||5R01CA144835-04 Interpret this number
||University Of Wisconsin-Madison
||Virtual Colonoscopy and Colorectal Cancer Screening
DESCRIPTION (provided by applicant): Colorectal cancer should be almost entirely preventable if identified and removed in the precancerous polyp stage yet almost 50,000 people each year die of this preventable disease. Rates of colorectal cancer screening are low partly because of the invasive nature of traditional optical colonoscopy (OC). Virtual colonoscopy (VC) is a less invasive test that is newly recognized by the American Cancer Society as one of the preferred methods for colorectal cancer screening and prevention. Although the need for additional approaches to improve colorectal cancer screening rates is clear, the role of VC remains controversial. In particular, it is not clear whether widespread VC availability is an important part of strategies to increase overall screening rates or, perhaps more importantly, as a tool to increase screening via cancer prevention (rather than simply cancer detection with a stool test after cancer develops). Addressing this question is increasingly critical given recognition that some areas of the country have extremely long waiting times for OC due to the nationwide shortage of gastroenterologists. The answer is not straightforward; there may be important barriers to VC use as a standard screening test. For example, primary care physicians (PCPs) may limit their use of VC because they wish to avoid the unplanned identification of new findings (e.g., extracolonic findings) that require additional follow-up for diagnosis or therapy. Consequently, individual PCPs likely have an integral role in the use of VC for patients. To inform policies and target interventions to improve practice, it is crucial to determine whether increased availability of VC improves screening and what factors influence the use of VC in typical clinical practice. The University of Wisconsin is the only center in the US with large numbers of patients who have commercial insurance coverage for both VC and OC screening, uniquely positioning us to address these questions. Our specific aims (1) examine whether patients with insurance coverage for VC have an increased likelihood of colorectal cancer screening or screening for cancer prevention versus cancer detection, (2) determine which PCP characteristics and key experiences affect the use of VC over time, and (3) identify if certain perceptions of VC advantages and disadvantages affect the extent to which a PCP uses VC. We will use data from our electronic health record to identify ~38,000 patients eligible for colorectal cancer screening each year from 2003 to 2010, assign patients to one of ~450 PCPs, and track these patients to determine their use of different screening tests. Because four major local insurers cover VC for screening, while other insurers do not, we will have comparison groups of patients who do and do not have coverage for colorectal cancer screening with VC. PCPs practices and experiences will be assessed using aggregated data from the electronic health record, a PCP survey, and data from the American Medical Association. Overall, the results from our investigation will impact strategies to increase colorectal cancer screening, to change the percent of patients screened for cancer prevention vs. detection, and to optimize the use of VC as a screening tool.