• AAFP Joins CMS to Launch Artificial Intelligence Challenge

    Goal Is Better Health IT, Improved Patient Outcomes

    March 29, 2019 03:29 pm Sheri Porter – On March 27, the AAFP, CMS and the Laura and John Arnold Foundation launched a national competition called the Artificial Intelligence Health Outcomes Challenge aimed at unleashing innovative health IT solutions to better predict health outcomes and improve health care quality.

    In a press release posted the same day, CMS Administrator Seema Verma, M.P.H., praised the initiative as "an opportunity for innovators to demonstrate how artificial intelligence (AI) tools -- such as deep learning and neural networks -- can be used to predict unplanned hospital and skilled nursing facility admissions and adverse events."

    She said AI could be a successful health care tool, but with these caveats: "It must not only enhance the predictability of illnesses and diseases, but also enable providers to focus more time with patients."

    AAFP Board Chair Michael Munger, M.D., of Overland Park, Kan., participated in the announcement ceremony at the Johns Hopkins University Applied Physics Laboratory in Laurel, Md. In remarks from the podium, he told the audience that health IT has so far failed to achieve promises made to America's physicians.

    "Electronic health records (EHRs) have not improved efficiencies in our practices, but rather have become barriers to overcome. Our members tell us this is one of the top three greatest challenges facing them in practice today," said Munger.

    Story Highlights

    The AAFP, CMS and the Laura and John Arnold Foundation recently launched the Artificial Intelligence Health Outcomes Challenge.

    AAFP Board Chair Michael Munger, M.D., said at the launch event that the Academy joined in sponsoring the challenge based on the expectation that the advancements it uncovers will lead to smarter health IT for members and, ultimately, better outcomes for patients.

    Potential participants are invited to submit applications now, and at the end of the yearlong challenge, the grand prize winner will receive a $1 million award.

    "Family medicine is a data-intensive specialty. We need health IT to provide the data at the point of care in a way that is efficient, effective and enhances patient care," he added.

    Munger outlined clear expectations from the Academy.

    "We see a future with use of artificial intelligence where raw patient notes and claims data can provide the analytics that become discrete actionable data that is able to drive clinical decisions and measure quality."

    Bottom line, the AAFP is a partner in this project because it's good medicine and good for family physicians.

    "We feel the advancements uncovered in this challenge program will lead to smarter health IT for our members and, most importantly, better outcomes for our patients," said Munger.

    Challenge Details

    According to a CMS Fact Sheet, the challenge will run for about one year and incorporate several stages:

    • During the launch stage, now through June 18, potential participants can submit online applications with information about their proposed solutions. As many as 20 applicants will be selected to advance to stage one.
    • In stage one, from summer to fall 2019, participants will design and test their proposed solutions using certain Medicare claims data sets. As many as five participants will claim $80,000 each and advance to stage two.
    • Stage two finalists will have access to additional Medicare claims data as they refine their solutions. The grand prize winner will receive $1 million; $250,000 goes to the runner-up.

    Dates are subject to change.

    Read the official public announcement about the challenge for additional details.

    Comments From CMMI Director

    At the 2019 Health Datapalooza in Washington, D.C., on March 28, CMS' Center for Medicare and Medicaid Innovation (CMMI) Director Adam Boehler said the goals of the challenge were twofold: "One is to save lives, and the second is to allow doctors to spend time with their patients."

    He noted his appreciation to the AAFP for joining forces with CMMI.

    "I was very happy to partner with a physician organization like the American Academy of Family Physicians," said Boehler, because the Academy recognizes that AI and advanced analytics humanize medicine and "allow physicians to focus on time with their patients, not on data entry."

    Boehler compared the practice of medicine in today's world with future medical care that could get a gigantic boost from AI. He pointed out that in the current system, a physician panel of 2,500 patients translates to a sample size of just that -- 2,500 patients.

    "One of the major goals of this initiative is to take the 2,500 panel and move it to hundreds of millions of interactions," he said.

    That means patients would have the benefit of AI recognition of untold numbers of previous patient interactions.

    "And we want to bring that (capacity) to mainstream physicians everywhere," said Boehler, because that advantage will "help us to save lives and change patient care.'