Hypotonic intravenous solutions in children

Review General

Hypotonic intravenous solutions in children Stephen D Playfor 1. Introduction: evolution of the concept of maintenance requirements 2. Physiology of osmolality and tonicity 3. The association between hypotonic fluid administration and hyponatraemia 4. Postoperative hyponatraemia 5. Iatrogenic hyponatraemia in other clinical settings 6. Pathophysiology of hyponatraemia 7. Management of acute symptomatic hyponatraemia 8. Conclusion 9. Expert opinion

Consultant Paediatric Intensivist, Honorary Clinical Lecturer in Paediatric Intensive Care Medicine, Paediatric Intensive Care Unit, Royal Manchester Children’s Hospital, Hospital Road, Pendlebury, Manchester M27 4HA, UK

The use of hypotonic intravenous solutions, especially 4% dextrose/0.18% saline, remains standard practice in many paediatric units in the UK. The practice of prescribing hypotonic intravenous fluids derives from the work of investigators in the 1950s, who produced arbitrarily-derived formulae for calculating the maintenance requirements for water and electrolytes in hospitalised patients. Combining these values led to the widespread acceptance of hypotonic solutions such as 4% dextrose/0.18% saline as ‘standard maintenance’ parenteral fluids. Unfortunately, these calculations do not account for the effects of antidiuretic hormone, the secretion of which is stimulated by many factors encountered during acute illness and especially in the perioperative period. In this setting, the administration of hypotonic intravenous fluids results in the retention of free water and the development of hyponatraemia. The routine administration of hypotonic intravenous fluids has been shown to be associated with severe morbidity and the deaths of many previously healthy children. The problem is compounded by the fact that 4% dextrose/0.18% saline is labelled as ‘isotonic’. Whilst this solution is isosmolar compared to plasma, lack of osmotically effective solutes means that it is hypotonic with reference to the cell membrane. There is no justification for the routine administration of hypotonic intravenous fluids. Keywords: hyponatraemia, hypotonic fluids, intravenous fluids Expert Opin. Drug Saf. (2004) 3(1):

Introduction: evolution of the concept of maintenance requirements 1.

Ashley Publications www.ashley-pub.com

The principles of prescribing parenteral maintenance fluids in children were developed by M Holliday and W Segar, who published a simple formula for calculating the water requirements of hospitalised patients in 1957 [1]. Holliday and Segar calculated water requirements as a direct product of energy expenditure; basal metabolic rate and an estimate of normal energy expenditure were used to arbitrarily define the energy requirement of hospitalised patients. Their assumption was that hospitalised infants have energy requirements that approach those in health, whilst hospitalised children and adults expend less energy, halfway between basal and normal levels. The relationship between water requirements and energy expenditure was described as follows; an allowance of 50 ml of water per 100 kcal consumed would replace insensible loss of water. An additional 66.7 ml of water per 100 kcal consumed would replace the average renal loss and allow for the excretion of approximately isotonic urine. Given that water of oxidation was estimated to provide 16.7 ml of water per 100 kcal consumed, the remaining 100 ml of water per 100 kcal consumed would need to be supplied to patients receiving parenteral fluid therapy. 2004 © Ashley Publications Ltd ISSN 1474-0338

1

Hypotonic intravenous solutions in children

Thus, the formula derived by Holliday and Segar is 100 ml/kg/day for an infant weighing from 0 to 10 kg, 1000 ml + 50 ml/kg/day for a child weighing from 10 to 20 kg, 1500 ml + 20 ml/kg/day for a child weighing > 20 kg. Accepted values for maintenance electrolyte requirements are also arbitrary. Empirically, 1 – 3 mmol of sodium and potassium are required for each 100 kcal consumed. Early investigators produced similar figures for maintenance electrolyte requirements through various calculations, and these were accepted as they broadly reflected the electrolyte composition of human breast milk and cows’ milk. In 1951, Gamble and co-workers reported that administration of carbohydrate to hospitalised fasting patients markedly reduced the metabolic solute load by reducing protein catabolism [2]. The minimal amount of glucose required by an infant to prevent protein catabolism has been calculated as 3 g/kg/ day and 1.5 g/kg/day in adults. Solutions containing 4% dextrose therefore, theoretically provide adequate carbohydrates that prevent gluconeogenesis, protein catabolism and ketogenesis. These values for maintenance requirements of water and electrolytes have led to the acceptance of commercially available solutions such as 4% dextrose/0.18% saline as ‘standard maintenance’ parenteral fluids. The strength of the formula developed by Holliday and Segar is its simplicity and ease of practical application. This is coupled with the innate ability of the kidney to handle a wide range of water and solute loads, as Holliday himself alluded to [3]: ‘I think people have been confused by attempting to make something very precise which does not need to be so. Like weighing a truck on an analytical balance.’ 2.

The association between hypotonic fluid administration and hyponatraemia 3.

Physiology of osmolality and tonicity

A mole is the amount of a substance that contains 6.022 x 1023 molecules. The mass in grams of one mole of a substance is the same as the number of atomic mass units in one molecule of that substance. For example, 1 mole of NaCl = 23g + 35.5g = 58.5g, and 1 mmol (1/1000 of a mole) = 58.5 mg. One osmole equals the molecular weight of the substance in grams divided by the number of freely moving particles each molecule liberates in solution. The osmolal concentration of a substance in a fluid is measured by the degree to which it depresses the freezing point by 1.86 °C. The number of milliosmoles per litre in a solution equals the freezing point depression divided by 0.00186. The osmolality of a solution is the number of osmoles of solute per kilogram of solvent, whereas the osmolarity is the number of osmoles per litre of solution. Osmolality is a measure of the number of particles present in solution and is independent of the size or weight of the particles. Therefore, the osmolarity of a solution is affected by the volume of the various solutes within it and the temperature, whilst the osmolality is not.

2

Serum osmolality can either be measured in the laboratory or may be calculated by adding the concentrations of solutes present within the solution. Any difference between these two values is termed the osmolar gap. The osmotic pressure is the hydrostatic pressure required to oppose the movement of water through a semipermeable membrane in response to an osmotic gradient, where different particle concentrations exist on either side of the membrane. Tonicity is a frequently used term that is open to misunderstanding, as it may be defined in different ways. The most useful definition is that tonicity is the effective osmolality and is equal to the sum of the concentrations of solutes in a solution that have the capacity to exert an osmotic force across a semipermeable membrane. If a solute is able to freely cross a membrane, these solutes are ineffective at exerting an osmotic force across this membrane once they are freely distributed. Tonicity is therefore equal to the osmolality less the concentration of such ineffective solutes. It is clear that tonicity is a property of a solution with reference to a particular membrane, whilst osmolality is a property of a solution that is independent of any membrane. For example, 5% dextrose is initially isosmolar with plasma but, in normal conditions, glucose is a permeant and ineffective solute, which readily enters cells. 5% dextrose is therefore isosmolar with plasma but hypotonic with reference to the cell membrane. Table 1 provides examples of the commonly used intravenous fluids in the UK, with details of their sodium content, osmolality compared to plasma, and tonicity compared to the cell membrane using the above definitions.

Hyponatraemia is reported to be the most common electrolyte disorder amongst hospitalised patients [4]; a large proportion of these cases are iatrogenic. Many physiological stimuli encountered during acute illness result in the non-osmotic release of antidiuretic hormone (ADH); these include pyrexia, nausea, pain, anxiety, reduced circulating volume (especially if by > 8%), drugs, CNS infections, metabolic and endocrine disorders such as hypothyroidism, hypoadrenalism and porphyria, and the postoperative state. The administration of hypotonic intravenous fluids in these circumstances results in the retention of free water and the development of hyponatraemia [5]. ADH may also directly potentiate brain damage by inhibiting sodium efflux from brain cells during the hyponatraemic response and by reducing cerebral blood flow. It has been calculated that a typical intravenous ‘maintenance’ fluid regimen of 100 ml/kg/day of 4% dextrose/0.18% saline would result in the accumulation of ∼ 50 ml/kg/day of electrolyte free water (if it is assumed that half of the water intake will be excreted renally or as insensible losses in an acutely ill child when ADH acts). This constitutes a gain of ∼ 8% in electrolyte free water compared to the total body water (600 ml/kg total body water + 50 ml/kg electrolyte free

Expert Opin. Drug Saf. (2004) 3(1)

Playfor

Table 1. Features of commonly used intravenous fluids in the UK. Solution

Osmolality (mOsmol/L)

Sodium content (mequiv/L)

Osmolality (compared to plasma)

Tonicity (compared to cell membrane)

0.9% saline

308

154

Isosmolar

Isotonic

0.45% saline

154

77

Hyposmolar

Hypotonic

5% dextrose/0.45% saline 432

75

Hyperosmolar

Hypotonic

5% dextrose

-

Isosmolar

Hypotonic

4% dextrose/0.18% saline 284

31

Isosmolar

Hypotonic

Ringer’s lactate

273

130

Isosmolar

Isotonic

4.5% human albumin solution

275

100 – 160

Isosmolar

Isotonic

278

water), which would proportionately decrease the serum sodium concentration from 140 to 129 mmol/l after 24 h [6]. 4.

Postoperative hyponatraemia

As early as 1905, Pringle and co-workers reported on the occurrence of oliguria following surgery under general anaesthesia regardless of the volume of fluid administered [7]. In 1935, Helwig and co-workers reported the first case of autopsy to confirm the presence of cerebral oedema following postoperative hyponatraemic encephalopathy [8]. Helwig described this syndrome as ‘water intoxication’ and went onto document the same clinical, biochemical and pathological features in a rabbit following the rectal administration of water. This was followed by the reports of Coller in 1944, and later by Wilkinson, that patients were unable to excrete large loads of sodium and water administered intravenously in the postoperative period [9-11]. In 1952, Zimmerman and colleagues reported 17 cases of convulsions after surgery with 3 fatal cases. The mean age of the patients was 67.4 years, and 16 cases had a serum sodium concentration of between 94 and 126 mmol/l [12]. In a series of papers published in 1958, Moore described oliguria and fluid retention as being normal after trauma or surgery and that administration of hypotonic saline would cause lowering of the serum sodium concentration and coma [13]. In 1979, Thomas and Morgan reported on 24 postoperative patients who received either isotonic or hypotonic saline solutions [14]. They found that, whilst the urinary ADH level was high in both groups, the serum sodium concentration only fell in the group administered hypotonic saline. In 1983, Burrows and co-workers conducted a similar study in 24 paediatric patients undergoing spinal surgery. All patients receiving hypotonic saline solutions experienced a significant fall in the serum sodium concentrations (6.2 ± 2.9 mmol/l) and the authors recommended the use of isotonic saline solutions in the postoperative period [15].

In 1988, Cowley and colleagues reported on a series of ten children undergoing spinal surgery who demonstrated fluid retention when administered hypotonic saline in the postoperative period. One suffered generalised seizures and was found to have a serum sodium concentration of 122 mmol/l, whilst another suffered a respiratory arrest and died with a serum sodium concentration of 118 mmol/l [16]. There are many other case reports describing the development of hyponatraemia in paediatric patients following surgery associated with the administration of hypotonic intravenous fluids [17-20]. Judd and co-workers studied the levels of ADH following tonsillectomy in children, along with plasma renin activity [21]. In the study, patients received intravenous isotonic saline during and after the procedure and had lower levels of plasma ADH and lower urine osmolalities. These findings suggest that hypovolaemia is the principle stimulus for ADH release following surgery. Under physiological conditions, the usual stimulus for ADH release is a rise in plasma osmolality or tonicity. Hypovolaemia is a particularly potent non-osmolar factor leading to ADH release, particularly where the circulating volume is reduced by > 8%. There is also evidence that osmolar and non-osmolar stimuli for the release of ADH interact [22]. Thus, patients who are starved and deprived of fluids and who are then exposed to the vasodilatory effects of anaesthesia and to blood loss and third space loss experience a powerful stimulation for ADH secretion. These factors, together with the effects of drugs, intercurrent illness, pain, vomiting and sweating, mean that the perioperative period may include several days of increased ADH secretion. If peri- and postoperative fluid replacement is with hypotonic solutions, there will be an increased risk of hyponatraemia. Table 2 shows paired daily plasma and urine osmolalities in a series of patients who had undergone uncomplicated abdominal surgery and who were administered hypotonic fluids in the postoperative period [23]. Even when an isotonic or near isotonic solution is administered postoperatively, patients may become hyponatraemic. In a study by Steele and co-workers, 21 out of 22 patients who


3


Table 2. Variation in plasma and urine osmolality in the days following surgery. Days after operation

0

1

2

3

4

5

Plasma osmolality 282 (mOsmol/L)

279

273

273

270

270

Urine osmolality (mOsmol/L)

696

610

358

489

611

579

Used with permission from: ALLISON SP: Metabolic aspects of intensive care. Br. J. Hosp. Med. (1974) 861-872.

received these types of fluid showed a fall in serum sodium concentration from a mean of 140 mmol/l to a mean of 136 mmol/l [24]. Most of these patients were excreting a hypertonic urine, with a mean sodium concentration of 150 mmol/l, and were in positive fluid balance. The implication was that ADH was acting in a process referred to by the authors as ‘desalination.’

Iatrogenic hyponatraemia in other clinical settings 5.

There are also several case reports describing the occurrence of hyponatraemia in children following the use of hypotonic fluids in common childhood illnesses. Arieff and co-workers reported the cases of 16 children who suffered hyponatraemic encephalopathy after receiving hypotonic intravenous fluids following minor surgical procedures or during common childhood infections; highlighting that cases such as these do not only occur after surgery. Of these 16 children, 15 died and 1 was left with profound neurological deficit [25]. It has also been suggested that iatrogenic hyponatraemia, through the administration of hypotonic intravenous fluids, may be responsible for the deterioration of children with La Crosse encephalitis [26]. The author has also recently described the case of a 13month-old girl with no significant past medical history who was admitted to hospital following a 48-h history of diarrhoea and vomiting [27]. Clinical examination revealed lethargy and signs of moderate dehydration, and her initial serum sodium concentration was 137 mmol/l. She was given 150 ml i.v. of 4.5% human albumin solution and commenced on intravenous fluids using 4% dextrose/0.18% saline at a rate of 105 ml/kg/day. Soon she was able to tolerate oral fluids and drank 350 ml of water and cordial over the next few hours. Twelve hours after admission, the child suffered a 3-min generalised convulsion. At this time, her serum sodium concentration had fallen to 120 mmol/l. Within the first 12 h of admission, the child had received a total of 950 ml of fluid (114 ml/kg), including 600 ml of hypotonic intravenous fluids. Unfortunately, she subsequently had a further respiratory arrest and developed bilateral fixed, dilated pupils. The child was admitted to the regional paediatric intensive care unit, where she died. An extensive post-mortem examination

4

revealed diffuse cerebral swelling with necrosis of the cerebellar tonsils. All bacteriological and viral studies were negative. The administration of hypotonic intravenous fluids during resuscitation has also been associated with the development of severe hyponatraemia, cerebral oedema and death in children in the accident and emergency department [28]. Gregorio and co-workers also described a previously healthy 4½-year-old boy who presented with watery diarrhoea and vomiting secondary to rotavirus infection [29]. On admission, he appeared mildly dehydrated and his serum sodium level was 136 mmol/l. He was rehydrated with intravenous fluids using 5% dextrose/0.45% saline. Four hours later he became unresponsive and his serum sodium concentration was found to be 118 mmol/l. A magnetic resonance imaging (MRI) brain scan revealed multiple pathogenic foci, and a diagnosis of central pontine myelinosis (CPM) was made. The boy suffered moderate-to-severe neurological impairment. More recently, Hanna and co-workers have conducted a retrospective review of the incidence and evolution of hyponatraemia associated with respiratory syncytial virus bronchiolitis [30]. They studied a total of 130 infants, of whom 33% had a serum sodium level of < 136 mmol/l and 11% had a serum sodium level of < 130 mmol/l. Four infants suffered hyponatraemic seizures at admission, three of whom had received hypotonic intravenous fluids prior to admission. 6.

Pathophysiology of hyponatraemia

Severe hyponatraemia may produce a variety of clinical features as water leaves the extracellular fluid pool, enters brain cells and results in cell swelling. These include anorexia, nausea, headache, lethargy, apathy, disorientation, agitation, delirium, seizures, focal neurological deficits, pathological reflexes, and respiratory arrest. Progressive cerebral swelling may lead to raised intracranial pressure, tentorial herniation and death. Frequently, the correct diagnosis is not promptly considered and patients are commonly administered antibiotics and antiviral agents and sent for neuroimaging studies before the hyponatraemia is properly addressed. When exposed to hyponatraemia, the brain is protected by a series of hormonal and physical responses. There is decreased ADH secretion in response to reduced plasma osmolality, which reduces thirst and renal free water absorption. There is an active reduction in cerebral blood and cere-


Playfor

brospinal fluid (CSF) fluid pools, and a cellular response characterised by the extrusion of intracellular osmolytes to reduce intracellular water and reach a new osmotic equilibrium. Sodium, potassium, chloride, amino acids, polyalcohols, and methylamines all contribute to volume regulation. The movement of ions occurs during the early phase, and movement of organic osmolytes occurs during the late phase of the regulatory process. These responses occur within 48 h and enable effective volume regulation during hyponatraemia. It has also been suggested that this adaptive accumulation of organic osmolytes (especially glutamate and aspartate) within the extracellular fluid of the brain may cause or aggravate seizure activity in the hyponatraemic state. It is the effectiveness of the adaptive process that leads to the observation of a much lower incidence of neurological symptoms as well as lower mortality in chronic, rather than acute, hyponatraemia. It is clear that the rate of fall of plasma sodium is usually more strongly correlated with morbidity and mortality than the absolute decrease. Paediatric patients are at particularly high risk for developing symptomatic hyponatraemia, as they tend to develop hyponatraemic encephalopathy at higher serum sodium concentrations than adults. Children also have a poorer prognosis, and this is probably due to a combination of physical and physiological differences. Children have a higher brain:skull size ratio than adults; their brains reach adult size by 6 years of age, 10 years before their skulls attain their final adult dimensions. Also, in older adults there is a progressive loss of brain volume, whilst the volume inside the skull remains constant. Brain water content is twice as high in children than in adults and the intracerebral volume of CSF is 10% higher in adults than in children, allowing for greater adaptation by reduction of and expansion into the CSF pool [25]. Whilst among adults the risk of developing postoperative hyponatraemic encephalopathy is equal for men and women, menstruant women are 25 times more likely to suffer severe morbidity or death compared with men or postmenopausal women. This may relate to hormonal factors, and it has been suggested that oestrogen impairs brain volume regulation in menstruant women. In a study of 65 patients with postoperative hyponatraemic encephalopathy by Ayus and co-workers [31], 97% of the 34 patients who died or had permanent brain damage were women. However, it has been highlighted that more than half of the men in the study were undergoing urological procedures, a well-recognised setting for the development of hyponatraemia in which prompt recognition and treatment may have lead to a lower mortality and morbidity. It has also been suggested that the overaggressive correction of hyponatraemia may itself lead to brain injury. Rapid normalisation of serum sodium concentration may lead to CPM or to a related and more diffuse condition known as osmotic demyelination syndrome or pontine and extrapontine myelinolysis. CPM is rare, with few cases described in children. It was first described in 1959 and is a demyelinating process that affects the central pons and may involve other areas of white

matter. It tends to present with an altered level of consciousness followed by spastic quadriparesis and pseudobulbar palsy. It is often described as a complication following rapid changes in sodium levels, particularly in the setting of an underlying disease process such as liver disease. The susceptibility to demyelination increases with the severity and duration of the pre-existing hyponatraemia and myelinolysis is rare in cases of mild or acute hyponatraemia.

Management of acute symptomatic hyponatraemia 7.

Acute symptomatic hyponatraemia is a medical emergency that must be treated promptly and aggressively. Treatment is as follows: • Maintain the airway; endotracheal intubation and mechanical ventilation if required, administer 100% oxygen. • Ensure adequate circulation. • Discontinue any hypotonic intravenous infusions. • Administer hypertonic saline to rapidly correct serum sodium concentration using the formula described below. • Admit to intensive care facility. • Assess fluid and tonicity balance of patient. • Re-assess frequently. Hypertonic saline solution should be used to rapidly correct the serum sodium concentration in acute symptomatic hyponatraemia; 3% sodium chloride solution (513 mmol/l) is frequently used in this setting. The amount of sodium required is calculated according to the following formula: mmol of Na required = (desired serum Na - present serum Na) x 0.6 x weight (kg) Sodium should be corrected to a desired serum sodium concentration of 130 mmol/l. In the setting of acute but symptom-free hyponatraemia or chronic hyponatraemia, hypertonic saline infusions should not be used. In these circumstances, any hypotonic intravenous infusions should be discontinued and the aim should be to correct the serum sodium concentration much more slowly, by 0.5 – 2 mmol/l/h. There have been attempts to prevent this problem from arising: following the death of two children who were administered 5% dextrose in the accident and emergency department [28], the paediatric accident and emergency department at the Primary Children’s Medical Center of Salt Lake City, USA removed 5% dextrose as a stock item. In addition, all emergency medical system protocols within the city and the county removed all reference to 5% dextrose and other hypotonic fluids. Also, at a hospital in the UK, Cosgrove and co-workers described how clinical practice was changed after a near miss [32]. A 3-year-old child with gastroenteritis was rehydrated with intravenous fluids using 4% dextrose/0.18% saline at a ‘maintenance’ rate. Overnight, her serum sodium concentration fell from 136 to 127 mmol/l, although she remained


5


symptom free. Following this critical incident, stocks of 4% dextrose/0.18% saline were removed from the ward area. In the 3 years after this action was taken no further instances of iatrogenic hyponatraemia were reported. 8.

Conclusion

The use of hypotonic intravenous solutions is associated with the development of profound hyponatraemia, which can result in severe neurological morbidity and death. This effect is mediated by the release of ADH during illness and is particularly common in the perioperative period. 9.

Expert opinion

The use of hypotonic intravenous solutions, especially 4% dextrose/0.18% saline, remains standard practice in many paediatric units. This practice is based on arbitrary calculations of normal physiological requirements for salt and water made in healthy children > 40 years ago. Every year, otherwise healthy children tragically suffer brain damage and die solely because of the inappropriate administration of these fluids. Healthy children rarely require intravenous fluids; in acute illness, and especially in the perioperative setting, there are many triggers to the secretion of ADH, which inhibits the ability to excrete electrolyte-free water. The key to avoiding hyponatraemia in this setting is to maintain tonicity balance; the physician must match what the patient is excreting, in terms of both volume and electrolyte content, with what is being administered.

Bibliography Papers of special note have been highlighted as either of interest (•) or of considerable interest (••) to readers. 1.

••

2.

3.

4.

6

HOLLIDAY M, SEGAR W: The maintenance need for water in parenteral fluid therapy. Pediatrics (1957) 19:823-832. Landmark paper providing a simple calculation for maintenance requirements of water and electrolytes. GAMBLE JL: Lane Medical Lectures. Companion of water and electrolytes in the organization of body fluids. Stanford University Publication (1951) Volume V, Number 1. CHESNEY RW: The maintenance need for water in parenteral fluid therapy. Pediatrics (1998) 102:399. ARIEFF AI: Effects of water, acid–base and electrolyte disorders on the central nervous system. In: Fluid, electrolyte and acid–base

5.

•

There is no justification for the routine administration of hypotonic intravenous fluids. Hypotonic intravenous fluids should not be stocked in the prehospital setting, accident and emergency departments or general ward areas. Their use should be restricted to intensive care units and some specialist ward areas. There are good reasons for recommending the exclusive use of isotonic intravenous fluids in all sick patients [33]. Peri- and postoperative patients, for whom there are 100 years of data describing the harmful effects of hypotonic fluids, should be administered 0.9% saline, 5% dextrose/ 0.9% saline or Hartmann’s Solution. Some patients with ongoing free water losses or a previously established deficit may require a more hypotonic fluid. The use of 5% dextrose/ 0.45% saline may therefore be justified in the sick, non-operative patient, although it must be remembered that this solution is hypotonic with reference to the cell membrane, provides 50% electrolyte-free water (compared to 0% with 0.9% saline or 5% dextrose/0.9% saline) and its use may be associated with occasional cases of severe neurological morbidity [29]. The continued labelling of products such as 4% dextrose/ 0.18% saline as ‘isotonic’ is misleading and perpetuates this problem. This solution is isosmolar compared to plasma but lack of osmotically-effective solutes means that it is hypotonic with reference to the cell membrane. Changing entrenched practice calls for concerted educational efforts by the responsible authorities; this would be reinforced if labelling of tonicity to plasma were replaced with more clinically relevant labelling of tonicity to the cell membrane, and warnings that the use of such products is associated with the development of severe hyponatraemia.

disorders. Arieff AI, Defronzo RA (Eds), Churchill Livingstone, New York (1985):992-1007.

9.

COLLER FA, CAMPBELL KN, VAUCHAN AH et al.: Postoperative salt intolerance. Ann. Surg. (1944) 119:533.

HALBERTHAL M, HALPERIN ML, BOHN D: Acute hyponatraemia in children admitted to hospital: retrospective analysis of factors contributing to its development and resolution. Br. Med. J. (2001) 322:780-782. Thorough retrospective review of hyponatraemia due to hypotonic fluid administration.

10.

WILKINSON AW, BILLING BH, NAGY C et al.: Excretion of chloride and sodium after surgical operations. Lancet (1949) 1:640.

11.

WILKINSON AW, BILLING BH, NAGY C et al.: Excretion of potassium after partial gastrectomy. Lancet (1950) 2:135.

12.

ZIMMERMAN B, WANGENSTENN OH: Observations on water intoxication in surgical patients. Surgery (1952) 31:654-659.

13.

MOORE FD: Common patterns of water and electrolyte change in injury, surgery and disease. N. Engl. J. Med. (1958) 258:277-285, 325-333, 337-384, 427-433.

14.

THOMAS TH, MORGAN DB: Post-surgical hyponatraemia: the role of intravenous fluids and arginine vasopressin. Br. J. Surg. (1979) 66:540-542.

6.

DURWARD A, TIBBY SM, MURDOCH IA: Hyponatraemia can be caused by standard fluid regimens. Br. Med. J. (2000) 320:943.

7.

PRINGLE H, MAUNSELL RCB, PRINGLE S: Effects of ether anaesthesia on renal activity. Br. Med. J. (1905) 2:942.

8.

HELWIG FC, SCHUTZ CB, CURRY DE: Water intoxication: report of a fatal human case with clinical, pathologic and experimental studies. JAMA (1935) 104:1569-1575. Expert Opin. Drug Saf. (2004) 3(1)

Playfor

15.

BURROWS FA, SHUTACK JG, CRONE RK: Inappropriate secretion of antidiuretic hormone in a postsurgical pediatric population. Crit. Care Med. (1983) 11:527-531.

16.

COWLEY DM, PABARI M, SINTON TJ et al.: Pathogenesis of postoperative hyponatraemia following correction of scoliosis in children. Aust. NZ. J. Surg. (1988) 58:485-489.

17.

18.

•

19.

ARMOUR A: Dilutional hyponatraemia: a cause of massive fatal intraoperative cerebral oedema in a child undergoing renal transplantation. J. Clin. Pathol. (1997) 50:444-446. HUGHES PD, MCNICOL D, MUTTON PM, FLYNN GJ, TUCK R, YORKE P: Postoperative hyponatraemic encephalopathy: water intoxication. Aust. NZ. J. Surg. (1998) 68:165-168. Very thorough literature review describing the historical observations of postoperative hyponatraemia. MCRAE RG, WEISSBURG AJ, CHANG KW: Iatrogenic hyponatremia: a cause of death following pediatric tonsillectomy. Int. J. Pediatr. Otorhinolaryngol. (1994) 30:227-232.

20.

BOHN D: Children are another group at risk of hyponatraemia perioperatively. Br. Med. J. (1999) 319:1269.

21.

JUDD BA, HAYCOCK GB, DALTON RN, CHANTLER C: Antidiuretic hormone following surgery in children. Acta Paediatr. Scand. (1990) 79:461-466.

22.

ROBERTSON GL, BERL T: Water metabolism. In: The kidney (3rd edn). Brenner BM, Rector FC (Eds), WB Saunders, Philadelphia (1986):385.

23.

ALLISON SP: Metabolic aspects of intensive care. Br. J. Hosp. Med. (1974) 861-872.

24.

25.

••

26.

27.

28.

•

STEELE A, GOWRISHANKAR M, ABRAHMSON S, MAZER D, FELDMAN R, HALPERIN ML: Postoperative hyponatremia despite isotonic saline infusion. Ann. Intern. Med. (1997) 126:20-25. ARIEFF AI, AYUS JC, FRASER CL: Hyponatraemia and death or permanent brain damage in healthy children. Br. Med. J. (1992) 304:1218-1222. Large study identifying the high morbidity and mortality associated with hyponatraemia due to hypotonic fluid administration.

29.

GREGORIO L, SUTTON CL, LEE DA: Central pontine myelinolysis in a previously healthy 4-year-old child with acute rotavirus gastroenteritis. Pediatrics (1997) 99:738-743.

30.

HANNA S, TIBBY SM, DURWARD A, MURDOCH IA: Incidence of hyponatraemia and hyponatraemic seizures in severe respiratory syncytial virus bronchiolitis. Acta Paediatr. (2003) 92:430-434.

31.

AYUS JC, WHEELER JM, ARIEFF AI: Postoperative hyponatremic encephalopathy in menstruant women. Ann. Intern. Med. (1992) 117:891-897.

32.

COSGROVE M, WARDHAUGH A: Iatrogenic hyponatraemia. Arch. Dis. Child. (2003). E-letter.

33.

MORITZ ML, AYUS JC: Prevention of hospital-acquired hyponatraemia: a case for using isotonic saline. Pediatrics (2003) 111:227-230. Articulate account of the clinical problem and the rationale for prescribing isotonic fluids.

MORITZ ML, AYUS JC: La Crosse encephalitis in children. N. Engl. J. Med. (2001) 345:148.

•

PLAYFOR S: Fatal iatrogenic hyponatraemia. Arch. Dis. Child. (2003) 88:646-647.

Affiliation

JACKSON J, BOLTE RG: Risks of intravenous administration of hypotonic fluids for pediatric patients in ED and prehospital settings: let’s remove the handle from the pump. Am. J. Emerg. Med. (2000) 18:269-270. Important paper describing a positive intervention to reduce the inappropriate administration of hypotonic fluids.


Stephen D Playfor MBBS DCH MRCP MRCPCH MD Consultant Paediatric Intensivist, Honorary Clinical Lecturer in Paediatric Intensive Care Medicine, Paediatric Intensive Care Unit, Royal Manchester Children’s Hospital, Hospital Road, Pendlebury, Manchester M27 4HA, UK Tel: +44 0161 727 2978; Fax: +44 0161 727 2198; E-mail: [email protected]

7