#SER2019 plenary session with @sandrogalea @_MiguelHernan Reconciling social epidemiology and causal inference 
A big thank you to @UNCpublichealth for sponsoring this years John Cassel lecture
Paper in press regarding this work!
It is this “orderly arrangement into chains of inference” which intrigues me
Social epidemiology is concerned with the health effects of forces that are ‘above the skin’
Quantitative causal inference helps us draw conclusions about causal effects
3 misconceptions may explain the divide between social epi and causal inference
1. Social exposures are qualitatively different from other exposures and should play by different rules
2. The goal of causal inference is to identify ‘causes’
3. Causal inference requires exposures that can be experientially manipulated
Using income as an example you can think of a hypothetical experiment that will allow you to operationalize the effect of income on health
But what do we do about factors such as race that are difficult to operationalize in this experimental framework. And remember that this is not limited to social epi factors...
Bridging the gap...3 implications
1. If interested in changing the world we should prioritize actionable causal inferences
2. This helps clarify the methods that we need to clarify causal effects
3. This clarifies how we identify and measure confounders
The rest of the quote...
• • •
Missing some Tweet in this thread? You can try to
force a refresh
