Austin-based ClosedLoop.ai, a healthcare data science platform, announced Friday it won the Centers for Medicare & Medicaid Services (CMS) Artificial Intelligence Health Outcomes Challenge. The healthcare-focused AI challenge began in March of 2019 with the winner announced in late April. ClosedLoop won $1.6 million after competing against 300 other organizations.
The CMS challenge asked participants to create “explainable artificial intelligence solutions to help front-line clinicians understand and trust AI-driven data feedback,” according to a news release. The final product needed to demonstrate how AI could predict unplanned hospital visits and adverse events, a $200 billion problem that impacts nearly 32 percent of Medicare beneficiaries.
ClosedLoop’s AI-powered platform currently helps doctors predict health outcomes, keep patients healthy and find sparse resources. The platform impacts more than three million patients each day, according to the release.
“Our Patient Health Forecasts (PHF) were key to winning the challenge,” Dave DeCaprio, co-founder and CTO of ClosedLoop, said in a statement. “We reimagined the entire concept into a comprehensive and personalized risk forecast that could be delivered directly into a clinician workflow.”
Each PHF highlights and explains how key variables can impact a patient’s risk. By showcasing relevant clinical information and specific interventions to physicians, ClosedLoop’s forecasts can help reduce unnecessary costs, improve outcomes and prevent adverse events. This updated PHF version is in private beta for existing ClosedLoop customers and select partners.
“For many physicians and nurses, we learn by case studies or ‘anecdotes,’” Dr. Jim Walton, Genesis Physician Group CEO and ClosedLoop customer, said in a statement. “When we see a patient on a particular medication, or with a particular diagnosis, we are often reminded of a unique case study or anecdote of a previous patient. Those mental images can bias us, without question, as busy clinicians don’t have time to review years of patient data to help us understand if our mental image is accurate for a patient with a similar set of symptoms.”
“ClosedLoop’s Patient Health Forecast, if we can learn to trust it,” Walton continued, “can help reverse this tendency, and say to all clinicians interested in practicing excellent care: ‘Trust the machine to help you find the relevant information — the needles in the haystack — and disabuse you of your anecdotal reflex.’”