Pittsburgh-based UPMC has created a machine-learning platform that can accurately predict which patients are at high risk of complications after surgery, according to a news release shared with Becker's.
UPMC physicians and researchers trained the algorithm on the medical records of more than 1.25 million surgical patients, looking out for those who had major cerebral or cardiac events like a stroke or heart attack after a procedure, and validated it against 200,000 patients who had surgery at UPMC. They published their results July 7 in JAMA Network Open.
"We designed our model with the healthcare worker in mind," said Oscar Marroquin, MD, chief healthcare data and analytics officer at UPMC, in a July 7 news release. "Since our model is completely automated and can make educated predictions even if some data are missing, it adds almost no additional burden to clinicians while providing them a reliable and useful tool."
UPMC has employed the model at its 20 hospitals for about two years. Each morning, the program reads patients' EHRs and flags those who are at risk for post-surgery complications, connecting them with care coordination, prehabilitation or a referral to the UPMC Center for Perioperative Care.
The researchers say their artificial intelligence model is more accurate than the American College of Surgeon's National Surgical Quality Improvement Program, or ACS NSQIP, which requires clinicians to input data manually.