Development of a large language model-based knowledge graph for chemotherapy-induced nausea and vomiting in breast cancer and its implications for nursing.
Yu Liu, Jingjing Chen, Xianhui Lin and 2 others
PMID 41367590WHAT IT FOUND
A prototype knowledge graph linked 238 chemotherapy-induced nausea and vomiting concepts from 47 evidence sources, but it was not tested with nurses or patients, so it cannot yet change bedside care.
Key findings
01The graph kept 238 validated entities, 242 relations and 244 triples after expert review of text from 47 studies.
02After supervised fine-tuning, the model's reported F1 scores were 82.97 for naming entities and 85.54 for extracting relations.
03The study did not test the graph in real clinical scenarios, such as decision support or patient symptom management.
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
No patients, nurses or clinical systems were tested, so it does not show whether the graph helps bedside decision-making or symptom control. The model still needed experts to remove wrong or made-up items, so the graph's reliability depends on manual checking. The extraction scores were not perfect, especially for linking relationships, so some relevant information may be missed or incorrectly connected.
Declared interests
Supported by provincial and university grants; authors declared no competing financial interests or personal relationships.
The easy way to misread this
Do not read this as a tested nursing decision-support tool. The study built and expert-checked a knowledge graph and measured text extraction, but it did not test whether nurses using it improve chemotherapy-induced nausea and vomiting care or patient outcomes.