Under normal physiologic conditions, water comprises 70% of lean body mass in children and adults and 75% in infants. Two-thirds of the fluid is intracellular and one-third is extracellular. Of the extracellular fluid, 75% is interstitial and 25% is intravascular. Fluid that is lost from the body often has an electrolyte composition similar to plasma. Most of the fluid deficit during the early stages of dehydration is from the extracellular space, but with time, the fluid losses equilibrate and fluid leaves the intracellular space. During the recovery phase, fluid administered to the patient is in the extracellular space and will need time to re-equilibrate with the intracellular space.

Gastroenteritis can be caused by viral, bacterial, or parasitic pathogens. The predominant cause in developed countries is viral. Rotavirus is known to cause more serious illness than other viral causes. Worldwide, all children have had at least one episode of rotavirus gastroenteritis by 5 years of age.

Gastroenteritis typically starts with the acute onset of vomiting, which may be persistent and severe. Clinicians should consider other etiologies in a child presenting with vomiting alone, but the subsequent development of diarrhea may clarify a diagnosis of viral gastroenteritis. The overall course of illness typically lasts about 5 to 7 days.

Children with dehydration from gastroenteritis may develop electrolyte abnormalities. Approximately one-third of patients who are moderately dehydrated have hypoglycemia. These children appear listless. Less commonly, patients have elevated serum sodium. This hypernatremia represents a free water deficit. The sodium acts osmotically to draw water from the cellular compartment extracellularly. This results in a state of intracellular dehydration. The acute manifestations of hypernatremia include early neurologic irritability (e.g., seizures). The degree of dehydration may be more difficult to assess in these children because most of the clinical findings of dehydration are caused by decreased extracellular fluid; one clue may be the presence of “doughy” skin. Conversely, hyponatremia results in intracellular swelling. These patients also may present with neurologic symptomatology, primarily seizures, lethargy, and coma.

Evaluation of dehydration focuses on a thorough history and physical examination. Historical points to consider include possible exposures to other children who have been sick, timing of the symptoms, number of episodes of vomiting and diarrhea, or presence of blood or bile. Exposure to well water, camping, or recent travel may raise the suspicion for an etiology other than viral gastroenteritis. High fever with bloody diarrhea would be concerning for bacterial dysentery.

A thorough physical examination can aid in confirming the diagnosis. The examiner should assess the overall appearance of the child for the presence of lethargy or listlessness that may be associated with dehydration. Other common examination features include an assessment of the urine output, the presence and quality of tear production, the quality of the mucous membranes (moist, tacky, or dry), the capillary refill (measured at the distal fingertip, considered prolonged if >2 seconds in an warm ambient environment), assessment of the skin turgor, assessment of the heart rate and distal pulses, as well as assessment of the fontanelle in the infant. Research on assessing the degree of dehydration has been helpful in creating a framework within which to evaluate a dehydrated patient. A recent meta-analysis presents data that are helpful (Table 25-1). If a finding is present, it may increase the likelihood of dehydration. For example, the likelihood ratio for the presence of prolonged capillary refill is 4.1 with a 95% confidence interval that ranges from 1.7 to 9.8. We can interpret this to mean that if prolonged capillary refill is present, then the odds of the child having at least 5% dehydration are 4.1 times higher than if prolonged capillary refill was not present. The examination findings in the meta-analysis that significantly increase the likelihood of dehydration include prolonged capillary refill, abnormal skin turgor, and abnormal respiratory pattern. Conversely, some findings if they are absent decrease the likelihood of the patient being 5% dehydrated. For example, the likelihood ratio for the absence of poor overall appearance is 0.46 with a 95% confidence interval that ranges from 0.34 to 0.61. We can interpret this to mean that the odds are less than half as likely for the patient to be at least 5% dehydrated if there is a normal general appearance. The findings that would decrease the likelihood of significant dehydration if they were absent include dry mucous membranes, sunken eyes, and poor overall appearance.