Impact of multimodal education management on postoperative rehabilitation after total knee arthroplasty: A machine learning-based prediction model study

多模式教育管理对全膝关节置换术后康复的影响:基于机器学习的预测模型研究

阅读:1

Abstract

This study aimed to evaluate the impact of multimodal education management using illustrated pathway with video education (IPVE) on rehabilitation quality after total knee arthroplasty. A retrospective cohort study design was adopted. Patients were grouped based on the median of final SF-36 quality of life scores. LASSO regression was used to screen predictive variables, and multivariate logistic regression was used to construct prediction models. The predictive performance of 5 machine learning algorithms was compared, and model efficacy was evaluated using ROC curves, calibration curves, and decision curve analysis. SHAP method was used to analyze feature importance. Multimodal education management using IPVE was significantly associated with better rehabilitation quality after total knee arthroplasty. A total of 223 patients who underwent total knee arthroplasty from October 2022 to January 2025 were included, with 121 cases (54.3%) in the high-quality rehabilitation group and 102 cases (45.7%) in the low-quality group. LASSO regression identified 4 key predictive variables: age, IPVE implementation, knee range of motion at discharge, and final knee function score. Multivariate logistic regression analysis showed that each 1-year increase in age reduced the probability of high-quality rehabilitation by 17.2% (P<.001), IPVE implementation was significantly associated with better rehabilitation quality (P<.001), each 1° increase in knee range of motion at discharge increased the probability of high-quality rehabilitation by 17.3% (P<.001), and each 1-point increase in final knee function score increased the probability of high-quality rehabilitation by 11.2% (P = .043). The random forest model performed best, with the AUC, sensitivity, specificity, accuracy, and F1 score all reaching 1.000, whereas the traditional logistic regression model had an AUC of 0.924. SHAP analysis showed that age was the most important predictive feature, and implementation of IPVE had a significant impact on rehabilitation quality. Multimodal educational management using an illustrated pathway combined with video-based education was significantly associated with improved rehabilitation quality after total knee arthroplasty.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。