Article Text


Epidemiology poster session 6: Preventive intervention: Community action
P1-S6.45 Impact of a community level, diffusion based HIV/STI intervention on heterosexually-identified, socially marginalised men in urban, coastal Peru
  1. K Konda1,
  2. A Lescano2,
  3. D Celentano3,
  4. A Maiorana4,
  5. X Salazar5,
  6. A M Rosasco5,
  7. S Montano2,
  8. T Kochel2,
  9. E Hall2,
  10. T Coates1,
  11. C Caceres5,
  12. S Kegeles4,
  13. NIMH Collaborative HIV/STD Prevention Trial Group6
  1. 1UCLA, Lima, Peru
  2. 2NAMRU-6, Callao, Peru
  3. 3Johns Hopkins University, Baltimore, USA
  4. 4UCSF, USA
  5. 5UPCH, Lima, Peru
  6. 6NIMH, USA


Objectives The NIMH Collaborative HIV/STD Intervention Trial implemented a 2-year intervention based on the theory of diffusion of innovations to decrease unprotected sex with non-primary partners among high-risk populations including heterosexually-identified, socially marginalised men in urban, coastal Peru.

Methods 20 communities in 3 cities were randomised to the intervention or comparison arm. In intervention communities, 15% of the target population was trained as Community Popular Opinion Leaders (CPOLs) to deliver HIV prevention messages to their peers. In all communities, yearly assessments of the study cohort, independent of the intervention, gathered information on sexual behaviour, provided pre-test counselling, and testing for HIV/STIs at baseline, 1 year, and 2 years after randomisation. This analysis included the 2146 heterosexually-identified men who fulfilled the trial's inclusion criteria (frequented community social venues, were aged 18 to 40, and had sex in the past 6 months).

Results Unprotected sex with a non-primary partner decreased by 10% in both study arms among heterosexually identified men. Almost half of the participants, in both the intervention and in the comparison arms, reported unprotected intercourse with a non-primary partner at the final follow-up (47.6% vs 48.7%, p=0.694) and there was no difference after controlling for baseline behaviour, OR 1.04, (95% CI 0.79—to 1.38). STI incidence was also did not differ by randomisation condition, IRR 1.14 (95% CI 0.77—to 1.68) comparing incidence in intervention vs control communities.

Discussion We speculate that the lack of effect of the intervention among the heterosexually-identified men may be attributable to 1) an insufficiently strong intervention approach, 2) this group of men did not feel vulnerable to HIV or STI, or 3) the repeated biomedical testing and ancillary required counselling were as potent as the CPOL intervention. Future HIV prevention trials with this population should try to tailor the intervention to this group's risk perceptions and behaviours.

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.