Matrix Medical Network has partnered with Berkeley Data Ventures to improve the availability and ease of COVID-19 testing by combining the power of machine learning and pooled testing. The new solution, driven by machine learning algorithms, will enable more individuals to be tested, more often, through strategic pooled testing.
“Frequent testing, for example at a school or workplace, is both costly and a real challenge given limited testing capacity,” said Dr. Daniel Castillo, Chief Medical Officer, Matrix Medical Network. “Our partnership with Berkeley Data Ventures combines our clinical knowledge and their machine learning expertise to tackle this problem by establishing new and smarter protocols that allow more frequent testing.”
Matrix has administered more than 30,000 coronavirus tests since the onset of the pandemic. Of those who tested positive, many were asymptomatic. These asymptomatic, and pre-symptomatic, carriers add an additional layer to the challenge of detecting infected individuals. The algorithms devised by Berkeley Data Ventures leverage key data to identify the optimal testing strategy by assessing individual prevalence and correlation risk to show who would, and who would not, benefit from pooled testing.
“Pooled testing using machine learning allows us to get on the offensive of COVID-19 testing by allowing for efficient testing at high frequency,” said Ned Augenblick, Associate Professor Haas School of Business, University of California, Berkeley. “One of the keys to the pooled testing algorithm is understanding who is more likely to be positive and separating them from those that are likely to be negative.”
Pooled testing can drastically help businesses boost testing capabilities. For example, if 20 individuals with a high likelihood of being negative were all pooled together for testing and no infection is detected, all of them are cleared using one test versus 20 individual tests. The strategy is to put people who are unlikely to be infected into a larger pool, and those that are likely to be infected into a much smaller pool, or even an individual test.
As businesses, universities and other organizations look to operate safely within the environment of COVID-19, frequent testing, along with the use of PPE and social distancing, is one of the most crucial strategies to have in place.
Matrix Medical Network is a leader in supporting the needs of at-risk populations. Matrix has worked with millions of individuals across the country to assess and help them manage their health risks through a network of more than 3,000 clinicians and fleet of mobile health clinics. The clinical services organization’s unmatched network of healthcare professionals meets individuals where they live and work to assess health and safety, identify and close care gaps and offer life-changing services that activate them to manage their own health.
Matrix has deep roots in clinical assessment and care management services that improve health outcomes, while helping clients manage risks and connect individuals to more options for the care and support they need. Matrix combines leading-edge technologies and proprietary platforms to harness the massive amounts of data captured to drive better clinical decision-making, improving outcomes and satisfaction while reducing the cost of care.
Berkeley Data Ventures was founded to help solve some of the biggest challenges in health care. The team brings years of academic experience studying behavioral economics, machine learning, causal inference and how these tools apply to health and health care as well as deep industry and government experience. Berkeley Data Ventures partners with companies trying to create truly unique solutions that generate massive value and change the health care system by bridging the gap between data analytics and strategy to build technology solutions and deploy them with measurable impact.