Tracker trialsBMJ 2000; 320 doi: https://doi.org/10.1136/bmj.320.7251.1727/a (Published 24 June 2000) Cite this as: BMJ 2000;320:1727
Introduction of resistance testing might be an inappropriate use of resources
- David Dunn, senior statistician (, )
- Sheena McCormack, trial physician,
- Abdel Babiker, head of HIV Division,
- Janet Darbyshire, director
EDITOR—Lilford et al argue for starting randomised studies of new health technologies as early as possible, even if the technology is in a phase of rapid development.1 Many of the points discussed are highly relevant to assays measuring drug resistance in patients with HIV infection.
These assays are currently primarily used as research tools, but recent articles and clinical guidelines have recommended that they should be routinely performed to guide the selection of antiretroviral drugs in the management of patients with HIV infection.2 The basis for these recommendations is not clear: the arguments are more complex than they first seem, and the empirical evidence that resistance testing improves clinical outcome is limited.3
The most commonly used form of resistance testing entails DNA sequencing of the reverse transcriptase and protease genes. But quality assurance studies have found that current methods frequently fail to identify key mutations associated with resistance.4 Moreover, it is often difficult to decide how to use the result of the resistance assay, since the influence of viral polymorphisms on in vivo response to the many combination drug regimens available is poorly understood. It is likely that resistance testing will ultimately improve the selection of drug regimens and become cost effective as the accuracy and interpretation of assays improve. There is no certainty, however, that this point has been crossed, and the widespread introduction of resistance testing at this time could be an inappropriate use of scarce healthresources.
Lilford et al propose flexible randomised trials—where duration is not predetermined and frequent interim analyses are conducted explicitly—and recognise that the effectiveness of a health intervention may change over time and they aim to monitor such changes.1 Testing for resistance of HIV may be appropriately evaluated by this type of study, although, as in other areas, convincing the medical community of the merits of this approach and securing funding may be problematic.5
A randomised trial would collect the information required for analyses to elucidate the clinical significance of viral mutations. These analyses could be performed during the trial without compromising the main comparison of resistance testing versus no testing. In principle, the findings from these analyses could influence the interpretation of resistance assays performed later in the study. This would be a strength, not a weakness.
Competing interests The MRC Clinical Trials Unit is currently coordinating a (conventional) randomised controlled trial of HIV resistance testing.
Continuous process of trial and review is needed
- Hazel Thornton, founding chairman, Consumers' Advisory Group for Clinical Trials ()
EDITOR—With reference to the paper by Lilford et al, successful implementation of tracker trials would require development of a more flexible approach to research not only by the medical profession but also by prospective participants and the commercial sector.1 The public could well be attracted to the proposition of methodically evaluated introduction of new technologies as well as skill in their use, particularly during the learning curve, following growing awareness of such problems through media coverage of, for example, the Bristol case cited by Lilford et al.
This proposal for overlap of audit and trial may be an ideal opportunity not just for flexible research but for flexible consent procedures where the current notion of trial participants being guinea pigs could be turned on its head. The public is coming to appreciate that it is those patients who are the subject of poorly monitored interventions who, in retrospect, are the guinea pigs. The medical profession acknowledges that patients in trials do better—for whatever reasons. A rigorous, standard continuous process of trial and review, discarding the inferior intervention and identifying the poor performer, would be a demonstration of the constant striving for improvement through research (rather than “breakthroughs”) that would surely serve to create a new attitude and a more positive general understanding of the striving for clinical excellence.
If any notion of imposition is to be avoided it is essential to involve potential participants, namely the general public, in consideration of this new approach at an early stage. As stakeholders in the NHS, patients have a vested interest in such methods that weed out ineffective treatments by continuous evaluation. Their contribution as active partners on steering committees providing the users' viewpoint is essential in the constant iterative learning process that a tracker trial would constitute.