Angharad Griffiths. Gastroenteritis, Don't Forget the Bubbles, 2020. Available at:
https://doi.org/10.31440/DFTB.28790
Conor is a 10 kg, 13 month-old boy who’s presented to the ED with a 24-hour history of diarrhoea and vomiting. He has had 5 episodes of non-bloody, non-bilious vomits. Since waking up this morning has two episodes of loose/watery non-bloody malodorous stools. They have not ‘flooded’ the nappy but were quite large. He is taking sips of fluid (mixtures of water, milk, and juice being offered) and has only eaten half a digestive biscuit so far today. He has had a fairly large wet nappy last night, but not since, though it’s now difficult to tell as his last nappy was dirty. He is alert and looking around while being carried but is upset on leaving his mother’s arms. He cries with tears, has a normal heart rate but his mother is worried about his dry lips. She was told by a healthcare worker neighbour that he will “need a drip”. CRT, HR, and BP are normal. His temperature is 37.8. His nappy is dry and has been on for 3 hours now. His capillary glucose measurement is 3.2. You decide he’s probably mildly dehydrated.
Introduction
Gastroenteritis (GE) is the presence or diarrhoea or vomiting (or both) that may or may not be accompanied by fever, abdominal pain and anorexia. Diarrhoea is the passage of excessively liquidy or frequent stools with liquid high water content. Although often felt to be a common minor illness presentation, it is a major cause of childhood mortality and morbidity, causing millions of deaths worldwide in children in low and middle-income countries; of all child deaths from gastroenteritis 78% occur in Africa and South-East Asia.
Gastroenteritis accounts for a huge proportion of GP and ED presentations. In Europe, acute gastroenteritis the third commonest cause of hospital admission, accounting for between 4-17% of admissions. In Australia, gastroenteritis caused by rotavirus alone accounts for 115,000 GP visits, 22,000 ED visits and 10,000 hospital admissions a year, with an estimated cost of 30m Australian Dollars (£12m, €18m). In the UK, 20% of GP consultations in the under 5’s are for GE.
It is imperative that the child with gastroenteritis is differentiated from more sinister causes of vomiting. The presence of diarrhoea is reassuring but doesn’t exclude other intra-abdominal causes. The same can be said for pain out of proportion with gastroenteritis, distension, peritoneal signs or localised tenderness.
Most cases are not associated with complications but when complications do occur, the commonest are electrolyte disturbance and metabolic acidosis. Supplementary fluids through oral or intravenous routes are the most effective way to avoid these complications.
Gastroenteritis in low and middle-income countries can present differently, has different aetiologies, is often managed differently, and is a larger burden to healthcare systems in general than in high-income countries. This post will focus on gastroenteritis in high-income countries. For more information about comparisons of guidelines across the world; Vecchio et al (2016) is an interesting read.
This is not meant to provide a clinical practice guideline; rather an overview of the illness. Many (if not all!) paediatric emergency departments or general paediatric units have their own guidelines.
Pathophysiology
Worldwide, the commonest causes are viral pathogens, most commonly rotaviruses and noroviruses. Viral infections cause damage to the small bowel enterocytes with resultant low-grade fevers and watery diarrhoea – classically without blood. Rotavirus strains are seasonal and vary within different geographical areas. The peak age for these infections is between 6 months and 2 years. Children with poor nutrition are at higher risk of acquiring gastroenteritis and developing dehydration and complications.
Children with bacterial gastroenteritis are more likely to have bloody stool.
Escherichia coli and Shigella dysenteriae can be complicated by haemolytic uraemic syndrome (HUS). This is an acute onset, microangiopathic haemolytic anaemia, thrombocytopaenia, acute renal impairment and multisystem involvement. (Just to confuse things, HUS can present in the absence of bloody diarrhoea.)
Pathogens can be generalised into four groups:
- Viral (70% of cases): Rotavirus, Norovirus, Adenovirus, Enterovirus
- Bacterial (10-20% of cases): Campylobacter jejuni, Salmonella spp, Escherichia coli, Shigella spp, Yersinia enterocolitica.
- Protozoa (unusual, accounting for <10%): Cryptosporidium, Giardia lamblia, Entamoeba histolytica
- Helminths (very unusual): Strongyloides stercoralis
Transmission
Pathogens are spread mainly via the faeco-oral route, acquired by ingesting contaminated food or drink. Water may be contaminated with bacteria, viruses, or protozoa. Undercooked (or inappropriately stored/cooked) meats and seafood are common culprits of bacterial pathogens. Bacterial contaminants can produce toxins (e.g. Bacillus cereus in re-warmed rice or Staphylococcus aureus in ice-cream).
Pathogens causing gastroenteritis can also be transmitted without the patient being symptomatic.
Assessment
Gastroenteritis is a clinical diagnosis. Enquire about sick/infectious contacts and potential sources (recent travel, food). Enquire about the frequency of symptoms and intake of fluids. Note the frequency of urination. Note other things that may cause diarrhoea e.g. recent use of enteral antibiotics or chronic constipation with overflow diarrhoea the presenting feature.
In the presence of signs such as high fever, long duration of symptoms, severe abdominal pain or bilious vomiting; review the diagnosis and do not immediately label as gastroenteritis.
Oral hydration fluids
Most children are not dehydrated and can tolerate oral fluids and so can be managed at home. Take a look at Nikki Abela’s DFTB19 talk on top tips for a high yield dehydration assessment.
When children are only mildly to moderately dehydrated, as a general rule they can be treated with oral / enteral rehydration with low osmolality oral rehydration solution (ORS). Worldwide, ORS is recognised as first line therapy and treating mild to moderate dehydration with enteral rehydration is supported by the WHO, European Society for Paediatric Gastroenterology and the American Academy of Paediatrics. The WHO recommends a low osmolality (hypo-osmolar) solution, usually containing sodium, potassium, chloride, carbohydrate (glucose) and a base. Low osmolarity solutions reduce the need for IV fluids, reduce stool output and reduce vomiting frequency.
But… a major limitation to the use of ORS is its taste – and this is where apple juice comes in. For minimally dehydrated patients, half-strength apple juice is associated with fewer treatment failures compared to ORS and could suit as a more palatable alternative. Take a look at a sweet summary (pun intended!) of the “apple juice trial”.
Breastfeeding should continue and a child can be supplemented with ORS if this is needed. Children can go back to a normal diet after the illness has passed.
Enteral (oral / NG) versus IV hydration
Most studies show that enteral rehydration with ORS is just as effective as IV hydration in mild to moderate dehydration with a 2006 Cochrane analysis concluding that enteral rehydration is as effective if not better than IV rehydration with fewer adverse events and a shorter hospital stay. It is also less invasive (even with NG placement) and anecdotally satisfaction is greater amongst parents. It is very safe.
Enteral rehydration only fails in approximately 1 in 20-25 children.
Barriers to oral rehydration include unfamiliarity with the benefits, misconception that it takes longer than IV therapy, and that it has a high failure rate.
Contraindications to enteral rehydration include haemodynamic instability, abdominal distension, concern over ileus, absent bowel sounds, or impaired airway reflexes.
IV therapy is more invasive and involves placing and maintaining IV access. There are also iatrogenic complications including electrolyte disturbance should inappropriate fluids / composition / volume / rate be used.
But… in severely dehydrated children, put away the ORS and apple juice. They will need IV rehydration as first line.
Antiemetics
How can we support enteral fluids? Well, children who receive Ondansetron are less likely to vomit, have greater oral intake and are less likely to require IV hydration. A Cochrane review demonstrates that Ondansetron also increases the proportion of children who stop vomiting when compared to placebo [RR1.4] and reduces the proportion of children needing IV therapy (and therefore admission rate) [RR 0.41]. Median length of stay is also shorter in the ED.
Reported side effects are rare with very few reported side effects other than a few cases of increased frequency of diarrhoea.
Antiemetics alleviate vomiting by acting on the ChemoReceptor Trigger Zone and vomiting centre. Ondansetron is a 5HT3 receptor antagonist. This class of antiemetics have fewer adverse effects (than dopamine antagonists, anticholinergics, antihistamines and corticosteroids) and can be safely used in children. The NICE guideline discusses its off-licence use (at time of publication it’s licence was for post-operative nausea and vomiting and chemotherapy induced vomiting).
Ondansetron prolongs the QT interval. Recommendations are it should be avoided in those with long QT and should be used in caution where there may be electrolyte imbalance (severe dehydration) or on other QT-prolonging medication.
Ondansetron is relatively cheap £1.71 for 10 4mg tablets and is available in oro-dispersible form (though these are much more expensive at £36 for 10x4mg tablets) and liquid (£36.82 for 40mg [50ml] bottle).
Probiotics
An ESPGHAN working group position paper on the use of probiotics in acute paediatric gastroenteritis concludes that:
- Effects seen in clinical trial are probiotic strain specific (this makes ‘trial-life’ difficult to replicate in ‘real-life’).
- A lack of evidence now doesn’t mean that there won’t be evidence sometime in the future.
- Safety profile of certain strains cannot be extrapolated to other strains.
- Studies that report benefits in certain doses in certain settings have insufficient evidence to support a health benefit at lower doses and different setting.
…the jury’s still out.
Other therapies
Antibiotics and anti-diarrhoeal agents aren’t routinely recommended in the management of paediatric gastroenteritis.
For gastroenteritis in high income countries, the WHO does not recommend adding zinc to a treatment regimen (it is for gastroenteritis in low and middle income countries).
Investigations
Routine lab testing in mild and moderate gastroenteritis is of little value in these patients and should be avoided unless clinically indicated.
This goes for stool samples too. Stool cultures are not routinely indicated in immunocompetent children with non-bloody diarrhoea.
Confirmation of viral gastroenteritis after the child has been discharged from the ED, and likely on the road to recovery at home, adds very little to (A) the clinical diagnosis of viral gastroenteritis in the ED, (B) the management plan and (C) the clinical outcome.
Should the investigation influence management, then stool sampling may be of benefit. This could be applicable where an outbreak may be suspected in school or creche, where there may be a public health benefit.
Stool samples should be sent in cases of bloody diarrhoea, immunodeficiency and recent foreign travel.
How about tests for dehydration? Sadly there is no one test that correlates clinically with dehydration. Urine specific gravity in infants is unreliable because the kidney reaches adult concentrating abilities after the age of 1. Also, the child often doesn’t begin urinating until rehydration has begun.
And glucose? Well, almost 10% of GE patients aged 1 month to 5 years in high income countries present with hypoglycaemia. Risk factors for hypoglycaemia on presentation include a longer duration of vomiting and increased frequency of vomiting. It would be reasonable to consider point of care glucose testing at triage for young children as identifying hypoglycaemia on clinical ground alone is difficult in this age group.
Prevention
The key to reducing the burden (and generally for an all-round happier life!) is in the prevention of acute gastroenteritis. Rotavirus vaccination is now commonplace thought the antibodies, the UK & Ireland and other countries around the world. It is very effective.
In the home and in the ED…Handwashing, handwashing, handwashing!
Vaccination leads to a profound reduction in presentations and admissions and a fall in overall seasonal workload, often within the first year after the introduction of universal vaccination against rotavirus. Even though only those under 1 year old are generally vaccinated, it has been shown to contribute to a significant herd effect with fewer cases than expected in older children. In Scotland, where initial vaccine uptake was 93- 94% during the first 2 years, annual rotavirus confirmed gastroenteritis cases fell by 84.7%, bed days reduced by 91% (from 325 to just 29), without any documented cases of intussusception. Reductions were seen across all age groups despite only infants receiving the vaccine. Similar results can be seen in other areas of the UK and Ireland.
The not to miss bits
- Do not assume isolated vomiting in a child is gastroenteritis. Consider other causes -these very widely from inborn errors of metabolism to diabetes mellitus, surgical obstruction to urinary tract infections. If you’d like to hear more, check out Dani’s talk on vomiting in children in DFTB Essentials.
- Beware chronic diarrhoea in an infant – do they have malabsorption or is this a presentation of IBD or an immunodeficiency?
- Beware the non-thriving child with diarrhoea.
- And beware chronic diarrhoea.
But what happened to Conor?
Conor was given a cup of Dioralyte ORS and his favourite beaker filled with Dioralyte. His mum was encouraged to give him syringes of 5 mls of Dioralyte frequently or for him to take sips from his beaker and was asked to document on a piece of paper how many he received. He vomited after 30 minutes of this therapy.
You give him a dose of Ondansetron and place an NG tube and give him 100mls (10ml/kg) over 1 hour after deciding he does not need rapid rehydration but slightly more than normal maintenance. He then receives maintenance volumes of Dioralyte via his NG, which he tolerates well and then starts to take his own sips from his beaker.
He does not vomit in the ED again, has one episode of loose stools, passes urine, and is tolerating fluids orally. He’s smiling at you! You feel he can be discharged and council his mum regarding regular fluid intake, choice of fluids, of any red flags, and encouraged to return in the event of any concern.
Conor’s Dad calls to say that Conor’s 3 year old sister at home is now vomiting too! But it’s OK – He’s not too worried about her and Conor’s Mum has advised his Dad to start giving her regular sips of Dioralyte at home…
References
Colletti JE, Brown KM, Sharieff GQ, Barata IA, Ishimine P. The Management of Children with Gastroenteritis and Dehydration in the Emergency Department. J Emerg Med [Internet]. 2010;38(5):686–98. Available from: https://dx.doi.org/10.1016/j.jemermed.2008.06.015
Elliott EJ. Acute gastroenteritis in children. Br Med J. 2007;334(7583):35–40.
Vecchio A Lo, Dias A, Berkley JA, Boey C, Cohen MB, Cruchet S, et al. Comparison of Recommendations in Clinical Practice Guidelines for Acute Gastroenteritis in Children. Gastroenterology. 2016;63(2):226–35.
Freedman SB, Willan AR, Boutis K, Schuh S. Effect of dilute apple juice and preferred fluids vs electrolyte maintenance solution on treatment failure among children with mild gastroenteritis: A randomized clinical trial. JAMA – J Am Med Assoc. 2016;315(18):1966–74.
BK F, A H, JC C. Enteral vs Intravenous regydration therapy for children with gastroenteritis: A meta-analysis of randomized controlled trials. Arch Paediatr Adolesc. 2004;158(1):483–90.
Hartling L, Bellemare S, Wiebe N, Kf R, Tp K, Wr C, et al. Oral versus intravenous rehydration for treating dehydration due to gastroenteritis in children (Review). 2006;
Fedorowicz Z, Jagannath V, Carter B. Antiemetics for reducing vomiting related to acute gastroenteritis in children and adolescents. [Internet]. Cochrane database of systematic reviews. 2011. Available from: https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD005506.pub5/full
NICE. Management of vomiting omiting in children and y young oung people with gastroenteritis : ondansetron. NICE GUIDELINES. 2014. p. 1–20.
Szajewska H, Guarino A, Hojsak I, Indrio F, Kolacek S, Shamir R, et al. Use of Probiotics for Management of Acute Gastroenteritis : A Position Paper by the ESPGHAN Working Group for Probiotics and Prebiotics. 2014;58(4):531–9.
Forrest R, Jones L, Willocks L, Hardie A, Templeton K. Impact of the introduction of rotavirus vaccination on paediatric hospital admissions , Lothian , Scotland : a retrospective observational study. 2017;323–7.
MARLOW RD, MUIR P, VIPOND I, TROTTER CL FA. Assessing the impacts from the first year of rotavirus vaccination in the UK. Arch Dis Child. 2015;100(Supl 3):A30.
Petechiae in Children – the PiC Study
Tessa Davis. Petechiae in Children – the PiC Study, Don't Forget the Bubbles, 2020. Available at:
https://doi.org/10.31440/DFTB.30782
Today the Lancet has published the long-awaited results of the Petechiae in Children (PiC) study. Team DFTB got our hands on a pre-publication copy to read, summarise, and analyse for you. So let’s get to it.
This PERUKI study by Waterfield et al. is a prospective, multicentre cohort study:
Waterfield T, Maney J-A, Fairley D, Lyttle MD, McKenna JP, Roland D, Corr M, McFetridge L, Mitchell H, Woolfall K, Lynn F, Patenall B, Shields MD, Validating clinical practice guidelines for the management of children with non-blanching rashes in the UK (PiC): a prospective, multicentre cohort study, The Lancet, 2020
Why is this study needed?
We are all somewhat terrified of children with fever and a non-blanching rash. We don’t want to miss meningococcal sepsis. Current guidelines are based on data from before the introduction of the Meningococcal B (2015) and C (1999) vaccines and consider a prevalence of 10-20% of meningococcal infection in children with fever and non-blanching rash.
Who were the patients?
The paper looked at children under 18 years old presenting to 37 Paediatric Emergency Departments in the UK over a 16 month period. Children were included if they had a fever (>38oC) and new onset of a non-blanching rash or features suggestive of meningococcal infection. Children were excluded if they had a pre-existing haematological condition or if they already had a diagnosis of Henoch-Schonlein purpura.
1513 patients were screened. 179 were excluded due to not meeting the criteria, not consenting, or a language barrier. Five that were enrolled had incomplete data leaving 1329 children were enrolled and included – the median age was 24 months, and 59% were male. Most children were vaccinated with 73% having had at least one dose of the Meningococcal B vaccine, and 77% having had at least one dose of the Meningococcal C vaccine.
What was the intervention?
There was no intervention here. Included patients were recruited at the point of meeting the criteria, using ‘recruitment prior to consent‘ and then consent was obtained soon after (usually within 24 hours). Data were collected contemporaneously: patient symptoms, blood test results, and treatment. A positive case was identified by being positive on PCR, or with a positive growth from another body sample (e.g. blood culture, or CSF). Patients were also checked for re-attendance to the hospital within 7 days. Results were also confirmed with the Public Health Agency – as meningococcal disease is a notifiable condition, this was a good method of picking up any missed cases.
What were the outcomes measured?
The primary outcome was assessing the performance of eight clinical guidelines on identifying children with invasive meningococcal disease (NICE meningitis (CG102); NICE sepsis (NG51); London; Chester; Bristol; Nottingham; Newcastle-Birmingham-Liverpool; and Glasgow).
The secondary outcomes were: performance of the eight guidelines in identifying children with other bacterial infections; and also looking at a cost comparison of each of the eight guidelines.
What were the results?
Of all 1334 children, 86% had a blood test and 45% had antibiotics. For patients admitted to hospital, the median length of stay was one night. 11 patients were admitted to PICU (2%) and two patients died (<1%).
Eight of these 11 PICU patients had N. meningitidis as did both of the patients who died. Seven patients had invasive bacterial infection (five with pneumococcal infection, one with E. Coli, and one with Group A Strep).
19 (1%) of patients in the cohort had meningococcal disease. 17 of these had N. meningitidis B, one had N. meningitidis C, and one had N. meningitidis W. Overall there were 26 patients (2%) with invasive bacterial infection (19 with meningococcal disease and 7 with an invasive bacterial infection).
346 patients (26%) did not have standard testing, and of these 19 patients (5%) had one unplanned re-attendance within seven days. However, none of these required readmission, antibiotics, or bacterial infection.
And how did the guidelines do?
All eight guidelines identified all of the 19 cases of meningococcal disease and all 26 cases of invasive bacterial infection (so the sensitivity of all of them is 100%). Specificity varied though. The NICE sepsis guideline stratified every patient as having a bacterial infection and therefore had a specificity of zero, making it the lowest specificity out of all the guidelines (closely followed by NICE meningitis guidelines with a specificity of 1%). This strategy clearly has cost implications too which is why the two NICE guidelines were also the most expensive per patient (£660.41 for the NICE sepsis guidelines).
Coming out top of the guideline ranking was the Barts Health NHS Trust guideline with a sensitivity of 100%, a specificity of 36%, and a cost of £490.29. This makes it the most accurate and also the cheapest.
Here’s the Barts Health NHS Trust guideline:
What about when we don’t follow the guidelines?
In practice, the guidelines were adhered to in 46% of the patients in the cohort. Deviation from guidelines resulted in fewer antibiotics being given. However, it also resulted in two patients being discharged with early meningococcal disease (they were subsequently treated and did not need PICU admission). Clinician decision-making increased the specificity (i.e. clinicians treated fewer people with antibiotics who didn’t have an invasive bacterial infection), but unfortunately reduced the sensitivity to 89%. Clinician decision-making did have the lowest cost per patient.
You’ve heard the facts, but how good was the paper?
As Ken Milne says…let’s get nerdy (and consider the CASP checklist for cohort studies)
Does the study address a clearly focused issue?
Yes.
Was the cohort recruited in an acceptable way?
Research without prior consent was used to avoid recruitment delaying any treatment plans. However consent was obtained as soon as possible after inclusion in the study (usually within 24 hours).
Was exposure accurately measured to minimise bias?
Yes. Objective measurements were used for a blood test and PCR results. Risk factors for meningococcal disease are subjective and were based on contemporaneous clinical assessment – but this is what we do in practice so is a good reflection.
Was the outcome accurately measured to minimise bias?
Yes. Note, however, that two patients with meningococcal disease were not included – one was not enrolled and the other was deemed by local staff to be inappropriate for inclusion.
Have the authors identified all-important confounding factors?
Yes.
Was the follow up of the subjects complete and accurate?
Yes, and also results were also checked with the Public Health Agency which would have allowed pick up of any missed meningococcal positive results.
What are the results?
There is a 1% prevalence of meningococcal disease in a mainly immunised population of children with fever and a non-blanching rash. The Barts Health NHS Trust guideline was the most accurate out of all the guidelines and with the lowest cost per patient.
Do you believe the results?
Yes.
Can the results be applied to a local population?
Yes. However, they would not be transferrable to populations with lower rates of vaccine uptake or a higher disease prevalence. The data was not shared on whether those with meningococcal disease were unimmunised or not, and therefore it would be prudent to be more cautious if your patient is not vaccinated.
Do the results fit with the other evidence available?
Previous data was from prior to the meningococcal vaccination so this is the first and largest study since then.
What did the authors conclude and what can we take away from this study?
Since the advent of a vaccination programme and increased vaccine uptake, the rates of meningococcal disease are lower. Although previous data suggested 10-20% of children with fever and a non-blanching rash had meningococcal disease, in fact this study shows that only 1% had meningococcal disease.
Using a cautious guideline like NICE results in a lower specificity and higher cost. Tailored guidelines can increase the specificity and reduce the cost per patient without compromising on 100% sensitivity. The Barts Health NHS Trust guideline was the top performing guideline.
And finally, a comment from the authors themselves:
From Tom Waterfield:
“The Petechiae in Children study represents the best available evidence regarding the assessment and management of febrile children with non-blanching rashes in the UK and clearly demonstrates that a lighter touch, tailored approach, is favourable to a test/treat all approach as currently advised by NICE. Moving to a tailored approach will reduce the need for invasive procedures, improve antimicrobial stewardship and save money.
In vaccinated populations where the prevalence of invasive meningococcal disease is low the presence of Petechiae alone should no longer be viewed as a red flag and should not be used to justify immediate treatment with broad spectrum antibiotics. The emphasis and teaching should shift away from worrying about all non-blanching rashes with greater emphasis on the importance of identifying purpuric rashes as they confer the greatest risk of invasive meningococcal disease.
Finally the PiC study demonstrates the importance of well designed prospective research studies in identifying risk factors for sepsis. Traditional approaches utilising retrospective reporting of symptoms from convenience samples of children with sepsis results in an over estimation the risks. This in turn leads to the development of overly aggressive clinical practice guidelines that are poorly adhered to. “
Note from Tessa: I am an employee of Barts Health but was not involved in the PiC study or in writing the Barts Health NHS Trust guideline.