A simulation study on uncertainties associated with back calculation methodology.

Ravanan, R.; Venkatesan, P.

International Journal of Information Sciences and Computing; 2009; 3; 47-52.

Abstract: The Backcalculation methodology depends on correct information about incubation period distribution, infection density and AIDS incidence. The uncertainties in any of the above three components results in the uncertainty of HIV/AIDS estimates obtained by Backcalculation methodology. The incubation period of HIV is very long and is highly variable within and between groups and also sensitive to the choice of incubation distribution. The infection density also adds to the uncertainty of the projected HIV/AIDS estimates. Most of the time, the non-homogenous Poisson process assumption for HIV infection curve is violated. The backcalculation also heavily depends on AIDS incidence data over a period of time. The reporting delay and under-reporting further adds to the uncertainties. This paper attempts using simulation techniques to examine the variations associated with backcalculation estimates for several key parameters of HIV/AIDS models. This study also attempts to quantify the extend of uncertainties due to variations in the incubation period and infection curve.

Keywords: Backcalculation Methodology – Uncertainty – Incubation Period and Infection Curve

Back to List of publications / Home