I first encountered meta-analysis in the late 1970s. A famous paper by Peto et al. (1977) had included one page on combining information from different trials, observing that “different trials which each compare the same 2 treatments may best be pooled, to give an overall treatment comparison, by analysing the data as though the separate trials were each retrospective strata within one large trial.” Also influential for me at that time was Richard Peto’s (1978) paper in a supplement to the journal Biomedicine.
Stimulated by these and other articles that presented, combined or discussed combining data from multiple studies in 1981 I gave a talk on “Evaluating a series of clinical trials of the same treatment” at the International Epidemiological Association conference in Edinburgh. Over the next two years I extended the material into a long paper with the same title, which was never (quite) completed.