COVID-19 Machine-Learning Algorithm Secures FDA Emergency Use Authorization

(Oakland, CA, USA), a machine learning diagnostic algorithm company that is targeting early disease intervention to improve patient care outcomes. The COViage system analyzes patient data from electronic health records (EHR) systems and gives healthcare providers advance notification of patients who are predicted to experience unstable blood pressure or respiratory decline requiring mechanical ventilation.

COViage was evaluated in a clinical trial that enrolled 197 patients who visited the emergency department or were admitted to the hospital at five US hospitals between March 24, 2020 and May 4, 2020. Evaluable patients had confirmed COVID-19 diagnoses and their first set of vital sign and lab measurements were taken within two hours of arrival or admission. Data were analyzed by COViage and the standard of care Modified Early Warning Score (MEWS) for comparison. The outcome of respiratory decompensation leading to mechanical ventilation, defined as invasive ventilation requiring endotracheal tube or mechanical ventilation not including BIPAP or CPAP, was assessed 24 hours after model predictions were made. The COViage algorithm achieved an area under the receiver operator characteristic curve of 87% compared to 64% by MEWS (a 36% increase), demonstrating a substantially higher sensitivity and specificity.

“COVID-19 remains a significant public health emergency both in the U.S. and around the globe, and we are encouraged that by receiving this EUA, our machine learning algorithm can help caregivers diagnose critical conditions resulting from COVID-19 earlier and more accurately,” said Ritankar Das, president and chief executive officer of Dascena. “The early identification of patients at risk of respiratory decompensation or hemodynamic instability would enable physicians to more aggressively monitor these patients in a controlled environment and provide earlier treatment.”

Related Links:
Dascena, Inc.