The Use of Chatgpt by Pediatric Physicians in the Management of Childhood Asthma in the City of Marrakech (Morocco)

Abstract

Pediatric asthma is a chronic disease requiring continuous management, where digital tools, such as ChatGPT, are beginning to play a vital role, in optimizing physicians' efficiency, reducing their workload, and improving communication with patients. However, its integration raises questions about the accuracy of the information provided and data security. In addition, the use of ChatGPT in clinical practice remains little explored compared to its use in the field of researsch. In this context, this study aims to describe the use of ChatGPT by pediatricians, assess its advantages and limitations, and suggest recommendations for optimal integration of this tool in the context of pediatric asthma. A prospective survey was conducted for one month at the mother-child hospital in Marrakech. Data were collected online using a structured questionnaire from all pediatric doctors working in this establishment. Data analysis was performed using descriptive methods in addition to bivariate and multivariate analyses, with SPSS software. The study included 53 pediatric physicians, represented mainly by women (94.3%). Approximately 56,6% of doctors used ChatGPT in the management of children with asthma, especially for abstracting (41.5%), translating (35.84%), and asking medical questions (22.64%). The main reasons for use were; simplification of administrative tasks (44.11%) and the rapid access to information (35.29%). Although 57% of physicians were satisfied with ChatGPT’s use, limitations such as lack of customization (26.6%) and reliance on technology (31.3%) were noted. User doctors were the ones who were familiar with this technology and who encountered more difficulties in the management of pediatric asthma (p<0.001). Although recent, ChatGPT is already widely adopted by pediatricians for some specific tasks in the management of childhood asthma, such as writing summaries, translating, and answering questions, while leaving clinical decisions under the control of physicians. This highlights both the effectiveness of the tool and the need for strict human supervision and enhanced security measures.