Risks and consequences of travel burden on prophylactic granulocyte colony-stimulating factor administration and incidence of febrile neutropenia in an aged Medicare population.

Objective: Granulocyte colony-stimulating factors (G-CSFs) decrease the incidence of febrile neutropenia (FN) in patients receiving myelosuppressive chemotherapy. This study examines the impact patient travel burden has on administration of prophylactic G-CSFs and the subsequent impact on FN incidence.
Methods: Medicare claims data were used to identify a cohort of beneficiaries age 65+ with non-myeloid cancers at high risk for FN between January 2012 and December 2014. Driving distance and time were calculated from patient residence ZIP code to the location of G-CSF and/or chemotherapy administration. Regression models were used to estimate the odds of G-CSF prophylaxis relative to patient driving distance and time, and odds of FN incidence relative to timing of G-CSF administration (optimal [days 2-4 after chemotherapy], sub-optimal [same day], or none).
Results: The 52,389 study patients had a mean age of 73.5 years, and were 82% female and 89% white race; 49% had female breast cancer, 12% lung cancer, 15% ovarian cancer, and 24% non-Hodgkin's lymphoma. Of these high FN risk patients, 69% had at least one prophylactic G-CSF administration within at least one chemotherapy cycle. The percentage of patients receiving prophylactic G-CSFs in the first cycle was 56%. Median travel time was slightly longer for patients who did not receive G-CSFs and patients receiving short-acting vs long-acting G-CSFs. The odds of receiving no G-CSFs were 26-52% higher (depending on cancer type) for patients with a >80-min one-way travel time, compared to patients traveling <20-min. Concurrently, the odds of FN (using a "narrow" definition) were 18-93% higher for patients who did not receive G-CSFs in the first cycle of chemotherapy.
Conclusions: Travel burden, linked to clinic visits for G-CSF administration following myelosuppressive chemotherapy, is associated with sub-optimal use of G-CSF prophylaxis, which may result in a higher incidence of FN.