OBJECTIVES After initial therapy with imatinib, chronic myelogenous leukemia (CML) patients who do not completely respond may require dose escalation or switching to another BCR/ABL kinase inhibitor to achieve the desired response. This study compared healthcare resource utilization associated with either escalation of imatinib dose or switching to dasatinib. METHODS Two large administrative claims databases were combined (MarketScan and Ingenix Impact, January 1999-March 2008) to identify patients diagnosed with CML (ICD-9 code: 205.1). Patients initiated with imatinib who were continuously enrolled 6 months prior to and at least one month following their first dose increase or switch to dasatinib were selected. Patients who switched to dasatinib before reaching imatinib 800 mg/day (switchers) and the non-switchers who increased imatinib dose to >400 mg/day (dose escalators) were then followed until treatment discontinuation or end of eligibility. Negative binomial regression models were used to compare resource utilization associated with dose escalators vs. switchers, controlling for demographics, baseline imatinib treatment patterns, therapies, adverse events, and resource utilization. Cox regression models were used to study hospice services and stem cell transplants during the follow-up period among patients without such prior event. RESULTS Among CML patients who initiated on imatinib, 474 dose escalators and 175 dasatinib switchers were identified. Compared to dose escalators, switchers had significantly more frequent inpatient visits (incidence rate ratio [IRR]=3.37, p=.005), emergency room visits (IRR=1.80, p=.018), and outpatient visits (IRR 1.38 p<.001). Although low in absolute rates, switchers had substantially higher risks of hospice use (hazard ratio [HR]= 14.55, p=.066) and stem cell transplant (HR=8.71, p=0.006), indicating deteriorated clinical outcomes. CONCLUSIONS Imatinib-treated CML patients who switched to dasatinib are associated with significantly more intensive resource utilization and adverse clinical outcomes than those who escalated to higher doses. Further studies are warranted to examine the causality of the differences in resource utilization and clinical outcomes.