Box 10.1: Estimating the impact and cost–effectiveness of selected recommendations using mathematical models: results from the independent HIV Modelling Consortium
Consolidated ARV guidelines, June 2013
As HIV programme managers work to implement these guidelines, they may face complex choices on how to optimally allocate resources for HIV treatment: for example, determining the relative allocation of resources for scaling up HIV testing and linkage to care and for increasing access to ART based on expanded eligibility criteria.
The HIV Modelling Consortium, an independent group of research institutions, used multiple independent mathematical models based on data sets from four countries with different types of epidemics and current ART coverage – India, South Africa, Viet Nam and Zambia – to examine the health benefits, costs and cost–effectiveness of various strategies for expanding eligibility for ART as well as testing and access to HIV care (19) (Web Annex). In each case, a range of potential options was investigated, including various thresholds for ART eligibility (CD4 cell count less than 500 cells/mm3, all people with HIV and specific key populations), assuming current as well as expanded patterns of HIV testing and linkage to care. The costs and both individual and preventive health benefits associated with each intervention were estimated, including changes in HIV incidence, reductions in the loss of healthy life-years and costs over time. These models also considered the relative cost–effectiveness of strategies, highlighting which of them, for a given budget, would be expected to maximize health gains.
Expanding the criterion for ART eligibility to CD4 cell count ≤500 cells/mm3 was found to be highly cost-effective in low- and middle-income settings. However, combining expanded eligibility with a large increase in HIV testing and linkage to care produced the greatest benefits, especially in settings with low ART coverage. Expanding ART eligibility to all adults with HIV (irrespective of CD4 count) was less cost-effective than expanding the criterion to ≤500 cells/mm3 because of less immediate improvements in health as the CD4 threshold for initiating ART is increased.
The modelling results should be interpreted in light of some important limitations. Many of these conclusions could change substantially depending on cost assumptions, especially those related to testing and counselling and to retaining people in pre-ART care. Moreover, the models did not consider how the estimated impact and cost–effectiveness of the various interventions would change if they were combined or only partly implemented. Models also did not address potential trade-offs with non-antiretroviral interventions, and several important issues were not covered, such as treatment of children (section 6.2.4 discusses the relative cost-effectiveness of various treatment options for children).