Key result
Machine learning model predicts length of stay after hip or knee arthroplasty with ~79% accuracy.
Why the study?
The growth of arthroplasty procedures requires innovative strategies to reduce inpatient length of stay.
Can a machine learning model predict length of stay in patients undergoing hip or knee arthroplasties?
Observational (n=1,517)
No
Can a machine learning model predict length of stay in patients undergoing hip or knee arthroplasties?
Effect estimate: AUC 0.7899
A machine learning model using pre-operative clinical data can predict length of stay after hip or knee arthroplasty with an AUC of 0.7899.
May support preoperative LOS planning in arthroplasty; leaves open prospective validation and outcome impact.
Background: The growth of arthroplasty procedures requires innovative strategies to reduce inpatients Length of Stay (LOS). This study aims to develop a machine learning prediction model that may aid to predict LOS after hip or knee arthroplasties. Methods: A collection of all the clinical notes of patients undergoing elective primary or revision arthroplasty from 1 January 2019 to 31 December 2019 was performed. The hospitalization has been classified as “Short LOS” if it was less than or equal to 6 days and “Long LOS” if it was greater than 7 days. Clinical data coming from pre-operative laboratory analysis, vital parameters, demographic characteristics of patients were screened. Final data have been used to train a Logistic Regression model with the aim of predicting short or long LOS. Results: Final dataset was composed of 1517 patients (795 “LONG”, 722 “SHORT”, p = 0.3196) with a total of 1541 admissions (729 “LONG”, 812 “SHORT”, p < 0.000). Complete model had a prediction efficacy of 78,99% (AUC 0.7899). Conclusions: Machine learning may facilitate day-by-day clinical practice predicting which patients are suitable for a shorter LOS, from those with a longer LOS in which a cautious approach could be recommended.
No takes yet. Share an insight, caveat, or question.
Matteo et al. (2024) conducted an observational in Hip or knee arthroplasties (n=1,517). Logistic Regression machine learning prediction model was evaluated on Prediction of short (≤6 days) or long (>7 days) length of stay (AUC 0.7899). A logistic regression machine learning model predicted short versus long length of stay after hip or knee arthroplasty with an efficacy of 78.99% (AUC 0.7899).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: