OBJECTIVES: Colorectal cancer screening is a cost-effective method to reduce incidence and mortality. In a population-based setting, many factors may contribute to the performance of screening. This study investigated factors that influence the quality of cancer screening.

METHODS: The national colorectal cancer screening by biennial faecal occult blood test has been implemented since 1992 in Taiwan. In this study, we linked the colorectal cancer screening database (2010-2014) with Cancer Registry and Death Registry (till 2015). Possible influencing factors during screening progress were identified. The outcome variables included detection rate, proportion of early stage diagnosis and survival rates. The logistic regression and Cox regression were adopted to identify and to estimate factors for outcome performance.

RESULTS: About 5.26 million screenings were performed during 2010-2014. Among them, 7.7% were referred for positive results of occult blood tests, of which 64.7% were further confirmed by colonoscopy. The second phase results showed that colorectal cancer, polyps accounted for 3.7% and 49.3%, respectively. Of the 9341 records found in the cancer registration, 66.2% (6448 records) were in stages 1-2.

Higher detection rates of colorectal cancer and polyps were in males (3.8%), hospitals (3.5%), elderly (3.5% or above), and people with family history for colorectal cancer (3.3%). Referred patients who actually complete colonoscopy had a lower risk of death than no referral cases, while in the screening results was colorectal cancer (OR=0.49, 95%CI=0.45-0.54).

CONCLUSIONS: We identify that age, family history, hospital and local health center in current cancer screening program are associated with higher detection rates or lower mortality. Our results can contribute to a future risk-stratified modification of our current colorectal cancer screening.