LASSO回归筛选骨质疏松性椎体压缩骨折患者术后残余疼痛的预测因子及模型构建

LASSO regression to identify predictors and model construction for postoperative residual pain in patients with osteoporotic vertebral compression fracture

目的探讨骨质疏松性椎体压缩骨折(OVCF)患者经皮穿刺椎体成形术(PVP)术后残余疼痛的因素,并基于LASSO模型筛选相关因素构建Nomgram模型,旨在为早期识别术后残余疼痛高危人群提供参考。方法采用回顾性研究,将2022年1月—2023年12月杭州市富阳中医骨伤医院PVP术后存在残余疼痛的60例OVCF患者纳入残余疼痛组,将同期无残余疼痛的60例患者纳入无残余疼痛组。比较2组基线资料、围手术期指标;采用LASSO回归筛选相关危险因素进行logistic回归分析,构建Nomgram模型并进行内部验证。结果通过LASSO回归模型筛选出6个潜在的相关因素,经logistic回归分析,结果显示,TLF损伤、骨水泥渗漏、关节突关节侵扰、骨水泥注入量多是影响OVCF患者PVP术后残余疼痛的独立危险因素(OR>1,P<0.05),骨密度(BMD)T值高、术后椎体高度恢复率高是保护因素(OR<1,P<0.05);Nomgram模型结果显示,C-index值为0.926,模型具有良好的区分度,校准曲线整体趋势接近理想曲线,ROC曲线显示,列线图预测OVCF患者PVP术后残余疼痛的AUC为0.926,预测价值较高。结论基于BMD T值、TLF损伤、骨水泥渗漏、关节突关节侵扰、骨水泥注入量、术后椎体高度恢复率构建OVCF患者PVP术后残余疼痛Nomgram模型,可有效预测OVCF患者PVP术后残余疼痛的风险。

Objective To investigate factors of residual pain following percutaneous vertebroplasty(PVP)in patients with osteoporotic vertebral compression fracture(OVCF)and to construct a Nomgram model based on LASSO regression for identifying relevant factors.This aims to provide reference for early identification of high-risk groups for postoperative residual pain.Methods A retrospective study was conducted on the efficacy of acceptance commitment therapy combined with a collaborative care model in esophageal cancer patients undergoing radiotherapy and chemotherapy.Clinical data of 60 OVCF patients with residual pain after PVP surgery from January 2022 to December 2023 in Fuyang Traditional Chinese Medicine Orthopedics and Trauma Hospital were included in the residual pain group,while clinical data of 60 patients without residual pain during the same period were included in the no residual pain group.Compare the baseline data and perioperative indicators of the two groups.Utilize LASSO regression to identify relevant risk factors and perform logistic regression analysis to develop a nomogram model with internal validation.Results Six potential related factors were screened using the LASSO regression model.The results of logistic regression analysis showed that TLF injury,bone cement leakage,facet joint invasion,and excessive bone cement injection were independent risk factors affecting residual pain in OVCF patients after PVP surgery(OR>1,P<0.05).In contrast,high BMD T values and significant postoperative vertebral height recovery rate were identified as protective factors(OR<1,P<0.05).The constructed Nomgram model exhibited a C-index value of 0.926,indicating strong discrimination.The overall trend of the calibration curve was close to the ideal curve.The ROC curve analysis predicted an AUC of 0.926 for residual pain after PVP in OVCF patients,indicating high predictive validity.Conclusion A Nomogram model incorporating BMD T value,TLF injury,bone cement leakage,facet joint invasion,bone cement injection volume,and postoperative vertebral height recovery rate,can effectively predict the risk of residual pain in OVCF patients after PVP surgery.