CAJ | 학술논문

目的系统评价磁共振成像(MRI)对于类风湿关节炎(RA)的诊断效能.方法计算机检索国内外各主要数据库,辅以手工检索和其他检索方法,收集关于MRI诊断RA的对照试验,检索时间 1996-2009年.按照纳入标准筛选文献并提取资料,采用Meta-disc软件合并敏感性,特异性,阳性似然比,阴性似然比,诊断性试验比值比,进行异质性分析(Q-检验,I2),当P>0.10,且I2≤50%时无异质性.结果共纳入12项随机对照试验.结果显示病程≤2年的RA患者,滑膜炎、骨侵蚀和腱鞘炎诊断的敏感度分别为81%～100%、48%、100%~67%~96%,特异度分别为64%～89%、16%~100%~21%~74%.病程>2年的RA患者,滑膜炎、骨侵蚀诊断的敏感度分别为91%、84%,特异度分别为70%、81%.结论 MRI显示滑膜炎对于早期和中晚期RA有诊断意义,骨侵蚀和腱鞘炎对于早期RA的诊断意义有待进一步研究.纳入研究样本量小,对照组病例缺乏代表性,建议设计严谨且更大样本病例的临床研究.
목적 계통평개자공진성상(MRI)대우류풍습관절염(RA)적진단효능.방법 계산궤검색국내외각주요수거고,보이수공검색화기타검색방법,수집관우MRI진단RA적대조시험,검색시간 1996-2009년.안조납입표준사선문헌병제취자료,채용Meta-disc연건합병민감성,특이성,양성사연비,음성사연비,진단성시험비치비,진행이질성분석(Q-검험,I2),당P>0.10,차I2≤50%시무이질성.결과 공납입12항수궤대조시험.결과 현시병정≤2년적RA환자,활막염、골침식화건초염진단적민감도분별위81%～100%、48%、100%~67%~96%,특이도분별위64%～89%、16%~100%~21%~74%.병정>2년적RA환자,활막염、골침식진단적민감도분별위91%、84%,특이도분별위70%、81%.결론 MRI현시활막염대우조기화중만기RA유진단의의,골침식화건초염대우조기RA적진단의의유대진일보연구.납입연구양본량소,대조조병례결핍대표성,건의설계엄근차경대양본병례적림상연구.
objective To evaluate the diagnostic efficiency of magnetic resonance imaging(MRI)for rheumatoid arthritis(RA).Methods The major international databases was searched by computer and Other methods to collect control studies about MRI for the diagnosis of RA,the searching deadline was December 2009.Data were screened and extracted by inclusion criteria.Meta-disc software Was used for statistics,including sensitivity,specificity,positive likelihood ratio,negative likelihood ratio,diagnostic odds ratio(OR)and heterogeneity analysis.Results Twelve random control tests were included.Meta-analysis indicated that sensitivities of vities of synovitis,bone-erosion and tendenosynovitis were 81%~100%,48%～100%and 67%～96%while specificities of them were 64%～89%,16%~100%and 21%-74%respectively for RA patients(≤2vears).Sensitivities of synovitis and bone-erosion were 91%and 84%while specificities of them were 70%and 81%for RA Datients (>2 years).Conclusion Synovitis shown by MRI is helpful for the diagnosis of earlv and med-to-late RA.Diagnostic values of bone-erosion and tendonosynovitis demonstrated by MRI for earlv RA were not clear.The sample size of the included studies is small and some studies lack of control groupsput the conclusion of this meta-analysis liable for bias.We suggest that better designed and larger sample clinical studies are necessary.