Acute Hyperglycemia and Hypoglycemia

Giuditta Angelini

Douglas B. Coursin

CASE SUMMARY

A 52-year-old man with a past medical history of diabetes mellitus presents with chest pain, shortness of breath, nausea, and a cold leg which has been symptomatic for 3 hours. Review of systems indicates that he has polyuria. On physical examination, he is diaphoretic and tachypneic. He is slightly confused and has fruity breath. His pulse is 120, and blood pressure is 90/50. His breathing is labored and deep, with a respiratory rate in the mid-30s. The cardiopulmonary examination is otherwise unremarkable. His skin is cool to the touch, but his right leg is mottled. No pulses are obtainable in that extremity. His electrocardiogram shows some minor degree of ST depression in leads II, III, and aVF that do not quite reach 1 mm. His chest radiograph is unremarkable. Laboratory abnormalities are: Glucose 655 mg per dL, sodium 132 mmol per L, potassium 5.5 mmol per L, magnesium 1.2 mmol per L, phosphate 3.0 mmol per L, creatinine 3.2 mg per dL, troponin 0.7 ng per mL, and a blood gas pH of 6.9 with a base excess of -21. In addition, urine is positive for ketones, and his serum betahydroxybutyrate is 2.9. The patient receives 8 units of regular insulin as a bolus, and will start on a continuous insulin infusion at a rate of 4 units per hour. He receives judicious fluid resuscitation with normal saline to maintain an adequate blood pressure and a urine output of at least 0.5 mL/kg/hour. Central venous and arterial access are obtained. After a moderate amount of fluid resuscitation, cautious introduction of intravenous metoprolol or esmolol will be attempted. A heparin infusion will be started. Electrolytes will be corrected. While these therapies are ongoing and the patient has some degree of stability, surgery to relieve the poor perfusion to his right leg will be undertaken. Eventually, the patient will reside in the intensive care unit (ICU). He will have ongoing surveillance for cardiac ischemia, intravascular volume status, insulin administration, and electrolyte correction.

What Is Diabetes Mellitus?

Diabetes mellitus develops from the failure of the body to produce an adequate amount of insulin to avoid an increase in serum glucose. Type 1 is an autoimmune disease of the β-pancreatic cells resulting in essentially little, if any, endogenous insulin production. However, 95% of diabetics have type 2 disease. In this disease, the patient has some inadequacy of insulin production, often accompanied by resistance to the action of insulin.¹

The prevalence of diabetes has increased over the last decade. The 40% increase is almost entirely attributed to type 2 diabetes. The trend has been associated with an increase in obesity.¹ In fact, there is some suggestion that obesity, hypertension, and hyperlipidemia may be clustered together with insulin resistance as part of a metabolic syndrome. It is unclear whether patients with this syndrome carry an increase in complications such as cardiovascular disease.²

What Is Diabetic Ketoacidosis?

Diabetic ketoacidosis (DKA) is the most common hyperglycemic emergency. The presence of a high serum blood glucose, low serum pH, and increase in ketone production are the cornerstone of pathophysiology. The annual incidence of DKA per the Centers for Disease Control is 3 to 8 per 1000 persons. The mortality from DKA is 2% to 5% in developed countries and 6% to 24% in underdeveloped countries. These statistics have worsened over the last 20 years.³

Although previous conventional wisdom reported that DKA only developed in type 1 diabetics, it does present in type 2 patients but with a much lower incidence. Only approximately 5% to 10% of diabetic have type 1. The
increased incidence of DKA is related to the increased prevalence of type 2 diabetes. DKA can occur in any setting where a patient has absolute or relative insulin deficiency.⁴ Type 1 diabetic patients have a near-total absence of endogenous insulin. During times of stress, such as an infection, patients may fail to administer the appropriate amount of exogenous insulin. Type 2 diabetic patients have peripheral insulin resistance, in addition to varying degrees of inadequate endogenous insulin production.⁵ In both situations, circulating carbohydrates are not utilized because of insufficient functioning insulin.

In addition, increased levels of counterregulatory hormones are also present and result in insulin resistance. Glucagon, catecholamines, cortisol, and growth hormone stimulate lipolysis, glycogenolysis, gluconeogenesis, and proteolysis. An increased stimulation of these ketogenic pathways develops. At some point, ketones cannot be utilized peripherally and begin to accumulate. As a result, there is hyperglycemia and an accumulation of ketoacids. β-hydroxybutyric acid is the ketoacid typically measured in plasma.⁶ Proinflammatory cytokines have also been shown to be elevated in DKA. Interleukin (IL)-10, IL-8, IL-6, IL-1β, and tumor necrosis factor α (TNF-α) are increased and can further exacerbate insulin resistance and aggravate hyperglycemia. After treatment with insulin and resolution of acidosis, most levels of inflammatory mediators decrease. In addition, markers of cardiovascular risk, such as homocysteine, are also elevated and respond to insulin, which suggest the potential for increased risk of cardiovascular problems secondary to DKA.⁷

In contradistinction to DKA, the hyperosmolar hyperglycemic state typically associated with type 2 diabetes usually results in more severe electrolyte disturbances and glucouric-induced volume depletion. Owing to the relatively minimal degree of acidosis, the patient experiences less severe symptoms and thus presents later in the course. Therefore, blood sugar may be significantly higher as is serum osmolarity, but the serum pH is relatively normal. The initial therapy for DKA and hyperosmolar states is similar. Aggressive fluid resuscitation and electrolyte replacement are paramount. Insulin therapy may be more important in DKA, but still has a role in hyperosmolar states as well.⁸

The physical examination in a patient with DKA frequently demonstrates mental status changes and cardiovascular effects that are consistent with dehydration. These can be extreme, depending on the degree of dehydration, but more so depending on the severity of the inciting cause of the episode. The deep, rapid breathing secondary to acidosis is called Kussmaul respirations. Diabetic keotacidosis can also be associated with delayed gastric emptying or intestinal ileus.⁸

Initial fluid resuscitation should include a 1-L bolus of normal saline over an hour or less, followed by 250 to 500 mL per hour continuous infusion. Blood glucose should be checked hourly during resuscitation. For a patient in DKA, insulin should also be started with a bolus of 0.15 units per kg, followed by an infusion. Table 42.1 is one example of an insulin infusion protocol. For patients in a hyperosmolar hyperglycemic state, hyperglycemia may drop with fluid resuscitation alone. Insulin should be considered in much smaller doses only after initial resuscitation has occurred in these patients. Once blood glucose falls below 250 mg per dL, fluid resuscitation should continue, with dextrose in normal saline at a slower rate of 150 to 250 mL per hour. The endpoint for DKA should be clearing ketones from the urine, which is slower than serum. The endpoint for hyperosmolar hyperglycemic state is when the plasma osmolarity is <310 mOsm per kg.⁸

Patients frequently have hyperkalemia when they present in DKA due to the extracellular potassium shifts secondary to acidosis. However, total body potassium is depleted on average of 3 to 5 mmol per kg body weight.³ Therefore, potassium supplementation should begin as soon as potassium level falls below 5.4 mEq per L.⁸ The combination of clearing the acidosis and administering insulin will drive potassium into the cells, making it safe to start replacing the deficit. It is important to assure that urine output is maintained, but potassium clearance is not affected unless the creatinine clearance is <15 mL per minute. This is due to extensive tubular secretion of potassium, which exceeds its glomerular filtration rate.⁹

The patient with DKA often also requires magnesium replacement as part of their treatment. Losses of this divalent cation are frequently underappreciated. While hypomagnesemia is certainly associated with deficiency, a normal serum magnesium does not rule it out.¹⁰ The only way to accurately assess total body magnesium is through tissue levels that are not readily available. Diuresis results in increased renal magnesium wasting, and most patients with DKA have had an osmotic diuresis from hyperglycemia. In addition, up to 40% of hospitalized patients are likely to be magnesium-deficient, and 25% to 39% of diabetics in the general population are also likely to be magnesium-deficient secondary to chronic magnesiumuria.¹⁰ Therefore, a patient with DKA will need magnesium replacement. Since magnesium is difficult to accurately measure, and a significant amount is lost in the urine despite ongoing deficiency, high doses (up to 8 g in 24 hours) are required.¹¹ Rapidly infused magnesium is renally excreted at a high rate, and therefore slow continuous infusion provides more complete replacement of deficits. Significant renal dysfunction, which would limit potassium replacement, should also limit magnesium replacement. Recent investigation into the effects of magnesium deficiency suggests that it may play a role in insulin resistance and producing diabetic complications such as cardiovascular disease.¹⁰

DKA is commonly associated with phosphate depletion that can be as high as 1.5 mmol per kg body weight.³ Patients can present initially with hyperphosphatemia, which is also related to cellular shifts that will reverse with the initiation of DKA therapy. Phosphate repletion does not need to be prophylactic, but should be initiated once the level falls to <1.5 mg per dL to prevent complications such as white cell, red cell, and platelet dysfunction,
as well as respiratory muscle weakness or rhabdomyolysis.⁸ Calcium levels should also be monitored during treatment and replaced as necessary.

TABLE 42.1 Example of Insulin Infusion Protocol

INITIATION
▪ Stop all previous hypoglycemic therapy
▪ Regular insulin 250 units/250 mL of D5 (normal saline if patient has ketoacidosis)
Blood Glucose	Action
≤200 mg/dL	Initiate infusion at 2 units/h
>200 mg/dL	Administer 4 unit bolus and initiate at 4 units/h
GLUCOSE MONITORING
▪ Measure glucose hourly
▪ If stable between 100 and 150 mg/dL for 3 h, change glucose checks to every 2 h for 2 checks, then every 4 h if remains in this range
TITRATION
Blood Glucose	Action
≤40 mg/dL	▪ Stop infusion for 30 min
	▪ Administer 50 mL of D50 by intravenous push
	▪ Recheck blood glucose in 15 min
	▪ Repeat until serum glucose is 70 mg/dL or higher
	▪ Restart infusion at 50% of previous rate
41-70 mg/dL	▪ Administer 25 mL of D50 by intravenous push
	▪ Recheck blood glucose in 15 min
	▪ Repeat until serum glucose is 70 mg/dL or higher
	▪ Reduce infusion to 50% of previous rate
71-99 mg/dL	Decrease rate by 50% of previous rate
100-150 mg/dL	No change
151-200 mg/dL	Increase rate by 1 unit/h
201-250 mg/dL	Increase rate by 2 units/h
>250 mg/dL	Increase rate by 3 units/h; bolus 3 units IV push

Acidosis should resolve with resuscitation and insulin therapy. Bicarbonate administration is rarely necessary and should never be considered routine. It can rapidly induce hypokalemia, and frequently results in metabolic alkalosis after DKA resolves. Sodium bicarbonate may be used if the pH is <7.0 to stimulate cardiac inotropy and reverse peripheral vasodilatation.⁸ Small doses such as 25 to 50 mEq should be used. Bicarbonate administration has not been shown to improve outcome with pH values between 6.9 and 7.1.¹²

▪ CEREBRAL EDEMA

One of the devastating complications of DKA is cerebral edema. Cerebral edema develops more commonly in children (approximately 1%) compared to adults. It is associated with 40% to 90% mortality. Although previously hypothesized, the rate of change in glucose or sodium and the amount of fluid resuscitation has not been shown to be associated with the incidence of cerebral edema. The one variable that has been associated with cerebral edema is the administration of bicarbonate.¹³

What Causes Diabetic Ketoacidosis?

The initial presentation of type 1 diabetes is the direct cause in approximately 25% of cases of DKA. In an additional 25% of patients, there is no proximal cause.³ Noncompliance with diabetic therapy, such as inadequate insulin, accounts for approximately 20%. Other major precipitating causes include surgery, trauma, burns, myocardial ischemia, stroke, pancreatitis, and thyroid storm, which altogether constitute approximately 6%.⁶

Virtually, any type of physical or psychological stress can produce a hormonal response that results in hyperglycemia and subsequent DKA. Medications can also be a cause. While corticosteroids are an obvious culprit, β-blockers, calcium channel blockers, antipsychotics, phenytoin, diuretics, and cimetidine can also induce a glycemic crisis.⁸ The most common precipitating cause worldwide is infection, constituting approximately 30% of cases. Pneumonia and urinary tract infections are the most common infectious causes of DKA.

Pneumonia constitutes a major proportion of the infections that predispose to DKA. It is unclear whether diabetes is an independent risk factor for respiratory disease. However, it is associated with an increase
in the number of respiratory infections caused by Staphylococcus aureus, gram-negative organisms, and Mycobacterium tuberculosis. In addition, increased morbidity and mortality are associated with Streptococcus pneumoniae and influenza pneumonia in patients with DKA.¹⁴

Urinary tract infections are the second, most common infections known to be associated with the development of DKA. It is not clear that diabetes predisposes to this type of infection, but there is an increase in upper tract infections compared to those in the lower tract. Pyelonephritis, especially bilateral, develops in approximately 80% of diabetics with urinary tract infections. Diabetic patients are also more likely to have fungi as a cause of their bladder or renal parenchymal infections.¹⁴

Several types of infections occur mostly in diabetic patients compared to the general population. The incidence of foot ulcers is approximately 2% per year. Fifteen percent of these patients develop osteomyelitis and, of these, 15% require amputation.¹⁵ Foot ulcers appear to be a small problem compared with expected complications; yet, the afflicted diabetic risks a statistically significant increase in mortality.¹⁵

Diabetes is a risk factor for Salmonella enteritidis. Nonpregnant adults with diabetes are more likely to have group B streptococcal infection. The incidence of tuberculosis is three to four times higher in diabetic patients. Finally, superficial candidal infections have an increased frequency in people with diabetes.¹⁴

Malignant otitis externa resulting from an invasive infection of Pseudomonas aeruginosa can extend intracranially and cause cranial osteomyelitis. Ketoacidosis is the most significant risk factor for rhinocerebral mucormycosis. This is a necrotic fungal infection of the nasal turbinates caused by Rhizopus oryzae.¹⁴ Both are relatively subtle infections that can rapidly progress to a life-threatening illness. Urgent identification, surgical debridement, and long-term antimicrobial therapy are needed to improve survival. Diabetics are also more prone to emphysematous infections of the gallbladder, renal parenchyma, and bladder. These are syndromes that involve gas-forming organisms and often require surgical removal of the affected organ if the patient continues to demonstrate symptoms despite adequate antimicrobial therapy.¹⁴

Why Is the Catabolic Response to Surgery Similar to Diabetic Ketoacidosis?

The endocrine effects of surgery have been well described. Typically, the hypothalamic stimulation of the sympathoadrenergic system is most commonly identified.¹⁴ The release of norepinephrine causes the subsequent rise in heart rate and blood pressure. However, there are more subtle effects of the pituitary that are likely more wideranging. Increases in hormone levels of all of the following can be expected:

▪ Adrenocorticotropic growth hormone
▪ Thyroid stimulating, β-endorphin
▪ Prolactin
▪ Gonadotrophins
▪ Arginine vasopressin
▪ Cortisol
▪ Aldosterone
▪ Glucagon

The levels of insulin and thyroxine are usually decreased.¹⁶

Glucose uptake and use by cells is inhibited; the liver is stimulated to increase glycogen breakdown; and proteolysis and lipolysis are increased—all of these processes culminate in hyperglycemia. Macrophages and neutrophils are inhibited from accumulating in areas of inflammation, and a tendency to retain fluid develops. Finally, oxygen consumption is increased.¹⁶ These effects are undesirable after surgery and are very similar to what occurs in DKA, including the cytokine release. Therefore, problems with glucose control may be more prevalent in diabetic patients undergoing surgery.¹⁷ These issues may develop in patients who were not previously known to be diabetic. Although it seems counterintuitive to administer insulin to a nondiabetic patient who will likely not need insulin after the stress of surgery abates, changing this hormone milieu may have a widespread effect. Patients may not only do better after surgery, but they may also have fewer complications. In fact, insulin infusions have been shown to improve outcome after surgery, whereas administering growth hormone has been shown to increase mortality in selected catabolic patients.¹⁸

Why Should Insulin Be Used in the Perioperative Period?

Over the last 20 years, increasing information, especially from the Diabetes Control and Complications Trial (DCCT), has suggested that maintaining blood glucose as close to normal as possible in ambulatory diabetic patients who require insulin is beneficial in decreasing long-term complications.¹⁹ More recently, maintaining normoglycemia in stressed hospitalized patients that may or may not have diabetes has been examined very closely. There is increasing evidence that normoglycemia may decrease morbidity and mortality in a wide range of disease states, ranging from myocardial infarction and stroke to trauma and ICU patients.

Stroke and hyperglycemia have likely been studied the longest, with the most evidence showing poor outcome, worse recovery, and an elevated risk of mortality. Some evidence suggests that a blood sugar above 150 mg per dL in the first 24 hours predicts an adverse effect in survival for up to a month after the event.²⁰ More recently, patients with an acute myocardial infarction have been shown to have a higher risk of death if the blood glucose is 110 to 150 mg per dL or higher and an increased risk of congestive heart failure or cardiogenic
shock if the blood glucose is 150 to 180 mg per dL or higher.²¹ Trauma patients are also a group that has shown some survival benefit from better glycemic control. These patients may be more susceptible to the deleterious effects of hyperglycemia than other patients in a surgical ICU.²²,²³

In 2001, Professor Van den Berghe et al. released a study performed in surgical ICU patients where blood glucose was maintained at or below 110 mg per dL.²⁴ Before this, maintaining the serum glucose below 200 mg per dL was generally considered satisfactory. Essentially, the goal was to avoid problems such as osmotic diuresis, electrolyte abnormalities, and acid-base imbalances associated with progressively uncontrolled hyperglycemia, with the added benefit that neutrophil and macrophage function was likely improved at a lower level as well.²⁵ Aggressive insulin management represents a paradigm shift.

In this prospective, randomized study that was dominated by postoperative cardiac surgical patients (approximately two thirds of the patients), all of whom initially required mechanical ventilation, one group of patients were maintained between 180 mg per dL and 200 mg per dL (conventional) compared to the other group that was maintained no higher than 110 mg per dL. Mortality in the ICU was decreased by 32%. Overall in-hospital mortality was also decreased by 34%. Bloodstream infections, renal failure requiring dialysis or hemofiltration, red cell transfusion, and critical illness polyneuropathy were all decreased by 40% to 50%. There was also a tendency to require the ventilator for fewer days. All the patients were in the surgical ICU, and therefore were likely undergoing some of the above alterations in their hormonal milieu. Because most of the patients were not diabetic, they were likely experiencing postoperative insulin resistance.²⁴ This is somewhat similar to what was found in the DCCT trial in that any degree of normoglycemia maintained at any time point in patients who were either newly diagnosed or long-term, poorly controlled diabetics produced some benefit.¹⁹ This is the first time that prospective evidence showed benefit in a nondiabetic patient.

Only gold members can continue reading. Log In or Register to continue