重置DAS28CRP临界值对类风湿关节炎患者疾病活动度评价的影响

Effect of Resetting the DAS28CRP Thresholds on the Assessment of Disease Activity in Patients with Rheumatoid Arthritis

目的评价用C反应蛋白计算的28关节疾病活动指数(disease activity score 28 C-reactive protein,DAS28CRP)、用红细胞沉降率计算的28关节疾病活动指数(disease activity score 28 erythrocyte sedimentation rate,DAS28ESR)、简化疾病活动指数(simplified disease activity index,SDAI)、临床疾病活动指数(clinical disease activity index,CDAI)、健康评估问卷(health assessment questionnaire,HAQ)及健康调查简表(the MOS item short from health survey,SF-36)对类风湿关节炎疾病活动度评估价值。方法连续纳入131例类风湿关节炎患者,计算DAS28CRP、DAS28ESR、SDAI、CDAI、HAQ及SF-36指数,评估患者疾病活动度及生活质量。用Pearson相关分析上述任意两种方法及DAS28CRP与DAS28ESR独有成分[0.36×ln(CRP+1)+0.96]和[0.70×ln(ESR)]间的相关性,并用Bland altam plot分析DAS28CRP与DAS28ESR的一致性,进一步用Kappa值描述DAS28CRP、DAS28CRP1分别与DAS28ESR及SDAI内部的一致性。结果四种疾病活动度评估方法的结果相关性较好,其与HAQ呈负相关(均r<0.00,P<0.001),与SF-36呈正相关(均r>0.00,P<0.001)。DAS28CRP与DAS28ESR独有成分(r=0.618,P<0.001)相关性较其本身(r=0.948,P<0.001)相关性低,在Bland altam plot分析中,虽然有97.7%的点落在差值95%的一致性范围内,但其错误百分率33%超过临床上可接受范围30%。而采用Fleiscmann等提出的DAS28CRP临界值与DAS28ESR及SDAI进行一致性分析时,其Kappa值分别从0.543和0.346提高到0.602和0.448,一致率分别从68.7%和54.96%提高到86.26%和81.70%。结论临床病情评估时,可以采用不同方法对患者进行实时评估,通过重置DAS28CRP评估临界值,可以提高其与DAS28ESR间的一致性,以便进一步准确评估患者病情。

Objective To testify the correlation of DAS28-CRP, DAS28-ESR, SDAI, CDAI, HAQ and SF-36 as well as the unique components of DAS28-CRP and DAS28-ESR. Furthermore, to explore the consistency between the DAS28-CRP with thresholds validated for DAS28-ESR and the new proposed and DAS28-ESR, SDAI. Methods 131 rheumatoid arthritis （RA） patients were included and the DAS28- CRP, DAS28-ESR, SDAI, CDAI, HAQ and SF-36 were calculated. Statistical analysis was carried out by Pearson＇s correlation, Bland altam plot and kappa coefficient. Results Four method for disease activity assessment were strongly correlated with each other, which was negatively correlated with HAQ （r〈0. 00, P〈0. 001 ） and positively correlated with SF-36 （r〉0. 00, P〈0. 001）. The unique components of DAS28- CRP and DAS28-ESR （r= 0. 618, P〈0. 001） had a relatively lower correlation than themselves （r= 0. 948, P〈0. 001 ）. In Bland altam plot analysis, although 97.7% of point data fell between the upper and lower bounds of the limits of agreement, the percentage error （33%） was higher than the acceptable proportion of 30%. The kappa of agreement between DAS28-CRP with newly proposed thresholds and DAS28-ESR （Kappa= 0. 602） , SDAI （Kappa=0. 448） was higher than DAS28-CRP with validated thresholds for DAS28-ESR （Kappa= 0.543） and between them （Kappa= 0. 346）. The proportion of agreement also increased from 68.7% to 86.26%, and from 54.96% to 81.70% respec- tively. Conclusion We can use different method to make a real-time assessment of the disease activity level of a patient. Using the DAS28-CRP with threshold values validated for DAS28-ESR may lead to er- rors in determining of disease activity and therefore may lead to errors in the management of patients with rheumatoid arthritis.