How would you apply causal inference techniques to observational healthcare data to estimate treatment effects?
