Journal Club Grand Rounds

1
Journal Club Grand Rounds

• Nilda Franco, MD
• Richard Gerkin, MD
• Kristin Manley, DO
• Cheryl W. OMalley, MD

2
Revamped Journal Club

• Teach critical appraisal JAMA Users Guides
• Review 2 important articles around a common theme
• One retro landmark
• One recent
• Statistical Principle
• Review how to apply these studies to current
  practice

3
November 2001
4
What we knew

• Hyperglycemia in critically ill patients was
  common
• Hyperglycemia was associated with worse outcomes
• Physiologic rationale for improvements with
  glycemic control.
• Outpatient control benefited type 1 and type 2 DM
• Diabetes and Insulin-Glucose infusion in Acute MI
  (DIGAMI)

5
First DIGAMI Insulin Therapy Improves Outcomes
in Patients with MI

• IV insulin in AMI X 24 hours
• Then subcutaneous for 3 months
• Target glucose 126-180mg/dL
• 29 reduction in mortality- 1 year
• 28 reduction in mortality- 3.4 year
• Which component (type of insulin) was responsible
  for the reduction?

6
Intensive Insulin Therapy in Critically Ill
Patients

• Known famously as the Van Den Berghe trial
• Published in the New England Journal of Medicine
  in November 2001

7
Hypothesis of the study

• Study was to determine whether normalization of
  blood glucose levels with intensive insulin
  therapy reduces mortality and morbidity in
  critically ill patients.

8
Study design

• Prospective, randomized controlled trial at one
  university hospital in Belgium.
• Involved adults admitted to the surgical ICU
  between 2/2/2000 and 1/18/2001 who were
  mechanically ventilated.
• A total of 1548 patients were enrolled
• On admission to the ICU patients were randomly
  assigned to either intensive or conventional
  insulin therapy

9
Study design continued.

• Assignments to the treatment groups were made
  with the use of sealed envelopes
• Permuted block randomization and stratification
  according to the type of illness were used to
  balance the two groups

10
. Van den Berghe G et al. N Engl J Med
20013451359-1367
11
Study design continued.

• Pts were given IV glucose 200-300gm over the
  first 24 hours followed by parental/enteral/combin
  ed feeding
• Undiluted arterial blood was used obtain the
  whole blood glucose on admission and Q4hours
• The dose of the insulin was adjusted by a team of
  nurses and a study physician that were not
  involved in the clinical care of the patients.
• The laboratory staff, electrophysiologist, and
  pathologist involved in the patients care were
  unaware of the treatment assignments of the
  patients.

12
Study design Conventional Group

• The group had a continuous infusion of insulin
  mixed with normal saline and administered via
  pump if the glucose level exceeded 215 mg/dL.
• Whole blood glucose levels were maintained
  between 180-200 mg/dL

13
Study design Intensive Treatment

• In the group the insulin infusion was started if
  the blood glucose level exceeded 110 mg/dL.
• Whole blood glucose levels were maintained
  between 80-110 mg/dL

14
What Data Was Collected?

• Baseline demographics, information to calculate
  APACHE scores, and Therapeutic Intervention
  Scoring System scores, blood glucose levels.
• Blood cultures were obtained if there was a
  temperature of 38.5 degrees Celsius or above.
• Weekly electromyographic screening for critical
  illness polyneuropathy was performed if duration
  of ICU stay was greater than 1 week

15
Outcome Measures of the Study

• Death from any cause during ICU stay
• Number of days in ICU, prolonged stay or need for
  readmission
• Ventilatory support
• Renal replacement therapy
• Inotropic/pressor support
• Critical-illness polyneuropathy
• C-reactive protein, WBC, body temperature
• Bloodstream infection
• Use of antibiotics for more than 10 days
• Hyperbilirubinemia

16
Results of the Study

• Intensive Insulin Group
• Almost all required insulin.
• AM glucose was 103 mg/dL on average.
• 39 patients affected by hypoglycemia
• Conventional Treatment Group
• 39 required insulin therapy
• Average AM blood glucose was 153 mg/dL.
• The remainder of the group had average AM glucose
  of 140 mg/dL.

17
Results

• The trial had interim analyses of the overall
  mortality that were investigated at 3 month
  intervals. During the 4th interval it was
  determined that the conventional treatment was
  inferior and the study should be terminated
  early.

18
Results on Mortality

• 35 patients in the intensive-treatment group died
  during ICU stay as compared to 63 patients in the
  conventional group.
• Apparent Risk reduction of 42
• Greatest reduction in deaths were those due to
  multiple-organ failure and associated sepsis

19
Leuven Study Intensive Insulin Therapy in the
Surgical ICU
100 96 92 88 84 80 0
100 96 92 88 84 80 0
Intensive treatment
Intensive treatment
Conventional treatment
Survival in ICU,
In-Hospital Survival,
Conventional treatment
Mortality ? 42, Plt.04
Mortality ? 34, Plt.01
0
20
40
60
80
100
120
140
160
0
50
100
150
200
250
Days After Admission
Days After Admission
Van den Berghe G, et al. N Engl J Med. 2001.
20
Risk Factors for Increased Mortality

• APACHE II score of 9 or higher during the first
  24 hours
• Older in age
• Admission involving non-cardiac surgery
• Outside hospital referral
• In the conventional insulin group without a prior
  history of diabetes or hyperglycemia

21
Intensive Insulin Therapy in Critically Ill
Surgical Patients Morbidity and Mortality
Benefits
Mortality
Sepsis
Dialysis
Polyneuropathy
Blood Transfusion
N 1,548
34
41
44
46
Reduction()
50
van den Berghe G, et al. N Engl J Med.
200134513591367.
22
Van den Berghe Morbidity Data
30
Glucose 180-200
25
Glucose 80-110
20
of Patients
15
10
5
0
gt14d on Ventilator
gt14d in ICU
Peak Cr. gt2.5
ICU Sepsis
Hyperbilirubinemia
Antibiotics gt10d
Polyneuropathy
23
Are the Results of the Study Valid?

• Was the Assignment of the Patients to Treatment
  Randomized?
• The assignment to the two groups was randomized
  using permuted blocks.

24
Were All Patients Who Entered the Trial Properly
Accounted for and Attributed at Its Conclusion?

• Patients were accounted for but the trial was
  stopped early secondary to results showing a
  better outcome with mortality in the intensive
  insulin group.

25
Were Patients, Their Clinicians, and Study
Personnel Blind to Treatment?

• Mostly-The clinicians involved could not be
  completely blinded because the blood sugar
  monitoring basically revealed the patient group.
  To minimize the bias the physician and ICU nurses
  involved in the insulin adjustment were blinded
  to important outcome measures and did not take
  part in clinical decisions.
• Lab personnel, EMG physician, pathologist were
  all blinded to the treatment groups of the
  patients.

26
Were the Groups Similar at the Start of the Trial?

• Yes. Secondary to the randomization with permuted
  blocks-the groups had no significant difference
  between them at the beginning of the trial.

27
Aside From the Experimental Intervention, Were
the Groups Treated Equally?

• The only differences in the treatments between
  the two groups were the insulin therapy and the
  antibiotics for the patients. The patients in the
  intensive insulin group required a shorter
  duration and less frequency of use of the
  antibiotics since there was an overall lower
  significant outcome of sepsis.

28
What Were the Results?

• How Large was the Treatment Effect?
• Is significant with the risk reduction of death
  being 42 (ARR 8.0-4.6 3.4)
• Number needed to treat
• 1/ARR 1/0.034 29.4

29
How Precise Was the Estimate of Treatment Effect?

• Precision is the relative width of the confidence
  interval
• For mortality, the risk reduction was 42 (95 CI
  22-62). This is slightly wide.

30
Were all Clinically Important Outcomes Considered?

• Yes
• Mortality
• Morbidity (Table 4)
• Hypoglycemia

31
Are the Likely Treatment Benefits Worth the
Potential Harms and Costs?

• According to this study results the treatment
  benefits do outweigh the harm and costs
• 39 patients in the intensive insulin control
  group did have hypoglycemic episodes
• The study using whole arterial blood with the
  measurement of the glucose likely was protective
  of hypoglycemic events as usually whole blood
  glucose is measured 15 lower than plasma
  glucose
• The study also gave the patients 24 hours of IV
  glucose at the start of the study which probably
  helped lower the adverse outcome of severe
  hypoglycemia and death

32
Can the Results be Applied to My Patient Care?

• -Single European center
• -Largely surgical patients

33
Examples of clinical trials and other studies
demonstrating better outcomes with improved
glycemic control.
Setting BG Threshold Outcomes
General medical and surgicalUmpierrez Pomposelli gt189 gt230gt220 ? Mortality ICU LOS? Infection rates
Acute MICapes Bolk lt100.8 vs 199.8gt109.8 gt124 ? Mortality? Mortality CHF shock
Cardiac surgeryFurnary Zerr gt150 ? Mortality? Sternal wound infection
StrokeWilliams Bruno gt13090-180 ? Mortality (HR 1.87) LOS60 ? Penumbra salvage56 cm2 ? Infarct size
No diabetes Diabetes
34
Even more conditions have better outcomes with
better glycemic control

• Chemotherapy outcomes (remission, survival, and
  infection rates) Weiser et al. Cancer 2004 Mar
  15100(6)1179-85
• Reduced rates of kidney transplant rejection
  Mathew et al, Transplantation 2001 Oct
  1572(7)1321-4
• Better outcomes in a mixed med / surgical real
  world ICU setting Krinsley JS. Mayo Clin Proc.
  2004799921000.
• Better outcomes in community acquired pneumonia
• The weight of evidence supports a goal of good
  glycemic control for all inpatients!

35
American Association of Clinical
EndocrinologistsInpatient Consensus Conference
- 2003

• Conference Conclusions
• Improved glycemic control during hospitalization
  results in improved clinical outcomes
• Diabetes management during hospitalization needs
  to become a much greater priority
• Upper Limit Inpatient Glycemic Targets
• ICU 110 mg/dl (6.1 mmol/L)
• Non-critical care
• Pre-prandial 110 mg/dl (6.1 mM)
• Maximum 180 mg/dL (10 mM)

AACE Position Statement on Inpatient Diabetes and
Metabolic Control Endocrine Practice 10 (1)
77-82, 2004
36
Targets established in Dec 2003
ICU Non-ICU Preprandial Non-ICU Maximal
AACE/ACE ADA 110 mg/dL 100 mg/dL 110 mg/dL 90-130 mg/dL 180 mg/dL 180 mg/dL

• AACE Position Statement on Inpatient Diabetes and
  Metabolic Control
• Endocrine Practice 10 (1) 77-82, 2004
• Garber AJ et al. American College of
  Endocrinology Task Force on Inpatient Diabetes
  Metabolic Control. Endocr Pract. 20041077-82.
  American Diabetes Association. Diabetes Care.
  200528S4-S36.

37
History of Inpatient Glucose Control 2006

• ACE, AACE ADA plus numerous co-sponsoring
  organizations Call to Action Consensus
  Conference
• Addressing systemic implementation barriers
• Inpatient hyperglycemia a quality-of-care
  measure
• ACE/ADA Task Force on Inpatient Diabetes. Endocr
  Pract. 200612458-468.

38
Joint CommissionInpatient Diabetes Certification

• Specific staff education requirements
• Written blood glucose monitoring protocols
• Plans for treatment of hypoglycemia and
  hyperglycemia
• Data collection of incidences of hypoglycemia
• Diabetes self-care education
• Identified program champion or team

http//www.jointcommission.org/certificationprogra
ms/inpatientdiabetes accessed May 20, 2010.
39
Intensive Insulin Therapy in the Medical ICU

• Greet Van den Berghe, M.D., Ph.D., and the Leuven
  Group
• N Engl J Med, Volume 3545449-461, February 2,
  2006

• RCT of insulin infusion to goal of 80-110 mg/dL
  vs usual therapy (180-200 mg/dL).
• 1,200 patients randomized
• A priori outcome of interest patients in MICU
  for gt 3 days
• Only 17 were diabetic

40
Conclusions MICU study

• Intention to treat Intensive insulin therapy
  significantly reduced overall morbidity but not
  mortality.
• Morbidity Reduced (newly acquired kidney injury,
  ICU days, hospital days and days on the vent.)
• Predefined population analysis (ICU gt 3 d)
• In-house mortality reduced (ARR 9.5-NNT10)
• ICU mortality reduced (ARR 7.2-NNT 14) p.05
• BUT, More deaths (18.8 vs 26.8) in patients in
  ICU lt 3 days
• -Once adjusted for patients with care limited or
  withdrawn in the first 72 hours it was (37.8 vs
  33.5-p.1)
• More studies needed

41
Glucontrol Study (abstract info)

• Mixed population of ICU patients (medical,
  scheduled and trauma surgery)
• N 3500, multicenter, Europe
• Target glucose
• 80 110 mg/dl vs. 140 180 mg/dl
• Endpoint in-hospital and 28 day mortality
• Start October 2004

42
GLUCONTROL
Group A (n 550)
Group B (n 551)
P
Age, yr
65 (51-74)
65 (51 74)
0.9207
Sex ratio, M/F
352/198
338/213
0.3827
41.2 32.7 18.1 7.9
Category Medical Scheduled Surgery
Emergency Surgery Trauma
42.9 31.3 18.1 7.7
0.9437
Philippe Devos, MD Jean-Charles Preiser MD,
PhD University Hospital of Liège - Belgium
43
GLUCONTROL
300

p lt 0.0001
250
200
Blood glucose, mg/dl
147 mg/dl
150
119 mg/dl
100
50
Group A
Group B
44
GLUCONTROL
Group A (n 550) Group B (n 551) P
ICU death rate, 16.7 15.2 0.5022
Hospital death rate, 24.5 20.7 0.1452
Day 28 death rate, 19.5 16.2 0.1685
Hypoglycemia 8.6 vs
4
Median (IQR)
45

• VISEP
• Brunkhorst et al, N Engl J Med 358125-39, 2008

46
VISEP Trial

• Study Aim 600 subjects with sepsis randomized
  to conventional or intensive insulin therapy
• Methods
• Conventional Therapy CII started at BG gt 200
  mg/dl and adjusted to maintain a BG 180 - 200
  mg/dl.
• Intensive Therapy group CII started at BG gt 110
  mg/dl and adjusted to maintain BG 80 -110 mg/dl
  (Leuvens protocol)
• Primary Outcomes
• Mortality (28 days) and morbidity (sequential
  organ failure dysfunction, SOFA)
• Safety end-point hypoglycemia (BGlt40 mg/dl)

47
VISEP Trial
Mortality rate,
Data from 488 patients IIT goal 80 110
mg/dL mean BG 112 mg/dl CIT goal 180 200
mg/dL mean BG 151 mg/dl
Brunkhorst et al, N Engl J Med 358125-39, 2008
48
Infusion Protocols Variable Hypoglycemia Rates
49
Wiener, R. S. et al. JAMA 2008300933-944.
50
Association of Tight Glucose Control vs Usual
Care With Outcomes Among Critically Ill Adults
Subgroup studies RR
Hospital Mortality Very tight (lt110) Moderately tight (lt150) Overall 14 13 27 0.90 (0.77-1.08) 0.99 (0.83-1.18) 0.93 (0.85-1.03)
Septicemia Very tight (lt110) Moderately tight(lt150) Overall 4 5 9 0.80 (0.57-1.11) 0.64(0.41-1) 0.76 (0.59-.97)
New need for Dialysis Very tight (lt110) Moderately tight(lt150) Overall 4 5 9 0.95(.7-1.29) 0.98(0.59-1.61) 0.96 (0.76-1.20)

• Wiener, R. S. et al. JAMA 2008300933-944.

51
Meta-Analysis results
Subgroup studies RR
Severe Hypoglycemia Very tight Moderately tight Overall 11 4 15 5.23(4.12-6.54) 4.37(2.19-5.72) 5.13 (4.09-5.42)
Percent of patients with gt1 BG Less than 40 13.7
of patients VS 2.5 NNH8.9
52
Bringing it home Whats been happening at Good
Sam?

• IV order set
• Early 2004- multidisciplinary group
• May 2004- subcut order set
• Oct 2005- larger scale data collection
• 2006- revisions in protocol and data analysis
• June 2007- new tube feed order set
• Dec 2007- IV order set standardized and data
  collection
• Jan 2008-System-wide initiative targeting
  glycemic control (70-140 mg/dL)
  in the ICU

53
(No Transcript)
54
BGSMC Second Floor ICUs a win- win
55
How low do we need to go? Tight/intensive vs
Moderate/good

• Waiting for NICE SUGAR

56
Original Article Intensive versus Conventional
Glucose Control in Critically Ill Patients
The NICE-SUGAR Study Investigators
N Engl J Med Volume 360(13)1283-1297 March 26,
2009
57
Hypothesis of the Study

• That intensive glucose control reduces
  mortality at 90 days

58
Assessment, Randomization, and Follow-up of the
Study Patients
The NICE-SUGAR Study Investigators. N Engl J Med
20093601283-1297
59
Baseline characteristics
The NICE-SUGAR Study Investigators. N Engl J Med
20093601283-1297
60
Methods

• RCT involving medical and surgical ICUs of 42
  hospitals in New Zealand, Australia and North
  America
• Eligible patients were those expected to require
  treatment in the ICU on 3 or more consecutive
  days
• Control of blood sugar was achieved with IV
  infusion of insulin in saline

61
Methods

• The trial intervention was discontinued once a
  patient was eating or discharged from the ICU but
  it was resumed if the patient was readmitted to
  the ICU within 90 days.
• It was discontinued permanently at the time of
  death or 90 days after randomization, whichever
  occurred first.

62
Methods

• Blood samples for glucose measurement were
  obtained by arterial catheters whenever possible,
  the use of capillary samples was discouraged
• Blood glucose levels were measured with the use
  of point of care or arterial blood gas analyzer
  or laboratory analyzer in routine use at each
  center.

63
Calorie Administration, and Corticosteroid
Treatment, According to Treatment Group

• Calorie administration 891 in intensive control
  group and 872 in conventional control group
  (mostly by enteral route, but also by parenteral
  route and as IV glucose)
• 34.6 in intensive control group and 31.7 in the
  conventional control group received
  corticosteroids main indication was septic
  shock

64
NICE-SUGAR Study Design
6104 ICUa patients

• Conventional
• IV insulin if BG gt180 mg/dL
• Target 140-180 mg/dL

Intensive IV insulin if BG gt108 mg/dL Target
81-108 mg/dL
69 insulin BG 144 mg/dL
97 insulin BG 115 mg/dL
Primary Outcome 90-day mortality
aOne third were surgical patients and two thirds
were medical patients 20 had diabetes.
NICE-SUGAR Study Investigators, et al. N Engl J
Med. 2009.
65
NICE-SUGAR

• Intensive control vs conventional control
• Mortality 27.5 vs 24.9 P 0.02
• ARR27.5-24.9 2.6 ?NNH 38
• Severe hypoglycemia (BG 40 mg/dl) 6.8 vs 0.5
  Plt0.001
• No significant difference between the two
  treatment groups in the median number of days in
  the ICU or hospital

The NICE-SUGAR Study Investigators. N Engl J Med.
20093601283-1297.
66
Conclusion

• In this large, international, randomized trial,
  we found that intensive glucose control increased
  mortality among adults in the ICU a blood
  glucose target of 180 mg or less per deciliter
  resulted in lower mortality than did a target of
  81 to 108 mg per deciliter

67

• Are the test results of the study valid?

68

• Was the assignment to patients randomized?
• Yes, 6104 patients underwent randomization,
  3054 were assigned to undergo intensive control
  and 3050 to undergo conventional control. Data
  with regard to primary outcome was available for
  3010 and 3012 respectively.

69

• Were all patients who entered the trial properly
  accounted for and attributed at its conclusion?
• Yes all the patients were accounted for.
  Study treatment was discontinued prematurely on
  304/3054 in the intensive control group and
  225/3050 in the conventional control group.
  Patients were withdrawn because of patients
  request or surrogate request, physicians request,
  were switched to palliative care, had other
  reason, or were withdrawn because of serious
  adverse event.

70

• Was follow up complete?
• At the completion of the trial data on vital
  status 90 days after randomization were
  unavailable for 82 of the 6104 patients (1.4),
  44 in the intensive control group and 38 in the
  conventional control group.
• For 74 of those 82 patients the vital status data
  were missing because consent had been withheld or
  withdrawn

71

• Were patients analyzed in groups to which they
  were randomized?
• Yes, they were analyzed in either the
  intensive glucose control group or the
  conventional glucose control group.

72

• Were the groups similar at the start of the
  trial?
• Yes. According to Table 1, the groups were
  similar in the measured baseline characteristics.

73

• Aside from experimental intervention, were the
  groups treated equally?
• Yes, they were both in the same environment they
  were allowed to eat the same diet (by enteral
  route, parenteral rout and IV glucose) for a
  total of 891 calories for the intensive glucose
  control group and 871 calories for the
  conventional glucose control group.
• Both groups also received corticosteroid
  treatment.

74

• Blood sample for glucose measurement was obtained
  by means of arterial catheters in both groups
  whenever possible.
• The use of capillary sample was discouraged.
• Blood glucose level was measured via the use of
  point of care or arterial blood gas analyzers in
  both groups

75

• What were the results?
•  

76

• How large was the treatment effect?
• 90 days after randomization 27.5 in the
  intensive control group had died as compared to
  24.9 in the conventional control group,
  therefore intensive glucose control increased
  mortality among adults in the ICU. The absolute
  difference in mortality was 2.6. The odds ratio
  was 1.14

77

• How precise was the estimate of the treatment?
• Precision is defined by the tightness of the
  confidence interval of the treatment effect.
• For the odds ratio of 1.14, the 95 CI was 1.02
  to 1.28. This odds ratio is significant because
  it does not include 1.00.
• It is somewhat precise, because it says the
  true odds ratio is likely from 1.02 to 1.28.

78
NNT

• What is your number?

79
Absolute Risk (AR)

• The incidence of disease in a group (either the
  exposed or nonexposed group)
• Absolute Risk Reduction (ARR) is the disease
  incidence that can be attributed to an exposure

80
Absolute Risk (AR)

• AR Incidence in each group
• ARR Incidence in exposed group Incidence in
  unexposed group

81
Smoking Example
ARR 50/500 35/500 15/500 0.03
82
Number Needed to Treat(NNT)

• The number of patients that need to be treated to
  prevent one additional case of the disease
• NNT 1/ARR

83
Smoking Example
NNT 1/ARR 1
1 33.3
50/500 35/500 15/500
84
Toxic Example
ARR 100/1000 90/1000 10/1000
85
Number Needed to Treat(NNH)

• The number of patients that need to be treated to
  cause one additional bad outcome
• NNH 1/ARR

86
Toxic Example
NNH 1/ARR 1
1 100
100/1000 90/1000 10/1000
87
NNTWhat is a good number?

• How serious is the outcome?
• How expensive is the treatment
• How serious are side effects of the treatment?
• What is the NNH for a side effect?

88
Now What?
89
BGSMC Response to NICE SUGAR
90
Glycemic Control in the NICE-SUGAR Leuven 1
Studies
Glycemic Target Results Mean (SD) Values in Target
NICE-SUGAR 81-108 118 (25) 25
LEUVEN I 80-110 103 (19) 53
Control Target Control Results Mean (SD) Overlap Within 2 SD of interventional mean
NICE-SUGAR 140-180 145 (26) 80
LEUVEN I 180-210 153 (33) 30
91
RCT Glycemic Control Targets in Critically Ill
Patients
Intensive Target Range Mean achieved
Conventional Target Range Mean yellow bar

Van den Berghe 1
103

112
VISEP

118
Glucontrol
118
NICE SUGAR
BG lt40 70 100 130 140 160 180
200 250 299 400
92
Intensive Insulin Therapy Mortality in ICU
Patients Meta-analysis of 26 Articles
no. deaths/total n
Risk ratio

• ICU
• Mixed
• Medical
• Surgical
• All

IIT 1351/5051 253/724
77/1037 1681/6812
Control 1301/5089 270/736 110/935
1681/6760
(95 CI) 0.99 (0.87-1.12) 1.00 (0.78-1.28) 0.63
(0.44-0.91) 0.93 (0.83-1.04)
Griesdale DE, et al. CMAJ. 2009.
93
Intensive Insulin Therapy and Mortality in
Critically Ill Patients A Meta-analysis
No. Events/Total No. Patients No. Events/Total No. Patients No. Events/Total No. Patients No. Events/Total No. Patients
Study IIT Control Risk ratio (95 CI)
Van den Berghe et al 39/765 6/783 6.65 (2.83-15.62)
Henderson et al 7/32 1/35 7.66 (1.00-58.86)
Bland et al 1/5 1/5 1.00 (0.08-11.93)
Van den Berghe et al 111/595 19/605 5.94 (3.70-9.54)
Mitchell et al 5/35 0/35 11.00 (0.63-191.69)
Azevedo et al 27/168 6/169 4.53 (1.92-10.68)
De La Rosa et al 21/254 2/250 10.33 (2.45-43.61)
Devos et al 54/550 15/551 3.61(2.06-6.31)
Oksanen et al 7/39 1/51 9.15 (1.17-71.35)
Brunkhorst et al 42/247 12/290 4.11(2.2-7.63)
Iapichino et al 8/45 3/45 2.67 (0.76-9.41)
Arabi et al 76/266 8/257 9.18 (4.52-18.63)
Mackenzie et al 50/121 9/119 5.46 (2.82-10.60)
NICE-SUGAR 206/3016 15/3014 13.72 (8.15-23.12)
Overall 654/6138 98/6209 5.99 (4.47-8.03)
Hypoglycemic Events
Favors IIT Favors Control
0.1
1
10
Risk Ratio (95 CI)
Griesdale DE, et al. CMAJ. 2009.
94
Intensive Insulin Therapy Mortality in ICU
Patients Meta-analysis of 26 Articles

• IIT had no effect on overall risk of death
• IIT may benefit patients in SICU
• 6-fold increased risk of severe hypoglycemia
  among IIT patients compared with controls
• Risk for hypoglycemic events did not differ by
  type of ICU or by intensity of insulin therapy

Griesdale DE, et al. CMAJ. 2009.
95
History of Inpatient Glucose Control 2009

• Updated AACE/ADA Consensus Statement
• Identifies reasonable, achievable, and safe
  glycemic targets
• Describes protocols, procedures, and system
  improvements

Moghissi ES et al AACE/ADA Inpatient Glycemic
Control Consensus Panel. Endocr Pract.
200915353-369. http//www.aace.com/pub/pdf/guide
lines/InpatientGlycemicControlConsensusStatement.p
df.
96
Updated AACE/ADA Consensus Statement 2009
NICE Sugar
118
lt40 70 100 130 140 160 180 200
250 299 400
2009 AACE/ADA goals

• Identifies reasonable, achievable, and safe
  glycemic targets
• Describes protocols, procedures, and system
  improvements

Moghissi ES et al AACE/ADA Inpatient Glycemic
Control Consensus Panel. Endocr Pract.
200915353-369. http//www.aace.com/pub/pdf/guide
lines/InpatientGlycemicControlConsensusStatement.p
df.
97
AACE/ADA Target Glucose Level in ICU Patients-
revised after NICE SUGAR

• ICU setting
• Starting threshold of no higher than 180 mg/dL
• Once IV insulin is started, the glucose level
  should be maintained between 140 and 180 mg/dL
• Lower glucose targets (110-140 mg/dL) may be
  appropriate in selected patients  
• Targets lt110 mg/dL or gt180 mg/dL are not
  recommended

Moghissi ES, et al. Endocr Pract. 2009.
98
Recommendations for Managing Inpatient
Hyperglycemia
Antihyperglycemic Therapy
Insulin Recommended
Oral Agents Not Generally Recommended
SC Insulin Non-critically ill patients
IV Insulin Critically ill ICU patients
Clement S, et al. Diabetes Care. 2004 Moghissi
ES, et al. Endocr Pract. 2009.
99
Will the results help me to care for my patients?

• Choose a glycemic target that can be reached
  safely
• e.g., 140 mg/dL
• Avoid excessive severe hypoglycemia
• Use appropriate data monitoring tools
• Glycemic values
• Clinical and financial outcomes
• Minimize glycemic variability

Krinsley JS, et al. Crit Care. 2008.
100
Thank you

• Nilda Franco
• Kristin Manley
• Richard Gerkin

• Jonathan Frunzi
• Carl Houghton
• Peter McKellar

101
Feedback/Questions

• Please complete evaluation
• Need volunteers for the next 1-2 JC Grand Rounds
• Need suggestions for articles
• Other Questions?