Statistics and ethics in medical research: study design.

The term "design" encompasses all the structural aspects of a study, notably the definition of the study sample, size of sample, method of treatment allocation, type of statistical design (randomised, cross-over, sequential, etc), and choice of outcome measures. The importance of this stage cannot be overemphasised since no amount of clever analysis later will be able to compensate for major design flaws. In this article I will consider the relation between design and ethics in observational studies and clinical trials, but I will defer the problem of sample size until the next article.