全膝关节置换术后膝关节功能恢复情况预测模型构建与验证-东南大学学报医学版

全膝关节置换术后膝关节功能恢复情况预测模型构建与验证

单位：山西医科大学附属运城市中心医院关节外科, 山西运城 044000

分类号：R687.4;R195.1

出版年·卷·期（页码）：2026·45·第三期（456-465）

摘要：

目的: 探讨全膝关节置换术(TKA)后膝关节功能恢复情况预测模型构建与验证,以期为临床工作提供参考。方法: 选取2021年8月至2023年8月于山西医科大学附属运城市中心医院接受初次TKA的238例患者为研究对象,随访12个月,剔除失访后,根据美国特种外科医院膝关节(HSS)评分系统评估膝关节功能恢复情况,并分别纳入恢复良好组、恢复不良组。统计所有患者膝关节功能恢复情况并分组,比较2组基线资料,LASSO回归及交叉验证法筛选变量,Logistic回归分析TKA后膝关节功能恢复不良的影响因素,构建术后膝关节功能恢复不良的Nomogram预测模型,并采用受试者工作特征(ROC)曲线、校准曲线、决策曲线对模型预测效能进行验证。结果: TKA术后12个月共233例成功完成随访,其中157例膝关节功能恢复良好,76例恢复不良,两组年龄、体质量指数(BMI)、糖尿病占比、骨质疏松症占比、术前HSS评分、术中行软组织平衡术占比、术后疼痛视觉模拟量表(VAS)评分、术后最大屈膝角度比较,差异有统计学意义(P＜0.05);LASSO回归筛选出年龄、骨质疏松症、术前HSS评分、术中行软组织平衡术、术后VAS评分、术后最大屈膝角度6个特征变量,Logistic回归分析显示,以上6个特征变量均是TKA后膝关节功能恢复不良的影响因素(P＜0.05);基于多因素结果构建Nomogram预测模型,内部及外部验证结果显示,在训练集与验证集中,ROC曲线显示该模型预测AUC分别为0.900(95%CI:0.857~0.943)、0.888(95%CI:0.795~0.981),校准曲线显示该模型膝关节功能恢复不良与临床实际一致性较高,Dxy分别为0.695、0.642,Brier分数分别为0.162、0.175,Hosmer-Lemeshow检验显示拟合良好(训练集:χ²=7.32,P=0.502;验证集: χ²=7.055,P=0.508),决策曲线显示,阈值概率0.2~0.8范围内运用该模型进行预测的净获益显著高于全部干预和全部不干预策略,且该模型预测结果与临床实际的符合率为95.00%,Kappa值为0.886(95%CI:0.633~0.979)(P＜0.05)。结论: 年龄、骨质疏松症、术前HSS评分、术中行软组织平衡术、术后VAS评分、术后最大屈膝角度均与TKA术后膝关节功能恢复不良密切相关,基于以上因素构建的Nomogram模型具有良好预测效能及临床适用性,可作为临床精准预测及个体化干预的辅助工具。

Objective: To explore the prediction model construction and verification of knee joint function recovery after total knee arthroplasty(TKA), so as to provide reference for clinical work. Methods: A total of 238 patients who underwent initial TKA in Yuncheng Central Hospital Affiliated to Shanxi Medical University from August 2021 to August 2023 were selected as the study subjects. After 12 months of follow-up, the patients were excluded after loss of follow-up. The recovery of knee joint function was evaluated according to the Hospital for Special Surgery(HSS) knee score system, and the patients were divided into good group and poor group. The recovery of knee joint function in all patients was counted and grouped. The baseline data of the two groups were compared. The variables were screened by LASSO regression and cross-validation. Logistic regression was used to analyze the influencing factors of poor recovery of knee joint function after TKA. The Nomogram prediction model of poor postoperative knee function recovery was constructed, and the receiver operating characteristic(ROC) curve, calibration curve and decision curve were used to verify the prediction efficiency of the model. Results: At 12 months after TKA, a total of 233 patients were successfully followed up. Among them, 157 patients had good recovery of knee joint function and 76 patients had poor recovery. There were significant differences in age, body mass index(BMI), proportion of diabetes mellitus, proportion of osteoporosis, preoperative HSS score, proportion of intraoperative soft tissue balance, postoperative pain visual analogue scale(VAS) score and postoperative maximum knee flexion angle between the two groups(P<0.05). Six characteristic variables including age, osteoporosis, preoperative HSS score, intraoperative soft tissue balance, postoperative VAS score and postoperative maximum knee flexion angle were screened out by LASSO regression. Logistic regression analysis showed that the above six characteristic variables were the influencing factors of poor recovery of knee joint function after TKA(P<0.05).The Nomogram prediction model was constructed based on the multi-factor results. The internal and external validation results showed that in the training set and the validation set, the ROC curve showed that the predicted AUC of the model were 0.900(95%CI: 0.857-0.943) and 0.888(95%CI: 0.795-0.981), respectively. The calibration curve showed that the poor recovery of knee joint function of the model was highly consistent with clinical practice, Dxy: 0.695, 0.642, Brier score: 0.162, 0.175. The Hosmer-Lemeshow test showed a good fit(training set: χ²=7.32, P=0.502; validation set: χ²=7.055, P=0.508). The decision curve showed that the net benefit predicted by the model was significantly higher than that of all intervention and all non-intervention strategies in the range of threshold probability 0.2-0.8. The coincidence rate between the predicted results of the model and the clinical practice was 95.00%, and the Kappa value was 0.886(95%CI: 0.633-0.979)(P<0.05). Conclusion: Age, osteoporosis, preoperative HSS score, intraoperative soft tissue balance, postoperative VAS score, and postoperative maximum knee flexion angle are closely related to the poor recovery of knee joint function after TKA. The Nomogram model based on the above factors has good predictive efficacy and clinical applicability, and can be used as an auxiliary tool for clinical accurate prediction and individualized intervention.

参考文献：

[1] 王杨,彭大霖,朱裕成,等.关节线参数对机械对齐全膝关节置换术中截骨量的影响[J].东南大学学报(医学版),2024,43(2):243-247.
[2] ARNOLD M,LUSTIG S,PASQUIER G,et al.Total knee arthroplasty:where are we after the "kinematic alignment" wave?[J].Orthop Traumatol Surg Res,2024,110(2):103838.
[3] DEFRANCE M J,SCUDERI G R.Are 20% of patients actually dissatisfied following total knee arthroplasty? a systematic review of the literature[J].J Arthroplasty,2023,38(3):594-599.
[4] ALI E,ADEDOYIN G,SIVAYOGANATHAN S,et al.Total knee arthroplasty in patients who have skeletal dysplasia:a center's experience with a mean 9-year follow-up[J].J Arthroplasty,2024,39(11):2745-2754.
[5] MEENA A,HOSER C,ABERMANN E,et al.Total knee arthroplasty improves sports activity and the patient-reported functional outcome at mid-term follow-up[J].Knee Surg Phys Traumatol Arthrosc,2023,31(3):1795-1802.
[6] RUBERTO R A,FARAH O,LOPEZ C,et al.Perioperative intravenous glucocorticoids in total joint arthroplasty:a systematic review[J].J Am Acad Orthop Surg,2023,31(2):e94-e106.
[7] PALANNE R,RANTASALO M,VAKKURI A,et al.Fat tissue is a poor predictor of 1 year outcomes after total knee arthroplasty:a secondary analysis of a randomized clinical trial[J].Scand J Surg,2023,112(1):22-32.
[8] 冯泽宇,许嘉政,任海潮,等.处在学习曲线内的机器人辅助全膝关节置换术是否会对关节功能产生不良影响:一项倾向性评分研究[J].中华保健医学杂志,2025,27(3):465-470.
[9] NARIN S,UNVER B,BAKIRHAN S,et al.Crosscultural adaptation reliability and validity of the Turkish version of the Hospital for Special Surgery HSS Knee Score[J].Acta Orthop Traumatol Turc,2014,48(3):241-248.
[10] 赵猛,孙翠云,卫斐然,等.三维振动疗法在人工全膝关节置换术后康复治疗中的应用进展[J].现代医学,2024,52(9):1471-1474.
[11] BERGSTEIN V E,WEINBLATT A I,TAYLOR W L,et al.Total knee arthroplasty survivorship and outcomes in young patients:a review of the literature and 40-year update to a longitudinal study[J].Arch Orthop Trauma Surg,2024,144(9):4077-4083.
[12] MACDESSI S J,OUSSEDIK S,ABDEL M P,et al.The language of knee alignment:updated definitions and considerations for reporting outcomes in total knee arthroplasty[J].Bone Joint J,2023,105-B(2):101-108.
[13] ONGUN P,KAYA C,KIZILAY Y O,et al.Quality-of-life outcomes and influencing factors following robotic-assisted versus manual total knee arthroplasty:a comparative analysis[J].J Rob Surg,2025,19(1):527-532.
[14] 王子豪,李昕华,蒋慧萍,等.全膝关节置换术后短期膝关节功能及其影响因素[J].中国康复理论与实践,2024,30(1):111-118.
[15] LI Y,LUO R,LUO S,et al.Influencing factors analysis of asymmetry in knee adduction moment among patients with unilateral knee osteoarthritis[J].BMC Musculoskelet Disord,2024,25(1):832-837.
[16] ZHENG Y,PENG L R,ZHAO H G.Analysis of risk factors for postoperative functional recovery of fractures around knee joint[J].China J Orthop Traumatol,2024,37(12):1219-1223.
[17] YIN Q,LIU Y,YIN F,et al.The role of the infrapatellar fat pad in osteoporotic knee osteoarthritis:bone-fat crosstalk:cross-sectional studies[J].J Orthop Surg Res,2025,20(1):969-975.
[18] WHITING P S,HARE K,KRUEGER D,et al.Periprosthetic fractures are osteoporotic fractures:missed opportunities for osteoporosis diagnosis[J].Osteoporos Int,2024,35(7):1165-1171.
[19] 高迎娜,李慧娟.膝骨性关节炎全膝关节置换术术后关节功能康复水平的因素分析及预测模型建立[J].临床医学,2025,45(8):35-40.
[20] ZHANG H,WANG J,WANG B,et al.Effect of lower-limb isokinetic muscle strengthening on knee function and joint contact force in knee osteoarthritis patients awaiting total knee arthroplasty:study protocol for a randomized controlled trial[J].J Orthop Surg Res,2024,19(1):677-683.
[21] JIAO S,FENG Z,HUANG J,et al.Enhanced recovery after surgery combined with quantitative rehabilitation training in early rehabilitation after total knee replacement:a randomized controlled trial[J].Eur J Phys Rehabil Med,2024,60(1):74-83.
[22] 欧锐金,莫楚鸿,谭建基,等.膝关节置换术与内镜下膝关节镜清理术对膝骨性关节炎患者Lysholm评分、HSS评分的影响[J].中国伤残医学,2023,31(4):1-5.
[23] SHEN M,LI H,LIU H,et al.The impact of soft tissue release on radiological and clinical outcomes in functional alignment robot-assisted total knee arthroplasty[J].J Orthop Surg Res,2025,20(1):940-945.
[24] DHUNGANA H,JANGID S,GOYAL M.Alignment techniques in total knee arthroplasty:where do we stand today?[J].Chin Med Sci J,2024,39(3):217-225.
[25] 曹常松,关健,侯颖周,等.两种不同术式治疗内侧单间室膝关节骨性关节炎患者对髌骨高度相关指数及HSS、VAS评分的影响[J].海南医学,2025,36(13):1890-1894.
[26] 王权,邵强.关节镜手术治疗中老年SDLM患者的中期效果及安全性研究[J].浙江创伤外科,2023,28(5):952-954,958.
[27] 徐浩煌,李泽龙,辜海洋,等.全膝关节置换术后患者长期生活质量现状及影响因素分析[J].中国医药指南,2025,23(28):65-68.
[28] ZHOU J,ZHAO F,CHEN C,et al.Influencing factors of flexion contracture among patients with knee osteoarthritis:a multi-center retrospective cross-sectional study[J].Sci Rep,2025,15(1):32144.
[29] YU W,ZOU D,TAN J,et al.Medial-pivot total knee arthroplasty enhances tibiofemoral axial rotation stability in weight-bearing mid-range flexion compared to posterior-stabilised system[J].Knee Surg Phys Traumatol Arthrosc,2024,32(8):2075-2086.
[30] ABBAS N,KHATTAK S K,FAHEEM M U,et al.Safe zone of joint line elevation for the treatment of knee flexion contracture preventing mid-flexion instability in total knee replacement[J].J Ayub Med Coll Abbottabad,2024,36(2):234-239.
[31] ÁLVAREZ BENITO M.A bimodal nomogram:a non-invasive tool to assist breast radiologists in decision-making[J].Eur Radiol,2024,34(4):2605-2607.

服务与反馈：

【文章下载】【发表评论】【查看评论】【加入收藏】

提示：您还未登录，请登录！点此登录