Evaluation of Information Generated by ChatGPT on Preventing Peripheral Venous Catheter-Related Infections.
Seda Pehlivan, Derya Akça Doğan, Öznur Erbay Dallı
PMID 41761913WHAT IT FOUND
No ChatGPT version clearly beat the others on peripheral venous catheter infection questions.
GPT-3.5 and GPT-4o scored 5 of 10; GPT-4 scored 3. All missed hand hygiene and aseptic technique.
Key findings
01GPT-3.5 and GPT-4o scored 5 points on the 10-question infection knowledge form, while GPT-4 scored 3 points.
02All three models incorrectly answered questions on hand hygiene, aseptic technique, catheter site dressing type, and replacement of administration sets for neither lipid emulsions nor blood product infusions.
03No significant difference was found between the models for accuracy or completeness of explanations.
STILL TO COME
How it was doneWhat they foundWhat it means for RNs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study tested only 10 questions about peripheral venous catheter infection prevention, so it does not show how ChatGPT performs across other nursing topics or real patient care. No nurses, patients, or clinical outcomes were involved, so it cannot show whether using ChatGPT changes care or patient safety. The questions were fixed and asked on one date, so results may not reflect how clinicians use AI in variable clinical situations. The form relies on predefined CDC-based questions, which may not capture dynamic bedside decisions.
Declared interests
The authors report no funding and declare no conflicts of interest.
The easy way to misread this
Do not read the similar scores across GPT versions as evidence that ChatGPT is reliable for peripheral venous catheter infection prevention. The models still gave wrong answers on hand hygiene, aseptic technique, dressing type, and administration set replacement, and no patient outcomes were tested.