Clinical Question: Is breast self-examination effective in reducing breast cancer mortality?

Setting: Population-based

Study Design: Randomized controlled trial (single-blinded)

Synopsis: Although preliminary results of this study were reported in 1997, the current study represents a re-analysis of data after a rigorous cleaning of the dataset. To conduct the study, investigators went to Shanghai to find a large group of women who had never heard of breast self-examination. The researchers randomly assigned women working in more than 500 factories (cluster randomization) to be trained in self-examination (132,979 women) or to a control group (133,085 women). The self-examination group received intensive training with reinforcement sessions one and three years afterward. They also received reminders and were shown videos in the work-place, and the employers were encouraged to provide additional reminders. The researchers assessed self-examination proficiency using silicone breast models. Every one to two months, members of the research team visited the factories to inquire about deaths and transfers. Whenever a woman reported finding a breast lump, the factory medical worker did the initial evaluation and then referred the woman to a surgeon if breast cancer was suspected. The researchers also used the cancer registry to track outcomes of the women. The main outcome of interest was breast cancer mortality. Researchers who were unaware of group assignment were used to confirm diagnoses of breast cancer and to determine the cause of death. Women in the self-examination group attended more than 80 percent of the sessions. The proficiency testing demonstrated only modest improvement in the women’s self-examination skills. Assessments performed one year after training demonstrated decline in their skill. During 10 years of follow-up, 4 percent of women in the self-examination group died, compared with 4.5 percent in the control group (number needed to treat: 200; P < 0.001). Rates of breast cancer mortality were identical (0.12 percent) in each group.

Bottom Line: This massive study, conducted over 10 years, demonstrates that a formal program of teaching breast self-examination has no effect on breast cancer mortality but may produce a small reduction in overall mortality. (Level of Evidence: 1b)