膝关节镜术后慢性期疼痛预测模型的构建与验证

Construction and Validation of a Pain Prediction Model for the Chronic Stage After Knee Arthroscopy

目的探究膝关节镜术后慢性期中、重度疼痛的危险因素,构建并验证预测模型。方法收集2019年9月至2021年2月在唐都医院收治行膝关节镜手术的236例患者临床资料,根据术后3个月时的疼痛数字评分(numeric rating scale,NRS),将NRS≥4分患者纳入疼痛组、NRS<4分患者纳入无痛组。术后慢性期疼痛组29例,其中男13例,女16例;平均年龄(62.03±9.52)岁。无痛组207例,其中男114例,女93例;平均年龄(40.00±12.46)岁。采用多因素Logistic回归分析比较两组患者术后慢性期疼痛(chronic postsurgical pain,CPSP)的独立危险因素。根据筛选得到临床指标,应用R语言构建列线图预测模型,采用BootStrap自助抽样法进行内部验证,通过受试者工作特征(receiver operating characteristic,ROC)曲线,校准曲线评价模型的预测能力。结果单因素分析结果显示,术后慢性期疼痛危险因素为年龄、身体质量指数(body mass index,BMI)、关节软骨损伤Outerbridge分级、术后72 h NRS≥4分、是否患有糖尿病。多因素分析结果显示,高龄、BMI≥24 kg/m^(2)、Outerbridge分级Ⅲ~Ⅳ级、术后72 h NRS≥4分是导致膝关节镜术后慢性期疼痛的独立危险因素。基于多因素分析结果构建列线图预测模型,内部验证显示C-index为0.968,ROC曲线下面积为0.969,校准曲线的预测结果和实际观察结果一致性良好。结论高龄、BMI≥24 kg/m^(2)、Outerbridge分级Ⅲ~Ⅳ级、术后72 h NRS≥4分是膝关节镜术后慢性期疼痛的危险因素,基于上述因素构建的列线图模型具有较高的准确度和临床实用性。

Objective To explore the risk factors for moderate to severe pain in the chronic stage after knee arthroscopy and to construct and validate a prediction model.Methods Clinical data were collected from^(2)36 patients who underwent knee arthroscopic surgery at Tangdu Hospital between September 2019 and February 2021.Patients with NRS≥4 points were included in the pain group,and those with NRS<4 points were included in the pain free group according to the numerical rating scale(NRS)for pain at 3 months after surgery.29 patients,including 13 males and 16 females,with a mean age of(62.03±9.52)years were in the postoperative chronic stage pain group and 207 patients,including 114 males and 93 females,with a mean age of(40.00±12.46)years.Multivariate logistic regression analysis was used to compare the independent risk factors for chronic postoperative pain(CPSP)between the two groups.Based on the clinical indexes obtained from screening,a nomogram prediction model was constructed in R language,and bootstrap bootstrap sampling was used for internal validation.The predictive ability of the model was evaluated by receiver operating characteristic(ROC)curve and calibration curve.Results The results of univariate analysis showed that the risk factors for pain in the chronic stage after surgery were age,body mass index(BMI),Outerbridge grade of articular cartilage damage,NRS≥4 points at 72 h after surgery,and having diabetes mellitus or not.In multivariate analysis,advanced age,BMI≥24 kg/m^(2),Outerbridge gradesⅢtoⅣ,and NRS≥4 points at 72 h postoperatively were independent risk factors for pain in the chronic stage after knee arthroscopy.A nomogram prediction model was constructed based on the results of multivariate analysis,and internal validation showed that the C-index was 0.968,the area under the ROC curve was 0.969,and the predictive results of the calibration curve and the actual observation were in good agreement.Conclusion Advanced age,BMI≥24 kg/m^(2),Outerbridge gradesⅢtoⅣ,and NRS≥4 points at 72 h after knee arthroscopy are the risk factors for pain in the chronic stage,and the nomogram model constructed based on the above factors has high accuracy and clinical usefulness.