Article Text


Epidemiology poster session 6: Preventive intervention
P1-S6.03 Why are so many of our biomedical and behavioural prevention trials failing?
  1. D Celentano,
  2. C Beyrer
  1. Johns Hopkins Bloomberg School of Public Health, Baltimroe, USA


Background With a few exceptions, most biomedical and behavioural randomised, controlled prevention trials have failed to demonstrate efficacy. The recently reported iPrEx Trial on pre-exposure prophylaxis, the Thai prophylactic vaccine trials and the CAPRISA 004 trial of a female-controlled microbicide demonstrated only modest protection from HIV acquisition in high-risk populations, although both attained statistical significance. Understanding why most trials fail is important to designing trials that may have greater success in the field.

Methods We conducted a desk review of all Phase III biomedical and behavioural prevention trials reported in the past 10 years, including vaccines, microbicides, barrier methods, circumcision and behavioural prevention interventions, and assessed possible factors associated with success and failure to achieve anticipated outcomes.

Results Of the hundreds of prevention trials conducted over the past decade, only six reported significant findings, of which three were male circumcision trials, with one each for prophylactic vaccines, Truvada for PREP, and TDF+FTC microbicide gel. Most biomedical prevention trials rely on patient reports of adherence to use, and generally adherence is significantly lower than advised. Condom use among treatment arms are also usually lower than in control arms, which when linked with insufficient product use, leads to increased risk of disease acquisition. Repeated behavioural prevention counselling, HIV testing, STD detection and treatment all lead to diminished reports of risk behaviour among both intervention and control participants, diminishing the statistical power of the planned trials. IRB requirements may lead to control conditions that provide prevention services dramatically greater than is involved in “usual care”. RCTs which randomise individuals can also lead to significant cross-exposure among treatment arms.

Conclusions Alternative to the RCT Conclusions: Design alternatives to the individual or community randomised controlled trial may be required in the future. Increased attention to improving adherence to recommended use of prevention approaches which must repeatedly conducted (eg, use of a microbicide before each sex act), and alternatives to enriching control conditions above that of usual care should be considered, while maintaining fidelity to ethical conduct of research.

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.