Predicting the Development of Surgery-Related Pressure Injury Using a Machine Learning Algorithm Model.
Ji-Yu Cai, Man-Li Zha, Yi-Ping Song and 1 others
PMID 33351552WHAT IT FOUND
The model missed most patients who developed surgery-related pressure injury, but correctly identified all patients who did not develop it.
Key findings
01The model correctly identified only 3 patients who developed surgery-related pressure injury and misclassified 34 as not developing it.
02All 112 patients without surgery-related pressure injury were correctly predicted as not developing it.
03The primary performance indicator was an area under the curve of 0.806, with sensitivity 8.11% and specificity 100%.
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 data came from one hospital, so the model may not generalize to other settings. The model had very low sensitivity. It missed most patients who developed pressure injury. The authors note that not collecting data prospectively may affect model performance. The sample included only patients who had cardiac or aortic surgery.
Declared interests
The authors declared no conflicts of interest.
The easy way to misread this
Do not read the model as clinically usable because its overall performance score was 0.806. It correctly identified only 3 patients who developed pressure injury and missed 34, so it is not sensitive enough for prevention decisions.