To the Editor.
—The meta-analysis published by Dr Kerlikowske and colleagues1 has added little to our knowledge of whether mammography screening for breast cancer is efficacious in women 40 to 49 years old. While the statistical technique of meta-analysis is useful for pooling data from multiple trials, its efficacy is reduced when all randomized clinical trials (RCTs) included are seriously flawed. In addition, a comprehensive meta-analysis of existing published data was recently published by Elwood et al,2 which more thoroughly reviewed the strengths and limitations of the existing literature and reached similar conclusions. No matter how it is summarized, the existing literature in this area is inadequate to make conclusions about the efficacy of screening women 40 to 49 years of age with mammography.Problems with RCTs to date include the following: (1) Only one of the RCTs was designed specifically to assess the effectiveness of mammography screening