Prediction of respiratory failure in COVID-19 infection and understanding pathophysiological mechanisms

Funder: University of Oxford

PI: Sophie Yacoub

Location of activity: Hanoi, Ho Chi Minh City, and Cu Chi, Viet Nam and Jakarta, Indonesia

In this observational study, our primary aim is to utilise the expertise we have developed in OUCRU Viet Nam to develop dynamic models that predict disease progression using real-time, longitudinal clinico-physiological data and specific blood biomarkers. Achieving this requires research in early infection and across the disease severity spectrum. In many countries, this is difficult due to the high burden of severe cases. In Viet Nam, the early identification of SARS-CoV-2 infections is routine due to a highly efficient test, trace and quarantine system.

At OUCRU Viet Nam, in collaboration with Oxford University, we are using point-of-care ultrasound and physiological monitoring via simple, low-cost wearable devices to develop artificial intelligence (AI) systems to better manage critically ill patients.

In this project, we intend to use conventional statistical techniques and AI to develop dynamic predictive models that will enable the identification of patients at risk of disease progression in COVID-19. We will build on our expertise using data from wearable devices, point-of-care heart and lung ultrasound, and specific blood biomarkers.

This study is currently recruiting in two sites in Ho Chi Minh City (Hospital for Tropical Diseases and Cu Chi Hospital) as well as in the National Hospital for Tropical Diseases in Hanoi and the Pasar Minggu Hospital in Jakarta, Indonesia.