Article Text

Download PDFPDF

HIV testing to the test: does HIV testing promote HIV prevention in HIV-uninfected adults?
  1. Nora E Rosenberg1,2
  1. 1Department of Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, USA
  2. 2University of North Carolina Project, Lilongwe, Malawi
  1. Correspondence to Dr Nora E Rosenberg, Research Assistant Professor, University of North Carolina Project, Tidziwe Center, Private Bag A-104, Lilongwe, Malawi; Nora_Rosenberg{at}

Statistics from

In their article ‘How do HIV-negative individuals in sub-Saharan Africa change their sexual risk behavior upon learning their serostatus? A systematic review’ Ramachandran and colleagues address a critical public health question.1 They review articles that compare the sexual behaviours of HIV-uninfected adults before and after HIV testing to assess whether HIV testing lowers HIV acquisition risk. They include articles from Southern and Eastern Africa from 2003 to 2013, the first 10 years of the President's Emergency Plan for AIDS Relief, when HIV testing and the larger HIV response were being scaled up in the region.2

It is remarkable that there is a need for this review—that in 2016 the answer to this seemingly simple question is not known. Several reviews addressed this question in earlier periods and found that HIV-uninfected persons who test alone do not consistently change their behaviour, but HIV-uninfected persons in HIV-discordant relationships who undergo couple HIV counselling and testing (CHCT) do change their behaviour, especially increasing their condom use.3–5

In Ramachandran's review, the new time period is meaningful. This period represents a change in the volume of HIV testing and also the paradigm of HIV testing—from stand-alone, client-initiated, opt-in models to clinic-based, provider-initiated, opt-out models. Simultaneously, life-saving antiretroviral therapy was becoming widely available, changing the face of the epidemic from a terminal illness to a manageable chronic disease. Would these new models of HIV testing, which de-emphasised counselling, result in less behaviour change than HIV testing from the previous periods? Would better HIV prognoses diminish motivation to remain uninfected among those who tested HIV-negative?

In spite of the new context and paradigms, the findings remain the same: HIV testing does not consistently lead to behaviour change among HIV-uninfected persons who test alone. For many behaviours, the direction of behaviour change is inconsistent and the magnitude is modest. For example, the six studies assessing consistent condom use range from a 10.6% decrease to a 7.6% increase. The four studies assessing abstinence range from a 5.3% decrease to a 10.9% increase. And the change in the number of sex acts ranges from a 15.7% decrease to a 9.4% increase.

However, just as in the previous period, HIV-uninfected persons in HIV-discordant couples undergoing CHCT display substantial behaviour change, especially with respect to unprotected sex. In a Ugandan cohort of HIV-discordant couples, 40% initiated condom use following CHCT, and this behaviour was strongly associated with lower HIV incidence.6 In a second Ugandan HIV-discordant couple cohort, a sexual behaviour risk score decreased by 100% following CHCT.7 Among HIV-discordant couples enrolled in two large biomedical HIV prevention trials, unprotected sex in the last month decreased from 30% to 7%8 and 27% to 9% following CHCT and ongoing testing.9 These findings are not emphasised in the review. Instead, the authors highlight modest increases in the proportion reporting unprotected extramarital relationships,6 ,8 and an increase in the number of unprotected sexual acts with outside partners.9

The key question for CHCT at a population level is whether riskier behaviours with new outside partners undermine the safer behaviours with existing HIV-positive partners. This is unlikely. One study directly assesses this question and finds that following CHCT, HIV incidence is the same among those becoming sexually active with new outside partners (3.0 cases per 100 person-years), and those remaining sexually active with their HIV-positive partners (2.9 cases per 100 person-years).8 Notably, these post-CHCT incidence rates are considerably lower than HIV incidence rates in HIV-discordant couples, who have not undergone CHCT at all (10–15 cases per 100 person-years).10

The observation that HIV-uninfected persons display minimal behaviour change when testing alone, but substantial behaviour change when testing with HIV-positive partners led WHO to issue guidance recommending CHCT, as well as research into increasing CHCT demand and use.11 Since then, a number of studies have assessed strategies to increase CHCT use in sub-Saharan Africa. For example, in an antenatal clinic in Malawi, three-quarters of HIV-infected pregnant women who initially presented alone were able to return with male partners for CHCT when supported by an invitation and a male partner tracing intervention.12 In a Kenyan antenatal setting, home visits by an HIV counsellor led to 85% uptake of CHCT.13 In Rwanda and Zambia, network leaders and agents of change were able to influence thousands of persons to seek CHCT.14 ,15

The period under observation, the first decade of the intensive HIV response in sub-Saharan Africa, reflects one of the largest public health victories in world history. Millions of people started life-saving antiretroviral therapy following tens of millions of HIV tests.1 Unfortunately, for the largest group of testers, HIV-uninfected persons receiving their HIV test results alone, HIV testing had little impact on risk behaviours. In this next decade, the new challenge will be to ensure that those who test HIV-negative remain HIV-negative. Supporting the scale-up of CHCT to identify HIV-discordant couples will play an integral role in addressing this challenge.


View Abstract


  • Funding National Institute of Mental Health (K99MH104154-01A1).

  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

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.

Linked Articles