Skip to content

Code supplementing AI for Good Lab's work on long COVID sequelae using EHR data

License

Notifications You must be signed in to change notification settings

microsoft/causal-impact-long-covid

Repository files navigation

Introduction

Code supplementing AI for Good Lab's work on long COVID sequelae using EHR data

This project aimed to discover long COVID sequelae (symptoms) from Electronic Health Records (EHR) data using causal impact analysis of time series and analyze its association with social determinants of health The goal is to understand symptoms of long COVID and how it affects underprivileged communities for a better healthcare. This aligns with AI for Health initiative pillar of Microsoft.

Getting Started

Please use environment.yml file with conda for installing the required python packages

Notebooks

CausalImpactLongCOVID: shows the steps to discover long COVID sequelae from aggregrated time series data of symptoms of EHR

DescribeDemographics: shows the analysis of demographics (social determinants of health, SDOH) in association with long COVID

In addition to the code snippets, the notebooks also contain additional results, such as long COVID sequelae with False Discovery Rate (FDR) adjustment

This repository accompanies the paper titled "Using data science and a health equity lens to identify long-COVID sequelae among medically underserved populations."

About

Code supplementing AI for Good Lab's work on long COVID sequelae using EHR data

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published