CMStatistics 2022: Start Registration
View Submission - CMStatistics
Title: Elaborated ontologies for causal inference in resource-limited settings Authors:  Aaron Sarvet - EPFL (Switzerland) [presenting]
Abstract: Emerging scarcity requires new policies for triaging limited resources. However, common-sense counterfactual targets are often impossible to articulate under standard causal models. We will briefly review these standard causal models and discuss their limitations. Then, to make progress, we will elaborate a general potential outcomes-based framework for evaluating the effects of strategies for allocating a fixed supply of limited resources in a longitudinal setting. We will provide non-parametric conditions that allow the identification of counterfactual outcomes from the observation of a single cluster ($n=1$) of patients, and motivate semi-parametric estimators based on likelihood ratio weights. As an illustration, we will consider the estimation of survival under counterfactual rules for ventilator triage (including both initiation and termination) in an intensive care unit over the course of a COVID-19 epidemic.