Purpose of review: It has been common practice in randomized trials of HIV medicines to classify switches away from the original therapy as failures in analyses of virological effect, in line with an HIV-RNA measurement above a given level of quantification. This approach precludes the ability to identify the possible effects of a given therapy on those of a subsequent therapy. This review explores whether there have been changes in the reporting of randomized trials since the importance of continuous follow-up throughout the study period was initially raised 2 years ago.
Recent findings: Follow-up is still likely to be discontinued at a premature switch from study medication in a large number of the randomized trials published in 2002-2003. However, some studies, all initiated by investigators, did follow patients throughout the study period. In three of the studies, the proportions of patients with virological failure assessed with and without data after the premature discontinuation of randomized therapy could be elicited. Substantial differences were seen in the comparisons of two highly active antiretroviral therapy regimens according to the choice of analytical approach. In all three studies significant differences were observed between the regimens according to one approach, but not to the other.
Summary: The notation of treatment switch equals failure leads to an imprecise measurement of virological effect, and complete follow-up throughout the study period should be strongly encouraged, thus enabling several supplementary analyses of the virological effect of the treatment strategies being compared.