Less than optimal treatment for colorectal cancer in Alabama: Do disparities exist?

B84

Background: Studies of cancer treatment utilization are mostly based on populations covered by the SEER registry which does not include Southern States except for parts of Georgia. Alabama is among the poorest states in the US with 45% of the population living in rural areas (vs. 21% of US population). Alabamians face more challenges in accessing health care because of low income, transportation barriers, and location of medical facilities. For some cancers, black residents have lower incidence rates but higher mortality rates than white residents; this may be explained by lower access to quality care.
 Objective: To describe treatment for early stage colorectal cancer in elderly cancer patients residing in Alabama, and to identify demographic factors (age, gender, race) associated with initiation and completion of treatment.
 Methods: We obtained Medicare administrative claims for 3,137 individuals identified through the Alabama Statewide Cancer Registry as being diagnosed with stages I-III colorectal cancer in 1999-2003. Treatment was identified using International Classification of Disease-v.9 (ICD-9) procedure and Current Procedure Terminology (CPT) codes. Treatment duration was determined using the dates of first and last claim. Logistic regression modeling identified factors associated with the likelihood of initiating and not completing treatment while adjusting for stage of disease, comorbidities and concurrent chemotherapy-radiation therapy (XRT). Non-completion of adjuvant chemotherapy was defined as fewer than 5 months of treatment, and of XRT as fewer than 5 weeks.
 Results: Among 2,464 colon cancer patients, 94% received surgery, 32.5% chemotherapy (60% of stage III patients) and 3% XRT. There were no differences by gender or race in treatment initiation. Patients older than 75 were significantly less likely to receive chemotherapy than younger ones (22.0% vs. 44.7%, p= 0.0001). Over 25% of patients did not complete chemotherapy; women were more likely to discontinue than men (29.6% vs 23.3%, p= 0.06) while there were no race or age differences.
 Among 673 rectal cancer patients, 88% received surgery, 41% chemotherapy and 38% radiation. Among stage III patients, 64.5% and 50% started chemotherapy and XRT, respectively. There were no differences in chemotherapy or XRT initiation by gender or race. Patients younger than 75 were more likely than older ones to start chemotherapy (46.9% vs. 34.0%, p= 0.002). Of patients on chemotherapy, 30-50% did not complete treatment depending on whether chemotherapy was given alone or with XRT. There were no differences in dropout by race and age. Women were more likely to discontinue chemotherapy than men (51.0% vs. 26.4%, respectively, p= 0.001). Over 20% of patients on XRT failed to complete treatment regardless of whether XRT was given before or after surgery. There were no differences by race, gender or age.
 Conclusions: A significant portion of colorectal cancer patients in Alabama did not complete a standard course of treatment. Minorities were not more at risk of dropping out of treatment, while women were less likely to complete chemotherapy than men. Age affected initiation of treatment but not its completion. Results may be limited by the small sample sizes especially for rectal cancer. More studies are needed to address reasons why Alabama cancer patients may receive less than optimal treatment for colorectal cancer.