摘要目的:检验和评价骨质疏松性胸腰椎骨折（OTLF）评分及分型系统的可信度及临床应用效果。方法:采用多中心回顾性病例系列研究分析2021年1月至2022年6月西安交通大学附属红会医院等8家医院收治的530例OTLF患者的临床资料，其中男212例，女318例；年龄55~90岁［（72.6±10.8）岁］。美国脊髓损伤协会（ASIA）分级：C级4例，D级18例。根据骨质疏松性胸腰椎损伤分类及严重程度（OTLICS）评分，本组患者OTLICS评分均>4分，需行手术治疗。急性症状性OTLF（ASOTLF）分型410例，其中Ⅰ型24例，ⅡA型159例，ⅡB型47例，ⅡC型31例，ⅢA型136例，Ⅲ B型8例，Ⅳ型（无神经症状）2例，Ⅳ型（合并神经症状）3例。陈旧性症状性OTLF（CSOTLF）分型120例，其中Ⅰ型62例，Ⅱ型21例，Ⅲ型17例，Ⅳ型（全身麻醉下可复位）3例，Ⅳ型（全身麻醉下无法复位）9例，Ⅴ型（全身麻醉下可复位）1例，Ⅴ型（合并神经症状）5例，Ⅴ型（全身麻醉下无法复位）2例。患者分别采用经皮椎体成形术（PVP）、体位复位+PVP、经皮椎体后凸成形术（PKP）、后路切开复位植骨融合钉道强化内固定术、后路切开复位减压植骨融合钉道强化内固定术、后路切开复位截骨矫形植骨融合钉道强化内固定术。采用加权Kappa方法检验OTLICS评分、ASOTLF分型及CSOTLF分型的观察者间及观察者内可信度。比较术前、术后1个月及末次随访时视觉模拟评分（VAS）、Oswestry功能障碍指数（ODI）、美国脊髓损伤协会（ASIA）分级。观察术后并发症发生情况。结果:OTLICS评分观察者间平均评分一致性的百分比为93.4%，可信度Kappa值均值为0.86；观察者内平均评分一致性的百分比为93.0%，可信度Kappa值均值为0.86。ASOTLF分型观察者间平均分型一致性的百分比为94.2%，可信度Kappa值均值为0.84；观察者内平均分型一致性的百分比为92.5%，可信度Kappa值均值为0.83。CSOTLF分型观察者间平均评分一致性的百分比为91.9%，可信度Kappa值均值为0.80；观察者内平均评分一致性的百分比为91.3%，可信度Kappa值均值为0.81。患者均获随访6~12个月［（9.0±2.1）个月］。术后1个月及末次随访时ASOTLF分型和CSOTLF分型患者VAS、ODI较术前明显降低（ P均<0.05）。其中ASOTLF分型ⅡA型、ⅡB型、ⅡC型、ⅢA型和Ⅳ型患者末次随访时VAS较术后1个月明显降低，Ⅰ型、ⅡA型、ⅡB型、ⅢA型、ⅢB型和Ⅳ型患者末次随访时ODI较术后1个月明显降低；CSOTLF分型Ⅱ型、Ⅲ型、Ⅳ型和Ⅴ型患者末次随访时VAS较术后1个月明显降低，所有分型患者末次随访时ODI较术后1个月明显降低（ P均<0.05）。末次随访时2例ASIA分级C级恢复至D级，其余患者均恢复至E级（ P<0.01）。随访过程中未发生重要血管或神经损伤及内固定失败。发生骨水泥渗漏18例，均未出现相关临床症状。发生新发椎体骨折35例，给予对症治疗后，均恢复良好。 结论:OTLICS评分、ASOTLF分型及CSOTLF分型系统具有较高的可信度，依据该评分及分型系统对不同损伤程度患者采取阶梯化治疗可减轻疼痛、促进脊柱功能恢复并减少并发症，对临床治疗方法选择具有一定指导意义。

abstractsObjective:To test and evaluate the reliability and clinical effectiveness of osteoporotic thoracolumbar fracture (OTLF) scoring and classification system.Methods:A multicenter retrospective case series study was conducted to analyze the clinical data of 530 OTLF patients admitted to 8 hospitals including Honghui Hospital Affiliated to Xi'an Jiaotong University from January 2021 to June 2022. There were 212 males and 318 females, aged 55-90 years [(72.6±10.8)years]. There were 4 patients with grade C and 18 with grade D according to American Spinal Injury Association (ASIA) classification. According to the osteoporotic thoracolumbar injury classification and severity (OTLICS) score, all patients had an OTLICS score over 4 points and required surgical treatment. Among them, 410 patients had acute symptomatic OTLF (ASOTLF), including 24 patients with type I, 159 type IIA, 47 type IIB, 31 type IIC, 136 type IIIA, 8 type IIIB, 2 type IV (absence of neurological symptoms) and 3 type IV (presence of neurological symptoms), and 120 patients had chronic symptomatic OTLF (CSOTLF), including 62 patients with type I, 21 type II, 17 type III, 3 type IV (reducible under general anesthesia), 9 type IV (not reducible under general anesthesia), 1 type V (reducible under general anesthesia), 5 type V (presence of neurological symptoms), and 2 type V (not reducible under general anesthesia). Surgical procedures included percutaneous vertebroplasty (PVP), positional repositioning plus PVP, percutaneous kyphoplasty (PKP), posterior open reduction combined with bone graft fusion and bone cement augmented screw internal fixation, posterior open reduction combined with decompression, bone graft fusion and bone cement augmented screw internal fixation, and posterior open reduction combined with osteotomy and orthopedics, bone graft fusion and bone cement augmented screw internal fixation. A weighted Kappa was used to test the interobserver and intraobserver reliability of the OTLICS score, the ASOTLF classification, and the CSOTLF classification. The visual analog scale (VAS), Oswestry disability index (ODI), ASIA classification were compared before, at 1 month after surgery and at the last follow-up. Incidence of postoperative complications was observed.Results:The percentage of mean interobserver agreement for OTLICS staging was 93.4%, with a mean confidence Kappa value of 0.86, and the percentage of mean intraobserver agreement was 93.0%, with a mean confidence kappa value of 0.86. The percentage of mean interobserver agreement for ASOTLF staging was 94.2%, with a mean confidence Kappa value of 0.84, and the percentage of mean intraobserver agreement was 92.5%, with a mean confidence Kappa value of 0.83. The percentage of mean interobserver agreement for CSOTLF subtyping was 91.9%, with a mean confidence Kappa value of 0.80, and the percentage of mean intraobserver agreement was 91.3%, with a mean confidence Kappa value of 0.81. All the patients were followed up for 6-12 months [(9.0±2.1)months]. The VAS and ODI scores were significantly lower in patients with ASOTLF and CSOTLF classifications at 1 month after surgery and at the last follow-up than those before surgery (all P<0.05). The VAS scores in patients with ASOTLF types IIA, IIB, IIC, IIIA, and IV were significantly lower at the last follow-up than that at 1 month after surgery; the ODI scores in patients with ASOTLF types I, IIA, IIB, IIIA, IIIB and IV were significantly lower at the last follow-up than those at 1 month after surgery. The VAS scores in patients with CSOTLF types II, III, IV, and V were significantly lower at the last follow-up than those at 1 month after surgery, and the ODI scores in patients with all CSOTLF types were significantly lower at the last follow-up than those at 1 month after surgery (all P<0.05). Two patients with ASIA grade C recovered to grade D, and the rest recovered to grade E at the last follow-up ( P<0.01). No major vessel or nerve injury or internal fixation failure was found during follow-up. There were 18 patients with cement leakage, none of whom showed relevant clinical symptoms. There were 35 patients with new vertebral fractures, all of whom recovered well after symptomatic treatment. Conclusions:The OTLICS score, ASOTLF classification and CSOTLF classification have a high degree of reliability. Application of stepwise treatment for patients with different levels of injury according to the scoring and classification system can reduce pain, promote recovery of the spinal function, and reduce complications, which is of some significance in guiding the selection of clinical treatment.