BACKGROUND: CDC’s HIV Optimization and Prevention Economics (HOPE) model has been applied to estimate the effects of reaching national goals on HIV incidence in the United States and to estimate the national HIV effective reproduction number (Re). Re, the average number of secondary infections from an infected person in a population where some persons have already been infected, is a fundamental epidemiologic concept used to study the potential spread of infectious disease. Given that the first model application focuses on the policy question of reaching national goals while the second approach estimates a theoretical epidemiologic concept, policy makers should know if the implications of the two approaches are similar.
METHODS: The HOPE model simulates the sexually active U.S. population aged 13 to 64 years stratified into 195 subpopulations by transmission category, sex, race/ethnicity, age group, male circumcision status, and HIV risk level. People transition from being uninfected into 23 infected compartments defined by disease progression and continuum-of-care stage. We first used the model to estimate the impact on HIV incidence of reaching the goals of 90% of persons with HIV having diagnosed infection, 85% linked to care, and 80% of those diagnosed achieving viral suppression by 2020. In the second application, we estimated Re over this same period by integrating subpopulation transmission rates among persons with HIV with their transition rates by disease and continuum stage to determine whether Re was significantly below its threshold value of 1.0, which would indicate the possibility of disease eradication. We then examined the effect of changes in continuum flow rates on Re.
RESULTS: In the first application, achieving and maintaining the goal of viral suppression resulted in the greatest reduction in HIV incidence. Reaching the 2020 viral suppression goal required the use of a combination of strategies: increasing the probabilities of antiretroviral therapy (ART) prescription and viral suppression and decreasing the probability of losing viral suppression after having achieved it. The importance of keeping persons on ART from becoming non-adherent was a key finding of this analysis, which included rates of infected persons transitioning from being prescribed ART to becoming virally suppressed either immediately or after a delay, and reverse transitions indicating the loss of viral suppression. In the second application, the Re estimate was slightly below 1.0, suggesting that intensified prevention efforts to improve the rates of progression along the HIV care continuum are needed to ensure disease elimination, and that decreases in prevention efforts could result in rebounding incidence trends. Preventing the loss of viral suppression after achieving it was the one intervention that could reduce the estimated Re to a value substantially below the 1.0 threshold.
CONCLUSTIONS/IMPLICATAIONS: The conclusions of these two disparate applications of the same model were similar. A focus on achieving and, perhaps more important, maintaining viral suppression among HIV-infected persons appears to yield the greatest impact on reducing HIV incidence. Policy makers can have increased confidence in the results of one modeling approach with confirmation of results by an alternative analysis.
