Programmatic Theme: Translational Bioinformatics

Abstract: Hypotension in critical care settings is a life-threatening emergency that must be recognized and treated early. While fluid bolus therapy and vasopressors are common treatments, it is often unclear which interventions to give, in what amounts, and for how long. Observational data in the form of electronic health records can provide a source for helping inform these choices from past events, but often it is not possible to identify a single best strategy from observational data alone. In such situations, we argue it is important to expose the collection of plausible options to a provider. To this end, we develop SODA-RL: Safely Optimized, Diverse, and Accurate Reinforcement Learning, to identify distinct treatment options that are supported in the data. We demonstrate SODA-RL on a cohort of 10,142 ICU stays where hypotension presented. Our learned policies perform comparably to the observed physician behaviors, while providing different, plausible alternatives for treatment decisions.

Learning Objective: Our work relates to the following listed learning objectives:

- Recognize and select state-of-the-art informatics approaches, theories, and methods relevant to translational bioinformatics (TBI), clinical research informatics (CRI), informatics implementation science, and data science;
- Formulate research hypotheses and projects related to informatics and the health data sciences based on existing datasets or on the aggregation of data from disparate sources;

Authors:

Joseph Futoma (Presenter)
Harvard University

Muhammad Masood, Harvard University
Finale Doshi-Velez, Harvard University