Global study to analyze ICU patient data to guide COVID-19 treatments

SYDNEY - A global study of intensive care units (ICU) will use artificial intelligence (AI) to uncover the best standards of treatment for COVID-19, and help health systems worldwide cope with critically-ill patients.

Revealed on Wednesday, the Australian-led study aims to examine the COVID-19 patient data from 300 ICUs around the globe in the hopes of shedding light on which treatments work best.

University of Queensland (UoQ) researcher Professor John Fraser, who is also the ICU Director at Brisbane's St Andrews Hospital, said the COVID-19 Critical Care Consortium Study is the first of its kind in the world.

"Frontline doctors and nurses need evidence to guide them, especially when faced with COVID-19 patients who already have a chronic disease like diabetes, but at present, clinicians have nothing," Fraser said.

The study aims to analyze potentially tens of thousands of patients on six continents, to create predictive models and easily access information to guide medical workers on ICU treatments.

To make sense of the data, researchers will use a cutting edge AI tool, co-developed by UoQ and IBM.

Clinicians can use the new tool to quickly record, share and compare a range of treatment factors, including vital sign measures, the use of mechanical ventilation, duration of stay in ICU, and survival rates.

"By leveraging this data, we can enhance ICU patient care, improve COVID-19 understanding in doctors and nurses and guide future treatments of the disease," Fraser said.

"Ultimately, this study will give clinicians the decision support mechanism they need to instantly determine the most appropriate COVID-19 treatment and increase ICU survival rates."