摘要
目的:通过meta、ROC分析以及按病变部位、性质进行的亚组分析分析对目前诊断梗阻性黄疸的非侵入性影像诊断方法(US,CT和MRCP)进行对比研究。方法:1.采用medline检索。纳入标准为:(a)US、CT和MRCP诊断梗阻性黄疸性疾病的文献(b)病理检查、术中所见或临床、实验室检查结果作为诊断金标准。(c)能够直接或间接获得每个影像方法的真、假阳性数,真、假阴性数。提取数据、文献质量评估通过kappa分析进行一致性检验。统计分析采用漏斗图、SROC分析方法以及协变量分析。2.疑胆胰系疾患接受MRCP检查患者105例,其中同时做US检查者65例。另有同期CT资料59例,其中同时做US检查者31例。盲法与金标准对比,计算出各诊断方法的真阳性率和假阳性率,ROC分析其诊断效能。同时按病变部位、性质分别计算MR-CP、US及CT的敏感度、特异度和似然比等指标进行比较分析。结果:1.漏斗图US相关文献分布形状略不规则,CT、MRCP相关文献分布形状类似漏斗形。SROC曲线图MRCP线最靠近左上角,诊断效能高于US和CT.MRCP的Q*值(0.9256)高于US(0.8765)和CT(0.8606)。三者间经检验无显著性差别,MRCP和CT间检验Z=0.33,双侧P>0.25。协变量分析未见对诊断效能有显著性影响因素。2.ROC分析显示,MRCP的曲线最靠近左上角,US次之,CT在最下面,三者的曲线下面积(AZ)分别为0.985,0.981,0.901,均大于0.9,MRCP与CT间离均差(Z)为0.75,双侧P>0.25。MRCP、US和CT诊断胆胰系恶性占位、结石的敏感度分别为100%、83%、82%;92%、71%、76%。经检验,MRCP与US和CT间有显著性差异,p<0.05。结论:经meta、ROC分析,认为MRCP在诊断梗阻性黄疸疾病中具有优势,诊断效能高于US和CT。
Objective： To perform a meta- analysis and Receiver Operating Characteristic curves（ROC）to compare current noninvasive imaging methods, such as ultrasonography （US）, computed tomography （CT）, magnetic resonance cholangiopancreatography （MR（P） in the deteetion of obstructive jaundice disease. Mathods： 1. A MEDLINE literature search was performed. Articles were included ff （a） US, CT and/or MR（P were performed for evaluation of Obstructive jaundice disease （b） Pathological results, operating findings or clinical and laboratorial examination results were the reference standards and （c） absolute numbers of true - positive, false - negative, true - negative and false - positive resuits were available or derivable. The k value was calculated as a measure of agreement between extracted variables and quality score. Funnel plot and Summary receiver operating characteristic （SROC） were obtained and a covariate analysis was used to evaluate the influence of patient or studyrelated factors on sensitivity. 2. We collected 105 patients suspected with cholangiopancreatic disease. All patients were detected with MR. Of them, 65 patients underwent US examination. Additional computed tomography for 59 patients, of them, Ultrasound was performed in 31 patients. True positive rate and false positive rate were calculated, respectively. We analysed the data with ROC. According to location and cause of the obstruction, sensitivity, specificity and likelihood ratio were calculated, respectively. Results： 1. The funnel plot demonstrated a symmetric funnel - shaped distribution for the CT and MR（P studies. SROC analysis demonstrated better discriminatory power for MRCP than for US and CT. The value of Q ＊ for MR（P（0.9256）was higher than that for US（0.8765）and CT（0. 8606）, but there was no significant difference between MRCP and CT （Z = 0.33, P 〉 0.05）. Covariate factors had no influence on sensitivity by covariate analysis. 2. ROC curve showed MRCP curve lied in top left comer and demonstrated better discriminatory power for MRCP than for US and CT. Areas under the ROC curve（AZ） of Ultrasound, CT and MRCP were 0.981, 0.901 and 0.985, all AZ values of them, higher than 0.9. But there was no significant difference between MRCP and CT （Z= 0.75, P〉 0.05）. The sensitivity of MRCP in distinguishing the various pancreato- biliary tumors and stones was significantly higher than that of US and CT （p 〈 0.05）. Conclusion： MRCP is considered to be superior to US and CT for the diagnosis of obstructive jaundice disease.
出处
《现代生物医学进展》
CAS
2006年第8期31-34,共4页
Progress in Modern Biomedicine
