Predictive Models of HIV Survival

• Activity Number: 650
• Type: Contributed
• Date/Time: Thursday, August 8, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Georgiy Bobashev*+ and Jacob Norton and Olga Tousova
• Companies: RTI International and NCSU and Pavlov Medical University
• Keywords: predictive survival ; simulation ; HIV ; disease transmission ; risk ; predictive model

Abstract

Predictive modeling considers the results of association analysis as inputs and uses them to predict future outcomes. Often the risks of disease acquisition from a single event, such as an unprotected sex act, are small and are difficult to interpret. However, when risky behaviors are repeated it possible to forecast survival from disease acquisition. In our study we consider the development of survival trajectories based on risk-bearing events that can occur at different time scales. We illustrate our model using an HIV example. We consider existing estimates of HIV transmission risks through injecting and sexual contacts to develop a predictive survival model for an individual who is exposed to HIV through intimate relationships. We simulate survival curves for a number of behavioral scenarios and discuss sources of simulated uncertainty. We apply the model to a longitudinal study of HIV-discordant pairs and discuss model validation based on order statistics.