Russo's research is primarily driven by the design and analysis of complex biomedical studies. He has mainly used Bayesian modeling to analyze data. Part of his work focuses on developing scalable computational methods to estimate models, such as using Variational Bayes techniques.

Russo is currently working on methods for signal detection in high-dimensional data motivated by learning the safety profiles of drugs by analyzing claims data and Electronic Health Records. Additionally, he is broadly interested in clinical trial design and analysis, particularly using Bayesian decision theory.