A model-based approach to estimate the AIDS-free time distribution in homosexual men using longitudinal data

Area of Science:

• Biostatistics
• Epidemiology
• Mathematical Modeling

Background:

• Estimating time from human immunodeficiency virus (HIV) seroconversion to acquired immunodeficiency syndrome (AIDS) diagnosis is crucial for understanding disease progression.
• Longitudinal data from HIV-positive individuals presents challenges due to interval-censored seroconversion times and right-censored endpoints.

Purpose of the Study:

• To develop and apply a model-based approach for estimating the time from HIV seroconversion to AIDS diagnosis.
• To identify time-dependent covariates that influence the progression from HIV infection to AIDS.

Main Methods:

• A discrete hazard rate (DHR) model, a generalized linear model with a complementary log-log link, was proposed for interval-censored data.
• Classification trees were used for covariate screening, and imputation was employed for missing seroconversion times.
• Maximum-likelihood estimation and the jackknife procedure were used for parameter estimation and precision assessment.

Main Results:

• The final DHR model incorporated CD4% count, hemoglobin levels, p24 antigen presence, and their interaction as significant predictors of AIDS progression.
• The model successfully estimated the discrete survival distribution of AIDS-free time for the studied cohort.

Conclusions:

• The developed DHR model provides a robust method for analyzing time-to-AIDS in HIV-infected individuals with complex data censoring.
• Key infection progression markers significantly influence the rate of AIDS development, aiding in prognosis and clinical management.