) in most other countries. These two scales are directly related to the relationship 18 mg / dL = 1 mM. The entire range of blood glucose values ​​is referred to in most publications as 1.1-33.3 mM, which is believed to encompass virtually all values ​​observed. As recommended by the DCCT (see DCCT Research Group (1993) The Effect Of Intensive Treatment of Diabetes On the Development and Progression of Long-Term Complications of Insulin-Dependent Diabetes Mellitus. New England Journal of Medicine, 329, pp. 978-986). The blood glucose target range, also known as the normal blood glucose range, for people with diabetes is 3.9-10 mM, hypoglycaemia occurs when blood glucose levels fall below 3.9 mM, and hyperglycaemia occurs when the blood glucose level falls below 3.9 mM. blood glucose rises above 10 mM. Unfortunately, this scale is numerically asymmetric - the range of hyperglycaemia (10-33.3 mM) is wider than the range of hypoglycaemia (1.1-3.9 mM), and the range of normal blood glucose (3.9-10 mM) is not is situated in the middle of this scale. The inventors corrected this asymmetry by introducing the transformation, f (BG), which is a continuous function defined in the range of blood glucose (BG) (1.1-33.3), having an analytical form with two parameters:

ί (ΒΘ, α, β) = [(ln (BG))^and - β], α, β> 0 where this function meets the assumptions:

A1: f (33.3, α, β) = α, β) i

A2: f (10.0, α, β) = -f (3.9, α, β).

The function f (.) Is then multiplied by the third scale conversion parameter to determine the minimum and maximum values ​​of the blood glucose range after transformation - and λ / Ϊ0, respectively. These values ​​are convenient because a random variable with a standard normal distribution has 99.8% of its values ​​in the range [- Vl0 i]. If the blood glucose concentration is measured in mM / L, then after the numerical solution to the assumptions A1 and A2, the parameters of the function f (BG, a, β) are α = 1.026, β = 1.861, and the scale conversion parameter is γ = 1.794. If, on the other hand, blood glucose is measured in mg / dL, the calculated parameters are α = 1.084, β = 5.381 and γ = 1.509.

Thus, when blood glucose is measured in mmol / L then the symmetrization transformation is a function of f (BG) = 1.794 [(ln (BG))^1,026 -1.861], and when the blood glucose concentration is measured in mg / dl, then the symmetrization transformation is a function of f (BG) = 1.509 [(In (BG)) 1^,084 - 5,381 ].

Based on this symmetrization transformation f (.), The inventors introduced the Low Blood Glucose Index - a new measure for assessing the risk of hypoglycemia based on blood glucose self-monitoring readings (see Cox DJ, Kovatchev BP, Julian DM, Gonder-Frederick LA, Polonsky WH , Schlundt DG, Clarke WL: Frequency of Severe Hypoglycemia In IDDM Can Be Predicted From Self-Monitoring Blood Glucose Data, Journal of Clinical Endocrinology and Metabolism, 79: 16591662, 1994 and Kovatchev BP, Cox DJ, Gonder-Frederick LA, Young-Hyman D, Schlundt D, Clarke WL. Assessment of Risk for Severe Hypoglycemia Among Adults With IDDM: Validation of the Low Blood Glucose Index, Diabetes Care 21: 1870-1875, 1998). Based on a series of blood glucose self-monitoring data, the low blood glucose index is calculated as the mean of 10-f (BG), determined for f (BG) <0 and otherwise 0. A high blood glucose index, computed symmetrically with the low blood glucose index, has also been proposed, but this index has not found practical application.

By using the low blood glucose index in the regression model, the inventors were able to account for 40% of the high hypoglycemic episode variance in the next 6 months based on the history of severe hypoglycemic episodes and blood glucose self-monitoring data, and later increase this predictability to 46% (see Kovatchev BP , Straume M, Farhi LS, Cox DJ: Estimating the Speed ​​of Blood

PL 198 154 B1

Glucose Transitions and its Relationship With Severe Hypoglycemia. Diabetes, 48: Supplement 1, A363,1999).

In addition, the inventors found some data on HbA1c and blood glucose self-monitoring (see Kovatchev BP, Cox DJ, Straume M, Farhy LS. Association of Self-monitoring Blood Glucose Profiles with Glycosylated Hemoglobin: Methods in Enzymology, Vol. 321: Numerical Computer Methods) , Part C, Michael Johnson and Ludvig Brand, Eds, Academic Press, NY; 2000).

These studies have become part of the theoretical background of the present invention. To put this theory into practice, several key theoretical components have been added, including as described below. In particular, these methods have been developed to use HbA1 assessment_C. when predicting the long-term and short-term risk of hypoglycaemia. The development of these methods was based, without limitation, on a detailed analysis of data from 867 people with diabetes, which included more than 300,000 blood glucose self-monitoring readings, severe hypoglycemic recordings, and HbA1c determinations.

The inventors have therefore sought to reduce the above-mentioned limitations inherent in conventional methods, and thus provide simple and reliable methods that are suitable both for assessing glycemic control in patients and their risk of developing hypoglycemia, so that the methods can be used in their daily living environment.

The invention relates to a method for controlling a patient's HbA1c based on blood glucose data collected over a period of time, characterized in that a weighted bias towards high blood glucose (WR) and an estimated rate of change in blood glucose (Dr) are calculated. based on the collected blood glucose data; and HbA1c is estimated using a specific mathematical formula from the calculated WR and Dr.

Preferably, WR is mathematically calculated from a series of blood glucose readings x_1f X2, ... x_nat times t-ι, t₂, ..., vol_nas:

WR = 1Σ wr (x,; 1) ni = i where:

wr (BG; b) = 10.f (BG)^b if f (EG)> 0, otherwise 0, b = 1, represents the weight parameter and the computed DR is mathematically defined as:

Dr = mean of sk + 1 - sk, where sk = 10.S (k + t |)² for k = 0, 1.. ., t_n-i.e.,

S (i.e. = f (Xj), for j = 1 ,.., N.

Preferably, the estimated HbA1c value from the monitored blood glucose data is mathematically defined as:

estimated HBA1c = 0.9008 (WR) -0.8207 (DR) +6.7489.

Preferably, certain categories of estimated HbA1 value are defined_C.with each of these categories of estimated HbA1c being a range of the estimated HbA1c value; and assigning said estimated HbA1c value to at least one of said estimated HbA1c value categories. The following categories of estimated HbA1c are defined as follows:

in Category 1, the estimated HbA1c value is less than about 7.8; in Category 2, the estimated HbA1c value is from about 7.8 to about 8.5;

in Category 3, the estimated HbA1c value is from about 8.5 to about 9.0;

in Category 4, the estimated HbA1c value is from about 9.0 to about 9.6;

in Category 5, the estimated HbA1c value is from about 9.6 to about 10.3;

in Category 6, the estimated HbA1c value is from about 10.3 to about 11.0; in Category 7, the estimated HbA1c is above about 11.0.

Predicted confidence intervals are defined for the relevant categories of estimated HbA1c values, these predicted confidence intervals being determined as follows:

category 1 corresponds to a predicted HbA1c value of less than about 8.0; category 2 corresponds to a predicted HbA1c value in the range of 8.0-8.5; category 3 corresponds to a predicted HbA1c value in the range 8.5-9.0;

Category 4 corresponds to the predicted HbA value_1c in the range of 9.0-9.5; category 5 corresponds to a predicted HbA1c value in the range 9.5-10.1; category 6 corresponds to a predicted HbA1c value in the range 10.1-11.0; and category 7 corresponds to a predicted HbA1c value of greater than about 11.0.

Moreover, the predicted confidence intervals corresponding to said value are defined

HbA1c, with each of these predicted confidence intervals representing a specific range of values ​​for HbA1c. The confidence intervals of the predicted HbA1c values ​​have a confidence level of approximately 95%.

In another aspect of the embodiment, a method of controlling a patient's HbA1c based on blood glucose data collected over a period of time is characterized by calculating a weighted bias towards high blood glucose WR and an estimated rate of change in blood glucose Dr based on said of accumulated blood glucose data, the HbA1 is assessed_C. using a specific mathematical formula based on said calculated values ​​of WR and Dr, and a specific confidence interval is provided for classifying the estimated HbA1c value. The confidence interval is from about 85% to about 95%.

In a further aspect of the invention, a method for assessing the long-term likelihood of severe hypoglycaemia in a subject from blood glucose data collected over a period of time is characterized in that a weighted bias towards low blood glucose WL and an estimated rate of decline in blood glucose are calculated. DrDn based on the collected blood glucose data, and assessing the number of future severe hypoglycemic episodes using a specific mathematical formula from said calculated WL and DrDn values. The calculated WL is mathematically determined from a series of blood glucose readings x_1f x₂, ... x_nat times t-ι, t₂,. .., vol_n as:

wl = 1Σ wl (x,; 2) ni = i where:

wl (BG; a) = 10.f (BG)^and if f (BG) <0, otherwise 0, a = 2 represents the weight parameter and the computed DR is mathematically defined as:

DrDn = average of sk + 1k - s_k, if sk <k + 1 where:

sk = 10.S (k + t1) 2 for k = 0.1 ,. .., tn-L,

S (ie) = f (Xj), for j = 1 ,. . ., n.

The estimated number of future severe hypoglycaemic episodes (EstNSH) is mathematically defined as:

EstNSH = 3.3613 (WL) - 4.3427 (DrDn) - 1.2716.

Preferably, certain categories of EstNSH are defined, each of which is a range of EstNSH values; and assigning said EstNSH number to at least one of said EstNSH categories.

The listed EstNSH categories are defined as follows:

Category 1 in which the EstNSH value is less than about 0.775; Category 2, in which the value of the EstNSH number is 0.775-3.750; Category 3 in which the value of the EstNSH number is 3.750-7.000; category 4 in which the value of the number EstNSH is greater than 7.0.

The probability of a selected number of severe hypoglycemic episodes is determined for each of the designated EstNSH categories as appropriate, whereby the listed probability and the corresponding selected number of severe hypoglycemic episodes are defined as follows:

Category 1 corresponds approximately. A 90% chance of having approximately 0 severe hypoglycaemic episodes and about a 10% chance of developing about 1 or more severe hypoglycaemic episodes within a specified time period;

Category 2 has an approximately 50% chance of experiencing approximately 0 severe hypoglycaemic episodes and approximately 25% likelihood of developing approximately 1-2 severe hypoglycaemic episodes within a specified time period;

Category 3 has approximately 25% probability of having approximately 0 severe hypoglycemic episodes, approximately 25% probability of having approximately 1-2 severe hypoglycemic episodes

With hypoglycaemia and an approximately 50% chance of experiencing more than 2 episodes of severe hypoglycaemia within a specified time period; and Category 4 has an approximately 20% probability of experiencing approximately 0-2 severe hypoglycemic episodes and about 80% probability of experiencing more than 2 severe hypoglycemic episodes within a specified time period;

Preferably, the probability of a selected number of severe hypoglycemic episodes for each of the designated EstNSH categories is defined, and at least one probability of a selected number of severe hypoglycemic episodes is determined depending on the EstNSH category to which the EstNSH value belongs.

In a further aspect of the solution, a method for assessing the long-term likelihood of severe hypoglycaemia in a patient from blood glucose data collected over a period of time is characterized by computing a weighted bias towards low blood glucose WL and an estimated rate of decline in blood glucose. DrDn based on the collected blood glucose data, the number of future severe hypoglycemic episodes is assessed using a specified mathematical formula from the calculated WL and DrDn values ​​listed, and the probability of a selected number of severe hypoglycemic episodes occurring against the estimated high hypoglycemic episodes listed.

In a further aspect, a method for assessing the short-term risk of severe hypoglycaemia in a patient from the blood glucose data over a specified time period is characterized in that a weighted bias towards low blood glucose WL is calculated, determined Max (wL) by calculating the maximum values ​​wl (BG; 2), the risk value is determined by determining the geometric mean WL and Max (wl) over a specified period of time, wherein said risk value is mathematically defined as:

risk value = - / WL (24) · Max (wl).

The calculated WL is determined mathematically from a series of blood glucose readings x₁, Χ2, .... x_nat times ti, t2, ..., t_n as:

ⁿ

WL = - Σ wl (x,; 2) ni = i where:

wl (BG; a) = 10.f (BG)^and if f (BG) <0 otherwise 0, a = 2 represents the weight parameter.

Preferably, a predetermined risk threshold is applied and said defined risk is compared to that predetermined risk threshold.

If the specified risk value is greater than the threshold value, then the short-term risk of a hypoglycaemic episode is considered high; and if the specified risk value is below the threshold value, the short-term risk of developing a hypoglycaemic episode is defined as low.

The short term means approximately 24 hours or 12-72 hours, or approximately 17 or 12-25.

The device for the control of diabetes in a patient from the data of the self-test according to the invention, is characterized in that the output of the glycometer is connected to a microprocessor to which, through a keyboard decoder, a keyboard, memory and a clock are connected, and to the control output of the microprocessor, through the display controller , a display is connected, and a modem is connected to the microprocessor output via the I / O port, to which, via a personal computer, a central computer is included.

These and other objects along with the advantages and features of the invention described herein will become more apparent from the description, drawings, and claims.

The above and other objects, features and advantages of the present invention, as well as the invention itself, will be more fully understood from the following description of the preferred embodiments.

1 is a flowchart of a method for calculating the confidence intervals of an estimated HbA._1cand predicted HbAi_c in accordance with the present invention, Fig. 2 is a flowchart of a method for calculating an estimated number of future severe hypoglycemic episodes and their associated

According to the present invention, Fig. 3 is a flowchart of a method for calculating the estimated short-term risk of an impending severe hypoglycaemia according to the present invention, Fig. 4 is a graphical representation of the typical blood glucose disturbances observed before and after a severe hypoglycaemic episode, Fig. . 5 illustrates the performance of a method to predict short term severe hypoglycemia by inserting 10 week data for patient A (upper part) and patient B (lower part). Highly hypoglycemic episodes are marked with a triangle and the black line represents the risk value. When the risk threshold is exceeded, this method shows the next high risk period (gray bar), Fig. 6 is a functional block diagram of a computer system for implementing the present invention, Fig. 7-9 are block diagrams of alternative embodiments of the present invention relating to processors, telecommunications links, and systems.

The invention enables, but is not limited to, the development of accurate methods for assessing glycemic control in diabetic patients, and includes system software and software code for use in calculating the key components of the method. Methods of the Invention for HbA1 Assessment_C., the long-term likelihood of severe hypoglycaemia and the short-term risk of hypoglycaemia are also based on the wealth of data collected, as will be discussed later in this specification. Finally, aspects of these methods can be combined in a structured display or matrix.

Stationary measures of blood glucose deviation

According to the inventors' theory of blood glucose symmetrization (see Kovatchev BP, Straume M, Cox DJ, Farhi LS. Risk Analysis of Blood Glucose Data: A Quantitative Approach to Optimizing the Control of Insulin Dependent Diabetes. J of Theoretical Medicine, 3: 1-10,2001), the natural clinical center of the blood glucose measurement scale is at a blood glucose level of 112.5 mg / dL (6.25 mmol / L) - a safe value for glycemic control in diabetics.

After taking this clinical blood glucose scale, weighted deviations to the left (towards hypoglycaemia) and to the right (towards hyperglycaemia) are calculated. The weighting degree of these deviations will be represented by the parameters a and b as follows:

wl (BG; a) = 10-f (BG)^andif f (BG) <0 and otherwise 0, wr (BG; b) = 10-f (BG)^l'if f (BG)> 0 and otherwise 0 where f (BG) is the blood glucose symmetrization function described in the prior art section. The weight parameters a and b may be different or the same for the left and right deviation. Data analyzes carried out by the inventors have shown that the optimal parameter values ​​in a practical application are a = 2 (parameter value used in the calculation of the low blood glucose index) and b = 1. For the range of blood glucose readings X1, X2, ... Xn, the weighted average deviations left and right from the clinical center of the blood glucose scale are defined as:

WL = - ΣΖ wl (Xp2) and WR = —z, wr (X; 1) i = i Π i = i

These two measures of blood glucose deviation are independent of the timing of blood glucose readings and are therefore stationary. To capture the dynamics of change in blood glucose, measures of the rate of change in blood glucose were introduced as described below.

Calculating risky rate of change in blood glucose

Denote the n readings of the patient's blood glucose self-monitoring at times t), t2, ... t by Χ1, X2, ... Xn_n,. These data are then transformed by calculating the values ​​of f (X1, f (X2), f (Xn) and plotting a third degree curve S (t) passing through the points (tf (X1)), feffe)). .. (tnf (Xn)). The function S (t) is then a continuous function defined over the entire interval [t1, tn] so that S (t_and) = f (X_and) for j = 1, ... n. The set of values ​​sk = 10.S (k + t1 for k = - t_1f hence interpolated values ​​are obtained at one-hour intervals.

Then consider all pairs of sk numbers with consecutive indices: C.₀= (p₀, S1), C1 = (s1, s2),

C2 = (s2, s3), ...... and let us prezz M._up bbió r wsyystcich ar r C_ksuch that Sk> S_{k +}i, and by M._on the set of all Ck pairs such that Sk <S_k+1.

Finally, let DrDn be the average of the numbers sk + 1 - Sk assuming that Ck e M_dn, and Dr is the average of the numbers Sk + rSk provided that Ck e Mup + M_dn.

PL 198 154 B1

The DrDn and Dr numbers measure the rate of change in blood glucose in the 'risk area', such as the rate of change in risk associated with a change in blood glucose. In addition, DrDn is a measure of the rate of change in blood glucose only when blood glucose is falling, that is, DrDn estimates how quickly the risk may increase as blood glucose is falling, while Dr is a measure of overall susceptibility of blood glucose to fluctuations. . In addition, it is assured that DrDn will be associated with the risk of hypoglycemia (if the patient's blood glucose can fall rapidly then the risk of hypoglycemia is greater), while Dr. will be associated with overall blood glucose stability.

Program code

The first is for blood glucose readings in mmol / L, in which case the variable is BGMM. The second is for glucose readings in mg / dL in which case the variable is BGMG.

If blood glucose is measured in mmol / L, then each blood glucose reading is first transformed as follows:

SCALE1 = (ln (BGMM)) ** 1.026 - 1.861

RISK1 = 32.185 * SCALE1 * SCALE1

If blood glucose is measured in mg / dL, then each blood glucose reading is first transformed as follows:

SCALE2 = (In (BGMG)) ** 1.08405 - 5.381

RISK2 = 22.765 * SCALE2 * SCALE2

In addition, the weighted left and right deviations are calculated as follows:

WL = OWL = 0

IF (SCALE1 1e 0.0) WL = RISK1

WR = 0

IF (SCALE1 gt 0.0) WR = sqrt (RISK1)

Assuming that blood glucose readings are spaced evenly over time or are interpolated at hourly intervals, the rate of change in blood glucose is calculated as follows:

Dr = RISK1 (BG) = RISK1 (BG-1)

DrDn = 0

IF (SCALE 1e 0.0 and Dr gt 0) DrDn = Dr

Finally, the aggregation trajectory through all the patient's blood glucose readings will give:

WL = mean (WL)

WR = mean (WH)

Dr = mean (Dr), and DrDn = mean (DrDn)

How HbA1 is assessed_C.

A preferred example of the HbA1c evaluation method 100 according to the invention is shown in Fig. 1. In a first step 102, blood glucose self-monitoring data is collected for a predetermined time. For example, blood glucose self-monitoring data is collected over 4-6 weeks with a frequency of 3-5 blood glucose measurements per day, which data are transformed using the code or formulas shown above. Different formulas can be used if blood glucose measurements are recorded as mg / dL or mmol / L. One skilled in the art will appreciate that different levels, durations and frequencies may be used. In step 104, the weighted bias towards high blood glucose (WR) and the estimated rate of change in blood glucose (Dr) are calculated using the formula / code described above. In step 106, the estimate of HbA1c from the self-check data is computed using a linear function: EstHBA1c = 0.9008 * WR-0.8207 * DR + 6.7489. It should be noted that the coefficients for this function are obtained from data from 867 people with diabetes, and further data collection may lead to an update of these coefficients. In step 108, the categories of estimated HbA1c, representing the range of values ​​of the estimated HbA1c, are defined according to Table 1.

Table 1: Defining categories based on EstHBA1c:

<td>EstHBA1c</td><td> &lt;7,8</td><td> 7,8-8,5</td><td> 8,5-9,0</td><td> 9,0-9,6</td><td> 9,6-10,3</td><td> 10,3-11,0</td><td> &gt;11,0</td>

<td>Category</td><td> 1</td><td> 2</td><td> 3</td><td> 4</td><td> 5</td><td> 6</td><td> 7</td>

PL 198 154 B1

In step 110 according to Table 2, the predicted confidence intervals for the respective HbA1c value are determined.

Table 2: Predicted 95% confidence intervals for the HbA1c classification

<td>Category</td><td> 1</td><td> 2</td><td> 3</td><td> 4</td><td> 5</td><td> 6</td><td> 7</td>

<td>HBA1c</td><td> &lt;8,0</td><td> 8,0-8,5</td><td> 8,5-9,0</td><td> 9,0-9,5</td><td> 9,5-10,1</td><td> 10,1-11,0</td><td> &gt;11,0</td>

In step 112, the estimated HbA1c value from step 106 is set into one of the categories provided in Table 1 and / or Table 2.

Empirical rationale for HbA1c assessment

The HbA1c ranges in Table 2 are based on extensive research. To justify these ranges, we analyzed SMBG and HbA1c data from 867 diabetic patients. All patients were instructed to use their memory blood glucose meters for 6 months and to measure their blood glucose two to four times a day. 5-8 HbA1 tests were conducted at the same time_C. in each patient. Data from the memory meter were downloaded electronically and saved on a computer for further analysis. This procedure resulted in a database of over 300,000 blood glucose self-monitoring readings and 4,180 HbA1c test results collected over 6 months. An analysis of variance was performed to compare HbA1c across the seven categories listed in Table 1. The five categories showed a very significant difference with F = 91 and p &lt;0.00001. In addition, the mean HbA1c value showed a significant difference for each pair of Duncan ranges with p &lt;0.01.

In addition, 95% confidence intervals were calculated for the mean HbA1c in each of these seven categories. These confidence intervals were used as the basis for calculating the HbA1c intervals shown in Table 2. Subsequent analysis of the classification strength of this method showed that the method was well protected against serious errors such as misclassification of HbA1c into Category 1, 2, or 3 based on blood glucose self-monitoring results, while the actual HbA1c value was greater than 9.5 %, or putting HbA1c into categories 5, 6, or 7, while actual HbA1c was below 9.0%.

In summary, after the initial 4-6 weeks of self-monitoring of blood glucose, a computerized method calculates an estimated range for HbA1c values ​​that can be used to track changes in a patient's glycemic control for high blood glucose.

A method of assessing the long-term likelihood of developing severe hypoglycaemia

A preferred embodiment of the inventive method 200 for estimating the long-term likelihood of severe hypoglycemia is shown in Figure 2. In a first step 202, blood glucose self-monitoring data is collected over a predetermined time period. For example, blood glucose self-monitoring data is collected over 4-6 weeks with a frequency of 3-5 blood glucose measurements per day, and is transformed by the code or formulas shown above. Different formulas should be used depending on whether the blood glucose measurements are recorded as mg / dL or mmol / L. One skilled in the art will appreciate that different levels, durations and frequencies can be used. In step 204, WL and DrDn are calculated using the formula / code shown above. In step 206, an estimated number of future severe hypoglycemic episodes is calculated using the linear function:

EstNSH = 3.3613 * WL - 4.3427 * DrDn - 1.2716.

It should be noted that the coefficients for this function are derived from data obtained from 181 people with diabetes, and that further data collection may update these coefficients. It should further be noted that this formula gives a single estimate of the number of future severe hypoglycemic episodes and that additional methods, discussed below, are creating categories with confidence intervals for extended clinical use. In step 208, the categories of estimated number of severe hypoglycemic episodes (estNSH) representing a certain range of estNSH values ​​are determined according to Table 3.

Table 3: EstNSH classification:

<td>EstNSH</td><td> &lt;0,775</td><td> 0,775-3,750</td><td> 3,750-7,000</td><td> &gt;7,000</td>

<td>Category</td><td> 1</td><td> 2</td><td> 3</td><td> 4</td>

PL 198 154 B1

In step 210, depending on the estNSH category, a probability of 0.1-2 or more than 2 severe hypoglycemic episodes in the next six months is determined as shown in Table 4.

Table 4:

Probability of 0.1-2 or more than 2 episodes of severe hypoglycaemia in the next 6 months

<td></td><td>Category</td><td>Category</td><td>Category</td><td>Category</td>

<td></td><td> 1</td><td> 2</td><td> 3</td><td> 4</td>

<td>0 SH</td><td> 90%</td><td> 50%</td><td> 25%</td><td rowspan="2"> &lt;20%</td>

<td>1-2 SH</td><td rowspan="2"> 10%</td><td> 25%</td><td> 25%</td>

<td>&gt; 2 SH</td><td> 25%</td><td> 50%</td><td> &gt;80%</td>

In step 212, the EstNSH value of step 206 is assigned to one of the categories listed in Table 3 and / or Table 4.

Empirical rationale for assessing the long-term likelihood of severe hypoglycaemia

181 Type I adults with diabetes (mean age 37 years, disease duration 18 years) used memory blood glucose meters to collect more than 34,000 blood glucose self-monitoring results in one month. Then, for the next 6 months, they kept a diary of every occurrence of severe hypoglycemia. The blood glucose self-monitoring data were mathematically transformed and a linear regression model was used to predict future severe hypoglycemia, resulting in a very significant model (F = 36.3, p &lt;0.0001) and a multiple of R 55%.

All patients were categorized into four categories using the subject's long-term prediction of severe hypoglycemia. The mean number of future severe hypoglycemic episodes in categories 1, 2, 3, and 4 were 0.3, respectively; 2.0; 5.0 and 9.75. The analysis of variance showed very significant differences between these categories, F = 19.0, p &lt;0.0001.

In summary, the linear combination of a low blood glucose index and the rate of blood glucose decline as measured in the 'risk area' provides an accurate assessment of the long-term risk of severe hypoglycaemia. As this is based on the recording of blood glucose self-monitoring results, which are automatically stored in many rebound glucometers, an effective and clinically useful indicator of patient glycemic control in the low blood glucose range has been obtained.

A method of assessing the short-term (within 24 hours) risk of hypoglycaemia

A preferred embodiment of the method 300 for assessing the short-term risk of severe hypoglycemia according to the invention is shown in Fig. 3. In a first step 302, blood glucose self-monitoring data is collected for a predetermined short period of time. For example, blood glucose self-monitoring data is collected over 24 hours with a frequency of 3-5 blood glucose measurements per day - 4 or more readings as the nominal level according to the data analysis. One skilled in the art will appreciate that different levels, periods (durations), and frequencies may be used. In step 304, WL (24) and Max (wl) are calculated from all readings collected during the previous 24 hours, with the maximum value of wl (BG; 2) being Max (wl).

In step 306, the risk value is determined by taking the geometric mean of WL and Max (wl) over 24 hours, this risk value being mathematically determined by the following formula:

Risk (24) = 7WL (24) Max (wl)

In step 308, a risk threshold value is determined. In step 310, the estimated risk value is compared to the risk threshold value. For example, if the risk threshold is set to 17, then if Risk (24)&gt; 17, based on the previous 24 hours blood glucose self-monitoring data, the result is a high risk of impending hypoglycaemia in the patient. In other words, it is a decision rule that considers the 24-hour time of collecting blood glucose self-monitoring data and judges whether there is a likelihood of

This time precedes an incoming episode of hypoglycemia. Threshold 17 is derived from a large dataset, however it should be recognized that this value may be updated after further data collection or for additional purposes.

Empirical rationale for assessing the short-term risk of hypoglycemia

Eighty-five people were recruited by advertising in newspapers, diabetes clinics, and addressing them directly. The admission criteria were as follows: 1) age 21-60 years; 2) type I diabetes mellitus for at least two years and insulin use since diagnosis; 3) at least two documented episodes of severe hypoglycaemia in the past year; and 4) the routine use of blood glucose self-monitoring devices for diabetes monitoring. Participants were instructed to use the meter 3-5 times a day and to record any severe hypoglycemic episodes, including their exact dates and times, in monthly diaries. Severe hypoglycemia was defined as severe neuroglycopenia that causes stupor or loss of consciousness and precludes self-treatment. For each patient, the study was conducted for 6-8 months and the patient's glucose meter readings were downloaded monthly and a diary of any major episodes of hypoglycemia was taken. The memory capacity of the meters was sufficient and the download of results was frequent enough so that no blood glucose data was lost. There were no changes in the subjects' diabetes management, nor was there any additional treatment during the study.

During this study, a total of 75,495 blood glucose self-monitoring readings were downloaded from the participants' glucose meters (mean 4.0 ± 1.5 per day per patient), and 399 were recorded in patient diaries (4.7 ± 6.0 per patient) episodes of severe hypoglycaemia. One important finding was that there was measurable disturbance in blood glucose levels before and after episodes of moderate or severe hypoglycemia. During the 24 hours before the episode of severe hypoglycaemia, the low blood glucose index (e.g., WL) increased (p &lt;0.001), the mean blood glucose value was lower (p = 0.001), and the blood glucose variance increased (p = 0.001). Within 24 hours after the episode of severe hypoglycaemia, the low blood glucose index and the blood glucose variance remained increased (p &lt;0.001), but the mean blood glucose value returned to its baseline.

Figure 4 is a graphical representation of the typical blood glucose disturbance observed before and after a severe hypoglycemic episode. During 48-24 hours before the episode of severe hypoglycaemia, the mean blood glucose value decreased and the blood glucose variance increased. During the 24 hours immediately before the episode of severe hypoglycaemia, the mean blood glucose value continued to decrease and the blood glucose variance continued to increase. Within 24 hours after the episode of severe hypoglycaemia, mean blood glucose levels normalized, but blood glucose variance remained strongly increased. Both the mean blood glucose concentration and the variance in blood glucose returned to their baseline within 48 hours after the episode of severe hypoglycaemia.

According to the invention, the perturbations shown in Fig. 4 are quantified on the basis of blood glucose self-monitoring data to allow an assessment of the short-term risk of developing hypoglycemia. The cut-off value Risk (24) = 17 is obtained from optimization with the following constraints: 1) this method had to predict the maximum percentage of severe hypoglycemic episodes, i.e. define as at risk the maximum percentage of the time 24 hours before the severe hypoglycemic episode, and 2) to avoid overestimation , this method had to mark as risky no more than 15% of the total study time (on average one day per week). The risk cutoff value of 17 was kept constant for all study subjects. The reason for selecting a value of 15% was to prevent patients from being irritated by being overloaded with "false alarms" and then from ignoring "true alarms". In practice, the patient's physician may select a different value depending on the severity of the patient's diabetes and the specific goals.

The following example illustrates the operation of this algorithm on blood glucose self-monitoring data in two study participants. Figure 5 shows ten week data for patient A (upper part) and patient B (lower part). Severe hypoglycaemic episodes are marked with triangles. The black curve shows the value of the risk. When the risk threshold (horizontal Risk line = 17) is exceeded, this algorithm marks the next high risk period (gray bar). For patient A, seven out of nine severe hypoglycaemic episodes are predicted, and there are five false alarms, eg, high-risk periods that did not cause severe hypoglycaemia. For patient B, there are three false alarms and only one severe hypoglycaemic episode is predicted. It is evident that the risk values ​​for patient B have more variations and are greater variations than the risk values ​​for patient A. For both patients, all severe hypoglycemia episodes were associated with supercritical risk values, and approximately half of all severe abnormalities were associated with at least one episode of severe hypoglycemia.

In all study participants, 44% of all recorded episodes of severe hypoglycaemia were preceded within 24 hours by a high-risk period, and 50% were preceded by a high-risk period within 48 hours. If only periods with at least three or at least four blood glucose self-monitoring measurements are taken into account, the accuracy of the latter prediction increases to 53% and 57% respectively. Subsequent analysis of blood glucose values ​​during or immediately following high-risk periods not followed by severe hypoglycaemic episodes, that is, during or immediately after false alarms, showed that the patient mean minimum blood glucose values ​​were 2.3 ± 0.2 mmol / L versus 5.9 ± 1.7 mmol / L (t = 19.5, p &lt;0.0001) for all risk-free periods, including all severe hypoglycaemic episodes, which remained unaccounted for. This means that although symptomatic severe hypoglycaemia did not occur, blood glucose values ​​after the high-risk periods were noticeably low.

In summary, the inventors simulated the operation of the short-term risk method on a 6-month string of blood glucose self-monitoring readings in 85 people with type I diabetes. With four or more blood glucose self-monitoring readings per day, at least 50% of all severe hypoglycemic episodes could be predicted. Even if there was no symptomatic severe hypoglycemia, the algorithm predicted moderate hypoglycemia episodes.

Integrating these three ways

The three methods of the invention, discussed above and illustrated in Figures 1-3, use the same series of blood glucose self-monitoring data. From the point of view of the blood glucose self-monitoring device, the unified display or matrix of results of these three methods may be similar to the table below:

<td></td><td></td><td colspan="7">EstHBA categories (Algorithm 1)</td>

<td></td><td></td><td> 1</td><td> 2</td><td> 3</td><td> 4</td><td> 5</td><td> 6</td><td> 7</td>

<td rowspan="4">Categories EstNSH</td><td> 1</td><td></td><td></td><td></td><td>Ss 1</td><td></td><td></td><td></td>

<td> 2</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> 3</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> 4</td><td>Ss 2</td><td></td><td></td><td></td><td></td><td></td><td></td>

For example, a patient 1 (Ss 1) result in the table above indicates that this person is likely to have HbA1_C. between 9-9.5% and has a 90% chance not to develop severe hypoglycaemia in the next six months. A patient 2 (Ss 2) result means that this person is likely to have HbA1c below 8% and has a greater than 80% chance of experiencing at least three episodes of severe hypoglycaemia in the next six months.

In addition to such a scoreboard, the short-term risk determination method provides a continuous tracking of the risk of impending hypoglycemia and can be used to sound alert when the risk becomes high.

The method of the invention may be implemented on hardware, software, or a combination thereof, and may be implemented on one or more computer systems or other processing systems, such as a personal digital assistant. In an exemplary embodiment, the invention is implemented with software running on a general purpose computer 900 as shown in FIG. 6. Computer system 600 includes one or more processors, such as processor 604. Processor 604 is coupled to telecommunications infrastructure 606 (e.g., a communications bus, link bus, or network). Computer system 600 includes an imaging interface 602 that directs graphics, text, and other data from telecommunications infrastructure 606 (or a buffer memory not shown) to be displayed on display 630.

The computer system 600 also includes a main memory 608, preferably a random access memory (RAM), and may also include a secondary memory 610. The auxiliary memory 610 may include, for example, a hard disk drive 612 and / or a removable memory drive 614, represented by

The removable memory drive 614 reads from and / or writes to the removable memory unit 618 in a known manner. The removable memory assembly 618 represents a floppy disk, magnetic tape, optical disk, etc. that is read by and written to by the removable memory drive 614. The removable memory assembly 618 includes a computer-usable storage medium on which computer software and / or data are stored.

In alternative embodiments, auxiliary memory 610 may include other means for loading computer programs or other commands into the computer system 600. Such means may include, for example, a removable memory unit 622 and an interface 620. Examples of such removable memory units / interfaces include a program cassette and a cassette interface (as found in video game devices), a removable memory device (e.g. ROM, PROM, EPROM and EEPROM) and the corresponding socket and other removable memory units 622 and interfaces 620 that allow the transfer of software and data from removable memory assembly 622 to the computer system 600.

Computer system 600 may also include a communication interface 624. This communication interface 624 allows software and data to be transferred between computer system 600 and external devices. Examples of the communication interface 624 may include a modem, a network interface (such as an Ethernet card), a communication port, a slot, and a PCMCIA card, etc. The software and data carried over the communications interface 624 take the form of signals 625, which may be electronic, electromagnetic, optical, or other signals capable of being received through the communications interface 624. The signals 628 are provided to the communications interface 624 over the communications path (i.e., channel). 626. Channel 626 carries signals 628 and may be a wire or cable, optical fiber, telephone line, cellular link, radio link, and other telecommunications channels.

As used herein, the terms "computer program medium" and "computer usable medium" are generally used to refer to means such as a removable memory drive 914, a hard disk installed in a hard disk drive 612, and signals 628. Such computer program articles are providing software to a computer system 600. The invention includes such articles comprising a computer program.

Computer programs (also called logical computer control) are stored in main memory 608 and / or auxiliary memory 610. Computer programs may also be received via communication interface 624. Such computer programs, when executed, allow the computer system 600 to perform the features of the subject matter. invention. In particular, computer programs, when executed, enable processor 604 to perform the tasks of the present invention. Such computer programs represent drivers of the computer system 600.

In an exemplary embodiment where the invention is implemented using software, the software may be stored in a computer program product and loaded onto the computer system 600 via removable memory drive 614, hard disk 612, or communication interface 624. Logic control (software) when performed by processor 604, causes the processor 604 to perform the functions of the invention described herein.

In another embodiment, the invention is mainly implemented in software using, for example, hardware components such as application specific integrated circuits (ASICs). It will be obvious to those skilled in the art to realize the hardware device for performing the tasks described herein.

In yet another embodiment, the invention is implemented using a combination of both hardware and software.

In an embodiment of the software of the invention, the methods described above are implemented in the SPSS control language, but may be implemented in other programs, e.g., but not limited to, in the C ++ programming language.

Figure 7-9 is a block diagram of alternative embodiments of the invention. Figure 7 shows a block diagram of a system 710 that essentially includes a glucometer 728 used by patient 712 to record, inter alia, insulin doses and measured blood glucose (BG) values. Data received from meter 728 is preferably transferred via appropriate links 714 or teletransmission modem 732 to a processing station or circuit, such as personal computer 740, personal digital assistant, or cellular telephone. For example, the recorded data may be stored in the meter 728 and may be directly downloaded to the personal computer 740 via a suitable interface cable. An example is the ONE TOUCH monitoring system from LifeScan, Inc., which is compatible with IN TOUCH software which includes an interface cable for downloading data to a personal computer.

The computer or personal digital assistant 740 includes the software and hardware necessary to process, analyze, and interpret diabetes patient self-monitoring data according to defined flow sequences (described in detail above) and to generate an appropriate data interpretation output. Preferably, the results of the analysis and interpretation of patient data obtained by the computer 740 are displayed as a printout from a printer connected to the personal computer 740. Alternatively, the results of the data interpretation procedure may be directly displayed on a monitor connected to the computer 740.

Figure 8 shows a block diagram of an alternative embodiment of a diabetic patient management system in the form of a patient operated device 810 having a housing small enough for the device 810 to be hand held and carried by the patient. On the surface of the housing 810 is a blood glucose test strip guide (not shown). The strip serves to receive a blood sample from patient 812. The apparatus includes a microprocessor 822 and memory 824 connected to microprocessor 822. The microprocessor 22 is designed to execute a computer program stored in memory 824 to perform various calculations and control functions detailed above. Keyboard 816 is coupled to microprocessor 822 via a standard keyboard decoder 826. Display 814 is coupled to microprocessor 822 via display controller 830. Microprocessor 822 is connected to display controller 830 via an interface, and this display controller 830 updates and refreshes display 814 with control by microprocessor 822. Speaker 854 and clock 856 are also connected to microprocessor 822. Speaker 854 is controlled by the microprocessor and emits acoustic signals. , alerting the patient to the possibility of a future hypoglycemia. Clock 856 gives the current date and time to the microprocessor 822.

Memory 824 also stores patient 812 blood glucose values, insulin dose values, insulin types, and parameter values ​​used by the microprocessor 822 to calculate future blood glucose values, additional insulin doses, and carbohydrate supplements. Each blood glucose value and insulin dose value is stored in the memory 824 with the date and time. Memory 824 is preferably a non-volatile memory such as an electrically erasable permanent memory (EEPROM).

The device 810 also includes a glucometer 828 attached to the microprocessor. This 828 glucometer is designed to measure blood samples obtained on blood glucose test strips and to determine blood glucose values ​​from blood sample measurements. As mentioned previously, such glucometers are known. The meter 828 is preferably of the type that produces digital values ​​that are fed directly to the microprocessor 822. Alternatively, the 828 meter may be of a type that produces analog values. In this alternate embodiment, meter 828 is coupled to microprocessor 822 through an analog-to-digital converter (not shown).

Device 810 further includes an I / O port 834, preferably a serial port, which is connected to microprocessor 822. Port 834 is connected to modem 832 via an interface, preferably a standard RS232 interface. Modem 832 is provided to provide communication between device 810 and personal computer 840 or health care professional computer 838 via telecommunications network 836. Specific techniques for connecting electronic devices through wires are known. Another alternative example is the Bluetooth connectivity technique.

Alternatively, Figure 9 is a block diagram of an alternate embodiment with a diabetic patient management system that is a patient operated device 910, similar to that shown in Figure 8, having a housing preferably miniature enough for device 910 to be hand held and carried. by the patient. However, this embodiment includes a separate or detachable blood glucose meter or blood glucose determination mechanism 928.

The embodiments described herein are applicable to data transmission networks such as the Internet to make estimates, estimates, and information available to any processor or computer at any remote location as shown in Figs. 6-9 and / or in US Patent No. 5,851,186 (Wood). Alternatively, patients located in remote locations may have data up to16

Blood glucose sent to a central healthcare facility or other remote site.

Overall, the invention proposes a computerized data analysis method and system for the simultaneous evaluation of two of the most important components of glycemic control in people with diabetes: HbA_1c and the risk of hypoglycemia. Among other things, this method provides three sets of output signals by using only routine blood glucose self-monitoring data.

Potential use of the method, system and computer program product of the invention provides, but is not limited to, the following advantages. First, the invention extends existing home blood glucose monitoring devices by creating and displaying: 1) HbA categories assessed_1c, 2) the assessed likelihood of severe hypoglycaemia in the next six months, and 3) the assessed short-term risk of hypoglycaemia (i.e., within the next 24 hours). The latter may include warnings, such as an alarm, that signals an upcoming episode of hypoglycaemia. These three components can also be integrated to provide continuous information on glycemic control in people with diabetes and to improve monitoring of their risk of hypoglycemia.

A second advantage is that the invention extends existing software or hardware for reproducing blood glucose self-monitoring data. Such software or hardware is produced by virtually every home blood glucose monitoring device manufacturer and is typically used by patients and healthcare professionals to interpret blood glucose self-monitoring data. The methods and system of the invention may be directly incorporated into existing home blood glucose monitors, or used to extend software that recreates blood glucose self-monitoring data by introducing a data interpretation component capable of predicting both HbA1_C. as well as periods of increased risk of hypoglycaemia.

Yet another advantage is that the invention allows the accuracy of home blood glucose monitoring devices to be assessed for both low and high blood glucose levels and the entire blood glucose scale.

Moreover, another advantage is that the invention enables the effectiveness of various diabetes treatment systems.

Since diabetics throughout their lives face the problem of optimizing the maintenance of tight glycemic control without increasing their risk of hypoglycemia, the present invention alleviates this problem by using its simple and reliable methods, i.e. the invention is able to evaluate both patient and glycemic control. their risk of hypoglycaemia, and at the same time allows it to be used in the everyday environment of diabetics.

Additionally, the invention creates the missing link by proposing three different but compatible algorithms for the assessment of HbA1c and the risk of hypoglycaemia based on blood glucose self-monitoring data, to use them to predict the short-term and long-term risks of hypoglycaemia and the long-term risk of hyperglycaemia.

Finally, another advantage is that the invention makes it possible to evaluate the effectiveness of new insulin or insulin delivery devices. Any manufacturer or researcher of insulin or insulin delivery devices can use the practice of the invention to validate the relative success of proposed or tested types of insulin or insulin delivery devices.

The invention may be applied in any other specific form without departing from the spirit and its essential characteristics. The foregoing embodiments are therefore intended to be illustrative in all respects rather than as a limitation of the invention described herein. The scope of the invention is thus set forth in the appended claims and not in the above description, and any changes that occur to the meaning and scope of equivalence of the claims are intended to be included in the invention.

Designation list

102 collecting data on glucose concentration in qv and over a specified period

104 calculating WR (weighted for hypergicemia) and Dr (estimating the rate of change in blood glucose) from blood glucose data

106 low-low HbA values_1c uuncej and WR and Dr

By classifying the estimated HbA1 value_C. in the appropriate categories define the predicted confidence intervals for the corresponding HbA1c values ​​assign HbA1c to at least one HbA1c confidence interval to obtain a patient's HbA1c assessment collect blood glucose data over a specified time calculation of WL (weighted deviation towards low blood glucose (hypoglycemia)) and DrDn (estimated rate of blood glucose decline) based on blood glucose data estimated future number episodes of severe hypoglycemia (EstNSH) classifying EstNSH into specific categories defining the probability of severe hypoglycemia episodes in relation to the EstNSH category assigning the probability of future severe hypoglycemia episodes of at least one category of upcoming severe hypoglycemia episodes, To obtain an assessment of the patient's long-term likelihood of an impending hypoglycaemia Collect blood glucose data over a specified period of time Calculate WL and determine Max (wL) Determine risk value Provide a specified risk threshold value Compare a specified risk value with a specified risk threshold value to receive the short-term risk of an impending episode of severe hypoglycaemia in a patient display interface processor telecommunications infrastructure memory main memory auxiliary hard disk drive removable memory drive unit interface removable memory unit communication interface communication path display patient glucometer modem computer computer personal computer or personal assistant digital patient display keyboard microprocessor memory decoder keypads glucometer display controller modem I / O port computer central computer personal computer or personal assistant digital clock patient display keyboard

PL 198 154 B1

922 microprocessor

924 memory

926 keyboard decoder

928 glucometer

930 display driver

932 modem

934 I / O port

938 kamuste reesttalyy

940 kamuste rssobistylukssobistksysstasydigisty

956 bricks