Cox Proportional Hazards Model with Cross-Sectional HIV Prevalence Data

This study applies the Cox proportional hazards model for failure times to the HIV-prevalence dataset collected over four years among women in Cape Town, South Africa. Given the assumption that women of different ages may have distinct HIV hazards functions that are proportional to each other, a semi-parametric approach is employed to estimate the risk parameters. The main advantage of this approach is that no distributional assumptions to the baseline hazards function are required.