Assessment and therapy of selected endocrine disorders

Anesthesiology Clin N Am
22 (2004) 93 – 123
Assessment and therapy of selected
endocrine disorders
Lisa E. Connery, MDa,*, Douglas B. Coursin, MDb
Departments of Surgery, Internal Medicine, and Anesthesiology, Long Island Jewish Medical Center,
270 – 05 76th Avenue, New Hyde Park, NY 11040, USA
Department of Anesthesiology, University of Wisconsin Medical School, B6/319 UW CSC,
Madison, WI 53792, USA
Anesthesiologists care for the entire spectrum of patients ranging from complex newborns to the most compromised geriatric patient. Clinicians must adapt
techniques and anticipate special problems and complications that may arise
when dealing with specific patient populations. Because patients with various
endocrine disorders frequently are at risk for adverse perioperative events caused
by their primary endocrinopathy or secondary complications, preemptive management should be directed when available by evidence-based guidelines.
Preanesthesia evaluation entails review of medical records, the patient interview, a focused physical examination, and judicious use of preoperative testing.
This assessment then guides further consultation or specific specialized tests and
facilitates development of the anesthetic plan. Preoperative assessment is complicated by a number of issues. With the prevalence of same day surgeries, the
patient is often first seen by the anesthesiologist on the day of surgery. This presupposes that an appropriate preoperative evaluation has been performed. There
is often pressure on the anesthesiologist to proceed with even elective surgeries if
information is missing or incomplete. A sea change has occurred in the area of
preoperative assessment from performing a battery of tests on patients regardless
of their medical history or the type of the procedure being performed, to the
current situation of directed and individualized testing. This has come about from
the findings that routine batteries of tests performed preoperatively on asymptomatic patients are often unlikely to be of significant benefit to the patient, and
may, occasionally, harm the patient when incidental findings are unnecessarily
and aggressively pursued. In general it is best not to order a test unless a result of
the test would change the management plan. Preoperative assessment is not yet a
science and further research is needed. The American Society of Anesthesiolo* Corresponding author.
E-mail address: [email protected] (L.E. Connery).
0889-8537/04/$ – see front matter D 2004 Elsevier Inc. All rights reserved.
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
gists task force on preanesthesia assessment concluded that based on the available
scientific literature, decision-making parameters for ordering specific preoperative tests, and the timing of such tests, could not be unequivocally determined [1].
In the evaluation of patients with various endocrine disorders, much of the
‘‘classic’’ preoperative assessment remains empiric and, unfortunately, is not prospectively established.
Endocrine disorders range from the ubiquitous, diabetes, to the rare pheochromocytoma. All endocrinopathies may contribute to serious adverse events or
a labile perioperative course if they are advanced, underappreciated, or inadequately managed. Although present guidelines and recommendations are often
empiric, this review focuses on current concepts of preoperative assessment and
perioperative management of diabetes, and also discusses some specific thyroid
and adrenal pathologies.
Diabetes mellitus
Diabetes mellitus is the most commonly encountered endocrinopathy in western society, developing in roughly 15 to 20 million Americans or 7% to 8% of the
population of the United States. Ninety percent of the diabetics in the United
States have type 2 diabetes, and the remainder type 1 [2,3]. There is strong
longitudinal data that long-term glycemic control limits development of some
complications of diabetes. Recently a growing body of evidence supports the
benefits of euglycemia in selected patients, particularly critically ill cardiac surgical patients and patients with CNS or myocardial ischemia or infarction. The
clinical care of diabetes therefore should focus on appropriate diagnosis, risk
stratification, and therapeutic intervention. The days of laisse faire attitude about
glucose control are behind us.
Type 1 diabetes results from the destruction of insulin-producing pancreatic
beta cells. This destruction is mediated by autoimmune and other mechanisms.
Onset of disease peaks in the teenage years [3]. Type 2 diabetes is characterized
by defective insulin secretion or use that may occur with excessive hepatic gluconeogenesis. Both genetic and environmental factors play a role in the development of this disease, which until recent years, was generally regarded as a
disease of adult onset. At its current rate of increase, type 2 diabetes is projected
to develop in a quarter to one third of the American population within the next
25 years. The increasing incidence of obesity and a sedentary lifestyle in our
society has been blamed for causing the prevalence of type 2 diabetes to surge,
and for the disease to have its onset at a younger age. Although patients with type
2 diabetes may or may not require insulin to optimize their care, the systemic complications of the disease remain the same as with type 1 diabetes [4].
Hepatic and peripheral resistance to insulin result in increased hepatic glucose
production, and decreased muscle uptake of glucose [3].
Diabetics undergo interventional procedures and surgery more commonly than
their non-diabetic counterparts. End-organ diseases resulting from diabetes often
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
result in the need for major surgical intervention, such as vascular and coronary
bypass procedures, amputations, and renal and pancreas transplants. Coronary
artery disease, peripheral vascular disease, renal insufficiency, gastroparesis, and
autonomic neuropathy are common sequelae of longstanding diabetes and should
be identified during preoperative evaluation. The multisystem destructive effects
of diabetes places these patients at greater perioperative risk for complications
such as stroke, myocardial infarction (MI), and worsening renal insufficiency.
Wound complications and infection related to poor healing are also more likely
than in nondiabetics. Recent data suggests control of postoperative hyperglycemia may be beneficial in cardiac surgical and critically ill patients.
Coronary artery disease is prevalent in longstanding diabetics. Type 1diabetes
may manifest symptoms of coronary artery disease at a young age. Diabetics with
coronary disease frequently have atypical symptoms of ischemic heart disease,
and often have silent ischemia. Perioperative MI has an associated mortality rate
of 40% to 70% and is greater for diabetics than their non-diabetic counterparts.
Diabetics who sustain a perioperative infarct are also less responsive to therapeutic intervention than their non-diabetic counterparts.
Significant cardiac disease can be identified through the clinical history, including an assessment of functional capacity, physical examination, 12-lead
electrocardiogram and individualized diagnostic evaluation based on initial
findings and risk stratification. Patients unable to perform a 4 metabolic
equivalent (MET) workload equivalent are at increased risk for perioperative
cardiac events. Major, intermediate, and minor clinical predictors have been
identified and are reviewed in the article by Dr. Fleisher in this issue. The
presence of diabetes mellitus is classified as an intermediate clinical predictor for
increased perioperative risk for adverse cardiovascular events (MI, congestive
heart failure, or death). The presence of additional risk factors, such as hyperlipidemia, smoking, hypertension, and a family history of heart disease compound those odds.
Preoperative electrocardiograms may reveal evidence of ischemia, prior MI,
rhythm disturbance, or conduction delay. An abnormal electrocardiogram is more
informative when an earlier study is available for comparison, and is more useful
in patients who are classified as intermediate and high risk.
The practice of evidence-based medicine forces us to critically examine the
efficacy of treatment plans in a scientific manner. For example, class I evidence,
based on accepted prospective randomized trials, exists for obtaining an EKG for
all patients who have an intermediate risk factor including diabetes, scheduled for
an intermediate or high-risk procedure [5].
The American College of Cardiology/American Heart Association (ACC/
AHA) updated its guidelines in 2002 for the perioperative cardiovascular
evaluation of patients undergoing noncardiac surgery [5]. The committee emphasized that few patients will require revascularization procedures to minimize
their risk before undergoing noncardiac surgery, unless the revascularization
would have been warranted irrespective of the surgery being planned. The need
for further preoperative cardiac evaluation is tailored to the circumstances and is
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
dependent on a number of variables, including the presence or absence of clinical
predictors of disease in concert with the degree of risk of the surgical procedure.
If a revascularization procedure is not deemed necessary, patients should be
considered for perioperative medical therapy.
Although often underused in the diabetic because of concerns about limiting
the response and recognition of hypoglycemia, class I evidence exists that
patients who have been on beta-blockers in the recent past for the treatment of
angina, hypertension, or arrhythmias, should receive them perioperatively. In addition, diabetics have been shown in a large retrospective database review, to
have improved long-term survival when maintained on chronic beta-blockade
after sustaining a MI [6]. Beta-blockade is also beneficial for high-risk patients,
including those with diabetes before undergoing major vascular surgery. A
class II recommendation, based on less rigorous studies than those qualifying
as Class I (Class IIa data is still collected prospectively or is clearly reliable data
from retrospective analyses), advises the use of beta-blockers in the perioperative
period for patients with untreated hypertension, major risks for, or known coronary artery disease, in the absence of any contraindication to beta-blockade [6].
In an emergency, one generally should proceed to the operating room and
moderate risk as able. In an elective situation, if a patient has had a recent MI and
post infarct risk stratification does not indicate that there is significant myocardium at risk, the ACC/AHA committee advises that it may be best to wait for 4 to
6 weeks before proceeding with the procedure. However, there are not many
clinical trials that support a firm guideline on what timeframe would be best [5].
Many diabetics are treated with angiotensin converting enzyme (ACE)
inhibitors as part of an antihypertensive regimen or to limit or potentially reverse
albuminuria, because ACE inhibitors and tight glycemic control have been found
to reduce the rate of progression of diabetic nephropathy [4]. Patients chronically
taking ACE inhibitors have been reported to develop hypotension refractory to
ephedrine administration while undergoing general anesthesia [7]. Preoperative
instructions ordinarily advise the patient to continue their antihypertensive
medications on the morning of surgery. Although this is still an area of controversy, consideration should be given to advising the patient to withhold the
ACE inhibitor on the day of surgery. This might be especially pertinent in the
longstanding diabetic, who is more likely to have diffuse vasculopathy and hence
may be more at risk for ischemic complications should problematic hypotension
develop during anesthesia.
The presence of significant renal disease may make the patient prone to
volume overload intraoperatively. Crystalloid solutions containing potassium
should be avoided in the diabetic with advanced renal disease and hyperkalemia.
The effects of renally excreted medications such as aminosteroid neuromuscular
blocking agents may be prolonged and should be judiciously dosed, monitored,
and reversed. The half-life of insulin is also prolonged, which may make the
patient more prone to hypoglycemia [2]. Finally, patients with baseline renal
insufficiency are more prone to develop acute renal failure in the perioperative
period [8].
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Blood glucose and outcome
The traditional approach to diabetic treatment in the hospital focused on maintaining modest control of hyperglycemia with the philosophy that aggressive
treatment and the risk of hypoglycemia and neuroglycopenic insult were potentially more detrimental than moderate hyperglycemia. Blood glucoses in the
200-mg/dL range were considered reasonable. However, it was subsequently
found that blood sugars greater than 200 to 250 mg/dL are more likely to be
associated with adverse outcomes. Leukocyte function and the ability of immunoglobulins to fix complement have been shown to be impaired with blood
glucoses in this range [9]. The diffuse vascular disease that is often present in
diabetics also impairs the ability of the circulation to deliver adequate oxygen and
substrates to the injured area. Interestingly, short-term hyperglycemia has been
shown to impair the ability of the endothelium to vasodilate. Studies have shown
that hyperglycemia increases endothelial release of endothelin-1, a potent vasoconstrictor, and may reduce availability of the vasodilator nitric oxide [10].
Hyperglycemia inhibits collateral coronary blood flow and inhibits the development of new collateral vessels [11]. Recent studies have indicated that patients
treated with intensive regimens (‘‘tight control’’) that result in euglycemia versus
standard regimens, have lower rates of morbidity and mortality [11,12].
Mortality rates have been found to be directly correlated with the degree of
elevation of blood glucose or HbA1c levels in both critically ill patients and in
those admitted with acute MI [12]. The diabetes and insulin glucose infusion in
acute myocardial infarction (DIGAMI) trial found mortality rates of 35%, 40%,
and 55% in patients presenting with blood glucose concentrations of less than
235, 235 to 298, and greater than 298 mg/dL respectively, at the time of admission with an MI. This compares quite unfavorably to the 9% mortality rate
found previously for patients presenting with normal blood glucose values at the
time of admission for acute MI [13].
Surprisingly, increased mortality rates have also been noted in the presence of
blood glucose ranges that were previously considered to be only mildly elevated.
Fasting blood glucose levels of even 110 mg/dL are associated with an increased
relative risk of cardiovascular events, such as cerebrovascular accident, MI, and
sudden death [11]. More importantly, tight postoperative glycemic control maintaining blood glucose in the range of 80 to 120 mg/dL has been shown to
significantly decrease mortality in these patient populations with no noted adverse consequences, compared with traditional, less aggressive diabetic management (eg, maintaining blood glucose in the range of 180 to 200 mg/dl) [11,13].
Although it is easy to extrapolate this data to the intraoperative management of
the diabetic, clinical trials substantiating a benefit from tighter operative glycemic
control are lacking at this time.
Airway considerations in the diabetic
One third of patients with longstanding type 1diabetes may be difficult to
intubate [14]. Diffuse glycosylation of proteins in patients with longstanding
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
diabetes may result in ‘‘stiff joint syndrome.’’ This is manifested by the appearance of tight, waxy skin, short stature, and joint rigidity. Inability of the patient to approximate the palms and fingers together is called the ‘‘prayer sign.’’
This is believed to be a manifestation of stiff joint syndrome and an indication
that the patient may be difficult to intubate. Stiff joints in the diabetic result in
decrease of atlantoaxial mobility. This may result in difficult laryngoscopy [14].
Gastroparesis is also more common in patients with autonomic neuropathy.
Awake fiberoptic intubation may be a safer way to proceed with these patients,
who thus have a combination of a potentially difficult airway and are at greater
risk for aspiration. Sodium citrate, 30 mL orally just before proceeding to the
operating room will help to decrease gastric acidity, and metoclopramide 10 mg
intravenously may be administered a half-hour before induction to improve gastric emptying. H2 blockers such as ranitidine or famotidine may also be administered, but require more than 45 minutes to reduce gastric acidity. Patients with
diabetes on occasion will develop a neuropathy of the vagus and the recurrent
laryngeal nerves, resulting in bilateral vocal fold immobility. Symptoms of this
disorder include the insidious onset of dysphonia and stridor [15].
Autonomic dysfunction
Autonomic neuropathy has been reported to be present in 20% to 40% of
diabetics [16]. Identifying the presence of autonomic neuropathy preoperatively
may influence the anesthetic plan. The presence of the prayer sign, peripheral
neuropathy, orthostatic changes in blood pressure, loss of normal respiratory
variations of the heart rate, and resting tachycardia are suggestive findings on
physical examination that are simple and expeditious to perform [17]. Patients
with autonomic neuropathy are more likely to have problems with intraoperative
blood pressure lability [17]. Invasive arterial monitoring may be helpful to more
closely monitor such blood pressure variability.
A study by Kitamura and colleagues [16] found that diabetics with autonomic
neuropathy developed significantly lower core temperatures under general
anesthesia for abdominal surgery than control nondiabetic patients and diabetics
without autonomic neuropathy. Patients with autonomic neuropathy were found
to have impaired vasoconstriction, which was felt to be responsible for the predisposition to hypothermia. Perioperative hypothermia has been found to be
associated with complications such as poor wound healing. More aggressive intraoperative measures to maintain normothermia may be warranted.
Preoperative instructions
Perioperative control of diabetes is complicated by the induction of counterregulatory hormones by surgical stress, infection, and concurrent corticosteroid
administration. This is compounded by a situation where the patient is either
NPO, unable to tolerate their normal diet, or has sporadic caloric intake. Ideally,
diabetic patients should have their surgeries scheduled for early in the morning to
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
reduce the time period of fasting and to minimize the disturbances in the patient’s regimen.
Type 1 diabetics
Patients with type 1 diabetes will always require at least basal amounts of
insulin, even when fasting. This prevents the onset of ketosis. Regular insulin
is usually held on the morning of surgery, unless the blood glucose is greater
than 200 mg/dL. Sliding scale insulin should not be used as a sole substitute
for an insulin infusion or a partial dose of intermediate- acting insulin. Doing
so risks precipitating ketoacidosis [18]. Table 1 [3,17] reviews various insulin
preparations whereas Table 2 [9,17] provides guidance on insulin dosing pre
and intraoperatively.
External continuous subcutaneous insulin pumps administer insulin preparations through a subcutaneous needle. A reservoir is filled with short-acting
insulins, either lispro, aspart, regular insulin, or buffered regular insulin (Velosin
SR). The pump may be programmed to have a variable output throughout the day
and night, and can administer boluses [18]. These short acting insulins can be
discontinued, and the patient converted to a regular insulin infusion perioperatively when judged appropriate (Box 1) [17].
Type 2 diabetes
Type 2 diabetes is frequently associated with obesity and is characterized by the
presence of insulin resistance. Patients may be treated with diet, oral medications,
insulin or combinations of the three. The goals of medical therapy of type 2
diabetes are to stimulate insulin production by the pancreas, decrease peripheral
and hepatic resistance to insulin, and modulate hepatic gluconeogenesis [3].
Five classes of oral hypoglycemic agents are currently available and highlighted in Box 2 [3,9,11,17 –19].
Several classes are associated with increased risk of perioperative hypoglycemia, including sulfonylureas, glitazone compounds, and the biguanide, metformin. With the exception of acarbose, the alpha-glucosidase inhibitor, patients
should be instructed to routinely hold their oral agent on the morning of surgery.
Type 2 diabetics with inadequate glycemic control with diet and oral agents
require the addition of insulin to their regimen. Type 2 diabetics with insulin
resistance may require much higher doses of insulin than expected to achieve
euglycemia. Diet controlled type 2 diabetics may be able to maintain normoglycemia perioperatively simply by avoiding dextrose in intravenous fluids [18].
However, many patients with type 2 diabetes who did not require insulin at home
may require insulin perioperatively to maintain glycemic control [18].
Subcutaneous insulin is generally not advised for intraoperative glucose
control because of its potentially erratic absorption secondary to altered regional
blood flow, tissue edema, or fluid shifts during surgery [20]. If an insulin infusion
(see Box 1) is chosen as the method to control blood glucose intraoperatively in
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Table 1
Insulin preparations and guidelines
Insulin lispro
Insulin aspart
(Novolog) [3]
The altered amino acid
sequence of these analogs
favors rapid subq absorption
and rapid onset. Usually
administered immediately
prior to a meal. May be used
subq or via an insulin pump,
but not recommended for
continuous infusion
Short acting
Regular insulin Administered around 30 – 60
min before meals. Main
insulin used in continuous
iv infusions.
NPH (neutral Made by adding protamine
to insulin
Made by adding zinc to an
acetate buffered solution
of insulin
Zinc suspension of insulin.
Its peak levels are less
prominent than those of NPH.
Another insulin analog, It
most closely provides a
constant, basal level of insulin.
It is administered at 10 PM.
It cannot be mixed with other
types of insulin in the same
syringe. It is less likely to
cause hypoglycemia. [3]
Although long-acting
insulins are often held or
halved in dose before
surgery, glargine may be
administered as usual to
provide basal insulin levels
during surgery. [17]
Ultralente and glargine are
used as basal insulin regimens
70/30: 70% NPH/30% regular
Premixed/combination Premixed
formulas help 50/50: 50% NPH/ 50% regular
to reduce the
75/25: 75% insulin neutral
protamine lispro (NPL, similar
likelihood of
mixing errors. to NPH) and 25% insulin lispro.
This mixture is intermediate
in action and is usually
administered twice a day,
and is also known as
Humalog mix.
Onset 5 – 15 min
Peak effects:
60 – 120 minutes
Duration: 4 – 5 hours. [3]
Onset 30 – 60 mins
Peak: 2 – 4 h
Duration: 6 – 8 h
Onset 1 – 3 h
Peak 4 – 6 h
Duration 12 – 14 h
Onset 1 – 3 h
Peak 4 – 8 h
Duration 12 – 20 h
Onset 2 – 4 hours.
Peak 14 – 18 hours.
Lasts 18 – 24 hours.
Onset of 1 – 2 h
Duration: 20 – 24 h
70/30 usually given
before breakfast
50/50 before dinner
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Table 2
Insulin protocols on day of surgery
Insulin infusion
Hold all insulins the morning of surgery. Obtain a blood glucose and start an
insulin infusion at 1 – 2 units per h, along with D5W or D5 0.45 NS at
75 – 100 cc/h, approximately 5 grams of glucose per hour.
Administration of glucose during surgery helps to prevent ketosis,
hypoglycemia, and catabolism. [17]
The dextrose solution is not intended for volume replacement. Any additional
fluid necessary for volume resuscitation should not contain dextrose.
Blood glucose should be checked every 1 – 2 h, and the insulin infusion
adjusted to achieve a blood glucose between 100 – 150 mg/dl [9]
Intermediate-acting Give half to two thirds of intermediate to long acting insulin on the morning
insulin use
of surgery.
Dose with regular insulin intravenously, from 1 – 4 units/h, with a goal blood
glucose of 100 – 150 mg/dL [9]
Glucose- insulin-potassium (GIK) regimen: Patients with normal renal function
supplementation and normal potassium levels may receive dextrose containing fluids with
additional potassium (10 – 20 mEq/L) in addition to the insulin infusion.
Insulin pump
Options are to turn the pump off and use a continuous insulin infusion or
continue pump at a basal rate supplemented with dextrose and potassium as
needed with rate adjustment based on serial blood glucose measurements.
the type 2 diabetic, the patient should receive their last normal dose of insulin the
evening before surgery, and the insulin infusion should generally be initiated
around 2 hours before surgery to allow for equilibration (See options in Table 2).
In the immediate postoperative period, blood glucose should be checked every
1 to 2 hours for several hours in type 1 diabetics and every 4 hours for type 2
diabetics [17]. Dextrose should be infused and insulin administered as directed by
laboratory data.
The sympathetic response provoked by surgical trauma results in neuroendocrine changes, including increased serum levels of catecholamines, adrenocorticotrophic hormone (ACTH) and cortisol. The stress reaction thus produces an
environment that favors catabolism and gluconeogenesis. The relative insulin
resistance and insulin deficiency invoked by surgical stress predisposes to hyperglycemia and also enhances lipolysis, which may lead to ketosis and acidosis in
type I diabetics [21].
The renal threshold for glucose resorption in normal kidneys is between
10 mmol/L and 11.1 mmol/L (180 mg/dL) [21]. Higher glucose levels may
result in osmotic diuresis and compromised wound healing and therefore require
appropriate insulin therapy.
Diabetic ketoacidosis and surgery
Acutely ill type 1 diabetics who present for emergency surgery may present in
diabetic ketoacidosis. Diabetic ketoacidosis may occur even in the absence of
significant hyperglycemia, provoked by inadequate availability of insulin at a
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Box 1. Insulin infusion protocol
Standard insulin infusion orders
1. Stop all previous insulin orders.
2. Draw stat serum glucose and compare to glucose meter
measurement within 15 minutes.
3. Prepare solution of regular insulin 250 units/250 mL
(1 unit/1 mL) in:
(Check one) __ Dextrose 5% or __sodium chloride 0.9% (Use if
diabetic ketoacidosis is present; may switch to Dextrose 5%
when glucose is <250 mg/dL.)
a. Run 50 mL of the solution through the tubing and waste.
Should also be done with each tubing change.
b. If admitted to a medical/surgical unit use the micro drip
infusion pump.
c. Target range for blood glucose __– __ mg/dL (suggest
100– 160).
d. If blood glucose is not decreasing by at least 50 mg/dL/h,
notify physician.
e. If blood glucose is less than 200 mg/dL, initiate drip
at __ units/h (suggest 1– 2 units/h).
f. If blood glucose is greater than or equal to 201 mg/dL,
administer IV bolus of __ units regular insulin
(suggest 2 – 4 units).
g. Check blood glucose after 1 hour.
Maintenance titration
1. Do not stop the infusion!
2. Perform glucose meter measurement every hour, adjust
infusion rate as follows:
<70 mg/dL Decrease rate by __units/hr (suggest 1– 2 unit/h)
and administer 12.5 G (1/2 vial) of D50 IV (25 mL).
70 – 100 mg/dL Decrease rate by __ units/h (suggest
0.5-1.0 unit / hour).
100 – 130 mg/dL No change.
131 – 160 mg/dL Increase rate by __ units/h (suggest
0.5 unit/hour).
161 – 200 mg/dL Increase rate by __ units/h (suggest
1.0 unit/hour).
201 – 250 mg/dL Increase rate by __ units/h (suggest
1.5 unit/hour).
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
>251 mg/dL Increase rate by __ units/h, give __ unit IV
bolus (suggest 2 units/h & 2 units bolus).
Notify MD
3. When feasible, all other drips should be in normal saline.
4. When glucose remains low (<70 mg/dL) for two consecutive
tests please page MD.
5. If the glucose is consistently within the goal and the infusion
rate is stable, glucose meter measurements may be changed
to every 2 –4 hours.
1. For conversion to subcutaneous insulin, refer to the new
insulin orders.
2. Infusion may only be stopped 30 minutes after the administration of the scheduled subcutaneous insulin dose.
time when there is increased demand [17]. Type 2 diabetics are far less ketosis
prone, but elderly patients with poor oral intake may develop a nonketotic, hyperglycemic, hyperosmolar state [17]. Urine or serum ketones should be measured if the blood glucose is greater than 240 mg/dL in a type 1 diabetic patient
[18]. Therapy of evolving diabetic ketoacidosis centers on identifying and
eliminating the initiating event, providing adequate insulin therapy, fluid resuscitation and correction of electrolyte abnormalities [22].
Perioperative glycemic control
Several recent studies resulted in various clinicians advocating euglycemia as
the goal of perioperative glucose control. The single center prospective controlled
study by Van Den Berghe and colleagues in 1548 patients compared euglycemic
or ‘‘tight control’’ of blood glucose between 80 and 110 mg/dL with conventional
therapy. Investigators only initiated insulin therapy in the control group if the
blood glucose was greater than 210 mg/dL, aiming for a treated level of 150 to
180 mg/dL. Most of those reported were surgical patients with cardiac surgical
patients representing roughly 60% of enrollees. The tightly controlled group
required insulin infusions quite often, but had significant decrease in mortality
(4.6% versus 8% for controls). The effect was most notable in patients who
required prolonged intensive care, longer than 5 days. For this group, patients had
fewer blood stream infections, a lower incidence of renal failure requiring
dialysis, required less red cell transfusion therapy, and had a lower incidence of
critical illness polyneuropathy. Various mechanisms have been proposed to
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Box 2. Oral hypoglycemic agents
Sulfonylureas: First generation long acting sulfonylureas: chlorpropamide (Diabinese), tolbutamide (Orinase), tolazamide (Tolinase), acetohexamide (Dymelor). Second-generation sulfonylureas
include glyburide (Diabeta, Micronase), glipizide (Glucotrol), and
Sulfonylureas work by closing the adenosine triphosphate regulated potassium (Katp) channel in pancreatic islet cells.
Closure of this channel results in insulin release. This mechanism is clinically relevant because ischemic preconditioning
(whereby myocardium that has been subjected to brief episodes
of ischemia is more resistant to infarction) is dependent on
activation and resultant opening of Katp channels. This physiology is a potential explanation for the increased mortality rate
noted in diabetic patients being treated with sulfonylureas after
angioplasty for acute myocardial infarction, and their worsened
appearance of myocardial ischemia as determined by dipyridamole stress endocardiography [11].
Sulfonylureas become less effective with progression of the
disease [3].
Long-acting sulfonylureas are best discontinued 48 –72 hours
before surgery [17].
Shorter acting sulfonylureas may be held the night before or
the morning of surgery, to minimize the risk of perioperative
hypoglycemia [17].
Meglitinides are nonsulfonylurea insulin secretagogues such
as repaglinide (Prandin), and nateglinide (Starlix), and have a short
duration of action geared towards treating postprandial hyperglycemia [3].
Insulin sensitizers act to decrease hepatic gluconeogenesis and
reduce peripheral insulin resistance and increase insulin uptake into
muscle. Thiazolidinediones and biguanides are the two classes of
insulin sensitizer currently available.
Thiazolidinediones: rosiglitazone (Avandia) piaglitazone (Actose)
These drugs improve insulin uptake and decrease hepatic gluconeogenesis by activating genes that stimulate transcription
for proteins that enhance cellular insulin action.
The onset of action of these drugs is about 6 weeks. They may
result in preserved pancreatic beta cell function and improve
endothelial function.
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Another drug in the thiazolidinedione class, trogliazone, was
associated with cases of hepatic failure and was removed from
the market. Sporadic cases have been associated with rosiglitazone also. It is recommended that liver function be assessed every 2 months in patients taking this class of drugs
[3], thus it may be wise to obtain liver function studies in the
perioperative period.
These may be given on the morning of surgery, although missing a dose will not have much of an effect because of the long
duration of action of this class of drugs.
Thiazolinediones may also be continued postoperatively as
long as hepatic function and cardiac function are normal. The
liver metabolizes them.
The biguanide, phenformin, was banned in the United States in
the 1970s because of the potential development of lactic
Metformin (Glucophage), the only biguanide currently available
in the US, is commonly administered as part of a treatment
regimen for type 2 diabetes. Although it’s association with
lactic acidosis is much less common than that of phenformin,
lactic acidosis may occur in the perioperative period with
patients receiving this drug [19]. Risk factors for developing
lactic acidosis related to metformin use include the presence
of renal or hepatic insufficiency, a fasting state, surgery, and
age >60 years [19].
Metformin should be discontinued at least 24 h prior to surgery
or before performing any studies involving administration of
radiographic contrast [3].
Metformin should be held for 48 h after major surgery until it is
clear that the patient has not suffered a decline in renal function postoperatively [9].
Alpha-glucosidase inhibitors (acarbose [Precose], miglitol [Glyset])
These agents act by blocking the breakdown and absorption of
complex carbohydrates, and are effective only when administered with food. This helps to reduce postprandial hyperglycemia, but has no effect on fasting blood glucose [3,9,18]. It
should be noted that oral sucrose, maltose, and starch will not
be effective for treatment of hypoglycemia in patients taking
this drug [3].
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
explain the benefits of normal blood glucose. These include maintenance of
normal white blood cell and macrophage function, positive trophic and anabolic
effects of insulin, improved erythropoiesis and decreased hemolysis, reduced
cholestasis, and less axonal dysfunction and damage associated with hyperglycemia and insulin deficiency. A recent retrospective single center study by Krinsley
corroborated Van Den Berghe’s work in a larger, but more diverse medical and
surgical evaluation of over 1800 patients [23]. He showed that even a modest
increase in the mean blood glucose level above normal in critically ill medical,
surgical, and coronary care patients substantially increased hospital mortality.
Other evidence supporting the deleterious effects of hyperglycemia in postoperative patients include reports of decreased sternal infections in cardiac surgical patients who had tighter intraoperative and postoperative glucose control,
improved survival in acute MI patients who had their blood glucose controlled,
and worsened outcome after stroke in patients who were concurrently hyperglycemic [24 – 26].
More data is likely to be forthcoming regarding the potential benefits of
euglycemia in the perioperative period and a better understanding of the exact
mechanisms that generate improved outcome and limited morbidity.
Thyroid disease
Hypothyroidism is a common condition in the United States, affecting approximately 1% of all patients and 5% of the population over age 50 [27]. Hypothyroidism develops 10 times more often commonly in females than males. The
most likely cause of hypothyroidism is iatrogenic secondary to either surgical
resection or radioactive ablation of the gland as part of treatment for hyperthyroidism. Hashimoto’s thyroiditis is the most common noniatrogenic cause of
hypothyroidism. If Hashimoto’s is present, one must seek other autoimmune diseases such as systemic lupus erythematosus, rheumatoid arthritis, primary adrenal
insufficiency (AI), pernicious anemia, diabetes mellitus, and Sjogren’s syndrome.
Clinical findings
The diagnosis and treatment of hypothyroidism are facilitated by timely
history, physical examination and application of modern thyroid hormone assays.
Fatigue, memory loss, headaches, thinning hair, lethargy, constipation, cold
intolerance, and anorexia are the most common symptoms whereas weight gain
may be the most noticeable of the multitude of presenting physical signs.
Hypothyroidism should be considered in inactive elderly patients who have
become increasingly withdrawn and selected patients who live sedentary lives.
Cardiovascular effects of hypothyroidism include a decrease in cardiac output, as
a result of a decreased stroke volume and heart rate, caused by the loss of the
inotropic and chronotropic effects of thyroid hormone. The decrease in circula-
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
tion to the skin results in its cool and pale appearance. Pericardial effusion may be
present with severe hypothyroidism, but will rarely result in tamponade. Patients
with hypothyroidism may also have obstructive sleep apnea and depressed
respiratory drive. Pleural effusions may be observed. Renal water excretion is
impaired, resulting in hyponatremia and an increase in total body water, although
plasma volume tends to be reduced [28].
Treatment of hypothyroidism
Correction of the hypothyroid state entails thyroid hormone replacement, most
commonly with tetraiodothyronine (T4) (levothyroxine). Triiodothyronin, T3, the
most active thyroid moiety, is subsequently formed in the body by intracellular
conversion of T4 to T3. Levothyroxine, a common replacement agent, has a halflife of 6 to 7 days; thus missing a morning dose on the day of surgery has little
impact on the patient [29]. Levothyroxine prescriptions seem to be prone to
transcription errors. Prescribers occasionally will misplace the decimal point
when ordering levothyroxine in milligrams, resulting in a 10-fold error in dosage.
Errors have also been reported when the dose has been converted to micrograms
from milligrams. It has been suggested that levothyroxine be ordered in micrograms, rather than milligrams, to avoid decimal point errors and conversion
errors. The anesthesiologist needs to be alert to the possibility of thyroid medication errors.
Intravenous forms of levothyroxine are available, if a patient is unable to take
oral medication for a prolonged period of time, say greater than 5 to 7 days. The
intravenous dose is 100% bioavailable, whereas oral doses are only 50% bioavailable. Thus, when converting from oral to intravenous doses of levothyroxine, the dose should be halved.
The hypothyroid patient with coronary artery disease
Timing of initiation of thyroid hormone replacement in patients with coexisting coronary artery disease and hypothyroidism remains an area of controversy
and clinical judgment. The competing concerns involve the risk for precipitating
unstable angina or an MI by increasing the patient’s metabolic rate and cardiac
work with levothyroxine, versus precipitating myxedema coma, heart failure, or
neurologic complications of severe hypothyroidism should the patient proceed
for cardiac intervention before hormone replacement is initiated. It is generally
believed that most patients may proceed with cardiac surgery before replacing
thyroid hormone [30].
The surgical patient with hypothyroidism
There is not convincing evidence that mild to moderate hypothyroidism
necessitates postponing elective surgery. Elective surgery should be delayed for
patients with more advanced, severe hypothyroidism. If a patient with severe
hypothyroidism (manifestations being myxedema coma, pericardial effusion,
heart failure or extremely low levels of thyroid hormone) requires urgent surgery,
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
they should receive intravenous T4 or T3 perioperatively, in addition to glucocorticoids as a certain proportion of patients with hypothyroidism may also have
occult AI. If urgent treatment is deemed necessary, 200 to 500 micrograms (mcg)
of intravenous levothyroxine (T4) may be administered slowly. Subsequent daily
replacement doses are 50 to 100 (mcg) intravenously (IV) per day [20]. Treatment
of the hypothyroidism without providing steroid replacement may precipitate
adrenal crisis [20,30].
Myxedema coma
Patients may develop myxedema coma postoperatively [30]. Myxedema coma
is uncommon but carries a significant mortality rate of up to 60%. It can be
precipitated in a patient with hypothyroidism by several medications, events or
environmental factors, including hypothermia, trauma, infections, cerebrovascular accidents, anesthetics, sedatives and analgesics, amiodarone, and lithium carbonate [27]. Myxedema coma is manifested by depressed mental status, delirium
or coma, hypothermia, bradycardia, and hypopnea. CO2 narcosis caused by hypoventilation may also be contributory to the changes seen in mental status.
Pericardial and pleural effusions may be present. Myocardial contractility is
decreased, resulting in a diminished stroke volume and low cardiac output.
Cardiac tamponade may also occur in the presence of a significant pericardial
effusion. Echocardiography should be considered in hemodynamically unstable
patients. Despite an increase in total body water, intravascular volume is
decreased, which in combination with the depressed cardiac output makes the
patient susceptible to hypotension should they become vasodilated. External
warming is thus inadvisable because the resulting vasodilation may precipitate
cardiovascular collapse. Laboratory findings include an elevated TSH and
depressed levels of free T4 and T3, with the exception of patients who have
hypothyroidism caused by pituitary disease whereupon TSH levels will be
normal or low. Hyponatremia is common. This can also exacerbate mental status
changes and occasionally promote seizure activity. Hypoglycemia can be present
and may be a manifestation of associated AI, because of autoimmune causes or
secondary AI caused by pituitary disease. Initiation of levothyroxine replacement
can be complicated by the onset of arrhythmias. T4 is converted to T3 in the
periphery by deiodinases, but the activity of these enzymes is reduced in the
presence of hypothyroidism. Since T3 is the active form of thyroid hormone,
initiating thyroid hormone replacement with T4 rather than T3 will result in a
more gradual onset of effect. However, using T3 alone for acute replacement may
precipitate arrhythmias because of its more rapid onset. Different protocols have
thus been advised, with varying combinations of T3, T4, or both, depending on
the age of the patient and the suspicion for the presence of coronary artery
disease. Initial loading doses of intravenous T4 are generally in the range of
200 to 500 mcg. Typical (low) doses of intravenous T3 are 10 mcg every 8 hours,
until the patient regains consciousness. In general, the use of intravenous T3
alone for replacement is not recommended [27].
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Grave’s disease is an autoimmune disorder that results in excessive thyroid
hormone production because of abnormal stimulation of the thyroid gland by
thyroid stimulating hormone (TSH) receptor antibodies. Hyperthyroidism may be
clinically overt, manifested by signs and symptoms of hyperthyroidism (tachycardia, tremor, weight loss), goiter, and opthalmopathy. Gastrointestinal symptoms,
such as nausea, vomiting, and diarrhea may also be present. At times, patients may
only have subclinical or biochemical evidence of hyperthyroidism. Subclinical
hyperthyroidism is manifested by increased nocturnal pulse rates, frequent atrial
premature beats, or in the elderly, the onset of atrial fibrillation. Total and free T4
and T3 levels are normal or elevated, but TSH is suppressed [28]. Postpartum
patients may also develop relapses of Grave’s disease. The anesthesiologist
needs to remain alert to the possibility of an undiagnosed case of thyrotoxicosis
in patients with unexplained tachycardia, ectopic beats, fever, tremor or other
suggestive signs and symptoms. Laboratory evidence of thyrotoxicosis includes an
elevated free T4 level and suppressed TSH levels. Some patients will have normal
free T4 levels, but will have suppressed TSH levels and an elevated T3 level [28].
Treatment of hyperthyroidism
Surgical treatment of Grave’s disease is not as common as it was in the past,
with options for medical treatment and radioactive iodine being available, safe,
and effective. Radioactive iodine is usually the treatment of choice, and generally
achieves a euthyroid state within 6 to 18 weeks. The incidence of hypothyroidism
after radioiodine treatment is at least 50% by 10 years after treatment [28].
Pregnant women presenting with hyperthyroidism are generally placed on medical regimens because of the risks of surgery and the concerns of using radioactive
iodine for the fetus.
Beta-blockers, antithyroid medications, and iodides are used in various combinations for the patient being medically treated for hyperthyroidism. Goals of
medical treatment are to reduce symptomatology caused by beta-adrenergic
excess with beta-blockers, and to reduce thyroid hormone production with the
antithyroid agents and iodides. Pregnant women presenting with hyperthyroidism
are generally placed on medical regimens because of the risks of surgery and the
concerns of using radioactive iodine for the fetus.
Beta-blockers inhibit peripheral conversion of T4 to T3, the most active thyroid
hormone, and help to limit the adrenergic effects of hyperthyroidism. The goal of
beta-blockade should be to achieve control of the heart rate to less than 90 [20].
Propranolol or more beta-1 selective agents such as atenolol or metoprolol may be
used [28].
Antithyroid medications such as propylthiouracil (PTU) and methimazole are
actively transported into the thyroid gland and decrease thyroid hormone
synthesis. PTU also inhibits peripheral conversion of T4 to T3, although this
effect may not be clinically significant. Starting doses of PTU are 100 to 300 mg
by mouth daily, and are generally adjusted downwards as treatment progresses.
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Methimazole achieves a euthyroid state more rapidly than PTU, and is less likely
to be associated with agranulocytosis, hepatitis, or vasculitis. Initial doses of
methimazole are 10 to 20 mg by mouth daily [20,28,30]. Thionamides only block
formation of new thyroid hormone. Iodines block the release of thyroid hormone
from the colloid space [30].
Thyroidectomy is considered when patients have a large goiter, have failed or
are intolerant of antithyroid medications, or refuse radioactive iodine. Surgery is
also performed in patients with Grave’s disease that have cold nodules and have a
suspicion for thyroid cancer. A euthyroid state should be achieved before surgery
to avoid precipitating thyroid storm [28].
The hyperthyroid patient presenting for surgery
Patients with medically treated hyperthyroidism should take their antithyroid
medications on the morning of surgery [29]. In the event of the need for emergent surgery in a patient who has thyrotoxicosis, emergency preparation entails
administering an antithyroid medication, such as PTU or methimazole (Tapazole), followed by iodide. Iodides should not be administered first, as supplemental iodide provides more substrate for thyroid hormone synthesis. This can
result in an increased amount of thyroid hormone released from the thyroid gland
and potentially precipitate thyroid storm. One should wait for 2 to 3 hours after
administering the thionamide before initiating iodines [27].
PTU may be administered as a loading dose of 1 g orally or by nasogastric
(NG) tube and later followed by a dose of 200 mg orally or NG every 6 hours.
Iopanoic acid, an oral contrast agent, (Telepaque) is considered the iodine of
choice. This is administered as 1 gram every 8 hours for the first 24 hours, then
500 mg twice daily. Lugol’s iodine or saturated solution of potassium iodide
(SSKI) are alternatives. These are administered as oral drops in a dose of
4 to 8 drops every 6 to 8 hours [27]. Patients with thyrotoxicosis are also at
risk for AI and should receive stress doses of corticosteroids [20,30]. Glucocorticoids also have the effect of decreasing peripheral conversion of T4 to T3.
Anesthetic agents that are vagolytic or sympathomimetic should be avoided in
patients with thyrotoxicosis.
Thyroid storm
Thyroid storm may occur in the perioperative period in patients who have
undiagnosed or undertreated hyperthyroidism. It is also more commonly seen in
poor or underserved populations who have limited access to medical care [27].
Malignant hyperthermia and pheochromocytoma crisis are in the differential
diagnosis. Thyroid storm is recognized by the onset of fever, tachycardia, and
delirium, which may progress to cardiovascular collapse and death. It may be
necessary to treat the patient based on clinical suspicion alone while waiting for
confirmatory thyroid function tests. The laboratory value itself does not distinguish between hyperthyroidism and thyroid storm per se. Thyroid storm remains
a clinical diagnosis. Thyroid storm carries a mortality rate of 10% to 75% and
mandates treatment in the intensive care unit (ICU) [27].
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Treatment encompasses thionamides, beta-blockers, antipyretics, nutritional
support with dextrose and vitamins, and treatment of cardiac complications such
as atrial fibrillation and high output heart failure. Acetylsalicylic acid should not
be used as an antipyretic in patients with thyroid storm, as it interferes with the
protein binding of T4 and T3 and can thus increase free thyroid hormone
concentrations [30]. Methods to enhance thyroid hormone clearance may also
be used, including cholestyramine to bind the hormone and clear it through the
gastrointestinal tract. Rarely, charcoal hemoperfusion, hemodialysis, or plasmapheresis may be needed to increase thyroid hormone clearance (Table 3) [27,31].
Adrenal insufficiency
Hypothalamic – pituitary-adrenal axis
The hypothalamic – pituitary-adrenal axis (HPA) regulates adrenal output of
glucocorticoids, which are intricately involved with the metabolism and production of nutritional substrates and maintenance and regulation of immune and
circulatory function. Hypothalamic release of corticotrophin releasing hormone
stimulates the pituitary to produce ACTH. ACTH output from the anterior
pituitary subsequently stimulates the adrenal cortex to produce cortisol, which
completes the cycle by providing negative feedback for CRH and ACTH release
[32]. Normal cortisol output of the adrenal gland in nonstressed conditions is
between 15 mg and 30 mg per day [32,33].
The intact HPA axis is responsive to stressors such as surgery, trauma, burns,
exercise, and psychologic trauma. Cortisol output amplifies proportionate to the
degree of stress, and may increase to 60 to 100 mg/m2 per day [32,33].
Glucocorticoids affect the circulation by facilitating the effects of catecholamines
such as norepinephrine and epinephrine on vascular tone and by their positive
inotropic effects [32]. Glucocorticoids have an inhibitory effect on endothelial
production of prostacyclin (PGI2) [34]. Relative glucocorticoid deficiency thus
allows enhanced PGI2 production, which results in a vasodilated state [34].
The ability of the adrenal axis to increase its output in the presence of stressful circumstances and enhance substrate availability and cardiac output to accommodate the increase in energy demand is essential for the survival of the organism.
Etiology of adrenal insufficiency
Endogenous causes of AI are uncommon disorders. Primary AI results from
the destruction or malfunction of more than 90% of the adrenal cortex.
Tuberculosis used to be the most common etiology of primary AI. With the
wane of tuberculosis, autoimmune adrenalitis became the most likely primary
cause in the United States. However, 30% of patients with advanced HIV
infections develop primary AI [33]. Aldosterone production is deficient in
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Table 3
Management of thyroid storma
Atenolol or metoprolol
Propylthiouracil (PTU)
Iodinated contrast agents
(administer only after PTU
or methimazole given)
Iopanoic acid or ipodate
Lugol’s solution or
Saturated solution of
potassium iodide
(SSKI) or
Sodium iodide
Hydrocortisone or
Dose and medication route
250 mg – 1,000 microg/kg iv bolus
followed by 50 – 300 mg/kg/min
1 mg iv as needed, can convert to
60 – 80 mg mouth (po) or
nasogastric(NG) tube every 4 hours
Adjust for equivalent doses
to propranolol
Antagonizes effect of
increased adrenergic tone
and inhibits conversion of
T4 to T3
800 – 1000 mg orally immediately,
then 200 mg every 4 hours by
mouth or NG
30 mg by mouth immediately, then
30 mg every 6 h by mouth or NG
Blocks new thyroid
hormone synthesis,
blocks T4 to T3
(PTU only)
0.5 – 1.0 g/day by mouth or NG
Blocks T4 to T3
conversion, block thyroid
hormone release
(via iodine release)
10 drops three times per d by mouth
or NG
5 drops every 6 h by mouth or NG
Blocks thyroid hormone
0.5 – 1.0 g every 12 h
50 mg every 6 h IV
2 mg every 6 h IV
Blocks conversion of T4
to T3 and provide stress
doses of glucocorticoids
Supportive therapy in an intensive care unit to maintain blood pressure, heart rate, and
temperature, while aggressive identification of a precipitating etiology is undertaken. Medications as
outlined above are used to control thyroid hormone release and peripheral activity.
Modified from Ginsberg J. Diagnosis and management of Graves disease. CMAJ 2003;168(5):584;
with permission of the publisher, 2003 Canadian Medical Association.
patients with primary AI also, and may be apparent because of the presence of
hyponatremia and hyperkalemia.
Secondary AI results from pituitary disease that has resulted in reduced ACTH
production. Tertiary AI is caused by hypothalamic disease, or is iatrogenic in
nature. Iatrogenic AI because of exogenous administration of corticosteroids is
the most common cause overall of AI in the general population. Because
aldosterone production is primarily regulated by the renin-angiotensin system,
hypoaldosteronemia is not expected in AI of secondary and tertiary causes, and
the electrolyte disturbances are generally not seen [20,33,34].
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Screening of the surgical patient for adrenal insufficiency
Preoperative screening should help identify individuals who are at risk for
HPA axis suppression. Categories of patients include patients receiving chronic
corticosteroids for conditions other than primary AI (eg, transplant patients),
patients who may be receiving corticosteroids intermittently or who had a course
of steroids within the past year, those being treated for diagnosed AI, and patients
with undiagnosed/suspected AI.
One may suspect the possibility that a patient may have received a course of
corticosteroids in the last year by noting certain diagnoses, such as asthma,
inflammatory bowel disease, collagen vascular disease, rheumatoid arthritis, and
in those who have had central nervous system tumors, or neurosurgery. Patients
may also develop adrenal suppression from the use of topical corticosteroids for
dermatologic disorders, especially when applied to large skin surface areas, and
with the use of occlusive dressings and glucocorticoids preparations of higher
potency. Adrenal suppression may also occasionally occur with the use of potent
inhaled glucocorticoids [34].
Chronic corticosteroid administration, with the exception of low doses and
alternate day regimens, will suppress the HPA axis. The duration of treatment with
corticosteroids that results in HPA axis suppression, and the daily threshold dose
necessary for this to occur has not been precisely defined. Generally, however,
patients who have received doses in excess of 20 mg per day for more than 5 days
may be considered to be at risk for adrenal suppression [34].
Patients who have been receiving doses of prednisone or prednisone equivalent in the physiologic range are not considered to have HPA axis suppression.
This corresponds to doses of prednisone 5 mg per day, hydrocortisone 25 mg per
day, and dexamethasone 0.75 mg per day. Patients who have been receiving
steroid doses in excess of the physiologic range, but less than 20 mg of
prednisone or equivalent per day, may be expected to have HPA axis suppression
after 4 weeks of treatment [34]. Alternate day regimens, especially when dosing
the corticosteroid in the morning rather than the evening, are less likely to cause
HPA axis suppression.
Recovery from HPA axis suppression after removal from endogenous (Cushing’s syndrome) or exogenous corticosteroids has been shown to occur over a
time course of approximately 9 months. Pituitary function normalizes first, with
ACTH secretion resuming its diurnal pattern. Return of adrenocortical function is
more gradual. The response to provocative testing generally normalizes 9 months
or more after withdrawal from glucocorticoid therapy [34].
Patients with chronic primary AI present with fatigue, weight loss, nausea,
vomiting, and diarrhea. Hyperpigmentation may be seen. Electrolyte abnormalities, eosinophilia, and hypoglycemia may be noted on laboratory examination.
Patients with primary AI generally receive routine replacement doses of hydrocortisone in the range of 20 to 30 mg daily in divided doses, with most of it
administered in the morning. Such patients may still develop symptomatic AI
when subjected to the stresses of surgery and the perioperative period because
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
they remain unable to compensate for the increased cortisol levels necessitated by
these circumstances.
Diagnostic provocative testing of the HPA axis can be performed if one
suspects undiagnosed primary AI, but these circumstances are rare. Cosyntropin,
a synthetic ACTH, stimulation testing assesses the ability of the adrenal gland to
accelerate production of cortisol in response to a surge in ACTH levels. Baseline
cortisol levels are drawn, and then a high dose, 0.25 mg of cosyntropin, is administered. Cortisol levels are repeated at 1/2 hour and again at 1 hour after
the dose. Baseline cortisol levels in normal, non-stressed individuals are in the
range of 6 to 18 mg/dL. Cortisol levels are expected to increase to more than 18 to
20 mg/dL after the stimulation test in non-stressed individuals. Critically ill,
stressed patients are expected to have higher than normal baseline cortisol levels.
Acute adrenal insufficiency
Acute AI is manifested by nausea, vomiting, and hypotension and may
progress to cardiovascular collapse. Hyponatremia and hyperkalemia may be
present. Shock caused by acute AI is typically distributive, associated with a low
systemic vascular resistance. The shock state is frequently unresponsive to catecholamines. Acute AI can mimic septic shock. Some degree of hypovolemia may
be present also. A high index of suspicion for acute AI needs to be maintained in
this scenario in patients who have not been previously diagnosed with AI. To
circumvent this outcome, supplemental corticosteroids should be administered to
patients identified to be at risk.
Once one has identified the patient at risk for HPA suppression, the dose of
corticosteroid supplementation with hydrocortisone or equivalent (Tables 4,5)
[33] should be determined on an individualized manner, based on an estimate of
the degree of stress anticipated to occur in association with an illness or planned
surgical procedure. Excessive supplementation is unnecessary and may predispose
the patient to untoward side effects such as hyperglycemia, poor wound healing,
catabolism, and corticosteroid psychosis. Under normal circumstances, cortisol
production increases for approximately 2 days after surgery and loses its normal
diurnal variation. Afterwards, cortisol levels generally decline to the normal range
and resume a diurnal pattern [33].
Two other specialized groups have an increased risk for AI with evidence of
inadequate cosyntropin responsiveness and improvement with corticosteroid
supplementation. These include a subgroup of septic shock patients and elderly
(>55 years old) patients who are vasopressor dependent post operatively after
general surgical procedures despite adequate volume resuscitation [35,36].
Annane and colleagues [35] performed a French multicenter, randomized, prospective, double-blind study in 299 septic shock patients septic shock patients
who were vasopressor and fluid dependent for less than 8 hours. A cosyntropin
stimulation test was performed to identify responders (>9 mg/dL increase after
cosyntropin) or non-responders (<9 mg/dL after stimulation). All patients were
then randomized to receive intermittent intravenous (IV) boluses of hydrocorti-
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Table 4
Guidelines for adrenal supplementation therapy
Medical or surgical stress
Corticosteroid dosage
Inguinal hernia repair
Mild febrile illness
Mild-moderate nausea/vomiting
Open cholecystectomy
Significant febrile illness
Severe gastroenteritis
Major cardiothoracic surgery
Whipple procedure
Liver resection
25 mg of hydrocortisone or 5 mg of methylprednisolone
intravenous (IV) on day of procedure only
50 – 75 mg of hydrocortisone or 10 – 15 mg of
methylprednisolone IV on day of procedure
Taper quickly over 1 – 2 days to usual dose
100 – 150 mg of hydrocortisone or 25 – 30 mg of
methylprednisolone IV on day of procedure
Rapid taper to usual dose over next 1 – 2 days
From Coursin D, Wood K. Corticosteroid supplementation in adrenal insufficiency. JAMA 2002;287:
236 – 40; with permission.
sone 50 mg every 6 hours and daily 50 mcg enterally of fludrocortisone or placebo. Patients who were non-responders and were treated with both glucocorticoid and mineralocorticoid had faster recovery and improved survival. It is
important to emphasize that responders treated with glucocorticoids had no
benefit and may have been compromised by steroid therapy [35].
Rivers et al [35] described a ‘‘transient’’ AI developing in surgical patients
older than 55 years of age who had sepsis and hypotension. In a prospective,
Table 5
Comparative steroid potency (mg basis)a
Steroid preparation
(equivalent to cortisol)
0.1 – 0.2
0.1 – 0.2
1/2 life (h)
18 – 36
18 – 36
36 – 54
18 – 36
Supplementation via the intravenous route is preferred for those who are NPO, have unpredictable
or poor absorption of medications, or have major stresses or critical illness. Prednisone and cortisone are
not recommended in patients who are unable to methylate these preparations into an active form.
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
observational study, Rivers and colleagues [36] evaluated 104 patients with a
mean age of 65.2 +/ 16.9 years who underwent surgery and were admitted to the
ICU with hypotension requiring vasopressors despite adequate fluid resuscitation.
After high dose synthetic cosyntropin stimulation testing, AI was defined as a
baseline serum cortisol less than 20 mg/dL with an increase in cortisol of less than
9 mg/dL after stimulation, or functional (relative) hypoadrenalism, was identified
by a serum cortisol less than 30 mg/dL with change in cortisol of less than 9 mg/dL.
The latter was present in almost a third of patients (32.7%), a far higher incidence
than predicted for the general surgical population. The patients with relative AI
benefited from corticosteroid therapy.
Pheochromocytomas are chromaffin cell tumors that secrete excessive amounts
of catecholamines. Patients develop secondary hypertension, which is often
paroxysmal, and the classic triad of headaches, sweating, and palpitations.
Anxiety and panic attacks may also be seen [36]. Most pheochromocytomas are
located in the adrenal medulla. Classical teaching is that 10% of patients will have
bilateral adrenal tumors, 10% will be extraadrenal, and less than 10% are
malignant [38,39]. Most pheochromocytomas are sporadic in nature. Pheochromocytomas may also be familial, inherited in an autosomal dominant manner, or
may be seen with VonHippel Lindau syndrome. Approximately 16% will be
associated with other endocrine disorders, such as multiple endocrine syndrome
type 2 (MEN II), which is comprised of medullary thyroid carcinoma, pheochromocytoma, and parathyroid hyperplasia. Pheochromocytoma has also been
associated with neurofibromatosis type I [37,40].
Diagnosis of pheochromocytoma
Lenders and colleagues [41] compared the sensitivity and specificity of plasma
free metanephrines, plasma catecholamines, urinary catecholamines, urinary total
and fractionated metanephrines, and urinary vanillylmandelic acid (VMA) for the
diagnosis of pheochromocytoma. They found that plasma free metanephrines and
urinary fractionated metanephrines had the greatest sensitivity for the diagnosis,
with 99% and 97% sensitivities, respectively. Urinary vanillylmandelic acid had
the lowest sensitivity at 65%. The greatest specificity was found with urinary
VMA (95%) and urinary total metanephrines (93%). The lowest specificity was
seen with urinary fractionated metanephrines at 69%. The authors concluded that
plasma free metanephrines should be the first test used when evaluating the
patient for a possible pheochromocytoma, and that a negative test would virtually
exclude this diagnosis.
False positive test results may be seen in the presence of caffeic acid (present
in coffee), tricyclic antidepressants, and phenoxybenzamine [42]. Patients with
positive biochemical screening tests for pheochromocytoma should then proceed
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
to an imaging study to attempt to localize the tumor [37]. Imaging tests include
MRI or CT scans, nuclear imaging studies such as metaiodobenzyl guanine
(MBIG), and 6-(18 F) fluorodopamine positron emission tomography. CT scans
are superior for detecting adrenal tumors, whereas MRI scans are superior in the
detection of extraadrenal tumors. Neither of these two techniques is very specific
for pheochromocytoma. Nuclear scans for pheochromocytoma are more specific
but are limited in availability. At times biochemical screening tests are positive,
but conventional scans are unable to localize the tumor.
Nuclear testing may be helpful in such situations [37] (see reference [37] for
algorithms to diagnose pheochromocytoma). The clonidine suppression test and
glucagon stimulation test may be useful in some circumstances, but should be
performed with caution because of their potential to result in severe hypotension
or hypertension, respectively.
Cardiac manifestations of pheochromocytoma
Excessively high levels of catecholamines have been demonstrated to have
toxic effects on the myocardium. Not surprisingly, patients with pheochromocytoma may be noted to have significant baseline ECG changes. Some may also
present with chest pain and ECG changes suspicious for ischemia. Despite striking
repolarization changes, many patients who proceed to coronary angiography
preoperatively are found to have normal coronary arteries. Patients presenting
with chest pain, ECG changes, and known pheochromocytoma may be better
served by proceeding to angiography rather than thrombolysis [43]. Some believe
the ECG changes are, in fact, a manifestation of a toxic myocarditis. Such changes
have often been noted to resolve rapidly in the postoperative period. In addition
to ECG changes that seem suspicious for MI, many patients with pheochromocytoma are noted to have a long QTc interval, which may predispose the patient to
ventricular arrhythmias. Liao and colleagues [43] noted a 16% incidence of
significant QTc prolongation in patients presenting for pheochromocytoma
resection, and an 80% incidence in those who were evaluated with coronary
angiography before surgery. The QTc intervals also normalized after surgery.
Pheochromocytoma multisystem crisis
Occasionally, patients may present with pheochromocytoma multisystem
crisis. This syndrome is manifested by wide swings in blood pressure, fever,
metabolic acidosis, (which may be lactic acidosis or hyperglycemic and hypersosmolar nonketotic in nature), renal insufficiency, myocarditis, respiratory
failure, and pulmonary edema. Mental status changes may also be present. Lactic
acidosis may be caused by tissue hypoperfusion, or possibly by an increase in
hepatic lactate production secondary to glycogenolysis. Tumor necrosis may also
play a role in this presentation, which is often associated with large, right-sided
adrenal masses [43].
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Surgical resection of pheochromocytoma
Although most of these tumors are not malignant, surgery is necessary to
eliminate the humoral disturbances and limit growth of the mass. Surgical resection is curative in up to 90% of cases [37,44]. Pheochromocytoma resections
have mortality rates reported between 0% and 6.5% in recent years (a considerable
improvement from the 20% mortality rate noted in 1951), and morbidity rates of
3% to 36%. This may be more reflective of risk associated with open, rather than
laparoscopic resections [44,45].
Historically, two open surgical approaches have been commonly performed for
pheochromocytoma resection. The midline abdominal approach was advantageous for thorough exploration of the abdominal cavity and contralateral adrenal
gland or bilateral resection. This approach also facilitates proximal vascular control and enables early ligation of the adrenal vein. The flank incision has the
advantage of decreasing the likelihood of adhesions or organ injury (eg, splenic
injury), but greatly limits exploration [40].
The first laparoscopic adrenalectomy for pheochromocytoma was performed in
1992 [40,46]. Bentrem and colleagues [40] reviewed the outcomes of patients who
underwent open, laparoscopic, or laparoscopic-assisted resections of adrenal or
extraadrenal pheochromocytomas by the same surgeon at Northwestern Memorial
Hospital between 1997 and 2001. Both laparoscopic and laparoscopic-assisted
procedures were performed in the lateral decubitus position. The surgeon planned
laparoscopic-assisted procedures when it was unclear if the procedure could be
completed by way of a laparoscopic technique alone. After initial laparoscopic
dissection, a subcostal incision was made while still in the lateral decubitus position. Seventeen patients, who had similar weight, age, and preoperative preparation
were studied. Patients who had laparoscopic resections had a smaller mean tumor
size (4.2 cm) versus the open (6.7 cm) and laparoscopic-assisted (6.3 cm) groups,
because of the selection criteria used. All extraadrenal tumors were removed with
an open technique. They found that laparoscopic resections were longer than open
procedures (average operative time of 218 versus 202 minutes) but were associated
with less blood loss (187 versus 562 mL), and the patients had a shorter hospital
stay. Exposure is more difficult with the laparoscopic technique, but the patient
benefits postoperatively with significantly less incisional pain. Laparoscopic
assisted procedures, on the other hand, were even longer than laparoscopic cases
(at 260 minutes) and were associated with more blood loss than the open technique
(average of 925 mL). These patients were also not discharged earlier than those in
the open group. The investigators concluded that laparoscopic adrenalectomies
were preferable to open resections, although they still advised open resections for
tumors greater than 6 cm in size and those extraadrenal in location. They also
concluded that one should proceed directly to open rather than laparoscopicassisted adrenalectomies if it was unclear preoperatively whether the tumor would
be able to be completely resected laparoscopically.
Open procedures also remain preferred for patients in whom malignancy is
suspected based on CT, MRI, or MBIG nuclear scan findings of periaortic
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
adenopathy. Patient selection for open versus laparoscopic procedures should
be individualized.
Preparing the patient with pheochromocytoma for surgery entails the institution
of alpha and beta-adrenergic blockade. Preoperative preparation with alpha
antagonists has been credited with much of the decrease in perioperative mortality
that has been demonstrated over the past few decades. Phenoxybenzamine, a longacting noncompetitive alpha-adrenergic antagonist, is initiated at least 1 week and
generally for a period of 2 to 4 weeks, before surgery. Beta-blockade is initiated in
patients more than 3 days before surgery in patients who have persistent
tachycardia or reflex tachycardia related to initiation of alpha blockade, or who
are having arrhythmias. It is important to initiate alpha adrenergic blockade before
beta adrenergic blockade to avoid a situation of ‘‘unopposed alpha’’ agonism
whereby the patient suffers from intense vasoconstriction from the alpha adrenergic excess and is at risk for extreme hypertension and increases in myocardial
workload. The pharmacologic adrenergic blockade helps to blunt the intense
surges in blood pressure that occur with surgery and tumor manipulation.
Preoperative preparation with alpha and beta-blockade also allows the heart to
recuperate before surgery from catecholamine-induced stress and cardiomyopathy.
Intravascular volume is also decreased in patients with pheochromocytoma.
This is manifested by hemoconcentration and orthostatic changes in blood pressure. Alpha adrenergic-mediated vasoconstriction and, possibly altered capillary
permeability is felt to be responsible for these findings [39].
Treatment with metyrosine (alpha-methyl-para-tyrosine) preoperatively results
in depletion of tumor catecholamine stores caused by competitive inhibition of
tyrosine hydroxylase, and gives rise to decreased blood pressure lability, and
decreased blood loss intraoperatively [37]. Alpha adrenergic blockade enables the
patient to have repleted intravascular volume. If this has been successful, one
expects the hemoconcentration to resolve or improve before surgery. The presence
or absence of orthostasis, and changes in the hematocrit should be assessed at the
time of the preoperative visit.
Certain medications should be avoided in patients with pheochromocytoma,
including tricyclic antidepressants, metoclopramide, droperidol, and naloxone [39].
Intraoperative management
Various anesthetic techniques have been used for pheochromocytoma resections, a full discussion of which is beyond the scope of this article. Short acting
antihypertensive infusions are preferred for blood pressure control, because of
anticipated hemodynamic lability before tumor resection and the frequency of
hypotension afterwards. Nitroprusside is commonly used. Others have used
fenoldopam for this purpose, which has the advantage of not having toxic
metabolites. Fenoldopam acts by stimulation of dopamine-1 receptors that cause
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
peripheral vasodilation while simultaneously increasing renal blood flow. Dose
ranges are 0.2 to 0.8 mg/kg/min. Fenoldopam also results in a diuresis and
natriuresis, however, which may not be desirable in these patients who are often
volume contracted [47]. Intraoperative use of magnesium sulfate with administration of a 40 to 60 mg/kg bolus before intubation, followed by a 2 g/h infusion
has also been advocated, because of inhibition of catecholamine release from the
adrenal medulla, and antiarrhythmic and vasodilatory properties [48].
Pheochromocytoma and surgical outcomes
Despite careful preoperative preparation for surgery, many patients will still
have a labile perioperative course. In a retrospective study by Plouin [44], patients with higher systolic blood pressures, recurrent tumor resections, and higher
urine metanephrine concentrations were more likely to have complications or die
in the perioperative period. Plasma catecholamine levels did not correlate with
risk. Kinney and colleagues [45] did a retrospective review of all patients presenting for an initial pheochromocytoma or paraganglionoma resection at the
Mayo Clinic. Of the 149 patients in the review period, there were no perioperative deaths. Almost 30% of patients had sustained episodes of hypertension or
hypotension intraoperatively, despite most having been premedicated with both
alpha and beta adrenergic blockers. A correlation was noted between large tumor
size, higher levels of urinary metanephrines and catecholamines, and a prolonged
surgical time and perioperative complications. However, the incidence of significant perioperative complications even in the presence of labile intraoperative
blood pressure was surprisingly small in this review. None of the patients suffered
MI or stroke. There were also no reports of ventricular dysrhythmias intraoperatively. Postoperative complications occurred in 6.3% of patients, mostly
made up of a 4.2% incidence of prolonged intubation, and a 1.4% incidence of
renal dysfunction.
Diabetes remains the most commonly encountered endocrinopathy with the
incidence of type 2 doubling in the past decade. The prevalence of diabetes is
projected to continue to increase dramatically over the next several decades
unless major public health initiatives are successful in stemming this growth.
Both type 1 and 2 diabetics more frequently require surgical and critical care than
their non-diabetic counterparts. Type 1 and 2 diabetics also sustain greater perioperative morbidity and mortality. Careful preoperative assessment and appropriate perioperative intervention may limit this.
There is increasing evidence that maintenance of normal blood glucose in the
perioperative period and during critical illness is beneficial for diabetic and nondiabetic patients. More data will hopefully be forthcoming to substantiate recent
reports and identify the mechanisms of improved outcome.
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
Thyroid disease remains a commonly encountered pathology that is more
readily identified and controlled in the modern era of radioimmune assays of
thyroid hormone and successful medical and surgical therapies. Severe hypothyroidism and thyroid storm are associated with significant increases in perioperative
morbidity and mortality. Recognition of these entities or those at risk for developing them post operatively is crucial in initiating timely and effective therapy.
Primary AI is uncommon, but results in glucocorticoid and mineralocorticoid
deficiency. Tertiary AI is far more common, most often secondary to iatrogenic
therapy with exogenous glucocorticoids for the management of chronic diseases
such as connective tissue disorders, anti-rejection regimes, and severe asthma.
Glucocorticoid replacement or supplementation is needed on a case-by-case basis
and should be individualized based on chronic steroid dose, duration, and stress
of the surgical procedure. Perioperative steroid dosing regimes now recommend
lower doses for shorter periods than previously suggested. More recently AI has
been recognized in two populations, elderly patients undergoing major surgery
and a subgroup of patients with septic shock. Timely diagnosis using synthetic
ACTH stimulation testing and stress glucocorticoid, and possibly mineralocorticoid therapy, seems to reverse these processes and improve recovery.
Although uncommon, patients with pheochromocytoma who undergo open or
laparoscopic resections remain diagnostic and therapeutic challenges. Perioperative outcome seems to have improved, in part, related to newer therapies and less
invasive surgeries when indicated.
The appropriate preoperative assessment and management of patients with
various endocrinopathies is important to optimize outcome and limit avoidable
complications. Hopefully additional evidence based guidelines will be forthcoming particularly in caring for the ever increasingly encountered perioperative diabetic.
[1] American Society of Anesthesiologists Task Force on Preanesthesia Evaluation. Practice advisory for preanesthesia evaluation: a report by the American Society of Anesthesiologists Task
Force on preanesthesia evaluation. Anesthesiology 2002;96(2):485 – 96.
[2] Glister B, Vigersky R. Perioperative management of type 1 diabetes mellitus. Endocrinol Metab
Clinics 2003;32(2):411 – 36.
[3] Weiland D, White R. Diabetes mellitus. Clini Fam Pract 2002;4(3):703 [Serial on-line. Available
from MD consult. Accessed August 8, 2003].
[4] Nathan D. Clinical review 146: the impact of clinical trials on the treatment of diabetes mellitus.
J Clin Endocrinol Metab 2002;87(5):1929 – 37.
[5] Eagle K, Berger P, Calkins H, Chaitman BR, Ewy GA, Fleischmann KE, et al. ACC/AHA
guideline update for perioperative cardiovascular evaluation for noncardiac surgery- executive
summary. A report of the American College of Cardiology/American Heart Association Task
Force on practice guidelines (committee to update the 1996 guidelines on perioperative cardiovascular evaluation for noncardiac surgery). Anesth Analg 2002;94:1052 – 64.
[6] Gottleib SS, McCarter RJ, Vogel RA. Effect of beta-blockade on mortality among high-risk and
low-risk patients after MI. N Engl J Med 1998;339(8):489 – 97.
[7] Boccara G, Outtara A, Godet G, Dufresne E, Bertrand M, Riou B, et al. Terlipressin versus
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
norepinephrine to correct refractory arterial hypotension after general anesthesia in patients chronically treated with renin-angiotensin system inhibitors. Anesthesiology 2003;98(6):1338 – 44.
Ryckwaert F, Boccara G, Frappier JM, Colson PH. Incidence, risk factors, and prognosis of a
moderate increase in plasma creatinine early after cardiac surgery. Crit Care Med 2002;30(7):
1495 – 8.
Metchick L, Petit W, Inzucchi S. Inpatient management of diabetes mellitus. Am J Med 2002;
113(4):317 – 23.
Bagg W, Ferri C, Desideri G, Gamble G, Ockelford P, Braatvedt GD. The influences of obesity
and glycemic control on endothelial activation in patients with type 2 diabetes. J Clin Endocrin
and Metab 2001;86(11):491 – 7.
Gu W, Pagel P, Warltier DC, Kersten JR. Modifying cardiovascular risk in diabetes mellitus.
Anesthesiology 2003;98(3):774 – 9.
Van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, et al. Intensive
insulin therapy in critically ill patients. N Engl J Med 2001;345(19):1359 – 67.
Malmberg K. Prospective randomised study of intensive insulin treatment on long term survival
after acute myocardial infarction in patients with diabetes mellitus. DIGAMI (diabetes mellitus,
insulin glucose infusion in myocardial infarction) study group. BMJ 1997;314:1512 – 5.
Doyle DJ, Arellano R. Upper airway diseases and airway management: a synopsis. Anesth Clin
North Am 2002;20(4):767 – 87.
Gardner GM. Posterior glottic stenosis and bilateral vocal fold immobility: diagnosis and treatment. Otolaryngologic Clin N Am 2000;33(4):855 – 78.
Kitamura A, Hoshino T, Kon T, Ogawa R. Patients with diabetic neuropathy are at risk of a
greater intraoperative reduction in core temperature. Anesthesiology 2000;92(5):1311 – 8.
Marks J. Perioperative management of diabetes. Am Fam Phys 2003;67(1):93 – 100.
Levetan CS. Hospital management of diabetes. Endocrinol Metab Clin N Am 2000;29(4):
745 – 70.
Luft FC. Lactic acidosis update for critical care clinicians. J Am Soc Nephrol 2001;12(Suppl 17):
S15 – 9.
Mercado DL. Perioperative medication management. Med Clin N Am 2003;87(1):41 – 57.
Hemmerling T, Schmid MC, Schmidt J, Kern S, Jacobi KE. Comparison of a continuous
glucose-insulin-potassium infusion versus intermittent bolus application of insulin on perioperative glucose control and hormone status in insulin-treated type 2 diabetics. J Clin Anesth 2001;
13(4):293 – 300.
Angelini G, Ketzler JT, Coursin DB. Perioperative care of the diabetic. ASA Refresher Courses
in Anesthesiology 2001;29:1 – 9.
Krinsley JS. Association between hyperglycemia and increased hospital mortality in a heterogeneous population of critically ill patients. Mayo Clin Proceed 2003;78:1471 – 8.
Furnary AP, Zerr KJ, Grunkemeier GL, Starr A. Continuous intravenous infusion reduces the
incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures.
Ann Thorac Surg 1999;67:352 – 62.
Weir CJ, Murray GD, Dyker AG, Lees KR. Is hyperglycemia an independent predictor of
poor outcome after acute stroke? Results of a long-term follow up study. BMJ 1997;314:
1303 – 6.
Sala J, Maisa Gonzalez de Molina FJ, Fernandez-Real JM, Gil M, Bosch D, Ricart W, et al.
Short-term mortality of MI patients with diabetes or hyperglycaemia during admission. J Epidemiol Comm Health 2002;56(9):707 – 12.
Ringel MD. Management of hypothyroidism and hyperthyroidism in the intensive care unit. Crit
Care Clin 2001;17(1):59 – 74.
Larsen RP, Davies TF. Hypothyroidism and thyroiditis. In: Larsen PR, Kronenberg HM, Shlomo
M, Polonsky KS, editors. Williams textbook of endocrinology. 10th edition. Philadelphia: WB
Saunders; 2003. p. 423 – 8.
Spell III NO. Stopping and starting medications in the perioperative period. Med Clin N Am
2001;85(5):1127 – 8.
L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123
[30] Schiff R, Welsh G. Perioperative evaluation and management of the patient with endocrine
dysfunction. Med Clin N Am 2003;87(1):175 – 92.
[31] Ginsberg J. Diagnosis and management of Grave’s disease. CMAJ 2003;168(5):575 – 85.
[32] Jabbour SA. Steroids and the surgical patient. Med Clin N Am 2001;85(5):1140 – 7.
[33] Coursin D, Wood K. Corticosteroid supplementation in adrenal insufficiency. JAMA 2002;
287(2):236 – 40.
[34] Axelrod L. Perioperative management of patients treated with glucocorticoids. Endocrinol Metab
Clin N Am 2003;32(2):367 – 83.
[35] Annane D. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality
in patients with septic shock. JAMA 2002;288(7):862 – 71.
[36] Rivers EP, Gaspari M, Saad GA. Adrenal insufficiency in high-risk surgical ICU patients. Chest
2001;119(3):889 – 96.
[37] Pacak K, Linehan WM, Eisenhofer G, Walther MM, Goldstein DS. Recent advances in genetics,
localization, and treatment of pheochromocytoma. Ann Intern Med 2001;134(4):315 – 29.
[38] Caputo C, Fishbane S, Shapiro L, Courgi RG, Kostadinov S, Donovan V, et al. Pheochromocytoma multisystem crisis in a patient with multiple endocrine neoplasia type II B and pyelonephritis. Am J Kidney Dis 2002;39(6):E23.
[39] Bogolioubov A, Keefe D, Groeger J. Circulatory shock. Crit Care Clin 2001;17(3):697 – 719.
[40] Bentrem D, Pappas SG, Ahuja Y, Murayama KM, Angelos P. Contemporary surgical management of pheochromocytoma. Am J Surg 2002;184(6):621 – 5.
[41] Lenders J, Pacak K, Walther MM, Linehan WM, Mannelli M, Friberg P, et al. Biochemical
diagnosis of pheochromocytoma: which test is best? JAMA 2002;287(11):1427 – 34.
[42] McHenry C. What’s new in general surgery: endocrine surgery. J Am Coll Surg 2002;195(3):
364 – 71.
[43] Liao WB, Liu CF, Chiang CW, Kung CT, Lee CW. Cardiovascular manifestations of pheochromocytoma. Am J Emerg Med 2000;18(5):622 – 5.
[44] Plouin PF. Factors associated with perioperative morbidity and mortality in patients with pheochromocytoma: analysis of 165 operations at a single center. J Clin Endocrin Metab 2001;86(4):
1480 – 6.
[45] Kinney M, Warner ME, van Heerden JA, Horlocker TT, Young Jr WF, Schroeder DR, et al.
Perianesthetic risks and outcomes of pheochromocytoma and paraganglionoma resection.
Anesth Analg 2000;91(5):1118 – 23.
[46] Gagner M, Lacroix A, Bolte E. Laparoscopic adrenalectomy in Cushing’s syndrome and pheochromocytoma. N Engl J Med 1992;327(14):1033.
[47] Cooper Z, Mihm F. Blood pressure control with fenoldopam during excision of a pheochromocytoma. Anesthesiology 1999;91(2):558 – 60.
[48] Pivalizza E. Magnesium sulfate and epidural anesthesia in pheochromocytoma and severe coronary artery disease. Anesth Analg 1995;81(2):414 – 6.