HOME
Return to Table of Contents
197 WHAT ATTRIBUTES OF COLORECTAL CANCER SCREENING TESTS INFLUENCE PRIMARY CARE PHYSICIANS' RECOMMENDATIONS FOR AVERAGE RISK SCREENING? SK Kromm, GR Currie, RJ Hilsden, SJ Heitman, AM Irving, J Bergman, D Sadowski, DA Marshall, SE McGregor BACKGROUND: Physicians have the potential to influence the screening activities of their patients. The purpose of this study was to determine the attributes of CRC screening tests that are most important to physicians when recommending a screening test to their patients.
University of Calgary, Calgary; Alberta Cancer Board; University of Alberta, Edmonton, Alberta; McMaster University, Hamilton, Ontario
METHODS: 156 Alberta primary care physicians (88 men) participated in a discrete choice experiment (DCE) administered by mail survey (155 completed the DCE questions). The DCE involved a series of 8 choices between two hypothetical screening tests and not being screened. Each test was described by 7 attributes: accuracy, need to travel for test, waiting time, test procedure, ability to find polyps, need for follow-up test and testing interval. The physicians were asked to indicate which screening test (if any) they would recommend to a hypothetical 55-year-old patient with no family or personal history and no symptoms of polyps or CRC. Four of the choices were for a male patient and four were for a female patient. Responses were analyzed using multinomial logit regression. Two choices in each questionnaire were identical except for the gender of the patient.
RESULTS: Physicians preferred to recommend screening to their patients over no screening regardless of test characteristics. When choosing a screening test, physicians took almost all of the test attributes into account: all test attributes were significant, except for the frequency of the test. Test accuracy was the most important influence of physician's recommendations followed by ability to find polyps, need for travel, need for follow-up test, test preparation and procedure, and waiting time. All else being equal, the type of test preparation and procedure physicians preferred to recommend was one closely resembling FOBT. Patient gender was not a significant influence on the screening test recommended.
CONCLUSION: All test characteristics (except test frequency) influenced physician's recommendation. As was found in our survey of the general population, the two most important attributes were test accuracy and ability to find polyps. The patient's gender did not influence which screening test the physician would recommend. The next step in this research is to undertake further comparison between physician preferences with those of potential patients.