Report of the scientific meeting on some developments in the methodology of controlled clinical trials in cancer

Publisher Website
Google Scholar
Abstract
This paper summarizes discussions which took place recently between statisticians and medical practitioners about methodological developments in clinical trials in cancer. Some recent mathematical theories for the analysis of censored survival data enable comparisons to be made between the effects of different treatments without rigid assumptions about the shape of the survival curve, and taking into account various disturbing variables. The use of computers for data processing is increasingly important, and great benefits can be derived from the use of general computer programs. Many clinical studies of new agents are of limited value because of the inadequacy or absence of controls. The value of randomization is well known, but some attention is being given to the possibility that randomization might be dispensed with in certain circumstances. Another possibility is the use of special methods of allocation which tend to place the majority of patients on the better treatment. These various suggestions for new approaches to the design of trials are especially important in view of the large number of possible questions to be investigated by trials and the limited resources available.

This publication has 14 references indexed in Scilit: