CMS names AI Health Outcomes Challenge winner, runner-up

May 4, 2021

The Centers for Medicare & Medicaid Services (CMS) announced that ClosedLoop.ai was selected the winner and Geisinger the runner-up in the agency’s Artificial Intelligence (AI) Health Outcomes Challenge.

This multi-stage competition, operated by the CMS Innovation Center in collaboration with the American Academy of Family Physicians (AAFP) and Arnold Ventures, began in 2019 with the aim of accelerating development of AI solutions for predicting patient health outcomes for Medicare beneficiaries for potential use by the Innovation Center. The competition has significantly broadened the Innovation Center’s knowledge of AI and its potential applications, which will help the CMS Innovation Center better design and implement future models as part of its mission to test innovative payment and service delivery models.

Out of more than 300 initial entries, ClosedLoop.ai was chosen as the grand prize winner and will receive up to $1 million in prize money. Based in Austin, Texas, the company is a leader in developing healthcare data science solutions. Challenge Runner-Up Geisinger, a non-profit healthcare system based in Danville, Pennsylvania, will receive up to $230,000 in prize money.

The CMS AI Health Outcomes Challenge attracted innovators from all sectors – not just from healthcare – to harness AI solutions to predict health outcomes for Medicare beneficiaries. From the initial group of more than 300 entries, the Challenge progressed through several stages, and participants were narrowed down to the top 25 and then seven finalists.

ClosedLoop.ai and Geisinger were among the seven finalists chosen in October 2020 to take part in the final stage of the competition. In this last stage, the participants further refined the solutions that they developed in the prior stage to help predict unplanned hospital and skilled nursing facility admissions and adverse events, and additionally developed predictive algorithms to identify beneficiaries at risk of mortality in 12 months. The finalists worked to address sources of bias in their solutions that could have the potential to affect health disparities in their submissions. The finalists also demonstrated how their solutions would easily explain the predictions to clinicians.

“Avoidable hospitalizations and skilled nursing facility stays are bad for patients and make our healthcare system costlier and less sustainable for everyone,” said Mark Miller, executive vice president of health care at Arnold Ventures. “We are eager to see how the winners of the competition use new data approaches to identify solutions to improve care in Medicare."

Each finalist in the AI Health Outcomes Challenge exhibited superior performance on at least one required element of the competition. ClosedLoop.ai and Geisinger were distinguished by their consistent strong performance across all competition elements while generating the best prediction accuracy results.

To select the winner and runner-up, CMS conducted a rigorous evaluation process, supported by a team of AI scientists. Clinicians from the AAFP reviewed and scored the explainability element. Submissions were reviewed and the winners selected by a panel of CMS senior leadership. Of the $1.65 million in total prizes to participants, Arnold Ventures will contribute up to $300,000 and the AAFP is contributing up to $340,000.

CMS has the release.