A decision aid to support informed choices about bowel cancer screening among adults with low education: randomised controlled trial

Abstract

Objective: To determine whether a decision aid designed for adults with low education and literacy can support informed choice and involvement in decisions about screening for bowel cancer.

Design: Randomised controlled trial.

Setting: Areas in New South Wales, Australia identified as socioeconomically disadvantaged (low education attainment, high unemployment, and unskilled occupations).

Participants: 572 adults aged between 55 and 64 with low educational attainment, eligible for bowel cancer screening.

Intervention: Patient decision aid comprising a paper based interactive booklet (with and without a question prompt list) and a DVD, presenting quantitative risk information on the possible outcomes of screening using faecal occult blood testing compared with no testing. The control group received standard information developed for the Australian national bowel screening programme. All materials and a faecal occult blood test kit were posted directly to people's homes.

Main outcome measures: Informed choice (adequate knowledge and consistency between attitudes and screening behaviour) and preferences for involvement in screening decisions.

Results: Participants who received the decision aid showed higher levels of knowledge than the controls; the mean score (maximum score 12) for the decision aid group was 6.50 (95% confidence interval 6.15 to 6.84) and for the control group was 4.10 (3.85 to 4.36; P<0.001). Attitudes towards screening were less positive in the decision aid group, with 51% of the participants expressing favourable attitudes compared with 65% of participants in the control group (14% difference, 95% confidence interval 5% to 23%; P=0.002). The participation rate for screening was reduced in the decision aid group: completion of faecal occult blood testing was 59% v 75% in the control group (16% difference, 8% to 24%; P=0.001). The decision aid increased the proportion of participants who made an informed choice, from 12% in the control group to 34% in the decision aid group (22% difference, 15% to 29%; P<0.001). More participants in the decision aid group had no decisional conflict about the screening decision compared with the controls (51% v 38%; P=0.02). The groups did not differ for general anxiety or worry about bowel cancer.

Conclusions: Tailored decision support information can be effective in supporting informed choices and greater involvement in decisions about faecal occult blood testing among adults with low levels of education, without increasing anxiety or worry about developing bowel cancer. Using a decision aid to make an informed choice may, however, lead to lower uptake of screening. Trial registration ClinicalTrials.gov NCT00765869 and Australian New Zealand Clinical Trials Registry 12608000011381.

Flow of participants through trial