Article Text

Download PDFPDF
Managing febrile neutropenia in the UK in 2020
  1. Sheena Guram1,
  2. Aditi Vedi2
  1. 1 Department of Paediatric Haematology and Oncology, University College London Hospitals NHS Foundation Trust, London, UK
  2. 2 Department of Paediatric Oncology, Cambridge University Hospitals NHS Foundation Trust, Cambridge, Cambridgeshire, UK
  1. Correspondence to Dr Sheena Guram, Department of Paediatric Haematology and Oncology, University College London Hospitals NHS Foundation Trust, London, London, UK; s.guram{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Neutropenic sepsis is a life-threatening medical emergency in paediatric haematology and oncology patients. It is a complication of myelosuppressive therapy used to treat children with cancer. Improvement in supportive care in oncology has seen the mortality from febrile neutropenia (FN) fall from 40% to 1%–3% in the last 50 years,1 with the risk of FN in children being reported as low as 0.4%–1%.2 However, many children will still undergo frequent and long inpatient admissions to the hospital, due to fever and neutropenia, while remaining clinically well with no identifiable source of infection. This new guidance draws on evidence from the Predicting Infectious Complications in Children with Cancer collaboration (PICNICC+) collaboration between UK, Australian and Swiss groups to propose a risk-stratified approach to managing these patients and ultimately reducing the time spent in the hospital.3 4 Clinicians and families may find this desirable, particularly during the COVID-19 pandemic while trying to reduce hospital time and keeping patients at home where possible.

Information about the current guideline

This guideline has been written by the Children’s Cancer and Leukaemia Group using/adapting the Australian low-risk FN programme.5 Principal treatment centres and their associated paediatric oncology shared care units should consider whether they wish to adopt the protocol as written or to adapt it to fit their needs. The reader should check whether their local hospital policies support ambulatory care management of patients with suspected FN who are at low risk of sepsis.

This guidance uses the validated Australia–UK–Swiss (‘AUS’) Score to risk stratify patients (see table 1) using three variables: chemotherapy intensity, total white cell count and platelets at admission.

Table 1

AUS Score risk stratification

Previous guidelines

There is no previous guidance in the UK for the risk stratification of FN in paediatric haematology/oncology patients. The current National Institute for Health and Care Excellence guidance (2012) defines FN as fever of ≥38°C and neutrophils of less than 0.5×109/L in a patient undergoing anticancer treatment and recommends that they are treated with initial intravenous piperacillin with tazobactam, if not penicillin allergic.6 Links to the current and previous guidelines can be found in box 1.

Box 1


Key issues


  • This is a clinical guideline for children who are clinically well with fever and risk of neutropenia. Children who are haemodynamically unstable or septic should be treated as per local protocol for neutropenic sepsis.

  • All children presenting with fever who are at risk of neutropenia should receive empirical broad-spectrum intravenous antibiotics within an hour of presentation without waiting for laboratory results, including the full blood count (FBC).

  • The AUS scoring system allows patients to be risk stratified based on their diagnosis, treatment regimen and FBC parameters. The patient’s risk group stipulates a minimum period of hospital observation prior to consideration of discharge on oral antibiotics (table 2).

  • C-reactive protein (CRP) is not a variable in the AUS Score and decisions to treat, admit or discharge home are safely made without assessing CRP.

Table 2

Period of observation for each AUS Score risk group

Antibiotic treatment

  • All children with presumed FN should initially be treated with empirical antibiotics (intravenous piperacillin with tazobactam) within 1 hour of presentation or development of symptoms on the ward/department.

  • This ideally should occur after blood cultures are drawn on presentation but without waiting for results of any blood tests, including FBC.

  • If first-line antibiotics are contraindicated, ciprofloxacin (or other locally agreed antibiotic combinations) can be used under guidance from the local microbiology team.

Location of treatment

  • The AUS Score, along with suitability for home/parent-led care, should be used to risk stratify the patients into very low-risk, low-risk, moderate-risk and higher-risk groups (see table 1).

  • Patients should be managed/observed in the hospital for the minimum period according to their risk score (see table 2).

  • Variations in minimal period of hospital observation in the very low-risk and low-risk groups may include an overnight stay for observation. In the moderate-risk and high-risk groups, this may include ongoing hospital care until afebrile for 24 hours (table 2).

  • In addition to low-risk stratification, the patients should meet the eligibility criteria for discharge home/parent-led care (see table 3).

  • Eligible patients should be discharged home on oral antibiotics if they can tolerate them and there are no concerns regarding absorption. Mucositis, vomiting and severe diarrhoea may impair absorption of oral antibiotics.

Table 3

Eligibility for care in the community (must answer yes to all of the criteria)

Duration of antibiotics

  • Oral antibiotics for discharge should be provided for 5 days with instructions to stop when clinically well, blood cultures are negative and the patient is apyrexial for greater than 24 hours.

  • A combination of oral ciprofloxacin and oral co-amoxiclav is a suggested antibiotic regimen; however, antibiotic choice should require discussion with the local microbiology team in each centre before implementing this guidance.


  • All patients should be followed up by daily telephone review by an experienced clinician until antibiotics have been stopped.

  • If the patient is still requiring antibiotics on day 5, they should be reviewed clinically.

  • The patient’s carer should check the child’s temperature every 4–6 hours while awake.

  • Children with bacteraemia should be readmitted for intravenous antibiotics, especially if they have a central venous access device in situ.

When should I bring patients back for medical review?

  • Ongoing fever >72 hours from presentation or if new fever occurs after being afebrile for 24 hours.

  • The patient feels unwell or has new symptoms.

  • The parents are concerned about their child.

  • Positive blood culture.

  • Severe or persistent pain.

  • Chills/rigours/shaking.

  • Poor oral intake or concerns about dehydration.

When should I readmit patients for intravenous antibiotic therapy?

  • Fever of ≥38°C on day 5.

  • Clinically unwell.

  • Infection requiring inpatient care.

What evidence is there to support this guideline?

The Swiss Paediatric Oncology Group (SPOG) undertook a prospective multicentre study to derive a score to predict the risk of adverse events in paediatric oncology patients with FN.3 This score used four variables: chemotherapy intensity, haemoglobin level, leucocyte and platelet count at presentation. Their study found that this scoring system was accurate in predicting adverse events in FN.

The Australian–PICNICC+ study was a prospective multicentre study which identified a bloodstream infection in 111 out of 858 FN events (13%).7 Using the data from the Australian study, SPOG rule has been modified to produce the ‘AUS Score’, which uses three variables: white cell count, platelet count and chemotherapy intensity to produce a score from 0 to 3 (see table 2). The AUS Score has been validated using the Australian–PICNICC+ datasets, which include 1500 FN episodes from the UK, Europe, North and South America. These data have yet to be published.

What should I start doing?

  • Consider use of this validated risk score (see table 2) to risk stratify children with fever and risk of neutropenia who are clinically stable at presentation.

  • Consider outpatient or ambulatory reduced-intensity treatment where appropriate.

  • Educate patients, families and clinicians and ensure appropriate follow-up measures.

What should I continue to do?

  • Children presenting with fever and risk of neutropenia should be promptly assessed and empirical intravenous antibiotics should be administered within 1 hour without waiting for the FBC result.

  • Treat clinically unwell patients using local sepsis management guidelines and broad-spectrum antimicrobials as indicated.

What should I stop doing?

  • CRP has not been shown to be a useful indicator of infection in oncology patients8 and is not validated in this subgroup of children. It is not used to calculate the AUS Score and should not be used to assess these children.

Take home messages

  • Risk stratification of febrile neutropenic patients who are haemodynamically stable at presentation is inherently safe and supported for use by the Children’s Cancer and Leukaemia Group Supportive Care Group.

  • It is important to be aware of this guidance, but local hospital policies must be used. The reader should check whether their guidance is supported by their local principal treatment centre and paediatric oncology shared care unit.

  • Clinical review is paramount and clinicians should feel confident that their patient is clinically stable, appropriate for home/parent-led care and able to tolerate oral antibiotics before discharge home.

  • Avoidance of unnecessary, frequent and lengthy hospital admissions (particularly in this patient group) is important to improve quality of life and to reduce burden on hospital beds and healthcare costs.

  • This may be particularly desirable during the COVID-19 pandemic to keep immunocompromised patients who are clinically well away from the hospital.

Ethics statements

Patient consent for publication



  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.