MUSC student used artificial intelligence to find patients at risk for COVID complications – Charleston Post Courier

When the COVID-19 pandemic forced the Medical University of South Carolina to suspend clinical rotations used to train students in health care settings, Alan Snyder had an idea.

He was third-year medical student this past spring and was in the middle of a dermatology rotation when he was sent home and required to transition to online learning. But instead of sitting back and letting the pandemic run its course, Snyder devised a new way to reach patients.

Using an artificial intelligence model developed by Jvion, a Georgia-based health care AI firm, Snyder combined Census tract information with the MUSC patient database to identify thousands of adults who faced a high risk of developing serious complications should they contract COVID-19.

But this was just the start. Snyder then mobilized more than 150 volunteers, ranging from students to retirees, across the state to call these at-risk patients and educate them about coronavirus safety protocols. Over the course of 41 days last year, these volunteers made 1,370 calls to 814 patients and were able to help more than 50 percent of "extremely high-risk patients" activate their online MyChart accounts, which are used to store electronic medical records and connect patients with their providers.

Not only that, in a few cases, the volunteers associated with the project were able to pinpoint cases of elder abuse and medical emergencies and connect patients with social workers and care almost immediately.

"Necessity brings out innovation," said Snyder, now completing his fourth and final year at MUSC. "I was scared. How could I help other people? Thats my job as a health care professional. It seems like something that was useful for my time."

Dr. Lancer Scott, the section chief of emergency medicine at the VA hospital in Charleston and a faculty member at MUSC, said when he first heard about Snyder's idea for the project, "the hairs stood up on the back of my neck."

"Find me a group in this state that is working on preventive measures on the front end of COVID instead of contact tracing on the back end," Scott said. That's what makes Snyder's idea so unique, he said.

"Its not just a model for COVID. ... It's a model for vulnerable populations," Scott said. "If there was a Nobel Prize for medical students, Alan should get it. Its really amazing."

Snyder presented his results (remotely) during the virtual American Medical Association Research Symposium in December. He is currently applying to residency programs and crunching data collecting through his project to determine if the outreach to patients did, in fact, prevent any COVID-related hospitalizations or deaths.

"There are a lot of people behind the scenes who put their heart and time into this," Snyder said. "Its something that Im really proud of."

Reach Lauren Sausser at 843-937-5598.

Follow this link:

MUSC student used artificial intelligence to find patients at risk for COVID complications - Charleston Post Courier