Tria Federal Datathon Transforms Open-Source Data into Healthcare Insights

We believe in the transformative power of data to shape better healthcare policies and improve population health outcomes. In alignment with this mission, Tria Federal’s Data Practice team recently hosted “Data for Good,” an internal datathon that gave employees hands-on experience using open-source data to tackle real-world healthcare challenges.

Why a Datathon?

Tria’s Data Practice team initiated the datathon with a clear vision: to inspire participants to define a healthcare problem, explore open-source datasets, and derive meaningful insights. Working with de-identified health data, participants had the opportunity to uncover trends and craft solutions to some of today’s most pressing healthcare challenges.

Our straightforward, accessible approach encouraged broad participation across various expertise levels. We focused on proprietary tools and fundamental and exploratory data analysis techniques, leveraging familiar platforms like Microsoft Excel, Google BigQuery, and Python, as well as advanced analytic platforms like Snowflake Streamlit.

“The goal was not just to crunch numbers but to draw patterns, identify correlations, and tell compelling data-driven stories that could drive impactful decision-making,” said datathon organizer Nikita Patel , a senior data quality analyst at Tria.

Making Data Work for Everyone

To set participants up for success, Patel kicked off the datathon with an introduction to open-source datasets, data analysis techniques, and analytical tools. Attendees were encouraged to start with a problem statement, identify relevant data sources, and select the tools they were most comfortable using.

“By opening Data for Good to all employees—with any title, experience or background—we ensured that anyone with a passion for data could participate,” Patel said. “We focused on practical statistical techniques, such as pattern recognition and relational analysis, making it easier for participants to translate their findings into actionable solutions.”

The following entries were recognized for displaying creativity and analytical skills:

🥇 First Prize: How RSV Inoculation Influences ER Visits

The winning entry examined whether maternal respiratory syncytial virus (RSV) vaccinations correlated with a reduction in emergency department visits over time. Using Microsoft Power BI, the employee applied the Shapiro-Wilk test for normal distribution and Kendall’s tau for correlation analysis, offering valuable insights that could inform future vaccination strategies.

🥈 Second Prize: Key Drivers of Social Determinants of Health

This team used Tableau for exploratory data analysis and Python for machine learning techniques—including multivariate linear regression, decision trees, random forests, and gradient boosting—to identify key drivers impacting social determinants of health. Their findings offered a fresh perspective on how societal factors influence healthcare outcomes.

🥉 Third Prize: Cost and Quality Performance in the ACO REACH Model

Using scatterplots in Excel, this analysis assessed whether Accountable Care Organizations (ACOs) were meeting both cost and quality performance goals. By identifying correlations in the data, the employee provided valuable insights into how ACOs can optimize their healthcare delivery models.

🎖 Innovative Entry: Correlation Between COVID-19 Infections and Deaths in Nursing Homes

Addressing a critical public health issue, this entry analyzed the relationship between COVID-19 infections and mortality rates in nursing homes. The individual utilized Tableau line charts to visualize trends, shedding light on an area that continues to be a focal point in population health research.

Expanding the Datathon Initiative

The success of Data for Good reaffirmed Tria’s belief that data can drive meaningful change in healthcare. Future events will continue to encourage collaboration, explore real-world data, and provide opportunities to partner with industry and federal stakeholders to amplify impact.

We extend our deepest gratitude to datathon participants for their enthusiasm, hard work, and innovative thinking. The insights they generated have the potential to influence healthcare policies and strategies in meaningful ways.