Pediatric Annals

Diabetic Ketoacidosis in Childhood

Arlan L Rosenbloom, MD; Desmond A Schatz, MD

Abstract

Diabetic ketoacidosis (DKA) is the most common cause of hospitalization of children with diabetes and of diabetes-related death. Most of the deaths can be attributed to intracerebral crises.1 Over the past 25 years, there has been a substantial reduction in the proportion of newly diagnosed diabetic patients hospitalized for DKA, falling from 75% to less than 25%, and recurrent DKA in established patients has been reduced in frequency by the intervention of multidisciplinary teams.2-3 Unfortunately, there is no evidence of a decrease in case fatality below the 1% to 2% achieved in the early 1970s, despite improvements in fluid and electrolyte and insulin therapy and more careful monitoring.1,3

The typical pediatrie practice will include only a few patients with diabetes, and the general pediatrician is unlikely to encounter more than a single episode of DKA every few years. This article, therefore, is presented as something of a checklist for imprinting and review at the time of the event, which should be reinforced by more comprehensive, readily available texts when time permits. No set of guidelines can operate without concomitant clinical judgment by the attending physician, who must not assume that either the patient or the treatment regimen will follow the book. Based on experience and personal review of more than 50 cases of DKA with disastrous outcomes, we advise our physician colleagues to follow four basic rules in dealing with DKA:

* Admit patients only to a unit in which neurological status and vital signs can be monitored frequently and blood glucose levels can be measured hourly.

* Personally evaluate the patient early on admission and frequently thereafter.

* Keep good records, including rationalization for therapeutic decisions, and a flow sheet of measures pertinent to those decisions.

* Develop a relationship with a pediatrie diabetes specialist you trust and call him or her with any questions, including whether or not the patient needs to be transferred to a specialized unit.

We do not know what the case fatality rates are in general hospitals versus teaching hospitals, but recog' nizing that even the latter are dealing with case fatalities of 1% to 2% should alert the pediatrician to treat DKA as he or she would any other potentially fatal condition.4

DEFINITION

The combination of hyperglycemia (greater than 200 mg/dL), hyperketonemia (high serum ketones, acetone or betahydroxybutyrate levels), or large keton' uria, with acidosis (venous pH <7.3 or serum bicar' bonate <15 mEq/L) defines DKA.

CAUSE

Diabetic ketoacidosis is always caused by insulin deficiency, either relative or absolute. Many previ' ously undiagnosed patients have been seen in physidans' offices or emergency rooms with inadequate evaluation to make the diagnosis before they became critically ill. A high index of suspicion is particularly important for infants and young children. An inter' esting phenomenon is the occasional marked delay in diagnosis seen in medical families and in the siblings or offspring of people with diabetes, reflecting denial. Remember that a simple urine dipstick test may turn out to be lifesaving by preventing the initial episode of ketoacidosis, particularly in the infant.1

In the established patient, DKA results from:

* failing to take insulin, the most common cause of recurrent DKA, particularly in adolescents5;

* acute stress from trauma, a febrile illness, or severe psychological stress, which increases counterregulatory hormones such as glucagon, epinephrine, cortisol, or growth hormone; or

* poor sick-day management, particularly not giving insulin because the child is not eating or failing to increase the insulin dose during the illness, as dictated by blood glucose monitoring.

PRESENTATION

The various signs and symptoms associated with diabetic ketoacidosis reflect the disturbed metabolic status, and understanding…

Diabetic ketoacidosis (DKA) is the most common cause of hospitalization of children with diabetes and of diabetes-related death. Most of the deaths can be attributed to intracerebral crises.1 Over the past 25 years, there has been a substantial reduction in the proportion of newly diagnosed diabetic patients hospitalized for DKA, falling from 75% to less than 25%, and recurrent DKA in established patients has been reduced in frequency by the intervention of multidisciplinary teams.2-3 Unfortunately, there is no evidence of a decrease in case fatality below the 1% to 2% achieved in the early 1970s, despite improvements in fluid and electrolyte and insulin therapy and more careful monitoring.1,3

The typical pediatrie practice will include only a few patients with diabetes, and the general pediatrician is unlikely to encounter more than a single episode of DKA every few years. This article, therefore, is presented as something of a checklist for imprinting and review at the time of the event, which should be reinforced by more comprehensive, readily available texts when time permits. No set of guidelines can operate without concomitant clinical judgment by the attending physician, who must not assume that either the patient or the treatment regimen will follow the book. Based on experience and personal review of more than 50 cases of DKA with disastrous outcomes, we advise our physician colleagues to follow four basic rules in dealing with DKA:

* Admit patients only to a unit in which neurological status and vital signs can be monitored frequently and blood glucose levels can be measured hourly.

* Personally evaluate the patient early on admission and frequently thereafter.

* Keep good records, including rationalization for therapeutic decisions, and a flow sheet of measures pertinent to those decisions.

* Develop a relationship with a pediatrie diabetes specialist you trust and call him or her with any questions, including whether or not the patient needs to be transferred to a specialized unit.

We do not know what the case fatality rates are in general hospitals versus teaching hospitals, but recog' nizing that even the latter are dealing with case fatalities of 1% to 2% should alert the pediatrician to treat DKA as he or she would any other potentially fatal condition.4

DEFINITION

The combination of hyperglycemia (greater than 200 mg/dL), hyperketonemia (high serum ketones, acetone or betahydroxybutyrate levels), or large keton' uria, with acidosis (venous pH <7.3 or serum bicar' bonate <15 mEq/L) defines DKA.

CAUSE

Diabetic ketoacidosis is always caused by insulin deficiency, either relative or absolute. Many previ' ously undiagnosed patients have been seen in physidans' offices or emergency rooms with inadequate evaluation to make the diagnosis before they became critically ill. A high index of suspicion is particularly important for infants and young children. An inter' esting phenomenon is the occasional marked delay in diagnosis seen in medical families and in the siblings or offspring of people with diabetes, reflecting denial. Remember that a simple urine dipstick test may turn out to be lifesaving by preventing the initial episode of ketoacidosis, particularly in the infant.1

In the established patient, DKA results from:

* failing to take insulin, the most common cause of recurrent DKA, particularly in adolescents5;

* acute stress from trauma, a febrile illness, or severe psychological stress, which increases counterregulatory hormones such as glucagon, epinephrine, cortisol, or growth hormone; or

* poor sick-day management, particularly not giving insulin because the child is not eating or failing to increase the insulin dose during the illness, as dictated by blood glucose monitoring.

PRESENTATION

The various signs and symptoms associated with diabetic ketoacidosis reflect the disturbed metabolic status, and understanding their basis is essential to rational treatment.

Hyperglycemia. Insulin deficiency results in decreased glucose uptake with tissue starvation resulting in proteolysis and lipolysis and thereby providing amino acids and glycerol for gluconeogenesis, enhanced by a counterregulatory hormone response to both the precipitating and tissue starvation stress. In the liver, insulin deficiency results in glycogenolysis and enhanced gluconeogenesis, also stimulated by counterregulatory hormones.

Dehydration and thirst. Osmotic diuresis due to hyperglycemia and hyperketonemia, hyperventilation, or vomiting as part of the primary precipitating illness or resulting from the ketosis; because dehydration is usually hyperosmolar and mostly intracellular, dehydration may be underestimated on. clinical examination.

Acidosis. Ketonemia from overproduction of ketones and lactic acidosis from tissue hypoperfusion.

Rapid deep respiration (Kussmaul breathing). Compensatory response to metabolic acidosis, contributing to dehydration.

Coma. Due to hyperosmolality, not acidosis; calculated osmolality greater than 320 mOsm/L is associated with coma.6

Hyperlipidemia. Due to counterregulatory hormone response plus hypoinsulinemia-mediated lipolysis.

Electrolyte disturbances. Spuriously low sodium due to osmolar dilution by glucose and sodium-free lipid fraction; nonetheless, sodium deficit is estimated at 10 mEq/kg body weight. Potassium may be misleadingly normal because of acidosis-related exudation from tissues and obligatory urinary loses; estimate total potassium deficit at 5 mEq/kg.

Other observations.

* Blood urea nitrogen is elevated as a result of dehydration.

* Creatinine level may be falsely reported as elevated because of interference by ketones in the autoanalyzer methodology for measuring creatinine levels.

* Serum ketone levels are high, with betahydroxybutyrate being the most abundant. Serum dilutions for ketones are of no value.

* Elevated white blood cell count with a shift to the left is a stress response and not helpful in diagnosing intercurrent infection.

* Abdominal pain and tenderness, with ileus, are usually nonspecific and are lessened with improvement of the metabolic state. If they are not, they need to be evaluated as with any other acute abdominal problem.

* Increased blood pressure and heart rate are caused by constricted circulatory volume and stress.

Table

TABLE 1Monitoring Treatment of Diabetic Ketoacldosls

TABLE 1

Monitoring Treatment of Diabetic Ketoacldosls

TREATMENT

The patient with a pH >7.25 who does not have persistent vomiting can be treated and observed in the emergency room over a few hours without hospital admission. The goals of treatment are to: 1) restore perfusion, which in and of itself will increase glucose use in the peripheral tissues and reverse the progressive acidosis; 2) decrease blood sugar level and stop ketogenesis by giving insulin, which will reverse proteolysis and lipolysis, and stimulate glucose uptake and processing; 3) replace electrolyte deficits; and 4) avoid intracerebral complications, hypoglycemia, and hypokalemia.

Fluid Therapy

* Generally assume 10% dehydration (100 mL/kg) and up to 15% in infants.

* Provide 20 mL/kg 0.9% sodium chloride in the first 1 to 2 hours as expansion fluids.

* Calculate maintenance in the usual fashion (eg, 1000 mL for the first 10 kg + 50 mL/kg for next 10 kg + 20 mL/kg over 20 kg).

* Calculate the remainder of replacement fluids needed after the loading dose has been given based on 10% dehydration and administer this amount along with maintenance fluids over the subsequent 22 to 23 hours.

* If the serum osmolality (calculated or measured) is >320 mOsm/L, correct it over 36 hours; if it is >340 mOsm/L, correct it over 48 hours.

* Provide potassium intravenously as half KCL and half KPO. (to replenish low phosphate levels and to decrease the risk of hyperchloremia) after the serum potassium has fallen below 6 mEq/L or urine output has been established.

* The administration of bicarbonate is rarely indicated since there is no evidence that it facilitates metabolic recovery. It should be given, however, when there is hypoventilation, but never by push as this can produce dangerous hypokalemia.

Insulin

* Insulin can be given immediately at the time of initial fluid expansion or it can be withheld until this is completed, giving a more realistic starting glucose level.

* Hourly administration of 0.1 U of insulin per kilogram of body weight can be given as a continuous infusion, using a calibrated pump.

* An initial bolus of 0.1 U/kg can be given when the infusion is started.

* It may be more convenient in some settings to administer initially 0.1 U/kg by intravenous (IV) route and 0.1 U/kg by intramuscular (IM) route with subsequent doses of 0. 1 U/kg by IM or subcutaneous (SC) route hourly. In adults, there does not seem to be any difference whether insulin is administered by the IV, IM, or SC route after the first 2 hours of treatment.7

* There is no evidence that low-dose insulin as currently used results in any less frequency of hypoglycemia, hypokalemia, or cerebral edema than treatment with much higher doses. The principal advantages of low-dose therapy given as continuous IV or hourly IM or SC injections are that there is frequent contact with the patient and that hourly blood glucose levels are determined.

* High blood glucose levels should drop 50 to 150 mg/dL/hour; if they do not, the insulin dose should be increased. This is rarely necessary.

* When the blood glucose level falls to 250 mg/dL, the IV fluids should be changed to contain 5% dextrose.

* If the blood glucose level fells below 150 mg/dL with the 5% dextrose solution running, the insulin dose should be reduced to .05 U/kg/hour or, preferably, the IV fluid dextrose concentration should be increased to 10%.

* Do not stop insulin or reduce it below .05 U/kg/hour because a continuous supply of insulin is needed to prevent ketosis and permit continued anabolism.

Monitoring

* A flow sheet is essential to record at hourly intervals the clinical signs, laboratory values, and fluid and insulin administered as noted in Table 1.

* The nursing staff must be given clear guidelines for calling the attending physician, such as those listed in Table 2.

* Electrocardiogram monitoring must be done and hourly serum potassium levels measured if the initial potassium level is <3 or >6 mEq/L.

* Mannitol should be kept at the bedside for the first 24 hours in quantities sufficient to give 1 to 2 g/kg body weight.

* Catheterization is not justified simply to monitor urine output. The occasional patient in DKA may have bladder atony and require initial catheterization, but rarely is an indwelling catheter needed.

COMPLICATIONS

The complications of diabetic ketoacidosis include the preventable problems of hypoglycemia, persistent acidosis, and hypokalemia, which should be seen rarely, if ever, following the guidelines of treatment above. Intracerebral complications and rare opportunistic infections can occur despite appropriate management of the acute episode.

Hypoglycemia

* Infrequent with hourly blood glucose monitoring.

* With the presence of an IV line, severe hypoglycemia is best treated with IV glucose. Glucagon can produce ketosis and nausea with vomiting, particularly in children.

Persistent Acidosis

* Defined as persistence of a bicarbonate value of less than 10 mmol/L after 8 to 10 hours of treatment.

* The usual cause is inadequate insulin often demonstrated by persistent hyperglycémie. Check insulin dilutions and rates of administration, and consider inadequate absorption if it is being given SC or IM. Also consider insulin resistance due to unusually high counterregulatory hormones with concomitant febrile illness or other severe stress.

* Switching from SC or IM administration may be needed. If the patient is receiving insulin IV, a fresh solution should be made up every 6 hours.

* Extremely rare causes are lactic acidosis caused by an episode of hypotension or apnea or inadequate renal competency in the handling of hydrogen ion following a shock-like episode.

Hypokalemia

* Extracellular potassium concentrations fell as a result of treatment with potassium reentering cells.

* If initial serum potassium is less than 3 mEq/L, potassium must be in initial expansion fluids without waiting for demonstration of renal function, and insulin should not be administered until after the initial bolus fluids have been given.

Table

TABLE 2Signs and Symptoms of Intracerebral Crisis During Treatment of Diabetic Ketoacidosis

TABLE 2

Signs and Symptoms of Intracerebral Crisis During Treatment of Diabetic Ketoacidosis

Intracerebral Complications

These comprise the most serious and frequent complications of DKA and can occur despite adherence to the above guidelines. There is no evidence for reduction in the frequency of intracerebral complications with the advent of low-dose insulin use by continuous infusion or the shift to more isotonic initial fluids. Although the classic picture is one of metabolic and clinical improvement with sudden deterioration, there may be several hours of decreasing sensorium and change in vital signs as noted in Table 2, and some children will be admitted in coma and never recover. An occasional patient may experience sudden neurologic collapse very early in treatment or even before treatment, possibly due to thrombosis. In the majority of patients who develop intracerebral complications, there is a sufficient warning period to permit administration of mannitol to reduce edema, and if indicated by respiratory distress, intubation/ hyperventilation. When this is accomplished before respiratory arrest, there is a greater than 50% chance of survival either in the normal state or with disability that does not prevent independence.1 Localized basilar edema is often seen on initial computed tomography scans, without generalized edema.

Because undiagnosed patients and children younger than 5 years of age are the most susceptible to intracerebral complications, the early recognition and prevention of the development of ketoacidosis is the most important means of preventing this complication.

Mucor Infection

Opportunistic infection with mucormycosis is a rare and frequently fetal complication of recurrent ketoacidosis, involving the upper respiratory tract and sinuses with erosion into the brain. We have seen two cases in the past 10 years, one fatal and one severely disabling.8

TRANSITION

Patients who have had correction of their metabolic state should resume normal eating and drinking and start an outpatient insulin regimen as rapidly as possible. Maintaining an intravenous line and chasing elevated blood glucose levels with regular insulin doses for a day or longer is a costly and pointless exercise.

* Intravenous fluids can be stopped 1 to 2 hours after substantial consumption and retention of oral fluids.

* Intermediate-acting insulin can be started when the patient no longer needs IV fluids.

* Wait until the presupper or prebreaîdast time to start or restart intermediate-acting insulin. Until then, give regular insulin 0.25 U/kg SC every 4 to 6 hours, stop the infusion 30 to 60 minutes after the first dose.

- Give established patients their usual morning or afternoon dose unless the dosage was previously inappropriate. The infusion can be stopped an hour later.

- Give the new patient 0.5 U/kg intermediateacting insulin and stop the infusion an hour later. If the blood glucose level is >220 mg/dL, this can be combined with 0.1 U/kg of regular insulin. Repeat the regular insulin dose every 4 to 6 hours as dictated by the blood glucose levels.

* Do not keep patients in the hospital to adjust insulin dosage because their diet, activity, and psychosocial environment are not like those at home.

REFERENCES

1. Rosenbioom AL. Inrracercbrai crises during treatment of diabetic ltetoac idosis. Diabetes Gire. 1 990; 1 3:22 -3 Ì.

2. Giordano B, Rosenblootn AL, Heller DR, Weber FT, Gomales R. Grgic A. Regional services foi chiltlren and youth with diabetes. fWiairics. 1977:60:492-498.

3. Travii LB, Kalia A. Diabetic ketoacidosis. In: Travii LB1 BmuhanJ BH, Schreiner BJ, «fa. Dwtetei Meittatí m CAiiÄirn antf Aiuticene, Riílattelphia, PA: WBSaiinJersCo; 1987:147-168.

4. Schau CW, Rosenhloom AL. Diabetic keioacidosis: management tactics in young patients Journal oj" Criticai IHmrss. 1988;3:3M5.

5. Melone Jl, Root AW. Plasma free insulin concentrations: keystone to effective management of diabetes meliitus in children, j Pediatt. 1981;99:862-867.

6. FulopM.TannenbaumH, DreyerN. Ketoiichyperosmolar coma. Lancet. 1973;2;635639.

7. Fisher J, Shahshahani M, Kitabchi A. Diabetic ketoacidosis: Im* dose insulin therapy by various routes. N Eng i Mea. 1 97 7 ¿97:238-243.

8. Geffken G, Johnson S, Silverstein ], Rosenbloom A. The death uf a child with diabetes from neglect. CIm Peàaa. 1992;31 :325-330.

TABLE 1

Monitoring Treatment of Diabetic Ketoacldosls

TABLE 2

Signs and Symptoms of Intracerebral Crisis During Treatment of Diabetic Ketoacidosis

10.3928/0090-4481-19940601-06

Sign up to receive

Journal E-contents