Personalized Predictions for Changes in Knee Pain Among Patients With Osteoarthritis Participating in Supervised Exercise and Education: Prognostic Model Study.
Mahdie Rafiei, Supratim Das, Mohammad Bakhtiari and 5 others
PMID 40116731WHAT IT FOUND
A new model using six simple questions predicts knee pain changes after GLA:D therapy for 58% of patients.
This is about 7% more accurate than telling everyone to expect the average improvement. The model is not yet ready for clinical use because the authors state the accuracy is too preliminary.
Key findings
01The concise model correctly predicted 58% of cases, showing a 7% improvement over using average values for prediction.
02The concise model uses six variables: baseline pain, duration of symptoms, EQ-5D score, 40-meter walking time, age, and BMI.
03The authors state the current model performance is too preliminary to change clinical practice or guidelines.
STILL TO COME
How it was doneWhat they foundWhat it means for PTs
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 model is still incorrect for nearly half of the patients. The study relies on registry data where patients self-selected into the program, which may introduce bias compared to a randomized trial population. The authors explicitly state the performance is too preliminary for clinical implementation. Predictions do not account for factors like patient preference or constant pain, which were not in the dataset. The model was validated only on the GLA:D program and may not apply to other exercise therapies.
Declared interests
One author is a co-founder of GLA:D, the program being studied. Another author is on the editorial board of a journal related to the outcome measures used and is also a co-founder of GLA:D. The study used registry data and did not report external funding.
The easy way to misread this
Do not use this model to tell patients they will not benefit from GLA:D if their predicted change is small. The model is wrong for a large proportion of patients, and the authors state it is not yet ready for clinical decision-making.