Innovations to Assess Program Attribution and Calculate Return on Investment in Large-Scale Health Programs

• Activity Number: 560 - Using Large Healthcare Databases and Modern Statistical Methods to Impact Health Policy
• Type: Topic Contributed
• Date/Time: Wednesday, July 31, 2019 : 2:00 PM to 3:50 PM
• Sponsor: ENAR
• Author(s): Jiaqi Li* and Ping Yu
• Companies: Booz Allen Hamilton and Booz Allen Hamilton
• Keywords: Health Policy; Market Share Analysis; Overlapping Program; Attribution ; Evaluation; Causal Inference

Abstract

A major challenge in evaluating large-scale healthcare quality improvement programs has been how to attribute observed outcomes to a given program when multiple national programs and policies with similar goals and activities are in action simultaneously. A simple difference-in-difference analysis often does not estimate program’s impact accurately in a complex healthcare environment saturated with quality improvement programs. We propose an innovative method to consider healthcare quality improvement initiatives as a market and evaluate each program as a share of the total market. Our proposed methodology is based on existing multiple treatment causal inference estimation by Lechner 2001, Cuong 2008, Bradley and Migali 2011, and Lopez and Gutman 2017.