Introduction

Clinical trials have been conducted in patients with AIDS or HIV for ten years. There is no obvious reason why the principles underlying good clinical trial design should be any different for trials on HIV than for any other disease. However, the social setting of the infection has demanded that some of these principles should be more clearly justified. Attempts have been made by some trialists to conduct trials that avoid the need for randomisation or placebos, or attempt to use early signs of disease progression or laboratory markers rather than death as an end-point. This paper will illustrate these issues by comparing the design and analysis of two large European studies, coordinated by the MRC at the Brompton Hospital (in collaboration with INSERM in Paris), with other trials studying the same questions.

Summary of some of the key trials

Trials on the early use of Zidovudine

Zidovudine is an anti-HIV drug which was shown in one study to reduce mortality in patients with advanced HIV disease over a median study time of four months. On the basis of this study, the drug has been licensed and widely used.