Community Vulnerability Index

Background and Motivation

The unprecedented impact of COVID has had a devastating impact on the US. Unfortunately, the populations who are hardest hit were already struggling before the pandemic; this includes rural, homeless, incarcerated, immigrant, and low-income communities. When it comes to understanding the current landscape and allocating resources, it does not make sense to conduct business as usual. These unique circumstances call for a unique approach.

Many organizations and individuals want to help their communities in the most beneficial manner possible, but lack the in-house data expertise to quantitatively assess local need. We want to fill this gap.

Project Description

CVI is an open source project led by a team of Machine Learning Engineers, Data Scientists, Public Health Researchers, and Web Developers. We aim to help donors and nonprofits make better decisions by using COVID-specific data, as well as metrics designed to measure and predict areas of vulnerability. We have compiled a comprehensive dataset of public health, socio-economic, demographic, infrastructure, and COVID data at the county level for the entire US and constructed 8 literature-review backed metrics to assess a range of community risk and needs: Risk for Severe Economic Impact, Likelihood of Severe COVID Case Complications, Need for Mobile Health Resources, Lack of Information Access, Need for Food Services, Likelihood of Overwhelming the Healthcare System, Community Connectedness, Need for Mental Health Resources.

Intended Impact

This project is our main focus right now. We have completed a full beta version of the project including an initial county level dataset, 8 linear metrics, and an interactive web dashboard. We are currently conducting partner interviews and user testing to define how to best refine our dashboard to be most useful to stakeholders. Additionally, our Lit Review team has undertaken a comprehensive literature review to identify gaps in our datasets and corrections to the 8 metrics. Part of this project would be finding, cleaning, and incorporating those new datasets. We also are considering adding additional metrics and are undertaking ML studies to further improve existing metrics.

Internal Stakeholders

The data team (led by Stephanie Santo), the machine learning team (led by Savannah Thais), and the dashboard team (currently adding new lead).

Internal People Available During the Project

Our entire organization is available to help volunteers depending on which part of the project they want to focus on. In particular we highlight: - Dr. Savannah Thais: Founder and Machine Learning Researcher - Cassye Durkee: Project Manager - Diep Hoang: Software Intern - Stephanie Santo: Data Lead

Project tasks