特发性脊柱侧凸PUMC分型的可重复性及可信度研究

Prospective study of interobserver and intraobserver repeatability and reliability of PUMC classification for idiopathic scoliosis

目的:探讨特发性脊柱侧凸PUMC(协和)分型系统的一致性,并对影响分型一致性的因素进行分析。方法:随机选取南京鼓楼医院2004年～2006年治疗的80例特发性脊柱侧凸患者,其中男性15例,女性65例,平均年龄14.6岁(10～18岁),均有完整的术前站立位正侧位X线片及仰卧位左右Bending片共四张。由4名脊柱外科专科医生分别根据PUMC(协和)分型标准进行分型,2周后此4位医生分别对这些患者的X线片再次进行分型,收集分型结果分别作可信度和可重复性分析,计算Kappa值检验一致性并对影响分型一致性的因素进行统计分析。结果:80例患者均可用PUMC分型,分型可信度平均Kappa值为0.801,可重复性Kappa值为0.878。PUMC分型不一致中包括上胸弯的界定18次;胸弯明显时,代偿性与结构性腰弯的界定15次;腰弯明显时,代偿性与结构性胸弯的界定18次;单弯顶点的判断20次;测量角度的差异18次。结论:PUMC(协和)分型学习曲线相对较短,易于掌握且具有良好的可信度和可重复性。

Objective：To determine the interobserver and intraobserver repeatability and reliability of PUMC classification for idiopathic scoliosis and review the factors correlating to this.Method：Eighty patients with idiopathic scoliosis treated in Drum Tower hospital from 2004 to 2006 were randomly chosen for study,which included 15 males and 65 females with a mean age of 14.6 years old （range,10 to 18 years old）.Pre-operative standing AP and lateral view X-ray as well as supine lateral bending were taken in all patients.Four spine surgeons independently assign the curve type to each patients following the guidance described by Qiu Guixing et al.The procedure was repeated two weeks later with the patients presented in different order.Kappa value was used to determine the intraobserver and interobserver repeatability and reliability.Result：All curves could be assigned into the PUMC classification system.A mean kappa value of 0.801 for interobserver reliability were determined,and a mean kappa of 0.878 for intraobserver repeatability.The disagreement evidenced included the determination of the upper thoracic curve（18 times）,distinguishment between structural and nonstructural lumbar curve secondary to thoracic major curve （15 times）,distinguishment between structural and nonstructural thoracic curve secondary to lumbar major curve （18 times）,determination of apic vertebral from single curve（ 18 times） and angular variation due to each curve.Conelnsion：PUMC new classification has relative short-time learning curve and is more reliable and repeatability for practical application.