Delayed diagnosis of a transient ischemic attack caused by ChatGPT

Abstract

SummaryTechniques of artificial intelligence (AI) are increasingly used in the treatment of patients, such as providing a diagnosis in radiological imaging, improving workflow by triaging patients or providing an expert opinion based on clinical symptoms; however, such AI techniques also hold intrinsic risks as AI algorithms may point in the wrong direction and constitute a black box without explaining the reason for the decision-making process.This article outlines a case where an erroneous ChatGPT diagnosis, relied upon by the patient to evaluate symptoms, led to a significant treatment delay and a potentially life-threatening situation. With this case, we would like to point out the typical risks posed by the widespread application of AI tools not intended for medical decision-making.

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

“…present a case report of a patient who relied on ChatGPT inaccurate diagnosis, and that resulted in a substantial delay in treatment and a potentially life-threatening circumstance [49]. These findings raise important clinical and ethical considerations, especially in the sensitive setting of a PICU, where decisions can have significant consequences.…”

“…The ability of ChatGPT to provide appropriate and equitable medical advice was evaluated by Nastasi et al, as they presented ChatGPT with various clinical vignettes and found that while ChatGPT’s responses were largely in line with clinical guidelines, it did not consistently offer personalized medical advice [48]. AI algorithms carry inherent dangers that may lead to incorrect conclusions and operate as “black box” systems that do not elucidate the rationale behind decision-making [49]. Saenger et al present a case report of a patient who relied on ChatGPT inaccurate diagnosis, and that resulted in a substantial delay in treatment and a potentially life-threatening circumstance [49].…”

“…AI algorithms carry inherent dangers that may lead to incorrect conclusions and operate as “black box” systems that do not elucidate the rationale behind decision-making [49]. Saenger et al present a case report of a patient who relied on ChatGPT inaccurate diagnosis, and that resulted in a substantial delay in treatment and a potentially life-threatening circumstance [49]. These findings raise important clinical and ethical considerations, especially in the sensitive setting of a PICU, where decisions can have significant consequences.…”

The ethics of ChatGPT in medicine and healthcare: a systematic review on Large Language Models (LLMs)

“…Paramount concerns include the propensity of LLMs to disseminate inadequate information, the input of sensitive health information or patient data, which raises significant privacy issues 24 , and the perpetuation of harmful gender, cultural or racial biases 27 – 30 , well known from machine learning algorithms 31 , especially in healthcare 32 . Case reports have documented that ChatGPT has already caused actual damage, potentially life-threatening for patients 33 .…”