A parametric survival model with an interval-censored covariate

Area of Science:

• Biostatistics
• Epidemiology
• Survival Analysis

Background:

• Injecting drug use is a significant risk factor for Human Immunodeficiency Virus (HIV) infection.
• HIV infection can lead to Acquired Immunodeficiency Syndrome (AIDS), a critical stage of the disease.
• Understanding the time intervals between these events is crucial for public health interventions.

Purpose of the Study:

• To develop and apply a parametric survival model incorporating interval-censored covariates.
• To investigate the association between the time from first injecting drug use to HIV infection and the subsequent AIDS incubation period.
• To analyze data from a cohort of injecting drug users in Badalona, Spain.

Main Methods:

• Development of a novel parametric survival model.
• Inclusion of an interval-censored covariate representing the time to HIV infection.
• Utilizing a numerical solver and the AMPL mathematical programming language for likelihood function maximization.
• Handling doubly-censored data for time until AIDS onset.

Main Results:

• The study successfully implemented a specialized survival model for complex time-to-event data.
• The model allowed for the analysis of associations where key events are not precisely observed.
• Numerical optimization techniques were effective in maximizing the likelihood function.

Conclusions:

• The proposed parametric survival model is suitable for analyzing interval-censored data in epidemiological studies.
• The methodology provides a framework for understanding disease progression in populations with risk factors like injecting drug use.
• Further application of this model can enhance insights into HIV/AIDS natural history and inform prevention strategies.