Background

Worldwide, it has been estimated that approximately one in five paediatric deaths can be attributed to dehydration related to gastroenteritis.1 2 The increased fluid losses from vomiting and diarrhoea can result in rapid fluid and electrolyte shifts leading to dysregulation of physiological mechanisms such as thirst. In extremis, this can lead to volume depletion and dangerous electrolyte imbalances.1 An understanding of the symptoms, signs and investigations used to assess and manage dehydration is vital for any clinician caring for acutely unwell children. This interpretation therefore focuses entirely on the management of dehydration in children with an acute gastrointestinal illness. Other, less common, causes of dehydration (eg, diabetic ketoacidosis, renal failure, cardiac failure, liver disease and postoperative care) are not covered.

Throughout this interpretation, dehydration will be described as mild (<5% of weight loss), moderate (5%–10% weight loss) or severe (>10% weight loss).3

What are the symptoms and signs of dehydration in a child?

A number of clinical features were identified from the literature as predictive of dehydration. These are outlined in table 1 and include decreased skin turgor, capillary refill time of >2 s, ill appearance, absent tears, abnormal respiration pattern, dry mucous membranes, sunken eyes, abnormal (weak or feeble) radial pulse, tachycardia and decreased urine output.4–7 The absence of these signs, among children with gastroenteritis, was highly predictive of the child not having dehydration with a negative predictive value (NPV) of over 90%.4–7

Unfortunately, the symptoms and signs listed above and in table 1 were not specific for dehydration; that is, the presence of one or more of the symptoms could not be used to accurately confirm dehydration as present. The most specific symptoms and signs of dehydration were reported as prolonged capillary refill time of >2 s (positive predictive value (PPV) 0.57–0.65) and decreased skin turgor (PPV 0.52–0.57). The remaining symptoms and signs were non-specific for dehydration with a PPV of less than 0.5.4–7

Based on these findings, it is reasonable to assume that, in a child with acute gastroenteritis, if all of the symptoms and signs listed are absent, then dehydration is unlikely. The presence of any of the features listed should alert the clinician to the possibility of dehydration while also being mindful that many of the features are non-specific and may indicate other illnesses.

One approach to rising awareness and to standardise care in this area has been the development of clinical decision tools. One such tool validated for use in children (aged 1 month–3 years) is the Clinical Dehydration Scale (CDS),8 summarised in table 2. Another useful resource is the National Institute for Health and Care Excellence (NICE) clinical guideline 84: ‘Diarrhoea and vomiting caused by gastroenteritis in under 5 s: diagnosis and management’.9

Which symptoms and signs are best for predicting the severity of dehydration?

Predicting the severity of dehydration is challenging and requires the clinician to interpret and combine a mixture of symptoms and signs. When taking the history, it is important to ask about the number of episodes of diarrhoea and/or vomiting, as well as asking about intake, duration of illness and estimated urine output. Of the reported symptoms, it is the number of episodes of diarrhoea and/or vomiting that correlates best with severity.10

During the examination, it is important to make a global assessment of the child’s health while also specifically examining for skin turgor, capillary refill time, dryness of mucous membranes, appearance of eyes (sunken or not) and the presence of tears. The individual predictive value of these signs has been discussed already (table 1), and the presence of any one feature should alert the clinician to possibility of dehydration.4–7 10–12 The greater the number of features, the greater the likely degree of dehydration.6

The CDS provides a pragmatic approach to assessing severity (table 2) and can be used to aide decision making.4 8 Similarly, NICE CG84 provides guidance on assessing severity with features of dehydration listed, including five ‘red flag’ signs of ‘reduced responsiveness, reduced skin turgor, tachycardia, tachypnoea and sunken eyes’.9

Can laboratory investigations be used to assess the severity of dehydration?

NICE does not recommend routine blood testing in the assessment of dehydration in children. Blood testing is only recommended for children who require intravenous therapy, have signs of shock or where hyponatraemia/hypernatraemia is suspected.9 When testing is required, NICE only recommends measuring blood sugar, electrolytes, urea, creatinine and a blood gas. This approach is supported by the available evidence summarised in table 3.7 10 12–18

Measuring serum urea and creatinine levels is helpful when assessing severe dehydration only with very high serum urea values (greater than 16.7 mmol/L) and creatinine values (greater than 80 μmol/L) reliably predicting severe dehydration.10 13 Unfortunately, urea and creatinine levels are unhelpful in mild and moderate dehydrations.

Of the tests advised by NICE,9 the blood gas (specifically bicarbonate levels) correlates the best with severity, as shown in table 4.12 13 16 17 As a general trend, the lower the venous bicarbonate levels, the greater the severity of dehydration.10

What about other tests?

There are a number of other tests that have been suggested as helpful in the assessment of dehydration in children. These include urinary testing (sodium and specific gravity) and blood tests (blood urea nitrogen (‘BUN’), BUN to creatinine ratio and uric acid).12–14 17 18

Urinary testing for specific gravity is of almost no value when assessing for dehydration and should not be routinely used.18 Urinary sodium levels may be helpful with a urine sodium of less than 90 mmol/L being shown to be highly sensitive and specific for dehydration.17 A urinary sodium may therefore be useful in children where the traditional clinical assessment of hydration may be difficult, for example, dehydration in conditions such as nephrotic syndrome.

Blood testing for BUN and BUN/creatinine levels may also be helpful in some. An elevated BUN greater than 45 mg/dL (16 mmol/L) is indicative of at least moderate dehydration,12 whereas a BUN to creatinine ratio of less than 20 is indicative of an absence of dehydration.14

Experimental testing

The digitally measured capillary refill time (DCRT) is an innovative approach for predicting dehydration severity. The test requires use of specific software to analyse ‘frame-by-frame’ the video of one of the child’s fingertips both before and after a light pressure is applied for 5 s. Studies have demonstrated that a DCRT of >0.4 s was 100% sensitive (95% CI 75% to 100%) and had a specificity of 91% (95% CI 82% to 97%) for predicting the presence of at least moderate dehydration.19

Ultrasound (US) scanning has been proposed as a quick method to determine the severity of dehydration in children. Two different US measures (aorta to inferior vena cava (IVC) ratio and IVC inspiratory collapse) have been studied as methods of estimating dehydration in children. The aorta to IVC ratio demonstrated a sensitivity of 93% and specificity of 59% compared with 93% and 35% for IVC inspiratory collapse.20 The use of US remains experimental but could become of greater clinical relevance as point-of-care US becomes more widely available.20