Vienna (dpa) – Doctors and nurses perform better overall at emergency department triage than artificial intelligence (AI), which displayed lower "overall accuracy” in a series of hospital tests. According to researchers speaking at the European Emergency Medicine Congress in Vienna, AI gave accurate responses to triage-related questions and scenarios about half the time, while nurses achieved around 65% and doctors over 70%. At a time when AI grows increasingly capable, the findings run counter to indications that many human skills could in future be rivalled by software. In tests carried out at the emergency department of Vilnius University Hospital, the doctors and nurses were asked to use the Manchester Triage System to classify patients from case studies in one of five categories from most to least urgent. "Sensitivity – how well it identified true urgent cases – for AI was also lower at 58.3% compared to nurses, who scored 73.8%, and doctors, who scored 83.0%," said Renata Jukneviciene of Vilnius University. The same cases were analysed by ChatGPT, which, although falling short of nurses and doctors overall, did outperform nurses in the first triage category — the most urgent or potentially life-threatening cases presented in an emergency ward. "We conducted this study to address the growing issue of overcrowding in the emergency department and the escalating workload of nurses," Jukneviciene explained. "Given the rapid development of AI tools like ChatGPT, we aimed to explore whether AI could support triage decision-making, improve efficiency and reduce the burden on staff in emergency settings," she added. While AI "has potential" for use in hospitals, according to Barbra Backus, an Amsterdam-based emergency doctor, it comes with "limitations." "This study shows very clearly that it cannot replace trained medical staff for triaging patients coming in to emergency departments," said Backus, who chairs the conference's abstract selection committee. The following information is not intended for publication dpa spr coh
(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)
