Mentored Scientist Award

Atrial Fibrillation in HIV Infection

Award date
Award cycle
Award amount - Direct


Objective: The purpose of this study is to investigate whether there is an association between HIV infection and atrial fibrillation (AF), the most common cardiac arrhythmia. Despite the importance of AF and its associated morbidity and mortality, the incidence, treatment and outcomes of AF in the setting of HIV infection have never been formally studied.

Design: We will perform a retrospective cohort study examining the rate of developing AF in HIV-infected compared to HIV-uninfected veterans. Additionally, we will investigate the association of AF and stroke within an HIV-infected population and compare the strength of association of AF with stroke in HIV-infected versus HIV-uninfected veterans. The target population of the study will consist of a cohort of all HIV-infected patients in the Clinical Case Registry (CCR) database of the Veteran's Health Administration (VA) documented with HIV between January 1, 1984 and December 31, 2009. A contemporary control group of HIV-uninfected controls from the entire VA health care system will be specified to compare predictors and outcomes of interest using the VA Information Resource Center (VIReC).

Duration of Study: Aim 1 will examine the association of HIV infection with incident AF, while Aim 2 will investigate the association of AF with adverse health complications such as stroke among veterans with HIV infection. Both Aims will be studied over 1 year, the duration of this proposal.

Statistical Analysis of Data: For time-to-event analyses, we will use Kaplan-Meier survival curves and the log-rank test for univariate analyses and Cox proportional hazard regression for multivariate analysis. Certain covariates (hypertension, diabetes, congestive heart failure, etc.) will be modeled using the last-observation carry-forward principle. Other variables, such as lab measurements, will be modeled as time dependent covariates. Overall adjusted hazard ratios will be reported.