Fluid management for children presenting to the emergency department: Guidelines for clinical practice

Cardiovascular 

The physical immaturity and smaller size of the infant compared to the adult result in significant structural and functional differences in the cardiovascular system and the maintenance of fluid balance. These differences impact on the response to hypovolaemia and dehydration of the infant and child and in turn this effects clinical assessment, the interpretation of findings and the strategies used to restore the circulation and manage fluid balance.1, [2]

Organ perfusion is dependent on a number of factors including cardiac output, which is a function of heart rate and stroke volume. Adult heart rate and stroke volume are variable and tightly controlled to ensure adequate perfusion. However, fewer actomyosin elements and mitochondria render the infant myocardium unable to increase contractility resulting in a relatively fixed stroke volume. For this reason, infants have a greater reliance on increases in heart rate to increase cardiac output when compared to the adult.1, [2]

Release of calcium from the sarcoplasmic reticulum in the infant myocardium is less efficient than from adult myocardium making infant contractility more dependent on extracellular calcium concentrations, which has implications for fluid administration and electrolyte monitoring.[2], 3

Infant myocardial compliance is also lower than the adults and so for the same ventricular filling volume infants will experience higher atrial pressures. This has a negative effect on preload and also restricts capacity to increase stroke volume.2 This helps explain the infants relatively poor response to aggressive fluid resuscitation which aims to augment cardiac output.1

Despite this the cardiac index is at its highest in infancy (300ml/(minkg) compared to 70–80ml/(minkg) in adults), which is a reflection of a much higher resting heart rate.1 This serves to meet the substantially higher metabolic requirements of the infant.

Blood pressure increases with age and is predominantly a result of a corresponding increase in systemic vascular resistance.1 Assessment of blood pressure will be discussed more extensively in one of the following sections.

Fluid and electrolytes 

The distribution of body water across fluid compartments also varies with age. However, after approximately 6 months of age total body water remains constant at around 60–65ml/kg having fallen from 80ml/kg at birth and the distribution between intracellular and extracellular fluid resembles the adult.2 The difference in total body fluid and distribution are partly responsible for the infants increased susceptibility to dehydration. Additionally, infant circulating blood volume is higher than the circulating blood volume of an adult (70–80ml/kg compared to 60–70ml/kg).1

The glomerular filtration rate (GFR) and renal blood flow of the infant are lower than the adults GFR and the infant kidney has less capacity to concentrate the urine and excrete a solute load than the adult kidney. Renal immaturity limits the capacity of the kidneys to respond aggressively to hypovolaemia, low cardiac output and dehydration.2 In addition, the infant is less able to acidify the urine and is more sensitive to hyperchloraemic acidosis resulting from large volume resuscitation with normal saline.[2], 3

Accuracy of assessment parameters 

Recognition of these signs and correctly identifying them as evidence of hypovolaemia is essential as a drop in blood pressure is unlikely to be evident until approximately 30% of circulating volume is lost.[2], 3

The utility of clinical signs for the detection of hypovolaemia is dependent on their ability to detect hypovolaemia early and whether these signs are similarly recognized and interpreted by clinicians. There is some evidence to indicate that assessment of indirect measures of shock such as capillary refill time, pulse strength and tissue turgor is at best only moderately reliable.13

Capillary refill 

Assessment of capillary refill time is recommended as a non-invasive measure of peripheral perfusion and an indirect indicator of circulating intravascular volume.1 Clinicians are cautioned about relying exclusively on this parameter, as the evidence for its sensitivity for shock in young children is not convincing. Leonard and Beattie's study demonstrated no correlation between prolonged capillary refill time (greater than 3s) and bacterial infection.14 These findings are consistent with findings from a study comparing capillary refill time with invasive cardiovascular indices, which showed poor correlation between the two.15

Studies focusing on the value of capillary refill time for estimating severity of dehydration have had greater success and indicate that it is likely to be a valuable assessment tool.16, 17 Gorelick et al. has shown that capillary refill time is not affected by the presence of fever.18 Other studies indicate that sensitivity and specificity are likely to be effected by ambient temperature and the technique and site of measuring CRT.15, 16, 19, 20, 21

However, in the absence of more suitable parameters, capillary refill time assessment remains a means to assessing peripheral perfusion and to some extent provides an estimate of fluid balance for children presenting to the ED and two seconds is the accepted upper limit of normal.

Blood pressure 

As was described earlier a drop in BP is a sign of substantial intravascular volume loss. The majority of children with acute illness do not reach this stage and our aim is for early recognition of hypovolaemia before hypotension occurs.

Hypotension is defined as a systolic blood pressure (SBP) lower than the 5th percentile for age and Table 1 provides accepted values for different age groups.1, 22, 23 Variation in published values exists and this is likely to be a result of the lack of evidence supporting these recommendations. Using blood pressure measurements in well children to calculate the 5th percentile, Haque and Zaritsky make one of the first and only attempts to substantiate the values used to define hypotension.22 The results of their work indicate that most published recommendations for the lowest acceptable systolic blood pressure are actually higher than the 5th percentile for well children. Additionally, they reveal that height also impacts significantly on blood pressure, which has not been previously recognized. The authors developed a simple formula from their data to assist the clinician to calculate systolic blood pressure for children of different ages (1–17 years):

Despite this study producing SBP values founded on the best evidence base to date, the authors conclude that adoption of a higher threshold for minimum acceptable blood pressure for unwell children is a clinically rationale approach. Like capillary refill time, blood pressure monitoring is a valuable assessment tool but treatment decisions should not rest solely on this parameter, particularly where blood pressure measurement is considered normal.

Blood pressure measurements are performed less frequently on children than adults presenting to the emergency department. This is likely to reflect the practical difficulties measuring the BP in young children and infants and concerns about the accuracy of the measurement. Young children are rarely cooperative and variation in their size effects cuff selection which along with the level of anxiety, the position of the arm in relation to the heart and the type of manometer will affect the result.24 Clark et al. studied the effect of two methods for estimating the appropriate cuff size for children aged 5 days to 22 years and found deficiencies in all methods studied.25 A cuff most closely representing 40% of the circumference of the upper arm gives a measurement closest to invasive radial blood pressure values.

Laboratory values 

Base deficit is considered a measure of the adequacy of cardiac output and tissue oxygenation. Given the urgency of resuscitation of the circulation and the potential delay accessing laboratory results, it has no role in guiding immediate fluid resuscitation of hypovolaemic children. Base deficit is better used as an indicator of the success of earlier treatment and a guide to ongoing resuscitation. Bicarbonate has been repeatedly shown to be the most sensitive measure of the severity of dehydration17, 26, 27, 28, 29 while other parameter such as specific gravity of urine, the presence of ketones and the urine output during rehydration are not good indicators of the severity of dehydration.26, 30 However, the evidence supporting laboratory tests is not sufficiently compelling and recommendations for assessment of dehydration severity rarely include routine laboratory testing.

Dehydration assessment tools 

Dehydration is best detected and quantified by changes in total body weight.3 However, accurate recent weights are often not available for young children presenting to the emergency department and clinicians must rely on assessment of the clinical features of dehydration to determine the severity. Study authors repeatedly conclude from results that clinicians poorly estimate dehydrations severity, with one Australian study revealing overestimation by as much as 3.2%.17, 29 This provides persuasive support for serial weight measurements for children presenting to the ED with the potential for fluid loss.

There is considerable overlap between the signs considered likely to indicate dehydration and those indicating hypovolaemia. Furthermore, hypovolaemia is associated with severe dehydration resulting in these children exhibiting the signs of hypovolaemic shock. Significant attention has been placed on identifying clinical signs, which are sensitive and specific for dehydration and its severity.

A history of fluid loss (diarrhea and vomiting), low urine output and poor fluid intake are considered indicative of the likely presence of dehydration. However, a meta-analysis summarizing data from three studies show that these symptoms have low predictive value for 5% dehydration.17, 31, 32, 33 Of greater value to hydration assessment is parental report of normal urine output, which positively predictive for normal hydration.33

Three signs are shown to have positive predictive value for detecting dehydration and they are prolonged capillary refill time, abnormal skin turgor and abnormal respiratory pattern.16, 17, 26, 33, 34 Otieno et al.13 report that skin turgor is the single most reliable sign of dehydration in children. Signs such as the absence of tears, cool extremities, sunken eyes, dry mucus membranes, ill appearance and a weak pulse may be of some clinical value to the assessment of hydration but results are variable and demonstrate weaker predictive value.

Given the variability in study results, the rational approach for clinicians is to approximate severity of dehydration based on the number of signs of dehydration identified rather than the presence of specific individual signs. Several authors have published results demonstrating that dehydration assessment tools comprised of several signs have greater predictive value than individual signs.33, 34 Friedman et al. identified the combination of general appearance, eyes (sunken), mucous membranes, and tears as likely to give the most sensitive indication of dehydration severity.35 Table 4 provides details for the Gorelick et al. [18] model for estimating dehydration severity.

Hypovolaemia 

Hypotension has been repeatedly demonstrated to be associated with higher rates of mortality36, 37 and aggressive early normalization of blood pressure values improves survival.1, [2], 3, 38 Only small numbers of children present with hypotension while much larger numbers reveal signs of hypovolaemia requiring fluid resuscitation.

The aim of treatment is to increase circulating blood volume and fluids are selected based on their capacity to achieve this. Hypotonic fluids do not remain in the intravascular space and are therefore not suitable for resuscitation. Colloids were considered likely to remain in the circulation for longer than isotonic crystalloids and for many years they were the mainstay of fluid resuscitation.

Initially, albumin was viewed as the ideal colloid; however early study results prompted concern that resuscitation with albumin increased mortality. Recommendations were qualified to advocate for albumin as a more appropriate volume expander in specific circumstances, e.g., sepsis and while less appropriate in others, e.g., head trauma.39, 40

The results of the recently updated Cochrane review show that isotonic crystalloids and colloids are equally suited to restore circulating blood volume despite the theoretical advantage attributed to colloids and that an increase in adverse events is not associated with specific fluids.41 Crystalloid fluids are widely available and inexpensive making them the fluid type of choice.39, 41 Normal saline has been the most commonly recommended crystalloid for fluid resuscitation. However, overuse carries some risk in the form of hyperchloraemic acidosis. More recently Hartmann's solution is gaining favour as this solution unlike normal saline contains lactate which when converted to bicarbonate acts as a buffer.3

Volume recommendations for treatment of hypovolaemia most frequently support 20ml/kg boluses of fluid.1, [2], 3 This choice is largely arbitrary and founded on a study which suggests that signs of hypovolaemic shock are exhibited with loss of approximately 30% of circulating volume, which equates to approximately 20ml/kg.3 Fluid will only remain in circulation for a relatively short time and repeat bolus of fluid are likely to be required particularly where there is a total body fluid deficit or large volume shifts as in sepsis. Aggressive fluid administration of up to 80–100ml/kg over an hour may be required to achieve adequate perfusion in children with sepsis.

Direct intravenous access is the most common route of administration of fluids, however in children with profound hypovolaemia and poor peripheral perfusion gaining access can be extremely difficult. Intraosseous access (IO) is the alternative and clinicians should have a low threshold (one to two brief attempts to cite an intravenous (IV) cannula) for using this route for administration of fluids to the shocked child.3 Intraosseous insertion is much easier than IV insertion and increasingly this is made even easier with the availability of devices such as bone drills and bone guns.

Fluids given via an IO will not run on a gravity drip set and must be pumped under pressure. This can be achieved using manual blood pump sets, syringes and pressure bags. Rapid delivery of fluids via small IV cannulae will also require higher pressure. For smaller volumes fluid boluses can be given by ‘pushing’ fluids with a 20 or 50ml syringe connected to a three-way tap in the fluid line (see Fig. 1). Administration of larger volumes will require use of infusion pumps, pressure bags or blood pump sets.

View Large Image Download to PowerPoint

Figure 1. Line set up to administer bolus of fluid via an intraosseous cannula or a small gauge IV cannula: (A) withdrawing fluid from fluid bag into syringe; (B) injecting fluid into the patient.

Dehydration 

Intravenous replacement of fluid is not considered the ideal method of fluid administration for dehydration in the absence of associated hypovolaemia. Evidence from a meta-analysis and a systematic review demonstrate that oral replacement of the fluid deficit is achieved at least as effectively as intravenous replacement of the deficit42, 43 and may be safer than intravenous rehydration.43 Oral fluid replacement largely avoids the potential fluid, electrolyte and acid–base imbalances that can occur with intravenous fluid management.

Oral rehydration can be safely and effectively achieved using oral rehydration solutions (ORS) based on the recommendations of the World Health Organisation (Table 2). In children admitted to hospital with diarrhoea, reduced osmolarity ORS when compared to WHO standard ORS is associated with fewer unscheduled intravenous fluid infusions, lower stool volume post-randomization, and less vomiting. No additional risk of developing hyponatraemia when compared with WHO standard ORS was detected.44 This is now the standard composition of ORS used in developed countries. Table 3 summarises the composition of a commercial oral rehydration solution (ORS) and a number of commonly available fluids.45 Fruit juices and soft drinks require considerable dilution (one part to four parts water) if they are to be used as an alternative to ORS for children refusing to drink ORS.46

Table 2. Composition of a commercial ORS and commonly available drinks

	Solution	Na (mmol/l)	K (mmol/l)	Carbohydrate (mmol/l)	Osmolarity (mmol/kg)
	WHO recommendation	90	20	111	310
	Reduced osmolarity solution	75	20	75	245
	Hydrolyte	55	20	80	230
	Soft drink	2	0	700	750
	Apple juice	3	32	690	730
	Broth	250	8	0	500
	Sports drink	20	3	255	340
	Tea	0	0	0	5

Note: WHO, World Health Organisation. Source: Hahn, Kim, and Garner (2002) and Gastanaduy and Begue (1999).44, 45

Table 3. Recommended hourly rate for oral or nasogastric rehydration

	Weight on admission (kg)	Degree of dehydration
		Moderate (4–6%)		Severe (>7%)
		ml/h 0–6h	ml/h 7–24h	ml/h 0–6h	ml/h 7–24h
	3.0	25	20	45	20
	4.0	35	30	60	30
	5.0	45	35	75	35
	6.0	55	40	90	40
	7.0	60	45	100	45
	8.0	70	50	115	50
	9.0	80	55	130	55
	10	90	60	150	60
	12	105	65	175	65
	15	135	70	220	70
	20	175	85	290	85
	30	260	90	440	90

Source: Royal Children's Hospital (Melbourne) (2004).46

It is recommended that oral fluid replacement for children with moderate to severe dehydration is given over 6h to prevent fluid and electrolyte imbalances and Table 4 documents the volume of fluid administered by weight for children with moderate and severe dehydration.46 However, when electrolyte imbalance is not a concern, rapid rehydration regimens advocating 25ml/(kgh) to replace fluid losses over 3–4h are increasingly used for rehydration.

Table 4. Assessment tool to determine dehydration severity

	Signs of dehydration
	Decreased skin elasticity
	Capillary refill>2s
	General appearance (ill)
	Absence of tears
	Abnormal respirations
	Dry mucus membranes
	Sunken eyes
	Abnormal radial pulse
	Tachycardia (HR>150)
	Decreased urine output

Fewer than three signs=mild dehydration=three to six signs: moderate dehydration; greater than six signs=severe dehydration. Source: Gorelick, Shaw, and Murphy (1997).33

A small study compared children's tolerance of frozen ORS and ORS given as a fluid provides evidence for the increasing use of ORS icy poles in ED and at home.47 Children in the study were more likely to tolerate frozen ORS than conventional ORS. Children with only mild dehydration can usually be encouraged to drink sufficient amounts of fluid to maintain adequate hydration. Those children with dehydration who refuse to drink, are not tolerating oral fluids or are exhibiting signs of moderate to severe dehydration can in most circumstances be successfully rehydrated enterally via a nasogastric (NG) tube.

Nasogastric tube insertion can be safely and quickly achieved in the ED to provide for fluid replacement.[2], 3 Children requiring nasogastric rehydration are usually in the infant and toddler age group and are therefore not cooperative. This means restraining children during insertion and using strategies to prevent them pulling the tube out and means to alleviate their distress. The use of local anaesthetic agents does not appear frequently in practice recommendations. However, it is accepted that NG tube insertion is one of the most painful procedures performed in ED.48, 49 Fine bore tubes (eg: french guage 8) and continuous administration of fluid by feeding pump or manual drip set will be best tolerated.

Maintenance fluid management 

The purpose of maintenance fluids is to provide sufficient water and electrolyte to allow for normal urine output at a similar osmolality to extracellular fluid, where oral intake is insufficient to meet requirements. Children with acute illness often drink poorly and have higher losses (insensible, vomiting and diarrhea). A common water calculation for maintenance fluids for healthy children is provided in Table 5.[2], 3, 39 However, calculations were based on data collected from healthy children. As explained earlier, acute illness is often associated with an increase in ADH secretion resulting in water retention and varying levels of hyponatraemia. Therefore, maintenance fluid regimens based on the needs of healthy children are considered unsuitable for acutely unwell children. Less water should be prescribed to children suffering illness where water retention is possible or a positive water balance would be harmful, e.g., brain injury and severe respiratory disease. Hypotonic solutions are no longer recommended and maintenance volumes should be reduced by one third.[2], 3, 39 The accepted view is that isotonic solutions such as normal saline are more appropriate than hypotonic solutions and are now advocated for maintenance fluid therapy.[2], 3, 39, 50, 51

Table 5. Water calculation for healthy children per day and per hour

		Body weight
		<10kg	10–20kg	>20kg
	Daily	100ml/kg	1000ml+50ml/kg for each kg >10kg	1500ml+20ml/kg for each kg >20kg
	Hourly	4ml/(kgh)	40ml/h+2ml/kg for each kg >10kg	60ml+1ml/kg for each kg >20kg

Source: Henning (1995) and Hazell and Wilkins (2006).[2], 3