Abstract
This article provides additional follow-up data of two cohorts from the Malmö Mammographic Screening Trial (MMST). The first cohort, MMST I, contained 7,984 women under age 50 at entry into MMST who were born between 1927 and 1932. Half were assigned to a control group and were not invited for examination until four years after the code was broken in the MMST in 1988. The second cohort, MMST II, contained 17,786 women born between 1933 and 1945. Fifty four percent of these women were randomly invited to screening between 1978 and 1990. The remaining 46%—the control group—was invited to screening between 1991 and 1994. Nine screening rounds were completed in MMST I, and a mean of five rounds were completed in MMST II; the screening interval ranged from 18 to 24 months. The effect of screening on breast cancer mortality was assessed by pooling the two cohorts. At the end of follow-up—December 1993 for MMST I and December 1995 for MMST II—there was a statistically significant 36% reduction in breast cancer mortality in the intervention groups (relative risk = 0.64; 95% CI: 0.45-0.89; P = 0.009). A harm-benefit analysis showed, however, that for every two breast cancer deaths prevented, one clinically insignificant cancer was diagnosed; for each breast cancer death prevented, 63 cancer-free women had been called back for further examinations; and for every 20 lives saved, one radiation-induced breast cancer death may have occurred. Recommendations for screening must therefore weigh mortality benefits against these negative effects.