Physicians' Colorectal Cancer Screening Discussion and Recommendation Patterns

Background: Primary care physician (PCP) actions are pivotal to colorectal cancer (CRC) screening performance, and guidelines recommend discussion with patients about test options and potential benefits and harms. This article profiles patterns of discussion about and recommendations for screening and explores potential associations with multilevel factors (patient, clinician, practice, and environment). Methods: In 2009, we analyzed data from 1,266 physicians responding to the 2006-2007 National Survey of Primary Care Physicians' Recommendations and Practices for Breast, Cervical, Colorectal, and Lung Cancer Screening (absolute response rate = 69.3%; cooperation rate = 75.0%). Descriptive statistics examined physicians' reports of discussion and recommendations. Multivariate analyses assessed the associations of these practices with multilevel factors. Results: Although few respondents reported discussion of all options, 46% usually discuss more than one option; the vast majority of these respondents discuss fecal occult blood testing (FOBT) and colonoscopy (49%) or FOBT, sigmoidoscopy, and colonoscopy (32%). Of physicians who discuss more than one option, a majority reported usually recommending one or more test options, most commonly colonoscopy alone (43%) and FOBT and colonoscopy (43%). Several personal characteristics (specialty), perceived patient characteristics (prefer physician to decide), practice characteristics (geographic location), and community barriers (specialist availability) were independently associated with discussion and/or recommendation patterns. Conclusions: PCPs do not discuss the full menu of test options, but many report selecting one or two options for discussion and recommendation. To ensure that patients' perspectives and concerns are elicited and considered, patient decision-making approaches should be considered. Impact: Attention to informed decision making in CRC screening will be important for enhancing patient-centered quality care. Cancer Epidemiol Biomarkers Prev; 20(3); 509-21. (C)2011 AACR.