INSULIN OPTIMIZATION SYSTEMS AND TESTING METHODS WITH ADJUSTED EXIT CRITERION ACCOUNTING FOR SYSTEM NOISE ASSOCIATED WITH BIOMARKERS

• US 20110313674A1
• Filed: 06/18/2010
• Published: 12/22/2011
• Est. Priority Date: 06/18/2010
• Status: Active Grant

• Current Assignee: Roche Diabetes Care Inc. (Roche Holding AG)
• Original Assignee: Roche Diagnostics Operations Incorporated (Roche Holding AG)
• Inventors: Duke, David L., Soni, Abhishek, Bousamra, Steven, Percival, Matthew W.
• Granted Patent: US 8,532,933 B2
• Time in Patent Office: Days
• US Class Current: 702/19
• CPC Class Codes: A61B 5/14532 for measuring glucose, e.g. by tissue impedance measurement
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass

Abstract

Embodiments of a testing method for optimizing a therapy to a diabetic patient comprise collecting at least one sampling set of biomarker data, computing a probability distribution function, a hazard function, a risk function, and a risk value for the sampling set of biomarker data wherein, wherein the probability distribution function is calculated to approximate the probability distribution of the biomarker data, the hazard function is a function which yields higher hazard values for biomarker readings in the sampling set indicative of higher risk of complications, the risk function is the product of the probability distribution function and the hazard function, and the risk value is calculated by the integral of the risk function, minimizing the risk value by adjusting the diabetic patient'"'"'s therapy, and exiting the testing method when the risk value for at least one sampling set is minimized to an optimal risk level.

Claims

1. A testing method for optimizing a therapy to a diabetic patient comprising:

collecting at least one sampling set of biomarker data, wherein each sampling set comprises one or more sampling instances recorded over a collection period and each sampling instance comprises one or more biomarker readings;

computing a probability distribution function, a hazard function, a risk function, and a risk value for the sampling set of biomarker data wherein,

the probability distribution function is calculated to approximate the probability distribution of the biomarker data,

the hazard function is a function which yields higher hazard values for biomarker readings in the sampling set indicative of higher risk of complications,

the risk function is the product of the probability distribution function and the hazard function, and

the risk value is calculated by the integral of the risk function;

minimizing the risk value by adjusting the diabetic patient'"'"'s therapy; and

exiting the testing method when the risk value for at least one sampling set is minimized to an optimal risk level.

2. The testing method of claim 1 wherein the diabetic patient'"'"'s therapy for minimizing the risk value is selected from the group of adjusting an insulin dosage parameter, adjusting diabetic patient behaviors to reduce biomarker variability, adjusting a target biomarker level, or combinations thereof.

3. The testing method of claim 2, wherein the insulin dosage parameter is selected from the group consisting of a basal dosage parameter, an insulin to carbohydrate parameter, an insulin sensitivity parameter, a meal rise parameter, a meal offset parameter, an insulin active parameter, and combinations thereof.

4. The testing method of claim 2 wherein the risk level is minimized to an optimal level when the mean is at or substantially close to the target biomarker level or range, or to an adjusted target biomarker level, and the standard deviation of the sampling set falls within a similar noise range of a prior sampling set.

5. The testing method of claim 1 wherein the optimal risk level may be set by a physician.

6. The testing method of claim 1 wherein the probability distribution function is calculated from the mean and standard deviation of the sampling set.

7. The testing method of claim 1 wherein the probability distribution function is calculated using a kernel density estimator.

8. The testing method of claim 1 wherein the hazard function is a function which yields higher hazard values for biomarker readings in the sampling set indicative of hyperglycemia or hypoglycemia and yields hazard values at or near zero at a target biomarker level or within a target biomarker range.

9. The testing method of claim 1 wherein for minimization of the risk value to the optimal level, there are no hazard values indicative of a hyperglycemic or hypoglycemic event.

10. The testing method of claim 1 wherein the collection period for the sampling set is defined as multiple sampling instances within a day, multiple sampling instances within a week, multiple sampling instances within consecutive weeks, or multiple sampling instances on consecutive days within a week.

11. The testing method of claim 1 wherein each sampling instance comprises the biomarker reading and other contextual data associated with the biomarker reading, wherein the contextual data is selected from the group consisting of the time of collection, the date of collection, the time when the last meal was consumed, stress, exercise, energy level, the time and dose of medications including insulin, the recommended amount of insulin, and combinations thereof.

12. The testing method of claim 2 wherein the insulin dosage is variable during the collection period.

13. The testing method of claim 2 wherein the insulin dosage is at a constant level during the collection period.

14. The testing method of claim 1 wherein the biomarker readings are fasting blood glucose readings.

15. The testing method of claim 1 wherein the biomarker reading includes information concerning a biomarker type selected from glucose, triglycerides, low density lipids, and high density lipids.

16. The testing method of claim 1 wherein the insulin is basal insulin.

17. The testing method of claim 2 further comprising adjusting the insulin dosage according to an insulin adjustment regimen, which is determined by comparing the mean of the current completed sampling set to the target biomarker level, the insulin adjustment regimen being defined as the number of insulin dosage adjustments required to achieve a target insulin level and the amount of each adjusted insulin dosage, wherein the target insulin level is the amount required to achieve the target biomarker level.

18. The testing method of claim 17 wherein the insulin adjustment regimen is determined by calculating the requisite insulin dosage to achieve the target biomarker level, D_target, is calculated by the equation D_target=D_k+m·

(B_target−

B_k), wherein

m is the rate of change from a first biomarker reading B<sub>k−

1</sub> to a subsequent second biomarker reading B_k based on the adjustment of insulin from a first dosage D<sub>k−

1</sub> to a subsequent second dosage D_k as defined by the equation

19. The testing method of claim 18 wherein an adjusted insulin dosage D_k+1 after adjustment by the insulin adjustment regimen is calculated by the equation D_k+1 Dk+λ

·

m·

(B_target−

B_k) wherein λ

is a regimen tuning parameter which ranges from 0 to 1 and corresponds to an aggressiveness of the insulin adjustment regimen.

20. The testing method of claim 19 wherein an aggressive insulin adjustment regimen adjusts an insulin dosage to the maximum level in a first adjustment, and includes a λ

value of 1.

21. The testing method of claim 19 wherein a less aggressive insulin adjustment regimen adjusts an insulin dosage to the maximum level incrementally over at least two dosage adjustments, and includes a λ

value of 0.5 or less.

22. The testing method of claim 18 wherein the adjusted insulin dosage cannot exceed a maximum insulin dosage set by a healthcare provider.

23. The testing method of claim 18 wherein the calculated insulin dosage to achieve the target biomarker level, D_target for the sampling set is assessed by a healthcare provider in comparison to a calculated D_target for a prior sampling set

24. The testing method of claim 15, wherein the calculated maximum dosage to achieve a target biomarker level D_target is assessed against a maximum insulin dosage set by a healthcare provider.

25. The testing method of claim 2 wherein the target biomarker level is adjusted according to a biomarker target adjustment regimen, the biomarker target adjustment regimen being defined as an adjustment or a sequence of adjustments to the target biomarker level to the level that minimizes the risk for the current amount of noise.

26. The testing method of claim 4 wherein the similar noise range is less than a 10% change from a previous amount of noise calculated.

27. The testing method of claim 1 wherein the hazard function H(B) is defined by the equation H(B)=(1.509(log(B)^1.0804−

5.381))², wherein B is a biomarker reading in the sampling set.

28. The testing method of claim 1 wherein the hazard function H(B) is defined by the equation H(B)=H_hypo(B)+H_hyper(B), wherein B is a biomarker reading in the sampling set, H_hypo(B) is the hazard function associated with hypoglycemic events, H_hyper(B) is the hazard function associated with hyperglycemic events.

29. The testing method of claim 27 wherein H_hypo(B) is

30. The testing method of claim 1 wherein the hazard function is an index which correlates a biomarker reading to a corresponding hazard value.

31. The testing method of claim 1 further comprising collecting one or more additional sampling sets of biomarker data when the risk level is not at an optimal level.

32. The testing method of claim 1 wherein the risk value J is obtained by the following equation

33. The testing method of claim 1 further comprising conducting a new testing plan after the risk is minimized to an optimal level.

34. The testing method of claim 1 further comprising meeting entry criteria before conducting the testing plan.

35. The testing method of claim 1 further comprising meeting adherence criteria throughout the collection period.

36. The testing method of claim 35 wherein the adherence criteria requires a fasting period before collection of the sampling set of biomarker data.

37. A method for guiding a diabetic patient through a testing plan directed to optimizing an administration dosage of insulin, the method utilizing a data processing system and comprising:

instructing the diabetic patient via a display unit to collect one or more sampling sets of biomarker data, wherein each sampling set comprises one or more sampling instances recorded over a collection period, each sampling instance comprises one or biomarker readings;

computing a probability distribution function, a hazard function, a risk function, and a risk value for the sampling set of biomarker data wherein,

the probability distribution function is calculated to approximate the probability distribution of the biomarker data,

the hazard function is a function which yields higher hazard values for biomarker readings in the sampling set indicative of higher risk of complications,

the risk function is the product of the probability distribution function and the hazard function, and

the risk value is calculated by the integral of the risk function;

instructing the patient to minimize the risk value by adjusting the patient'"'"'s therapy, or exit the testing method if the risk value for at least one sampling set is minimized to an optimal risk level.

38. The method of claim 37 wherein the diabetic patient'"'"'s therapy for minimizing the risk value is selected from the group consisting of adjusting the insulin dosage, adjusting diabetic patient behaviors to reduce biomarker variability, adjusting a target biomarker level, or combinations thereof.

39. The method of claim 37 wherein, for minimization of the risk value to the optimal level, there are no hazard values indicative of a hyperglycemic or hypoglycemic event.

40. The method of claim 37 wherein further comprising determining whether entry criteria for the diabetic patient to begin the testing plan has been met.

41. A device configured to guide a diabetic patient through a testing plan directed to optimizing an administration dosage of insulin, comprising:

a processor coupled to memory, wherein the memory comprises collection procedures; and

software having instructions that when executed by the processor causes the processor to;

instruct the diabetic patient to collect one or more sampling sets of biomarker data in accordance with the collection procedures, wherein each sampling set comprises one or more sampling instances recorded over a collection period, and each sampling instance comprises one or more biomarker readings; and

compute a probability distribution function, a hazard function, a risk function, and a risk value for the sampling set of biomarker data wherein,

the probability distribution function is calculated to approximate the probability distribution of the biomarker data,

the hazard function is a function which yields higher hazard values for biomarker readings in the sampling set indicative of higher risk of complications,

the risk function is the product of the probability distribution function and the hazard function, and

the risk value is calculated by the integral of the risk function;

instruct the diabetic patient to minimize the risk value by adjusting the patient'"'"'s therapy, or exit the testing method if the risk value for at least one sampling set is minimized to an optimal risk level.

42. The device of claim 41 wherein the diabetic patient'"'"'s therapy for minimizing the risk value is selected from the group consisting of adjusting the insulin dosage, adjusting diabetic patient behaviors to reduce biomarker variability, adjusting a target biomarker level, or combinations thereof.

43. The device of claim 41 wherein, for minimization of the risk value to the optimal level, there are no hazard values indicative of a hyperglycemic or hypoglycemic event.

44. The device of claim 41 wherein the collection device is a continuous glucose monitor to obtain time-resolved glucose information that is provided as biomarker data to the processor.

45. The device of claim 41 further comprising a therapy device configured to administer insulin to a diabetic patient.

46. The device of claim 41 wherein the therapy device is an insulin pen.

47. The device of claim 41 further comprising a lancet operable to pierce the skin of the diabetic patient to obtain a blood glucose biomarker.

48. The device of claim 41 further comprising a meter configured to measure one or more selected biomarkers.