Many children experience brief episodes of vomiting and diarrhoea due to mild gastroenteritis and are managed by their parents at home. Parents may not approach a healthcare professional at all. However, many do seek advice either ‘remotely’ (for example, NHS Direct) or through a face-to-face consultations. When children present to a healthcare professional, it is important to establish whether or not they truly have diarrhoea or vomiting. This chapter reviews the relevant published evidence and provides recommendations regarding the clinical and laboratory diagnosis of gastroenteritis. These recommendations are intended to take account of the particular setting in which the child presents, for example in the community, the general practitioner’s surgery, or the hospital emergency department.
3.1. Clinical diagnosis
Introduction
The sudden onset of diarrhoea with or without vomiting in a previously well child is usually due to gastroenteritis. The definition of diarrhoea may seem to be self-evident but, even in well infants and children, stool frequency and consistency vary considerably. For example, breastfed infants may have more frequent and softer stools than bottle-fed infants. Even in older children, confusion may occur – those with overflow faecal incontinence due to constipation are often mistakenly reported to have diarrhoea.
Vomiting may occur before the onset of diarrhoea. However, vomiting in isolation may be due to a wide range of other potentially serious conditions. In infants, vomiting must be distinguished from the normal phenomenon of regurgitation.
Although most children with acute-onset diarrhoea have gastroenteritis, occasionally it may occur in association with other disorders, such as non-gastrointestinal infections (for example, pneumonia), and surgical conditions (for example, acute appendicitis). In 2007, 7600 children presented to a paediatric emergency unit in England with suspected acute gastroenteritis.36 Of these, 60% were discharged for home treatment. A total of 3022 were admitted to an observation ward, and only 106 of these were subsequently transferred to a medical or surgical ward. Those children had a range of diagnoses, including gastroenteritis, non-specific abdominal pain, appendicitis and constipation. Other diagnoses were rare but included such diverse conditions as non-infective colitis, malabsorption, intestinal obstruction, inguinal hernia and pyloric stenosis.
Even when a presumptive diagnosis of gastroenteritis has been made at the outset, it is important to reconsider the diagnosis if the subsequent course of the illness is inconsistent with the condition.
This chapter gives recommendations regarding practical definitions for diarrhoea and vomiting, identifies key clinical pointers to conditions other than gastroenteritis and provides information on the natural history of the disease.
Clinical question
What definitions of diarrhoea and vomiting have been used previously?
Research studies on the incidence of gastroenteritis have employed various arbitrary definitions of diarrhoea based on the stool frequency and/or consistency. A change in these variables in the individual child has also been considered to be an important consideration. Examples of definitions that have been employed are listed below.
Diarrhoea is defined as a change in bowel habit for the individual child resulting in substantially more frequent and/or looser stools.17
Diarrhoea in children is the passage of unusually loose or watery stools, usually at least three times in 24 hours. It is acute if it has persisted for less than 10–14 days.20
Diarrhoea is an alteration in normal bowel movement characterised by an increase in the water content, volume or frequency of stools. A decrease in consistency (i.e. soft or liquid) and an increase in frequency of bowel movements to three or more stools per day have often been used as a definition for epidemiological investigations.37
In diarrhoea, stools contain more water than normal – they are also called loose or watery stools. They may also contain blood, in which case the diarrhoea is called dysentery. Acute diarrhoea starts suddenly and may continue for several days. It is caused by infection of the bowel.38
There is less variation regarding definitions of vomiting. It has been defined as the ejection of gastric contents up to and out of the mouth, brought about by a forceful contraction of the abdominal muscles and diaphragm.39
For the purpose of this guideline, the GDG thus decided to use the following definitions:
‘Diarrhoea refers to the passage of liquid or watery stools. In most cases there is an associated increase in stool frequency and volume.’
‘Vomiting is the forceful ejection of the stomach contents up to and out of the mouth.’
The GDG considered that, in children with gastroenteritis, the occurrence of diarrhoea is readily apparent from the sudden change in stool consistency to loose or watery stools. It is often associated with an increase in stool frequency, although this may not yet be evident at the time of presentation. Vomiting is partly a voluntary action and partly an involuntary reflex, and is characterised by the forceful ejection of gastric contents up to and from the mouth. Regurgitation is common in infants but is a recurring phenomenon and so can usually be distinguished from sudden onset vomiting due to gastroenteritis.
Recommendation on clinical diagnosis
Suspect gastroenteritis if there is a sudden change in stool consistency to loose or watery stools, and/or a sudden onset of vomiting.
Clinical question
What is the usual duration of diarrhoea and vomiting in children with gastroenteritis?
It was important to establish the usual duration of vomiting and diarrhoea in children with gastroenteritis. Persistence of diarrhoea for a longer period of time than expected might be an important pointer to an alternative diagnosis, for example a non-infectious gastrointestinal disorder such as coeliac disease. This information would also be important, however, when advising parents about the likely outcome of the illness, and how soon recovery might reasonably be expected.
Evidence overview
Many observational studies identified during the literature search did not provide data on the duration of diarrhoea and vomiting in children or they failed to differentiate between whether the reported duration was prior to study entry or during the study period. In other studies, some children were given treatment but the authors failed to provide information separately for the group which did not receive treatment. Hence information to answer this question was obtained predominantly from control subjects in RCTs. A total of 11 studies were included, six of which provided information on the duration of diarrhoea and five on vomiting. Two of these 11 studies were conducted in the UK.
Of the studies relevant to diarrhoea, five were conducted in a hospital setting and one in a primary care setting. Data on the mean duration of diarrhoea were obtained from the control subjects in four RCTs40–43 conducted in Poland, Israel, Colombia and South Africa. The sample size in these RCTs ranged from 28 to 120. Further information was obtained from a cross-sectional study from Kuwait,44 in which almost 50% of the children (n = 595) had bacterial infections. Unfortunately, the data in the Kuwait study was presented without standard deviation values. The primary-care-based study45 was a small RCT from Denmark (n = 19) that recruited participants with acute diarrhoea from a day-care centre. The results are presented in Table 3.1.
Table 3.1
Duration of diarrhoea in children.
Data on the mean duration of vomiting prior to hospital admission was given in three RCTs: 2.4 days (range 1–6),46 1.6 days (SD 1.2)47 and 2.5 days (SD 1.7).48 These RCTs were conducted in Australia (n = 59), Saudi Arabia (n = 150) and Finland (n = 65), respectively. Two small RCTs conducted in the UK provided data on vomiting during the hospital stay. In one RCT,49 the numbers of children with vomiting on days 1, 2 and 3 were 7/16, 3/16 and 2/16, respectively. In the other RCT,50 it was reported that, of 14 children, none experienced any vomiting over 4 days.
Evidence summary
Evidence for this question was marked by great variation among the studies regarding the settings, sample sizes and study period. In the five hospital-based studies, the mean duration of diarrhoea from the time of presentation ranged from 2.3 to 7.4 days. In the small community-based study, more than 60% of the children presenting with diarrhoea to a clinic recovered from the condition by day 5.
In three RCTs, the mean duration of vomiting prior to hospital admission varied from 1.6 to 2.5 days. Data from two small hospital-based RCTs indicated that vomiting had ceased in most patients within 2 or 3 days of admission.
GDG translation from evidence to recommendation
The GDG noted the lack of satisfactory data from the UK regarding the duration of diarrhoea and vomiting in children with gastroenteritis. The available data were obtained from clinical trials in which measurement of symptom duration was not the primary aim of the study. Many of the studies measured duration of diarrhoea only during the period of hospital admission, and so underestimation may have occurred. The only study based in a primary care setting was very small. Those members of the GDG with experience of managing gastroenteritis in the community believed that diarrhoea usually resolves within 5–7 days but occasionally may persist till 14 days. That view is consistent with the data presented in Table 3.1.
Recommendation on duration of diarrhoea and vomiting
Be aware that in children with gastroenteritis:
diarrhoea usually lasts for 5–7 days, and in most it stops within 2 weeks
vomiting usually lasts for 1–2 days, and in most it stops within 3 days.
Clinical question
What factors influence the natural history of gastroenteritis?
Five relevant studies were identified to address this question. Four cross-sectional studies with EL = 3 reported on the clinical features associated with various infective pathogens51–44 and three of them were conducted in Europe. The fourth study is a case–control study [EL = 2+] from India reporting on the risk factors for persistent diarrhoea.54
Evidence overview
Clinical features associated with infective pathogens
In a prospective cross-sectional study from Sweden,51 the clinical features associated with bacterial and viral gastroenteritis were compared in children (n = 393) presenting to the hospital. The study focused on duration and frequency of symptoms, fever and abdominal pain, and duration of hospital stay. Infection with rotavirus was characterised by sudden onset of vomiting, a high incidence of fever and dehydration and a mean duration of diarrhoea of 5.9 days. Enteric adenovirus was associated with longer lasting diarrhoea, with a mean of 10.8 days. Bacterial infections were associated with abdominal pain, bloody diarrhoea, prolonged diarrhoea (mean 14.1 days), leucocytosis and a raised erythrocyte sedimentation rate (ESR). [EL = 3]
In the second hospital-based prospective cross-sectional study from Italy,52 the clinical features at presentation were compared in 215 children with viral and bacterial gastroenteritis. Enteric pathogens were detected in 127/215 (59%) cases with viral infections accounting for almost 80% of these (single or double viral or co-infection with bacterial pathogens) while mono-bacterial infections were detected in the rest. No sample was positive for parasites. Vomiting and dehydration were more frequent in children with viral gastroenteritis (P < 0.01) compared with those without viral infection. The severity of dehydration (assessed by a 14-point severity score) was significantly higher in children infected with either astrovirus or rotavirus group A. Prolonged hospitalisation was also more likely to occur with rotavirus infection. [EL = 3]
A prospective UK study53 included 1148 children younger than 16 years admitted to a subregional infectious disease hospital with a diagnosis of gastroenteritis over a 1 year period. Of the admitted children, 55% (635/1148) were younger than 1 year while 5% were over 5 years of age. Admissions were predominantly from socially disadvantaged families (62% from social classes IV and V). At the time of admission, 8.8% (101/1148) of children were clinically dehydrated, with 1% assessed to have greater than 5% dehydration. Seventy-nine percent of children had a history of vomiting before admission and it was more common with rotavirus infection than bacterial pathogens (92% versus 54%; P < 0.001). Diarrhoea of bacterial, protozoal or mixed aetiology had a higher incidence of stool containing blood and/or mucus compared with rotavirus, and abdominal pain was more common in bacterial diarrhoea than diarrhoea of other aetiology (P < 0.001 for both). During the hospital stay, diarrhoea persisted for longer duration in children with bacterial, protozoal or mixed aetiology than in those with rotavirus. Fifty-two percent of rotavirus cases settled within 48 hours of admission compared with 39% for the other group, while 16% of the rotavirus group had diarrhoea persisting beyond 96 hours of admission compared with 31% for the other group (P < 0.05 for both comparisons). This study also gave information on the incidence of enteric pathogens and biochemical abnormalities detected during laboratory investigations, and that information has been included under the relevant sections. [EL = 3]
In a cross-sectional study from Kuwait,44 the duration of diarrhoea and clinical characteristics associated with various pathogens were examined. In total, 595 children (age range under 1 year to 12 years) admitted to the hospital with gastroenteritis were included. Two stool specimens were collected within 18 hours of admission to determine the pathogens responsible for gastroenteritis. Rotavirus was the most common pathogen detected, in 45% of children (34% pure isolates and the rest in combination with bacteria), followed by salmonella in 24% of children (17% pure isolates and 7% in combination with rotavirus). The mean duration of diarrhoea was longer in those with salmonella (12.3 days) than in those with rotavirus (4.8 days) or other bacterial pathogens detected alone or in combination with other infections (mean duration ranging from 6.8 to 7.9 days) (Table 3.2). Gastroenteritis with rotavirus infection had a self-limiting course with no mortality and low morbidity and was associated with few extra-intestinal manifestations, while salmonella infections had the highest frequency of milk intolerance (31%), malnutrition (36%), and associated features (convulsions in 20% and septicaemia in 17%). Abdominal pain and bloody diarrhoea were common in infections with shigella, salmonella and campylobacter. Features of extra-intestinal invasions (such as toxic look, fever > 3 days, poor feeding, hepatosplenomegaly and pneumonia) were seen almost exclusively in children with salmonella infections. [EL = 3]
Table 3.2
Duration of diarrhoea in 595 children with gastroenteritis.
Risk factors for persistent diarrhoea
A case–control study from India54 investigated possible risk factors for persistent diarrhoea (>14 days). They compared 170 cases with 340 controls (aged 1–23 months) who had been admitted to hospital. Cases were children with diarrhoea persisting for more than 14 days while the controls had all recovered within 7 days of admission. For each case, two age-matched controls were recruited. If the controls became cases either during hospital stay or follow-up, they were withdrawn and fresh controls recruited. The episode of diarrhoea was considered to be controlled when a child had no diarrhoea for 2 consecutive days. Fifteen potential risk factors were examined. After adjusting for co-variables by multiple logistic regression analysis, the factors independently associated with persistent diarrhoea were malnutrition (adjusted OR 2.9; 95% CI 1.9 to 4.5), stool with blood and/or mucus (adjusted OR 2.4; 95% CI 1.3 to 4.3), indiscriminate use of antibiotics (adjusted OR 2.4; 95% CI 1.6 to 3.9), stool frequency more than 10 per day (adjusted OR 1.8; 95% CI 1.2 to 2.8) and persistence of dehydration for more than 24 hours (adjusted OR 1.4; 95% CI 1.2 to 1.7). [EL = 2+]
Evidence summary
Results from three cross-sectional studies [EL = 3] suggested that viral gastroenteritis was associated with a shorter duration of diarrhoea than bacterial gastroenteritis. The fourth cross-sectional study from Italy did not report the duration separately for the viral and bacterial infections. There was consistent evidence from these studies to indicate that vomiting and dehydration were more common with viral gastroenteritis while bloody diarrhoea and abdominal pain were more often associated with bacterial gastroenteritis. A case–control study [EL = 2+] reported that malnutrition, bloody and/or mucoid stools, indiscriminate use of antibiotics, high stool frequency (>10 per day), and persistence of dehydration (>24 hours) were factors independently associated with an increased risk of persistent diarrhoea.
GDG translation from evidence to recommendation
Several cross-sectional studies have provided consistent evidence of differences in the manifestations of viral and bacterial gastroenteritis. Viral gastroenteritis is of shorter duration than bacterial gastroenteritis and associated with an increased risk of vomiting and dehydration. Bacterial gastroenteritis is more often associated with severe abdominal pain and sometimes with bloody diarrhoea.
While the GDG recognised that the study of risk factors for persistent diarrhoea was based on a population in India, the findings were nevertheless important, especially given the multi-ethnic composition of the UK population and the increasing frequency of overseas travel. Some potentially important risk factors for persistent diarrhoea were identified: persistence of dehydration for more than 24 hours after appropriate fluid therapy, presence of blood and/or mucus in the stool, stool frequency greater or equal to 10 per day, indiscriminate use of antimicrobials, weight loss and poor nutritional status.
Clinical question
In children presenting with diarrhoea and/or vomiting, what characteristics may suggest a diagnosis other than gastroenteritis?
Acute diarrhoea is not always due to an intestinal infection. Other non-enteric infections and non-infective gastrointestinal disorders may be responsible. Diarrhoea is a common side effect of antibiotic therapy. Many other drugs and certain dietary constituents (for example, sorbitol and xylitol), may occasionally be responsible and on rare occasions certain toxins (for example, organophosphate insecticides).55
Evidence overview
In the absence of comparative studies, the literature search for this question was undertaken to identify case series with sample sizes more than 100. Another source of information was the NICE published guideline Feverish illness in children.56 It had recommended signs and symptoms for identifying children with fever who are at high risk for serious illness. The list of various alternative diagnoses based upon the published evidence and the consensus view of the GDG are given in Table 3.3.
Table 3.3
Key symptoms and signs in the differential diagnosis of children presenting with diarrhoea and/or vomiting.
GDG translation from evidence to recommendation
Based on consensus, the GDG identified a number of key points that they considered were important in the history and examination of the child. It would be important to be aware of any history of contact with an individual with symptoms of gastroenteritis. Also, exposure of the child to a known source of enteric infection or a history of recent travel abroad would be important. Certain symptoms and signs might point to diagnosis other than gastroenteritis. For example, although fever may occur in children with gastroenteritis, high fever is somewhat unusual. Therefore the possibility of other disorders would require careful consideration in such cases.
The GDG also considered that the main conditions in the differential diagnosis of gastroenteritis include non-enteric infections, non-infective gastrointestinal disorders, abdominal surgical disorders and antibiotic-associated diarrhoea. It was agreed that, if vomiting persisted for more than 24 hours without diarrhoea, diagnoses other than gastroenteritis (for example, urinary tract infection) should be considered.
Recommendations on diagnosis
If you suspect gastroenteritis, ask about:
recent contact with someone with acute diarrhoea and/or vomiting and
exposure to a known source of enteric infection (possibly contaminated water or food) and
recent travel abroad.
Consider any of the following as possible indicators of diagnoses other than gastroenteritis:
fever:
- –
temperature of 38 °C or higher in children younger than 3 months
- –
temperature of 39 °C or higher in children aged 3 months or older
shortness of breath or tachypnoea
altered conscious state
neck stiffness
bulging fontanelle in infants
non-blanching rash
blood and/or mucus in stool
bilious (green) vomit
severe or localised abdominal pain
abdominal distension or rebound tenderness.
3.2. Laboratory investigations in diagnosis
Introduction
Most children with gastroenteritis do not require any laboratory investigations. Many infants and children experience brief episodes of diarrhoea and are managed by their parents without seeking professional advice. Even if advice is sought, healthcare professionals often consider that a clinical assessment is all that is required, and laboratory investigations are not undertaken. However, there may be particular circ*mstances when investigations may be helpful in diagnosis.
3.2.1. Stool microbiological investigation
In clinical practice, most children with gastroenteritis do not undergo any stool investigations and no attempt is made to identify the presumptive enteric pathogen. In some cases, however, microbiological investigation may be appropriate. There might be circ*mstances in which identification would be important. Some pathogens are of special significance. For example, amoebic dysentery would require antimicrobial therapy. Escherichia coli O157:H7 is associated with a risk of haemolytic uraemic syndrome (HUS) – a serious and potentially deadly disorder. Bloody diarrhoea may be caused by serious non-infective conditions such as inflammatory bowel disease (ulcerative colitis or Crohn’s disease) and isolation of a bacterial pathogen might therefore be diagnostically helpful.
It was therefore important to determine the frequency with which enteric pathogens were identified by stool microbiological investigation. An additional group for consideration was the child with diarrhoea recently returned from overseas travel in whom the likely pathogens might differ.
Clinical question
In what proportion of children with a clinical diagnosis of gastroenteritis is a causative enteric pathogen identified? What is the incidence of specific pathogens?
It was important to first determine how frequently enteric pathogens can be found in children with gastroenteritis in the UK. The evidence to inform this question was taken from eight published studies from England and Wales and the website of the Health Protection Agency (HPA).
Evidence overview
The first three studies53,57,58 gave information on the proportion of children with enteric pathogens isolated but all these studies were hospital based and conducted before 1990. Of the remaining studies, four59–62 gave information on viral pathogens isolated in children with gastroenteritis but they all provided regional data. The last published paper11 is a population-based study comparing the incidence of gastroenteritis in the community with that in patients presenting to their general practitioner, but in this paper the study population included both adults and children. Since no published data were available for the overall incidence of enteric pathogens in children with gastroenteritis from England and Wales, this information was collected from the website of the UK HPA (www.hpa.org.uk).
A prospective study53 included 1148 children younger than 16 years admitted to a sub-regional infectious disease hospital with a diagnosis of gastroenteritis over a 1 year period (1986–87). The demographic characteristics of the study population have been described in detail in Section 3.1 above. In this study, 8.8% of children (101/1148) were clinically dehydrated, with 1% assessed to have greater than 5% dehydration. Dehydrated children were more likely to have an enteric pathogen identified compared with those without clinical dehydration (61% versus 43%; P < 0.001). Stool examination identified enteric pathogens in altogether 44.6% of cases (512/1148), with rotavirus being the most common (in 31%). The common bacterial pathogens isolated were salmonella spp. (5%), campylobacter spp. (3.2%), and enteropathogenic E. coli (2%), while cryptosporidium (1.4%) was the most common protozoal organism detected. [EL = 3]
The second study was a prospective survey57 that recruited 447 children younger than 2 years and admitted to a hospital with gastroenteritis over a 1 year period (1981–82). Seventy-four percent of the children were younger than 1 year and two-thirds of under-1-year-olds were younger than 6 months. Pathogenic enteric organisms were isolated in 75% of cases (335/447), with viruses alone in 57%, bacteria alone in 6% and both viruses and bacteria isolated in 10% of cases. Rotavirus was the most common organism isolated, in 34% of cases (152/447), while all other viruses were detected in 53% of children. Enteropathogenic E. coli (6.9%), Campylobacter jejuni (5.1%), Clostridium difficile toxin (4.9%), salmonella spp. (4.3%) and shigella spp. (2%) were the main bacteria isolated. [EL = 3]
The third study58 included 215 children admitted to four paediatric units in south Wales with gastroenteritis over a 1 year period (1987–88). The age of the study population ranged from 2 weeks to 9 years and 61% of children were younger than 1 year. The primary aim of the study was to describe the clinical characteristics, incidence of complications, and management (pre-admission and hospital) of the patients. The authors did not specify the total number of cases with clinical dehydration, but overall only 7% were judged to be severely dehydrated. Pathogenic organisms were isolated in the stools of 58% of children (125/215) with virus alone in 30% (65/215). Among the viral pathogens, rotavirus was the most frequently isolated (83% of all viruses). Bacteria alone were found in 14% of cases, with campylobacter and E. coli being the most common. Cryptosporidium spp. were the most common parasites detected. [EL = 3]
A surveillance study59 reported viral agents isolated from clinical specimens in a regional public health laboratory in the UK during the winter of 1999–2000. Altogether 3172 specimens (stool or occasional vomit) of sporadic cases of gastroenteritis in children under the age of 7 years were sent to the laboratory. Samples were received from general practitioners (34%), hospitals (56%) and other public laboratories (9%) in the southwest of England and south Wales. Over the same period, 1360 specimens were also tested from 285 reported outbreaks of gastroenteritis, with 34% of these specimens coming from the community and the rest from the hospitals. A viral aetiology was confirmed in 37.7% of sporadic cases of gastroenteritis, with rotavirus being the most common viral agent detected, in 21.6% (685/3172) of them. The other enteric viruses detected were norovirus (10.3%), adenovirus (3.9%), astrovirus (3.1%) and calicivirus (0.2%). The prevalence of all the enteric viral agents was reported to be higher in the specimens from the community compared with the specimens from the hospitals. For the outbreaks, norovirus was the most common viral agent isolated (in 63.9% of outbreaks), followed by rotavirus (3.9%), adenovirus (0.4%) and astrovirus (0.4%). No viral agent was isolated in 32.6% of outbreaks. [EL = 3]
In another prospective study from a large hospital,60 stool samples of children (younger than 16 years) admitted to the hospital with a diagnosis of acute gastroenteritis or who developed the condition following hospitalisation were examined. Gastroenteritis was considered to be healthcare associated if symptoms developed 48 hours or more following admission. The study included a total of 243 subjects (87% of eligible subjects) who had clinical data and a stool specimen collected during the 5 month study period, and 37% (91/243) of these cases were judged to be healthcare associated while the rest were diagnosed to have community-acquired infection. Rotavirus was detected in altogether 29% of cases (71/243). The proportion of community-acquired cases testing positive for rotavirus was 36% (54/152), while for healthcare-associated cases the proportion was 19% (17/91). [EL = 3]
In the sixth study, conducted in East Anglia, England,62 stool samples collected during three consecutive winter seasons (2000–2003) were tested for the presence of viral pathogens. The sample population comprised 685 children younger than 6 years with symptoms of gastroenteritis: 223 children presenting to a general practitioner (part of a structured surveillance study evaluating burden of disease), 203 children referred by various general practices in the community to the hospital, and 259 children admitted to the hospital as inpatients or attending the accident and emergency department. A viral agent was detected in 53.4% of samples (366/685). A significantly greater number of children from the structured surveillance study group had a viral pathogen detected in their stool specimen compared with the community cohort (68.6% versus 51.2%; P < 0.05) or the hospital cohort (68.6% versus 42.1%; P < 0.05). The proportion of children from the community with a viral agent detected was also significantly higher compared with children in the hospital cohort (51.2% versus 42.1%; P < 0.05). Rotavirus was the most common pathogen isolated in each of the three cohorts: 40.4% in the structured surveillance study, 24.6% in the community cohort and 17.8% in the hospital cohort. The second most common viral pathogen isolated was norovirus in the surveillance study group (9.9%) and the hospital cohort group (9.7%), while in the community cohort group it was enteric adenovirus (8.9%). Multiple viral pathogens were detected in 8% of the samples and most of these (72.7%) involved rotavirus in combination with other viruses. [EL = 3]
As part of a prospective, multicentre study on the incidence of rotavirus in Europe,61 1010 stool samples were examined from children younger than 5 years who presented with gastroenteritis to a regional health service in the UK in the winter of 2004–2005. The results were reported in relation to the clinical setting in which they were seen: primary care (general practitioners and/or paediatricians), emergency department and hospital admission. The overall percentage of children with rotavirus-positive gastroenteritis was estimated to be 35.9%, with the incidence being almost the same for the two groups of children seen in the emergency department and hospital (60.0% and 60.7%, respectively). In a primary care setting, rotavirus was isolated in 31.9% of the samples. [EL = 3]
The last study was a population-based study11 undertaken to evaluate the incidence and aetiology of infectious intestinal disease in both adults and children presenting to general practitioners and in the community, and to establish how many of them are reported to the national laboratory-based surveillance. The study cohort included a population of 459 975 patients registered with 70 general practices in England, and this selected sample of population was representative of all the general practices nationally with respect to geographical location, urban and rural characteristics, and social deprivation index. To calculate incidence in the community, 200 people were randomly recruited from each general practice out of which 9776 people (an average of 140 people from each general practice) agreed to participate. These people were asked to return weekly postcards for 6 months declaring the absence of symptoms, and those with symptoms were asked to send a stool specimen to a public health laboratory. Eighty-two percent of the participants returned 22 or more of the 26 weekly postcards. For the general practice incidence data, all cases of infectious intestinal disease presenting to a general practitioner were eligible for inclusion irrespective of their age. The practices were randomly allocated to two arms: in the first arm, all patients (34 general practices) were asked to send stool specimens to a public health laboratory, while in the second arm (36 general practices), stool testing was conducted locally and information sought from the national surveillance database on those with a positive stool specimen. This step was taken to evaluate the completeness of the reporting system.
In the community, 781 cases were ascertained for infectious intestinal disease, with an incidence of 19.4 per 100 person-years (95% CI 18.1 to 20.8), while 8770 people presented to their general practice giving an incidence of 3.3 per 100 person-years (95% CI 2.94 to 3.75). The ratio of community incidence to general practice presentation was 5.8, suggesting that, for every case presenting to general practice with intestinal disease, almost six more cases were present in the community. The ratio was high for cases associated with E. coli non-O157, yersinia, rotavirus group C, C. difficile cytotoxin, aeromonas and for cases where no organism was isolated. In contrast, the ratio was lower for cases with salmonella and shigella infection, indicating that most people having these infections present to their general practitioners (Table 3.4). On comparing the results of the reporting system, it was found that cases of non-bacterial gastroenteritis were less likely to be reported to national surveillance. The rate ratio of community cases of gastroenteritis to the cases reaching national surveillance scheme was lower for bacterial pathogens (salmonella 3.2 : 1, campylobacter 7.6 : 1) compared with that of the viruses (rotavirus 35 : 1). [EL = 3]
Table 3.4
Incidence of infectious intestinal disease identified in patients from the community and those presenting to general practice.
The UK HPA is a non-departmental public body and its Centre for Infections carries out a range of work on the prevention of infectious disease. The remit of this body includes infectious disease surveillance, and it regularly updates data on enteric pathogens isolated in patients with gastroenteritis. The data are based on submitted laboratory reports, are stratified by regions, age group and year, and although incomplete can identify important trends. Table 3.5 lists the various pathogens identified in the stool samples of children over the period 2002–2006.
Table 3.5
Laboratory reports of enteric pathogen isolates, England and Wales (2002–2006) stratified by age; data from the HPA website.
Evidence summary
Although results from three hospital-based studies show variation in the proportion of children with gastroenteritis (45%, 75% and 58%) who had pathogenic enteric organisms isolated from their stool examination, rotavirus was detected as the most common cause of gastroenteritis in children in all the studies. Bacterial and protozoal organisms were detected less commonly. Other studies have identified norovirus and adenovirus as other common viral causes, with norovirus being more common than adenovirus. However, rotavirus was identified about four times more often and the results were similar from the community and from hospital settings.
The most frequent pathogens causing bacterial gastroenteritis reported in the published studies and the HPA website were campylobacter, salmonella and E. coli, while cryptosporidium was the most common protozoal infection detected. There were no published population-based data comparing the detection of viral pathogens with bacterial pathogens, but results from the population-based study suggest that cases of non-bacterial gastroenteritis are less likely to be reported to the national surveillance authority than the cases of bacterial gastroenteritis.
Clinical question
Which enteric pathogens are most commonly identified in children returning to the UK with traveller’s diarrhoea?
Gastroenteritis is the most commonly reported travel-associated illness in England and Wales. No published data were identified to provide information on this question. Information was again collected from the HPA website,64 as given in Table 3.6, but the age-specific data were not available.
Table 3.6
Laboratory reports of enteric pathogens isolated from patients with a history of recent travel abroad; data from the HPA website.
Evidence summary
Salmonella was the most commonly reported cause of gastroenteritis associated with overseas travel. The other commonly reported enteric pathogens were campylobacter, shigella and cryptosporidium. While salmonella, campylobacter and cryptosporidium were reported to be associated with travel in Europe, infections with shigella, giardia and entamoeba were associated with travel to the Indian subcontinent and sub-Saharan and southern Africa.
GDG translation from evidence to recommendations
Rotavirus is the predominant single pathogen responsible for gastroenteritis in children. Other viral pathogens, although individually less common, collectively accounted for many other cases. Norovirus has been recognised increasingly as an important cause of gastroenteritis. A range of other viruses, including adenovirus, astrovirus, and calicivirus may also be responsible. Bacterial infections, particularly campylobacter and salmonella species, and protozoal infections such as cryptosporidium may also cause gastroenteritis. In those with a history of recent overseas travel, the likelihood of infection with salmonella, campylobacter, cryptosporidium and other infections unusual in the UK is greater.
The GDG agreed that in, in keeping with current practice, children presenting with acute diarrhoea do not usually require stool microbiological investigation. Most have viral gastroenteritis. Even if the illness is due to a bacterial or protozoal enteric infection, most children do not require antimicrobial treatment (see Chapter 7 on antibiotic therapy), and so identification of a specific pathogen is not generally required.
However, the GDG agreed that in some circ*mstances microbiological investigation should be considered, as listed below.
If the clinician is in doubt about the diagnosis of gastroenteritis, isolation of a stool pathogen can provide diagnostic reassurance. Using routine microbiological techniques including stool microscopy, culture and standard viral detection methods, a causative pathogen can be identified in most patients with gastroenteritis.
It is also important to investigate those who present with bloody and/or mucoid diarrhoea, because this can be due to a range of important and potentially serious disorders.
- –
Various non-infective inflammatory bowel disorders may present with bloody stools. Infants (both breastfed and bottle-fed) may present with bloody diarrhoea due to non-specific colitis; this may be an allergic phenomenon, although the aetiology is often uncertain. Inflammatory bowel disease (ulcerative colitis or Crohn’s disease) may sometimes begin in early childhood, and the presentation can be acute and associated with bloody diarrhoea. Without evidence of an infective cause, it may be impossible to distinguish between bacterial or amoebic dysentery and a first episode of bloody diarrhoea due to ulcerative colitis or Crohn’s colitis.
- –
Various surgical disorders, such as intussusception, could also be associated with passage of blood and/or mucus. Although most children with bloody diarrhoea due to an enteric infection do not require antimicrobial therapy, this is not always so. Treatment is indicated for salmonella gastroenteritis in young infants and in immunocompromised children (see Chapter 7)
- –
Enteric infection with enterohaemorrhagic E. coli (mainly E. coli O157:H7) often presents with bloody diarrhoea. These children are at risk of developing HUS, a life-threatening condition. Early microbiological diagnosis is therefore important.
- –
Rarely, children may present with bloody diarrhoea associated with antibiotic-related pseudomembranous colitis, often due to C. difficile. Here C. difficile toxin may be detected in the stool.
It is also reasonable to consider investigating children in whom diarrhoea persists for more than 7 days, because certain treatable enteric infections (for example, giardiasis) may be responsible.
If a child presenting with acute diarrhoea is very ill and the possibility of septicaemia requires empiric antibiotic therapy, stool microbiological investigation would be indicated. Some patients with bacterial dysentery (salmonella or shigella) may have bacterial septicaemia, and stool culture could identify the responsible pathogen.
Various pathogens requiring antimicrobial treatment, such as Giardia lamblia, shigella and salmonella (in selected cases), are more commonly identified in patients returning from other countries. The GDG therefore agreed that stool microbiological testing should be considered in those with a history of recent overseas travel.
The GDG recognised the importance of providing appropriate and adequate clinical information to the laboratory, including the clinician’s suspicion of any unusual pathogen, in order to inform the investigative strategy. Certain organisms require special arrangements for collection or transport to optimise identification. Particular pathogens may sometimes require a targeted approach with specific laboratory techniques.
The GDG recognised that within hospitals and other institutions it may be important to gather data on the specific pathogens responsible for gastroenteritis, but policy on this is outside the scope of this guideline.
The GDG recognised that the public health authorities have access to local and national epidemiological data, and have a central role in the management of outbreaks of gastroenteritis. It is therefore important to discuss with the authority any suspected outbreak of gastroenteritis and its implications.
Recommendations on stool microbiological investigation in diagnosis
Consider performing stool microbiological investigations if:
the child has recently been abroad or
the diarrhoea has not improved by day 7 or
there is uncertainty about the diagnosis of gastroenteritis.
Perform stool microbiological investigations if:
you suspect septicaemia or
there is blood and/or mucus in the stool or
the child is immunocompromised.
Notify and act on the advice of the public health authorities if you suspect an outbreak of gastroenteritis.
If stool microbiology is performed:
collect, store and transport stool specimens as advised by the investigating laboratory
provide the laboratory with relevant clinical information.
3.2.2. Other laboratory investigations
The GDG examined evidence regarding the potential value of various blood tests in distinguishing bacterial from viral gastroenteritis. As discussed earlier, this might be important for clinical management in some patients.
Clinical question
How accurate are laboratory blood tests in distinguishing bacterial from viral gastroenteritis?
There is variation in practice regarding the use of laboratory blood tests in distinguishing between bacterial and viral causes of gastroenteritis, hence their accuracy in detecting these causes was sought.
Evidence overview
Four diagnostic studies were included in this section, one with EL = 2 and the rest with EL = 3. In the first three studies, the accuracy of acute-phase proteins was evaluated for detecting bacterial gastroenteritis, and C-reactive protein (CRP) was assessed in all the studies, while ESR, interleukin-6 (IL-6) and interleukin-8 (IL-8) were assessed in one study each. The last study evaluated the diagnostic ability of total and differential blood count in differentiating bacterial from viral causes of gastroenteritis.
The first study, from Italy,65 looked at the diagnostic accuracy of CRP and ESR measurements in the differentiation of bacterial and viral gastroenteritis. Over a 4 year period, it recruited 111 children aged between 1 and 60 months admitted to a hospital with acute diarrhoea lasting more than 12 hours but less than 15 days. Children with chronic gastrointestinal diseases such as cow’s milk protein intolerance, Crohn’s disease, gastro-oesophageal reflux or chronic diseases were excluded. After admission, all children had blood taken for the measurement of CRP and ESR levels, while stool culture was performed to detect bacterial aetiology and viruses detected by ELISA testing on the stool specimens. The accuracy of CRP in detecting bacterial or viral gastroenteritis was calculated at the cut-off values of 12, 20 and 35 mg/l, while elevated ESR was taken as value ≥ 25 mm/hour. Of the 111 children, 53 (48%) were diagnosed with bacterial gastroenteritis (mainly non-typhoidal salmonella), 35 (32%) had viral gastroenteritis and the remaining 21% had culture-negative infections. The mean CRP level in children with bacterial infections was significantly higher than in those with viral infections (P < 0.001) and culture-negative infections (P < 0.01). CRP levels were strongly associated with bacterial infections at all three cut-offs: 12 mg/dl (OR 25.8; 95% CI 7.6 to 87.9), 20 mg/l (OR 46.4; 95% CI 5.9 to 365) and 35 mg/l (OR 27; 95% CI 3.4 to 212). The specificity of CRP in detecting bacterial gastroenteritis was high at all the cut-off levels (89% at 12 mg/l and 97% at both 20 mg/l and 35 mg/l) but the highest sensitivity was 77% at 12 mg/dl, compared with 58% and 44% at the other two cut-off values, respectively. The area under receiver operating characteristic curve (AROC) at 12 mg/l was 0.83. Raised ESR levels (≥25 mm/hour) were also strongly associated with bacterial infections (OR 3.5; 95% CI 1.2 to 9.9) and showed a sensitivity of 42%, specificity of 83% and AROC of 0.62 for detecting them. Raised total leucocyte count did not show any statistically significant association with any of the three infections. [EL = 2]
The second study, from Taiwan/China,66 aimed to determine whether IL-6, IL-8 and CRP were useful diagnostic markers in differentiating bacterial from viral gastroenteritis. The study included 56 children (mean age 2.5 years) admitted with acute gastroenteritis, of whom 21 had rotavirus (by Rotaclone® test), 18 had bacterial infections (by stool culture with salmonella species isolated predominantly) while 17 children were recruited as controls. Children with chronic disease or history of persistent/intractable diarrhoea were excluded. No details were provided about the control group or exclusion criteria. The concentration of both CRP and IL-6 were significantly higher in children with bacterial gastroenteritis than in those with viral infections (P < 0.001) and the control group (P < 0.001). IL-8 concentrations were elevated in both bacterial and viral infections and there was no statistically significant difference in the levels between the two groups. Diagnostic accuracy results were analysed using ROC curves and it showed best results for CRP, with the AROC being 0.90 at the cut-off value of 2 mg/dl, followed by IL-6 with an AROC of 0.83 at the cut-off value of 10 pg/ml. At these cut-off values, the sensitivity and specificity of CRP in detecting bacterial gastroenteritis were 83% and 76%, respectively, while those of IL-6 were 78% and 86%, respectively. IL-8 was found to be of less diagnostic value, with an AROC of 0.68, sensitivity of 50% and specificity of 67% at the cut-off value 70 pg/ml. [EL = 3]
In the third study, from Israel,67 the ability of the quick-read CRP (QR-CRP) test to detect bacterial gastroenteritis was determined in a convenience sample of 44 children (age range 4 days to 17 years, median age of 2.4 years) admitted to the emergency department of a tertiary hospital. All children had symptoms of vomiting, diarrhoea more than three episodes and fever and underwent laboratory testing. Exclusion criteria were not defined. Bacterial aetiology was determined by stool culture while antigen testing was used to isolate rotavirus, but it was done in only 28 children. QR-CRP was performed at the bedside with a level of 8 mg/l or more considered as a positive test. Stool culture was positive for bacteria in eight children while rotavirus was isolated in 13 children. The mean CRP concentration was significantly higher in children with bacterial gastroenteritis than in those with viral gastroenteritis (P < 0.001). The ROC curve was used to calculate the diagnostic accuracy of QR-CRP. The best cut-off value derived from the ROC curve was 95 mg/l and, at this cut-off value, QR-CRP showed a sensitivity of 87%, specificity of 92% and AROC of 0.94 in detecting bacterial gastroenteritis. [EL = 3]
Another study from Israel68 evaluated the ability of total and differential leucocyte counts to differentiate bacterial from non-bacterial gastroenteritis infections. This study recruited 238 children admitted to hospital with gastroenteritis but further details about demographic characteristics were not specified. Bacterial pathogens were isolated by stool culture, along with testing for blood counts but no further details were provided about these tests. One hundred and ninety-two children had bacterial gastroenteritis (shigella in 130, salmonella and campylobacter in 25 each and E. coli in 12) while 46 children were in the non-bacterial group. The total white blood counts were similar between the aetiological groups but great variation was observed in the differential blood counts. The absolute band neutrophil count and the ratio of band neutrophils to total neutrophils were significantly higher in the shigella, salmonella and campylobacter groups compared with the E. coli and non-bacterial groups (P < 0.05 for all comparisons). Among all the bacterial pathogens, children with shigella had the highest values for both these parameters. It was found that band neutrophils to total neutrophils ratio of more than 0.10 could differentiate bacterial infections from E. coli and non-bacterial groups with a sensitivity of 84% and a specificity of 75%. [EL = 3]
Evidence summary
There was a lack of good-quality studies to evaluate the ability of laboratory tests to distinguish between bacterial and viral gastroenteritis. Evidence from three studies suggested that raised CRP levels had a high diagnostic ability in detecting bacterial causes, with AROC ranging between 0.83 and 0.94, but the studies used different cut-off values to define a positive CRP test. Other acute-phase proteins (IL-6 and IL-8) and raised ESR levels were found to be less accurate than CRP. The total leucocyte blood count was not helpful in distinguishing bacterial from non-bacterial/viral causes in two studies, while one study reported raised absolute neutrophil band cell count and the ratio of band neutrophils to total neutrophils count (ratio > 0.10) to be useful in distinguishing between the pathogens.
GDG translation from evidence to recommendations
There was evidence that in children with gastroenteritis an elevated CRP would support a diagnosis of bacterial rather than viral gastroenteritis. However, a normal CRP does not exclude the possibility of bacterial gastroenteritis. As discussed elsewhere (Chapter 7 on antibiotic therapy), in the UK most children with bacterial gastroenteritis do not require antibiotic treatment. However, infants younger than 6 months and immunocompromised children should be treated with antibiotics if they have salmonella gastroenteritis. In such vulnerable patients if bacterial gastroenteritis is clinically suspected, antibiotic therapy should be commenced while awaiting the results of stool microbiological investigations. It would not be safe to withhold antibiotic treatment based on a normal CRP result. As measurement of the CRP would rarely influence management, the GDG did not consider that its routine use would be cost-effective.
The GDG considered that, in keeping with normal clinical practice, a blood culture (the gold standard for septicaemia) should be performed prior to commencing antibiotic therapy in children with suspected or confirmed bacterial gastroenteritis.
Children with E. coli O157:H7 infection are at risk of developing HUS and the GDG considered that these children should be monitored for the development of microangiopathic haemolytic anaemia, thrombocytopenia and renal insufficiency. This should be done in consultation with an appropriate specialist.
Recommendations on other laboratory investigations in diagnosis
Perform a blood culture if giving antibiotic therapy.
In children with Escherichia coli O157:H7 infection, seek specialist advice on monitoring for haemolytic uraemic syndrome.
