Student seminar series
Link to join seminar: Hosted on Microsoft Teams
Causal Inference with Marginal Structural Models
Epidemiological and clinical studies often seek to discover the causal effect of medical or surgical exposure on a certain clinical outcome. However, the causal effect of interest may vary for patients with different backgrounds, such as ages or medical history. These factors that affect both the exposure and outcome are called confounders. It is important to select proper measures that quantify the causal effect of interest and fit appropriate models to adjust for confounding factors. This talk will present the mechanisms of Marginal Structural Models (MSM) and how it applies to one-time and recurrent events to estimate corresponding causal effect measures regarding confounders.