Description

Overview

Data science is a multidisciplinary approach that blends techniques from computer science, statistics, epidemiology, and other domains. It often focuses on large or novel data sources and the application of artificial intelligence domains such as machine learning or natural language processing. This particular focus on large and complex data sources has the intention to improve the measurement and prevention of a host of public health issues including injury and violence.

With a dedicated focus on this emerging field, we can continue to uncover ways to improve public health assessments, interventions, and other mechanisms to improve health, safety, and well-being for our communities.

Opportunity

To elevate the utility of data science in advancing public health practice, the Centers for Disease Control and Prevention, National Center for Injury Prevention and Control, in partnership with the American Public Health Association, has developed the Injury and Violence Prevention Data Science Demonstration Project. The overarching aim of this initiative is to enhance public health capacity and provide support to organizations engaged in data science work.

For this iteration of the program, we are seeking responses from organizations with existing data science projects looking to expand their efforts. This may include projects to improve the timeliness of health information, responding to public health threats in a more efficient manner, increasing the effectiveness of prevention programs/campaigns, or other topics.

Goals

• Enhance the data science workforce and capacity
• Strengthen data science partnerships
• Increase awareness of the public health benefits from data science

RFP
* Please review the RFP before beginning your application.
Click here to download a copy of the RFP. 

Program Components

Access to Data Science Subject Matter Experts

Each site will be assigned a CDC subject matter liaison and meet monthly to address support needed or other issues related to the implementation of the project.

Funding

Each site will be awarded up to $200,000 to assist with project implementation. Final award amount will be determined by an internal team and in some instances may be less than the amount requested.

Community of Practice

There will be opportunities for sites to engage with one another to share any lessons learned and best practices throughout the duration of the data science demonstration project program period.

Learning Labs

Sites will be able to participate in professional development opportunities related to data science.

Requirements

Eligibility Criteria

This opportunity is open to organizations or institutions currently working on data science projects focused on injury and violence prevention. If various groups representing multiple organizations are partnering on a collective effort, they are also eligible to apply. There are three main areas of focus for this RFP:

1) Using natural language processing or machine learning to improve the timeliness, efficiency, and automation of injury or violence surveillance systems or programs. For example, major injury mortality data systems, such as the National Violent Death Reporting System (NVDRS) and the State Unintentional Drug Overdose Reporting System (SUDORS), experience significant delays because information must be manually abstracted, entered, or curated. Furthermore, often the information needed to complete these surveillance systems comes from multiple data sources (health, law enforcement, laboratory, etc.) which are not linked or even sometimes available in standard electronic formats.

Opportunities exist to use machine learning, natural language processing, or other data science approaches to automate and streamline processes. For example, natural language process (NLP) approaches may help to identify and abstract certain variables in a semi-automated or automated way. Applicants could propose a data science focused approach to improve injury or violence mortality data surveillance systems (such as NVDRS,SUDORS) and the state and local systems and processes which compile and abstract information for them.

2) Lags in official data sources such as death certificate data hinder public health efforts to respond in a timely way.  Novel machine learning and prediction modeling approaches to nowcast (I.e., prediction of the present state absent gold standard data) injury trends can improve timely situational awareness, surveillance, and public health decision-making. Applicants are encouraged to propose a data science-focused approach to harness real-time novel data available at the state and local level to do nowcasting work.

3) Additionally, applicants could also focus on leveraging NLP or exploring large language models (LLM) to speed up the analyses of qualitative data.

Other potential areas of focus include:

• Use of predictive modeling or machine learning to improve service delivery
• Technology based interventions to address injury or violence
• Data linkage for injury or violence prevention
• Data science work related to advancing health equity

It is recommended that all proposed projects include collaboration with state or local health departments or a relevant local public health entity. Content areas of the proposed work should focus on violence, adverse childhood experiences, suicide, and/or drug overdose.

Scope of Work/Program Requirements

At the conclusion of this program, it is expected that participants will demonstrate enhanced capacity in data science and disseminate project findings.

Program participants are committing to the following:

• Participate in monthly meetings with the CDC/APHA project team
• Submit Interim Report
• Respond to Bi-monthly brief surveys
• Participate in Learning Labs
• Produce a final deliverable (e.g. report, case study, training, ) that incorporates plain and clear communication of complex data science concepts and methods
• Submit final report

Please note: We do not provide feedback on applications not accepted for the grant award.