Michael Barbella, Managing Editor05.16.24
Study data finds CLEW Medical Inc.’s U.S. Food and Drug Administration (FDA)-cleared artificial intelligence-driven models for predicting patient deterioration were five times more accurate than alerts from a leading telemedicine system, using data from two major U.S. health systems.
The study was conducted in ICUs by Craig Lilly, M.D., vice chair of Critical Care at UMass Memorial Medical Center, and David Kirk, M.D., chief Clinical Integration Officer at WakeMed Health and Hospitals, with support from CLEW advisory board physicians and data science experts. The prediction models used in the study are a part of CLEW’s intelligent clinical surveillance platform. The study compared performance of this system to the accuracy and utility of alerts in the most widely utilized telemedicine and bedside monitoring systems.
The CLEW system, which notifies clinical staff that a patient is likely to deteriorate up to eight hours before other monitoring systems would indicate, creates opportunities for early intervention, helping to decrease complications and mortality. The study describes “the transformation of even a small number of emergent responses to physiological instability events... would be expected to meaningfully improve ICU care delivery.” CLEW’s models provide predictions for high and low risk of respiratory failure and hemodynamic instability, two of the most common conditions affecting patient status in intensive care.
With notification that a patient is at high-risk of one of these situations, caregivers have time to intervene in advance of physiological signs of deterioration. This can help them potentially avoid emergency interventions, such as the use of a ventilator for acute respiratory distress. It also supports the health system with capacity management, by providing early identification of potential bottlenecks caused by unexpected deterioration.
In addition to demonstrating its superior accuracy, the study concluded the CLEW system generated 50-times fewer alarms than other leading systems. In busy and overburdened critical care environments, this can greatly reduce alarm fatigue and the associated cognitive burden on caregivers. Fewer alarms mean fewer interruptions, which as the study notes “creates a more calm and peaceful ICU environment.” The study found that on average, 98 percent of bedside monitoring alarms (147 of 150 per patient day) were false positives. Due to its less frequent interruptions and greater accuracy, it concluded that the use of the CLEW system “is one of the few available options that improves ICU burnout syndrome by reducing unnecessary workload.”
The CLEW system is claimed to be the first of its kind cleared by the FDA as a class II medical device. It is currently in use at a number of academic and health system networks.
CLEW offers the first FDA-cleared, class II medical device, AI-based clinical predictive models for critical care as a part of its intelligent clinical surveillance platform. Its proprietary models offer an optimal balance of precision and sensitivity, which results in fewer high-risk notifications, reducing alarm fatigue amongst providers and clinicians.
The study was conducted in ICUs by Craig Lilly, M.D., vice chair of Critical Care at UMass Memorial Medical Center, and David Kirk, M.D., chief Clinical Integration Officer at WakeMed Health and Hospitals, with support from CLEW advisory board physicians and data science experts. The prediction models used in the study are a part of CLEW’s intelligent clinical surveillance platform. The study compared performance of this system to the accuracy and utility of alerts in the most widely utilized telemedicine and bedside monitoring systems.
The CLEW system, which notifies clinical staff that a patient is likely to deteriorate up to eight hours before other monitoring systems would indicate, creates opportunities for early intervention, helping to decrease complications and mortality. The study describes “the transformation of even a small number of emergent responses to physiological instability events... would be expected to meaningfully improve ICU care delivery.” CLEW’s models provide predictions for high and low risk of respiratory failure and hemodynamic instability, two of the most common conditions affecting patient status in intensive care.
With notification that a patient is at high-risk of one of these situations, caregivers have time to intervene in advance of physiological signs of deterioration. This can help them potentially avoid emergency interventions, such as the use of a ventilator for acute respiratory distress. It also supports the health system with capacity management, by providing early identification of potential bottlenecks caused by unexpected deterioration.
In addition to demonstrating its superior accuracy, the study concluded the CLEW system generated 50-times fewer alarms than other leading systems. In busy and overburdened critical care environments, this can greatly reduce alarm fatigue and the associated cognitive burden on caregivers. Fewer alarms mean fewer interruptions, which as the study notes “creates a more calm and peaceful ICU environment.” The study found that on average, 98 percent of bedside monitoring alarms (147 of 150 per patient day) were false positives. Due to its less frequent interruptions and greater accuracy, it concluded that the use of the CLEW system “is one of the few available options that improves ICU burnout syndrome by reducing unnecessary workload.”
The CLEW system is claimed to be the first of its kind cleared by the FDA as a class II medical device. It is currently in use at a number of academic and health system networks.
CLEW offers the first FDA-cleared, class II medical device, AI-based clinical predictive models for critical care as a part of its intelligent clinical surveillance platform. Its proprietary models offer an optimal balance of precision and sensitivity, which results in fewer high-risk notifications, reducing alarm fatigue amongst providers and clinicians.