Retention in HIV Care and Survival: Estimating Association and Causal Effect

Retention in care is important for the health and well-being of HIV-infected persons. However, we do not know which measures of retention best predict key clinical outcomes such as mortality. In addition, existing studies of retention have only established an association between retention and mortality, rather than a causal effect. Using the San Francisco General Hospital HIV clinic population, the objective of the proposed study is twofold: 1) to determine which of five commonly used retention measure best predicts mortality, and; 2) to estimate the causal effect of missed visits on mortality. This clinic cohort study will employ Cox proportional hazards analysis in Aim 1 and a marginal structural modeling approach in Aim 2 to account for time-dependent confounders. Our hope is that the findings of Aim 1 can be used to develop a prognostic model of patients at risk for mortality based on their retention in care and that the findings of Aim 2 will serve as the justification for developing and implementing retention interventions to decrease mortality. We anticipate that this study will take one year to complete.