Longer Distance to a Specialized Treatment Center Does Not Adversely Affect Access to Treatment or Outcome in Acute Myeloid Leukemia: A National Population-Based Cohort Study

BACKGROUND Treatment of acute myeloid leukemia (AML) is widely centralized at specialized centers. Longer distances to a specialized treatment facility may affect patients' access to curative-intended treatment and ultimately survival. Few studies have focused on the potential distance decay association in hematological cancers and limitations include small sample size and lack of individual-level socioeconomic, clinical, and treatment information. AIM We designed a large national population-based cohort study of all AML patients diagnosed in Denmark between 2000-2014 (followup ending 2017) to investigate the effect of distance on treatment intensity and outcome considering individual-level clinical and socioeconomic factors. METHODS Demographics, clinical, and outcome data were obtained from the Danish National Acute Leukemia Registry (DNLR). Socioeconomic information was retrieved from registries at Statistics Denmark (Figure 1). Distance to specialized treatment centres was calculated using shortest route (Google Maps) from city center of habitation and categorized into groups. We investigated effects of distance to nearest specialized treatment center on chance of receiving intensive chemotherapy using logistic regression analysis (odds ratios; ORs). In intensive therapy patients, we calculated chance of complete remission (CR [ORs]) and in allogeneic transplantation (HSCT) candidates; we estimated chance of HSCT (ORs). Overall survival was calculated for all patients and in intensive therapy patients only using cox regression (Hazard Ratios; HRs). All results were adjusted for sex, age, and individual-level socioeconomic (ethnicity, income, education, and occupation) and disease-related factors (WBC, secondary or therapy-related AML, cytogenetics, performance status, and comorbidity). Results were given crude and adjusted with 95% confidence intervals (CI), and stratified by age (