Testing projects using social networks effective in identifying people with undiagnosed infection

This article is more than 13 years old. Click here for more recent articles on this topic

A peer-driven strategy in which people with HIV encouraged individuals in their social networks to take an HIV test has proven to be an efficient and effective way to reduce undiagnosed infection, US Centers for Disease Control and Prevention researchers report in the June issue of the American Journal of Public Health. The approach worked better than traditional partner notification, and the HIV prevalence in people coming forward for testing was high at over 5%.

Traditional partner notification involves people with HIV (often when newly diagnosed) identifying their sexual partners, who will be informed of their possible exposure to HIV and encouraged to be tested themselves. Partners may be contacted by healthcare staff (relatively common in the US) or by the person with diagnosed HIV (more common in the UK).

The approach taken by the Centers for Disease Control and Prevention (CDC) did not involve contacting sexual partners, but instead friends and other people in a social network thought to be at high risk of HIV infection. Those who came forward for testing were then also encouraged to recruit other people in their social network for testing.



A healthcare professional’s recommendation that a person sees another medical specialist or service.


An umbrella term for people whose gender identity and/or gender expression differs from the sex they were assigned at birth.

Although managed by the CDC, the testing projects were run by nine community organisations in seven American cities (San Francisco, Washington, Boston, New York, Philadelphia, Orlando, Lafayette). Each community organisation focused the project on specific populations with high rates of undiagnosed infection, including African-American and Hispanic communities, homeless people, men who have sex with men, heterosexuals at high risk, transgender people, and injecting drug users.

The programme worked as follows:

  • HIV-positive people who were already receiving services at the community organisation were invited by staff to participate in the project and to become ‘recruiters’.
  • Recruiters were briefed on their role and coached, using role-play techniques, in how to discuss testing with people they knew.
  • Recruiters were interviewed, in order to identify people in their social network who might be at high risk of HIV infection.
  • Recruiters approached their contacts and recommended HIV counselling and testing to them.
  • Those who tested positive were referred into appropriate services.
  • Both those who tested positive and those who tested negative but whose behaviour indicated a high risk of becoming infected were invited to become recruiters and begin the cycle once more.

Incentives were used to encourage participation, but varied from project to project. Typically each recruiter was rewarded with a gift voucher or public transport pass worth about $10 for each person they recruited, and the incentive for those attending testing and receiving their results had a value of around $5.

Across the nine projects, 422 people became recruiters and 3172 people were tested. Among those testing, 177 people were newly diagnosed with HIV, indicating a prevalence of 5.6% in those testing as part of the project. This is considerably higher than the prevalence at other CDC-supported testing sites (2.4% in community health centres, 2.2% in hospitals, 1.8% at outreach settings, etc.).

Moreover, with 177 new diagnoses for 422 recruiters, this suggests that it was necessary to engage 2.4 recruiters in order to make one new diagnosis. This compares very favourably with the average of nine partner notification interviews needed to make one new diagnosis.

On average, each recruiter facilitated the testing of around seven people, although one-third did not recruit anybody at all. Recruiters who were HIV-positive or were gay or bisexual men were more likely to get people with undiagnosed infection tested than other recruiters.

Just under two-thirds of the newly diagnosed people (111 of 177, 63%) were known to have followed up referrals and attended medical care or support services.

The researchers consider that although a limited number of people came forward for testing at each site, the number of people with undiagnosed infection was high. They believe that their findings “show that a peer-driven, network-oriented strategy can be a powerful tool for identifying persons with undiagnosed HIV infection in populations that may be difficult to reach with other methods”. They consider this to be “a significant public health achievement”.


Kimbrough LW et al. Accessing social networks with high rates of undiagnosed HIV infection: the Social Networks Demonstration Project Am J Public Health 99: 1093-1099, 2009.