Addressing an Extreme Positivity Violation to Distinguish the Causal Effects of Surgery and Anesthesia via Separable Effects

The lecture describes a warning from the U.S. Food and Drug Administration (FDA) regarding potential neurotoxic effects of anesthetics in the third trimester of pregnancy, for which there is only limited clinical evidence. A key question is whether these effects are caused by the anesthesia, the surgery, or both. Separating these influences is methodologically challenging because they are closely intertwined in observational data.

To solve this problem, a causal model of separable effects is used to identify the isolated influence of anesthesia. This approach is applied to a large Medicaid dataset and its robustness is further tested.

Caleb Miles is an assistant professor in the Department of Biostatistics at the Columbia University Mailman School of Public Health. He works on developing semiparametric methods for causal inference and applying them to problems in medicine and public health.