Objectives Aim was to assess 1) acceptance by women in the general public of overdetection (OD) in breast cancer screening, for different scenarios describing treatment after OD (mastectomy; lumpectomy; lumpectomy followed by radiotherapy; lumpectomy followed by radiotherapy and hormonal therapy), and 2) correlates of acceptance.

We recruited a random sample of Dutch and Australian women, stratified by age (45–75), through an online survey company. We assessed women’s preferences for either no screening or screening, for each treatment scenario (randomized order), using pictographs of important screening outcomes (breast cancer deaths averted, overdetected, cured irrespective of screening) for 1000 women screened over 25 years. For each scenario, we presented five pairs of numbers: deaths avoided (remained constant at 5)+number of women overdetected, varied to 0, 5, 30 (Australian model), 15 (Independent Panel’s estimate), and 2 (Dutch model). We assessed screening history, breast cancer in family/friends, prior attitudes, social norms, risk perception, worry, anticipated regret, perceived seriousness of diagnosis, comprehension of OD, health literacy, and numeracy. Using MANOVAs we compared the acceptance between treatment scenarios, and assessed associations and interactions with the other variables.

We received 854 responses, and deleted 51 due to inconsistent replies, leaving 400 Australian and 403 Dutch respondents. The majority (49%–57%) strongly preferred screening, even at the 5:30 ratio of deaths-avoided: OD; 10%–15% would never screen regardless of the ratio; around 15% would screen at 5:5. Only a slight effect (p=0.08) was seen for the effect of scenario, with lumpectomy having the highest acceptance of screening, then lumpectomy-radiotherapy, mastectomy, and lumpectomy-radiotherapy-hormones. Preference for screening was associated with having children and screening history. No effects were seen for health literacy and numeracy, but better comprehension of OD correlated with a less strong preference for screening (r=−0.35 to −0.40). The strongest psychological predictors were prior screening attitude (0.41–0.45), subjective norm (0.35–0.39), and anticipated regret (0.37–0.39). In a MANOVA, only comprehension, attitude and subjective norm remained significant.

We found very strong preferences for screening, even with 6 cancers overdetected for every death averted. To a large extent this was explained by poor comprehension of OD. This finding, and the insensitivity to treatment burden point to a general difficulty of questioning women about screening. Women have an overwhelming and uncritical feeling that screening is always beneficial. Better information and education are needed. Only once we create a good understanding of the drawbacks of screening can we assess a true harm-benefit trade-off.