Measuring the Impact of the Global Response to the AIDS Epidemic

Measuring the Impact of the Global Response to the AIDS Epidemic: Challenges and Future Directions.

In the Declaration of Commitment of the 2001 United Nations General Assembly Special Session on AIDS, all Member States agreed to a series of actions to address HIV. This article examines the availability of data to measure progress toward reducing HIV incidence and AIDS mortality and discusses the extent to which changes can be attributed to programs. Lacking a method to directly measure HIV incidence, trends in HIV prevalence among 15-year to 24-year olds and groups with high-risk behaviours are used as a proxy measure for incidence trends among adults in generalized and concentrated/low-level epidemics, respectively. Although there is limited empirical data on trends in new infections among children, progress in the treatment area is tracked through indicators for the percentage of people who remain on antiretroviral treatment 12 months after initiation and the coverage of antiretroviral treatment. Successive iterations of epidemiological models using surveillance data from pregnant women and groups with high-risk behaviour and data from national household surveys, demographic data and epidemiological assumptions have produced increasingly robust estimates of HIV prevalence, incidence and mortality. Globally, incidence has decreased among adults (accompanied by evidence of changes in behaviour in several countries) and children over the past decade. The decline in AIDS mortality is more recent. On the basis of the underlying logical framework and mathematical models, it is concluded that programs have contributed to a reduction in HIV incidence and AIDS mortality. More data are needed to reliably inform trends in HIV incidence and AIDS mortality in many countries to allow an assessment of progress against national and global targets. In addition, impact evaluation studies are needed to assess the relationship between changes in incidence and mortality and the HIV response and to determine the extent to which these changes can be attributed to specific programmatic interventions.

The decline in new infections (incidence) reported in the 2009 UNAIDS/WHO Epidemic Update is not translating into lower numbers of people living with HIV (prevalence) because antiretroviral treatment roll-out is reducing mortality. Thus, there are fewer new infections and fewer deaths but the size of the HIV epidemic in absolute numbers continues to grow. This article examines the availability and nature of data to track the HIV epidemic, the modelling undertaken to estimate trends, and the importance of robust evaluations of programmes and policies to explain what has contributed, and what has not, to the changes in incidence that we are observing. Although generally we are on track to be able to know in 2015 whether the HIV epidemic has been ‘halted and reversed’ (Millennium Development Goal 6), this analysis of UNGASS reporting shows that many countries need to improve their epidemiological data collection and most countries need to conduct carefully designed evaluation studies to know whether specific programmes have contributed to any observed changes.