脊柱侧凸椎体旋转半脱位三种测量方法的可重复性和可信度研究

Reproducibility and reliability analysis for the three manual radiographic measurements of rotatory subluxation in scoliosis

目的:分析MB法(midbody,MB)、EP法(endplate,EP)和C法(centroid,C)在椎体旋转半脱位(rotatory subluxation,RS)测量中的可重复性和可信度,探讨三种测量方法的影响因素。方法:回顾性分析2012年3月-2020年6月于南京大学医学院附属鼓楼医院行手术治疗的59例脊柱侧凸畸形患者,其中先天性脊柱侧凸15例,特发性脊柱侧凸9例,退变性脊柱侧凸16例,神经肌源性脊柱侧凸19例;男17例,女42例,平均年龄34.0±18.8岁。3名脊柱外科医师均使用MB法、EP法及C法分别间隔两周测量RS值,用组内相关系数(intraclass correlation coefficient,ICC)对三种测量方法结果进行一致性分析。结果:59例患者平均主弯Cobb角为64.2°±23.9°。RS节段分别为T4/5节段1例(1.7%)、T5/6节段2例(3.4%)、T6/7节段3例(5.1%)、T7/8节段3例(5.1%)、T9/10节段1例(1.7%)、T10/11节段1例(1.7%)、T12/L1节段6例(10.2%)、L1/2节段5例(8.5%)、L2/3节段9例(15.3%)、L3/4节段16例(27.1%)、L4/5节段12例(20.3%)。MB法、EP法、C法测量的RS值分别为16.5±7.1mm、8.5±4.2mm、14.2±6.3mm。三种方法的观察者内一致性分别为MB法0.77-0.91、EP法0.78-0.81、C法0.82-0.94;观察者间一致性分别为MB法0.87、EP法0.90、C法0.86。结论:MB法、EP法和C法在RS的测量中均有较高的可重复性及可信度,其中C法的组内一致性最高、EP法的组间一致性最高。

Objectives: To analyze the reproducibility and reliability of midbody(MB) method, endplate(EP method and centroid(C) method in the measurements of rotatory subluxation(RS), and to investigate their respective influencing factors. Methods: 59 spinal deformity patients with RS treated with surgery in Drum Tower Hospital affiliated to Nanjing University Medical School from March 2012 to June 2020 were retrospectively reviewed. There were 15 cases of congenital scoliosis, 9 cases of idiopathic scoliosis, 16 cases of degenerative scoliosis and 19 cases of neuromyogenic scoliosis. There were 17 males and 42 females, and the average age was 34.0±18.8 years. Three spinal surgeons each performed the measurements using the three methods with a time interval of 2 weeks. Subsequently, the intraclass correlation coefficient(ICC) was applied for reliability analysis. Results: The average main curve Cobb angle was 64.2°±23.9° for the cohort. The RS levels were T4/5 in 1 patient(1.7%), T5/6 in 2(3.4%), T6/7 in 3(5.1%), T7/8 in 3(5.1%), T9/10 in 1(1.7%),T10/11 in 1(1.7%), T12/L1 in 6(10.2%), L1/2 in 5(8.5%), L2/3 in 9(15.3%), L3/4 in 16(27.1%), and L4/5 in12(20.3%). The RS values measured by MB, EP and C methods were respectively 16.5 ±7.1mm, 8.5±4.2mm,and 14.2±6.3mm. The intra-observer reliabilities were 0.77-0.91 for MB method, 0.78-0.81 for EP method and 0.82-0.94 for C method, while the inter-observer reliabilities were 0.87 for MB method, 0.90 for EP method and 0.86 for C method. Conclusions: The MB method, EP method and C method had high reproducibility and reliability in the measurement of RS, among which C method showed the best intra-observer reliability and EP method showed the best inter-observer reliability.