Disparities in Healthcare Expenditure and Mortality by Neighborhood Income in Chronic Myeloid Leukemia

INTRODUCTION: Tyrosine-kinase inhibitors (TKIs) for treatment of chronic myeloid leukemia (CML) is financially burdensome. Patients with a low neighborhood income, whose treatment cost is high relative to or exceeds income, may become “priced out” of treatment resulting in decreased TKI initiation and an increased rate of death. Low neighborhood income patients may also forgo non-cancer treatment to maintain cancer treatment. We evaluated the trend in TKI drug price and out-of-pocket (OOP) expenditure across time, and whether there is a disparity in mortality and non-cancer healthcare expenditure between patients with high and low neighborhood income. METHODS: We used the Surveillance, Epidemiology, and End Results (SEER) – Medicare database to conduct our analyses. We calculated annual average TKI gross-payment and monthly OOP expenditure and compared annual changes to inflation indices. Low neighborhood income was defined as a census-tract income below the median. We measured the adjusted effect of neighborhood income on the rate of TKI initiation using a Cox-proportional hazards model. Similarly, among TKI initiators, we measured the effect of neighborhood income on the rate of mortality. We used inverse probability weighted partitioned regression to measure the adjusted association between neighborhood income and cumulative non-cancer expenditure over 60-months. RESULTS: Average annual gross payment per 30-day supply of a TKI increased by an average of 12.8% annually throughout the observation, which exceeded 2% medical inflation. There was no trend in median annual OOP expenditure per 30-day supply, which varied between $450 to $600. We found that of 503 CML cases, low neighborhood income was not associated with TKI initiation (HR=1.12, P=0.48). Among 354 TKI initiators, low neighborhood income was associated with an increased rate of CML-specific death (HR=3.24, P=0.04), and patients with low neighborhood income had an insignificant cumulative total non-cancer expenditure that was $99,110 (95% CI [-$659,485 to $275,113]) less. CONCLUSION: Escalating TKI price and high OOP expenditure places financial pressure on patients, which may manifest as an increased risk of CML-specific death in low-income patients, but not as forgoing non-cancer treatment. Patients may instead be forgoing unmeasured goods and services. Our findings support the need for policy interventions which mitigate the financial burden of cancer.