Automatic infusion system based on an adaptive patient model

• US 7,491,187 B2
• Filed: 03/24/2003
• Issued: 02/17/2009
• Est. Priority Date: 03/22/2002
• Status: Active Grant

• Current Assignee: KU Leuven Research & Development
• Original Assignee: KU Leuven Research & Development
• Inventors: Van Den Berghe, Greta, Berckmans, Daniel, De Smet, Frank, Pluymers, Bert, De Moor, Bart, Aerts, Jean-Marie
• Primary Examiner(s): Lucchesi; Nicholas D.
• Assistant Examiner(s): Stigell; Theodore J
• Application Number:
  US10/508,572
• Publication Number:
  US 20050171503A1
• Time in Patent Office: 2,157 Days
• Field of Search: 604 65- 67, 604/504, 700 28- 31, 700 44- 45, 702 22- 23, 702 85- 86, 702/104
• US Class Current: 604/66
• CPC Class Codes: A61M 2005/14208 with a programmable infusion control system, characterised by the infusion program
• CPC Class Codes: A61M 2205/6018 providing set-up signals for the apparatus configuration
• CPC Class Codes: A61M 2230/201 Glucose concentration
• CPC Class Codes: A61M 5/142 Pressure infusion, e.g. using pumps
• CPC Class Codes: A61M 5/1723 using feedback of body parameters, e.g. blood-sugar, pressure measurement of body parameters A61B5/00

Abstract

Present invention is a system of blood glucose monitoring and intensive insulin therapy in a ICU for strict maintenance of normoglycemia which reduces intensive care and hospital mortality and morbidity of critically ill adult patients. The findings of present study also reveal factors determining insulin doses needed to maintain normoglycemia as well as the impact of insulin dose versus blood glucose level on the observed outcome benefits have been established. The invention provides a control system that adapts the flow of the insulin infusion based on insulin requirement calculated by blood glucose levels and clinical parameters such as history of diabetes, Body Mass Index, blood glucose level on admission, reason of ICU admission, time in the ICU, type and severity of illness, caloric intake, obesity, drugs affecting insulin sensitivity). This automated insulin monitoring systems significantly reduces the workload and human resource management problems for intensive insulin therapy in patients in the ICU.

Claims

1. An apparatus comprising:

a sensor to produce a signal indicative of glycemia levels in a patient, said patient characterized by a patient profile,

a programmable actuator adapted to deliver a blood glucose regulator to said patient at an administration rate and

a control system adapted to regulate said administration rate of said blood glucose regulator,

wherein a model-based predictive controller based on an adaptive dynamic patient model predicts a future response of said patient'"'"'s glycemia to said administration rate of said blood glucose regulator as a function of time and compensates effects of external disturbance on said patients'"'"' glycemia by adjusting said administration rate of said blood glucose regulator to achieve a status of normoglycemia in said patient, which is a critically ill patient and

wherein said adaptive dynamic patient model comprising parameter settings corresponds to a model for which;

in an initial phase, the parameter settings are determined based on the patient profile comprising variables including on-admission parameters, which determine the class of the patient, whereby each class is associated with distinct parameters for the model; and

after the initial phase, the parameter settings are adapted by the incoming signal to more closely approximate the dynamic behaviour of the glycemia in the patient.

2. The apparatus of claim 1, wherein the controller pro-actively compensates a future effect of known current or future disturbances by using said adaptive dynamic patient model in order to minimise deviation from normoglycemia.

3. The apparatus of claim 1, wherein the control system furthermore comprises a continuous feedback of said patients'"'"' glycemia to control.

4. The apparatus of claim 1, wherein the controller is a feedback controller based on an adaptive dynamic patient model.

5. The apparatus of claim 1, wherein the external disturbance is one selected from the group of disturbances consisting of:

change in kind of medication, dose of medication, means of administration of medication, and change in caloric intake.

6. The apparatus of claim 1, wherein said adaptive dynamic patient model is based on a low-order (N)ARX-model.

7. The apparatus of claim 1, wherein said adaptive dynamic patient model is based on a state-space model.

8. The apparatus of claim 1, wherein said profile of said critically ill patient further comprises one or more of the following variables:

BMI, prior history of diabetes, reason for ICU admission.

9. The apparatus of claim 1, wherein said on-admission parameters are selected from the group consisting of:

APACHE-II on admission, on-admission glycemia, caloric intake on admission, concomitant medication on admission.

10. The apparatus of claim 1, wherein prefiltered input and output variables (i.e. instrumental variables) are used to estimate the model parameters.

11. A method for the treatment of a critically ill patient under a condition of insulin resistance or under a condition of increased glucose turnover, which comprises administering to a critically ill patient, characterized by a patient profile, a blood glucose regulator using an apparatus comprising:

a sensor to produce a signal indicative of the glycemia levels in said patient;

a programmable actuator adapted to deliver said blood glucose regulator to said patient at an administration rate; and

a control system adapted to regulate the administration level rate of said blood glucose regulator,

wherein said apparatus comprises a model-based predictive controller based on an adaptive dynamic patient model which predicts a future response of said patient'"'"'s glycemia to said administration rate of said blood glucose regulator as a function of time and compensates effects of external disturbance on said patient'"'"'s glycemia by adjusting said administration rate of said blood glucose regulator to achieve a status of normoglycemia in said critically ill patient and

wherein said adaptive dynamic patient model comprising parameter settings corresponds to a model for which;

in an initial phase, the parameter settings are determined based on the patient profile comprising variables including on-admission parameters, which determine the class of the patient, whereby each class is associated with distinct parameters for the model; and

after the initial phase, the parameter settings are adapted by the incoming signal to more closely approximate the dynamic behaviour of the glycemia in the patient.

12. The method of claim 11, wherein the blood glucose regulator is selected from insulin, an insulin analogue, an active derivative of insulin or a monomeric human insulin analogue.

13. The method of claim 11, wherein said blood glucose regulator is selected from the group consisting of:

glucagon like peptide 1 (a GLP-1), a GLP-1 analog, GLP-1 derivatives and pharmaceutically acceptable salts thereof, somatostatin, gastric inhibitor polypeptide, glucose-dependent insulinotropic peptide, bombesin, calcitonin gene-related peptide, gastrin-releasing peptide, cholinergic agonists, isoproterenol, and bethanechol.

14. The method of claim 11, wherein hyperglycemia is reduced to stable normoglycemia in a critically ill patient without causing hypoglycemia in an initial phase of less than 24 hours and normoglycemia is maintained under changing conditions of said critically ill patient (e.g., decreasing insulin resistance, other route of feeding, change in medication) or complications of said critically ill patient (e.g., concomitant infection).

15. The method of claim 11, wherein said apparatus is configured to maintain blood glucose between 60 mg/dl and 130 mg/dl in said critically ill patient.

16. The method of claim 11, wherein said apparatus is configured to maintain blood glucose between 80 and 110 mg/dl in said critically ill patient.

17. A method for the treatment of a critically ill patient, to prevent or decrease the incidence of blood stream infection, to reduce morbidity and mortality in the critical care, to prevent or decrease the incidence of prolonged inflammation, to prevent or decrease the incidence of acute renal failure, to prevent or decrease the incidence of polyneuropathy, to prevent or decrease the incidence of immune-mediated destruction of the beta cells, to prevent or decrease the incidence of systemic inflammatory response syndrome (SIRS) to prevent or decrease the incidence of sepsis, to prevent or decrease the incidence of endoneural edema, to prevent or decrease the incidence of phrenic nerves or to decrease dialysis or hemofiltration, which comprises administering a blood glucose regulator to a critically ill patient, characterized by a patient profile, using an apparatus comprising:

a sensor to produce a signal indicative of the glycemia levels in said patient;

a programmable actuator adapted to deliver a blood glucose regulator to said patient at an administration rate; and

a control system adapted to regulate the administration rate of said blood glucose regulator;

wherein said apparatus comprises a model-based predictive controller based on an adaptive dynamic patient model which predicts a future response of the patient'"'"'s glycemia to said administration rate of said blood glucose regulator as a function of time and compensates effects of external disturbance on said patient'"'"'s glycemia by adjusting said administration rate of the blood glucose regulator to achieve a status of normoglycemia in said critically ill patient and

wherein said adaptive dynamic patient model comprising parameter settings corresponds to a model for which;

in an initial phase, the parameter settings are determined based on the patient profile comprising parameters including on-admission parameters, which determine the class of the patient, whereby each class is associated with distinct parameters for the model; and

after the initial phase, the parameter settings are adapted by the incoming signal to more closely approximate the dynamic behaviour of the glycemia in the patient.