A machine learning model accurately predicted the risk of about 3 in 4 hospital-acquired pressure injuries, according to a new study.
Every year, about 2.5 million people develop a pressure injury, and 60,000 die from them in the U.S., according to past research. This condition annually costs health systems more than $26 billion, but if a 500-bed hospital used this predictive AI model, $18 million could be saved, according to the study.
Multiple machine learning methods were tested among about 35,000 EHRs spanning two academic hospitals over five years. The EHRs encompassed hospitalized patients at risk for pressure injury.
"Pressure injury prevention is a costly protocol to implement on a daily basis, and the existing tool for predicting pressure injuries is barely better than a coin flip," the study's lead author, William Padula, PhD, said in a news release. "We thought, there's got to be a better way of doing this. The question became, 'Could a computer do these risk assessments better than the nurses themselves at the bedside?'"
The model increased the prediction accuracy to nearly 75% — or a 20% improvement compared to current methods, including the Braden Scale.
The researchers hailed from the University of Southern California in Los Angeles, Johns Hopkins University in Baltimore and University Hospitals Cleveland Medical Center. Findings were published April 9 in BMJ Open.