Diabetic Ketoacidosis in Children

1
Diabetic Ketoacidosis in Children

• Keystone, July, 2008
• Arleta Rewers MD, PhD
• Robert Slover MD

2
Overview

• Review the incidence and pathophysiology of DKA
• Define the role of patient self-monitoring
  including blood ketones testing and the
  healthcare professional advice in preventing DKA
• Describe current approaches to the clinical
  diagnosis of DKA, including the role of ketone
  body levels
• List treatment options for DKA

3
Definition

• Hyperglycemia BG gt 200 mg/dl (11 mmol/l)
• (young or partially treated children, pregnant
  adolescents may present with euglycemic
  ketoacidosis)
• Venous pH lt7.3 and/or bicarbonate lt15 mmol/L
• mild DKA pH lt7.3 bicarbonate lt15
• moderate pH lt7.2 bicarbonate lt10
• severe pH lt7.1 bicarbonate lt 5
• Glucosuria and ketonuria/ketonemia (ß-HOB)

4
Incidence of DKA at onset

• Wide geographic variation in DKA rates at
  diabetes onset 15 -70
• More common in developing countries
• DKA rates inversely related to incidence of type
  1 diabetes

5
Diabetic Ketoacidosis at Diagnosis of DM in
Youth The SEARCH for Diabetes in Youth Study
Incidence of DKA at the time of diagnosis

• SEARCH is multicenter study
• In 2002 began population-based ascertainment of
  incident cases of DM in youth younger than 20
  years
• Incidence
• Overall - 25.5 (CI 23.9-27.1)
• Type 1 - 29.4 ( CI 27.5-31.3)
• Type 2 - 9.7 ( CI 7.1-12.2)

Rewers A et al., Pediatrics, May 2008
6
Risk factors for DKA at onset

• Age lt12 yrs
• No first degree diabetic relative
• Lower socioeconomic status
• High dose glucocorticoids, atypical
  antipsychotics, diazoxide and some
  immunosuppresive drugs
• Poor access to medical care
• Uninsured

7
Prevalence of hospitalization and DKA at onset
Colorado children, 1978-2001
Rewers et al., ADA 2003
8
DKA in children with established T1DM

• The risk of DKA varies from 110 to 1100 /p-yr
• Poor metabolic control or previous DKA ? risk
• Adolescent girls
• Children with psychiatric disorders, including
  those with eating disorders
• Lower socio-economic status
• Lacking appropriate insurance
• Inappropriate interruption of insulin pump therapy

9
Predictors of Acute Complications in Children
With Type 1 Diabetes A Rewers, HP
Chase, T MacKenzie, P Walravens, M Roback M
Rewers, RF Hamman, G Klingensmith
20022872511-2518
Cohort of 1,243 diabetic children from BDC -
age 0-19 years - residence in the six-county
Denver area - outpatient visits between
1/1/1996 - 1/1/2001 Average follow-up 3.2
years Total follow-up 4,000
person-years DKA events 320 DKA
incidence 8 / 100 person-years
10
Incidence of DKA in established patientsBDC
Cohort, 1996-2001
Incidence /100 p-yrs
p0.0006
p0.006
p0.06
Age years
11
Proportion of Children with Recurrent DKA
5

• 60 of DKA episodes occurred in 5 of children
  who had 2 or more events

of DKA events
12
Diabetic KetoAcidosis (DKA)

• 160,000 Admissions to private hospitals/year
• Cost over 1 billion annually
• 65 lt19 years old
• Main cause of death in children with diabetes
  (approximately 85)
• Cerebral edema in 69

13
Cost of hospitalization of a diabetic patient
11,123
6,055
HCIA-Sachs, 1998 Claims Data Warehouse,
represents 2.5MM lives and 150 health plans
14

Diabetes Care 2006 291150-1159
15
What are the presenting complaints?

• Gastro-enteritis
• Vomiting - but no diarrhea
• Dehydration - but excessive urine
  output !
• Respiratory distress
• But no lung findings

16
Signs of DKA

• Vomiting
• Increased urination
• Abdominal pain
• Fruity odor to breath
• Dry mouth and tongue
• Drowsiness
• Deep breathing
• Coma
• Death

17
Mortality in Children with DKA

• 0.15 USA
• 0.18 Canada
• 0.31 UK
• 80 of deaths occurs in association with signs of
  CE
• Other causes
• hypokalemia / hyperkalemia
• thrombosis
• intracranial bleeding, infarction
• sepsis and other infections, e.g., mucormycosis
• aspiration pneumonia
• pulmonary oedema, ARDS

18
Physical Exam

• Perfusion
• Vital Signs - including weight
• Hydration
• Mental Status
• Evidence for insulin resistance

19
Initial Laboratory Evaluation

• Venous pH
• BUN
• Serum Osmolality
• Phosphorus
• Calcium
• Anion Gap

• Glucose
• Ketones
• Sodium
• Potassium
• Chloride
• HCO3

Always perform in an ill child
20
Calculations

• Serum Osmolality
• 2NaK (glucose/18) BUN/2.8
• Serum Na
• Corrected Na
• measured Na (1.6)(glucose - 100)/100
• Anion Gap
• Na (ClHCO3)
• Normally 12/-2 mmol/L

21
Cerebral Edema

• Major cause of death in childhood DKA
• 20 with cerebral edema die
• 20 with mild to severe neurologic outcomes
• At risk
• Initial pH lt 7.1
• Baseline mental status abnormal
• Newly diagnosed, lt 5 years old
• Rapid rehydration (gt 50cc/ kg in first 4 hrs)
• Hypernatremia/ persistent hyponatremia

22
Age distribution of affected children
23
Time of onset of Neurological Compromise (hours)
Timing of Onset of Cerebral Edema in DKA
of patients
12-15
Muir A, et al, Diab Care. July 2004
24
Symptoms and signs of cerebral edema

• Headache
• Decreased or worsening level of consciousness
• Slowing of the HR
• Increase in BP
• Sudden onset/return of vomiting
• Warning signs occur before the onset of CE

25
Clinical Factors Associated with Cerebral Edema

• Prolonged Illness
• Severe acidosis - low PA CO2
• Severe dehydration
• Bicarbonate therapy
• Persistent hyponatremia
• Excessive fluid admistration

26
Cerebral edema

• CE occurs in 0.3- 1 of all episodes of DKA
• Initial 24 hours of treatment
• Younger children (lt 4 yrs)
• Delayed diagnosis
• Greater dehydration and acidosis, lower pCO2
• Insulin given before fluids

27
Etiology of CE

• Vasogenic - excessive accumulation of water and
  solutes in the interstitial space, due to
  dysfunction of the blood-brain barrier
• Cytotoxic - excessive accumulation of water and
  solutes in the intracellular space, due to
  dysfunction of cell-volume regulatory mechanisms
• Both forms may co-exist

28
Excessive Free Water

• Corrected Na Na(measured)1.6 (glucose-100)/100
• Calculated sodium is low and falling in many
  cases of cerebral edema
• ADH levels rise 5-50 times in DKA and contribute
  to increase in free water and hyponatremia

29
Cerebral Edema

• Know what to look for
• Altered mental status/ severe headache
• Recurrence of vomiting
• Changes in pupil size, seizures, bradycardia
• Clinical worsening despite improving lab values
• CT/ MRI changes may not be seen in early cerebral
  edema

30
Cerebral Edema Bedside Score
Caveat note that patient needs to be
significantly affected to meet diagnostic
criteria
Muir Diab Care 2004 271541-46
31
Timing of presentation of cerebral edema
32
Treatment of cerebral edema

• Mannitol 1 gram/ kg IV over 30 minutes
• Elevate the head of the bed
• Decrease IVF rate and insulin infusion rate
• Pediatric ICU management
• Do not delay treatment until radiographic
  evidence

33
Diagnosis and prevention of DKAin outpatients
34
Why do ketones develop?

• No carbohydrate intake
• fasting
• gastroenteritis
• Atkins diet, neonates fed high-fat milk
• Prolonged exercise, pregnancy
• Lack of insulin activity
• onset of diabetes (insufficient secretion)
• interruption of insulin delivery in established
  pt
• Increase in insulin resistance
• infection, illness, surgery, stress
• Alcohol, salicylate ingestion, inborn metabolic
  errors

35
Treatment of Mild DKA to Prevent Progression
Key Early Detection

• Check blood ketones (?-OHB) for a person with
  diabetes any time
• A SMBG is gt300 mg/dL (16.7 mmol/L)
• An illness or infection is present
• Unusual symptoms are present
• It is realized a shot/bolus was missed or bad
  insulin

36
Old Paradigm Check urine ketones New Paradigm
Check blood ?-OHB

• Blood ?-OHB tells you how you are doing at the
  time of the test. (Urine may have been in bladder
  for hrs)
• Urine ketone levels may not accurately reflect
  the severity of the ketonemia
• A person may not be able to void
• Some (teens) give false urine test results

37
Hand-held deviceAbbott/MediSense
38
Disadvantages to Urine Ketone Testing

• The results are not real time
• The readings are qualitative color comparisons
  indicating high, medium or low levels
• Short shelf life (typically 90 days on opening a
  vial)
• Sulfhydryl drugs, including the ACE inhibitor,
  Captopril, may cause false-positive results
• High doses of Vitamin C may cause false-negative
  results
• Method does not detect the major ketone body
  ?-hydroxybutyrate

39
Interpretation of Blood ?-OHB
?-OHB level (mmol/L) lt 0.6 normal
gt1.0 hyperketonemia 0.6-1.0 take extra
insulin fluids 1.0-1.5 as above recheck
in 1 hr and, if no improvement, call diabetes
provider 1.5-3.0 call diabetes provider
STAT gt 3.0 sick KETOACIDOSIS gt Go to ED
40
Sick Day Management A Randomized Clinical Trial
Laffel L, et al. Diabet Med 2005

• 123 participants, age 322 years
• 61 randomized to home blood ß-OHB testing
• 62 randomized to home urine Ketostix testing
• All participants trained on their sick-day
  guidelines
• Outcomes
• ER visits
• Hospitalizations

41
Patients who monitor blood ß-OHB test more
often than those who test for ketonuria
Laffel L, et al. Diabet Med 2005
42
Lower incidence rates of ER use/hospitalizations
in patients using blood ß-OHB monitoring vs.
Ketostix 6-month follow-up
p  0.05
Laffel L, et al. Diabet Med 2005
43
Use of Blood ?-hydroxybutyrate Levels at the
BedsideDuring Treatment of DKA

• ADA, June, 2007

44
Fat
Normal state postprandial
glucose
acetyl CoA
pyruvate
Krebs cycle
oxaloacetate
citrate
45
Fat
Normal state postprandial
lipase
?
fatty acids ( glycerol)
?
fatty acyl CoA
glucose
?-oxidation
acetyl CoA
pyruvate
Krebs cycle
oxaloacetate
citrate
insulin
46
Fat
Normal state postprandial
lipase
?
fatty acids ( glycerol)
?
fatty acyl CoA
?-oxidation
acetyl CoA
Krebs cycle
?
acetoacetyl CoA
HMGCoA synthase
acetoacetate
11
acetone ?-OHB
insulin
47
Fat
Ketosis in DKA - alternative source of energy
lipase
? fatty acids
glucose ?
fatty acyl CoA
?-oxidation
acetyl CoA
pyruvate
?
Krebs cycle
acetoacetyl CoA
?
HMGCoA synthase
acetoacetate
oxaloacetate
110
citrate
acetone ?-OHB
glucagon
insulin
48

• Is bedside ß-OHB monitoring using hand-held
  device as accurate as reference laboratory method
  ?

49
Correlation between venous whole blood ß-OHB
levels measured using Precision Xtra and serum
levels using Cobas Mira Plus (Roche)
Laboratory reference ß-OHB mg/dL
Bedside meter ß-OHB mmol/L
Rewers A et al. Diabet Technol Therapeutics,
20068671
50
Bland-Altman plot showing good agreement between
ß-OHB levels measured using Precision Xtra and
Cobas Mira Plus (Roche)Mean difference 0.18
(C.I. -1.18-1.53)
Rewers A et al. Diabet Technol Therapeutics,
20068671 also Byrne H, et al. 2000 Wallace
TM, et al. 2001 Ham MR, et al. 2004 Khan ASA,
et al. 2004
51
CONCLUSION

• Real-time bedside measurement of ?-OHB is
  generally as accurate as reference laboratory,
  especially at levels up to 3.0- 4.0 mmol/L

52

• Is capillary blood ß-OHB monitoring superior
  to testing urine for ketones ?

53
Measurement of Ketones

• Urine ketone measurements use a dip stick
  method based on a chemical reaction with
  acetoacetate. E.g., Chemstrip from Roche
  Clinistix, Ketostix , Keto-Diastix from Bayer)
• Blood ketone testing that specifically measures
  ß-hydroxybutyrate are available for use in the
  laboratory (e.g., Sigma, Cobos from Roche) and
  a hand-held meter (Abbott / MediSense)

54
Blood ß-OHB testing is superior to urine ketone
testing in detecting ketosis

Gold standard plasma ß-OHB by reference
laboratory method (KONE Delta Automatic Analyzer)
Guerci B , et al. Diabetes Care 2003 Similar
data Taboulet P et al. Eur J Emerg Med 2004
55
Advantages Blood ß-OHB vs. Urine Ketone Testing

• ?-OHB is a better marker of ketosis than
  acetoacetate
• ?-OHB is real-time while ketonuria is usually
  old news
• Ketonuria doesnt accurately reflect severity of
  ketonemia
• A dehydrated person may not be able to void
• Some people are too ill or exhausted to do the
  urine test
• Some patients (teens) give false urine sample
• Urine ketone strips spoil after opened gt6 months

56
?-hydroxybutyrate is a better indicator of
metabolic status when detecting and treating DKA







Schade DS, Eaton RP Special Topics in Endo and
Metab 198241-27
57
ß-OHB in diagnosis of DKAin ED
58
ß-OHB helps to diagnose DKA in patients with
known or new diabetes seen in ER
59
Capillary blood ß-OHB vs. venous pH in 118 newly
diagnosed children
no DKA compensated acidosis
DKA
3
?
?
?
?
0.5
7.25
Prisco F, et al. Pediatr Diabetes 2006
60

• Can bedside ß-OHB monitoring replace repeat
  measurements
• of pH, bicarbonate and pCO2 during treatment
  of DKA?

61
Correlation between baseline ß-OHB and other
biochemical indicators in 68 children with DKA
Pearson correlation coefficients (p lt0.05 for
all)
Rewers A et al. Diabet Technol Therapeutics,
20068671
62
Time series analysis showing that bedside ß-OHB
levels correlated very closely with
time-dependent levels of venous blood gases
plt0.001 plt0.0001
Rewers A et al. Diabet Technol Therapeutics,
20068671
63
CONCLUSION

• While the initial measurement of pH,
  bicarbonate and pCO2 is warranted, real-time
  bedside measurement of ?-OHB may replace repeat
  measurements of blood gases in treatment of DKA

64

• Can bedside ß-OHB monitoring shorten duration
  of DKA treatment ?

65


In most newly-diagnosed children with ketosis,
capillary ketonemia resolves sooner
than ketonuria
N 99
Prisco F, et al. Pediatr Diabetes 2006
66
In children with DKA, capillary ketonemia
resolves on average 11
hours sooner than ketonuria
(n40)
Example of an individual treatment profile
pH gt7.3 ß-OHB lt1.0
pH gt7.3 No ketonuria
i.v. insulin U kg/h
ß-OHB
Noyes KJ, et al. Pediatr Diabetes 2007,
confirming Vanelli M, et. Al. Diabetes Care 2003
67
CONCLUSIONS

• Real-time bedside measurement of ?-OHB may
  help to optimize treatment of DKA and shorten the
  duration of hospitalization

68
Initial Laboratory Evaluation

• Venous pH
• BUN
• Serum Osmolality
• Phosphorus
• Calcium

• Glucose
• Ketones
• Sodium
• Potassium
• Chloride
• HCO3

Always perform in an ill child
69
Treatment

• Monitoring
• Management requires close attention to detail
• Use a flowsheet to track vital signs labs, rates
  of insulin, fluids, dextrose
• Neurological status
• consider neuro checks q 1 hr
• How does the patient look TO YOU?
• Assess, reassess and then assess again

70
Treatment

• Consider ICU admission for closer monitoring if
• Severe DKA (pH lt 7.1 or lt 7.2 in young child)
• Altered level of consciousness
• Under age of 5 years
• Increased risk for cerebral edema
• Caution with meds that may alter mental status

71
Fluid Therapy for DKA

• Assume 10-15 dehydration
• Begin with a 10-20 ml/kg bolus of NS
• Replace calculated deficit evenly over 36 hours -
  generally 1.5 x maintenance for the next
  several hours is appropriate
• Do not exceed 40mls/kg in the initial 4 hours,
  or 4 L/m squared in 24 hours

72
DKA - Fluids

• Double bag system
• ¾ NS at 1.5 x M until glucose below 300 mg/dl
• D10 ¾ NS to be mixed with ¾ NS to achieve
  desired glucose concentration
• K supplementation
• 20mEq/L K Acetate 20mEq/L K Phosphate
• Ionized calcium is low, phosphorous should not be
  given
• early replacement and frequent monitoring
• Bicarbonate therapy is rarely, if ever, indicated

73
Insulin Therapy for DKA

• IV infusion with basal rate 0.1 U/kg/hr
• No initial insulin bolus it will decrease time
  to correction of the glucose, but does not alter
  the time to correction of acidosis
• It may decrease the serum osmolality more
  rapidly than desirable
• Ideal glucose decline is about 100 mg/hr
• Continue insulin until urinary (blood) ketones
  are cleared

74
Potassium

• Add potassium when Klt 5 and with urination
• K gt5.5 no potassium in IVF
• K 4.5 5.5 20 meq/L K
• K lt4.5 40 meq/L K

75
Phosphate the controversy

• Prevent depletion of RBC 2,3 DPG which will
  improve tissue oxygenation as acidosis is
  resolving
• May be useful in patients with anemia, CHF,
  pneumonia, hypoxia

76
Use of Bicarbonate in DKA

• Bicarbonate should be used only when
• there is severe depression of the
• circulatory system or cellular metabolism...
• Not recommended unless pH lt7.0, not even then,
  unless above true

77
DKA Cases

• 12 year old admitted with
• pH 7.0
• Na 136, K3.8, glucose 583mg/ dl
• She is oriented and conversant on admission, you
  follow the DKA protocol,
• 2 hours later she becomes difficult to arouse and
  is responsive only to deep pain. - What do you
  do?
• Presume cerebral edema
• Decrease fluid infusion to insensible losses
• Give mannitol 1 gm/kg

78
DKA Cases

• 6 y/o boy is admitted in severe DKA. The family
  has been traveling and he has been ill for
  several days.
• Initial pH7.0, K 3.7, glucose is 350mg.
• Despite replacement, his K now is 1.9 mg/dl -
  what do you do?
• A bolus of potassium at TCH is actually an
  infusion over an hour. An actual bolus of
  potassium into a central vein may be lethal

79
DKA Cases

• 16 year old boy is admitted in moderate to severe
  DKA (pH7.23), his weight is 230 lbs, his BG is
  1400, serum osm is 360 mOsm/L, what do you do?
• Monitor! Everything you can!

80
Successful Management

• Careful attention to detail
• Careful record keeping
• A detailed flow chart is essential
• Following the data recorded is also essential
• Repeated examination of the patient