ÌÒ×ÓÊÓƵ

Title: "From Randomized Clinical Trials to Real-World Data: Using Epidemiology to Bridge the Causal Inference Gap" Abstract: Randomized clinical trials are considered the gold standard for estimating causal effects of interventions on health outcomes. Randomization ensures that treatment groups differ only in their exposure. However, randomized clinical trials are often infeasible due to ethical constraints, cost, or practical limitations. For example, pregnant people and people living with HIV are frequently excluded from trials, limiting our understanding of treatment safety and effectiveness in these populations. When randomized trials are not feasible, observational data and appropriate methods are critical to estimate causal effects. However, analyses of observational data are susceptible to bias since treatment groups often differ in ways other than their exposure. Target trial emulation is one approach to overcome these challenges. In this framework, researchers specify the protocol of the hypothetical trial they would like to conduct (the target trial), and then attempt to emulate it using observational analyses of existing data. This approach helps address key challenges including confounding, selection bias, and misclassification bias. In this talk, I will present the mathematicalproof needed to estimate causal effects from observed data, introduce target trial emulation, and share an example from my research in perinatal and HIV epidemiology.