Iatrogenic hyponatremia in hospitalized children: Can it be avoided?

Abstract

Iatrogenic hyponatremia in hospitalized children is a common problem. It is usually caused by the administration of free water, either orally or through the prescription of hypotonic intravenous fluids. It can result in cerebral edema and death, and is most commonly reported in healthy children undergoing minor surgery. The current teachings and practical guidelines for maintenance fluid infusions are based on caloric expenditure data in healthy children that were derived and published more than 50 years ago. A re-evaluation of these data and more recent recognition that hospitalized children are vulnerable to hyponatremia, with its resulting morbidity and mortality rates, suggest that changes in paediatricians’ approach to fluid administration are necessary. There is no single fluid therapy that is optimal for all hospitalized children. A thorough assessment of the type of fluid, volume of fluid and electrolyte requirements based on individual patient requirements, plus rigorous monitoring, is required in any child receiving intravenous fluids. The present article reviews how hyponatremia occurs and makes recommendations for minimizing the risk of iatrogenic hyponatremia.

One of the most common tasks ascribed to paediatricians is prescribing fluids for hospitalized children. There are many indications for fluid administration in hospitalized children. While the need for administration of isotonic fluids to restore intravascular volume and correct hypotension is accepted, the choice of maintenance fluids in hospitalized children requires some scrutiny.

The traditional guideline for maintenance fluid infusion focuses on the need to replace insensible loss of water for heat dissipation, and is based on caloric expenditure data and deductions that were published more than 50 years ago (1,2). However, the assumptions and deductions are based on the requirements of healthy children, and have recently been challenged (3,4). Indeed, Holliday et al (5) recently modified their initial recommendations based on the recent controversy. The traditional approach to prescribing maintenance fluids in children should be re-evaluated based on the following:

• Hyponatremia is the most common electrolyte disorder in hospitalized patients (both adult and paediatric) (6–8);

• Danger is posed by iatrogenic hyponatremia in otherwise normal children (cerebral edema and death) (9); and

• Administration of hypotonic intravenous (IV) fluids is a major risk factor for developing hyponatremia (10).

In the current article, two cases are presented to illustrate iatrogenic hyponatremia, and some suggestions are provided on how to avoid it in hospitalized children.

CASE PRESENTATIONS

Case 1

A newborn baby weighing 2.2 kg developed chylothorax following repair of coarctation of the aorta, which was treated with chest drainage. Her feeds were changed to two-third strength Portagen (Mead Johnson Nutritionals, USA), (sodium content at full strength of 2.2 mmol/100 mL) at 150 mL/kg/day for six days. Her sodium level five days before the change in feed was 140 mmol/L. Her next serum sodium level, which was measured six days after the diagnosis and change to Portagen feed, was 111 mmol/L. By this time, she had also lost 250 g of her birth weight, but was otherwise asymptomatic.

Why did this baby develop severe hyponatremia?

In six days, the patient’s intake (oral and IV) was 168 mL of free water (dextrose 5% in water to maintain patency of the IV line) and 1453 mL of two-third strength Portagen. Her output consisted of 450 mL of chylous chest drainage (electrolyte concentration similar to serum) and 734 mL of stool and urine combined. Her total sodium balance consisted of an intake of sodium (Portagen = 22 mmol) minus losses (chyle = 58 mmol), plus at least 36 mmol in urine and stools. Case 1 demonstrated unusually large losses of a body fluid; in this case, chest tube drainage. The principles of fluid and electrolyte balance discussed can be applied to losses of any body fluid, such as nasogastric drainage or ostomy losses.

Case 2

A 13-day-old baby weighing 3.18 kg was admitted to hospital because of poor feeding, constipation and failure to thrive. There was no clinical evidence of dehydration. The patient’s admission serum electrolytes were sodium 132 mmol/L, potassium 4.6 mmol/L, urea 5 mmol/L and creatinine 18 μmol/L. An IV infusion was commenced at 20 mL/h of dextrose 5% in water with 2.5 mmol sodium chloride added. During hospitalization, the baby had a low-grade fever, poor feeding and infrequent bowel movements. The baby’s usual feeds were continued in hospital (Similac Advanced [Abbott Laboratories, USA] – sodium content 0.7 mmol/100 mL) prescribed at 100 mL/kg/day to 150 mL/kg/day in addition to the IV fluids. On day 4 of admission, the baby developed increasing abdominal distension with bilious vomiting. Feeds were stopped; IV fluids were changed to dextrose 5% in 0.45% sodium chloride and increased to 130 mL/kg/h. Urinary and stool losses to this point had not been measured. The following day, and after a surgical consultation, the baby was taken to the operating room where a diagnosis of Hirschsprung disease was made. Serum electrolytes on the day of surgery (five days after admission) were as follows: sodium 120 mmol/L, potassium 5.2 mmol/L, urea 3.6 mmol/L and creatinine 15 μmol/L. The last serum electrolyte blood test had been performed four days previously. The baby was transferred postoperatively to the paediatric intensive care unit for analgesia and correction of electrolyte disturbances.

Why did this baby develop severe hyponatremia?

The patient’s total sodium and fluid balance during hospitalization could not be calculated because the baby’s losses were not recorded. However, it is clear the child received hypotonic fluids in excess of requirements in the face of nonosmotic stimuli to antidiuretic hormone (ADH) secretion as a result of the acute illness.

Table 1 lists commonly prescribed IV fluids, and their electrolyte and free water contents.

TABLE 1.

Standard intravenous solutions

Solution	Sodium, mmol/L	Glucose, g/L	Osmolarity, mOsm/L	EFW, %
0.9% NaCl	154	0	308	0
0.45% NaCl, 5% dextrose	77	50	406	50
Ringer’s lactate	130	0	272	16
0.3% NaCl, 3.3% dextrose	53	33	269	66
0.2% NaCl, 5% dextrose	35	50	321	78
5% dextrose	0	50	252	100
3% NaCl	513	0	1027	0

EFW Electrolyte-free water

NaCl Sodium chloride

Factors leading to hyponatremia in these children included failure to monitor and adjust for excessive losses of body fluids, administration of electrolyte-free water through scheduled IV flushes and hypotonic fluids (oral or IV), excessive total fluid administration and failure to regularly monitor serum electrolytes.

PHYSIOLOGICAL BASIS FOR HYPONATREMIA

The above two cases represent failure to appreciate the physiological basis for iatrogenic hyponatremia. There are three requirements for the development of hyponatremia, not all of which are required in any individual patient (11). Most commonly, free water has to be administered and its excretion has to be impaired. In addition, greater urine losses of sodium and potassium relative to serum sodium may also exist. The primary source of electrolyte-free water in hospitalized children is the administration of hypotonic fluids, either intravenously or orally. Impaired free water excretion and greater urinary losses of sodium in hospitalized patients occur because they may have multiple nonosmotic stimuli for ADH secretion, even when the serum sodium level is less than 135 mmol/L (12–14). Some hyponatremic children may be both free water and sodium overloaded. These children require both water and sodium restriction, and careful diuresis and monitoring.

Traditional maintenance IV fluid rates may be excessive for sick hospitalized children due to nonosmotic ADH secretion, and thus they need to be individualized. A recent report (15) documented that the development of hyponatremia is unacceptably high in hospitalized children. The paper reviewed the literature and studied children who presented to the emergency department in a single institution over a three-month period and who had at least one serum sodium level measured (n=1586). The authors argue persuasively that the most important factor for hospital-acquired hyponatremia is iatrogenic. Release of ADH due to nonosmotic reasons may result in a small fall in serum sodium level, which leads to a suppression of thirst. In a nonhospitalized child, this limits free water intake by the child. In contrast, in hospitalized children, once IV fluids are administered, it is the physician rather than the patient who determines the patient’s water intake. The authors suggested that the original guidelines for maintenance fluid may not be applicable in the population of acutely ill children seen today because the complexity and the severity of illness seen in hospitalized children who receive IV fluid therapy (eg, leukemia and complex congenital heart disease) has radically changed; they further suggested the need for a revision of the current recommendations for IV fluid administration in hospitalized children (15). A systematic review of evidence from studies evaluating the safety of administering hypotonic versus isotonic IV maintenance fluids in hospitalized children has been performed (16). A meta-analysis of six studies, which met rigorous criteria, revealed that hypotonic solutions significantly increased the risk of developing acute hyponatremia (OR 17.22; 95% CI 8.67 to 34.2) and resulted in greater morbidity (16). Despite knowledge of the risks, symptomatic hyponatremia and deaths continue to occur (17,18).

POPULATIONS AT RISK FOR HYPONATREMIA

Linking energy expenditure to water losses in hospitalized children overestimates the need for maintenance fluid (19,20). A recent commentary (21) re-evaluated the factors used to calculate water and electrolyte requirements in Holliday and Segar’s article (2), and pointed out that the original calculations did not factor in the unpredictable effect of nonosmotic stimuli for ADH secretion in the acutely ill child (13,22). Those with meningitis, encephalitis, head injury, bronchiolitis, gastroenteritis and chronic lung disease of prematurity, and in association with chemotherapy (essentially any ill child), are at risk for excessive ADH secretion (23–31). Of special relevance in otherwise healthy children, the risk of hyponatremia in postsurgical patients is well recognized (13,22,32,33), and is compounded by the administration of hypotonic solutions. Anesthetic agents, pain, nausea and opiates are also associated with nonosmotic ADH secretion, and contribute to the development of hyponatremia in these children. Compounding this, the inconsistencies of perioperative fluid administration have been recognized as a cause for concern in the anesthesia literature (34).

DANGERS POSED BY HYPONATREMIA

While not all hospitalized children develop hyponatremia, age and skeletal muscle mass relative to body weight are important risk factors (19). Skeletal muscle contains close to 50% body water. The brain is 80% water, of which two-thirds is in the cells, which swell during hyponatremia. Children have a greater brain cell volume to skull volume than teenagers and adults (35). The combination of a smaller muscle mass, and a greater brain cell volume is believed to account for the greater risk of increased intracranial pressure and neurological damage occurring in children who develop hyponatremia (15,36–38).

Hyponatremia is not merely of cosmetic interest. The number of reported deaths and cases of significant neurological damage from hospital-acquired hyponatremia in children receiving hypotonic maintenance solution has increased over the past 10 years (36,39–42). Indeed, several reviews (12,20) have suggested the potential harm associated with the use of these solutions and recommend that their routine use in children be reconsidered. There is an ongoing public enquiry in Northern Ireland into the deaths of three children with hyponatremia (17). There are warnings on Royal College Web sites in the United Kingdom about the dangers of prescribing hypotonic IV fluids to children (43). IV guidelines were recently developed in Ontario in response to the Paediatric Death Review Committee of the Office of the Chief Coroner for Ontario (18). Despite these developments, standard paediatric medicine and anaesthesia texts and guidelines continue to recommend hypotonic maintenance solutions for paediatric patients (44–46).

RISKS ASSOCIATED WITH THE USE OF ISOTONIC FLUID

There have been several concerns raised with regard to the use of isotonic maintenance fluid. The belief that 0.9% saline may cause hypernatremia is not supported (13,14,33,47). To the contrary, 0.9% saline administration as a maintenance fluid does not eliminate the risk of hyponatremia. When large volumes of 0.9% saline are given due to high circulating levels of ADH, such as when restoring intravascular volume or during routine anesthesia, hyponatremia is well described. The subsequent diuresis results in a disproportionate excretion of sodium rather than free water, the so-called urinary desalination (48). Another cited risk of administering large volumes of 0.9% saline is hyperchloremic metabolic acidosis. This is common after large volumes of 0.9% saline are given during acute resuscitation, and mild degrees of hypernatremia can be seen. Several prospective studies (21) in children and adults attest to the safety of 0.9% saline, and a strong case can be made for the prophylactic use of 0.9% saline in all hospitalized children, unless they have demonstrated hypernatremia, rather than waiting for hyponatremia to develop with the traditional maintenance regimen. Considering the well-documented dangers of iatrogenic hyponatremia in hospitalized children, the risk benefit of 0.9% saline versus hypotonic solutions in our opinion falls strongly in favour of 0.9% saline. In addition, because traditional maintenance IV fluid rates may be excessive for sick hospitalized children due to nonosmotic ADH secretion, fluid restriction should be implemented after restoring intravascular volume until the child is stable. This will also significantly reduce the risk of iatrogenic hyponatremia.

RECOMMENDATIONS

• Every hospitalized child requires a thorough assessment of ongoing fluid and electrolyte requirements. Fluid balance should be rigorously monitored by regular weighing and daily monitoring of urea, creatinine and electrolyte levels. Less frequent serum sodium monitoring may be appropriate in a child with a stable medical condition, and after normal free water and sodium balance has been confirmed with daily serum sodium monitoring.

• Fluid prescription orders need to be individualized.

• Electrolyte content of body fluid losses may need to be evaluated.

• Children receiving enteral fluid or nutrition are also at risk of iatrogenic hyponatremia, and must have their serum electrolytes monitored, recognizing that all commercial enteral formulas are hypotonic.