ABSTRACT
This was a major worldwide effort involving 61 trials and almost 30,000 randomized women. It demonstrated the power and importance of collecting individual patient information on all the randomized trials worldwide related to a specific question. A detailed methodologic section provided a widely accessible description of and justification for overviews and dealt with issues about combining similar but not identical trials, and the need for large numbers to reliably detect moderate but important differences. Use of the ‘forest plot’ to graphically summarize the results of similar trials was introduced. The paper was particularly important in increasing the use of tamoxifen in early breast cancer which has led to worldwide improvements in survival.
