Breast self-examination and breast cancer: a note on postdisease reporting bias.

Recall bias in retrospective studies among cases reporting events which occur prior to diagnosis is difficult to measure. In this paper breast self-examination (BSE) practice as reported in a general population survey of 395 women is compared to data on BSE practice from six recent retrospective case-control studies of breast cancer. The focus is on the change in reported level of BSE practice which is associated with a change in the level of related study variables. The patterns of these differences in reported BSE practice are compared for the available categories of age, education, marital status, and menopausal status in each retrospective study to the corresponding pattern in the survey data. The patterns are similar except for the comparison by education. The other variables give no evidence of postdisease reporting bias. Factors other than postdisease reporting bias which might account for the education difference are discussed. Further development of a model for the impact of education on reporting bias is needed.