Gastroenteritis is caused by a wide range of pathogens including viruses, bacteria and parasites (as shown in Table 7.12.1). In developed countries, the majority of episodes are due to viruses, with rotavirus being by far the most common pathogen. The most common bacterial causes are Salmonella and Campylobacter. Shigella, Yersinia and Escherichia coli are less common, while Vibrio cholerae is seen in developing countries. Parasites such as Giardia and Cryptosporidium are sometimes the infective agent.

Viruses Bacteria Parasites Rotavirus Salmonella Cryptosporidium Norwalk virus Campylobacter jejuni Giardia lamblia Enteric adenovirus Shigella Entamoeba histolytica Escherichia coli Yersinia enterocolitica Vibrio cholerae

In general a bacterium is more likely to be the causative agent if:

There is blood or mucus in the stool.

There is significant abdominal pain.

There is a high-grade fever.

These infective agents cause gastroenteritis by one, or more mechanisms. Some directly invade the bowel wall, e.g. Salmonella or Shigella, some produce toxins prior to ingestion, e.g. Staphylococcus aureus, and some multiply and produce toxins within the gastrointestinal tract, e.g. Shigella.

In an uncomplicated acute bout of gastroenteritis, determining the causative pathogen is usually unnecessary, as this usually does not alter the management.

History

Most children with acute gastroenteritis present with a history of diarrhoea and vomiting. Abdominal pain and fever are sometimes accompanying symptoms. It is important to ask for specific details about these symptoms, to increase the certainty of the diagnosis of gastroenteritis and also to help assess the risk of dehydration in the child.

Diarrhoea refers to loose or liquid stools. The frequency, volume (small to voluminous) and consistency (semisolid to watery) and the presence of blood or mucus are all important. Similarly, the frequency and nature (bilious or non-bilious) of the vomiting are also important. The words ‘my child is vomiting everything he tries to drink’ can be used by parents to describe a child who has had either two or 20 vomits in the last 24 hours. It is important, therefore, to be specific in questioning.

The abdominal pain that can be associated with gastroenteritis is crampy in nature and commonly more pronounced in bacterial gastroenteritis. It is not accompanied by focal or significant findings on abdominal examination. If fever is a symptom, the height of the fever can be helpful. Although children with rotaviral gastroenteritis are often febrile, it is not usually high-grade. Such fevers are more likely in bacterial gastroenteritis or in other diagnoses altogether.

Specific details about the amount of fluid tolerated by the child are vital in assessing the risk of dehydration. Parents are usually able to estimate whether, over a 24-hour period, the child has taken one-quarter, one-half or three-quarters of normal intake. Intake that is consistently less than half of normal is of concern.

Specific questions about urine output (i.e. heaviness and frequency of wet nappies) are also important. Fewer than four wet nappies in 24 hours is of concern when assessing hydration status.

Questions about the level of alertness and activity of the child are also important. Lethargy may simply be an indication of significant dehydration in a child. However, more severe lethargy and drowsiness (particularly if out of proportion to that expected for the dehydration) may indicate another more serious diagnosis as the cause of the child’s symptoms, such as enteroviral meningitis.

Examination

The aim of the clinical examination is to exclude signs of an alternative cause of the symptoms, other than gastroenteritis (see Chapter 7.8) and to assess the degree of dehydration.

Occasionally pathogens causing gastroenteritis can cause extraintestinal disease. This can occur particularly with Shigella and Salmonella. Shigella can cause central nervous system irritation, which can manifest as encephalopathy or seizures. This may occur prior to the onset of diarrhoea. Salmonella can cause a bacteraemia, which can lead to focal infections including osteomyelitis and meningitis. The infant under 6 months of age is particularly at risk.

Assessment of dehydration

The accurate assessment of dehydration is difficult. Studies have shown that medical personnel tend to overestimate the degree of dehydration.² The gold standard in assessment of dehydration is a loss of weight compared with a recent pre-illness weight. For example, a 1-year-old weighing 10 kg a week ago who presents with gastroenteritis for 3 days and now weighs 9.5 kg, is approximately 5% dehydrated. However, a recent weight is rarely available in the ED. Therefore, tables such as Table 7.12.2 use a combination of clinical symptoms and signs to estimate the degree of dehydration.

It is important to remember that some of the signs and symptoms of dehydration can be affected by other factors. Mouth breathers have dry mucous membranes. Watery diarrhoea can make it difficult to assess if the child has a wet nappy from urine output. Crying can cause a sunken fontanelle to appear full. Tachycardia may be caused by a crying, anxious or febrile child; therefore it is the context and trend of the pulse that is important.

Various studies have attempted to validate combinations of these signs and symptoms with varying degrees of standardisation and scientific validity.²^–⁵ Difficulties arise as some of the signs are subjective and gold standards differ.

Gorelick et al ⁵ performed a good study in Philadelphia in 1997, which looked at the 10 clinical signs shown in Table 7.12.3. Using a gold standard of serial weight gain, they found that <3 signs correlated with <5% dehydration, 3–6 signs correlated with 5–9% dehydration and >7 signs correlated with >10% dehydration.

1. Decreased skin elasticity

2. Capillary refill >2 seconds

3. General appearance (ill-appearing, irritable, apathetic)

4. Absent tears

5. Abnormal respirations

6. Dry mucous membranes

7. Sunken eyes

8. Abnormal radial pulse

9. Tachycardia (HR >150)

10. Decreased urine output

Source: Gorelick et al⁵

Laboratory investigations in the assessment of dehydration

There are few data to support the usefulness of laboratory tests in the assessment of dehydration due to gastroenteritis. Dehydration is thought to typically cause a metabolic acidosis. It is true that vomiting can cause a metabolic acidosis by several mechanisms including volume depletion, lactic acidosis and starvation ketosis. However, isolated vomiting can also cause a metabolic alkalosis through loss of gastric acid. In addition, isolated diarrhoea can cause a metabolic acidosis through loss of bicarbonate in the stool, but the child who can increase oral intake to keep pace with the diarrhoeal losses may not actually be dehydrated. Despite these confounding factors, it has been shown that serum bicarbonate is significantly lower in children with moderate or severe dehydration (mean 14.5 mEq L^–1 and 10.3 mEq L^–1) than in children with mild dehydration (mean 18.9 mEq L^–1).³ Raised serum urea levels have also been thought to reflect dehydration, but with even less evidence.

From a practical point of view, it is not vital to put a specific percentage on the degree of dehydration, particularly when the accuracy of such a specific percentage is questionable. What is important is to allocate the child into a broad category of dehydration, for example, mild, moderate or severe, which determines what treatment should be commenced, and then to reassess the child and the degree of dehydration within a specific timeframe.

Investigations

In uncomplicated acute gastroenteritis, it is usually not necessary to perform any investigations, as they will not alter management. Indications to consider performing investigations include:

• uncertainty about diagnosis;

• severe dehydration;

• the child’s clinical course not progressing as anticipated;

• commencement of intravenous rehydration.

Stool cultures do not change management in simple gastroenteritis; therefore they are not usually warranted. If bacterial gastroenteritis is suspected a stool sample may be performed, although this will not change management in the majority of cases. A stool sample may be taken for public-health reasons, or to reassure the doctor that there is not another cause for the blood and mucus in the stool.

Urine culture is usually unnecessary when the diagnosis of gastroenteritis is clear. However, if there is little diarrhoea, particularly if the child is under 6 months of age, it may be necessary to exclude a urinary tract infection. It is important that a clean specimen is collected to avoid contamination by stool contents.

Electrolytes, urea and creatinine are usually unnecessary in children with gastroenteritis who are mild to moderately dehydrated and who are to receive oral fluids, as they will not usually alter management. If, however, the child does not improve as expected, or deteriorates despite rehydration, or the child was initially severely dehydrated, these tests along with a blood glucose level and perhaps other investigations, will be required.

A full blood count and blood culture are unnecessary in acute gastroenteritis. However, if the diagnosis is uncertain it may be reasonable to perform these tests. In the younger infant, particularly the under 3 month age group, if bacterial gastroenteritis is likely, the possibility of systemic Salmonella infection should be considered and blood culture and intravenous antibiotics considered.

If significant abdominal pain is present and surgical pathology such as intussusception or bowel obstruction seems more likely than simple gastroenteritis, other investigations such as abdominal X-rays, ultrasound or air enemas may be required.

Treatment

Mildly dehydrated

Children in developed countries presenting to EDs with gastroenteritis are often only mildly dehydrated or not dehydrated at all. Most can be managed as outpatients with oral fluids alone. Educating the parents in how to give fluids, and observing them giving a trial of fluids to the child, are important to empower the parents to provide ongoing oral rehydration at home.

The essential things to ensure are:

• the parent knows the appropriate oral fluids to give;

• the parent knows how, and is able, to give the oral fluids correctly to optimise success;

• the parent will have the child reviewed by the local doctor within 48 hours;

• the parent knows what worrying signs and symptoms to look for that would require representation.

If the above criteria cannot be satisfied, the child may need to be admitted. This is particularly so in the younger infant or those with profuse gastrointestinal losses, as they have a higher risk of becoming dehydrated. See ‘moderately dehydrated’ for further treatment options in those patients.

Appropriate oral fluids

Oral rehydration solutions (ORSs) are the appropriate fluids to rehydrate children with gastroenteritis. These are specifically designed fluids that contain the correct amounts of sodium, glucose and other electrolytes and are of the appropriate osmolality to maximise water absorption from the gut. They use the principle of glucose-facilitated sodium transport, whereby glucose enhances sodium and secondarily water transport across the mucosa of the upper intestine. The sodium and glucose concentrations and the osmolality are of vital importance.

The WHO ORS has a sodium concentration of 90 mmol L^–1. In developed countries with non-cholera diarrhoea, it is generally thought that 90 mmol L^–1 is a little high, as non-cholera gastroenteritis does not result in the same sodium losses that are seen in cholera. Many different ORSs with varying sodium concentrations have been developed. It has been shown ⁶ that water absorption across the lumen of the human intestine is maximal using solutions with a sodium concentration of 60 mmol L^–1 and this is the concentration recommended by the European Society of Paediatric Gastroenterology and Nutrition.⁷ Studies have also shown that hypo-osmolar solutions are most effective at promoting water absorption.⁸^–¹⁰

Studies have also examined rice-based ORSs and their effect on stool output and duration of diarrhoea when compared with glucose-based ORSs. Rice-based ORSs appear to have benefits in cholera diarrhoea but not in non-cholera diarrhoea.¹¹

The composition of various ORSs and other fluids is shown in Tables 7.12.4 and 7.12.5. Fruit juices and soft drinks are inappropriate because of the minimal sodium content and the excessive glucose content and hence excessive osmolality. Diluting these solutions will not address the grossly inadequate sodium content, nor will it result in an optimal glucose concentration or osmolality. Sports drinks are also inappropriate, with too low sodium levels and too high glucose levels and osmolalities.

Method of giving oral fluids

The important message is to give small amounts of fluid frequently, for example 0.5 mL kg^–1 every 5 minutes. The fluid can be measured in a syringe and given to the child either by syringe, teaspoon or cup. The child is far more likely to tolerate these small amounts of fluid than a whole bottle at once. If the child tolerates this fluid the parent can gradually increase the volume and decrease the frequency of the fluid offered. Success can be optimised in the ED setting by giving the parents a documentation chart to fill in, which shows the fluid given and any vomits or diarrhoea or urine passed.

It is important to educate the parents that seeing a doctor will not cure the vomiting and diarrhoea. Small, frequent amounts of fluid will hopefully minimise the vomiting, but will not reduce the diarrhoea. The aim is for the input to exceed the output enough to rehydrate and then maintain hydration. Parents will know if they are succeeding by observing their child for the following symptoms and signs.

Refeeding

• Breast-fed infants – breast-feeding should continue throughout the episode of gastroenteritis, including the rehydration phase.

• Formula-fed infants – full-strength normal formula should be started as soon as the infant is rehydrated. This can be supplemented with an ORS for ongoing losses, for example offer 10 mL kg^–1 after each diarrhoea. However, children with no dehydration and mild gastroenteritis are the least likely to take an ORS because of the salty taste. In this instance increased amounts of normal fluids in conjunction with age-appropriate feeding may be sufficient.

• Early refeeding – studies have shown that early refeeding does not worsen or prolong the duration of diarrhoea, nor increase vomiting or lactose intolerance, and leads to a significantly higher weight gain after rehydration.¹² Early reintroduction of age-appropriate foods is now recommended.

Moderately dehydrated

Some children who present to EDs with gastroenteritis are moderately dehydrated. These children can be rehydrated in several ways. Some are successfully rehydrated with oral fluids alone, as previously described. Some fail this and require additional intervention. Increasing numbers of hospitals in developed countries are using oral rehydration therapy (ORT) via continuous nasogastric infusion.^13,¹⁴ This is where an ORS is infused continuously down a nasogastric tube with a pump such as a Kangaroo pump. Nasogastric infusions should not be used when the child has an ileus or is comatose. Oral rehydration therapy has been the method of choice in developing countries since the 1970s. However, it has taken longer to become accepted in developed countries. This is despite numerous studies that show that it is as effective as intravenous rehydration but less expensive¹³^–²¹ and reduces lengths of hospital stay.¹³

Different regimens are used for continuous nasogastric rehydration. The European Society of Paediatric Gastroenterology and Nutrition recommends calculating the fluid deficit and replacing that over 4 hours.²² The American Academy of Pediatrics recommends that mildly dehydrated children receive 50 mL kg^–1 over 4 hours and moderately dehydrated children receive 100 mL kg^–1 over 4 hours.²³ Other regimes recommend a fixed volume, for example 40 mL kg^–1 over 4 hours for all mild to moderately dehydrated children followed by a reassessment and retrial of oral fluids.²⁴ This takes into account the tendency for medical officers to overestimate the degree of dehydration² and the desire to avoid subsequent over-hydration. The important thing to remember is that, whatever regimen is used to rehydrate the child, fluid status should be regularly reassessed.

Intravenous rehydration

Intravenous rehydration should be used to rehydrate children with gastroenteritis who fail nasogastric therapy or deteriorate.

The volume (mL) of replacement fluids is calculated using the formula: weight (kg) × dehydration (%) × 10. Maintenance fluids are then calculated, with one method being to give 100 mL kg^–1 for the first 10 kg, 50 mL kg^–1 for the second 10 kg and 20 mL kg^–1 for each subsequent kilogram. The volumes for rehydration and maintenance are then added together and divided by 24 to calculate the hourly rate.

The recommended fluid used to rehydrate children with gastroenteritis has changed over recent years. It is now recommended that for children, excluding neonates and infants <3 months of age, 0.9% (150 mmol L^–1) saline + 2.5% (or 5%) glucose should be used.^25,²⁶ Studies have shown that children with gastroenteritis have inappropriately high levels of antidiuretic hormone and increased incidence and risk of hyponatraemia²⁷ and that low sodium solutions cause hyponatraemia while solutions with a sodium content of 130–154 mmol L^–1 are protective.²⁸ As in nasogastric rehydration, it is important to regularly reassess the child’s fluid status. If the child remains on intravenous fluids for >24 hours, it is important to recheck the electrolytes.

If the child is significantly hypernatraemic the rate of infusion needs to be reduced to rehydrate the child over 48–72 hours. The aim is to avoid a rapid fall in the serum sodium level as this can precipitate cerebral oedema. It is still reasonable to start with 0.9% (150 mmol L^–1) saline + 2.5% (or 5%) glucose to prevent a rapid initial fall in serum sodium, but this may need to be changed to 0.45% (75 mmol L^–1) saline + 2.5% (or 5%) glucose, depending on progress of the electrolytes, which need to be rechecked frequently. If the child is hyponatraemic, 0.9% (150 mmol L^–1) saline + 2.5% (or 5%) glucose should be used. Senior paediatric advice should be sought if the serum sodium level is significantly abnormal.

Rapid intravenous rehydration: Interest in rapid intravenous rehydration for children with gastroenteritis has emerged in recent years in developed countries. It has long been used in developing countries.^29,³⁰ There have been a number of studies in developed countries over the last several years that have looked at this issue^24,^28,³¹^–³³ and shown that this seems to be a safe and effective way to rehydrate children with gastroenteritis who require intravenous therapy. An example of this type of regime is giving 0.9% (150 mmol L^–1) saline + 2.5% glucose at 10 mL kg^–1 hr^–1 for 4 hours.³⁴ It is important that lower sodium-containing fluids are not used.

Severely dehydrated

Although this is a relatively rare occurrence in developed countries, it is vital to diagnose and institute immediate treatment, as this is a medical emergency. The child is usually shocked and needs immediate intravenous access and resuscitation with a 20 mL kg^–1 bolus of normal saline or Hartmann’s solution. If the child remains shocked the bolus should be repeated. Senior advice should be sought about any child presenting in this condition. During insertion of the cannula, blood should be taken for electrolytes, urea, creatinine, glucose, full blood count and venous blood gas. Blood cultures can also be taken at the same time if indicated.

Glucose-containing fluids such as 0.45% (75 mmol L^–1) saline + 2.5% glucose or 0.225% (37.5 mmol L^–1) saline + 3.75% glucose should never be used for resuscitation boluses. Although these solutions are technically isotonic, the glucose is rapidly metabolised, so effectively it is as if a hypotonic solution has been given, with the resultant risk of rapid fluid shifts and cerebral oedema. It is important to differentiate these resuscitation boluses from rapid rehydration therapy. The resuscitation boluses are given over several minutes and do not contain glucose. The rapid rehydration fluids are given over several hours and do contain glucose.

After the initial resuscitation, replacement plus maintenance fluids are given as per above. It is important to reassess the child’s hydration status regularly and look for any indication that another diagnosis is the cause of the child’s condition. Electrolytes, urea and creatinine need to be regularly repeated while the child remains significantly unwell.

Other treatments

Anti-diarrhoeal and anti-emetic medications

Anti-diarrhoeal medications are not indicated in children with gastroenteritis. Most anti-emetic medications are also not indicated. There is little evidence of efficacy and a high incidence of side effects such as dystonic reactions and sedation in infants and young children. There have been some recent studies ³⁵^–³⁷ that show that ondansetron may have some clinical benefit in this setting by reducing vomiting, but they do not decrease the length of illness and may prolong the diarrhoea.³⁷ Experienced clinicians who wish to use this medication in this setting should generally limit its use to a single dose.

Disposition

In the past, many children were admitted to hospital with gastroenteritis when perhaps they did not need to be.³⁸ With the emergence of ‘Short Stay Wards’ or ‘Emergency Observation Units’, where patients can be admitted to a special area within the ED for a finite time of less than 24 hours, hospital inpatient admission rates for children with gastroenteritis have fallen.²⁴ In these units children can be rehydrated (orally or via rapid nasogastric or intravenous infusion) over a period of a few hours and then sent home after appropriate education and advice. This is provided that the medical officer is confident about the diagnosis of gastroenteritis and that the criteria set out in the ‘mildly dehydrated’ section are fulfilled. Care needs to be taken especially with young infants, as they become dehydrated more rapidly than older children and are more likely to have another diagnosis.

Paediatric staff in non-children’s hospitals can potentially be utilised to help manage these patients in conjunction with ED staff. An admission to, and discharge from, these ‘Short Stay Wards’ causes much less disruption to the child and family and allows hospital resources to be used for other patients requiring inpatient admissions. Some children, however, still require an inpatient admission for successful management.