Abstract
BACKGROUND: The colorectal cancer mortality rate in China has exceeded that in many developing countries and is expected to further increase owing to multiple factors, including the aging population. However, the optimal policy for colorectal cancer screening is unknown. METHODS: We synthesized the most up-to-date data using a 12-state Markov model populated with a cohort of Chinese men and women born during 1949-1988, and evaluated 16 conventional and 40 risk-tailored schemes for colorectal cancer screening, considering possible combinations of age (starting at 40 + years and ending at 75 years), frequency, and strategy (standard colonoscopy, fecal immunochemical testing with colonoscopy if positive, or risk-tailored). We projected the incidence and mortality of CRC, cost, and quality-adjusted life years for 2023-2038; and performed incremental cost-effectiveness, probability acceptability, and sensitivity analyses to identify the optimal scheme and the factors affecting this choice. RESULTS: By 2038, all standard colonoscopy, colonoscopy following fecal immunochemical testing, and risk-tailored schemes were effective in reshaping China's colorectal cancer trajectory, with relative reductions in colorectal cancer incidence and mortality rates of up to 34% and 33.7%, respectively, versus no screening. Two standard colonoscopy, one colonoscopy following fecal immunochemical testing, and four risk-tailored schemes were efficient using a starting age of 40 years. Among these options, a risk-tailored scheme (standard colonoscopy every 5 years for high-risk and annual fecal immunochemical testing screening for moderate-to-low-risk) had a high probability (31.1%) of being optimal (with ≥ 40% uptake for a high-risk population, in particular), given China's present per capita gross domestic product, and would yield the highest gain in quality-adjusted life years in 17 of 31 provinces. CONCLUSIONS: Our findings suggest the commencement of colorectal cancer screening at 40 years of age in China, and that risk-tailored and some conventional schemes would be effective and cost-efficient. These findings should be valuable for policy-making regarding cancer control and resource allocation.