Back to top anchor

Predicting acute asthma events using machine learning

Year:
2021
Duration:
22 months
Approved budget:
$29,498.00
Researchers:
Associate Professor Amy Chan
,
Assistant Professor Kebede Beyene
,
Miss Holly Tibble
,
Farhaan Mirza
Host:
The University of Auckland
Proposal type:
Health Delivery Research Activation Grant
Lay summary
New Zealand (NZ) has one of the highest rates of asthma in the world. Asthma is the third leading cause of death in NZ, and is a significant contributor to health inequity, with Māori having three times higher hospitalisation rates than non-Māori. Acute asthma events such as hospitalisation are avoidable if identified early, yet models that predict asthma events have poor predictive accuracy, partly due to a reliance on traditional statistical methods. In this proposal, we will use artificial intelligence (AI) methods to explore 10 years of data from NZ healthcare datasets to identify factors that predict acute asthma events. This will be the first time that AI methods, which enhance predictive accuracy, have been used to model asthma events in NZ. The work will inform future risk models, using health and non-health datasets, that can be used to identify changes in asthma risk early on and facilitate early intervention.