[Relationship between screening and colorectal cancer incidence: a systematic review and meta-analysis].

Objective To explore a relationship between screening and incidence for colorectal cancer (CRC). Methods Medline, Scopus, BIOSIS Previews, and the Cochrane Library were searched for studies, which included any randomized controlled trial (RCT) about CRC screening with a report on using endoscopy in screening arms and a usual care in control arms. Results Nine qualified trials, four one-time and five two-or-more-time screening programs, were included in this review with 997 131 persons randomized and a median follow-up of 10-28 years. In an intention-to-screen analysis, screening was associated with a reduced incidence of advanced CRCs by 10%-60%. Compared with control groups, incidence of CRC firstly increased and then decreased following a screening procedure. Five RCT with six screening arms reported incidences at different follow-up time points, and the incidence ratios at the last time of follow-up were 0.96 (95%CI: 0.86-1.06), 0.80 (95%CI: 0.70-0.92), 1.02 (95%CI: 0.91-1.14), 0.80 (95%CI: 0.70-0.90), 0.83 (95%CI: 0.73-0.94) and 0.94 (95%CI: 0.85-1.05) at the 9.0th, 10.9th, 13.0th, 18.0th, 18.0th, and 19.5th year after baseline, respectively. Compared with population-based screening, people in the hospital/clinic-based screenings had more benefits from CRC incidence reduction, particularly for those with high endoscopy rates, more lesions detected and resected by screening. Conclusions This study indicates that screening is likely associated with reduction of CRC incidence after 10 years since baseline screening, and this reduction seems to be highly affected by CRC risk of population screened, screening procedure program, and compliance to screening program including accepting endoscopy. Key words: Screening; Colorectal neoplasms; Incidence; Meta-analysis