Page 62 - 南京医科大学自然版

南京医科大学学报（自然科学版） 第45卷第1期
· 56 · Journal of Nanjing Medical University（Natural Sciences） 2025年1月

·临床研究·

单孔胸腔镜术后慢性疼痛预测模型的建立及验证

蒋 捷，刘 锋，王 波，王 芹，钟 健 *
南京医科大学附属脑科医院胸外科，江苏 南京 210029

［摘 要］ 目的：探讨单孔电视辅助胸腔镜手术（video⁃assisted thoracic surgery，VATS）后慢性疼痛的危险因素，建立预测模型
并加以验证。方法：回顾性分析2023年1—6月在南京医科大学附属脑科医院胸外科行单孔VATS术的302例患者，收集患者
临床资料，采用数字评分量表（numeric rating scale，NRS）评定患者术后3个月的疼痛程度，分为疼痛组及非疼痛组。按照7∶3比例
随机分为训练集（n=214）和验证集（n=88），对训练集采用单因素分析和多因素Logistic回归分析，并以此建立预测模型。采用
受试者工作特征（receiver operating characteristic，ROC）曲线评估模型区分度、校准曲线评估模型的一致性、决策曲线分析
（decision curve analysis，DCA）评估该模型的临床价值，并在验证集中校验模型。结果：多因素分析显示年龄（OR=0.925，95%
CI：0.872~0.981，P=0.009）、术后胸腔闭式引流时间（OR=1.273，95%CI：1.018~1.591，P=0.034）、术后第 1 天 C 反应蛋白

（C⁃reactive protein，CRP）（OR=1.090，95%CI：1.030~1.153，P=0.003）、术后第 1 天 NRS（OR=3.060，95%CI：1.879~4.981，P <
0.001）为单孔VATS后慢性疼痛的独立危险因素，基于此构建预测模型，ROC曲线下面积0.871（95% CI：0.799~0.943），最大约
登指数所对应的临界值为0.195，此时灵敏度为76.5%，特异度为85.6%，Bootstrap法抽样1 000次，校准曲线图预测的慢性疼痛
概率与实际风险高度一致。DCA曲线表明在全预测概率下均有正收益。结论：单孔VATS后，患者年龄、术后胸腔闭式引流时
间、术后第 1 天 CRP 值、术后第 1 天 NRS 评分值均为术后慢性疼痛的独立危险因素，该预测模型有助于准确预测术后慢性疼
痛，有较好的临床应用价值。
［关键词］ 电视辅助胸腔镜手术；术后慢性疼痛；NRS评分；预测模型；列线图
［中图分类号］ R655 ［文献标志码］ A ［文章编号］ 1007⁃4368（2025）01⁃56⁃08
doi：10.7655/NYDXBNSN240887

Construction and validation of a predictive model for chronic pain after single⁃port video⁃
assisted thoracic surgery

*
JIANG Jie，LIU Feng，WANG Bo，WANG Qin，ZHONG Jian
Department of Thoracic Surgery，the Affiliated Brain Hospital of Nanjing Medical University，Nanjing 210029，China

［Abstract］ Objective：To explore the risk factors for chronic pain after single ⁃ port video ⁃ assisted thoracic surgery（VATS）and
establish the predictive model to validate it. Methods：A retrospective analysis was conducted on 302 patients who underwent single⁃
port VATS at department of thoracic surgery，the Affiliated Brain Hospital of Nanjing Medical University from January to June 2023.
Clinical data of patients were collected. A numerical rating scale（NRS）was used to assess the degree of pain in patients three months
after surgery，and the patients were divided into the pain group and the non⁃pain group. The patients were also randomly divided into a
training set（n=214）and a validation set（n=88）at a 7∶3 ratio，and univariate analysis was performed on the training set. Multivariate
logistic regression analysis was used to establish a predictive model. Reciever operating characteristic（ROC）curve was used to
evaluate model discrimination，calibration curve was used to evaluate model consistency，decision curve analysis（DCA）was used to
evaluate the clinical value of the model，and the model was verified in the validation set. Results：The multivariate analysis showed
that age（OR=0.925，95%CI：0.872-0.981，P=0.009），postoperative closed chest drainage time（OR=1.273，95%CI：1.018-1.591，P=
0.034），C⁃reactive protein（CRP）value on the first day after surgery（OR=1.090，95%CI：1.030-1.153，P=0.003），and NRS score on
the first day after surgery（OR=3.060，95%CI：1.879-4.981，P < 0.001）were independent risk factors for chronic pain after single⁃port
thoracoscopy. Based on this，a prediction model was constructed，with an area under the ROC curve of 0.871（95%CI：0.799-0.943）
and the cut⁃off value corresponding to the maximum Yoden index was to 0.195 with a sensitivity of 76.5% and a specificity of 85.6%.

［基金项目］ 江苏省南京市卫生科技发展专项资金项目（YKK23148）
通信作者（Corresponding author），E⁃mail：1162014036@qq.com（ORCID：0009⁃0000⁃2265⁃5858）
∗