Six years of colorectal cancer mortality surveillance in the screening population for a risk stratified screening program.

Objective: To evaluate the impact of a colorectal cancer (CRC) risk predicting system on CRC mortality rates.

Method: An organized population screening program targeted at all the subjects (n = 102,076) at age 40-74 in nine towns of Jiashan county, China was conducted from 2007 to 2012. All of the screening participants were first triaged into high-risk & low-risk groups by a questionnaire and two fecal immunuochemical tests, only the high-risk subjects were subject to colonocopy. The screening participants were surveyed death caused by CRC for a total of six years after the enrollment. The CRC mortality in subgroups of the screening population was analyzed.

Results: A total of 82,184 (80.51 % of the targeted population) screening participants were identified. CRC death were recorded for 142 subjects (28.819 per 10 person-years). The age-adjusted relative risk(RR) of CRC death in the high-risk subjects (n = 12862, 84.48 per 10 person-years) was 3.92 (95 % CI = 2.81-5.49) compared with the low-risk subjects (n = 69322, 18.52 per 10 person-years). In the high-risk group, the age-adjusted RR of CRC death for those accepted colonoscopies (51.44 per 10 person-years) compared with those refused colonoscopies (187.94 per 10 person-years, P < 0.0001) was 0.34 (95 % CI = 0.21-0.56). The first three years after screening has seen the largest difference of CRC death hazard in both comparing groups.

Conclusion: The high-risk subjects triaged by the risk predicting system have a higher CRC mortality rate than the low-risk subjects, especially in the first three years after screening. Refusal of colonoscopy is risky behavior for the high-risk subject.