Case Study: Duke University School of Medicine accelerates COVID-19 research and uncovers care disparities with Optum's Data Science as a Service

How Duke University School of Medicine leveraged compliant, deidentified claims and social determinants of health data to research COVID-19 treatments.

Duke University School of Medicine needed to evaluate the real‑world effectiveness of COVID‑19 treatment regimens and to uncover socioeconomic disparities in care during the pandemic. To support that work they partnered with Optum, using Data Science as a Service (DSaaS) to access large-scale, deidentified claims and social determinants of health data.

Optum delivered DSaaS in a secure, hosted environment, giving Duke access to deidentified data on roughly 140 million patients plus layered SDoH insights and granular measures (medications, length of stay, time-to-treatment). The solution removed separate regulatory compliance reviews, revealed disparities by economic stability, race and ethnicity, and accelerated researchers’ ability to build and run new models by about 75%, enabling faster, more robust analysis and benchmarking.

Duke University School of Medicine

Michael Pencina

Vice Dean For Data Science and Information Technology