Glucose regulation in the ICU using a computerized algorithm

1
Glucose regulation in the ICUusing a
computerized algorithm

• R.J. Bosman

2
overview

• why more strict glucose regulation
• results
• future developments

3
(No Transcript)
4
1995

• Hyperglycaemia is beneficial by promoting
  cellular glucose uptake.
• A glucose concentration of 9-11 mmol/L
  recommended and acceptable..

Mizock BA. Alterations in carbohydrate
metabolism during stress a review of the
literature Am J Med. 1995 Jan98(1)75-84
5
2001
Mizock BA. Alterations in fuel metabolism in
critical illness hyperglycaemia. Best Pract Res
Clin Endocrinol Metab 200115(4)533-51.
6
improved outcome?
7
G vd Berghe et al NEJM 2001 345 1359-67
8
Van den Berghe - 2001

• Strict control of blood glucose in critically ill
  patients (4,4 - 6,1 mmol/L) confirmed a number of
  benefits
• lower mortality
• reduced ICU stay
• lower incidents of bacteraemia

9
mechanisms of hyperglycaemia
acute phase of critical illness
hyperglycaemia
10
glucose transporters
insulin independent insulin independent insuli
n independent insulin dependent
Glut 1 Glut 2 Glut 3 Glut 4

11
mechanisms of injury / protection

• endothelium / NO pathways?

12

-
-




-
-


-






-
-


Langouche et al.J Clin Invest. 2005 August 1
115(8) 22772286. doi 10.1172/JCI25385.
13
mechanisms of injury / protection
Langouche L et al J Clin Invest. 2005 August 1
115(8) 22772286. doi 10.1172/JCI25385.
14
mechanisms of injury / protection

• endothelium / NO pathways
• reduced defense mechanisms
• polymorphonuclear neutrophil dysfunction
• decreased intracellular bactericidal activity
• decreased opsonic activity

15
mechanisms of injury / protection
NADPH
NADP
A Perner et al ICM 2003 29 642-5
16
mechanisms of injury / protection
NADPH
NADP
6PGluconate
G6PDH ?
Glucose ?
G6P
NADP
A Perner et al ICM 2003 29 642-5
17
mechanisms of injury / protection
?
NADPH ?
NADP
6PGluconate
G6PDH ?
Glucose ?
G6P
NADP
A Perner et al ICM 2003 29 642-5
18
mechanisms of injury / protection

• endothelium / NO pathways
• reduced defense mechanisms
• anti-inflammatory effects
• through NF- k? regulated pathways
• lower CRP level in intensive insulin group

19
mechanisms of injury / protection

• endothelium / NO pathways
• reduced defense mechanisms
• anti-inflammatory effects
• insulin itself
• anabolic effect
• through PI3K pathway ? cell proliferation
• anti-apoptotic effect

20
mechanisms of injury / protection

• endothelium / NO pathways
• reduced defense mechanisms
• anti-inflammatory effects
• anti-apoptotic effects
• direct insulin effect

21
more strict regulation required
22
more strict regulation required

• feasibility?
• May 2002 POC glucose measurement on IC
• 2x Accucheck
• all 120 nurses trained
• renewed interest for regulation

23
several attempts.
24
several attempts.
25
several attempts.
26
free approach failed

• glucose must come down
• free approach failed
• ? guideline
• more frequent measurements
• result directly available
• short treatment loops

27
increased the POC capacity
28
guideline?
guideline requirements

• glucose in target range (4-7 mmol/L)
• reduce hypoglycaemias
• no fixed interval of measurements
• clinical acceptance

29
it aint simple
guideline?
30
guideline?
it aint simple
31
guideline?
guideline
32
study

• two main topics
• adherence to the guideline
• glucose regulation

33
guidelines
why adherence?

• reduce large variations and inequities in
  clinical care
• contribute to improved outcome
• low adherence

34
low adherence to guidelines
why adherence?

• knowledge barriers
• attitude barriers
• behavioral barriers

35
computerizing guidelines
why adherence?

• guideline content / algorithm
• human and organizational factors
• technical factors

36
study

• two main topics
• adherence to the guideline
• glucose regulation

37
glucose regulation
glucose regulation

• defined 5 ranges
• too low lt 2.5 mmol/l
• low 2.5 4.0 mmol/l
• normal 4.0 7.0 mmol/l
• high 7.0 8.5 mmol/l
• too high gt 8.5 mmol/l

38
glucose regulation
glucose regulation

• not number of measurements
• time in each range

39
glucose regulation
not number of measurements, but time in range

• defined 5 ranges
• too low lt 2.5 mmol/l
• low 2.5 4.1 mmol/l
• normal 4.2 7.0 mmol/l
• high 7.1 8.5 mmol/l
• too high gt 8.6 mmol/l

40
why time?
too high gt 8.5 high 7.0 - 8.5 normal 4.0
7.0 low 2.5- 4.0 too low lt 2.5
glucose regulation
5 glucose measurements
high 2 3 minute 6/1440 0.4 too high 54
minutes 54/1440 3.75
Le Floch JP, Escuyer P, Baudon D, Perlemuter L.
Blood glucose area under the curve.
Methodological aspects. Diabetes
Care. 1990131725
41
clinical setting

• 18 beds - closed format ICU
• non-academic teaching hospital
• 1900 admissions per year
• 1100 cardio-thoracic surgery
• information processing fully computerized since
  April 2001 with MetaVision (iMD-Soft)

42
research design

• before off on - off design
• before guideline implementation
• paper implementation (off)
• intervention (on)
• post-intervention only paper (off)
• all patients admitted to ICU are eligible
• randomized based on PatientID

43
research design

• guideline activation
• expected LOS gt 24 hours
• diabetes
• exclusion criteria
• induced hypothermia
• Glucose/Insulin/Potassium infusions
• ingestion of normal meals

44
research design

• primary end points
• improve adherence to the guideline
• improve time in normal range (4 7 mmol/L)
• reduce number of hypoglycaemic incidents

45
intervention

• computerized algorithm
• Decision Support Software
• (Event manager MetaVision iMD-Soft)
• timing of glucose measurements
• adherence to insulin advice in guideline

46
(No Transcript)
47
technical realization
client display
server Event Manager
client VB application
48
technical realization
client display
server Event Manager
client VB application
49
technical realization
client display
server Event Manager
client VB application
50
technical realization
client display
server Event Manager
client VB application
51
technical realization
client display
server Event Manager
client VB application
52
results

• guideline adherence
• timing next glucose measurement taken on time?
• insulin advice advice provided by guideline
  executed?
• glucose regulation
• time in range
• mean glucose values
• hypoglycaemic incidents

53
timing of glucose sample
results guideline adherence
too early on time too late
paper implementation 36.4 28.9 34.7
intervention group 34.3 40.2 25.5
post-intervention group 30.1 40.1 28.9
54
insulin advice followed
results guideline adherence
positive
paper implementation 56.3
intervention group 77.3
post-intervention group 42.4
55
time in range
results glucose regulation
Rood E, Bosman RJ et al JAMIA, 2005 12(2)
172-180
56
mean glucose values
results
no. of measurements mean value (median)
pre-implementation 5052 9.3 (10.8)
paper implementation 4634 7.6 (7.6)
intervention group 4823 7.0 (7.1)
post-intervention group 1154 7.5 (7.7)
57
hypoglycaemic incidents
results
of measurements no. per patient day
paper implementation 0.3 140
intervention group 0.6 118
post-intervention group 0.4 125
58
how to further improve guideline

• diabetic patients
• corticosteroids

59
diabetic patients
further improvement
mean glucose values diabetic non-diabetic
paper implementation 8.3 7.4
intervention group 7.6 6.9
post-intervention group 7.4 7.7
60
corticosteroids
further improvement
mean glucose values no steroids corticosteroids
paper implementation 7.7 7.5
intervention group 7.1 7.0
post-intervention group 8.0 7.3
61
future developments

• guideline improvement
• continuous glucose measurement
• closed loop infusion systems
• mitigate stress response of acute illness

62
study was done in 2003
further improvement

• guideline still used
• several improvements in guideline
• improvements in Event Manager

63
several attempts.
Sep/Oct2001 Nov/Dec2001 Jan/Feb 2002 Sep/Oct2002 Jun/Jul2005
no of patients admitted 322 319 322 347 356
no of glucose measurements 2770 2881 3163 3832 6873
mean glucose (median) 10.2 (9.5) 10.1 (9.5) 9.7 (9.1) 9.8 (8.9) 7.6 (6.9)
of patients treated insulin 24 37 67 69 74
no hypoglycaemic incidents 0 2 7 18 51
total amount of insulin administered (IU) 10198 16416 20974 35346 42736
all patients, irrespective of guideline activation
64
time in range
results
only patients where guideline was activated
65
conclusion

• guidelines improve glucose regulation
• computerizing and embedding the guideline in a
  CIS further improves glucose regulation and
  guideline adherence

66
(No Transcript)
67
verbetering outcome?
echter niet eenvoudig
68
(No Transcript)
69
Glucose measurement is too late
70
(No Transcript)
71
(No Transcript)