Simulation Modeling in HIV Prevention Intervention Research
This researcher is involved in two simulation-intensive projects in the area of HIV prevention intervention research.
Project one is a benchmarking analysis that is investigating optimal ways of using data from HIV seroconverter panels to estimate Mean Duration of Recent Infection (MDRI). The benchmarking exercise is simulating large numbers of datasets to run under various models to compute MDRI. This modeling exercise requires both MATLAB and R to execute the simulation code. Computing MDRI is critical in the development of assays for determining recency of infection, and therefore to have implementable ways for assessing HIV incidence without using logistically difficult and costly longitudinal cohort recruitment in resource poor settings where HIV prevalence is high.
Project two involves a simulation investigation of predicting, using longitudinal CD4 measurements, time to reaching a treatable CD4 threshold level (i.e. a CD4 measurement at which antiretroviral treatment can be initiated). The data are from a large household survey conducted in Botswana. The researcher is generating simulated datasets to investigate the performance of the prediction methods when time of seroconversion of the HIV infected person is unknown. These investigations make use of Bayesian inferential methods which are computationally intensive, and would be assisted by computational resources capable of quickly executing a large-scale simulation exercise.