Introduction
Neutropenia is a risk factor for certain infections that may appear as a frequent and severe complication in oncological patients; its only typical symptom is fever [
1].
This entity has classically been treated through hospitalized management and intravenous (i.v.) administration of wide-spectrum antibiotics in all patients. Currently, only a fraction of patients with febrile neutropenia present serious complications, and there is evidence of a diverse prognostic spectrum [
2‐
10]. In light of this fact, the main scientific societies at a national and international level, including the American Society for Clinical Oncology (ASCO) [
11] and the Spanish Society for Medical Oncology (SEOM) [
12], make a distinction between patients with febrile neutropenia who are at high or low risk, depending on defined criteria set out in globally validated scales like the ones from the Talcott group [
3], the Multinational Association of Supportive Care in Cancer (MASCC) [
13], and the clinical index of stable febrile neutropenia (CISNE) [
14]. Better understanding of these different risk profiles, plus the emergence of new care models like home hospital care, have raised the possibility of less aggressive treatment options, for instance simplified oral monotherapy and immediate or early discharge following an observation period.
This qualitative systematic review aimed to assess ambulatory care strategies in adult patients with febrile neutropenia through a triple aim, examining efficacy, cost-effectiveness, and quality of life. These results can inform the definition of the most adequate setting for managing these patients (ambulatory care, home hospitalization, or inpatient care); contribute to updating the action protocols in emergency services; and identify areas where there is uncertainty in the available evidence in order to trace new lines of research.
Discussion
Since at least the early 1990s, with Rubenstein et al. [
10], and possibly even before then, there has been recognition that the risk of complications is heterogeneous in populations with febrile neutropenia. The few studies available on this topic have been characterized by the diversity of their aims, approaches, and patient selection criteria. Factors contributing to this heterogeneity include evolving knowledge on the entity; the emergence of new oncological treatments (chemotherapies, CSFs, antibiotics); the growing tendency to simplify treatments to favor adherence and tolerability and to reduce adverse effects; and the decentralization of health care away from hospitals due to associated comorbidity, impaired quality of life, and increased cost.
Indeed, even the definition of fever itself and the best place to measure it have changed several times since Carl Wunderlich’s early work on it in 1868. This branching evolution is evident in the current definitions of the Infectious Diseases Society of America and other societies in South America, Europe, and Asia, and in the different ways that our included studies identify it; however, it is not necessarily reflected in the publication year of our included studies.
These changes are also reflected in the fact that Minotti et al. [
7] excluded patients with neutrophil counts under 300/μL because at that time, severe neutropenia, especially with neutrophil counts of less than 100/μL, was considered a marker of high risk. Later studies do not reflect the application of this criterion. Indeed, the interpretation of our results must consider the variations in the neutrophil count of included and excluded patients. Kern et al. [
1], Sebban et al. [
4], Innes et al. [
5], Hidalgo et al. [
6], and Malik et al. [
9] all included patients with moderate neutropenia (500–1000/μL) with the potential to become severe neutropenia (< 500/μL). However, only Malik et al. [
9] clarified that the patients whose neutrophil counts did not drop below the 500/μL threshold were excluded. However, given the inclusion of patients with critical levels of neutropenia in other studies, this distinction may be irrelevant.
Another fact suggestive of the diversity of selection criteria is that despite criticism of Talcott’s group and the MASCC and CISNE scales for not considering the neutrophil count as a criterion for risk assessment, the risk associated with this aspect does not seem to determine the appropriateness of managing neutropenia on an ambulatory basis. Indeed, good outcomes were achieved with simplified therapies in the ambulatory setting regardless of the criteria for selecting candidates. In the studies that did report the proportion of patients with neutrophil counts that dipped below 100/μL (Kern et al. [
2], Talcott et al. [
3], Hidalgo et al. [
6], Minotti et al. [
7], Malik et al. [
9], and Rubenstein et al. [
10]), these proportions were described as similar between groups in all studies but Rubenstein et al.’s, who did not specify.
Regarding the risk assessment scales, the ASCO 2018 [
11] guidelines recommend the use of MASCC followed by CISNE, while the SEOM 2018 [
12] guidelines prefer only the CISNE. However, the included studies did not evaluate the latter scale as a method to identify candidates for ambulatory care. Moreover, CISNE has not been validated in hematological patients, restricting the subgroups eligible for less aggressive interventions. Furthermore, Ahn et al. [
33] demonstrated that the MASCC had greater discriminatory power than the CISNE for detecting low-risk patients, with greater sensitivity and negative predictive value, resulting in a smaller likelihood of including patients at true high risk of complications in ambulatory regimes. Thus, in 2013, Ahn et al. [
34] already proposed procalcitonin values as an additional parameter for the MASCC to refine the probability of presenting bacteremia or septic shock, in patients at both low and high risk. Even within the latter group, the authors recognized heterogeneous prognoses, calling into question the need for indiscriminately applying aggressive management.
The evidence on antibiotic treatment approaches, both for double therapy with amoxycillin/clavulanic acid (or clindamycin in patients with allergies) plus quinolone and for monotherapy with fluoroquinolone, shows good efficacy outcomes in and outside of the hospital setting (Kern et al. [
2], Sebban et al. [
4], Hidalgo et al. [
6], Minotti et al. [
7], Malik et al. [
9]). Despite these results, the ASCO guidelines [
11] continue advising against fluoroquinolone monotherapy for ambulatory management of low-risk patients, while SEOM affirms that the only alternative in penicillin-allergic patients is hospital admission and i.v. treatment.
In 20 to 30% of febrile neutropenia cases, the germs responsible are identified, predominantly gram-positive rather than gram-negative bacteria, followed by polymicrobial and fungal infections (Malik et al. [
9], Rapoport et al. [
8], Hidalgo et al. [
6], Minotti et al. [
7], Sebban et al. [
4]). These results are consistent with other available studies in the literature.
With respect to the use of CSFs, it has been known for decades, as reported by studies like Mather et al. in 1994 [
35], that these interventions are capable of reducing the mean duration of the neutropenia by 1 day as well as the total time needed for resolution, although not the total days of fever. No study has explained how authors decided to administer this treatment, describing the criteria, type of neoplasm, level of neutrophils, or timing of administration (from the last session of chemotherapy or emergency presentation). Similarly, the literature has not explored how CSFs may have influenced the outcome (patient recovery or the success of the treatment strategy). There are also gaps in research investigating whether prescribing CSFs would result in savings given the implicit pharmacological costs associated with hospitalizing these patients in a hospital or their home, and whether it would be possible to reduce the duration of antibiotic therapy in patients with a fever of unknown origin and negative culture.
Several strategies for ambulatory follow-up have been evaluated, including home visits from health professionals, hotlines, scheduled telephone calls, and outpatient hospital visits. However, no studies have compared these interventions with each other to assess treatment response, adverse effects, cost-efficiency, or quality of life. These aspects could be the focus of future lines of research.
Although teenagers were not included as population in our systematic review, because we excluded studies with population under 19 years old in our criteria research, in many countries, they are treated around 16 years old in emergency departments as adults. Authors, such as Klastersky et al. [
36], carried on an early discharge strategy after 24 h, til 48 h in many of the episodes of febrile neutropenia in people that includeed 16 years old teenagers [
36].
Their results show that we could safely proceed in low-risk teenagers according to MASK score, in the same way as adults with an oral combination therapy of peniciline and quinolones, if not allergic, with a proposed followed up by phone, temperature control each/6 h, and blood control test each 48 h until accomplishing criteria of recovering. We should highlight in this investigation team, they first give a definition more than absence of hospital readmission as a criteria of succes of oral empirical therapy as 5 days without fever, no clinic symptoms, and signs of infection and pathogen erradication.
Within the sphere of ambulatory management, four studies proposed that patients receive hemograms every 2 to 3 days after discharge, at least until achieving neutrophil counts of 1000/μL, indicating mild neutropenia (Goodman et al. [
22], Kern et al. [
2], Sebban et al. [
4], Hocking et al. [
19]). Although several studies have evaluated the cost of ambulatory follow-up, there has been no calculation of the specific cost of such controls, which could generate substantial amounts of research information. It may be appropriate to restrict these control visits to cases with poor evolution or to limit them to one visit at the end of treatment.
Another strategy that could be analyzed for its efficacy and cost-efficiency would be the performance of a preliminary assessment of oncological patients’ household situation (including family support) by home hospital team, prior to initiating chemotherapy. This assessment could help facilitate decision-making in emergency departments around managing these patients.
Many studies have limited ambulatory approaches to patients who live at some distance from the reference center or hospital (Kern et al. [
2], Hidalgo et al. [
6], Talcott et al. [
3], Minotti et al. [
7], Rubenstein et al. [
10], Rolston et al. [
18], Lingaratnam et al. [
17]). To reduce the population of low-risk patients excluded from this strategy, future research could also assess the efficacy, patient satisfaction, and cost-efficiency associated with strategies in coordination with primary health care teams, for patients who live far from the hospital and whose only limitation for inclusion in an ambulatory regime is distance.
Available studies suggest than an ambulatory strategy for low-risk patients with febrile neutropenia is cost-efficient as well as effective. This is true for interventions assessed both in clinical trials (Hendricks et al. [
29], Elting et al. [
27]) and in prospective studies, for example, in those evaluating the implementation of new protocols compared to data from historical cohorts fulfilling the same criteria (Teh et al. [
30], Mayordomo et al. [
25]). Economic studies should be considered in light of potential differences in the source of economic data and their perspective. Establishing the most appropriate design to allow comparison with other studies can be complicated because health systems are not managed the same across different countries, nor is access to different therapeutic options the same in privately managed systems. Thus, in future studies, it will be necessary to establish quality criteria in order to favor reliability and comparability of the economic data obtained.
With regard to quality of life, Teuffel et al. [
37] suggested that patients prefer ambulatory strategies, although data did not show differences in the short quality of life evaluations performed in two clinical trials in people with febrile neutropenia. In any case, more structured trials that specifically aim to assess this domain are needed before drawing any firm conclusions.
Despite the positive evidence, fewer patients benefit from ambulatory strategies than expected (for clinical, socioeconomic, and geographic reasons). Health services still seem reluctant to implement the action protocols in low-risk patients that are recommended by national and international clinical practice guidelines, even though these tightly restrict the application of ambulatory models.
Available evidence on efficacy, quality of life, and cost-efficiency seems to support ambulatory approaches for low-risk patients with febrile neutropenia. The main societies and services involved in managing these patients should establish more standardized priorities and criteria to guide future research and reach more definitive conclusions. Studies are needed to analyze the facillitators and barriers for the instigation and maintenance of an ambulatory febrile neutropenia management programmes.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.