运动疼痛诱发试验对早期膝关节骨关节炎的诊断价值

The diagnostic value of motion pain induction test for early knee osteoarthritis

目的探讨运动疼痛诱发试验对早期膝关节骨关节炎(knee osteoarthritis,KOA)的诊断价值。方法采用横断面研究设计,数据来源于2018年杭州社区KOA健康管理项目,共纳入1816例符合标准的病例。根据KOA患病情况及程度,分为未患KOA(正常人群组530例),患KOA且Kellgren-Lawrence(KL)分期≤Ⅱ(早期组534例),患KOA且KL>Ⅱ(中晚期组752例)。将启动、蹲起、上下楼梯和做家务4项运动纳入诱发试验,统计指标包括年龄、性别、视觉模拟疼痛评分(visual analogue scale,VAS)。先进行相关性分析,根据结果绘制接受者操作特性(receiver operating characteristic,ROC)曲线,比较曲线下面积(area under the curve,AUC)值,并给出相应最佳临界值。通过倾向性评分匹配(propensity score matching,PSM)校正年龄、性别两项混杂因素,校正后平衡性检验P>0.05。通过Kappa分析评价运动疼痛诱发试验和X线片诊断早期膝骨关节炎的一致性。结果正常人群组、早期组、中晚期组年龄分别为(67.39±7.43)岁、(67.41±9.52)岁、(71.55±9.87)岁,三组性别的分布为男238例,女292例;男209例,女325例;男357例,女395例。正常人群组与早期组在年龄和性别分布上没有异质性(P>0.05),早期组与中晚期组在年龄和性别分布上有异质性(t=-0.034,P<0.05;χ2=8.80,P<0.05)。正常组、早期组、中晚期组启动痛VAS评分为0.16±0.37、2.70±1.69、3.68±2.10,蹲起痛VAS评分为0.42±0.49、2.88±1.44、4.01±2.08,上下楼梯痛VAS评分为0.47±0.50、2.87±1.38、3.82±1.98,做家务痛VAS评分为0.14±0.35、2.15±1.40、3.45±2.09,最严重疼痛VAS评分为0.51±0.50、3.59±1.48、4.68±2.01。正常人群组和早期组在上述疼痛中的差异有统计学意义(t=-33.81;t=-37.25;t=-37.66;t=-32.07;t=-45.41;P<0.05);在校正年龄和性别前,早期组中晚期组在上述疼痛中的差异有统计学意义(t=-8.93;t=-10.84;t=-9.56;t=-12.52;t=-10.64;P<0.05),校正年龄、性别后两者间启动疼痛的差异无统计学意义(P>0.05)。正常人群组和早期组ROC曲线结果显示最严重疼痛的AUC值最高,为0.98,最佳临界值为1分。校正混杂因素后早期组和中晚期组的ROC曲线结果显示最严重疼痛的AUC值最高,为0.72,最佳临界值为4分;由于早期组和中晚期组之间ROC曲线的AUC值偏低,使用获得的运动疼痛诱发试验诊断方法与X线片诊断方法对早期和中晚期患者进行一致性检验,Kappa值0.41(P<0.05)。结论运动疼痛诱发试验中最严重的疼痛评分>1分并且≤4分可初步诊断早期KOA,对鉴别正常人群和早期KOA患者的价值高并且近似于X线片检查;对区分早期和中晚期KOA患者的价值中等,与KL分期的一致性中等,在有条件的情况下还需完善影像学检查综合判断。

Objective To explore the diagnostic value of motion pain induction test for early knee osteoarthritis.Methods A cross-sectional study was conducted and the data came from The Project of Health Management of Knee osteoarthritis in Community in Hangzhou in 2018, and a total number of 1 816 people were included which were divided into normal group (n=530), early group (n=534) and middle-late group (n=752) by not sick, sick while Kellgren-Lawrence (KL) ≤Ⅱ and sick while KL>Ⅱ starting, squatting, walking up and down stairs and doing housework were included in the test, and the statistical indicators included age, gender and pain scores (visual analogue scale, VAS). Receiver operating characteristic (ROC) curves were mapped after the correlation analysis to obtain the cut-off points and compare their values of area under the curve (AUC). The confounders which included age and gender were corrected by propensity score matching (PSM) and the balance test is consistent with P>0.05 after the PSM. The Kappa analysis was used to verify the consistency of two diagnostic methods.Results The age of normal, early and medial-late groupwas 67.39±7.43, 67.41±9.52, 71.55±9.87. And the gender distribution of three groups was (238 male, 292 female), (209 male, 325 female), (357 male, 395 female). There was no heterogeneitybetween the normal group and early group in distribution (P>0.05) while there was heterogeneity between the early and medial-late group (t=-0.034, P<0.05;χ2=8.80, P<0.05). The VAS scoresof starting pain in three groups was 0.16±0.37,2.70±1.69, 3.68±2.10. The VAS scoresof squatting pain was 0.42±0.49, 2.88±1.44, 4.01±2.08. The VAS scoresof up and down stair pain was 0.47±0.50, 2.87±1.38, 3.82±1.98. The VAS scoresof housework pain was 0.14±0.35, 2.15±1.40, 3.45±2.09. The VAS scoresofmaximum pain was 0.51±0.50, 3.59±1.48, 4.68±2.01. And there was significant difference between normal and early groupin all kinds of pain (t=-33.81;t=-37.25;t=-37.66;t=-32.07;t=-45.41;P<0.05). The difference between early and medial-late group in all type of pain was significant (t=-8.93;t=-10.84;t=-9.56;t=-12.52;t=-10.64;P<0.05). The results were similar after adjusting for confounders except for the pain of starting (P>0.05). The results of ROC curve between normal and early group showed the maximum pain's AUC point was 0.98 and larger than others, and its cut-off point was 1. After adjusted, the results of ROC curve between early and medial-late group showed the maximum pain's AUC point was 0.72 which was larger than others and cut-off point was 4. For the AUC of ROC curve between early and medial-late groupwas lower, Kappa-test was used, and the Kappa point of two diagnostic methods was 0.41 (P<0.05).Conclusion The maximum of pain score in pain dimension>1 and ≤ 4 could be preliminarily diagnosed as early KOA. It had high value in separating early KOA from normal people and approximately similar to X-ray, and the value of identifying early and mid-late KOAwas moderateas well as the moderate consistency with KL stage. Comprehensive judgment of imaging examination should be improved when conditions are available.