Bernoulli Chief Analytics Officer John Zaleski, PhD, CAP, CPHIMS, will be presenting “Identifying actionable alarms in OSA patients receiving opioids” at the Machine Learning & AI in Healthcare conference May 3-4, in Boston.

Dr. Zaleski will discuss the use of adjustable, multi-variable thresholds involving combinatorial alarm signals to discriminate between actionable and non-actionable alarms without increasing risks to patient safety.

“Middleware designed to collect data at variable speeds and the use of precision alarms, which harness real-time patient data and notifications from individual devices in order to identify clinically relevant trends, sustained conditions and combinatorial indications are essential to continuous electronic monitoring (CEM),” said Zaleski.

Considered a best practice by the Joint Commission, the Anesthesia Patient Safety Foundation and other healthcare advocates and agencies, CEM is typically utilized in high-acuity settings, such as intensive care and med-surg units. However, the ability to combine analysis with real-time data at the point of collection makes enterprise-wide CEM a viable opportunity. “The ability to track patients throughout the hospital, continuously add new medical devices, and distribute real-time patient data to centralized dashboards and mobile devices should be a major consideration for health system seeking to achieve real-time healthcare capabilities,” he said.

Zaleski’s session is scheduled for 10 a.m. (ET) on May 4. The conference updates and news will be available on Twitter using the hashtag #MachineLearningHC.

Follow Bernoulli on Twitter and LinkedIn. Visit our Resource Center to download case studies, white papers and articles.