cancerous lesion in 71 % [36]. Much published scientific data confirm this high miss rate [33, 37]. The literature on mammographic errors and their impact on malpractice is neither as lengthy non as long-standing as that on chest misdiagnoses, but the findings are similar. A 1990 Physician Insurers Association of America study on breast cancer litigation revealed that mammography was an issue in 28% of cases [38].This study was published in the same year that HaIl [39] predicted that missed cancer on mammograms will replace missed cancer on chest nadiognaphs as the most com- mon cause of radiology litigation. Subsequent reports have borne out his prediction [40, 41]. Harvey et al. [42] found that retrospective review of mammograms disclosed evidence of carcinoma in up to 75% of the cases originally interpreted as normal. Other studies have found similar rates of disagreement among radiologic mammographic interpretations [43, 44]. Numerous publications document similarly high error and disagreement rates for a host of other techniques used in radiologic practice, including sonography in the diagnosis of ectopic pregnancy [45], angiography in determining degree of stenosis [46], MR angiography in assessing extent of dis- ease [47], thallium radionuclide heart scanning in ischemic changes [48], MR imaging of rotator cuff injury [49], and MR imaging of the extent of prostatic cancer [50]. In a study involving radiologists’interpretation of plain abdominal radiographs, Markus et al. [51] found that the agreement rate for certain findings was quite poor-as low as 1 0%. The barium enema miss rate in the diagnosis of colon carcinoma has been found to be as high as 25% [52]. Although these radiologic error and disagreement rates may seem so high as to cast doubt on the scientific validity of nadiology as a medical discipline, readers can achieve some comfort in being assured that comparative studies, albeit fewer, of other medical but nonradiologic fields disclose a similar prevalence of inaccuracy. Agreement among academic faculty physicians performing physical diagnosis of spleen enlargement [53], liver enlargement [54], abdominal ascites [55], and many other assorted physicalfindings [56] is quite poor, with the lowest rate of agreement being 11% among clinicians’ determinations of whispered pectoriloquy [57]. Large autopsy studies have uncovered frequent clinical errors and misdiagnoses [58]. Anderson et al. [59] found a missed-diagnosis rate as high as 47% for myocardial infarction in a medical school-affiliated hospital, concluding that errors are inevitable and the concept of necessary fallibility must be accepted. Wu et al. [60] reported that 90% of all house staff members admit to making mistakes. Recognition that error rates in radiology have not changed significantly in more than four decades should not lull radiologists into apathy about efforts to reduce the rates. Measures to improve accuracy have been discussed [61-63], but several should be emphasized here. Although Peterson, in a classic monograph [64], stated that radiologists should make every effort to interpret radiographs without the benefit of extraneous clinical information, most data suggest the contrary. More than 30 years ago, Schreiber [65] found that clinical history improves radiographic accuracy, and Doubilet and Herman [66] demonstrated that knowledge of pertinent clinical history increased the accuracy of chest radiographic interpretations from 16% to 72% for residents and from 38% to 84% for staff radiologists. In a series of articles, the most recent in 1995 [67], Berbaum and his group showed significant improvement in radiologists’ perceptual abilities and accuracy rates when patients’ histories, perunent clinical findings, and previous nadiologic studies and their reports were available at the time of interpretation. Referring to future technological developments, Greene [68] suggested that digital radiography may improve accuracy. Turner [69] wondered whether improved trainingcould further minimize mistakes but concluded that they are unlikelyto be significantlyreduced until we have “perfectdiagnostic tests”and “perfectobservers.” Downloaded from www.ajronline.org by 8. Copyright ARRS. For personal use only; all rights reserved AJR:165, October 1995 MALPRACTICE AND RADIOLOGISTS 787 Emerging Technologies A distinctive characteristic of radiology is the constant introduction of new technology and equipment. Techniques such as CT and MR imaging were developed in research laboratories, refined in schools, then rapidly diffused for general use throughout the radiologic community. Eventually, these technologies also find their way into the courts. The lag time between the introduction of a technology and the date the first lawsuit alleging its negligent use is filed in Cook County is changing. The first malpractice lawsuit specifically including sonography among its allegations was filed in 1982, more than a decade following the introduction of sonography into medical practice in Chicago. CT was first introduced into the Chicago area in 1974, but not until 1982 did it first become the subject of a malpractice suit. MR imaging was first installed in Cook