Mygind T, Dorph S, Blichert-Toft M, Lau Pedersen M, Siemssen O, Krag Jacobsen G
Acta Radiol Diagn (Stockh). 1984;25(3):189-93. doi: 10.1177/028418518402500306.
Six hundred mammograms from 307 consecutive symptomatic patients with a firm retrospective diagnosis were evaluated blindly and independently by two observers. Each observer identified 43 of 47 tumors (91 per cent) with predictive values of 0.63 and 0.68, respectively, for the positive statement, and 0.98 for the negative statement. The overall interobserver agreement with respect to malignancy was 94 per cent (Kappa 0.84) and concerning the diagnosis of simple cysts 91 per cent (Kappa 0.60). The frequency of subclinical cancer demonstrated by mammography in this material was higher than the prevalence of mammary carcinoma found by screening of risk group populations. Thus, mammography remains an objective and reproducible diagnostic method with distinct complementary value to clinical examination. By utilizing certain mammographic criteria for benign lesions it may be possible to reduce the number of biopsies presently performed on the basis of unspecific palpatory findings.
两名观察者对307例有明确回顾性诊断的连续有症状患者的600份乳房X线照片进行了盲法独立评估。每位观察者识别出47个肿瘤中的43个（91%），对阳性判断的预测值分别为0.63和0.68，对阴性判断的预测值为0.98。观察者间关于恶性肿瘤的总体一致性为94%（Kappa值0.84），关于单纯囊肿诊断的一致性为91%（Kappa值0.60）。该材料中乳房X线摄影显示的亚临床癌症发生率高于通过筛查高危人群发现的乳腺癌患病率。因此，乳房X线摄影仍然是一种客观且可重复的诊断方法，对临床检查具有明显的补充价值。通过利用某些乳房X线摄影的良性病变标准，有可能减少目前基于非特异性触诊结果进行的活检数量。
