Background: Individuals at elevated risk for colorectal cancer (CRC) are recommended to undergo regular surveillance colonoscopies. However, as these are invasive procedures and cause a large burden on healthcare resources, research has been looking into other non-invasive methods for surveillance. This study aimed to understand preferences for CRC surveillance in Australia.
Methods: Two types of discrete choice experiments (DCEs) were designed, an unlabelled DCE and a labelled DCE. For the labelled DCE, three test types were considered, including colonoscopy with home stool test, colonoscopy only, and home stool test only. The DCEs were described using the risk of missing a cancer over the next 10 years, and surveillance interval, with an additional two attributes in the unlabelled DCE including time commitment, and pain or discomfort during the test. 800 individuals were invited from a colonoscopy surveillance program in South Australia. DCE data were analysed under random utility theory.
Results: Among 416 respondents (50% female), 26% perceived their own risk of developing CRC to be small, and 32% to be large. The risk of missing a cancer and surveillance interval were the most important attributes in unlabelled DCEs. Preferences on the surveillance interval varied by the test type in labelled DCEs. For colonoscopy only test, respondents were averse to 10 yearly surveillances. Regarding home stool test only, respondents preferred to complete the test annually. With the combined test, respondents were more likely to accept having a colonoscopy every 10 years, provided the home stool test was offered annually or every two years in between. Preference heterogeneity was observed by respondents’ risk perception and risk attitude towards health.
Discussion: A better understanding of population preferences could inform the implement of a more effective surveillance program, which help to reduce the drop-out rates and improve the cost-effectiveness of the CRC surveillance in Australia.