Objectives: To determine if participation in colorectal cancer screening using faecal occult blood testing (FOBT) is affected by a restrictive diet and if it is associated with certain demographic variables.
Participants and setting: 1,203 residents of South Australia aged 50-69 years, with no "currently active bowel disease", randomly selected from a database of people willing to be contacted about unspecified health issues.
Design: Randomised controlled trial: participants were offered screening by immunochemical FOBT by mail in 1998. Half were randomly allocated to a group instructed to follow a low-peroxidase diet, as required for guaiac FOBT, while the other group was not so restricted.
Main outcome measures: Effect of diet restriction on participation (return of correctly completed FOBT sample cards within 15 weeks); time taken to return cards; relationships between participation and demographic variables.
Results: Participation rates were 65.9% (no-diet group) and 53.3% (diet group) (difference, 12.6%; 95% CI, 7.1%-18.1%). In the first week, rates of return as a proportion of all tests returned were 13.1% (no-diet) and 1.6% (diet) (difference, 11.5%; 95% CI, 8.6%-14.4%), increasing to 54.3% and 44.5%, respectively, after five weeks (difference, 9.8%; 95% CI, 4.2%-15.4%). Participation was significantly associated with older age (odds ratio, 1.40; 95% CI, 1.10-1.78), but not sex, Index of Social Disadvantage or rural versus urban address.
Conclusions: Dietary restrictions create a barrier to FOBT-based screening for colorectal cancer. The use of immunochemical rather than guaiac FOBT removes this barrier.