强直性脊柱炎髋关节结构受累严重程度的新评价方法的探索性研究

A new method to evaluate hip joint damage in ankylosing spondylitis

目的探索一种评价AS髋关节结构受累程度的新方法,并对该方法的可靠性、便捷性和对髋关节结构改变的敏感性进行验证。方法共纳入AS患者98例,依据随访分为随访1~2年、3~4年和5年及以上的3组,新方法从关节面的侵蚀、硬化和关节间隙狭窄这3个方面评价髋关节的放射学改变,由2位观察者按照新方法对研究对象在基线期和随访期内拍摄的骨盆正位片进行独立、盲法评分,通过计算组内相关系数评价观察观察者内和观察者间变异度,通过计算阅片平均用时评估其临床便捷性,采用配对t检验或非参数秩和检验分析随访前后各组组内评分的改变,组内均数的比较用方差分析或非参数秩和检验,组间率的比较采用χ^2检验。结果3组患者在基线期人口统计学指标和可能的髋关节受累危险因素方面差异均无统计学意义(P>0.05)。观察者间变异度为0.72,2位观察者的观察者内变异度分别为0.84和0.89,完成1张骨盆正位片的2个髋关节评分平均(33±10)s。仅5年及以上的随访的AS患者,其整体水平上髋关节的评分差异具有统计学意义,2例观察者的基线期、观察终点的评分分别为(6.0±2.7)分、(7.5±3.7)分(t=2.86,P<0.01)和(5.6±2.1)分、(7.1±3.6)分(Z=-2.99,P<0.01)。结论本研究建立了一种新的AS髋关节结构损伤程度的评价方法,这种方法可操作性强,有较好的可靠性,对髋关节放射学改变具有一定的敏感性,值得在更大样本量的研究中进一步验证。

Objective In order to assess the structure damage of hip joint in ankylosing spondylitis(AS), a new radiograph-based scoring method was developed according to the radiological characteristics of hip involvement in AS, as well referring to prior existing scoring indexes. Methods A new scoring method constituted of erosion, sclerosis and joint space narrowing was developed, pelvis anterior-posterior plain films acquired from patients with AS at baseline and follow-up were collected and assessed by two physicians who were trained in image reading by radiologists. All films were scored independently and blindly. Intra- and inter-reader reliability were assessed by intra-class correlation coefficient (ICC), the feasibility of this new scoring method was assessed by the mean time acquired to score a plain(two hips), its ability to detect the change of structure damage was assessed by the comparison of score differences between baseline and different follow-ups. The date were analyzed by paired-t test or nonparametric tests. Analysis of Variance(ANOVA) or nonparametric tests were utilized for the comparison of means of quantitative variables among the three groups, while Chi-square test for rates of categorical variables. Results No statistically significant differences existed in demographic data and suspected risk factors among the three groups at baseline (P>0.05). Intra-observer reliability was good (0.84 and 0.89), as well as the inter-observer reliability (0.72), the mean time needed to score was (33±10) seconds. Score changes were not statistically significant in the groups with follow-up duration of 1-2 and 3-4 years, but in the group of over 5 years, baseline/final scores assessed by the two observers were (6.0±2.7/7.5±3.7) and (5.6±2.1/7.1±3.6), respectively, both changes were statistically significant (t=2.86, Z=-2.99;P<0.01). Conclusion This new method is not only reproducible and easy to operate in clinic practice, but also can tell the changes of hip joint structure damage in the interval of over 5 years, further validation is requeired to demonstrate its discriminability in large populations.