| PurposeTo perform a meta-analysis to compare current noninvasive imaging methods ( ultrasonography [ US ] , computed tomography [ CT ] , magnetic resonance cholangiopancreatography [ MRCP ] ) in the detection of obstructive jaundice disease. Have a pilot study in evidence-Based Medicine of radiology. To evaluate the value of diagnostic ultrasound, computed tomography and MRCP in patients with cholangiopancreatic disease by Receiver Operating Characteristic curves (ROC) , and introduce the principle of ROC simply.Material and Method1. Evidence-Based Medical StudyA MEDLJNE literature search was performed of the English-language medical literature that were published between January 1980 and July 2003. Additional references were obtained from bibliographies of original articles, and experts in the field were consulted.Articles were included if ( a) US, CT and/or MRCP were performed for e-valuation of Obstructive jaundice disease ( b) pathological results or operating findings were the reference standard and (c) absolute numbers of true-positive, false-negative, true-negative and false-positive results were available or derivable.Articles were excluded if there was no blinded interpretation of the test result ; no standard of reference was used; data were presented as a case reports or a review; data were incomplete; no original data were reported, the absolute numbers of TP, FN, TN, and FP results either were not available or could notbe derived; authors reported on the same patients in more than one article.Extracted data included variables related to study design, patient characteristics, diagnostic imaging protocol,and absolute numbers of TP, FN, TN,and FP results. The natural logarithm of the diagnostic odds ratio (D) was calculated as foUows: D=ln[(TPxTN)/(FPxFN)]. This value represents a summary measure of the diagnostic performance per study (ie, the discriminatory power of the examination). This was the measure of interest in the summary ROC analysis.A quality score was defined on the basis of the following criteria: adequate description of the study group ( selection criteria were defined, and patient characteristics were presented) ; adequate description of diagnostic criteria; congruence of presented data. Studies fulfilling one criteria were given a quality score of 1, and a there-tiered rating of quality for diagnostic imaging studies would be gained.Two authors independently abstracted data sets and excluded data. Discrepancies in data extraction between the two authors were noted and resolved at consensus. The k value was calculated as a measure of agreement between extracted variables in the analysis.Funnel plot and Summary receiver operating characteristic [ SROC ] were obtained and a covariate analysis was used to evaluate the influence of patient or study-related factors on sensitivity.2. Clinical trialWe collected 105 patients who were suspected with cholangiopancreatic disease. Of all patients, 49 were male and 56 were female; their ages ranged from 1 years to 83 years. All patients were studied with MR(Elscint Prestige 2. OT MR and Philips 1. 5T MR). Of them, 65 patients underwent US examination ( Medison-Kret Voluson 730D US ). Additional computed tomography ( Marconi MX8000 CT) for 59 patients, Of them, Ultrasound was performed in 31 patients. The cause of pancreato-biliary abnormality was evaluated using a five-point scale to assign a confidence level; 1 definitely benign, 2 probably benign ,3indeterminate,4probably malignant, and 5definitely malignant. Pathological results, operating findings or clinical and laboratprial examination re-suits were the reference standard. Images were interpreted with being blinded to the reference standard. Taking1, (1) + 2, (1) + (2) + 3,(1) +2 +3 +4 as positive diagnosis, true positive rate and false positive rate were calculated respectively. We analysed the data with ROC function of soft ware SPSS11.0. According to location and cause of the obstruction , sensitivity, specificity and likelihood ratio of MRCP...
|
PDF Full Text Request