Explainable machine learning prediction of functional independence measure scores and gain in subacute stroke survivors.
Yuta Miyazaki, Junna Oba, Tetsuo Ishikawa and 8 others
PMID 41957775WHAT IT FOUND
Admission scores on basic daily tasks like bowel management strongly predicted discharge independence, while age and toileting skills predicted overall functional gain.
These machine learning models identified specific prognostic factors, offering a transparent way to forecast rehabilitation outcomes in subacute stroke survivors.
Key findings
01Admission FIM bowel management was the strongest predictor of FIM motor and total scores at discharge.
02For FIM motor and total gain, age, lower body dressing, and toileting exhibited high SHAP values.
03Age and cognitive-related FIM subitems at admission were identified as critical predictors for discharge FIM cognitive score and cognitive gain.
STILL TO COME
How it was doneWhat they foundWhat it means for PTsWhat it means for OTsWhat it means for SLPs
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 was single-center and retrospective, which limits generalizability and introduces potential selection bias. Discharge timing varied across participants, so predictions are not tied to a fixed time horizon. The models did not include deep learning algorithms, which might offer further accuracy improvements.
Declared interests
Funded by the Japan Society for the Promotion of Science. No other conflicts of interest are reported.
The easy way to misread this
Do not interpret the high predictive accuracy as a tool for individual patient decision-making without external validation. The study is retrospective and single-center, meaning the models have not been tested for generalizability in other populations or settings.
Summarised by AI from the full paper, without a clinician reviewing it. Check it against the source before it changes what you do. Read it on PubMed →