Stanford Medicine is leveraging Komodo Health's big data to study COVID outcomes

In a new partnership, Stanford Medicine is tapping into health tech firm Komodo Health’s platform to conduct research on COVID-19's health impacts.

Researchers at Stanford’s Center for Population Health Sciences will use Komodo’s Sentinel application, built on the company’s de-identified data of more than 330 million patients, to study population health. Though initial research will focus on short- and long-term COVID-19 outcomes through disparities in care and social determinants, it hopes to expand to span more broadly infectious disease, pediatrics, surgery, obstetrics and gynecology. 

“This collaboration with Stanford is key to accelerating innovation,” Ivy Weng, M.D., head of clinical development and real-world evidence at Komodo Health, said in an interview. “We both have aligned goals to better understand disparities and epidemiology.” Stanford is the first academic institution to partner with Komodo to conduct research specifically on COVID-19.

The health tech platform sees its data as uniquely comprehensive and representative of specific populations. Academic partnerships are key to improving clinician training and informing healthcare policy, Komodo said in a press release. 

Similarly, Stanford’s ability to conduct this research is enabled by Komodo’s data. Traditional methods of research using other data sets can take years. But Komodo’s data have already been cleaned and are more comprehensive, including big parts of the population typically missed in other data.

“It allows a much more rapid response to really important pressing medical and public health questions,” David Rehkopf, associate professor of epidemiology and population health at Stanford, said in an interview. Given how much urgency there is around better understanding long COVID, Stanford is aiming to have findings out in six months. 

“It’s also such a complicated, heterogeneous condition,” Rehkopf said. “Having this large size of data will really enable us to develop really specific phenotypes for different types of patients.”