Though the widening availability of COVID-19 vaccines appears to have largely put an end to the severe bed and equipment shortages U.S. hospitals faced just a few months ago, the U.S. government isn’t taking any chances.
The Biomedical Advanced Research and Development Authority (BARDA)—a division of the U.S. Department of Health and Human Services dedicated to dealing with bioterrorist attacks, nuclear incidents and, yes, viral outbreaks—has thrown its weight behind two startups developing tools to measure COVID patients’ risk of needing intensive emergency care.
With BARDA’s cost-share contract, PeraHealth will create a mobile and desktop app based on its FDA-cleared Rothman Index, which analyzes real-time electronic health record data like vital signs and lab results to identify the patients with the highest chance of deteriorating conditions.
The RI Risk Triage app will use machine learning to generate a risk assessment score for each patient entering a hospital’s emergency department for COVID treatment. Clinicians can then use that score to determine the most effective distribution of their facilities’ ventilators, ICU beds and other limited resources.
Additionally, even after the coronavirus pandemic ends, BARDA and PeraHealth noted that the RI Risk Triage app will be easily transitioned into use in other public health crises and for emergency department triage in general.
“This cost-share contract represents a critical milestone in our organization as we pivot to support patient triage in the emergency department for our hospital partners,” PeraHealth CEO Greg White said in a statement. “Emergency department capacities have experienced historic highs in the last 12 months, so developing effective clinical decision support countermeasures is essential.”
AgileMD, meanwhile, will receive $750,000 from BARDA to support the development of its own artificial-intelligence-powered risk assessment software. Like PeraHealth’s tool, AgileMD’s eCART system also continuously analyzes data from patients’ health records to predict whether their condition will worsen, then sends alerts to clinicians before it happens.
The eCART system monitors and looks for patterns among almost 100 variables in the EHR to make its predictions. The most telling of these is the amount of supplementary oxygen required to keep a patient’s blood oxygen at normal levels, AgileMD co-founder Dana Edelson, M.D., recently told IEEE Spectrum.
The eCART system has been in development since 2015, when it launched with the ability to analyze about 30 variables to determine whether patients will go into cardiac arrest, require a transfer to the ICU or die within the next eight hours.
With the onset of the COVID pandemic in early 2020, AgileMD sped up its plans to expand the platform to support even more health data, backed by another grant from startup incubator Matter Health in August. AgileMD used the Matter funding to back the process of submitting its software for FDA approval.
“This BARDA partnership continues that effort and will allow us to finalize regulatory work, pilot in 10 hospitals and rapidly make available the solution to many others with the ultimate objective of improving clinical decision support and patient safety,” Edelson said.