U.S. flag

An official website of the United States government

Algorithm Turns Medical Opinion into Code, and Now It’s Open Source

With an impending global health crisis that threatened to impact five million people across the DoD, Secretary of Defense Mark Esper approached DDS in early March for a technology solution to mitigate the spread of the novel coronavirus within our military community. Now, the base algorithm is available as open source code for anyone to use.

Leaders of pathology in all four branches had reason to be concerned: servicemen and women occupy close quarters on deployments; military families live in tight-knit communities on bases around the world; more employees report to work at the Pentagon than to any other municipal building in the world. An outbreak is a serious threat. Defense leadership reasoned that if service members could determine their symptoms early, they could limit exposing others to the virus.

We partnered with Dr. Donald Shell, Director of Disease Prevention, Disease Management and Population Health Policy and Oversight, to stand up MySymptoms.mil, an online tool to help flatten the curve.

MySymptoms.mil is essentially a translation of DoD medical opinion into code. The tool lets users anonymously self-assess the likelihood that they have contracted the virus based on a series of simple health-related questions. Users choose from a list of symptoms, pre-existing conditions, and core logic is used to calculate a score, from high to medium to low, as it relates to the likelihood of having COVID-19.

This is not the only symptom tracker available online, but it is the only one specifically designed for the Defense community. The value-added is that it’s based on expert opinions from the Defense Health Agency and the CDC.

Now we want to make that knowledge available to communities outside of the Department, too. That’s why we made the base algorithm of the tool open source for large organizations, the defense industrial base, and digital service providers to advance public health in their communities.

We will continue to maintain the tool as new knowledge of the virus is acquired. Additionally, the team is open to feedback on its tool by developers.

That’s the beauty of open source. In a time like this, we should all be sharing what we know about COVID-19 symptomatology. Open-sourcing was just the right thing to do.

MySymptoms.Mil will be showcased at the FedGeoDay 2020 virtual conference on Thursday, June 11, during a panel discussion on tech responses to COVID-19. Developers interested in using the code for their own symptom checkers can find it at npm @deptofdefense/covid19-calculator.