Fever in under 5s

Surv Surveillance eillance report 2016 – F Fe ever in under 5s (2013) NICE guideline CG160 Surveillance report Published: 24 April 2017 nice.org.uk

© NICE 2017. All rights reserved.

Surveillance report 2016 – Fever in under 5s (2013) NICE guideline CG160

Contents Surveillance decision ..................................................................................................................................................

3

Reason for the decision ..........................................................................................................................................................

3

Commentary on selected new evidence .............................................................................................................

6

Management by the paediatric specialist........................................................................................................................

6

Antipyretic interventions ......................................................................................................................................................

11

How we made the decision.......................................................................................................................................

16

New evidence .............................................................................................................................................................................

16

Views of topic experts.............................................................................................................................................................

16

Views of stakeholders .............................................................................................................................................................

16

NICE Surveillance programme project team .................................................................................................................

18


Page 2 of 18


Surv Surveillance eillance decision We will not update the guideline on fever in under 5s at this time. During surveillance, editorial or factual corrections were identified: We will add a footnote to recommendation 1.2.2.10 of the guideline, as well as the traffic light system for identifying risk of serious illness, to highlight that some vaccinations have been found to induce fever in children younger than 3 months. We will add a recommendation that cross-refers to the NICE guideline on sepsis: recognition, diagnosis and early management (NG51) so that clinicians can determine what considerations should be made, and what diagnostic tests should be performed, if they suspect that a febrile child has sepsis. We will add a recommendation to the non-paediatric section of the guideline highlighting that clinicians should not rely solely on a response to antipyretics to differentiate between serious and non-serious illness.

Reason for the decision We found 41 new studies through surveillance of this guideline. This included new evidence on symptoms and signs of specific illnesses that is consistent with current recommendations. We also identified new evidence or information in the following areas that was inconsistent with, or not covered by, current recommendations. We considered these areas in detail to determine the most appropriate action:

Clinical assessment of children with fe fevver Topic experts highlighted that a new protein-based meningococcal B vaccination induced fever in children less than 3-months old. Intelligence gathering revealed a Public Health England publication which highlights that fever is an adverse reaction associated with a new protein-based meningococcal B vaccine. Furthermore, an NHS England publication targeted at parents highlights that fever is a common side effect after some vaccinations. Topic experts pointed out that recommendation 1.2.2.10, in NICE guideline CG160 states that clinicians should recognise that children younger than 3 months with a temperature of 38°C or higher are in a high-risk group for


Page 3 of 18


serious illness. They expressed some concern that many children under 3 months who develop fever after immunisation will be admitted to hospital and subjected to a full septic screen because the guideline suggests that they are treated this way. Overall, topic experts felt that the addition of a footnote to the guideline recommendations would not address this issue. This feedback was considered in detail. No evidence was identified through surveillance that included children post vaccine aiming to identify the likelihood of serious bacterial infection in the presence of a postvaccine fever. Therefore, adding a footnote to NICE guideline CG160 to notify clinicians that certain vaccinations can induce fever in children younger than 3 months was considered to be the most appropriate action.

Management b byy the paediatric pr practitioner actitioner Topic experts highlighted that recommendations about lactate testing are in the NICE sepsis guideline (NG51) but not in NICE guideline CG160, and there is much cross-over between the 2 guidelines. They felt that although lactate testing is used to stratify risk in people who are thought to have sepsis, it is not used for detecting infection. Experts highlighted that children with sepsis comprise a subset of children presenting with fever. Furthermore, they pointed out that most trusts will screen for sepsis first, then fever, unless they are assured that the initial screening process in the fever guideline covers sepsis red flags. As a result, experts felt it is important that NICE guideline CG160 redirects clinicians to NICE guideline NG51 if they suspect a febrile child has sepsis. They suggested that a recommendation could be added to NICE guideline CG160; crossreferring to NICE guideline NG51 so that clinicians can determine what considerations should be made if they suspect that a febrile child has sepsis. Experts suggested the following recommendation could be added after 1.2.1.1 of NICE guideline CG160: "Think 'could this be sepsis?' and refer to the NICE sepsis guidance if a child presents with fever and symptoms or signs that indicate possible sepsis."

Management b byy the non-paediatric pr practitioner actitioner Although no evidence was found at any surveillance time point, substantive feedback from topic experts indicated that the guideline should be amended. Experts highlighted a coroner's report that indicated that the guideline provides advice for paediatric specialists (section 1.5) on assessment of children who respond, or fail to respond, to antipyretic therapy but no recommendations are provided in the advice for non-paediatric practitioners section (section 1.4). As a result, experts suggested that the section on management by non-paediatric practitioners should be amended to include a recommendation highlighting that when a child has been given antipyretics, clinicians should not rely on a decrease or lack of decrease in temperature to differentiate between serious


Page 4 of 18


and non-serious illness. Experts suggested that the following recommendation could be added to section 1.4: "When a child has been given antipyretics, do not rely on a decrease or lack of decrease in temperature to differentiate between serious and non-serious illness."

Other clinical areas We did not find any new evidence in areas not covered by the original guideline.

Equalities No equalities issues were identified during the surveillance process.

Ov Over erall all decision After considering all the new evidence and views of topic experts, we decided that an update is not necessary for this guideline. See how we made the decision for further information.


Page 5 of 18


Commentary on selected new e evidence vidence With advice from topic experts we selected 2 studies for further commentary.

Management by the paediatric specialist We selected the large prospective cohort study by Milcent et al. (2016) for full commentary because it assesses the diagnostic value of serum procalcitonin assays for detecting serious illness in febrile children under 5 years old.

What the guideline recommends Currently NICE guideline CG160 does not include procalcitonin testing in the list of laboratory investigations that could be used to detect serious illness in febrile children under 5 years old.

Methods The prospective cohort study by Milcent et al. (2016) assessed the diagnostic characteristics of procalcitonin assays for detecting serious bacterial infection and invasive bacterial infection in 2,047 infants aged between 7 and 91 days old. Investigators recruited infants from 15 French paediatric emergency departments between October 2008 and March 2011. Infants between 7 and 9 days old, with temperatures above 38°C, were included. None of the infants received antibiotics within 48 hours of presentation at the emergency department. Furthermore, all infants did not have any major comorbidities; such as, immunodeficiency, congenital abnormality, or chronic disease. Infants less than 7 days old were excluded because of evidence that they were likely to have early-onset sepsis related to perinatal factors. Furthermore, researchers justified excluding children under 7 days old by citing additional evidence that physiologic procalcitonin concentrations were higher during the first 3 days of life than after. Upon presentation at the emergency department, attending physicians classified children as well, minimally, moderately, or very ill. After the initial clinical examination, white blood cell, absolute neutrophil cell, and C-reactive protein were taken. Blood culture, urinanalysis, lumbar puncture, stool culture, and chest radiography, were also performed. Clinicians then used clinical criteria (symptoms and laboratory test results) to diagnose children as having serious bacterial infection or no infection. Serum samples were collected after the initial clinical examination for quantitative procalcitonin assays, which were performed retrospectively. Clinicians were blinded to procalcitonin test results at the time of diagnosing infants. Furthermore, laboratory assessors were


Page 6 of 18


blinded to clinical features and other laboratory test results. Serious bacterial infection was classified as the isolation of a bacterial pathogen from the blood culture, cerebrospinal fluid, a stool sample, or urine collected by catheterisation (greater than 50,000 colony-forming units per ml of a single pathogen) accompanied by pyuria (greater than 5 white blood cells per high-power field) and/or bacteriuria on microscopic analysis or a positive dipstick test result for leukocyte esterase or nitrite. Invasive bacterial infections were defined as pathogenic bacteria (including Streptococcus pneumonia, Neisseria meningitidis, groups A and B Streptococci, Staphylococcus aureus, and Escherichia coli. The diagnostic performance of the procalcitonin assay for detecting definite serious bacterial infections and invasive bacterial infections were investigated comparing areas under Receiver Operating Characteristic (ROC) curves, sensitivity and specificity, and likelihood ratios at various cut off values.

Results Of the 2,047 febrile children assessed in the study, diagnostic tests revealed that 1,691 infants had no infection, 217 had a possible serious blood infection and 139 had a definite serious blood infection. Of the 217 children with a possible serious blood infection, 173 (79.7%) had a urinary tract infection and 44 (20.3%) had pneumonia. Of the 139 children with a definite serious blood infection, 118 (84.9%) did not have an invasive bacterial infection: 115 of these patients had a urinary tract infection and 3 had gastroenteritis. Invasive bacterial infection was reported in 21 patients with serious blood infection: 8 had bacterial meningitis and 18 had bacteraemia.

Ar Areas eas under R ROC OC curv curves es When detecting serious bacterial infections, the area under the ROC curve for procalcitonin assays (area, 0.81; 95% CI, 0.75 to 0.86) was significantly greater than the area under the curve for white blood cell counts (area, 0.66; 95% confidence interval [CI], 0.58 to 0.73; p