Subcutaneous Insulin in Hospitalized Patients

1
Subcutaneous Insulin in Hospitalized Patients

• Cheryl W. OMalley, MD
• Cheryl.Omalley_at_bannerhealth.com

2
Welcome Glycemic Control Experts
3
Prepare yourself for the questions

• Why do we need to worry about glycemic control,
  hasnt that been proven to harm patients?
• Cant we just use their home regimens?
• We have become pretty good with sliding scale,
  Id prefer to just use that?
• I really dont know how to dose insulin for
  someone who is naïve, can you help?
• I dont want to use these expensive regimens
  because non of our patients can afford them when
  they go home

4
Why is this important despite recent
controversies

• Affects a large number of patients
• Robust physiologic and observational data that
  some sort of control matters
• Current consensus with the ADA/AACE is to target
  moderate control
• Traditionally we have used sliding scale which
  doesnt work
• Safety issues related to insulin

5
Current Recommended Glycemic Control Targets for
ICU 140-180 mg/dL
NICE Sugar
118
lt40 70 100 130 140 160 180 200
250 299 400
2009 AACE/ADA goals
6

• For the majority of noncritically ill patients
  treated with insulin, the premeal BG target
  should generally be lt140 mg/dl in conjunction
  with random BG lt180 mg/dl.
• Modify regimen if lt 100 mg/dL to minimize risk of
  hypoglycemia

• DIABETES CARE, VOLUME 32, NUMBER 6, JUNE 2009

7
Why cant we just use their home meds?
8
Hold Most Home Meds
Metformin No Rare risk of Lactic acidosis if contrast, renal insufficiency, poor perfusion states
Thiazoladinediones Maybe Long half life, NO hypoglycemia but cant use with CHF and causes edema.
Sulfonylureas No because of Hypoglycemia
DPP-4 Inhibitors (sitagliptin) Probably not rare hypoglycemia
GLP1 antagonist (Exenatide) Probably not nausea, rare hypoglycemia (when used with sulfonylurea)
Biggest problem cant adjust quickly enough to
needs in hospital.
9
Oral Agents in the Hospital

• Sulfonylureas
• Hypoglycemia (long acting)
• Metformin
• Lactic acidosis risk renal insufficiency,
  hypotension, CHF
• GI (nausea, abd. pain, diarrhea)
• Thiazoladinediones (TZDs) or glitazones
• Possible liver toxicity
• Fluid overload, CHF
• Inability to titrate (very slow onset of action)
• Only pioglitazone approved for use with insulin

10
I get pretty good control with sliding scale
11

Does this look familiar?
12
Case Sliding Scale Only
non-fasting blood sugar upon admission was 560,
the patient had a redraw at 605 and it was 369.
Diabetes mellitus type 2, uncontrolled. Once the
patient's blood sugar is better controlled, will
change Accu-Checks to q.a.c and q.h.s. and
cover with Apidra sliding scale insulin and
Lantus if necessary
13
Sliding Scale Alone Doesnt Work

• Sliding scale prospective cohort study
• 171 patients with type 1 DM
• 40 had BGgt300
• 23 lt70
• In 80 of patients, the orders written at
  admission were never changed despite poor control

Quele et al, Arch Intern Med 1997 157 545-552
14
A look at the real world Mayo Scottsdale

• Retrospective Analysis 2,916 discharges
• Teaching hospital (200 bed metro. Phoenix)
• LOS 3 or more days non-ICU
• Mean 1st 24 hours 170 ? stay 167 mg/dL? last 24 h
  165 mg/dL
• Highest tertile (mean 218 mg/dL)
• 46 still only on sliding scale regular insulin
• only 60 increased insulin doses

Cook CB, et al. J Hosp Med. 2007.
15

• Scheduled subcutaneous administration of insulin,
  with basal, nutritional, and correction
  components, is the preferred method for achieving
  and maintaining glucose control.

• DIABETES CARE, VOLUME 32, NUMBER 6, JUNE 2009

16
Recommendations for Managing Inpatient
Hyperglycemia
Antihyperglycemic Therapy
Insulin Recommended
Oral Agents Not Generally Recommended
SC Insulin scheduled basal/prandial/correction No
n-critically ill patients
IV Insulin Critically ill ICU patients
Clement S, et al. Diabetes Care. 2004 Moghissi
ES, et al. Endocr Pract. 2009.
17
Physiologic Insulin Secretion
Normal 24-Hour Profile
1. Nutritional Insulin Promote glucose
utilization
50
Insulin (µU/mL)
25
2. Basal Insulin Suppresses Glucose Production
Between Meals And Overnight
0
Breakfast Lunch Supper
150
100
Glucose (mg/dL)
50
3. Correction/ Supplemental Insulin Additional
insulin to treat hyperglycemia
0
7
8
9
10
11
12
1
2
3
4
5
6
7
8
9
A.M.
P.M.
Time of Day
18
Maintaining Physiologic Insulin Delivery in the
Hospital
Supplemental or stress insulin
Mealtime insulin (bolus)
Basal insulin
19
Human Insulins and AnalogsTypical Times of Action
Insulin Preparation Onset of Action Peak Duration of Action
Lispro Aspart Glulisine 15 minutes 1-2 hours 4-6 hours
Human regular 30-60 minutes 2-4 hours 6-8 hours
Human NPH/lente 2-4 hours 4-10 hours 12-20 hours
Glargine 2-4 hours minimal peak 24 hours
Detemir 3-4 hours minimal peak 12-24 hours (dose dependent)
20
Pharmacokinetics of Insulin Preparations
Short acting Analog
Regular
NPH
Glargine
Detemir
Insulin Effect
8 AM
8 AM
6 PM
N
10 PM
6-23
21
Pharmacokinetic Profiles of Insulin Types
Rapid Acting (Glulisine, lispro, aspart) 3-4 hours
Regular 6 hr
NPH 12-20 hr
Relative Plasma Insulin Level
Glargine (Lantus) 20-24 hr
2
6
12
18
24
Hours
22
Limitations of Human Regular Insulin

• Slow onset of action
• Requires inconvenient administration 20-40
  minutes prior to meal
• Risk of hypoglycemia if meal is further delayed
• Mismatch with postprandial hyperglycemic peak
• Long duration of activity
• Up to 12 hours duration
• Increased at higher dosages
• Potential for late postprandial hypoglycemia

23
Using Exogenous Insulin to Imitate Physiologic
Insulin Secretion Summary

• Basal insulin Use non-peaking, longer-acting
  insulins
• Glargine or detemir are preferred
• NPH also possible
• Nutritional insulin Depends on the type of
  nutrition
• Rapid-acting insulin is preferred when patients
  are eating meals
• Regular insulin also possible
• Correctional insulin Use rapid-acting (or
  regular) insulin
• Usually the same as the nutritional insulin

24
(No Transcript)
25
RABBIT-2 Trial Basal / Bolus arm

• D/C oral antidiabetic drugs on admission
• Starting total daily dose (TDD)
• 0.4 U/kg/d x BG between 140-200 mg/dL
• 0.5 U/kg/d x BG between 201-400 mg/dL
• TDD adjusted daily /- 20 for BG gt140 or lt 70
• Half of TDD as insulin glargine and half as
  rapid-acting insulin (lispro, aspart, glulisine)
• Insulin glargine - once daily, at the same
  time/day.
• Rapid-acting insulin- three equally divided doses
  (AC)

Smiley Umpierrez, Southern Med J, June 2006
26
Mean Blood Glucose Levels During Insulin Therapy







plt0.01
plt0.05
Day 3 P0.06
Umpierrez, Diabetes Care 30 2007
27
Blood Glucose Levels in Patients Who Failed
SSRI Transition to Basal Bolus Insulin
P NS
P 0.02





Failure was defined as 3 consecutive BG values gt
240 mg/dL during SSRI
Umpierrez, Diabetes Care 30 2007
28
DEAN Trial
Detemir with Aspart vs NPH with Regular Insulin
Therapy in the Inpatient Management of Patients
With Type 2 Diabetes

• 130 nonsurgical non-critically ill patients age
  18-80 with known type 2 diabetes admitted to
  noncritical care unit
• Half of patients were receiving insulin prior to
  admission and received similar outpatient insulin
  dose in the hospital
• Randomly assigned to
• Detemir once a day with premeal Aspart 3 times a
  day
• NPH and regular twice a day before breakfast and
  dinner
• Dosing
• 0.4 units per kg/day for BG 140-200
• 0.5 units per kg /day for BG gt 200
• Distribution of insulin
• Determir group 50 given as detemir and 50 as
  aspart
• NPH group 2/3 given as NPH and 1/3 as regular

Umpierrez et al. J Clin Endocrinol Metab. 94564
2009
29
DEAN-Trial
Detemir Novolog
NPH Regular
Blood glucose (mg/dL)
Duration of Therapy (days)
Data are SEM
Basal/bolus regimen Detemir was given once daily
and Novolog before meals. NPH/regular regimen
NPH and Regular insulin were given twice daily,
2/3 A.M., 1/3 P.M.
30
Percent of Glucose Values Within Target (lt140
mg/dl)
DEAN Trial
RABBIT-2 Trial

66
48
45


38
P lt 0.01
Umpierrez et al. JCEM, in press
Umpierrez et al. Diabetes Care 30218186, 2007
31
Rate of Hypoglycemia ( patients with BG lt 60
mg/dl)
DEAN Trial
RABBIT-2 Trial
32.8
25.4


3
3
Umpierrez et al. JCEM, in press
Umpierrez et al. Diabetes Care 30218186, 2007
32
DEAN Trial Discussion

• 50 of patients were on insulin prior to
  admission
• Detimir may need to be dosed bid
• NPH/R and detimir/aspart were equivalent in this
  study.
• Choice depends on physician preference, formulary
  choice, cost, and nursing considerations.

33
So, how can I figure out my patients doses?
34
3 Steps to using basal/bolus insulin in the
hospital

1. Determine total daily insulin dose
2. Divide up to 50 basal insulin, 50 bolus
3. Adjust daily

35
Step 1 Calculate Starting total daily dose (TDD)

• IV requirements
• Home dosebe careful of I use basal sliding
  scalehow many units of all types of insulin do
  they use on any given day
• Weight based 0.2-0.6 units/kg/day
• AACE slides said 0.2-0.4
• What we do at BGSMC
• 0.3 ESRD or
• 0.4 units/kg/day lean (BMI lt25)
• 0.5 units/kg/day overweight (BMI 25-30)
• 0.6 units/kg/day obese (BMI gt30)

36
But at home they eat poorly and here we are
giving them a diabetic diet
37
Physiologic Insulin Requirementsin Health and
Illness
Relative proportion of insulin
Clement, Braithwaite, Magee et al. Diabetes Care.
200427553-591.
38
How often do patients become NPO or have poor po
intake when hospitalized?
39
Step 2 Divide into Scheduled Basal vs.
Nutritional Insulin

• 40-50 should generally be basal (glargine,
  detimir, or NPH)
• Remaining 50-60 divided evenly and given to
  cover nutritional intake
• Rapid acting (lispro, aspart, glulisine) easier
  to match with meals in hospital
• Regular insulin also an option

40
Case Hypoglycemia Why?
Home regimen Glargine 120 qhs, 60 q
am Byettaheld at admit Glimipirideheld at admit
Glargine 120
Glargine 100
Glargine 60 units
41
Total Daily Dose of Insulin Divided to Match Needs

• 50 Bolus
• Rapid Acting
• Lispro
• Aspart
• Glulisine
• Short Acting
• Regular

0

• 50 Basal
• Glargine
• Detimir
• NPH

42
Basal insulin only when NPO
100 Basal
43
Continuing Nutritional Insulin when NPO
50 Basal
50 Bolus
44
Split mixed insulin when NPO
70 Basal
30 Bolus
45
Problems if you discontinue all scheduled insulin

• Sliding scale only
• DKA
• Severe uncontrolled hyperglycemia

46
The Sweet balance in NPO Patients
50 Basal
47
(No Transcript)
48
Step 3 Adjust Doses Daily

• If some BG were lt100 mg/dL
• Reduce TDD by 20-50
• Re-divide the new TDDI to preserve the desired
  ratio
• If some were over 180 mg/dL and none less than
  100 then
• Add up ALL of the insulin given in the last 24
  hours this was the real TDDI
• Add 10 to the TDDI from the prior day
• Re-divide the new TDDI to preserve the desired
  ratio

49
Insanity Doing the same thing over and over
again and expecting different results Albert
Einstein
50
(No Transcript)
51
Case 3 Daily Adjustments

• 47 y.o. HF with DM type 2 X 13 years
• Admitted for Pyelonephritis
• HbA1c 9.4 admission BG 370
• Home regimen metformin 500 mg bid

Glargine 35 units glulisine 12 with meals
Glargine 64 units glulisine 20 with meals
52
Management of Hyperglycemia is a safety concern
with risks due to

• Numerous insulin types with varying onset/peak
  and poor staff understanding.
• Changes in food/CHO intake
• Change in clinical status or medications
• Failure to adjust daily based on BG patterns
• Prolonged use of SSI as monotherapy
• Poor coordination of BG testing with insulin
  administration and meal delivery
• Poor communication during patient transfers
• Errors in order writing and transcription

53
Management of Hyperglycemia is a safety concern
with risks due to

• RISKS
• Numerous insulin types
• Changes in food/CHO intake
• Poor coordination of BG testing with insulin
  administration and meal delivery
• Errors in order writing and transcription

• SOLUTIONS
• Order sets
• Teams
• Limiting insulin options on order sets
• Include provisions for change in po intake

54
3 (1) Steps to Using Basal/Bolus Insulin in the
Hospital

1. Determine total daily insulin dose
2. Divide up to 50 basal insulin, 50 bolus
3. Reassess daily
4. USE YOUR HOSPITAL ORDER SETS

55
Effect of Structured Insulin Orders and an
Insulin Management Algorithm UCSD
5,530 patients with DM or Hyperglycemia and gt 7
POC Glucose readings TP3TP1

• RR Uncontrolled Patient-Day
• 0.77 (0.74 - 0.80)
• RR Uncontrolled Patient-Stay (70 controlled vs.
  60)
• 0.73 (0.66 - 0.81)
• RR Hypoglycemic Patient-Day (prevents 208 / year)
• 0.68 (0.59 - 0.80)
• RR Hypoglycemic Patient-Stay
• 0.77 (0.64 - 0.92)

Maynard G, et al. J Hosp Med. 2009.
56
UCSD Experience of 9,314 Patient-Stays with
Uncontrolled Hyperglycemia
57
Specific Situations
58
Challenging Clinical Situations

• Patient receiving corticosteroids
• Patient receiving TPN
• Patient on enteral nutritional support
• Continuous
• Intermittent

59
Corticosteroid Therapy

• Minimal elevation of fasting glucose
• Exaggeration of postprandial hyperglycemia
• Consider
• 70 prandial insulin, 30 basal insulin in
  patients with established diabetes history
• If already on insulin, add 10-20 to TDD
• Increase correction scale
• During the taper, be PROACTIVE

60
Frequency of hyperglycemia in patients receiving
high dose steroids
gt 1 BG gt 200 mg/dl
gt 2 BG gt 200 mg/dl
90
81
75

64
56
60
52
41
30
0
All
No Hx DM
Hx DM
Donihi A et al Endocrine Practice 12358, 296
61
One Suggested Approach for Treatment of
Hyperglycemia in Patients Receiving
Glucocorticoid Therapy
Administered in AM at time of prednisone
administration Glargine preferred if
dexamethasone used or Prednisone given twice a day
Clore JN, Thurber-Hay L. Endocrine Practice
15469 2009
62
How do Steroids Differ in their Effects?Steroid
Potency and Duration of Action

• 20mg/d of prednisone 80mg/d of hydrocortisone
  16mg/d of methylprednisolone 3mg/d
  dexamethasone)

63
TPN

• TPN commonly leads to hyperglycemia in absence of
  diabetes
• 75 of patients with type 2 diabetes not
  previously treated with insulin will require
  insulin with TPN
• Strategies-no studies comparing these
• Usual Method incremental doses added to TPN
• Preferred? separate IV infusion until
  requirements are known
• Other Basal/bolus

64
Glycemic Management of the Patient Receiving TPN

• Suggested
• In patients with known type 2 diabetes, add 1
  unit for each 10 Grams of carbohydrate in the
  solution
• Initiate Correctional Insulin Scale for BG gt 140
  mg/dl
• Add 60 to 100 of previous days correctional
  insulin dose to next days TPN solution
• Consider
• Add basal long or intermediate acting insulin at
  a dose of 0.2 to 0.4 units per kg per day

65
Enteral Nutrition

• High-fat formulas (monounsaturated fats) achieve
  better metabolic control that traditional
  high-carbohydrate preparations
• Blood glucose control may be attainable with
  long-acting subcutaneous insulin preparations-
  insulin glargine (with constant nutrition)
• Previous diabetes ¾ TDD
• Insulin naïve 0.6 units/kg

66
Continuous Enteral Tube Feeds

• Basal Less than or equal to 50 of TDDI
• Long acting at bedtime or morning
• Intermediate divided equally bid or tid
• Insulin drip
• Prandial/Nutritional 50 of TDDI
• Rapid acting every 4 hours
• Regular every 6 hours
• Correction
• Rapid acting every 4 hours
• Have a plan for if TF stopped to give dextrose,
  e.g. hang D10 at same rate as TF were running.

67
Korytkowski M, Salata R, Koerbel G et al
Diabetes Care 32594, 2009
68
Treatment Algorithm For Patients Receiving
Continuous Enteral Nutrition
Patient with no prior history diabetes started on
EN
2 BG gt 130 mg/dl
BG lt 130 mg/dl x 48 hrs
Discontinue BG Monitoring
Glargine 10 units Correction Insulin q6h
2 BG gt180 mg/dl in prior 24 hours Add 25-50
Correction Insulin to Glargine Administer
regular insulin q6h
All BG lt 130 mg/dl
Continue current regimen
69
Alternative Treatment Algorithm For Patients
Receiving Continuous Enteral Nutrition
Patient with no prior history diabetes started on
EN
2 BG gt 130 mg/dl
BG lt 130 mg/dl x 48 hrs
Discontinue BG Monitoring
Initiate correction insulin q6h
All BG lt 180 mg/dl
Continue regimen
2 BG gt180 mg/dl in prior 24 hours 1 BG gt 250
mg/dl in prior 24 hours
Start Scheduled Insulin Therapy
70
Blood Glucose Data on Participants According to
Group
CI correction insulin
There were no group differences in adverse events.
Korytkowski M, Salata R, Koerbel G et al
Diabetes Care 32594, 2009
71
Summary

• 50 of eligible subjects for this study had no
  previous history of type 2 diabetes or
  hyperglycemia
• Both glargine and correction insulin (CI) (with
  the addition of NPH) were effective at achieving
  glycemic control in these patients with careful
  glucose monitoring and adjustments of the insulin
  regimen
• 13/25 patients randomized to correction insulin
  alone required NPH insulin to achieve glycemic
  control
• No severe hypoglycemia events occurred during
  this study

Korytkowski M, Salata R, Koerbel G et al
Diabetes Care 32594, 2009
72
Glycemic Management of the Patient Receiving
Enteral Nutrition

• Continuous enteral nutrition (EN)
• Basal 40-50 of TDD as long or intermediate
  acting insulin given once twice a day
• Short acting 50-60 of TDD given q6h
• Cycled enteral nutrition
• Intermediate acting insulin given together with a
  rapid or short acting insulin with start of TF
• Rapid or short acting insulin administered q4 to
  6 hours for duration of EN administration
• Correctional insulin given for BG above goal
  range
• Bolus enteral nutrition
• Rapid acting analog or short acting insulin given
  prior to each bolus

73
Night time tube feeds

• Monitor blood sugars every 4 hours for the first
  few nights with supplemental scale coverage.
• After the dosing is determined
• Give a short acting insulin at the start of the
  tube feeds to cover the first several hours along
  with NPH to cover the rest of the night.

74
Bolus Tube Feeds

• Basal 50 of TDDI
• Long acting at bedtime or morning
• Intermediate bid (50/50 or 2/3 am and 1/3 pm) or
  at bedtime
• Insulin drip
• Prandial/Nutritional 50 of TDDI
• Rapid acting with each tube feeding
• Regular before each tube feeding
• Correction
• Rapid acting every 4 hours
• Regular every 6 hours

75
Specific Clinical Situations

• Patients with insulin pumps
• Patients who use CSII pump therapy in the
  outpatient setting can continue to use these
  devices as inpatients provided that they have the
  mental and physical capacity to do so.
• Availability of hospital personnel with
  expertise in CSII therapy is recommended
• A formal Inpatient Insulin Pump Protocol reduces
• confusion and treatment variability.

76
Inpatient CSII Protocol
An insulin pump should NEVER be discontinued
without initiation of either subcutaneous or
intravenous insulin.
If the pump is discontinued for any reason,
additional insulin (either IV or subcutaneous)
MUST be given 30 minutes prior to discontinuation.
Patient is to self-manage insulin pump and nurse
is to verify and document all basal rates and
bolus doses administered.
Insulin pumps must be discontinued for an MRI. If
the pump is interrupted for more than one hour,
another insulin source needs to be ordered.
Noschese ML et al Endocrine Practice 15415 2009
77
Hypoglycemic Events in Patients Continuing or
Stopping CSII Therapy During their Hospital Stay
Pump On
Pump Off
Blood glucose mg/dl
Bailon RM et al Endocrine Practice 1525 2009
78
Can U500 Regular Insulin Be Usedin the Hospital?
General Guidelines Inpatient use of U500 insulin
is reserved for patients who use this
concentrated form of regular insulin as
outpatients and who demonstrate a similar or
greater degree of insulin resistance at time of
hospital admission. To avoid dosing errors that
have potential for hypoglycemia, many hospitals
regulate the administration of U500 insulin by
requiring one or all of the following Order
written as volume to be given using a TB
syringe All doses prepared in pharmacy Alerts
in patient room and on patient medicine
administration record
79
Challenges
Improvements in glycemic control
Education Evidence DATA Systems Order
Sets Technology
Sliding Scale
80
The HOME stretch!

• But my patient wont be able to
  afford/manage/comply/etc with glargine/ rapid
  acting as an outpatient?

81
Factors Used for Selecting Discharge Therapy for
Patients with Known Diabetes

• Control at home and admission HbA1C
• Home regimen prior to admission
• Admission reason Hypoglycemia, Acute MI, Related
  to hyperglycemia (DKA, HHS, etc.)
• Physical limitations
• New co-morbidities that may limit prior oral
  therapy
• Hypoglycemia risk factors
• Treatment goals (I.e. hospice)
• Frequency of self monitoring
• Financial

82
Using the HbA1c and prior therapy to guide
outpatient therapy decisions
83
http//www.aace.com/pub/pdf/GlycemicControlAlgorit
hmPPT.pdf
84
Initiating Insulin
HbA1c gt7-8 on 2 agents, HbA1cgt10, ketonuria or
symptoms
Start with bedtime intermediate-acting or bedtime
or morning long-acting insulin Initiate with 10
units OR 0.2 units/kg OR basal dose in the
hospital
Patient titration Every 3 days, increase by 2
units until FBG lt 110 OR Every day increase by
1 unit until FBG lt 110 Physician titration Every
week by Treat to Target values
Continue treatment until goal reached
If hypoglycemia occurs, reduce bedtime dose by 4
units or 10
85
Case
Transition to subcut glargine 36 scheduled
Apidra 12 tid AC
Admission sliding scale apidra BMI gt35
Post op Insulin gtt
Discharge
86
Case Follow Up

• HgbA1c 13.4
• Discharged on
• Metformin 500 bid, instructed to increase to 1000
  mg bid in one week if not too much gi side
  effects
• Glipizide 5 mg bid
• Glargine 30 units q hs
• Diabetes education given
• HgbA1c 6.8 2.5 months later!

87
Selecting Discharge Therapy Take Home Messages

• Good to do something but dont get too aggressive
  because the time after discharge is high risk for
  hypoglycemia
• Once A1C is gt8.5 additional oral agents are
  unlikely to achieve goals
• Insulin at bedtime with or without oral agents is
  a good initial strategy
• Tailor glycemic target to individual

88
Additional Resources for Physician Education

• American Association of Clinical Endocrinology
  Inpatient glycemic control resource center
• Johns Hopkins Consultative Medicine Essentials
  for Hospitalists
• http//www.jhcape.com/betaX/site/article.cfm?ID6
• Quantia MD What is involved in the practical
  management of blood sugars postoperatively?
• http//quantiamd.com/player/rqdjtgk?cid53
• Quantia MD What is involved in the practical
  management of insulin preoperatively?
• http//quantiamd.com/player/rumyejs?cid53

89
Questions

• Cheryl.OMalley_at_bannerhealth.com

90
References

• Van den Berghe G, et al. Intensive insulin
  therapy in the critically ill patients. N Engl J
• Med. 20013451359-67.
• Brunkhorst FM, et al. Intensive insulin therapy
  and pentastarch resuscitation in severe sepsis. N
  Engl J Med. 2008358(2)125139.
• Intensive versus Conventional Glucose Control in
  Critically Ill Patients, N Engl J med 36013
  march 26, 2009
• Moghissi ES, et al. American Association of
  Clinical Endocrinologists and American Diabetes
  Association Consensus Statement on Inpatient
  Glycemic Control DIABETES CARE, VOLUME 32, NUMBER
  6, JUNE 2009
• Cook CB, et al. Inpatient Glucose Control A
  Glycemic Survey of 126 U.S. Hospitals Journal of
  Hospital Medicine Vol 4 No 9 November/December
  2009
• Queale WS et al, Ann Int Med, 1997 157
• Becker T et al., Clinical outcomes associated
  with the use of subcutaneous insulin-by-glucose
  sliding scales to manage hyperglycemia in
  hospitalized patients with pneumonia Diabetes
  Research and Clinical Practice 78 (2007) 392397
• Umpierrez GE, et al, Randomized Study of
  Basal-Bolus Insulin Therapy in the Inpatient
  Management of Patients With Type 2 Diabetes
  (RABBIT 2 Trial), Diabetes Care 30 2007