Precision medicine and causal inference – beyond the average treatment effect

Precision medicine and causal inference – beyond the average treatment effect

Epidemiologists often contrast the effects of two (or few) static treatment policies on an outcome to identify an average treatment effect for a population. However, patients are increasingly being promised personalized treatment plans. The fundamental problem of causal inference states that we cannot determine individual causal effects, so how shall we move forward in the era of precision medicine? This symposium will revisit epidemiologic concepts of interaction and effect modification (the assumption that effects are not homogeneous) and challenge attendees to consider the implications and nuances of their choice of effect measure and scale. Furthermore, it will present methods for defining interpretable treatment policies and estimating effects that vary by patient subgroup or by internal, dynamic patient characteristics. The discussant will tie together the major themes of the talks and connect them to current research and trends in the field of epidemiology.

Session Chair:
Catherine Lesko
, Johns Hopkins Bloomberg School of Public Health

Presenters:
Michele Jonsson-Funk, University of North Carolina
Yichi Zhang
, North Carolina State University
Maya Petersen, University of California, Berkeley

Discussant:
Tyler VanderWeele, Harvard TH Chan School of Public Health