肩袖损伤的超声诊断及预后预测

Ultrasonic diagnosis and prognosis prediction of rotator cuff injury

目的探讨超声对肩袖损伤的诊断价值,构建预测模型并分析其在肩袖损伤患者预后中的临床应用价值。方法回顾性分析2021年2月至2023年6月我院82例疑似肩袖损伤患者的临床资料和超声图像特征,评估超声与关节镜诊断结果的一致性。采用单因素及多因素Logistic回归分析筛选预测肩袖损伤患者预后的影响因素并构建预测模型;绘制受试者工作特征(ROC)曲线分析该模型预测肩袖损伤患者预后的效能。另选我院2023年7~11月收治的21例肩袖损伤患者进行临床外部验证。结果纳入的82例疑似肩袖损伤患者经关节镜检查确诊为肩袖损伤73例,超声诊断肩袖损伤的灵敏度为90.41%,特异度为88.89%,准确率为90.24%,与关节镜诊断结果的一致性中等(Kappa=0.614)。73例肩袖损伤患者预后良好者59例,预后不良者14例。单因素及多因素Logistic回归分析显示,年龄、肩袖撕裂程度、肩袖肌肉萎缩程度均为预测肩袖损伤患者预后的独立危险因素(OR=1.634、2.869、2.389,均P<0.05),由此构建关于肩袖损伤患者预后的预测模型。ROC曲线分析显示,该模型预测肩袖损伤患者预后的曲线下面积为0.805(95%可信区间:0.691~0.919,P<0.001),灵敏度为71.43%,特异度为77.97%。临床外部验证显示,该模型预测肩袖损伤患者预后的准确率为80.95%,与临床诊断结果的一致性中等(Kappa=0.481)。结论超声对肩袖损伤有较高的诊断价值;基于年龄、肩袖撕裂程度、肩袖肌肉萎缩程度构建的预测模型在预测肩袖损伤患者预后方面有一定的指导意义。

Objective To explore the value of ultrasound in the diagnosis of rotator cuff injury,and to analyze the clinical application value of constructed predictive model in predicting the prognosis in the patients with rotator cuff injury.Methods A retrospective analysis was conducted on the clinical data and ultrasound image features of 82 patients who were suspected of rotator cuff injury and admitted to the hospital from February 2021 to June 2023.The consistency between diagnostic results of musculoskeletal ultrasound and arthroscopy was evaluated.Univariate and multivariate Logistic regression analysis were used to screen the influencing factors for the prognosis of patients with rotator cuff injury,and a predictive model was constructed.Receiver operating characteristic(ROC)curve was drawn to analyze the efficacy of the model in predicting the prognosis in patients with rotator cuff injury.Another 21 patients with rotator cuff injury admitted to the hospital from July to November 2023 were selected for clinical external validation.Results Among the 82 included patients with suspected rotator cuff injury,73 were diagnosed as rotator cuff injury by arthroscopy.The sensitivity,specificity and accuracy of ultrasound for diagnosing rotator cuff injury were 90.41%,88.89%and 90.24%,respectively.The consistency with arthroscopy was at a middle level(Kappa=0.614).Among 73 patients with rotator cuff injury,59 patients had good prognosis and 14 patients had poor prognosis.Univariate and multivariate Logistic regression analysis showed that age,the degree of rotator cuff tear and the degree of rotator cuff muscle atrophy were independent risk factors for predicting the poor prognosis in patients with rotator cuff injury(OR=1.634,2.869,2.389,all P<0.05).A predictive model for predicting the prognosis in patients with rotator cuff injury was constructed.The area under the ROC curve of the model was 0.805(95%confidence interval:0.691~0.919,P<0.001),the sensitivity and specificity were 71.43%and 77.97%,respectively.Clinical external validation showed that the accuracy of the model for predicting the prognosis in patients with rotator cuff injury was 80.95%,and the consistency with clinical diagnosis results was at a moderate level(Kappa=0.481).Conclusion Ultrasound has high value in the diagnosis of rotator cfuff injury.The predictive model constructed on age,the degree of rotator cuff tear and the degree of rotator cuff muscle atrophy has a certain value in predicting the prognosis in patients with rotator cuff injury.