Medical departments at Prince Charles Hospital, a public hospital in Queensland, recently trialled what may be the world’s first electronic tool to assess the likely causes of delirium.
According to Metro North Health, the electronic tool uses a series of yes/no questions that help narrow down causes and select groups, such as infections, organ dysfunction or medications. It is built with algorithms that use ‘conventional principles’, combined with smart design features such as adaptive testing.
WHY IT MATTERS
Delirium, one of the most underdiagnosed conditions during hospital admissions, affects approximately one in five patients in Australia, mainly seniors. The onePparental costs the Australian healthcare system A$8.8 billion ($5.8 billion) per year.
The condition is often overlooked in about half of cases due to the many possible causes, says Dr. Eamonn Eeles, a geriatrician at TPCH and developer of the delirium assessment tool.
“Examination can be challenging due to drowsiness where a patient may be quiet or seemingly withdrawn. Alternatively, a patient may be irritable. In these situations, a diagnosis may be missed.”
Timely diagnosis cannot be overemphasized, as delays can “quadruple” the risk of mortality, Dr. Eeles.
Dr. Eeles has spent the past ten years developing the tool together with the TPCH Internal Medicine Service and the Dementia and Neuro Mental Health Research Unit at the University of Queensland Center for Clinical Research.
They are now working with other TCPH departments, including emergency medicine, to expand the application of the delirium assessment tool.
THE BIG TREND
The past year has been aAt least two studies have examined the application of AI models in assessing delirium. For example, a study from Johns Hopkins University revealed how AI models predicted delirium risks based on EMR data from 200,000 ICU stays. Another study used a supervised deep learning model, together with an EEG device, to better assess delirium.