A better understanding of healthcare practices during the treatment of children with cancer and FN can identify areas that require more studies and guide efforts for potential interventions. This is the first report of FN practices in pediatric oncology in Latin America that informs the result of an online survey of physicians. Participants represented most Latin American countries, and they provided input about healthcare practices when managing FN in pediatric oncology patients. The survey highlights concepts used for patient care practices corresponding to risk categorization, decision-point practices, and use of guidelines. These respondents represent the staff members of pediatric cancer units caring for these types of complications in Latin American hospitals, and they are mainly oncologists, infectious diseases physicians, and pediatricians. FN guidelines can guide the management of a patient with cancer and suspected infection. As expected, practices varied and aligned with local healthcare resources and the state of development, which differs across Latin America.8 Such variability can influence FN care delivery and outcomes. We found a need to increase the use of the guidelines that could improve resource utilization, which is critical in Latin American where the public healthcare investment was found below the 6% minimal recommended mark.8
The information obtained in the survey indicated that risk categorization is used by over two thirds of the respondents, and the risk-categorization schema of patients with FN was contained in their local guidelines; mostly, the schemas were based on published literature.1,2 Differences of FN outcomes in patients with hematologic or nonhematologic malignancies depend on the effect of underlying malignancy and/or treatment phases on the immune function.9 Currently, there are six validated risk-stratification schemas that aid clinical decision-making at the initial assessment.1 In Latin America, a frequent stratification is one proposed by Santolaya et al. 6,10 Essential components of this risk stratification include ANC, platelet count, C-reactive protein, and receipt of chemotherapy for fewer than 7 days. More recently, Haeusler GM et al. 11 prospectively validated nine FN clinical decision rules (CDRs) designed to predict infection or adverse outcomes. The investigators found that none of the rules perfectly differentiated children with FN at high or low risk of infection; however, the sensitivity of the CDRs improved at Day 2 of assessment. The overall recommendations from these studies and consensus are as follows: conduct a local validation of a chosen risk-stratification schema before institutional implementation; assess the institutional ability to support the CDRs within the selected schema (for example, testing C-reactive protein, IL-8, etc.); be aware that assessment on Day 2 increases the sensitivity of some CDRs; establish extra precautions for missed infection or adverse outcomes when choosing a CDR; and keep a record and perform reviews of the performance of the specific CDRs used to evaluate accuracy and safety within a specific clinical setting.11 Additionally, consistent use of CDRs might allow a comparison of performance between sites and possibly facilitate improved use of essential resources, including antimicrobials.12
Concepts for clinical decision making, such as fever definitions and methods for measuring temperature, were not homogenous among our respondents. These findings confirmed the variations in published guidelines.1,7,10 Temperature differs based on the body site where it is measured.13 14 In published guidelines, an oral temperature is used for defining fever. However, in our survey, axillary temperature was the preferred method for measuring body temperature. Axillary temperature can underestimate the oral temperature.14 In some pediatric oncology centers, a tympanic temperature of ≥39°C defines FN and marks a point for clinical decision making.15 The upper range of tympanic temperature is 0.5°C higher than that of oral temperature and 0.7°C higher than that of axillary temperature.14 We also found that respondents used various temperature limits to define fever in FN, ranging from 38°C to ≥38.5°C. A temperature result not only defines FN but importantly guides clinical care actions, such as admission to the hospital, initial and subsequent diagnostic studies, therapeutic interventions, as well as discharge from the hospital and follow-up evaluations. Amman et al. reviewed how fever definition influences the diagnosis of FN in patients with cancer.16 The investigators found that a definition with a lower limit (38°C vs. 39°C) could increase the diagnosis of FN by more than 37%. Therefore, optimizing temperature measurement, selecting the temperature limit that defines fever in FN consistently, and evaluating other associated clinical elements may affect hospitalization, length of stay, use of resources, and costs of FN therapy.
Respondents varied in their perceptions of neutropenia definitions used for clinical decisions in FN, departing from those definitions in published guidelines, which establish an ANC of 500/µL as a decision point value. Our finding confirmed previous studies results,17 where the definitions varied, even in similar geographic areas. Neutropenia-level values usually align with the frequency and severity of infections,3 and risks for bacterial and fungal infections are higher when the duration of neutropenia is longer than 1 week, and the ANC is less than 100/µL.3 The risk imposed by neutropenia is known to be influenced by the disease, the treatment and the FN event.18 The fact that more than one-third of our respondents used a higher than accepted ANC level to trigger a clinical decision, means that they treat more patients by admitting them more often, performing work-up more often, and providing more antibiotics that require a longer hospital stay. Therefore, using standard definitions of ANC in institutional FN guidelines makes sense. 12 Deviations from key recommendations can occur, despite guidelines being locally constructed.19 The adoption and implementation of practices with accepted definitions is a multistep process that requires the active and coordinated use of personnel and resources at the healthcare facilities.20 Additionally, for sustained adoption and implementation of practices of a recommended guideline, a reasonable degree of ideal circumstances, such as competent providers and optimal infrastructure, supplies, and organizational processes, might be required.21
The respondents also identified various usages of antibiotics in FN. Management of about half of the low-risk cases involved IV antibiotics. Current guidelines recommend the use of oral treatment for those at low risk of FN,1,2 which can decrease complications and costs of inpatient care. However, therapy might have to be given as in-patient, especially if patients cannot be monitored frequently. In low-income settings represented in our study, out of town families often do not have access to lodging near the healthcare facilities. and have minimal financial resources. 22,23 In the absence of shelters, hospitals become mandatory places for lodging of patients and their families. There is a growing interest and initiatives in Latin America to provide housing for patients and their families who must travel for cancer treatment (Liliana Vazquez, 2021 – 2023 SLAOP president, personal communication). Identification of patients at low risk of FN with potential for less intensive antibiotic management could decrease the burden of crowded hospitals in low-income settings.
Overall, our findings revealed the need to continue reviewing and addressing the gaps identified through our study, including standardization of definitions, diagnostics, and treatment. When the PRINCIPAL network formed in 2017, FN was a priority, and in 2018, select members sought to review the available literature to provide recommendations for FN management in the region. That work resulted in a guidance statement.12 Currently, the PRINCIPAL network serves as a forum for ongoing discussions, training, and mentoring and for identifying areas in need of further support and improvement.
Our study has several limitations. The survey targeted FN practices in Latin America, but the results might not represent practices of all regions. Latin America has a vast and heterogenous geographic area of diverse territorial size and population. While pediatric oncology units of smaller countries have better representation among our respondents, larger countries such as Brazil, Mexico, and Argentina have a large number of pediatric oncology units and might be less representative in our survey. Another limitation is a bias we might have introduced by using infectious disease societies and networks (for example, SLIPE and PRINCIPAL) for publicizing survey participation, resulting in less than half of the respondents being oncologists. However, the fact that 30% of the respondents were infectious disease specialists was important because they often participate in building guidelines and managing infections among children with cancer. Finally, our survey was in Spanish and the compromised participation of Portuguese speaking Brazil was evident from the survey respondents.
In conclusion, variability in the diagnosis and management of FN in the Latin American region might reflect the providers’ competencies and access to resources, such as clinical decision-making tools, antibiotics, and diagnostics.Despite a unified FN-management approach (i.e., that infectious etiology must be sought, and antibiotics must be initiated), variable concepts derived from guidelines and expert opinions have been used. Among these concepts are the definition of fever; type, number, and duration of antibiotics; risk-based initial management; and more recently, risk-based management for empiric antifungal therapy. It is also evident that it is important to have a consensus and local guidelines for FN, to standardize CDRs and clinical management to allow comparisons, and more importantly, to improve care. Networks of healthcare providers for pediatric oncology such as PRINCIPAL can play a key role in advancing these changes by facilitating discussions, building consensus, developing guidelines, generating data, and championing change within their institutions.12