The use of meta-analysis in epidemiology: diet and cancers of the breast and colon.

Abstract

The word "meta-analysis" has come to replace the simple descriptive phrase "scientific overview," and I fear has come to stay. The procedure was long overdue when introduced into medicine in the 1970s by Tom Chalmers in the United States and by Richard Peto and Iain Chalmers in the United Kingdom. Its objective was to contend more effectively with the apparent disparity of results obtained by multiple clinical trials, brought about by the expansion of the pharmaceutical industry and the growth in medical research. Sometimes the same treatment would be reported as beneficial, sometimes even to be harmful, and often to have no effect or to be no better than the standard treatment to which it was being compared, on the inadequate grounds that the differences observed were not statistically significant.