Diagnostic yield in a biennial Hemoccult-II screening program compared to a once-only screening with flexible sigmoidoscopy and Hemoccult-II

Background: Flexible sigmoidoscopy (FS) has a higher degree of sensitivity for detecting colorectal neoplasia in the left side of the colon than Hemoccult(R) (H-II). However, no randomized controlled trial has compared a single FS screening with a H-II screening program (annual or biennial) despite the well documented mortality reduction from colorectal cancer (CRC) in the latter. The aim was to compare the diagnostic yield of colorectal neoplasia in two aged-matched groups from two different randomized screening trials; one group screened by a single FS+H-II, the other with biennial H-II over the course of 16 years. Methods: 24,465 persons invited to participate in the Funen biennial H-II screening program were compared with 4,460 similar persons invited to another Funen trial using a single FS+H-II. Results: Compliance in the biennial H-II program was 65.5% during the first screening round compared to 39.8% for FS+H-II. The cumulative number of persons with positive tests was 8.2% (positive H-II) in the biennial H-II program during 16 years and 20.3% (polyps >3 mm in diameter seen at FS or positive H-II) for once-only FS+H-II. The diagnostic yield of CRC per 1,000 screened was 9.9 in the biennial H-II program and 6.6 after FS+H-II (6.5 and 2.7 per 1,000 invited). The yield of advanced adenomas (greater than or equal to l0mm and/or villous structure and/or severe dysplasia) was 2.3% in the H-II program and 3.3% after FS+H-II among the screened persons, but this difference disappeared when persons invited, but not necessarily screened, were compared (1.5% versus 1.3%). Conclusion: Screening with H-II in a biennial screening program during 16 years detected more CRCs than a single screening with FS+H-II and a similar number of advanced adenomas.