Kwong J S, Adler B D, Padley S P, Müller N L
Department of Radiology, University of British Columbia, Vancouver, Canada.
AJR Am J Roentgenol. 1993 Sep;161(3):519-22. doi: 10.2214/ajr.161.3.8352096.
The purpose of this study was to compare the sensitivities and diagnostic accuracies of chest radiography and CT in the assessment of diseases of the trachea and main bronchi in regard to detection, evaluation of focal vs diffuse disease, separation of benign from malignant disease, and accuracy of diagnosis.
The chest radiographs and CT scans of 35 patients with proved diseases of the trachea and main bronchi and five control subjects were randomly assessed by two independent observers in a blind study. The diagnoses included 25 malignant and 10 benign conditions. The malignant lesions included 24 primary tumors and one metastatic melanoma. The benign lesions included two benign tumors, two tracheal stenoses, one focal amyloidosis, one tracheomalacia, and four miscellaneous diffuse abnormalities. The observers were asked to determine if an abnormality was present, to classify it as focal or diffuse and as benign or malignant, and to list the three most likely diagnoses. Sensitivities were compared by using the McNemar test.
The sensitivity of both observers in detecting disease on the chest radiograph was 66%. The sensitivity on the CT scan was 97% for the first observer and 91% for the second observer (p < .01). Both observers were able to correctly classify the detected abnormalities as either focal or diffuse in 91% of cases on the chest radiographs and 97% of cases on the CT scans. The abnormalities were correctly classified as either benign or malignant in 78% of cases by both observers on the chest radiographs and in 85% and 78% of cases on the CT scans by the first observer and the second observer, respectively. The correct diagnosis was included in the list of the three most likely diagnoses in 61% of cases by both observers after reviewing the radiographs and in 56% of cases by one observer and 63% of cases by the other observer after reviewing the CT scans.
CT is superior to chest radiography in allowing detection of abnormalities of the major airways. Both CT and chest radiography are accurate for differentiating focal from diffuse disease, but neither technique is accurate for distinguishing benign from malignant disease or for establishing a specific diagnosis.
本研究旨在比较胸部X线摄影和CT在评估气管及主支气管疾病时,在疾病检测、局灶性与弥漫性疾病评估、良性与恶性疾病鉴别以及诊断准确性方面的敏感性和诊断准确性。
在一项盲法研究中,两名独立观察者对35例经证实患有气管及主支气管疾病的患者和5名对照者的胸部X线片和CT扫描进行随机评估。诊断包括25例恶性疾病和10例良性疾病。恶性病变包括24例原发性肿瘤和1例转移性黑色素瘤。良性病变包括2例良性肿瘤、2例气管狭窄、1例局灶性淀粉样变性、1例气管软化症和4例其他弥漫性异常。要求观察者确定是否存在异常,将其分类为局灶性或弥漫性、良性或恶性,并列出三种最可能的诊断。采用McNemar检验比较敏感性。
两名观察者在胸部X线片上检测疾病的敏感性均为66%。第一位观察者在CT扫描上的敏感性为97%,第二位观察者为91%(p <.01)。两名观察者均能在91%的胸部X线片病例和97%的CT扫描病例中将检测到的异常正确分类为局灶性或弥漫性。在胸部X线片上,两名观察者在78%的病例中将异常正确分类为良性或恶性;在CT扫描上,第一位观察者和第二位观察者分别在85%和78%的病例中将异常正确分类为良性或恶性。在查看X线片后,两名观察者在61%的病例中将正确诊断列入三种最可能的诊断列表中;在查看CT扫描后,一名观察者在56%的病例中将正确诊断列入三种最可能的诊断列表中,另一名观察者在63%的病例中将正确诊断列入三种最可能的诊断列表中。
在检测主要气道异常方面,CT优于胸部X线摄影。CT和胸部X线摄影在区分局灶性与弥漫性疾病方面都很准确,但两种技术在区分良性与恶性疾病或确立特定诊断方面都不准确。