2. By
DR: MOKHTAR , ABDELRAHMAN
Professor of Internal Medicine Mansoura
University
In hospital Management of
Hyperglycemia
3. Why ?
ā¢ Huge Implementation Gap
ā¢ Regulatory guidelines etc.
ā¢ Can be cost effective
ā¢ Inpatient hyperglycemia is very strongly
associated with poor outcomes
ā¢ Improved glycemic control is associated
with improved outcomes
Society of Hospital Medicine. http://www.hospitalmedicine.org/ResourceRoomRedesign/
pdf/GC_Workbook.pdf.
4. Hyperglycemia*: A Common Comorbidity
in Medical-Surgical Patients in Hospitals
64%
(1.7% mortality)
12%
(16% mortality)
26%
3% mortality
Normoglycemia
Known Diabetes
New Hyperglycemia
Umpierrez G et al, J Clin Endocrinol Metabol 87:978, 2002
n = 2,020
* Hyperglycemia: Fasting BG ļ³ 126 mg/dl
or Random BG ļ³ 200 mg/dl X 2
14. Glucose variability :
ā¢ Hermanides and associates 2010 , Published
the results of another large study in
Amsterdam that included 5728 patients
aiming at : Measuring GV over time in a large
strict glucose control-treated ICU population
across several ranges of mean glucose and
to investigate the association of GV and
mean glucose values with ICU and in-
hospital mortality.
ā¢ Variability was measured by , MAG ( Mean
absolute Glucose change per hour ) & SD.
15. Figure 1. Two fictitious patients with identical mean glucose and sd
but different patterns of variability expressed by mean absolute glucose
change (MAG).
16. ā¢ RESULT :
ā¢ In this retrospective cohort study, it has been shown
that GV, expressed as MAG, is highly associated with
ICU death in both high and low ranges of mean
glucose. In combination with a high mean glucose, GV
seems most detrimental.
ā¢ There appears to be a synergistic negative effect of
high mean glucose in combination with high GV.
ā¢ For those with persistently high mean glucose values
during admission, low GV seems protective.
17. ā¢ DISCUSSION : From a pathophysiological viewpoint, a
causal relationship can be substantiated; in vitro, varying
glucose levels have been shown to enhance cell
apoptosis.[24] In rats,
ā¢ Glycemic reperfusion after hypoglycemia caused neuronal
death,[25] and altering glucose levels were impairing
endothelial function in healthy volunteers.[26]
ā¢ Even more, tubulointerstitial cells exposed to intermittent
high glucose concentrations showed enhanced cell growth
and collagen syntheses compared with stable high glucose
concentrations.[27]
ā¢ Possibly, the adaptive cell mechanisms that are initiated in
case of constant hyperglycemia are ineffective when the
hyperglycemia is not constant but varying, explaining the
toxicity of GV.[28]
18. No hypo- or
hyperglycemia
ā¢Prevent fluid and electrolyte
abnormalities secondary to osmotic
diuresis
ā¢Improve WBC function
ā¢Improve gastric emptying
ā¢Decrease surgical complications
ā¢Earlier hospital dischange
ā¢Decreased post-MI mortality
ā¢Decreased post-CABG
morbidity and mortality
Target Glucose Levels
19. Managing Diabetes in the Hospital Presents
Different Challenges than Managing
Diabetes in the Outpatient Arena!
20. - Unstable lifestyle regimen:
Outpatient diabetic patients are generally
instructed to eat consistent amounts of
carbohydrate, take the prescribed doses of
insulin, and do regular exercise, each day.
- The hospital, however, results in a high level of
instability in these and other variables that impact
blood glucose.
21. The hospital is also associated with:
- Acute illness, āstress-relatedā hyperglycemia
- Use of medications that impact glycemic control
ā¢ COMMONLY
ASSOCIATED
ā¢ Steroids
ā¢ Catecholamines
ā¢ Tacrolimus
ā¢ Cyclosporine
ā¢ Gatifloxacin
ā¢ TPN
SIGNIFICANT but
LESS PROMINENT
ā¢ Oral contraceptive
pills
ā¢ Thiazides
ā¢ Atypical
antipsychotics
ā¢ Calcium-channel
blocking agents
ā¢ Protease inhibitors
23. Inpatient Diabetes
Goals
Who Cares
Just get patient home
Sliding Scales are fine
Avoid that scary
hypoglycemia
Inpatient Diabetes
Goals
Normal glucoses for
everyone
A high glucose means
failure
Sliding Scales are banned
Some hypoglycemia is
acceptable
In patient
Diabetes Goals
Appropriate
Glucose Control
Based on
physiology and
outcome studies
Diabetes is a
2ry Diag
24. ā¢ Oral agents can be continued in
stable patients with normal
nutritional intake, normal blood
glucose levels, and stable renal and
cardiac function. However, there
are several potential disadvantages
to using these medications in
hospital patients:
Oral Anti diabetes Agents in the Hospital
25. ā Disadvantages of most oral agents:
ā¢ Slow-acting/difficult to titrate
ā Disadvantages of insulin secretagogues (e.g.
sulfonylureas and meglitinides such as glyburide,
glypizide, repaglinide, etc.):
ā¢ Hypoglycemia if caloric intake is reduced
ā¢ Some are long-acting (hypoglycemia may be
prolonged)
ā Disadvantages of metformin:
ā¢ Lactic acidosis can occur when used in the setting of
renal dysfunction, circulatory compromise, or
hypoxemia
ā¢ Slow onset of action
ā¢ GI complications: Nausea, diarrhea
26. ā¢ _ Disadvantages of thiazoladinediones :
ā Slow onset of action (2-3 weeks)
ā Can cause fluid retention (particularly when used with
insulin), and increase risk for CHF
ā Disadvantages of alpha-glucosidase inhibitors (e.g.
acarbose, miglitol)
ā¢ Abdominal bloating and flatus
ā¢ Need pure glucose to treat hypoglycemia
ā Disadvantages of GLP-1 mimetics (e.g. exenatide)
ā¢ Newer agents without data to support use in the
hospital
ā¢ Abdominal bloating and nausea secondary to delayed
gastric emptying
30. ā¢ 65 year old twins
ā¢ Diabetes: on NPH 20 units and OHGs with
poor control, neither sees MD regularly
ā¢ Smokers
ā¢ At a āGentlemanās Clubā when both
developed chest pain. After 6 hoursā¦..
ā¢ Tom: goes to Hospital āAā
ā¢ Harry: Hospital āAā full, so Harry goes
cross town to Hospital āBā
31. Tom at Hospital āAā
ā¢ Admitted to CCU, MI confirmed
ā¢ Glucose 230 mg/dL
ā¢ No infusion started for 18 hours
ā¢ Infusion control poor, glycemic excursions
when Tom eats.
ā¢ Recurrent hypoglycemia, treated
inconsistently, especially with trips to
Radiology
ā¢ Finally controlled on infusion day 4.
32. Tom at Hospital āAā contād
ā¢ Transition to ward: Tom on sliding scale
ā¢ Recurrent hyperglycemia to 300
ā¢ Brief return to unit .
ā¢ Confusion with various insulin regimens as
Tom goes from eating to NPO several
times.
ā¢ No mention of hyperglycemia in discharge
summary
ā¢ Tom discharged on same meds as admit
ā¢ LOS 6 days, EF 35% at 1 month
33. Tom: 3 years later
ā¢ Follows up with Cardiology only .
ā¢ Glycemic control remains poor
ā¢ Recurrent CV events
ā¢ Recurrent hospitalizations
34. Harry at Hospital āBā
ā¢ Admit CCU, MI confirmed, glu 230 mg / dL
ā¢ Infusion started by protocol when glucose >
140 mg/dL x 2.
ā¢ Glycemic excursions with meals covered w/
subcutaneous RAA-I ( Rapid Acting
Analogue ) per protocol.
ā¢ Minor hypoglycemia covered routinely
ā¢ Transitioned to ward on basal / bolus
regimen, TDD of 80 units.
ā¢ A1C obtained: 10
35. Harry at Hospital āBā contād
ā¢ When Harry goes NPO for test, nurses
continue basal insulin, hold nutritional
insulin (as per protocol )
ā¢ Education on smoking cessation and DM
ā¢ Information about DM / glucose control
included in DC summary.
ā¢ Hospitalist arranges for PMC, discharge
regimen of Glargine 35 units, 10 units
RAA-I ( Rapid Acting Analogue I ) w/ meals
prescribed.
ā¢ LOS 5 days, EF at 1 month 45%
36. Harry: 3 years later
ā¢ Quits smoking
ā¢ A1c = 6.2
ā¢ Not re-hospitalized
ā¢ What does the story till ?
37. Not only the tool but how we are
using the tool .
39. Fear of hypoglycemia
ā¢ Physicians traditionally tend to look the
other way , believing that it is better to [do
no harm ] rather than risk of hypoglycemia
with more aggressive insulin therapy .
This often leads to clinical inertia.
40. Competing priorities :
ā¢ Unless the patient admitted for
hyperglycemic emergency ( DKA ,ā¦etc ) ,
the cause for admission ( procedure ,
surgery , asthma , etc..) usually occupies
the focus of care , pushing diabetes
control to the bottom of the problem list.
41. High Error rate Insulin is one of the top 5 .
ā¢ Order design :
Lacking basal insulinā¦ā¦DKA
( without basal insulin , blood glucose levels rise by about
45mg/dl/hr. )
Insulin Stacking ā¦ā¦.when SC insulin is given at
regular 4hrs or more frequent . Although it is rapid acting , it can be
active for as long as 8 hrs. Each subsequent dose has the effect of
stacking onto the previous dose , causing severe hypoglycemia
after frequent doses.
Order Implementation :
Confusing names of preparations eg , Lantus ,
lispro,,,etc.
(JCAHO Website, 2006)
46. Current Practice ā āBest Practiceā
ā¢ Dependence on non-physiologic insulin prescribing (as opposed
to insulin that mimics physiologic insulin secretion)
ā¢ Dependence on reactive strategies (e.g. sliding-scale insulin)
ā¢ Overemphasis on simplicity (particularly simplicity from the
perspective of the ordering physician)
ā¢ Overemphasis on avoidance of hypoglycemia
ā¢ Lack of standardization of insulin use in the hospital
47. Sliding Scale Insulin
ā¢ The perfect example of an insulin regimen that is NOT
considered part of the best practice of inpatient diabetes
management is
the use of sliding-scale insulin alone in
hospitalized patients.
ā¢ When using sliding scale insulin alone, insulin is
only given after metabolic control is
lost , instead of being given in an anticipatory manner.
Gregory Maynard MD, MS
48. Sliding Scale Alone Doesnāt
Work
ā¢ Sliding scale prospective cohort study
ā¢ Patients treated solely with SSI were 3X
more likely to have BG>300
ā¢ In 80% of patients, the orders written at
admission were never changed during
hospital stay despite poor control.
Quele et al, Arch Intern Med 1997: 157; 545-552
49. Sliding Scale Insulin
ā¢ āA common misconception is that a sliding
scale insulin regimen alone is sufficient for
diabetes managementā Lien, et al. Inpatients management of
Type 2 Diabetes Mellitus
ā¢ āThis autopilot approach as the sole mode
of treatment for inpatient hyperglycemia
has been strongly condemned.ā Abourizk, N.
Inpatient Diabetology
50. ā¢ Sliding scale insulin
ā This is a dirty word; we donāt use dirty words at
UCSD
ā āMindless medicine,ā āparalysis of thought,ā āaction
without benefit,ā āinsulin insanityā (Gregory
Maynard MD, MS)
ā¢ Evidence does not support this technique
without basal insulin; unacceptably high rates of
ā Hyperglycemia
ā Hypoglycemia and insulin stacking
ā Iatrogenic DKA in patients with type 1 DM
Umpierrez G et al. J Hosp Med. 2006; 1:141-4.
51. Why is sliding-scale insulin use so common if it is
such a bad practice?
ā¢ Sliding-scale insulin is ingrained
in the traditional practice of
medicine in hospitals, as it is easy
to order.
Namely, we kept using the tool that was easiest to grabā¦even though it
didnāt usually work or fit.
52. What is the āBest Practiceā for Managing
Diabetes and Hyperglycemia in the
Hospital?
ā¢ Anticipatory, physiologic insulin
dosing, prescribed as a
basal/bolus insulin regimen
e The right type of insulin .
e In the right amount .
e At the right time.
53. Basal-Bolus Concept The
ānormalā human adult
secretes about 25-30 units
of insulin a day.
As you can see, about Ā½
of this is in āBASALā
insulin. You can also see
that a ānormalā patient
would likely not exceed PG
= 150 mg/dL, even after a
meal. These ambient
glucose levels (euglycemia)
are reflected in a ānormalā
GHbA1c range of 4.5-6.5%
in the USA.
Breakfast Lunch Supper
Insulin
(ĀµU/mL)
Glucose
(mg/dL)
Basal glucose
150
100
50
0
7 8 9 101112 1 2 3 4 5 6 7 8 9
A.M. P.M.
Time of Day
Basal insulin
50
25
0
Nutritional glucose
Nutritional (prandial) insulin
Suppresses glucose
production between
meals and overnight
The 50/50
rule
Physiologic Insulin Secretion:
54. ā¢ Managing a
patient with
diabetes is :
An exercise in
mimicking the
normal
pancreas.
50 / 50
. In certain clinical situations (e.g.
continuous tube feeds, where there is a
risk that the patientās nutrition will be
abruptly stopped), it may be prudent to
give slightly less than Ā½ of the insulin as
basal insulin.
ā¢ In other situations (e.g. immediately
following cardiac surgery), the basal
insulin requirement may rise to as high as
60-80% of the total daily insulin
requirement.
55. A Physiologic Insulin Regimen
ā¢ Basal insulin
ā Long-acting insulin , to provide a constant background level
of insulin, whether the patient is eating or not .
ā¢ Nutritional or pre-meal / prandial insulin
ā Short-acting insulin given with meals in anticipation of
carbohydrate load glycaemic spike ( scheduled insulin).
ā¢ Correction or supplemental insulin
ā Short-acting insulin added to the predetermined nutritional
dose if pre meal glucose is still above target ( i.e as needed
reactive insulin ).
56. Insulin Requirements in
Health & Illness
Relative
proportion
of
insulin
requirement
(%)*
*Estimations for illustrative
purposes: requirements may
vary widely.
Diabetes Care 27:553-91, 2004.
Illness-Related
0
20
40
60
80
100
120
140
Correction
Nutritional
Prandial
Basal
57. Providing Exogenous Basal Insulin
ā¢ Long-acting, non-peaking insulin is preferred as it provides
continuous insulin action, even when the patient is fasting
ā¢ Required in ALL patients with type 1 diabetes
ā¢ Many patients with type 2 diabetes will require basal insulin
in the hospital
ā¢ Can be estimated to be about 1/2 of the total daily dose of
insulin (TDD)
58. Providing Exogenous Nutritional
Insulin
ā¢ Usually given as rapid-acting analogue (preferred in most
cases) or regular insulin, for those patients who are eating
meals
ā¢ Must be matched to the patientās nutrition.
ā¢ Should not be given to patients who are not receiving
nutrition (e.g. NPO).
ā¢ Can be estimated to be about Ā½ of the total daily dose of
insulin (TDD)
59. Providing Exogenous Correctional
Insulin
ā¢ Correctional insulin is extra insulin that is given to correct
hyperglycemia
ā¢ Usually rapid-acting or regular insulin (usually the same as the
nutritional insulin)
ā¢ Often written in a āsteppedā format that is used in addition to
basal and nutritional insulin
ā¢ Customized to the patient using an estimate of the patientās
insulin sensitivity
ā¢ If correctional insulin is required consistently, or in high doses, it
suggests a need to modify the basal and/or nutritional insulin
doses
60. Action Profiles of Bolus & Basal Insulins
Hours
NB: action curves are approximations for illustrative purposes. Actual patient
response will vary.
regular 2-4 hours
NPH 6-12 hours
lispro/aspart 1-2 hours
ļ§ BASAL INSULINS
detemir ~ 6-23 hours (dose dependant)
glargine ~ 20-26 hours
Mayfield, JA.. et al, Amer. Fam. Phys.; Aug. 2004, 70(3): 491 Plank, J. et.al. Diabetes Care, May 2005; 28(5): 1107-12
ļ§ BOLUS INSULINS
61. Which Patients Should be Treated with a Physiologic
Insulin Regimen?
During hospitalization
ā¢ Any patient with blood glucose levels consistently above the
target range
Immediately at the time of admission
ā¢ All patients with type 1 diabetes
ā¢ Patients with type 2 diabetes ifā¦
ā They are known to be insulin-requiring
ā They are known to be poorly controlled despite treatment with
significant doses of oral agents
ā They are known to require high doses of oral agents that will
be held in the hospital
62. A Stepwise Approach to Physiologic Insulin Dosing in the Hospital
1. Selecting a Non-ICU Glycemic Target For Your
Practice/Institution .
2. Estimate the amount of insulin the patient would need
over one day, if getting adequate nutrition = Total Daily
Dose (TDD)
3. Assess the patientās nutritional situation
4. Decide which components of insulin the patient will
require, and which percentage of the TDD each should
represent
5. Determine the correctional factor and put a scale for
correctional insulin.
6. Assess blood glucoses at least daily, adjusting insulin
doses as appropriate
64. STEP 2: Estimate the Amount of Insulin the Patient Would Need
Over One Day, If Getting Adequate Nutrition
Total Daily Dose (TDD)
ā¢ Calculate from insulin infusion amount
ā Recent steady state hourly rate x 20, for
example.
ā¢ Add up insulins taken at home, adjust for
glycemic control and other factors
ā¢ Calculate from weight, body habitus, other
factors
65. ā¢ Calculate starting total daily dose (TDD)
ā 0.3 units/kg/day (hypoglycemia risk factors, naĆÆve patient)
ā 0.4 units/kg/day (conservative for most patients)
ā 0.5 ā 0.6 units/kg/day (overweight to obese)
ā¢ Adjust TDD up or down based on
ā Past response to insulin
ā Presence of hyperglycemia inducing agents, stress
ā This Is very conservative and safe (adjust up as needed)
ā¢ Basal insulin = 40-50% of TDD
ā Glargine q HS or q AM, detemir in 1 or 2 doses
Weight based calculation
66. ā¢ Malnutrition and low
body weight
ā¢ Chronic renal failure
ā¢ Decreased oral intake,
failure to provide
nutrition or dextrose
infusion
ā¢ Advanced age
ā¢ Liver disease
ā¢ Beta-blockers
ā¢ Iatrogenic Risk Factors:
SSI, distractions, poor
regimens: disconnect
between testing,
administration of
insulin, and nutrition
ā¢ Known insulin resistance
recognized by high TDD of
insulin or obesity
ā¢ Medications: glucocorticoids,
catecholamines, tacrolimus,
cyclosporine
ā¢ Significant illness: āStress
responseā related to the
release of counter-regulatory
hormones
ā¢ Increases in nutritional intake
(e.g. restarting a diet, starting
enteral or parenteral
nutrition)
67. STEP 3 : Assess the Patientās Nutritional Situation
ā¢ Eating meals or receiving bolus tube feeds
ā¢ Eating meals but with unpredictable intake
ā¢ Getting continuous tube feeds
ā¢ Getting tube feeds for only part of the day
ā¢ Getting parenteral nutrition
ā¢ NPO
68. STEP 4: Decide Which Components of Insulin the Patient Will
Require, and Which Percentage of the TDD Each Should
Represent
ā¢
ā¢ In most cases, basal insulin should be provided
ā¢ In most cases, well-designed corrective insulin regimens should be provided
ā¢ When a patient is not receiving nutrition nutritional insulin should
not be given.
ā¢ Nutritional insulin needs must be matched to the actual nutritional intake.
69. STEP 5 :Determine the correctional factor and put a
scale for correctional insulin.
ā¢ How do we correct for preprandial
hyperglycemia?
ā¢ We use a SLIDING SCALE!!!
ā¢ Rules
ā Only given with meals
ā Do not use at bedtime or at 3am
ā Use the same type of short acting as your
SCHEDULED short acting
ā Add this to the amount of your SCHEDULED short
acting
71. Correctional Scale
<= 40 Units per day 40-80 Units per day > 80 Units per day
Pre-Meal
BG
Additional
Units
Pre-Meal
BG
Additional
Units
Pre-Meal
BG
Additional
Units
150-199 1 150-199 1 150-199 2
200-249 2 200-249 3 200-249 4
250-299 3 250-299 5 250-299 7
300-349 4 300-349 7 300-349 10
>349 5 >349 8 >349 12
72. STEP 6: Assess Blood Glucoses at Least Daily, Adjusting
Insulin Doses as Appropriate
ā¢ There is no āautopilotā insulin regimen for a
hospitalized patient!
ā¢ Perhaps more important than knowing
exactly how to modify an insulin program is
having the understanding that the program
must be regularly modified.
ā¢ At least once a day the insulin regimen
should be scrutinized by the medical team,
considering any changes in clinical condition
or diet, as well as the glycemic control that is
being achieved.
74. Have a Discharge Plan
Tailored to Patient!
ā¢ Diabetes and insulin education, survival skills:
START EARLY and repeat
ā¢ Follow up and community resources
ā¢ Patient and family can understand
ā¢ Reconcile medications
ā¢ Language, health literacy, and cultural barriers
ā¢ Use HbA1c
ā¢ Insulin requirement may decrease post discharge
75.
76.
77. Conclusion
ā¢ In the end, optimal metabolic control in the
hospital will only be achieved by frequent (at
least daily) assessment of a patientās blood
glucose, and frequent adjustment of the insulin
program.
ā¢ Making adjustments to an insulin regimen, based
on the glycemic control achieved, is more art
than science. It is difficult for any educational
module to allow you to master the art of insulin
adjustment across the range of possible patients
and circumstances.
78. Case # 1
ā¢ 56 year old man admitted with diabetic
foot infection, eating regular meals.
ā¢ Obese, weighs 100 kg
ā¢ Home regimen
ā 2 OHGās and 20 units of NPH q HS
ā¢ Baseline Control:
ā HbA1c of 10, RBS in ED 240 mg/dL
79. Accuchecks AC .
TDD: 100 kg x 0.6 units/kg/day = 60 units
Glargine (Lantus) Alternative
Basal:NPH (20 & 10 ) or Glargine 30 units HS.
Nutritional: Lispro 10 units q ac
Correction: Lispro per scale q ac .
81. Use 120% of yesterdayās total as new TDD (or 130%,
depending on the uniformity and degree of poor control)
82. Case #2:
ā¢ 60 yo man with DM 2, well controlled in ICU on
insulin infusion and continuous tube F at 40
ml/hour.
ā¢ Insulin Infusion rate 80 units in the last 24 hours,
3 units / hour over last 6 hours.
ā¢ Prior to hospitalization, baseline HbA1c was 8.7
on 40 units of 70/30 insulin per day and OHGs.
ā¢ Plan: Transfer to ward, continue enteral nutrition
ā¢ How do you transition this patient to a
subcutaneous insulin regimen?
83. Case 2: Transition to subcutaneous insulin
(enteral nutrition to continue)
ā¢ Safe Estimate of 24 hour requirement:
3 units / hour x 20 = 60 units
ā¢ 60 units represents the TDD: Basal and nutritional
insulin
ā¢ 50:50 Rule Example
ā Glargine 30 units = Basal
ā Regular 7 units q 6 h = Nutritional
ā¢ Correction dose of regular insulin also given along
with nutritional dose as needed.
ā¢ Glargine / Nutritional should be given BEFORE IV
insulin stopped
84. What if??? Enteral to PO
ā¢ Instead of continuing enteral nutrition on the
floor, you opt to stop enteral nutrition and start
patient on a soft diet?
ā¢ Glargine 30 units = Basal
ā¢ RAA 10 units q AC = Nutritional / Prandial
(IF you expect them to eat a full meal! )
ā¢ If po intake suspect at first, empirically reduce
nutritional RAA dose and give the dose just
AFTER the meal instead of just BEFORE the
meal.
ā¢ CORRECTION dose RAA insulin also needed.