全膝关节置换术患者术后6个月生活质量的影响因素分析及列线图预测模型的构建

关键词：全膝关节置换术  膝关节骨性关节炎  生活质量  预测模型  列线图

Keywords:total knee arthroplasty  knee osteoarthritis  quality of life  prediction model  nomogram

摘要：

      摘要 目的：探讨全膝关节置换术（TKA）患者术后6个月生活质量低于正常水平的影响因素和列线图预测模型的构建。 方法：选取2021年4月—2022年10月于青岛市市立医院行TKA的患者267例，收集患者入院及术后6个月内的临床资料，根据术后6个月随访时生活质量评估分为高质量组（136例）和非高质量组（131例），采用LASSO回归筛选预测因子，并建立多因素Logistic回归，构建TKA患者术后6个月生活质量预后的预测模型，绘制受试者工作特征曲线（ROC曲线）、校准曲线及决策曲线分析以评估其预测价值。 结果：共有131例（49.06%）TKA患者术后6个月生活质量低于正常水平。LASSO回归筛选出年龄、手术侧别、受教育程度、膝关节骨性关节炎（KOA）病程、围手术期康复情况、术前休息时视觉模拟评分（VAS）为预测因子；多因素Logistic回归分析显示年龄（OR， 1.08； 95% CI， 1.01—1.15）、KOA病程（OR， 1.13； 95% CI， 1.05—1.22）、术前休息时VAS（OR， 1.94； 95% CI， 1.20—3.15）、围手术期康复（OR， 0.42； 95% CI， 0.18—0.97）是TKA患者术后6个月生活质量低于正常水平的独立危险因素（P<0.05）。将多因素Logistic回归分析中有统计学意义的指标纳入列线图模型，进行内部验证，ROC曲线下面积为0.829（95%CI， 0.779—0.879），该模型的预测能力良好；校准曲线显示预测概率与实际概率有很好的一致性；决策曲线分析显示该模型适用于临床工作。 结论：年龄、KOA病程、术前休息时VAS和围手术期康复是TKA患者术后6个月生活质量低于正常水平的独立预测因素，且基于以上指标构建的列线图模型具有良好的预测效能。

Abstract:

      Abstract Objective：To explore the risk factors of quality of life (QoL) being below normal levels in patients undergoing total knee arthroplasty（TKA）at 6 months after surgery， and to construct a nomogram prediction model. Method：Two-hundred and sixty-seven patients who underwent TKA at Qingdao Hospital， University of Health and Rehabilitation Sciences（Qingdao Municipal Hospital） from April 2021 to October 2022 were selected. Clinical data were collected on admission and within 6 months after surgery. According to the QoL assessment at 6 months after surgery，they were divided into a high-quality group（n=136） and a non-high-quality group（n=131）. LASSO regression was used to screen the predictors， and multivariate Logistic regression was established to construct the predictive model of QoL of TKA patients at 6 months after surgery. Receiver operating characteristic curve（ROC curve）， calibration curve and decision curve were drawn to evaluate its predictive value. Result：A total of 131 （49.06%） patients had lower QoL than normal 6 months after TKA. LASSO regression selected age， surgical side， education level， duration of knee osteoarthritis （KOA）， perioperative recovery， and visual analog scale （VAS） during preoperative rest as predictors. Multivariate logistic regression analysis showed that age （OR， 1.08； 95% CI， 1.01—1.15）， duration of KOA （OR， 1.13； 95% CI， 1.05—1.22）， VAS during preoperative rest （OR， 1.94； 95% CI， 1.20—3.15）， and perioperative rehabilitation （OR， 0.42； 95% CI， 0.18—0.97） were independent risk factors for lower than normal QoL in TKA patients at 6 months after surgery （P<0.05）. Statistically significant indicators in multivariate logistic regression analysis were included in the nomogram model，which underwent internal validation. The area under the ROC curve was 0.829 （95% CI， 0.779—0.879）， indicating good predictive ability of the model. The calibration curve showed good consistency between the predicted probability and the actual probability. The decision curve analysis showed that this model is clinically applicable. Conclusion：Age， duration of KOA， VAS during preoperative rest， and perioperative rehabilitation are independent predictors of lower-than-normal QoL in TKA patients at 6 months after surgery， and the nomogram model based on the above indicators had good predictive performance.