Introduction to the Algorithm
The Health Equity Data Sandbox is a project that was born out of a Covid-19 pandemic experience when a touchless vital sign triage tool turned out to be ineffective on non-white patients. Homogenous data is a barrier to equitable innovation. Health DesignED: The Acute Care Design + Innovation Center in the Department of Emergency Medicine designed the Health Equity Data Sandbox with intentionally diverse data sets to bring together nontraditional data with healthcare data. People’s lived experiences are diverse and not easily captured by an electronic health record, the census survey, or even an individual's latest tweet or post to Instagram.
Application Roadmap
The sandbox will provide a centralized and secure environment for the discovery of knowledge and low-hypothesis exploration by combining datasets that are traditionally siloed and difficult to navigate independently, like patient encounter data, labs, claims; publicly available social determinant of health data like the American Community Survey (ACS) and Labor Force Statistics from the Current Population Survey; and mobility and IoT (internet of things) data. Insights, products, and tech enabled services will be generated from the sandbox to further the understanding of healthcare’s greatest equity challenges and drive the development of equity centered digital health solutions. This work is done with the support of the Community Outreach and Health Disparities Collaborative (1, 2).
Design and Implementation
The RADS2 team is involved in the technical implementation details for the project. We are creating the database schema and loading the data from external sources and Emory's EHR data. The data is being curated and loaded into Postgres tables on AWS. The data will be refreshed annually, in sync with the refresh of external data sets.