Making sense of patient experience data

8 Oct 2021

Digital Health hospital data concept map

University of Queensland researchers are pioneering the use of artificial intelligence to enable healthcare providers to better monitor and improve hospital care and the patient experience.

Supported by the UQ Global Change Institute (GCI), the research community is working to secure better patient outcomes through improved access to real-time patient outcome and experience data.

For some time, patients have been routinely invited to provide feedback on their healthcare experience by responding to a simple text message – not unlike rating a hotel room or restaurant online.

According to UQ Research Fellow Professor Jason Pole an efficient system was needed to allow healthcare providers to use the data constructively.

“Initially my UQ colleagues and I ran a small trial that demonstrated artificial intelligence could be used to process the patient feedback data in a way that was not only meaningful but could be done efficiently.”

The trial’s success led to Professor Pole and his UQ colleague Associate Professor Clair Sullivan examining the feasibility of pioneering this AI analysis technology at scale.

“The pilot study had demonstrated that an off-the-shelf AI tool, Leximancer, could be used to meaningfully interpret patient survey data quickly and with a high degree of accuracy,” Professor Pole said.

“We showed it was possible to drill down into the survey data and automatically extract comments and key words that offered tangible insights into understanding this type of patient response.”

Developed nearly twenty years ago at UQ by Dr Andrew Smith, Leximancer is now an internationally successful software platform that can identify key trends, concepts and ideas from large pieces of text.

The platform has already been used by customers in 140 countries and more than 2000 organisations, including universities, governments, non-for-profits, and commercials.

However, it has never previously been used for healthcare data at scale.

By generating an interactive topic list, the software allows users to understand the true meaning of text without human biases that could potentially distort the meaning.

GCI Director Professor Rachel Parker said the strength of the Digital Health Research Network was that it had drawn together a range of stakeholders from across disciplines to deliver integrated solutions to improving hospital care and patient experience.

Media: Professor Jason Pole, j.pole@uq.edu.au, 0400 605 106; GCI Communications, gcicomns@uq.edu.au, 0438 285 283

Latest