Basal-bolus insulin regimens combined with carbohydrate counting are attractive
regimens for middle and high school students. Because children often have erratic
eating habits, rapid-acting insulins are advantageous over regular insulin because
they can be injected directly before or immediately after a meal, accounting for the
portion of the meal a child actually consumed. J.C. should be started on
patients, insulin glargine may not last a full 24 hours; in this case, the dose of insulin
glargine should be divided and given twice daily, and then each adjusted based on
116 Although U-300 glargine or insulin degludec would last 24 hours,
they are not FDA-approved in children at this time.
When J.C. and her caregivers become skilled with carbohydrate counting, insulin
kinetics, dosing insulin based on her carbohydrate intake, and diabetes management,
the use of an insulin pump may be considered. The insulin pump therapy in the
pediatric population is increasing rapidly as it provides increased flexibility with
meal timing and has been shown to improve glycemic control and quality of life.
Young children (not just adolescents) are now recommended for consideration of
123 Family and adult support both at home and school is critical
for successful pump use until the child is able to manage his/her diabetes
CASE 53-4, QUESTION 4: Where should J.C. administer her insulin? Are the recommended sites of
injection different for children? Does the age of the child play a factor?
For infants with abundant SC tissue, injection sites are usually plentiful. For some
toddlers who have lost their “baby fat,” locating an appropriate site for injection can
be difficult. Injecting insulin into the abdomen of children with minimal SC
abdominal fat or in very young children may not be advisable. Rotation of injection
sites among arms, thighs, and the upper-outer quadrant of the buttock or hip area, as
well as the abdominal area in older children, may be beneficial. To achieve
consistent absorption, insulin injections can be patterned, for example, using the arms
for the morning injection and the thighs for the evening injection. Children and teens
should be cautioned to not consistently inject their insulin into a single area, which
may be more convenient for them.
125 Fatty deposits and scar tissue can develop
secondary to insulin action at the local tissue level. Insulin absorption from these
hypertrophied areas is generally poor and unpredictable, resulting in variability in
glycemic control. Insulin pen devices are very helpful for use in children because
they are easy to use, have the option for an even smaller needle size, and are less
intimidating (see Case 53-2, Question 7).
CASE 53-4, QUESTION 5: How often should J.C. monitor her BG?
The eventual goal for children with diabetes is self-management, with insulin
dosing decisions based on interpretation of BG results. Self-management skills and
basal-bolus insulin regimens rely on frequent SMBG. For children with Type 1
diabetes, four or more BG tests per day are generally necessary. Many home BG
meters allow for alternative site testing (arm or thigh), which decreases the
discomfort of fingersticks. Enthusiasm for frequent BG testing tends to wane with
duration of diabetes. However, families who are instructed on managing diabetes on
the basis of test results are better motivated to persevere with SMBG. CGM may also
be considered at some point for improved assessment of her metabolic control,
particularly to detect nocturnal hypoglycemia. J.C. should test her BG before each
meal and at bedtime, at a minimum. Additional tests should be performed whenever
J.C. experiences hypoglycemia, ketonuria, or when she becomes acutely ill.
units (~0.3 unit/kg). Has her diabetes gone into remission?
Approximately 20% to 30% of individuals with Type 1 diabetes go into a
remission phase (honeymoon period) within days to weeks of their diagnosis.
During this phase, which can last for weeks to months, insulin requirements can fall
well below the usual initial dose of 0.5 to 1.0 units/kg/day, and C-peptide can be
reduced. As illustrated by J.C., this presents clinically as markedly decreased insulin
requirements to maintain normoglycemia. J.C. should continue to perform SMBG and
closely monitor for rising BG concentrations, as β-cell destruction continues during
the honeymoon phase and she will eventually return to higher-insulin requirements.
How do the symptoms of hypoglycemia differ in a child compared with an adult? How can the risk of
hypoglycemia be minimized for J.C.?
hypoglycemia increases with attempts to maintain meticulous control of BG
It was previously thought that pediatric patients may be at risk of
cognitive impairment following severe episodes of hypoglycemia; however, current
literature does not support this finding. Common causes of hypoglycemia include
changes in carbohydrate intake, late or skipped meals or snacks, exercise or unusual
activity, and administration of excessive insulin. Because very young children may
not be able to identify or express symptoms of hypoglycemia, caretakers must
observe the child closely and identify symptoms or behaviors associated with a
falling BG. Symptoms of hypoglycemia may include crankiness, sudden crying,
restless sleep, or nightmares as seen in J.C.
Hypoglycemia is more frequent in children with lower A1C values, a prior history
of severe hypoglycemia, larger insulin doses, and younger children.
hypoglycemia is reported in 14% to 47% of children with Type 1 diabetes and is
thought to be related to impaired counter-regulatory response to hypoglycemia during
128 Bedtime BG levels are poor predictors of nocturnal hypoglycemia. J.C.’s
parents should be instructed to test her BG at 2 AM closely for the next few nights,
and then continue to check at least twice weekly. In children, insulin glargine can
exhibit a small peak effect during the initial 3 to 5 hours after administration,
increasing the risk for nocturnal hypoglycemia.
If this is the case, the insulin
glargine dose should be moved to dinnertime or in the morning. If this does not
correct the nocturnal hypoglycemia, then the dose should be reduced. Use of insulin
glargine is associated with less nocturnal hypoglycemia (and asymptomatic nocturnal
hypoglycemia) compared with NPH insulin in children and adolescents.
bedtime snack may also be needed. Treatment of hypoglycemia is addressed in Case
Insulin requirements always increase in the presence of an infection or acute
illness, even if food intake is diminished. Patients with Type 1 diabetes, such as
G.M., commonly decrease or eliminate insulin doses under these circumstances, and
it is in just this setting that ketoacidosis may occur.
Therefore, G.M. should be instructed to maintain her usual dose of insulin and test
her BG and urine ketones every 3 to 4 hours. If BG concentrations are above the
usual range, extra doses of her rapid-acting insulin should be administered according
to a prescribed algorithm based on her body’s sensitivity to insulin (e.g., 1 unit for
each 50 mg/dL above her BG target). People with Type 1 diabetes should be
instructed to test for ketones if their BG concentration is 300 mg/dL or higher. G.M.
should call her physician if her BG concentration remains more than 240 mg/dL after
three corrective insulin doses; if she has moderate-to-large amounts of ketones in her
urine or blood (if using a meter that can measure these); if she has been vomiting or
having diarrhea for longer than 6 hours; or if she begins to experience signs and
symptoms related to ketoacidosis (polyuria, polydipsia, dehydration, ketonuria, and a
fruity breath [see Case 53-10]). G.M. also should attempt to maintain her fluid,
mineral, and carbohydrate intake with easily digested food and fluids (Table 53-
Test your blood glucose more frequently: every 3–4 hours
example, 1–2 units for every 30–50 mg/dL over an agreed-on target glucose concentration (e.g., 150
testing especially when glucose readings exceed 300 mg/dL
4 hours). Foods such as gelatin, noncarbonated soft drinks, crackers, soup, and soda may be used
INSULIN REQUIREMENTS IN RENAL FAILURE
the anticipated effects of decreased renal function on M.B.’s insulin requirements?
The effects of renal failure on insulin requirements are complex, and under various
circumstances, insulin requirements may increase or decrease. The kidney is the most
important site of extrahepatic insulin metabolism and excretion. In nondiabetic
individuals, the liver extracts approximately 60% of insulin secreted endogenously
before it reaches the peripheral circulation.
63 Because exogenous insulin is delivered
directly to the periphery, the kidneys play a more important role in its elimination.
Insulin is filtered by the glomerulus and reabsorbed in the proximal tubules, where it
is destroyed enzymatically. The kidney also clears insulin from the peritubular
65,131 At that site, insulin can enhance the reabsorption of sodium, which
may account for the edema occasionally observed after the initiation of insulin
Diminished renal function can be accompanied by decreased clearance of
endogenous and exogenous insulin, resulting in increased plasma concentrations of
insulin. Therefore, M.B.’s insulin requirements may diminish as her renal disease
progresses. Patients with moderate degrees of renal failure (GFR >22.5 mL/minute)
remove 39% of insulin from arterial plasma, similar to normal subjects. In contrast,
patients with severe renal insufficiency (GFR <6 mL/minute) have a marked
reduction in insulin removal from arterial plasma (9%).
clearance in conjunction with nausea and decreased food intake associated with
uremia can lead to hypoglycemia in such individuals. In some patients with diabetes,
particularly those with residual endogenous insulin secretion (e.g., Type 2 diabetes),
glucose tolerance may normalize as renal function diminishes, eliminating the need
for insulin. In contrast, severe uremia is associated with glucose intolerance. This
seems to be related to tissue resistance to insulin secondary to an unknown factor that
As M.B.’s renal function worsens, a reduction in her insulin requirements should
MANAGEMENT OF THE HOSPITALIZED PATIENT
Patients with diabetes account for more than 1 in 5 hospital days in the United
States. Of the nearly $176 billion that is spent annually on diabetes, nearly half is
8 A clear, linear relationship exists between hyperglycemia
and adverse clinical outcomes in the hospitalized patient.
relationship exists regardless of a baseline diagnosis of diabetes at the time of
admission, and iatrogenic hyperglycemia in the hospital does not have the same
relationship morbidity that spontaneous hyperglycemia does.
observations have raised important questions about the relationship between
hyperglycemia and morbidity in the hospitalized patient.
Complex responses to acute illness including excess secretion of catecholamines
and cortisol result in peripheral insulin resistance and so-called stress
hyperglycemia. This makes it difficult to discern whether glycemia is a marker or a
mediator of adverse outcomes in the acutely ill patient. Accordingly, historic practice
had been to only aim for BG concentrations that prevent glucosuria (<200 mg/dL) and
the subsequent risk for dehydration in the hospitalized patient. However, beginning in
2001, a series of randomized trials tested glycemic control in the critically ill patient.
Significant changes have subsequently been made to practice recommendations for
both the critically and the noncritically ill hospitalized patient.
Although several trials beginning in the 1990s tested intensive insulin regimens in
patients with acute MI, they were small, placebo-controlled studies and reached
different conclusions that proved difficult to rectify.
den Berghe trials tested two different levels of glycemic control in a relatively large
number of surgical intensive care unit (ICU) patients with hyperglycemia with or
140 A liberal glucose control strategy (reduction only if BG
rose above 215 mg/dL) was compared with normalization of BG (80–110 mg/dL).
Overall, normalization of BG significantly reduced ICU mortality from 8.0% to
140 However, the same researchers were unable to replicate their findings in a
subsequent, similarly designed study in medical ICU patients with substantially
higher baseline mortality rates.
141 Although a substudy showed that an ICU stay of 3
days or longer was predictive of benefit from tight control, a subsequent and much
larger trial not only failed to confirm that finding but found increased mortality from a
blood sugar of 80 to 110 mg/dL compared with 140 to 180 mg/dL (27.5%, 829 of
3,010 vs. 24.9%, 751 of 3,012).
142 The incidence of severe hypoglycemia (<40
mg/dL) in the different study groups assigned to tight control was between 7% and
18% and did not explain the different findings between studies. However, several
important differences between these trials, which helped to inform current guidelines,
The 2001 van den Berghe trial used parenteral nutrition in all patients and allowed
for higher glucose values (insulin started if blood sugar exceeded 215 mg/dL) to
occur in the conventional glycemia arm.
It is therefore possible that the aggressive
insulin therapy in the tight control group helped to blunt the excessive glucotoxicity
that may have been occurring from the parenteral nutrition. In the second van den
Berghe trial as in the NICE-SUGAR study, parenteral nutrition was rarely used and
initiation of insulin therapy in the conventional arms began at blood sugar values
141 Additionally, in NICE-SUGAR, a more aggressive target
of less than 180 mg/dL was used rather than 180 to 200 mg/dL as in the two van den
140–142 Table 53-21 summarizes these three trials assessing level of
glucose control in critically ill patients.
Overall, numerous individual studies as well as meta-analyses have reached
different conclusions regarding whether or not tight control of BG is superior to
conventional control in the hospitalized, acutely ill patient.
In 2009, the ADA made substantial changes to its 2005 guideline on management
of inpatient hyperglycemia. Although existing randomized trials of glycemic control
have been performed in critical care settings, the ADA guideline included non-ICU
settings. To make recommendations for non-ICU patients, ADA relied on case series
and retrospective analyses, which will ultimately need to be subjected to
randomized, prospective trials.
In the meantime, recommendations for both
critically ill and noncritically ill hospitalized patients are the same: a premeal or
fasting BG target less than 140 mg/dL and random values of less than 180 mg/dL. A
tighter goal of 110 to 140 mgl/dl however may be beneficial in specific subgroups of
patients, such as open-heart surgery patients.
Summary Data of Three Major Trials of Intensive vs. Conventional Glycemic
Control with Insulin in Critically Ill Patients
Intensive Conventional Intensive Conventional CI)
1,548 80–110 180–200 103 153 ICU
1,200 80–110 180–200 111 153 Hospital
6,104 81–108 <180 115 145 90-day
CI, confidence interval; OR, odds ratio.
For perioperative insulin needs, A.G. should receive her usual basal insulin dose
(insulin glargine 24 units) on the night before surgery. If the basal insulin is normally
administered in the morning, the usual dose can still be given for patients with Type 1
diabetes; for those with Type 2 diabetes, 50% to 100% of the basal insulin is
administered the morning of surgery. Correction doses of rapid-acting insulin can be
administered the morning of surgery if the BG is more than 180 mg/dL.
A1C is not available, it can be measured to assess the patient’s glycemic control
Most insulin infusion protocols include the use of IV regular insulin and
maintenance IV fluids, either 5% dextrose in water (D5W) or D5W with 0.45%
normal saline (0.45% NaCl). For a patient requiring fluid restriction, 10% dextrose
149 The adjustment algorithms are used by nursing to
change the rate of infusion (in units/hour) depending on the BG level. Most often, the
insulin infusion is prepared in a solution of 1 unit/1 mL normal saline (e.g., 100 units
of regular insulin in 100 mL of 0.9% NaCl). A dedicated IV line is used for the
insulin infusion to avoid iatrogenic hypoglycemia. The insulin infusion is connected
to the maintenance IV containing dextrose (can be Y-connected). Because insulin
binds to plastic, the insulin solution should be flushed (e.g., with 20 mL) through the
IV tubing before the line is connected to the patient. An IV dextrose infusion is
maintained while a patient is on an insulin infusion. Most patients need 5 to 10 g of
glucose per hour (or D5W or D5W/0.45% NaCl at 100–200 mL/hour). Additional
maintenance fluids (and electrolytes) can be administered via a different port or line.
Some protocols include an initial bolus dose of insulin. The initial insulin infusion
rate is primarily based on current BG level and BMI; other factors such as body
weight, current daily insulin requirements, and renal function should be taken into
consideration. An initial rate of 1 unit/hour is common (can range from 0.5 units/hour
to ≥2 units/hour). The choice of initial infusion rate is not critical but should be
based on patient history. A rate of 0.5 units/hour is appropriate for a patient who has
never previously received insulin, whereas 2 units/hour would be appropriate for a
patient with known insulin-dependent diabetes. Adjustments in the insulin infusion
rate are determined by BG levels every 60 minutes until the BG is stable and close to
target. Then, the frequency of BG testing may be reduced to every 2 to 3 hours.
Algorithms should consider both the current and previous BG level, the rate of
change of the BG level, and the current infusion rate.
Insulin infusion should be started at least 2 to 3 hours before the surgery to titrate
to the desired level of glucose control. Examples of protocols are available on the
Institute for Healthcare Improvement’s website (available at
http://www.ihi.org/IHI/Topics/PatientSafety/MedicationSystems/Tools/), and
many are published in the medical literature.
Thus, A.G.’s usual SC insulin regimen should be discontinued, and she should be
initiated on an insulin infusion that is adjusted according to an algorithm. Throughout
the perioperative period, she should receive a minimum of 100 g of glucose daily to
Assessment of interfering substances with point-of-care BG testing is particularly
important for hospitalized patients. Some immunoglobulins and dialysates contain
nonglucose sugars (including maltose, xylose, and galactose), which can interfere
with glucose measurements with glucose dehydrogenase pyrroloquinolinequinone test
strips (will falsely elevate the reading, Table 53-16). BG concentrations should only
be performed by the laboratory in these patients.
QUESTION 1: G.O., a 42-year-old, slightly overweight (5 feet 11 inches, 200 pounds, BMI 27.9 kg/m
feet. G.O. also complained of impotence and “shooting pains” in both legs. A spot collection for
microalbuminuria was 450 mg of albumin/g creatinine (normal, <30 mg/g creatinine).
38 For the past several months, he has been treated with the following regimen: 14 to 18 units of
concentrations have been as follows:
Time Glucose Concentration (mg/dL)
hypoglycemia? What are the causes?
G.O.’s case illustrates one of the major hazards of aggressive BG targets and
intensive insulin therapy: hypoglycemia. Hypoglycemia is a fact of life for patients
with Type 1 diabetes, virtually all of whom experience a hypoglycemic episode at
one time or another. Nocturnal hypoglycemia is of particular concern. A syndrome
called “dead-in-bed” has been described for patients with Type 1 diabetes, who
experience repeated hypoglycemia and have an underlying cardiovascular pathology,
Hypoglycemia is a BG concentration less than70mg/dL, and its occurrence is
potentially fatal if not promptly recognized and treated. However, the exact level at
which a patient experiences symptoms is difficult to define. Clinical hypoglycemia is
associated with typical autonomic (neurogenic) and neuroglycopenic symptoms
relieved by the administration of a quickly absorbed carbohydrate.
Normal brain function depends on glucose, the exclusive fuel for cerebral
metabolism. Because the brain is unable to synthesize or store glucose, it must be
provided with a constant exogenous quantity via the brain’s blood supply. As BG
concentrations fall, a series of physiologic responses occur to restore glucose levels.
These responses create symptoms warning a patient to take corrective action by
consuming carbohydrates. If these counter-regulatory responses fail to alert the
patient and BG concentrations fall below a critical level, cognitive function becomes
impaired, and confusion and coma may ensue.
In people without diabetes, the peripheral responses to hypoglycemia are so
efficient that clinically important hypoglycemia probably never occurs. As glucose
levels fall between 50 and 60 mg/dL, a series of neuroendocrine events occur,
raising the plasma glucose concentration back toward normal by increasing hepatic
glucose output. The major hormone responsible for producing acute recovery from
insulin-induced hypoglycemia is glucagon; however, epinephrine alone also can
produce near-normal recovery. Rising levels of adrenergic and cholinergic hormones
generate warning symptoms of hypoglycemia. When hypoglycemia is prolonged,
growth hormone and cortisone play a greater role in producing recovery.
Patients with Type 1 diabetes who maintain insulin depots throughout the day are
predisposed to severe hypoglycemic reactions because deficiencies in the normal
feedback system occur with time. Glucagon secretion becomes deficient within the
first 2 to 5 years after diagnosis, and by 10 years or longer, epinephrine secretion
may become impaired. The latter defect leads to asymptomatic hypoglycemia or
hypoglycemic unawareness (see Case 53-9).
Certain circumstances predispose patients with Type 1 diabetes to severe
hypoglycemia. These include (a) a defective counter-regulatory hormonal response to
hypoglycemia (see Case 53-9), which may be further diminished with frequent
hypoglycemia, (b) medications such as β-blockers that diminish early warning signs
of impending hypoglycemia, (c) intensive insulin therapy that can alter secretion of
counter-regulatory hormones, (d) skipped meals or inadequate carbohydrate intake
relative to the insulin dose, (e) physical activity, and (e) excessive alcohol intake
The signs and symptoms associated with hypoglycemia vary in intensity according to
the presence of cognitive deficits and the patient’s ability to self-treat the reaction.
They vary substantially from one patient to another. Symptoms are conventionally
divided into two categories: neurogenic (or autonomic) and neuroglycopenic.
Comments
Post a Comment
اكتب تعليق حول الموضوع