Comparing nursing students and AI systems performance on the ability to solve basic life support and advanced cardiac...

The primary aim of this study was to compare graduating nursing students' performance on a Basic Life Support (BLS) and Advanced Cardiac Life Support (ACLS) exam to three AI systems: Gemini, ChatGPT and Copilot.

Several researchers from the medical field have shown outstanding performance of artificial intelligence (AI) systems such as ChatGPT in passing national licensure examinations. This achievement for AI systems is promising for students and instructors. In nursing, the evidence concerning this issue is limited. Furthermore, there is no previous study that have ever compared AI systems performance to nurses or nursing students.

This is a descriptive study.

A total of 120 third- and fourth-year nursing students enrolled in the "Emergency and Disaster Nursing Care" course, who were scheduled to take the "BLS and ACLS" examination, were invited to participate in the study. The students' exam scores were then compared with the scores generated by AI systems.

Mean students' score was 20.70 ± 3.39 (9-25) out of 25. AI systems performance on the same exam was as follows: Gemini scored 23 out of 25, ChatGPT 21 and Copilot 24. ChatGPT performed better than 55.8 % of graduate nursing students, Gemini better than 72.5 %, while Copilot performed better than 86.7 %.

Performance of AI systems compared with graduating nursing students is outstanding. Copilot is the best. These systems are incredibly smart, but they are not infallible. At present, expert human supervision is still required.