Article Text

other Versions


The appropriate evaluation of HIV prevention interventions: from experiment to full scale implementation.
  1. Timothy B Hallett (timothy.hallett{at}
  1. Imperial College London, United Kingdom
    1. Peter J White (p.white{at}
    1. Imperial College London, United Kingdom
      1. Geoff P Garnett (g.garnett{at}
      1. Imperial College London, United Kingdom


        Background: Preventing HIV infection is still an essential goal in tacking the HIV/AIDS pandemic. Remarkably little is known about how best to reduce HIV incidence because the majority of trials focus on the reduction of risk behaviours and assume an effect on HIV incidence.

        Objectives: To discuss the evidence for the effectiveness of HIV prevention strategies, exploring the different types of evidence available, including individual and community randomised controlled trials and observational studies.

        Results: Whilst providing a gold standard for evidence, trials have been limited in their scope and are difficult to interpret and generalise. There have been examples of national level successes in preventing HIV which have been detected in surveillance data and understood through behavioural and modelling studies. These have the advantage of being to scale and indicating effectiveness rather than efficacy.

        Conclusions: Whilst randomised trials are important because of their scientific rigor, it is also important that evidence from observational epidemiology is not overlooked. Only if good quality, consistent data are available can the history of the HIV epidemic be appropriately analysed.

        • HIV
        • Intervention
        • Mathematical Modelling
        • Randomised Control Trials
        • Sentinel Surveillance

        Statistics from

        Request Permissions

        If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.