• febrile
• neutropenia
• paediatric oncology

Serious infection affects around 20% of patients who present febrile and neutropenic.1 Febrile neutropenia (FN) is thus a well-recognised medical emergency. Addressing the challenges of both rapid delivery of antibiotics to patients at risk of life-threatening sepsis and appropriate antibiotic stewardship to reduce antibiotic overuse continues to keep this a priority for clinicians, patients and parents. Audits of FN practice have previously demonstrated variation in definitions for FN and its management.1 ,2 In September 2012, the National Institute for Health and Care Excellence (NICE) published the ‘Neutropenic sepsis: prevention and management of neutropenic sepsis in cancer patients’3 guideline and we sought to assess if this has reduced variations in practice.

This reaudit included 45 UK centres (14 children's cancer and leukaemia group principal treatment centres). The centres completed questionnaires regarding their FN policy and any admissions over a 2-week period in March 2015. One hundred and eight episodes from 30 centres were recorded ranging from 0 to 10 episodes per centre.

The results show continued variability in definition and management. The NICE suggested definition of FN (temperatures >38°C and neutrophil count of <0.5) is used by 21/43 (49%) centres (see table 1). Only 19/45 (42%) centres use risk stratification at admission (a modified Alexander approach in 16 centres, as suggested by NICE, with one centre using the Swiss Paediatric Oncology Group model4 and two setting up their own stratification system).

The NICE guideline suggests patients presenting with FN should have central and peripheral blood cultures and blood tests including C-reactive protein and lactate. Central blood cultures were sent in over 90% of episodes and 19% of episodes had peripheral blood cultures performed but only two centres routinely performed paired blood cultures. CRP was analysed in 83% of episodes but lactate in only 15% of admissions with six centres routinely testing lactate on presentation.

Thirty-seven per cent of centres use single agent piperacillin/tazobactam as initial antibiotic treatment, 34% use piperacillin/tazobactam with an aminoglycoside and 13% meropenem alone. Overall, 62% of inpatients received their antibiotics within 60 min of admission or febrile episode. Nine of 17 centres with multiple admissions over the 2-week period administered the first dose of antibiotic within 60 min in ≥80% of occasions.

Forty-four (40%) episodes were assessed as low risk at 48 h with 28 (64%) switching to oral antibiotics and 16 (36%) continuing intravenous therapy. Antibiotics were stopped upon resolution of fever irrespective of neutrophil count in 75%, while 14% appropriately continued antibiotics to treat a confirmed infection. In 11% of episodes, the reason for continuation of antibiotics was unclear.

The results show continuing variation in practice and a lack of adherence to NICE guidance. The variance has not significantly changed since the previous audit in 2012. Participating centres have been sent the results plus centre-specific analysis to see their results in comparison with national practice and guidelines. A reaudit is planned for 2016. The adoption of a national policy on neutropenic sepsis management based on the best available evidence of clinical effectiveness and cost-effectiveness requires local implementation but should improve outcomes.