Exploring Early Perceptions and Experiences of ChatGPT in Pediatric Critical Care: A Qualitative Study Among Healthcare Professionals

Mohamad-Hani Temsah,Noura Abouammoh,Mohammed Alsatrawi,Mohammed Almazyad,Fadi Aljamaan,Mariella Vargas-Gutierrez,Rebecca Hay,Muneera Al-Jelaify,Wejdan Alabdulkreem,Nawal Assiri,Ruaim Muaygil,Ibraheem Altamimi,Fatimah S. Alshahrani,Khalid Alhasan,Khalid H. Malki,Amr Jamal,Jaffar A. Al-Tawfiq,Ayman Al-Eyadhy
DOI: https://doi.org/10.1101/2024.03.18.24304453
2024-03-18
Abstract:This qualitative inquiry explores the initial impressions and firsthand encounters of healthcare professionals (HCPs) with ChatGPT, a Generative Pre-trained Transformer, within Pediatric Intensive Care Units (PICUs). Through focus group discussions held at a tertiary academic center, a diverse cadre of HCPs was engaged to ascertain their awareness, utilization patterns, perceived advantages, and apprehensions regarding ChatGPT. The analysis revealed three primary themes: understanding and ease of use of ChatGPT, its practical applications in clinical workflows for critically ill children and information retrieval, and the ethical considerations associated with its deployment. While participants praised ChatGPT for its engaging interface and potential to streamline tasks and provide prompt information, notable reservations surfaced regarding its limitations, particularly in medical accuracy, currency of data, and ethical implications. The findings suggest a cautious optimism towards integrating Generative Artificial Intelligence (GAI), like ChatGPT, in pediatric critical care, highlighting the need for balanced, informed, and transparent applications, with ongoing evaluation of GAI technologies in pediatric healthcare settings.
Health Informatics
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to understand the initial impressions and actual usage experiences of medical professionals in the Pediatric Intensive Care Unit (PICU) regarding ChatGPT, a Generative Pre - trained Transformer (GPT). Specifically, through focus group discussions, the research explored the following aspects: 1. **Understanding and Ease of Use of ChatGPT**: The research explored the extent to which medical professionals understand the functions and usage methods of ChatGPT, as well as the ease of use and limitations they discovered during actual operations. 2. **Application of ChatGPT in the Clinical Workflow**: The research analyzed the actual application of ChatGPT in handling critical cases of children, including its role in task organization, information retrieval, etc. 3. **Ethical Considerations**: The research focused on the ethical issues that may be raised by using ChatGPT in the pediatric intensive care environment, such as data privacy, patient autonomy, and changes in the doctor - patient relationship. Through the exploration of these aspects, the research aims to provide guidance and suggestions for more effective integration and application of generative artificial intelligence technology in pediatric intensive care in the future.