Non-Moderated Poster Abstract
Eposter Presentation
 
Accept format: PDF. The file size should not be more than 5MB
 
Accept format: PNG/JPG/WEBP. The file size should not be more than 2MB
 
Submitted
Abstract
The Clinical Integration Of ChatGPT Through An Augmented Patient Encounter In A Real World Urological Cohort
Podium Abstract
Clinical Research
AI in Urology
Author's Information
7
No more than 10 authors can be listed (as per the Good Publication Practice (GPP) Guidelines).
Please ensure the authors are listed in the right order.
Australia
Shane Qin shane.s.qin@gmail.com Austin Health Urology Melbourne Australia *
Emre Alpay emre.alpayy@gmail.com Austin Health Urology Melbourne Australia -
Bodie Chislett Bodiechislett@gmail.com Austin Health Urology Melbourne Australia -
Joseph Ischia ischiajj@gmail.com Austin Health Urology Melbourne Australia -
Luke Gibson lukedgibson@gmail.com Austin Health Urology Melbourne Australia -
Damien Bolton Damienmbolton@gmail.com Austin Health Urology Melbourne Australia -
Dixon Woon dixonwoon@gmail.com Austin Health Urology Melbourne Australia -
 
 
 
 
 
 
 
 
 
 
 
 
 
Abstract Content
The world is experiencing an artificial intelligence (AI) revolution, with generative AI technology having a transformative impact across various sectors including healthcare. Despite concerns, AI has the potential to revolutionize healthcare by reducing costs, improving patient comprehension, increasing access to health information and improving clinical workflow and efficiency. This study explores the use of ChatGPT, a generative AI chatbot, to aid patient education and streamline workflow through an augmented patient interaction in the outpatient setting.
A prospective feasibility study was conducted at a single institution involving patients referred with macroscopic or microscopic haematuria. Inclusion criteria comprised of patients over 18 years with haematuria and able to speak English. Patients were required to input predetermined prompts into ChatGPT on a computer including an explanation of haematuria; its causes, risks; the process and risks of a flexible cystoscopy; and alternative treatments. They then underwent a regular consultation with a clinician. Patients and clinicians both filled out surveys afterwards about their experience with ChatGPT. The accuracy, consistency and suitability of the information given by ChatGPT was also assessed.
In total, nine patients were recruited for the study. The majority found that ChatGPT was a useful resource in providing them with medical information about their presentation, however, most did not feel it could replace a clinician answering patient questions or providing informed consent for a procedure (see Table 1). This demonstrates that ChatGPT is an accurate and valuable resource for patient education when using the predetermined prompts. We identified no major errors in ChatGPT’s responses, and it provided clear and accurate information targeted at an appropriate level.
This feasibility study shows the potential of integrating ChatGPT into clinical practice for patient education, leading to benefits such as reduced clinician workload, improved patient comprehension, and increased community health literacy. ChatGPT has shown utility in aiding patient education and augmenting patient-clinician interactions, but is unlikely to completely replace clinician interactions in the near future.
Generative AI, ChatGPT, urology, consent, cystoscopy, haematuria
https://storage.unitedwebnetwork.com/files/1237/4e553a26b68bac15d0e0472022ff4c94.jpg
ChatGPT patient questionnaire responses
 
 
 
 
 
 
 
 
1859
 
Presentation Details