6.S963 Final Paper 07/01/24

Priya Bhirgoo
Siddhant Agrawal
Massachusetts Institute of Technology

Abstract:

Healthcare is an arena where the correlation vs. causation debate carries profound implications, influencing patient outcomes, public health policies, and medical advancements. The quest to discern causal relationships in healthcare—whether determining the impact of smoking on cancer risk, efficacy of a novel treatment for a disease, or evaluating the benefits of health insurance—underpins much of modern medical and health economics research. This paper reviews the spectrum of causal modeling approaches, from traditional Randomized Controlled Trials (RCTs) and observational techniques to innovative causal machine learning methods, showcasing their pivotal roles in shaping healthcare insights and policies. While RCTs remain the gold standard for establishing causality and methods using observational data require scrutiny, it is crucial not to dismiss the value of approaches based on observational data.

Full Paper

Disclaimer

This paper was written for Alfred Spector’s MIT Spring 2024 course 6.S963 Beyond Models – Applying Data Science/AI Effectively. It has not been peer-reviewed, and it may contain errors.