As featured on

One-year-old Decode Health aims to get ahead of the outbreak

One-year-old technology startup Decode Health has developed a predictive artificial intelligence platform that identifies emerging trends in the COVID-19 outbreak, providing governments and other organizations early insights to where resources need to be allocated to provide targeted outreach among high-risk populations.

Traditionally in the chronic disease space, Decode layered its data depository of electronic health records, insurance claims and social determinants of health on top of community-level COVID-19 data to build evolving risk assessment profiles that have so far shown 90 percent accuracy in predicting future outbreaks. CEO Chase Spurlock said the technology is especially useful in identifying trends among the populations most at-risk for poor health outcomes, and that the trends his 18-person team is identifying now could be an outline of how infectious diseases impact people with chronic health conditions in the future.

“It was very clear to us early on that our chronic disease work was colliding with the pandemic,” Spurlock said, noting health disparities among Black and Hispanic communities that became evident in the early phases of the pandemic.

Spurlock said the technology is adaptable to meet the needs of any level of government, businesses and schools, and has brought on an eight-person advisory board with experience across all realms of health care, which will help shape the company’s future application of the technology. For now, Spurlock said they are focused on getting the platform in the hands of decision-makers to help them get ahead of the outbreak where testing and contact tracing has failed.

“Those areas are opportunities to really think about policy changes, social distancing, going back into phases of reopening. So it’s this predictive framework that gives us a forward-looking view instead of a backward-facing view of this pandemic,” he said.

Decode Chief Strategy Officer Julia Polk has been testing the efficacy of the predictive platform, taking its risk indications and comparing to how the virus reacts in real-time. So far, she said, the team has been able to identify new outbreaks with more than 90 percent accuracy.

“We did some early work in late April where we identified the counties in the state of Tennessee that would have the highest tiered growth rate in cases. And over the course of seven days, we predicted 31 of the 33 counties who saw that,” she said.

Spurlock said the ability to integrate newly collected data from the COVID-19 pandemic into predictive analytics platforms could change the way policymakers approach mitigating outbreaks. Instead of imposing one-size-fits-all restrictions as a reaction to outbreaks, public health officials and businesses are able to get ahead of the virus and tackle the specific needs of populations where outbreaks and bad health outcomes are most prevalent.

“The future of this is how to prepare for the next outbreak,” he said.