基于LASSO回归的股骨颈骨折术后疼痛影响因素筛选研究-现代医学

基于LASSO回归的股骨颈骨折术后疼痛影响因素筛选研究

作者：吴琼陈澎

单位：南京市中心医院手术室, 江苏南京 210000

分类号：R683.4

出版年·卷·期（页码）：2025·53·第十一期（1774-1777）

摘要：

目的：基于LASSO回归筛选股骨颈骨折术后患者疼痛评分的关键影响因素。方法：收集我院2021年1月—2025年4月330例股骨颈骨折术后患者临床资料，以0~10分数字疼痛评分(NRS)为疼痛结局变量。采用Spearman秩相关分析各指标与疼痛评分之间的关系。所有变量通过LASSO回归模型进行变量筛选，并采用10折交叉验证确定最优正则化参数λ。计算模型的平均均方误差(MSE)、平均绝对误差(MAE)和决定系数(R²)评估预测性能。结果： LASSO回归筛选出14个关键变量，包括年龄、动脉血氧分压、吸入氧浓度、碱剩余、钙、体温、钠、乳酸、血糖、阴离子间隙、肌酐、呼吸频率、血氧饱和度及收缩压。最终模型R²为0.73，MSE为1.24，MAE为0.71，具有良好拟合效果。结论：术后患者的多项临床指标与疼痛评分存在相关性，基于LASSO回归可有效筛选关键影响因素，为术后疼痛评估和管理提供量化参考。

Objective: To identify key factors influencing postoperative pain scores in patients with femoral neck fractures using LASSO regression. Methods: A retrospective cohort of 330 patients with femoral neck fractures in our hospital from Jan. 2021 to Apr. 2025 was analyzed. The outcome was postoperative pain assessed by a 0-10 numerical rating scale(NRS). Spearman rank correlation was used for univariate analysis of the association between clinical variables and pain scores. All variables were screened using the LASSO regression model, and the optimal regularization parameter λ was determined by 10-fold cross-validation. Model performance was assessed using mean squared error(MSE), mean absolute error(MAE), and determination coefficient(R²). Results: Fourteen predictors were selected by LASSO: age, arterial oxygen partial pressure, fraction of inspired oxygen, base excess, calcium, temperature, sodium, lactate, glucose, anion gap, creatinine, respiratory rate, blood oxygen saturation, and systolic pressure. The final model achieved an R² of 0.73, MSE of 1.24, and MAE of 0.71, indicating a good fit. Conclusion: Multiple clinical indicators are correlated with postoperative pain scores. The LASSO regression model can effectively identify key predictors and provide a quantitative reference for postoperative pain assessment and management.

参考文献：

[1] SANGHAVI S,PATWARDHAN S,SHYAM A,et al.Nonunion in pediatric femoral neck fractures[J].J Bone Joint Surg Am,2020,102(11):1000-1010.
[2] JACKSON M,LEARMONTH I D.The treatment of nonunion after intracapsular fracture of the proximal femur[J].Clin Orthop Relat Res,2002(399):119-128.
[3] BERNSTEIN E M,KELSEY T J,COCHRAN G K,et al.Femoral neck stress fractures:an updated review[J].J Am Acad Orthop Surg,2022,30(7):302-311.
[4] HUTCHINSON H L,JAEKEL D J,LOVALD S T,et al.Multimodal pain management of femoral neck fractures treated with hemiarthroplasty[J].J Surg Orthop Adv,2019,28(1):58-62.
[5] FLETCHER J W A,SOMMER C,ECKARDT H,et al.Intracapsular femoral neck fractures-a surgical management algorithm[J].Medicina,2021,57(8):791.
[6] 高万露,汪小海.患者术后拒绝疼痛评分影响因素的logistic回归分析[J].数理医药学杂志,2014,27(3):331-333.
[7] ABU SHADEQUE MULLAH M,HANLEY J A,BENEDETTI A.LASSO type penalized spline regression for binary data[J].BMC Med Res Methodol,2021,21(1):83.
[8] WANG C,LI Q,SONG X,et al.Bayesian adaptive lasso for additive hazard regression with current status data[J].Stat Med,2019,38(20):3703-3718.
[9] KANG K,SONG X,HU X J,et al.Bayesian adaptive group lasso with semiparametric hidden Markov models[J].Stat Med,2019,38(9):1634-1650.
[10] KAWAGUCHI E S,DARST B F,WANG K,et al.Sign-based shrinkage based on an asymmetric LASSO penalty[J].J Data Sci,2021,19(3):429-449.
[11] SABOURIN J A,VALDAR W,NOBEL A B.A permutation approach for selecting the penalty parameter in penalized model selection[J].Biometrics,2015,71(4):1185-1194.
[12] 安亚杰,高峰,李玉莹,等.背侧海马乳酸激活BDNF-ERK1/2-Arc信号通路缓解神经病理性疼痛与记忆障碍共病[J].解剖学研究,2025,47(2):97-104,121.
[13] 赵雯.间歇性禁食通过上调前扣带回中的SIRT3降低乳酸含量缓解小鼠神经病理性疼痛[D].郑州:郑州大学,2023.
[14] 胡茵茵.星形胶质细胞-神经元乳酸穿梭在神经性疼痛中的作用[D].南昌:南昌大学,2022.
[15] 聂鸿飞.支链氨基酸代谢障碍在神经病理性疼痛中的作用机制研究[D].成都:四川大学,2021.
[16] 马丈玉.交感神经信号促进下颌关节骨关节炎疼痛的机制研究[D].西安:中国人民解放军空军军医大学,2025.
[17] 余帅江,沈鹏,赵新,等.交感神经松动加强技术治疗骨折术后复杂性区域疼痛综合征1例[J].四川医学,2021,42(11):1174-1177.
[18] 叶海涛,杨红军.交感神经、炎性细胞因子与神经病理性疼痛的关系[J].中国微侵袭神经外科杂志,2017,22(2):91-93.
[19] 张莉,吴鹏,黄韶敏,等.交感神经纤维芽生与神经病理性疼痛[J].国际骨科学杂志,2013,34(5):333-335.
[20] 赵薇,王碧,杨波,等.连续腰交感神经阻滞治疗下肢放射性疼痛的临床观察[J].中华保健医学杂志,2025,27(3):503-506.

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