Background: National mammography screening programmes are based on the results of randomised trials, but the quality of these trials has recently been questioned. The Swedish Two-County trial reported a 31% reduction in breast cancer mortality and was instrumental for the introduction of screening in many countries. In this trial, official Swedish health registries were used to identify breast cancers and breast cancer deaths in the study population.
Methods: We used data from the same registries to estimate the numbers of breast cancer cases and breast cancer deaths among the included women.
Results: Compared to official Swedish statistics we found that 192 breast cancer cases and 43 breast cancer deaths seem to be missing in the main publication of the Two-County trial; we found similar discrepancies in two updates of the trial. These large differences can hardly be explained by random fluctuations in the cancer occurrence.
Conclusion: The data reported for the Two-County trial are incomplete. Other data indicate that the mortality results in a recent report were flawed.