Impact of history of cancers among the close circle on cancer screening attendance. Results of the EDIFICE survey.

A49

Introduction The population-based EDIFICE survey is the first to focus on French population's adherence and attitude towards cancer screening (CS) for breast, colorectal and prostate cancers. In France, CS: - is organized over the whole country for breast cancer (1 mammography every other year for women aged 50 to 74 years); - is organized for colorectal cancer in 22 pilot departments since 2004 (Hemoccult^® test for men and women aged 50 to 74 years; - is not organized for prostate cancer. To our knowledge, EDIFICE is the 1^st-population-based survey to address this issue and compare the influence of history of cancers among the close circle (relatives, friends) on screening for breast, colorectal or prostate cancers. Methods This nationwide, observational telephone survey was carried out in France from January 18th to February 2nd, 2005 among a representative sample of 1504 subjects. Representativeness of the subject sample for sex, age, profession, region of residence and community size was ensured based on the statistics of the survey conducted by the French National Institute for Statistics and Economical Studies (INSEE). The survey questionnaire was administered by trained and independent interviewers using the Computer-Assisted Telephone Interview (CATI) technique. Main results 1) In univariate analyses, having undergone a screening test is significantly associated with a higher rate of persons affected with cancer in their close circle. For breast cancer, 73% of women who have undergone at least one mammography have an history of cancer among their close circle vs 49% for women with no screening test (p<0.01). These rates are 74% and 66% (p<0.05) for colorectal cancer, 71% and 58% (p<0.01) for prostate cancer. In multivariate analyses, however, knowing somebody having cancer was not associated with a higher rate of participation in screening tests for breast (OR=2.167;IC95%:0.953-4.927) and colorectal cancer (OR=1.122;IC95%:0.486-2.589). For prostate CS (Digital Rectal Examination and/or PSA determination), the corresponding OR was 4.93 (95% CI:1.68-14.42) and ranked second after the OR for consultation with an urologist. 2) There is a strong association between being in touch with persons having cancer and feeling concerned by CS (OR=4.5 for colorectal cancer, p<0.0001) Discussion and conclusion Based on our data, we assume that for breast and colorectal cancers the statistical weight of affected persons in the close circle is switched to other items, such as being concerned or feeling more motivated which appeared in the logistic model. It appears that high levels of concern and awareness can be achieved through information and communication in organized programs. In contrast, for prostate cancer for which there is no organized national program and associated communication, the existence of affected persons in the close circle is a strong momentum to be screened.