Tanu's Research Blog

Estimation of hazard of HIV-1 infection in vertically transmitted children using parametric and semiparametric model

In epidemiological studies, the analysis of event-time or survival data aims to describe the risk (hazard) function of event times in a population, the associated survival or cumulative incidence functions, and the effects of covariates on risk. Doubly interval-censored data arise in many studies of disease progression in which the initial event represents disease onset and the end event corresponds to some subsequent event such as death.This type of sampling scheme can result when the originating event is not directly observable but is detected via periodic screening.

For example, it occurs in the study of human immunodeficiency virus (HIV) infected haemophiliacs whose stored blood samples are screened for evidence of infection with HIV, the virus that causes acquired immunodeficiency syndrome (AIDS). In these studies, the induction period between infection with HIV and the failure event, onset of AIDS or risk of death, may be doubly censored.A variety of models have been developed for interval-censored data.

In fact, we applied a semiparametric model based on a method proposed by Kim, De Gruttola and Lagakos to the 130 vertically HIV-1-infected cases visiting the Paediatrics clinic in RML Hospital, New Delhi. In addition to Kim's model, we applied a parametric model viz. log-linear model.

Posted at 11:39PM Jun 18, 2008 by Tanushree Banerjee in Research  |  Comments[0]