Bolstering Data Science Expertise at the US Centers for Disease Control and Prevention (CDC)

Posted on January 31, 2024 by Mary Catherine Bertulfo

This entry is part 4 of 16 in the series Mar 2024

To support the needs of the data science workforce, CDC established the Data Science Upskilling program to train and support learning and growth in data science among staff.

As the volume and breadth of data to support public health increases, the need for skilled public health practitioners to collect, analyze, and disseminate findings from these data has increased. The United States government has identified data science as a key workforce component. In particular, the Department of Health and Human Services (HHS) has identified this as a priority. The Centers for Disease Control and Prevention (CDC) initiated an agencywide effort to support growth and development of data science and a data science workforce through its Data Modernization Initiative. “Advancing Data Science Among the Federal Public Health Workforce: The Data Science Upskilling Program, Centers for Disease Control and Prevention” describes CDC’s efforts to improve data science literacy among its public health workforce.

Meeting these goals requires not only investment in technology and policy solutions, but also in workforce development. Recruiting highly qualified data scientists to government settings, including public health agencies, can be challenging. Government positions often cannot match compensation offered in the private sector. As a larger number of industries compete for a relatively limited pool of skilled data scientists, meeting public health workforce needs also requires upskilling the existing workforce.

CDC’s Data Science Upskilling (DSU) Program

In 2019, CDC established the Data Science Upskilling (DSU) program. DSU was designed using evidence-based best practices for adult learning. Its goals are to increase agency data science capacity and promote a learning culture regarding data and its potential. DSU’s team-based learning program has trained approximately 360 learners from across CDC and advanced 92 high-priority data science projects.

Read the Article in JPHMP

DSU is a 10-month learning-in-place program that incorporates aspects of staff’s normal work projects and schedules into program activities. Staff and fellows across CDC apply to DSU as a team with a predefined data science project (either new or ongoing) that will serve as their applied learning activity during the program. Teams are interdisciplinary and generally include 2–5 people from the same work unit. Supervisory approval for all team members is required. During the 10-month program period, the DSU cohort of teams participates in structured and unstructured learning through computer language sessions known as boot camps, massive open online courses, and technical assistance (TA) sessions with experts in domains associated with their team projects (e.g., machine learning, data visualization, and natural language processing). Each cohort culminates in an agencywide seminar during which the teams present their results.

DSU Effects

Although DSU has existed for 5 years, the program has more than exceeded its goals. It has trained diverse talent throughout CDC, including epidemiologists, behavioral scientists, medical officers, laboratorians, and fellows. Annual evaluation results indicate a high level of engagement; teams reported that DSU improved their data science knowledge and skills and their confidence in making data science decisions. A majority of DSU alumni reported continued use of acquired skills after program completion.

DSU is considered a model program across U.S. government agencies. Several aspects of the program’s design contribute to its success. DSU upskilling is incorporated into participants’ existing daily work and normal work schedules; teams complete their DSU activities as part of their daily work. Therefore, staff are not required to complete upskilling activities in their free time, which can be difficult to schedule and prioritize. Additionally, because projects proposed by DSU teams are supported by their normal work unit supervisor, leadership supports the time needed to devote to DSU. Agency leaders see value in DSU by having new or existing high-priority data science projects move forward during the 10-month program. DSU contracts with data science experts for TA. This high-quality TA is on-demand during DSU and provides teams with rapid guidance and support to keep projects moving along. Additionally, TA is provided for all project domains and is not restricted to specific data science areas for which CDC has existing subject matter expertise.

Addressing the growing data science needs of public health will require creativity and innovation. Identifying incentives for data scientists to choose public health careers is important, but the pool of trained data scientists is limited, and the needs are substantial. Upskilling is a crucial addition to recruitment to meet growing needs of public health. The upskilling model shared in “Advancing Data Science Among the Federal Public Health Workforce: The Data Science Upskilling Program, Centers for Disease Control and Prevention” can help guide other public health organizations in meeting their data science workforce goals.

About the Author

Mary Catherine Bertulfo: Mary Catherine Bertulfo is an epidemiologist and DSU Program Lead. Her experience includes public health program management and surveillance. Mary Catherine holds a BA in Women and Gender Studies from the University of South Florida and an MPH in Epidemiology from the University of Florida.