Association of Colorectal Cancer Screening with Rural Residence and Travel Time

Charlton ME, Vaughan-Sarrazin MS, Kaboli PJ. Association of Colorectal Cancer Screening with Rural Residence and Travel Time. Poster session presented at: VA HSR&D Rural Health / VA Office of Rural Health Field-Based Meeting; 2010 May 6; Portland, ME.

Veterans Affairs (VA) guidelines for colorectal cancer (CRC) screening recommend beginning at age 50, unless high risk conditions warrant earlier screening. Screening modalities at VA Medical Centers (VAMCs) include double contrast barium enema (DCBE), flexible sigmoidoscopy (FS), and colonoscopy, but are not typically offered in community based outpatient clinics (CBOC). Fecal occult blood tests (FOBT) are available at both VAMCs and CBOCs. However, 42% of veterans live more than 90 miles from a VAMC and may receive most care at CBOCs resulting in barriers to CRC screening. To assess the potential impact of distance as a barrier, we evaluated CRC screening by rural residence and travel time to nearest VAMC. Methods Fiscal Year (FY) 2007 VA data files were used to identify unique VA users > = 50 years with 2 or more primary care visits during FY2007 (n = 2,551,038; mean age, 68.3 years), veteran rural residence, travel time to nearest VAMC, and CRC screening rates. Screening rates were normalized to reflect recommended screening intervals for each test (e.g., one-year colonoscopy rate represents 10% of 10-year colonoscopy rate). Logistic regression was used to evaluate relationships between screening and rural residence and travel time, controlling for demographics, comorbidity, and VA facility. Sub-analyses were performed in three groups: 1) high-risk patients with CRC, polyps, or inflammatory bowel disease; 2) colonoscopy after positive FOBT; and 3) Medicare-eligible veterans who may use non-VA services. Results Overall, 58%, 40%, and 2% of veterans resided in urban, rural, and highly rural areas, respectively; while 32%, 56%, and 13% of veterans lived within 30, 60 and 120 minutes from a VAMC, respectively. Overall (normalized) screening was 64%, with no significant variation by rurality or travel time. Among veterans 50-64 years, colonoscopy use generally declined (p < .001) with increasing travel time; relative to patients < 15 minutes from a VAMC, odds (OR) of colonoscopy were 1.05, 0.98, 0.90, 0.87, and 0.76 for patients 15-30, 30-60, 60-90, 90-120, and > 120 minutes from a VAMC. Use of FS and DCBE also decreased with increasing travel time. In contrast, odds of FOBT significantly increased with increasing travel time (0.99, 1.05, 1.14, 1.20, and 1.33, respectively). Odds of colonoscopy were slightly higher in rural veterans (OR = 1.03; p < .001), but 8% lower (OR = 0.92; p = .02) in highly rural veterans, compared to urban veterans. Conversely, odds of FOBT were higher for rural (OR = 1.15) and highly rural (OR = 1.35) veterans (p < .001) compared to urban veterans. Rural residence was not related to the use of FS or DCBE. Similar screening patterns were detected for veterans > 65 years and with high risk conditions. Of 478,052 veterans with FOBT, 40,517 (8.5%) were positive. Colonoscopy after positive FOBT was 28%, 21%, 22%, 22%, 17%, and 16% for patients within 15, 30, 60, 90, 120, or greater than 120 minutes, respectively. Implications We found little evidence for urban-rural disparities in overall CRC screening, but did find evidence of substitution of FOBT for colonoscopy and decreased colonoscopy follow-up of positive FOBT with increasing travel time. Impacts Distance from services may be a better marker for access than rural residence definitions. Further work should evaluate whether these variations in CRC screening result in outcome disparities.