Febrile neutropenia (FN) is a life-threatening side effect of myelosuppressive chemotherapy. The incidence and consequences of FN requiring inpatient care have been evaluated using healthcare claims or hospital administrative databases (Kuderer et al, Cancer 2006; Caggiano et al, Cancer 2005; Lyman et al, Eur J Cancer 1998). These sources did not include absolute neutrophil counts (ANC) and body temperature; thus the accuracy of case-ascertainment methods and findings is unknown. Moreover, none of these studies considered FN managed in the outpatient setting. Because some of these limitations may be overcome using electronic health records (EHR), a new study was undertaken.
Data were obtained from Humedica's National EHR-Derived Longitudinal Patient-Level Database (2007–2010), which includes comprehensive point-of-care information from EHR and administrative data stores across the continuum of care for ∼5 million patients. The study population included adult patients who initiated 1 or more new courses of myelosuppressive chemotherapy for the treatment of a solid tumor or non-Hodgkin's lymphoma (NHL). For each patient, each chemotherapy course and each cycle within each course was identified. FN was identified on a cycle-specific basis based on ANC <1.0 × 109/L and evidence of infection or fever (ie, temperature ≥38.3°C, diagnosis, or antibiotic use); inpatient diagnosis of neutropenia, fever, or infection; outpatient diagnosis of neutropenia and antibiotic use; or mention of FN in physician notes. Episodes of FN were categorized as inpatient or outpatient based on initial locus of care. Consequences of FN included hospital length of stay and mortality (inpatient cases only) and number of FN-related outpatient management visits. Means, percentages, and corresponding 95% confidence intervals (CIs) are reported below.
The study population included 2131 patients who received 2323 courses and 8999 cycles of chemotherapy. About 50% of patients were aged ≥65 years, and more patients were female (59.7%). The most common cancers were breast (23.0%), lung (19.9%), genitourinary (17.5%), NHL (10.7%), and colorectal (10.4%). The most common chemotherapy regimens were docetaxel/cyclophosphamide (TC; 33.9% of breast cancer patients); paclitaxel/carboplatin (PC; 42.9% of lung cancer and 51.1% of genitourinary cancer patients); cyclophosphamide/doxorubicin/vincristine/prednisone (CHOP; 42.0% of NHL patients); and fluorouracil/leucovorin/oxaliplatin (FOLFOX; 60.5% of colorectal cancer patients).
Among the 2131 patients in the study population, 401 patients experienced a total of 458 FN events, which occurred most frequently (41.0%) in cycle 1. Among the 2323 chemotherapy courses identified, the FN risk was 16.8% (95% CI: 15.3, 18.4). FN risk was highest in cycle 1 (8.1%; 95% CI: 7.1, 9.3) and cycle 2 (4.9%; 95% CI: 3.9, 6.0). Among the 8999 cycles of chemotherapy, 83.2% of FN events were initially treated in the inpatient setting and 16.8% were initially treated in the outpatient setting. Of events initially treated in the outpatient setting, 3.9% required subsequent hospitalization.
Among FN events initially treated in the inpatient setting, mean hospital length of stay was 8.4 (95% CI: 7.7, 9.1) days, and inpatient mortality was 8.1% (95% CI: 5.8, 11.1). Among FN events initially treated in the outpatient setting, the mean total number of FN-related outpatient management visits was 2.6 (95% CI: 2.1, 3.1); most encounters were in the physician's office (69.2%) or emergency department (26.9%).
Nearly 1 in 5 patients receiving myelosuppressive chemotherapy experienced FN. Most FN events (83.8%) required hospitalization either for initial treatment or subsequent to outpatient treatment, and mean hospital length of stay was greater than 8 days. Outpatient care alone was used to successfully treat 16.2% of FN events. Outpatient FN events required 2.6 outpatient management visits, most of which were in the physician's office.
Weycker:Amgen Inc: Research Funding. Barron:Amgen Inc.: Employment, Equity Ownership. Kartashov:Amgen Inc.: Research Funding. Legg:Amgen Inc. : Employment, Equity Ownership. Lyman:Amgen Inc: Research Funding.
Asterisk with author names denotes non-ASH members.