OBJECTIVES:  Fecal occult blood test (FOBT) has been widely used in preliminary screening for colorectal cancer (CRC). Moreover, high-risk factor questionnaire (HRFQ) and quantitative risk-assessment model (QRAM) were recommended to estimate the risk of CRC qualitatively and quantitatively in China. We compared the diagnostic value of CRC preliminary screening methods to provide a preferable one for screening strategy. METHODS: Individuals aged 40-74 years old were enrolled in the mass CRC screening program in Jiashan County, Zhejiang Province, China. CRC preliminary screening with FOBT and questionnaire (HRFQ or QRAM) was conducted to identify high-risk population. We linked the screening data to CRC Surveillance and Registry System, and CRC detection during the screening or later in the system was regarded as gold standard. We calculated the sensitivity, specificity and their confidence intervals (CIs) to assess the performance of CRC preliminary screening methods.

RESULTS: From 1 January 2007 to 31 December 2014, a total of 142 124 subjects participated in the screening program. FOBT had a sensitivity of 75.49% (95% CI: 69.84-80.39) and a specificity of 90.36% (95% CI: 90.17-90.54) for detecting CRC cases within the first follow-up year. QRAM was more sensitive (P<0.001) and less specific (P<0.001) than HRFQ. The sensitivity and specificity of FOBT along with HRFQ were 86.56% (95% CI: 81.81-90.22) and 81.37% (95% CI: 81.12-81.62), and those of FOBT along with QRAM were 88.93% (95% CI: 84.47-92.23) and 73.95% (95% CI: 73.67-74.23) at one year of follow-up. In addition, sensitivity of each screening method was at a downward trend as the follow-up period increased.

CONCLUSIONS: CRC preliminary screening with FOBT and QRAM in parallel has high sensitivity and satisfactory specificity for detecting CRC and is a useful strategy in mass screening programs.