The NCAG recommendations set a benchmark for maintaining existing cancer services and for development and improvement where required. Neutropenic sepsis is one of the key areas highlighted and cancer networks must work to provide an acute oncology service with standardised protocols and policies, ensuring that patients receive antibiotics within one hour of presentation (door-to-needle; NCAG, 2009
). Our study demonstrates that within SWLCN, chemotherapy-induced febrile neutropenia was generally recognised early and managed appropriately, however there remains room for improvement. Timely administration of appropriate antibiotics within 60
min requires particular attention. The seriousness of the condition is underlined by the fact that three patients died as a direct consequence of neutropenic sepsis, translating into a mortality rate of 4.2% for our series. This is in keeping with reported mortality rates of between 2 and 10% (Kuderer et al, 2006
; Klastersky and Paesmans, 2007
The NCEPOD report (NCEPOD, 2008
) revealed similar problems in delayed administration of antibiotics, the use of incorrect antibiotics and delay in both diagnosis and senior staff review. Recommendations include access to a local policy and clinical care pathway, management by experienced staff, as well as close liaison with A and E departments. It was apparent in our study that there was no common protocol that existed between the hospitals of the network, and this may have contributed towards delayed management and use of antibiotics, but also accounts for the variation of diagnostics undertaken. The National Comprehensive Cancer Network (NCCN) guidelines (Aapro et al, 2006
; Smith et al, 2006
) recommend routine blood tests and blood cultures for all patients, but other investigations such as chest x-rays, dip-stick urinalysis, urine and stool cultures, are left at the discretion of the assessing clinician. As time is a highly significant factor in the successful management of neutropenic sepsis, a network-based protocol would aid in expediting assessment and early treatment.
Almost all patients included in our study had well-established risk factors for the development of febrile neutropenia. American and European guidelines (Aapro et al, 2006
; Smith et al, 2006
) acknowledge the following factors as significant for consideration of prophylactic administration of haematopoietic growth factors: (1) chemotherapy regimens with a risk of >20% for the development of febrile neutropenia or (2) regimens with a risk between 10 and 20% in conjunction with certain patient-related risk factors such as stage of disease and age. Of our patients that fulfilled either of these criteria, only 21% and 43%, respectively, received G-CSF (). Primary prophylactic G-CSF can reduce the risk of febrile neutropenia by 46% and infection-related mortality by 45%, whilst also allowing improved relative dose intensity of chemotherapy (Kuderer et al, 2007
). In all, 50% of all episodes of febrile neutropenia occurred at or near the start of a chemotherapy course (cycles 1 and 2). This is in keeping with previous reports (Lyman, 2005
), which led to the recommendation for starting prophylactic measures ‘upfront' in appropriate risk groups (Aapro et al, 2006
; Smith et al, 2006
; NCCN, 2009
). Identification of patients who will benefit from G-CSF and improving access to these agents is an important process before the commencement of chemotherapy.
Risk stratification according to ASCO/EORTC guidelines and extent of prophylactic administration of G-CSF
Of note, one-third of patients with febrile neutropenia in our study were older than 65 years, and this may support the notion that increasing age is an independent predictor of the development of febrile neutropenia (Crawford et al, 2004
; Aapro et al, 2006
). There is sufficient evidence that prophylactic G-CSF reduces the incidence of chemotherapy-induced neutropenia, febrile neutropenia and infections in elderly patients (Repetto et al, 2003
). Increased use of growth factors may be particularly relevant in the elderly, in order to optimise their cancer care and quality of life. Older patients are frequently under-treated because of misperceptions of their frailty; however dose reductions may compromise treatment efficacy (Pietropaolo et al, 2003
; Lyman et al, 2004
). Growth factors may help these management dilemmas.
Use of G-CSF in treatment of febrile neutropenia is more controversial and although 63% of our patients received G-CSF as secondary treatment, this may not necessarily be representative of other networks. The ASCO and EORTC guidelines recommend colony-stimulating factors as treatment only for patients who are at high risk for infection-associated complications. A meta-analysis (Clark et al, 2005
) showed a significant reduction in the length of hospitalisation (hazard ratio: 0.63, 95% CI: 0.49–0.82, P
=0.0006) with the use of G-CSF, but only a marginally significant decrease in infection-related mortality (odds ratio: 0.51, 95% CI: 0.26–1.00, P
=0.05) and no significant reduction in overall mortality.
Primary antibiotic prophylaxis was used in 25% of cases in our study and its efficacy in reducing infections, febrile neutropenia and hospital admissions has been shown (Engels et al, 1998
; Cullen et al, 2005
). However there are valid concerns pertaining to possible increases in antimicrobial-resistant strains (Goossens et al, 2005
), and widespread use is therefore discouraged for standard chemotherapy regimens in most solid tumours. Use may be considered for intermediate risk groups, such as lymphoma patients and treatments in which prolonged (7–10 days) neutropenia is anticipated (Cullen et al, 2007
; NCCN, 2008
It is important to recognise that prevention of neutropenic sepsis and its complications begins with appropriate, considered commencement of chemotherapy. Of particular concern from the NCEPOD was the finding that in 19% of cases, the decision to give chemotherapy was inappropriate, mostly on the grounds of progressive disease and abnormal blood results. Furthermore, in 27% of cases, chemotherapy was judged to have hastened or caused death, with the peak of deaths occurring 11–15 days after chemotherapy and likely related to the development of neutropenic sepsis (NCEPOD, 2008
). For our data series, there was one case (1.4%) of death within 30 days of commencing chemotherapy, and with hindsight perhaps the use of second-line chemotherapy in this patient with metastatic SCLC was inappropriate. The NCAG recommendations provide a framework for a process of assessment, decision to treat, informed consent and prescription of chemotherapy, and emphasise the importance of detailed standardised consent forms and the involvement of senior trained oncology medical staff (NCAG, 2009
). The evaluation of these aspects was beyond the scope of this study but will be an area of focus within the network.
Our study was also limited in being able to capture patients receiving chemotherapy at SWLCN hospitals but who presented with febrile neutropenia at hospitals outside the network. The NCEPOD reported that 15% of admissions were to hospitals other than the one where chemotherapy had been administered (NCEPOD, 2008
). Peripheral hospitals may not be equipped with 24-h oncology services and similar levels of experience and it is therefore even more crucial for these sites to have access to well-designed protocols and to expert consultant advice. As outlined by NCAG, patient education regarding what to do in the event of a chemotherapy-induced complication is fundamental to ensuring they receive prompt appropriate care. Evaluating the efficacy of current patient education processes was also beyond the scope of this study, and in particular, it will be useful to assess the use of pre-designed ‘neutropenic sepsis' cards which patients present to their local A and E in order to expedite appropriate management.
All patients in our study were treated with intravenous antibiotics and were admitted for inpatient care. Recent advances in the management of febrile neutropenia have highlighted the value of risk stratification and the evolving role of oral antibiotics with early hospital discharge in low-risk patients (Freifeld et al, 1999
; Kern et al, 1999
; Hughes et al, 2002
). The Multinational Association for Supportive Care in Cancer has developed a risk index for the development of febrile neutropenia (Klastersky and Paesmans, 2007
) and studies have shown that outpatient oral antibiotics are feasible in low-risk patients (Innes et al, 2003
; Vidal et al, 2004
; Klastersky and Paesmans, 2007
). To our knowledge, this stratification is not routinely used in the UK (Innes et al, 2005
) but will be an increasing area of interest, with important implications on decreased inpatient costs and improved patient convenience with outpatient care.
Undoubtedly there remain several aspects of cancer care that will require ongoing audit and review, and our own study highlights the areas that interact and affect the quality of care for one single chemotherapy-induced complication – febrile neutropenia. Our study has provided an important and detailed insight into the incidence and management of chemotherapy-induced febrile neutropenia in a representative cancer network in the United Kingdom. Neutropenic complications in cancer patients are associated with substantial morbidity, mortality and healthcare costs, warranting research and audit in this area.
We support NCAG's initiative to encourage chemotherapy services to undertake regular self-assessment procedures to identify gaps in their performance and take urgent steps to improve quality of care. Their recommendations in combination with the NCEPOD data have aided SWLCN in analysing the results of this study and putting them in perspective. Areas of potential improvement are currently being addressed for the hospitals within our cancer network, based on the results from this study and on the recommendations set out by NCAG (see ). Improved uptake of prophylactic G-CSF according to ASCO and ESMO guidelines (particularly in patients above the age of 65), could reduce the risks of morbidity and mortality associated with neutropenic sepsis. Time intervals from arrival to completion of assessment and commencement of treatment for febrile neutropenia need improvement, and to achieve this, physician and nursing protocols to standardise and streamline clinical care pathways for the whole network are under consideration. It is hoped that the recommendation for NICE to provide a nationwide policy for management of neutropenic sepsis will lead to a standardised approach within and across networks. A dedicated specialist committee should review all clinical cases resulting in death following admission with febrile neutropenia. Finally, continued audit of current practices should be ongoing, in order to identify and rectify weaknesses whilst simultaneously consolidating current quality care.
Key areas for improvement and action points for SWLCN