The prevalence of women making an irrational decision will be compared to the participation rate in the Danish screening program. The study population will consist of a random sample of women aged 44-49. Similar to the American study, two types of information will be compared (detailed vs. non-detailed information about the harms of screening). Participants will be randomised to one of two information groups. Based on the American study, it is assumed that the willingness to participate in the hypothetical screening program will be 40 percent. To achieve the desired precision in this study, a standard error of 2.5 percent is required corresponding to a confidence interval spanning approximately 35-45 percent. This will require 384 women in each group, corresponding to a total study population of 768. The survey will be sent electronically (using "e-Boks") with a link to access the online survey trial. The survey will consist of screening and cancer related questions as well as questions on age, education, and health literacy. Five questions on appraisal of health information from the Health Literacy Questionnaire will also be included. After the data collection, data will be linked to data on income, residence, civil status, education, breast cancer diagnosis, and information on whether a participant is immigrant, descendent or of Danish origin.
The original plan for the main analysis was to compare the two groups of women randomized to different information groups (detailed vs. non-detailed information about the harms of screening). However, due to lower participation rate and therefore smaller sample size than expected, information group was disregarded in the main analysis. The two groups will still be compared in a secondary analysis. This decision was made before getting access to data from the questionnaire.
