Inserm U472, hôpital Paul Brousse, 16 avenue Paul-Vaillant-Couturier, 94807 Villejuif.
Clinical trials endpoints based on magnitude of reduction in HIV1 RNA levels provide an important complement to endpoints based on either virologic failure or percentage of patients below a threshold value. However, reductions in HIV1 RNA are often not completely observed since many patients have HIV1 RNA levels below the limit of quantification at the primary follow-up time. The crude method of analysing such data, is to define all HIV1 RNA levels below the limit of quantification to be equal to the limit of quantification. The crude method is widely used despite the inherent underestimation of such a method which can lead to underestimation of treatment difference with respect to HIV1 RNA reduction. Other survival methods have been suggested to estimate and compare HIV1 RNA reduction. Some methods allow take into account for missing data. These methods are applied to recent data from two clinical trials: protocols 073 (Novavir study) and 081 (Trianon study).