Influence of primary care use on population delivery of colorectal cancer screening.

OBJECTIVE: Colorectal cancer (CRC) screening is commonly initiated during primary care visits. Thus, at the population level, limited primary care attendance may constitute a substantial barrier to CRC screening uptake. Within a defined population, we quantified the percent of CRC screening underuse that is potentially explained by low use of primary care visits.
METHODS: Among 48,712 adults ages 50 to 78 years eligible for CRC screening within a Washington state health plan, we estimated the degree to which a lack of CRC screening in 2002 to 2003 (fecal occult blood testing, sigmoidoscopy, or colonoscopy) was attributable to low primary care use, expressed as the population attributable risk percent (PAR%) associated with 0 to 3 primary care visits during the 2-year period.
RESULTS: In analyses adjusted for age, comorbidity, nonprimary care visit use, and prior preventive service use, low primary care use in 2002 to 2003 was strongly associated with a lack of CRC screening among both women and men. However, a majority of unscreened women and men had > or =4 primary care visits. Thus, whether low primary care use was defined as 0, 0 to 1, 0 to 2, or 0 to 3 primary care visits, the PAR% associated with low primary care use was large in neither women (range, 3.0-6.8%) nor men (range: 5.6-11.5%).
CONCLUSIONS: Health plan outreach efforts to encourage primary care attendance would be unlikely to substantially increase population uptake of CRC screening. In similar settings, resources might be more fruitfully devoted to the optimization of screening delivery during primary care visits that patients already attend.