Resistance tests: what do clinical trials tell us?

Few routine laboratory tests have been evaluated by randomised controlled trials. Nevertheless, a number of such studies have been undertaken for genotypic and phenotypic HIV drug resistance assays. They are all based on a structure whereby patients are randomised to receive resistance tests or none (standards of care) at the time of acute viral failure, with the primary end points being the proportion with undetectable viral load at a time point 1224 weeks following randomisation. In other words, these studies assess the benefit afforded to clinicians by resistance data. We show that these studies vary in their results, and identify the potential factors associated with such variation. Although most studies demonstrate a clinical benefit of resistance testing, we argue that the opportunity for undertaking further such studies is probably gone and that effort should be focused on identifying more precisely the factors which determine the most cost-effective use of these expensive tests.