Symposium on Pediatric Neurology
Hypernatremia - Problems in Management Gwendolyn R. Hogan, MD.*
Hypernatremia, the increased concentration of solutes (sodium 150 mEq per liter) in the body fluids, can be classified into the following groups: (1) hypernatremia secondary to the loss of water in excess of solute as a result of diarrhea; (2) hypernatremia secondary to the excessive intake of solute as a result of accidental salt poisoning, concentrated formula, tube feedings without adequate water, and high electrolyte and solute solutions (boiled skim milk and Lytren) used in the management of diarrhea in small infants; and (3) hypernatremia secondary to central nervous system problems. In the third group there are two types: (a) isovolemic ("essential" hypernatremia) in which the patient has a sustained hypernatremia but is not volume depleted;3 characteristic findings include elevated serum sodium that is not corrected by fluid administration, intact antidiuretic hormone synthesis and renal tubular response to antidiuretic hormone, impaired osmotic regulation of secretion of antidiuretic hormone, correction of the hypernatremia with exogenous vasopressin, and little evidence of thirst; and (b) hypernatremia with a deficit of total body water as in patients with diabetes insipidus or as in patients with an absence of thirst. 5 Hypernatremia occurs most commonly as a result of excessive water loss due to diarrhea; however, with the improved care of children with diarrhea, specifically the discontinuance of the use of boiled skim milk, the incidence has decreased over the past few years. In the first six months of 1975, 39 children under the age of two were admitted to the Children's Hospital in Mississippi with a diagnosis of diarrhea and dehydration. Only four of these children had an initial sodium greater than 150 mEq per liter. It has been demonstrated in the experimental model and in the human that a series of events occurs with regularity when the serum sodium is elevated. Fluid from the intracellular space moves into the vascular compartment in an effort to establish normal osmolality. The vascular compartment is relatively well preserved with weight loss of "'Professor of Pediatrics; Associate Professor of Medicine (Neurology); Chief, Pediatric Neurology Service, University of Mississippi School of Medicine, Jackson, Mississippi
Pediatric Clinics of North America- VoL 23, No.3, August 1976
569
570
GWENDOLYN
R.
HOGAN
up to 20 per cent of body weight since only 6 to 7 per cent of the vascular volume is lost under these circumstances. 6 Thus, vascular collapse early in the course of the illness is rare. As the osmolality increases, the patient develops metabolic acidosis which may be the result of some change in cellular permeability or a dilutional factor. The acidosis is usually secondary to a shift of fluid to the extracellular space associated with a moderate loss of bicarbonate in the stool. Additional factors include the production of keto acids as a result of starvation, an actual increase in hydrogen ions, and the decreased excretion of hydrogen ions. It is pertinent that with acidosis the shift of potassium out of the cell is marked. Under these circumstances the intracellular potassium deficit is not reflected by the serum potassium. Hyperglycemia is a frequent component of hypernatremic dehydration. The degree of elevation of the blood glucose may reflect the severity of the stress and the metabolic imbalance. Some think the decrease in intracellular potassium may be a factor in decreased uptake of glucose by the cells. 4 It has been demonstrated experimentally that animals with chronic hypernatremia who have seizures during rehydration, have a blood glucose level in the dehydrated state that averages 100 mg per cent greater than those animals who do not convulse. s Therefore, if the blood glucose is significantly elevated, rehydrating fluids containing fructose instead of glucose would be ideal in that fructose to be utilized by the cells does not require the activity of cell membrane ATPase. Restoration of fluid balance will result in a return to normal of the blood glucose and pH as the excess hydrogen ions are excreted and the intracellular water is reconstituted. A mistaken diagnosis of diabetes mellitus can be fatal for the patient who receives insulin and large volumes of fluid because of resultant hypoglycemia secondary to the insulin and brain swelling secondary to water intoxication from volume overload. The clinical manifestations of hypernatremia will vary depending on the severity of dehydration and on the rate of onset of the hyperosmolar state. Since the vascular compartment is usually preserved, the skin feels "doughy" and the degree of dehydration is often underestimated clinically. With severe dehydration the child may be in shock. Respiratory rate is increased and the temperature is usually elevated. Central nervous system manifestations dominate the clinical picture. The child is irritable, restless, lethargic, or stuporous, muscle tone is increased, and the cry is high pitched. Seizures may occur if the child is severely dehydrated. The need to document the deficit, to calculate the replacement and maintenance requirements of fluids and electrolytes, and to correct the child's problem in an orderly and deliberate fashion is paramount. The evaluation of the child clinically and with laboratory data is urgent. Essentiallaboratory data include sodium, potassium, chloride, carbon dioxide combining power, glucose, blood urea nitrogen, and hematocrit. In most hospitals these parameters can be determined within a few minutes. Determination of osmolality, serum calcium, and magnesium, and a complete blood count will not alter the initial treatment and will be
571
HYPERNATREMIA-PROBLEMS IN MANAGEMENT
available in a few hours. The serum osmolality may be as much as 10 mOsm per liter higher than calculated from the serum sodium, blood urea nitrogen, and glucose. This may well represent the accumulation of amino acids, keto acids, and possibly sorbitol and fructose. 12 • 13 Urine specific gravity and pH are helpful. The serum sodium, potassium, chloride, carbon dioxide combining power, glucose, blood urea nitrogen, and hematocrit determinations should be repeated at 12 hour intervals during the early phase of rehydration and then daily until the child has recovered. Additional laboratory studies may be indicated such as blood and stool cultures, skull films, and lumbar puncture. The most difficult problems in rehydration usually occur in the infant who has hypernatremia secondary to excessive loss of water due to diarrhea. The infant is frequently febrile, moderately to severely dehydrated (10 to 15 per cent), acidotic, hyperglycemic, and mayor may not be in shock. The infant is often irritable and obtunded and seizures may have occurred. Restoring tonicity and avoiding the complications of water intoxication (seizures and brain swelling) may be difficult. The urgency of establishing an adequate blood volume in a patient who may be in shock is obvious and must be accomplished rapidly. This may necessitate the use of plasma, albumin, whole blood, normal saline and/or sodium bicarbonate. As a result a larger fluid volume than one would like must be given in a short period of time. In relation to the patient's osmolality, this fluid may be relatively hypotonic and can result in signs of water intoxication. It has been our clinical observation that 20 per cent of the calculated fluid deficit volume can usually be given within the first two hours if the patient is in shock or severely hypovolemic. The rehydration volume can be calculated by using the following formula: (Plasma osmolality per liter) x .6 (body wt in kg) 6 (b d 290 mOsm per liter -. 0 y
W
t·
III
kg) - Fl ·d d fi ·t· Ul e CI III
n1 ers.
Fluid deficit + 1500 ml per M2 for each 24 hours of rehydration equals rehydration fluid in liters. (This volume is calculated to run at a constant hourly rate over the predetermined number of days.) Serum osmolality may be calculated as follows: . . glucose BUN . 2 (serum sodIUm mEq per hter) + --}-8- + -::\- = osmolahty.
The total body sodium deficit in these patients is usually minimal, therefore the rehydrating fluid should not be markedly hypotonic in relation to the plasma, thus preventing the rapid production of a relative water intoxication. A solution containing 35 mEq per liter of sodium chloride, 35 mEq per liter of sodium bicarbonate, or, at times sodium lactate, and 15 to 20 mEq per liter of potassium in glucose or fructose would be ideal as an initial fluid. The severity of the acidosis will of course determine the necessity to correct the pH with additional sodium bicarbonate and/or lactate. In general, less than one third of the anion of the rehydration fluid should be base. A daily requirement of 2 mEq per kg of potassium should be added to the first day's fluid as soon as urine
572
GWENDOLYN
R.
HOGAN
output has been established or earlier if the potassium drops. The potassium concentration should not exceed 40 mEq per liter in the final solution. Hypocalcemia and hypomagnesemia are rarely problems in the management of acute diarrhea and dehydration. 2 If the child has a chronic problem or if intravenous therapy is required for more than 24 hours, this could become a symptomatic problem. Calcium is preferably administered slowly intravenously by the physician in a dose of up to 1 ml of 10 per cent calcium gluconate per kilogram of body weight. In our experience, if the magnesium is low, the serum calcium cannot be maintained at normal levels until the magnesium deficiency is also corrected. It is clear from available clinical and experimental data that as important, if not more important, than the nature of the solution is the rate of infusion.!' 2. 6-8 If fluids are well regulated and contain free water they can be given with no additional problems. In longstanding hypernatremia and in those patients with severe hypernatremia (sodium over 165 mEq per liter), rehydration is best accomplished over a 2 to 3 day period of time. If the sodium is not greater than 150 mEq per liter, the deficit can be corrected in a 24 hour period. If fluids can be taken orally after the first 24 hours of intravenous therapy, this is obviously the best physiological method of treatment. It is clear that the single most important factor in intravenous rehydration is the rate of infusion, not the specific solution. The management of hypernatremia secondary to excessive sodium intake requires free water in order to allow excretion of the excessive sodium. This is best achieved by a very slow rate of rehydration with a solution that is capable of furnishing free water but is not remarkably hypotonic in relation to the patient's serum. In severe cases one may begin with a solution of glucose in normal saline or if less severe (sodium less than 165 mEq per liter), a more hypotonic solution is used. Children with "essential" hypernatremia may have chronic mild elevations of the serum sodium (150 to 155 mEq per liter) without clinical manifestations of their hypertonic state. In this situation rehydration is best carried out by oral administration of fluids. Vasopressin is usually necessary to correct this condition; therefore, the major problem in management is the prevention of water intoxication. However, if the child has superimposed acute hypernatremia, then intravenous therapy (as outlined previously) is required. The major complications of hypernatremia are encountered in those patients with severe dehydration associated with diarrhea. Dehydration and hypovolemia produce shrinkage of the brain which can lead to tearing of the bridging vessels and resultant subarachnoid and subdural hemorrhage. Sludging results in small vessel occlusion, sagittal sinus thrombosis, and renal vein thrombosis. Seizures may occur in the severely dehydrated child as a result of metabolic derangement or infarction due to venous thrombosis. The child with sickle cell anemia presents a real challenge because of the necessity of rapid rehydration. These children may also require exchange transfusion prior to the usual course of rehydration.
HYPERNATREMIA-PROBLEMS IN MANAGEMENT
573
The single most important complication of rehydration therapy is the production of brain swelling secondary to water intoxication. As rehydration is instituted, water enters the cells readily, whereas solute equilibrium will require hours to days,6 and when carried out too rapidly, relative water intoxication is produced. All too often a child is evaluated properly but is rehydrated too vigorously initially and water intoxication results, manifested clinically by stupor, increased intracranial pressure, and seizures. The child may improve initially but then becomes sombulent, "jumpy," and develops generalized seizures. This may occur several hours after treatment has been initiated and frequently at a time when serum sodium is still elevated. It will, however, in almost all cases follow the administration of a rather large volume of fluid that is hypotonic when compared with the patient's serum. For example, a quantity of normal saline given to a patient with a serum sodium of 170 mEq per liter may well produce a relative water intoxication. When seizures occur under these circumstances it is best to treat them with 3 per cent saline calculated to raise the serum sodium 1 to 2 mEq per liter. This will frequently stop the seizures prior to the administration of diphenylhydantoin (Dilantin). Neurological complications dominate the clinical picture and the risk of permanent deficit is greater in the infant IO and in patients with a serum sodium of 160 mEq per liter or greater. 9 Permanent neurological deficit occurs in at least 11 to 15 per cent of these patients. The obvious neurological complications associated with hypernatremia result from shrinkage of the brain with occasional small hemorrhages or more commonly from venous thrombosis of small vessels in the brain resulting in focal signs and/or seizures and the occlusion of large veins or sinuses. These problems are further complicated in those children with sickle cell disease. It is clearly obvious and should be stated emphatically that the sick infant with dehydration and hypertonicity must be treated initially with intravenous fluids. There is no justification for the use of clysis in the management of the acutely ill hypernatremic infant. The complications of treatment, specifically, increased intracranial pressure, brain edema as a result of water intoxication, ischemia, and destruction of brain can be avoided by careful rehydration. Seizures may occur in the dehydrated state as a result of metabolic imbalance or brain injury secondary to venous thrombosis, water intoxication, or ischemia secondary to increased intracranial pressure. The treatment of the seizures will depend upon the underlying etiological factors.
SUMMARY Hypernatremia may be produced under several different circumstances but most frequently is the result of excessive water loss with diarrhea and the excessive solute load secondary to inappropriate preparation of formula. The clinical manifestations vary and depend primarily on the degree of dehydration and the rate at which the hyperosmolar
574
GWENDOLY N
R.
HOGAN
state has been reached. The management of the patient will, of course, depend upon the mechanism of hypernatremia and degree of dehydration and/or hypovolemia that is present. It seems clear that the exact nature of the rehydrating solution is not of major importance. The volume is of great concern but most vital seems to be the rate of rehydration. If rehydration is accomplished too rapidly the child becomes edematous, develops increased intracranial pressure, stupor, and convulsions. If fluids are given slowly and at a well regulated rate, these complications can usually be avoided. The patient should be monitored regularly with electrolytes, careful determination of weights, and records of intake and output. The rate of rehydration should be monitored to assure that the planned schedule of 24 to 72 hours (depending on the severity of the problem) is followed. Approximately 10 to 15 per cent of children with serum sodium of 160 mEq per liter or greater will have permanent neurological deficits.
REFERENCES 1. Banister, A.: Management of hypernatraemia in infancy. Arch. Dis. Child. 48 :825, 1973. 2. Bruck, E., Abal, G., and Aceto, T., Jr.: Pathogenesis and pathophysiology of hypertonic dehydration with diarrhea. Am. J. Dis. Child., 115:281-301, 1968. 3. DuRubertis, F. R., Michelis, M. F., and Davis, B. B.: "Essential" hypernatremia. Arch. Intern Med., 134:889-895,1974. 4. Finberg, L.: The management of the critically ill child with dehydration secondary to diarrhea. Pediatrics, 45: 1029-1036, 1970. 5. Goldberg, M., Weinstein, G., Adesman, J., et al.: Asymptomatic hypovolemic hypernatremia. Am. J. Med., 43:804-810,1967. 6. Hogan. G. R., Dodge, P. R., Gill, S. R., et al.: Pathogenesis of seizures occurring during restoration of plasma tonicity to normal in animals previously chronically hypernatremic. Pediatrics, 43:54-64,1969. 7. Hogan, G. R., Gill, S. R., Master, S., et al.: On the pathogenesis of seizures occurring during rehydration in chronic hypertonic dehyration. Trans. Am. Neurol. Assoc., 90:257260,1965. 8. Hogan, G. R., et al.: Unpublished data. 9. Macaulay, D., and Watson, M.: Hypernatremia in infants as a cause of brain damage. Arch. Dis. Child., 42:485-491,1967. 10. Morris-Jones, P. H., Houston, I. B., and Evans, R. C.: Prognosis of the neurological complications of acute hypernatremia. Lancet, 2:1385-1389,1967. 11. Nitzan, M., and Zelmanovsky, S.: Glucose intolerance in hypernatremic rats. Diabetes, 17:579-581,1968. 12. Prockop, L. D.: Hyperglycemia, polyol accumulation and increased intracranial pressure. Arch. Neurol., 25:126-140,1971. 13. Stevenson, R. E., and Bowyer, F. P.: Hyperglycemia with hyperosmolal dehydration in nondiabetic infants. J. Pediat., 77:818-823,1970.
Department of Pediatrics University of Mississippi School of Medicine 2500 North State Street Jackson, Mississippi 39216
Hypernatremia - Problems in Management Gwendolyn R. Hogan, MD.*
Hypernatremia, the increased concentration of solutes (sodium 150 mEq per liter) in the body fluids, can be classified into the following groups: (1) hypernatremia secondary to the loss of water in excess of solute as a result of diarrhea; (2) hypernatremia secondary to the excessive intake of solute as a result of accidental salt poisoning, concentrated formula, tube feedings without adequate water, and high electrolyte and solute solutions (boiled skim milk and Lytren) used in the management of diarrhea in small infants; and (3) hypernatremia secondary to central nervous system problems. In the third group there are two types: (a) isovolemic ("essential" hypernatremia) in which the patient has a sustained hypernatremia but is not volume depleted;3 characteristic findings include elevated serum sodium that is not corrected by fluid administration, intact antidiuretic hormone synthesis and renal tubular response to antidiuretic hormone, impaired osmotic regulation of secretion of antidiuretic hormone, correction of the hypernatremia with exogenous vasopressin, and little evidence of thirst; and (b) hypernatremia with a deficit of total body water as in patients with diabetes insipidus or as in patients with an absence of thirst. 5 Hypernatremia occurs most commonly as a result of excessive water loss due to diarrhea; however, with the improved care of children with diarrhea, specifically the discontinuance of the use of boiled skim milk, the incidence has decreased over the past few years. In the first six months of 1975, 39 children under the age of two were admitted to the Children's Hospital in Mississippi with a diagnosis of diarrhea and dehydration. Only four of these children had an initial sodium greater than 150 mEq per liter. It has been demonstrated in the experimental model and in the human that a series of events occurs with regularity when the serum sodium is elevated. Fluid from the intracellular space moves into the vascular compartment in an effort to establish normal osmolality. The vascular compartment is relatively well preserved with weight loss of "'Professor of Pediatrics; Associate Professor of Medicine (Neurology); Chief, Pediatric Neurology Service, University of Mississippi School of Medicine, Jackson, Mississippi
Pediatric Clinics of North America- VoL 23, No.3, August 1976
569
570
GWENDOLYN
R.
HOGAN
up to 20 per cent of body weight since only 6 to 7 per cent of the vascular volume is lost under these circumstances. 6 Thus, vascular collapse early in the course of the illness is rare. As the osmolality increases, the patient develops metabolic acidosis which may be the result of some change in cellular permeability or a dilutional factor. The acidosis is usually secondary to a shift of fluid to the extracellular space associated with a moderate loss of bicarbonate in the stool. Additional factors include the production of keto acids as a result of starvation, an actual increase in hydrogen ions, and the decreased excretion of hydrogen ions. It is pertinent that with acidosis the shift of potassium out of the cell is marked. Under these circumstances the intracellular potassium deficit is not reflected by the serum potassium. Hyperglycemia is a frequent component of hypernatremic dehydration. The degree of elevation of the blood glucose may reflect the severity of the stress and the metabolic imbalance. Some think the decrease in intracellular potassium may be a factor in decreased uptake of glucose by the cells. 4 It has been demonstrated experimentally that animals with chronic hypernatremia who have seizures during rehydration, have a blood glucose level in the dehydrated state that averages 100 mg per cent greater than those animals who do not convulse. s Therefore, if the blood glucose is significantly elevated, rehydrating fluids containing fructose instead of glucose would be ideal in that fructose to be utilized by the cells does not require the activity of cell membrane ATPase. Restoration of fluid balance will result in a return to normal of the blood glucose and pH as the excess hydrogen ions are excreted and the intracellular water is reconstituted. A mistaken diagnosis of diabetes mellitus can be fatal for the patient who receives insulin and large volumes of fluid because of resultant hypoglycemia secondary to the insulin and brain swelling secondary to water intoxication from volume overload. The clinical manifestations of hypernatremia will vary depending on the severity of dehydration and on the rate of onset of the hyperosmolar state. Since the vascular compartment is usually preserved, the skin feels "doughy" and the degree of dehydration is often underestimated clinically. With severe dehydration the child may be in shock. Respiratory rate is increased and the temperature is usually elevated. Central nervous system manifestations dominate the clinical picture. The child is irritable, restless, lethargic, or stuporous, muscle tone is increased, and the cry is high pitched. Seizures may occur if the child is severely dehydrated. The need to document the deficit, to calculate the replacement and maintenance requirements of fluids and electrolytes, and to correct the child's problem in an orderly and deliberate fashion is paramount. The evaluation of the child clinically and with laboratory data is urgent. Essentiallaboratory data include sodium, potassium, chloride, carbon dioxide combining power, glucose, blood urea nitrogen, and hematocrit. In most hospitals these parameters can be determined within a few minutes. Determination of osmolality, serum calcium, and magnesium, and a complete blood count will not alter the initial treatment and will be
571
HYPERNATREMIA-PROBLEMS IN MANAGEMENT
available in a few hours. The serum osmolality may be as much as 10 mOsm per liter higher than calculated from the serum sodium, blood urea nitrogen, and glucose. This may well represent the accumulation of amino acids, keto acids, and possibly sorbitol and fructose. 12 • 13 Urine specific gravity and pH are helpful. The serum sodium, potassium, chloride, carbon dioxide combining power, glucose, blood urea nitrogen, and hematocrit determinations should be repeated at 12 hour intervals during the early phase of rehydration and then daily until the child has recovered. Additional laboratory studies may be indicated such as blood and stool cultures, skull films, and lumbar puncture. The most difficult problems in rehydration usually occur in the infant who has hypernatremia secondary to excessive loss of water due to diarrhea. The infant is frequently febrile, moderately to severely dehydrated (10 to 15 per cent), acidotic, hyperglycemic, and mayor may not be in shock. The infant is often irritable and obtunded and seizures may have occurred. Restoring tonicity and avoiding the complications of water intoxication (seizures and brain swelling) may be difficult. The urgency of establishing an adequate blood volume in a patient who may be in shock is obvious and must be accomplished rapidly. This may necessitate the use of plasma, albumin, whole blood, normal saline and/or sodium bicarbonate. As a result a larger fluid volume than one would like must be given in a short period of time. In relation to the patient's osmolality, this fluid may be relatively hypotonic and can result in signs of water intoxication. It has been our clinical observation that 20 per cent of the calculated fluid deficit volume can usually be given within the first two hours if the patient is in shock or severely hypovolemic. The rehydration volume can be calculated by using the following formula: (Plasma osmolality per liter) x .6 (body wt in kg) 6 (b d 290 mOsm per liter -. 0 y
W
t·
III
kg) - Fl ·d d fi ·t· Ul e CI III
n1 ers.
Fluid deficit + 1500 ml per M2 for each 24 hours of rehydration equals rehydration fluid in liters. (This volume is calculated to run at a constant hourly rate over the predetermined number of days.) Serum osmolality may be calculated as follows: . . glucose BUN . 2 (serum sodIUm mEq per hter) + --}-8- + -::\- = osmolahty.
The total body sodium deficit in these patients is usually minimal, therefore the rehydrating fluid should not be markedly hypotonic in relation to the plasma, thus preventing the rapid production of a relative water intoxication. A solution containing 35 mEq per liter of sodium chloride, 35 mEq per liter of sodium bicarbonate, or, at times sodium lactate, and 15 to 20 mEq per liter of potassium in glucose or fructose would be ideal as an initial fluid. The severity of the acidosis will of course determine the necessity to correct the pH with additional sodium bicarbonate and/or lactate. In general, less than one third of the anion of the rehydration fluid should be base. A daily requirement of 2 mEq per kg of potassium should be added to the first day's fluid as soon as urine
572
GWENDOLYN
R.
HOGAN
output has been established or earlier if the potassium drops. The potassium concentration should not exceed 40 mEq per liter in the final solution. Hypocalcemia and hypomagnesemia are rarely problems in the management of acute diarrhea and dehydration. 2 If the child has a chronic problem or if intravenous therapy is required for more than 24 hours, this could become a symptomatic problem. Calcium is preferably administered slowly intravenously by the physician in a dose of up to 1 ml of 10 per cent calcium gluconate per kilogram of body weight. In our experience, if the magnesium is low, the serum calcium cannot be maintained at normal levels until the magnesium deficiency is also corrected. It is clear from available clinical and experimental data that as important, if not more important, than the nature of the solution is the rate of infusion.!' 2. 6-8 If fluids are well regulated and contain free water they can be given with no additional problems. In longstanding hypernatremia and in those patients with severe hypernatremia (sodium over 165 mEq per liter), rehydration is best accomplished over a 2 to 3 day period of time. If the sodium is not greater than 150 mEq per liter, the deficit can be corrected in a 24 hour period. If fluids can be taken orally after the first 24 hours of intravenous therapy, this is obviously the best physiological method of treatment. It is clear that the single most important factor in intravenous rehydration is the rate of infusion, not the specific solution. The management of hypernatremia secondary to excessive sodium intake requires free water in order to allow excretion of the excessive sodium. This is best achieved by a very slow rate of rehydration with a solution that is capable of furnishing free water but is not remarkably hypotonic in relation to the patient's serum. In severe cases one may begin with a solution of glucose in normal saline or if less severe (sodium less than 165 mEq per liter), a more hypotonic solution is used. Children with "essential" hypernatremia may have chronic mild elevations of the serum sodium (150 to 155 mEq per liter) without clinical manifestations of their hypertonic state. In this situation rehydration is best carried out by oral administration of fluids. Vasopressin is usually necessary to correct this condition; therefore, the major problem in management is the prevention of water intoxication. However, if the child has superimposed acute hypernatremia, then intravenous therapy (as outlined previously) is required. The major complications of hypernatremia are encountered in those patients with severe dehydration associated with diarrhea. Dehydration and hypovolemia produce shrinkage of the brain which can lead to tearing of the bridging vessels and resultant subarachnoid and subdural hemorrhage. Sludging results in small vessel occlusion, sagittal sinus thrombosis, and renal vein thrombosis. Seizures may occur in the severely dehydrated child as a result of metabolic derangement or infarction due to venous thrombosis. The child with sickle cell anemia presents a real challenge because of the necessity of rapid rehydration. These children may also require exchange transfusion prior to the usual course of rehydration.
HYPERNATREMIA-PROBLEMS IN MANAGEMENT
573
The single most important complication of rehydration therapy is the production of brain swelling secondary to water intoxication. As rehydration is instituted, water enters the cells readily, whereas solute equilibrium will require hours to days,6 and when carried out too rapidly, relative water intoxication is produced. All too often a child is evaluated properly but is rehydrated too vigorously initially and water intoxication results, manifested clinically by stupor, increased intracranial pressure, and seizures. The child may improve initially but then becomes sombulent, "jumpy," and develops generalized seizures. This may occur several hours after treatment has been initiated and frequently at a time when serum sodium is still elevated. It will, however, in almost all cases follow the administration of a rather large volume of fluid that is hypotonic when compared with the patient's serum. For example, a quantity of normal saline given to a patient with a serum sodium of 170 mEq per liter may well produce a relative water intoxication. When seizures occur under these circumstances it is best to treat them with 3 per cent saline calculated to raise the serum sodium 1 to 2 mEq per liter. This will frequently stop the seizures prior to the administration of diphenylhydantoin (Dilantin). Neurological complications dominate the clinical picture and the risk of permanent deficit is greater in the infant IO and in patients with a serum sodium of 160 mEq per liter or greater. 9 Permanent neurological deficit occurs in at least 11 to 15 per cent of these patients. The obvious neurological complications associated with hypernatremia result from shrinkage of the brain with occasional small hemorrhages or more commonly from venous thrombosis of small vessels in the brain resulting in focal signs and/or seizures and the occlusion of large veins or sinuses. These problems are further complicated in those children with sickle cell disease. It is clearly obvious and should be stated emphatically that the sick infant with dehydration and hypertonicity must be treated initially with intravenous fluids. There is no justification for the use of clysis in the management of the acutely ill hypernatremic infant. The complications of treatment, specifically, increased intracranial pressure, brain edema as a result of water intoxication, ischemia, and destruction of brain can be avoided by careful rehydration. Seizures may occur in the dehydrated state as a result of metabolic imbalance or brain injury secondary to venous thrombosis, water intoxication, or ischemia secondary to increased intracranial pressure. The treatment of the seizures will depend upon the underlying etiological factors.
SUMMARY Hypernatremia may be produced under several different circumstances but most frequently is the result of excessive water loss with diarrhea and the excessive solute load secondary to inappropriate preparation of formula. The clinical manifestations vary and depend primarily on the degree of dehydration and the rate at which the hyperosmolar
574
GWENDOLY N
R.
HOGAN
state has been reached. The management of the patient will, of course, depend upon the mechanism of hypernatremia and degree of dehydration and/or hypovolemia that is present. It seems clear that the exact nature of the rehydrating solution is not of major importance. The volume is of great concern but most vital seems to be the rate of rehydration. If rehydration is accomplished too rapidly the child becomes edematous, develops increased intracranial pressure, stupor, and convulsions. If fluids are given slowly and at a well regulated rate, these complications can usually be avoided. The patient should be monitored regularly with electrolytes, careful determination of weights, and records of intake and output. The rate of rehydration should be monitored to assure that the planned schedule of 24 to 72 hours (depending on the severity of the problem) is followed. Approximately 10 to 15 per cent of children with serum sodium of 160 mEq per liter or greater will have permanent neurological deficits.
REFERENCES 1. Banister, A.: Management of hypernatraemia in infancy. Arch. Dis. Child. 48 :825, 1973. 2. Bruck, E., Abal, G., and Aceto, T., Jr.: Pathogenesis and pathophysiology of hypertonic dehydration with diarrhea. Am. J. Dis. Child., 115:281-301, 1968. 3. DuRubertis, F. R., Michelis, M. F., and Davis, B. B.: "Essential" hypernatremia. Arch. Intern Med., 134:889-895,1974. 4. Finberg, L.: The management of the critically ill child with dehydration secondary to diarrhea. Pediatrics, 45: 1029-1036, 1970. 5. Goldberg, M., Weinstein, G., Adesman, J., et al.: Asymptomatic hypovolemic hypernatremia. Am. J. Med., 43:804-810,1967. 6. Hogan. G. R., Dodge, P. R., Gill, S. R., et al.: Pathogenesis of seizures occurring during restoration of plasma tonicity to normal in animals previously chronically hypernatremic. Pediatrics, 43:54-64,1969. 7. Hogan, G. R., Gill, S. R., Master, S., et al.: On the pathogenesis of seizures occurring during rehydration in chronic hypertonic dehyration. Trans. Am. Neurol. Assoc., 90:257260,1965. 8. Hogan, G. R., et al.: Unpublished data. 9. Macaulay, D., and Watson, M.: Hypernatremia in infants as a cause of brain damage. Arch. Dis. Child., 42:485-491,1967. 10. Morris-Jones, P. H., Houston, I. B., and Evans, R. C.: Prognosis of the neurological complications of acute hypernatremia. Lancet, 2:1385-1389,1967. 11. Nitzan, M., and Zelmanovsky, S.: Glucose intolerance in hypernatremic rats. Diabetes, 17:579-581,1968. 12. Prockop, L. D.: Hyperglycemia, polyol accumulation and increased intracranial pressure. Arch. Neurol., 25:126-140,1971. 13. Stevenson, R. E., and Bowyer, F. P.: Hyperglycemia with hyperosmolal dehydration in nondiabetic infants. J. Pediat., 77:818-823,1970.
Department of Pediatrics University of Mississippi School of Medicine 2500 North State Street Jackson, Mississippi 39216
