Health and Rehabilitation Sciences, Graduate Program, Faculty of Health Sciences, Western University, London, ON, Canada.
Department of Surgery, Division of Orthopaedic Surgery, University Hospital, London Health Sciences Centre, London, ON, Canada.
J Arthroplasty. 2022 Feb;37(2):267-273. doi: 10.1016/j.arth.2021.10.017. Epub 2021 Nov 2.
Approximately 20% of total knee arthroplasty (TKA) patients are found to be dissatisfied or unsure of their satisfaction at 1-year post-surgery. This study attempted to predict 1-year post-surgery dissatisfied/unsure TKA patients with pre-surgery and surgical variables using logistic regression and machine learning methods.
A retrospective analysis of patients who underwent primary TKA for osteoarthritis between 2012 and 2016 at a single institution was completed. Patients were split into satisfied and dissatisfied/unsure groups. Potential predictor variables included the following: demographic information, patella re-surfaced, posterior collateral ligament sacrificed, and subscales from the Knee Society Knee Scoring System, the Knee Society Clinical Rating System, the Western Ontario and McMaster Universities Osteoarthritis Index, and the 12-Item Short Form Health Survey version 2. Logistic regression and 6 different machine learning methods were used to create prediction models. Model performance was evaluated using discrimination (AUC [area under the receiver operating characteristic curve]) and calibration (Brier score, Cox intercept, and Cox slope) metrics.
There were 1432 eligible patients included in the analysis, 313 were considered to be dissatisfied/unsure. When evaluating discrimination, the logistic regression (AUC = 0.736) and extreme gradient boosted tree (AUC = 0.713) models performed best. When evaluating calibration, the logistic regression (Brier score = 0.141, Cox intercept = 0.241, and Cox slope = 1.31) and gradient boosted tree (Brier score = 0.149, Cox intercept = 0.054, and Cox slope = 1.158) models performed best.
The models developed in this study do not perform well enough as discriminatory tools to be used in a clinical setting. Further work needs to be done to improve the performance of pre-surgery TKA dissatisfaction prediction models.
大约 20%的全膝关节置换术（TKA）患者在术后 1 年时被发现不满意或不确定其满意度。本研究试图使用逻辑回归和机器学习方法，根据术前和手术变量预测术后 1 年不满意/不确定的 TKA 患者。
对 2012 年至 2016 年在一家单机构行初次 TKA 治疗骨关节炎的患者进行回顾性分析。患者分为满意组和不满意/不确定组。潜在预测变量包括以下内容：人口统计学信息、髌骨再表面化、后十字韧带牺牲，以及膝关节学会膝关节评分系统、膝关节学会临床评分系统、西部安大略省和麦克马斯特大学骨关节炎指数和 12 项简明健康调查问卷版本 2 的子量表。使用逻辑回归和 6 种不同的机器学习方法来创建预测模型。使用判别（AUC[受试者工作特征曲线下面积]）和校准（Brier 评分、Cox 截距和 Cox 斜率）指标评估模型性能。
共纳入 1432 例符合条件的患者，313 例被认为不满意/不确定。在评估判别能力时，逻辑回归（AUC=0.736）和极端梯度提升树（AUC=0.713）模型表现最佳。在评估校准能力时，逻辑回归（Brier 评分=0.141、Cox 截距=0.241、Cox 斜率=1.31）和梯度提升树（Brier 评分=0.149、Cox 截距=0.054、Cox 斜率=1.158）模型表现最佳。
本研究中开发的模型作为判别工具的性能不够好，不能在临床环境中使用。需要进一步努力提高术前 TKA 不满预测模型的性能。
