TUESDAY, June 19, 2018 (HealthDay News) -- A machine-learning algorithm can predict hypotension during surgery based on high-fidelity arterial pressure waveform analysis, according to a study published online June 11 in Anesthesiology.
Feras Hatib, Ph.D., from Edwards Lifesciences Critical Care in Irvine, Calif., and colleagues applied machine learning to arterial pressure waveforms and create an algorithm to predict hypotension. Development of the algorithm relied upon a retrospective cohort of 1,334 patient records (545,959 min of arterial waveform recording and 25,461 episodes of hypotension). Validation used a prospective, local hospital cohort consisting of 204 patient records (33,236 min of arterial waveform recording and 1,923 episodes of hypotension).
The researchers found that the algorithm predicted arterial hypotension 15 minutes before a hypotensive event with a sensitivity of 88 percent and specificity of 87 percent (area under the curve, 0.95); 10 minutes before with 89 percent sensitivity and 90 percent specificity (area under the curve, 0.95); and five minutes before with 92 percent sensitivity and 92 percent specificity (area under the curve, 0.97).
"Physicians haven't had a way to predict hypotension during surgery, so they have to be reactive, and treat it immediately without any prior warning. Being able to predict hypotension would allow physicians to be proactive instead of reactive," a coauthor said in a statement. "By finding a way to predict hypotension, we can avoid its complications, which can include postoperative heart attack and acute kidney injury that can lead to death in some cases."
Several authors disclosed financial ties to Edwards Lifesciences, which sponsored the study.
Abstract/Full Text (subscription or payment may be required)