Use of risk stratification to guide ambulatory management of neutropenic fever

Authors


  • Funding: These guidelines were developed independently through funding from Victorian Integrated Cancer Services. The Western and Central Melbourne Integrated Cancer Service (WCMICS), one of three metropolitan and five regional Integrated Cancer Services funded by the Victorian Department of Human Services to implement the Victorian Government's Cancer Services Framework across Victoria, administered this funding.

  • Conflict of interest: The following working group members are consultants or advisory committee members or receive honoraria, fees for service, or travel assistance (independent of research-related meetings) form; or have research or other associations with the organizations listed: Ian Davis – Novartis, Pfizer, GlaxoSmithKline, Schering-Plough; Andrew Grigg – Amgen, Gilead, Merck Sharp & Dohme, Novartis, Pfizer, Schering Plough, Sue Kirsa – Roche, Schering Plough, Amgen, Hospira, Alphapharm; Senthil Lingaratnam – Victorian Integrated Cancer Services; Orla Morrissey – Gilead, Merck Sharp & Dohme, Orphan, Pfizer, Schering Plough; William Renwick – Roche; Monica Slavin – Gilead, Merck Sharp & Dohme, Pfizer, Schering Plough; Brian Stein – Merck Serono; Constantine Tam – GSK, Roche; Karin Thursky – Gilead, Merck Sharp & Dohme, Pfizer, Schering Plough; Andrew Wei – Celgene, Hospira, Novartis.

Leon Worth, Department of Infectious Diseases, Peter MacCallum Cancer Centre, St. Andrew's Place, East Melbourne, Vic. 3002, Australia. Email: Leon.Worth@petermac.org

Abstract

Utilization of risk-stratification tools in the setting of neutropenic fever is currently limited by inadequate knowledge and lack of awareness. Within this context, the approach to management of low-risk patients with neutropenic fever is inconsistent with the available evidence across many Australian treating centres. These clinical guidelines define and clarify an accepted standard of care for this patient group given the current evidence base. The Multinational Association for Supportive Care in Cancer risk index is presented as the preferred risk assessment tool for determining patient risk. Suitability of ambulatory care within specific patient populations is discussed, with defined eligibility criteria provided to guide clinical decision-making. Detailed recommendations for implementing appropriate ambulatory strategies, such as early discharge and outpatient antibiotic therapy, are also provided. Due consideration is given to infrastructural requirements and other supportive measures at a resourcing and operational level. An analysis of the relevant health economics is also presented.

Background

Components of a clinical survey administered prior to the development of these guidelines explored risk stratification and current management of neutropenic fever in the ambulatory setting. Key results are published in an adjunct paper.1 The survey findings suggest that utilization of risk-stratification tools in the setting of neutropenic fever is currently limited by inadequate knowledge and lack of awareness, rather than a reluctance to apply them in a clinical context. Indeed, when surveyed physicians were provided with a specific haematology or oncology case scenario describing a patient with neutropenic fever at low risk (according to the Multinational Association for Supportive Care in Cancer (MASCC) risk index2) and suitable for administration of oral antibiotic therapy, only 60% of responding medical oncologists and 47.0% of responding haematologists believed that the case represented a patient at low risk. Further, the majority of respondents (86% of responding medical oncologists and 90% of responding haematologists) indicated that they would prescribe parenteral antibiotics for the low-risk patient described. These findings suggest that risk stratification of patients with neutropenic fever in Australia is currently not being performed according to accepted standardized criteria (e.g. MASCC risk index) and that the approach to management of low-risk patients with neutropenic fever is not consistent with available evidence. The following recommendations (see Table 1 for summary) aim to define and clarify an accepted standard of care for this patient group, and should be used in conjunction with the empiric therapy guidelines as published by Tam et al.3

Table 1.  Key practice points – risk assessment and ambulatory care
  1. AML, acute myeloid leukaemia; MASCC, Multinational Association for Supportive Care in Cancer; MRSA, methicillin-resistant Staphylococcus aureus; VRE, vancomycin-resistant enterococcal.

Assessing patient risk for medical complications in setting of neutropenic fever:
 • Patients who are clinically stable at onset of neutropenic fever should be assessed as low or high risk for developing medical complications using the MASCC risk index (grade B)
Suitability of ambulatory care for specific patient populations:
 • All patients with solid tumours or lymphoma receiving chemotherapy regimens that are expected to cause less than 7 days of neutropenia (e.g. R-CHOP-14/21) can be safely targeted for ambulatory management (grade A)
 • There is no evidence to support an ambulatory strategy in patients with high-grade lymphomas requiring high-intensity chemotherapy for example (R)CODOX-M/(R)IVAC
 • If suitable infrastructure is available (see in-text), patients with AML may be considered for inclusion to an ambulatory model of care; otherwise ambulatory management is not recommended for patients with haematological malignancies other than low- to intermediate-grade lymphomas.
 • Patients previously colonized, or colonized, with MRSA, VRE or multi-drug-resistant Gram-negative isolates should not be managed as outpatients, unless infection with these organisms has first been excluded and they are otherwise deemed ‘low-risk’ (expert opinion)
Appropriate ambulatory strategies
 • Institutional requirements: suitability of an ambulatory care-based approach will depend heavily upon the supportive infrastructure in place at a particular facility and the availability of relevant clinical services (detailed in-text)
 • Early discharge: a period of inpatient observation: 24 h for patients with solid tumour or low- to intermediate-grade lymphomas (grade B) , 48 h for patients with other haematological malignancies is recommended followed by early discharge and structured follow up (grade D)
 • Antibiotic therapy: dual therapy with amoxicillin-clavulanate and ciprofloxacin is recommended for the ambulatory care of patients without known drug allergy to these agents (grade A); low-risk haematology patients (other than low- to intermediate-grade lymphoma) should receive IV antibiotics for 48 h as an inpatient, prior to a switch to oral therapy (see in-text for caveats) (grade D)
 • Failure of ambulatory therapy: if fever recurs, inpatient management should be commenced rather than seeking an alternative oral antibiotic regimen

Recommendations for risk assessment and ambulatory care

Assessing patient risk for medical complications within the setting of neutropenic fever

The risk of a patient with neutropenic fever experiencing medical complications should be assessed using an accepted risk assessment tool. Patients who are clinically stable at onset of neutropenic fever should be assessed as low or high risk for developing medical complications using the MASCC risk index (grade B recommendation) (Fig. 1).2 The ‘Talcott’ and ‘modified Talcott’ approaches are not as rigorous, and have not been used as widely as the MASCC assessment strategy (Table 2), which has been validated in both patients with solid tumours and patients with haematological malignancy (level III evidence).4–6

Figure 1.

Management of clinically stable patients with neutropenic fever (detailed in Tam et al.3). MASCC, Multinational Association for Supportive Care in Cancer.

Table 2.  MASCC index score for identifying low-risk patients with neutropenic fever2
CharacteristicPoint score
  • The maximum value in this system is 26. A score of ≥21 predicts a ≤6% risk for severe complications and a very low mortality (<1%) in neutropenic febrile patients.

  • ‡Previous fungal infection: means demonstrated fungal infection or empirically treated suspected fungal infection.

  • §

    §Outpatient status: means onset of fever as an outpatient. MASCC, Multinational Association for Supportive Care in Cancer.

Burden of illness 
 No or mild symptoms5
 Moderate symptoms3
No hypotension5
No chronic obstructive pulmonary disease4
Solid tumour or no previous fungal infection4
Outpatient status§3
No dehydration3
Aged < 60 years2

Identifying patients at low risk

In addition to using the MASCC risk index, key clinical criteria for identifying patients with neutropenic fever at ‘low risk’ of medical complications, as cited in the available literature, are presented in Table 3.2,7–9

Table 3.  Clinical identification of a patient with neutropenic fever at ‘low risk’ of complications7,8,10
Appears well
Normal blood pressure
No dyspnoea/hypoxia
Eating and drinking well
No signs of focal infection

Risk and suitability of ambulatory care within specific patient populations

Patients with solid tumour

In general, most patients with solid tumours have an expected duration of neutropenia less than 7 days following chemotherapy, and are therefore considered low-risk in the setting of neutropenic fever and can be safely targeted for ambulatory management (level II evidence; grade A recommendation).9,11 However, higher-intensity chemotherapy or cumulative immunosuppressive effect of prior therapies for some solid tumours (e.g. advanced germ cell tumours, osteosarcoma and primitive neuroectodermal tumours) may warrant inpatient therapy for neutropenic fever. In the absence of specific data for these subpopulations, assessment of the expected degree and duration of neutropenia is advised, before classifying as low-risk. Although no published Australian data validate this strategy, the population of patients with solid tumours in Australia is not significantly different to many other international centres in terms of subclassification and stage of underlying malignancy, intensity of chemotherapy regimens, comorbidities and social milieu. Therefore, findings from such centres2,8,9,11 may be directly applied to Australian patients with solid tumours.

Patients with haematological malignancy

Lymphoma.  Patients with lymphoma receiving chemotherapy regimens that are likely to cause less than 7 days of neutropenia (e.g. R-CHOP-14) can be safely targeted for ambulatory management (level II evidence; grade A recommendation).9,11 However, there is no evidence to support this strategy in patients with high-grade lymphomas requiring high-intensity chemotherapy. Table 4 categorizes common chemotherapy regimens by risk, according to expert opinion.

Table 4.  Examples of common chemotherapy regimens for patients with lymphoma that may be suitable for ambulatory management
High-risk chemotherapy regimensLow- to intermediate-risk chemotherapy regimens
  • Individual institutions may include other low- to intermediate-risk chemotherapy regimens associated with higher rates of febrile neutropenia, according to their clinical resource capabilities.

(R)CODOX-M/(R)IVACR-CVP
(R)Hyper-CVAD/MTX-Ara-C(R)-CHOP-21
(R)IVAC(R)-CHOP-14
BEACOPP (esc)ABVD

Haematological malignancies other than low and intermediate grades of lymphoma.  There is limited evidence to support ambulatory management of patients with haematological malignancies other than low and intermediate grades of lymphoma, although one study did report positive outcomes with ambulatory care in a group of patients with acute myeloid leukaemia (AML).12 However, as chemotherapy dosing used in Australia may result in administration of higher-intensity regimens than those used in this particular study, the consensus group has reservations about extrapolating the study's findings to this patient population in Australia. Given that patients with AML may also require comprehensive monitoring and that outpatient care can differ substantially between centres, the decision to adopt an ambulatory approach should be informed by an institution's assessment of the overall risk to patient safety. Within this context, if suitable infrastructure is available (see later discussion), this patient population may be considered for inclusion to an ambulatory model of care at individual healthcare facilities (grade D recommendation).

Patients colonized with a multi-resistant organism

Any patient who is colonized with a multi-resistant organism (MRO) is at risk of developing infection with that same organism.13,14 Historically, this has been most widely reported with regard to methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococcal (VRE) isolates.15,16 The relationship between previous colonization and development of invasive infection may be more evident in high-risk rather than low-risk patient populations – for example, patients with AML and stem cell transplant patients.17–21 These patients often have other contributing risk factors for the development of invasive infection with MROs, such as the presence of a central venous catheter or higher grades of mucositis.

The relative prevalence of these organisms varies across different regions/states of Australia,14 between different healthcare facilities, and within individual hospital units.22 If a healthcare facility has detected (or screened) for an MRO, then monitoring and investigation must be performed to ensure that the patient is managed accountably and appropriate measures have been taken to exclude infection during an episode of neutropenic fever. As oral antibiotic regimens used for neutropenic fever (such as ciprofloxacin plus amoxicillin-clavulanate) do not have sufficient activity against many resistant organisms,23 we have adopted a conservative approach and recommend that patients previously colonized, or colonized, with MRSA, VRE or multi-drug-resistant Gram-negative isolates should not be managed as outpatients, unless infection with these organisms has first been excluded (e.g. blood and urine cultures) and they are otherwise deemed ‘low-risk’ (grade D recommendation).24

Appropriate ambulatory strategies

The consensus group support a considered approach to ambulatory care for low-risk patients and present the following recommendations as the accepted standard of care.

Early discharge

Experienced nursing and/or medical staff should assess all patients presenting with neutropenic fever and perform a full blood examination and septic work-up at the time of initial assessment. While there is evidence to support immediate outpatient treatment, predominantly with oral antibiotics, for patients with solid tumours and low- to intermediate-grade lymphomas (grade B recommendation),4,25–31 the supporting data for patients with other haematological malignancies are much fewer. Two haematology studies were identified, one administering parenteral antibiotics for at least 24 h prior to early discharge, the other for at least 48 h.24,32

Given the current diversity of available outpatient care, a universal recommendation to immediately discharge all low-risk patients with neutropenic fever is likely to represent an unacceptable risk, as robust outpatient support systems cannot be guaranteed. Thus, from the perspective of patient safety, a period of inpatient observation of at least 24 h for patients with solid tumour or low- to intermediate-grade lymphomas is recommended followed by early discharge (grade B recommendation).5,28,33 For patients with other haematological malignancy, a period of observation of at least 48 h is recommended followed by early discharge, if appropriate (grade D recommendation).24,32 This approach is consistent with published literature5,32–34 and will also enable the treating team to review preliminary culture results (blood, urine, etc.) before discharge. Note, however, that clinical studies using standardized risk assessment tools have not necessarily excluded patients with positive blood cultures.5,32,35 Additional eligibility criteria for early discharge and ambulatory care are described in Table 5.

Table 5.  Eligibility and exclusion criteria for ambulatory care of a patient with neutropenic fever at ‘low risk’ of complications (expert opinion)
  1. CVC, central venous catheter; HITH, hospital-in-the-home; MASCC, Multinational Association for Supportive Care in Cancer; MRSA, methicillin-resistant Staphylococcus aureus; UTI, urinary tract infection; VRE, vancomycin-resistant enterococcus.

Eligibility criteria
 • Patient assessed as ‘low-risk’ for complications using MASCC risk assessment tool
 • Solid tumour and/or lymphoma
 • Not on antibacterial prophylaxis
 • No antibiotics administered during the previous 7 days
 • Ability to swallow
 • Less than grade 2 mucositis according to National Cancer Institute Common Terminology Criteria for Adverse Events v4.02
 • Good education of patient and carer on reportable symptoms
 • Availability of a 24-h caregiver
 • Availability of a telephone
 • Availability of 24-h phone advice/emergency department review from treating hospital
 • Close proximity to an emergency department or treating hospital (transport within 1 h)
 • Suitably resourced follow-up assessment, such as daily telephone contact or HITH visits
 • Patient's preference
 • Physician's preference
 • No documented allergy to oral antibiotics
 • Psychosocial and cognitive situation no barrier to learning
 • No previous history of non-compliance or absconding from medical care
Exclusion criteria
 • Confirmed focus of infection (skin/soft-tissue, UTI, pneumonia, CVC-related infection, bacteraemia)
 • Indwelling catheters with the exception of implantable devices, such as infusaports
 • High-risk chemotherapy regimens
 • Chemotherapy-refractory disease
 • Multi-resistant organism colonization (MRSA, VRE)

Structured follow-up should be planned prior to an early discharge, for example through a hospital-in-the-home (HITH) unit. If a patient is unable to be discharged early for non-medical reasons (e.g. carer not present, not in proximity to treatment centre), the recommended approach to management should still be followed while they are an inpatient, that is, start or switch to oral antibiotic therapy as they would have in the ambulatory care setting, unless otherwise indicated.

Antibiotic therapy

For low-risk patients with solid tumours and low- or intermediate-grade lymphomas undergoing 24-h observation as an inpatient, it is reasonable to start oral antibiotics immediately or deliver the first dose of antibiotics parentally according to institution's preferred practice or as outlined in the empiric treatment guideline by Tam et al. (grade A recommendation).9 Haematology patients (other than low- to intermediate-risk lymphoma) selected as low-risk should receive IV antibiotics for 48 h as an inpatient, prior to a switch to oral therapy and early discharge, if appropriate (grade D recommendation).

There is insufficient evidence at this time to support the use of continuous infusion devices in the ambulatory care setting. Further, the vast majority of studies have used ciprofloxacin and amoxicillin-clavulanate as oral antibiotic therapy for low-risk patients with neutropenic fever.5,31,33,35–37 Interestingly, many have used a three times daily (TDS) dosing regimen, unlike the twice daily (BD) doses typically used for treatment of confirmed infection. One observational study (non-randomized) used a newer generation fluoroquinolone (moxifloxacin).4 This agent may indeed be advantageous, given the broader spectrum of antimicrobial activity (activity against Gram-positive organisms), and the potential to use monotherapy (may not need to use amoxicillin-clavulanate). However, further evidence to support the use of moxifloxacin is currently not available.

Consistent with the published evidence base, dual therapy with amoxicillin-clavulanate and ciprofloxacin is recommended for the ambulatory care of patients without known drug allergy to these agents (grade A recommendation);9 clindamycin plus ciprofloxacin for patients with a penicillin allergy (expert opinion);9,24,28 and monotherapy with amoxicillin-clavulanic acid for patients with a fluoroquinolone allergy (expert opinion). Note, however, as ciprofloxacin is neither Therapeutic Goods Administration (TGA)-approved nor listed on the Schedule of Pharmaceutical Benefits (PBS) for this indication, formulary approval by a hospital pharmacy would be required to enable routine dispensing. For community scripts, a non-PBS script would be necessary, requiring the patient to pay full price for this agent.

Duration of antibiotic therapy

Defervescence prior to neutrophil recovery.  If a patient becomes afebrile within 3–5 days of treatment, antibiotic therapy should be continued for a minimum of 7 days (grade D recommendation). This treatment duration has been used safely in clinical studies of oral antibiotic therapy38–40 and is supported by consensus expert opinion.41 If a causative microorganism is identified, therapy targeted to this pathogen should be commenced, or the initial antibiotic regimen continued, provided the patient has become afebrile within 3–5 days of starting treatment (grade D recommendation).

Defervescence after neutrophil recovery.  Consistent with previous consensus expert opinion,41 it is preferable to cease antibiotic treatment when the neutrophil count recovers to ≥0.5 × 109 cells/L (grade D recommendation). If the neutrophil count remains less than 0.5 × 109 cells/L and the duration of neutropenia is expected to be prolonged, the decision to discontinue or continue antibiotic therapy should be based upon clinical criteria; for example, consider discontinuing therapy if the patient can be carefully observed, mucous membranes and integument are intact, and there is no impending invasive procedure or ablative chemotherapy planned41 (grade D recommendation).

Suitability of institutional infrastructure to implement an ambulatory care-based approach

The suitability of an ambulatory care-based approach will depend heavily upon the supportive infrastructure in place at a particular facility and the availability of relevant clinical services. The consensus group recommends the following:

  • • Ensure a multidisciplinary ambulatory care team. A feasible approach would be to use the nurse-led or pharmacist-led clinic as a model of care delivery. This would entail the development of a face-to-face or telephone clinic staffed by a senior nurse trained in haematology or oncology or infectious diseases (clinical nurse specialist or nurse coordinator) or a senior haematology or oncology pharmacist practitioner. Alternatively, nursing staff with infectious diseases experience could be used in this role. The clinic coordinator must receive direct input from suitably trained medical staff at a consultant and registrar level (oncology, haematology or infectious diseases) and have the ability to refer patients to a social worker.
  • • Link with an HITH programme, where possible, to facilitate the delivery of community-based ambulatory services, for example home visits. In the absence of an HITH programme, the haematology or oncology nurse must be available to perform home visits and telephone follow-up. Alternatively, these duties may be negotiated with the clinic coordinator.
  • • Access to a 24-h phone service (or paging system) and facility for urgent medical assessment. This will usually require the support of an emergency department. A telephone (or paging) system must be available for outpatients (or carers) at all times to enable all patients (or carers) to speak directly with a haematology/oncology nurse or member of medical staff. The clinic coordinator or the on-call oncology/haematology medical service should facilitate this. These personnel should be accountable for direct liaison with outpatients, rather than covering after-hours medical staff or hospital units who are less experienced with treatment of neutropenic fever, for example emergency department.
  • • Provision of educational materials and advice. This includes the requirement for the patient (or carer) to self-monitor and record their temperature four times a day while at home.
  • • Timely follow-up of results. A haematology or oncology nurse or pharmacist or infectious disease nurse may follow up haematology and biochemistry results if the treating physician is notified of abnormal results. Daily haematology and biochemistry results are recommended for the first 3 days after onset of fever, and thereafter around Day 7 or every 2 days until neutrophil recovery. Communication of microbiology results may occur by a facility's usual mechanisms, for example, notify treating physician of positive blood culture results by phone.
  • • Early review. Should an outpatient require assessment, early review must be expedited. In most instances, an emergency department could facilitate this (and liaise with the on-site haematology or oncology unit). If available, a medical day unit, outpatient clinic or medical assessment unit could be used for this purpose. In these instances, appropriate facilities for physical examination and diagnostic evaluation must be available.

Failure of ambulatory therapy

Ambulatory therapy is considered to have failed if a low-risk patient with neutropenic fever requires readmission following early discharge. Common reasons for failure of ambulatory therapy cited in previous publications are listed in Table 6.5,28–30,33,35,42,43

Table 6.  Reasons for failure of ambulatory therapy following early discharge for low-risk neutropenic fever5,28–30,33,35,42,43
Patient continues to report feeling unwell
Recurrent or persistent fever >38°C
Positive blood culture requiring change to another antibiotic (oral or intravenous)
Non-adherence to medication or with instructions for when to contact hospital
Decline in ability to self-care/carer unable to manage/carer no longer available
Inability to tolerate oral antibiotics (e.g. nausea, vomiting, diarrhoea)
Worsening symptoms, such as shortness of breath, rash, hard chills, diarrhoea, mucositis, pain, cellulitis and hypotension
Medication toxicity
Significant decrease in food and fluid intake (i.e. 50% less than baseline)

The decision to readmit a patient should be determined by a review of the patient with due consideration to the defined eligibility criteria for ambulatory care (see Table 5). If the medical or psychosocial status of the patient significantly changes, or the fulfilment of eligibility criteria has changed from initial assessment, then readmission to hospital should be considered. The final decision will also be influenced by severity of symptoms, physician and patient preference and the ability of the institution to monitor and deliver appropriate levels of care in the ambulatory setting.

If fever recurs, inpatient management should be commenced rather than seeking an alternative oral antibiotic regimen (expert opinion). Transfer to inpatient care may be facilitated by review in a day centre or emergency department.

Conclusion

Risk stratification and ambulatory strategies are currently underused in the management of neutropenic fever. Risk assessment using a validated tool, such as the MASCC risk index, is key to identifying low-risk patients with neutropenic fever who may be suitable for ambulatory care. The group accepts that the resources provided to outpatient facilities, and thus the resulting standard of care, can vary significantly across institutions. These guidelines provide a minimum accepted standard for implementing ambulatory care strategies within the setting of low-risk neutropenic fever. Individual centres will need to assess their ability to deliver the ambulatory strategies described herein before adopting any recommendations into institutional protocols.

Acknowledgements

See those for entire neutropenic fever guidelines, as detailed in introduction by Lingaratnam et al.1

Ancillary