Discovering Anomalous Patterns of Care Using Health Insurance Claims

• Activity Number: 284 - Learning Individualized/Sub-Group Treatment Rules in Complex Settings
• Type: Topic-Contributed
• Date/Time: Wednesday, August 11, 2021 : 1:30 PM to 3:20 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Sriram Somanchi* and Edward McFowland III and Daniel B Neill
• Companies: University of Notre Dame and University of Minnesota and New York University
• Keywords: Causal Inference; Heterogenous Treatment Effect; Anomalous Pattern Detection; Healthcare

Abstract

This work provides a methodology to identify subpopulations for whom certain patterns of medical care have led to significantly anomalous health outcomes. We provide a general framework to identify these anomalous care patterns and provide an empirical analysis using health insurance data. We detect interventions for patient care (currently in terms of medications) that have significantly affected health outcomes either negatively (in which case they may represent suboptimal care that should be identified and corrected) or positively (in which case they may represent new, previously undiscovered best care practices). We hope that this work will eventually lead to better patient outcomes as measured by various factors such as the number of future hospital visits, length of stay in the hospital, fewer complications due to additional secondary diseases, etc. This will further help both in terms of improving patient health and reducing health care costs. This work's methodological contributions are developing novel machine learning methods to identify effective treatments for specific subpopulations from observational data. This is an important and challenging problem that i