|R. James Brenner, MD, JD
For the past decade, delay in the diagnosis of breast cancer has remained the most common reason that physicians are sued for malpractice; radiologists remain the most commonly named defendants. 1 The outcome of such cases depends on a complex interplay of factors, including the nature of the lesion, the ability of experts to persuade a jury regarding potential compliance with (or breach of) an established standard of care, and the strategic handling of the case by legal counsel. There is no shortage of commentary regarding the dangers inherent in breast imaging. These, coupled with strained third-party reimbursement, result in disincentives to enter the field. 2
This unfortunate scenario is complicated by two issues. First, there remains a lack of sufficient criteria to qualify expert witnesses who impugn the conduct of radiologists in the tort system, which relies on the sometimes illusory safeguard of rigorous crossexamination to ensure the sufficiency of testimony. 3 Even qualified experts (including some who feel that they comprehend legal procedure through prior case involvement) are often, and understandably, incompletely familiar with the quixotic nature of the legal system, adding to the misapplication of the trial process. Second, there may be insufficient training and education in the management of breast-imaging cases that are not particularly complex, but for which evidence- based approaches are ignored. Diagnosis appears to be based, at times, on limited experience, habit, and even irreconcilable intuition, when other rationales are available. Consequently, lesions that do not need biopsy are often considered suspicious, and lesions that warrant biopsy are, instead, subjected to surveillance.
For most purposes in breast imaging, the law calls for a standard of care based on whether a mammographically demonstrated abnormality shows sufficiently malignant features for a prudent radiologist, practicing under similar circumstances, to consider the lesion suspicious for malignancy. The operative standard involves whether the cancer eventually discovered was reasonably foreseeable. The benchmark requires no warranty of certainty. When lesions are appropriately evaluated and lack suspicious features, the need to exclude malignancy is neither medically nor legally grounded. Recent research 4 supports the notion that the failure to identify cancers that lack reasonable imaging criteria to indicate the diagnosis does not constitute legal negligence. During a lawsuit, however, the manner in which such information is presented by the defendant and expert witness may have considerable impact on how the case is resolved. Indeed, in certain states, there are no required pretrial depositions of expert witnesses, so the first opportunity to expose the strengths or weaknesses of the defendant’s position occurs in court.
Attorneys and expert witnesses can only address the images and report for a given case, so these two elements (rather than the subsequently expressed opinions) are of paramount importance in the disposition of an allegation of a breach of the standard of care. Attempting to avoid liability by overcalling in response to the fear of a lawsuit will not necessarily improve performance or sensitivity in recognizing breast cancer, but may decrease the cost-effectiveness of screening. 5
How, then, does one minimize risk without abandoning the field altogether? The answer is illustrated by a recent article 6 documenting adverse outcomes for lesions inappropriately subjected to mammographic surveillance rather than biopsy. Appropriate criteria and longitudinal studies 7 have validated surveillance of lesions that are probably benign as a reasonable, cost-effective alternative to immediate biopsy, with comparable outcomes. Nonetheless, subjecting growing or new solid masses in postmenopausal women to mammographic surveillance rather than biopsy-absent special circumstances such as trauma or infection– defeats the rationale for follow-up mammography, invites the kinds of adverse outcomes that were reported, and places the radiologist making such a recommendation in a difficult- to-defend position.
Retrospective studies and reviews 8 have evaluated the causes of false-negative mammograms. As a general rule, about a third of interval cancers are not seen on prior mammograms, another third have nonspecific or subthreshold features, and the final third are simply not detected or are misinterpreted. All three of these areas may lend themselves to legal redress, and it serves the radiologist to recognize where potential liability lies (and where it does not lie).
It is not infrequently the case that breast lumps either will not be identified or will be insufficiently characterized on mammography to allow meaningful management recommendations to be made. Ultrasound is a common and useful adjunct in such situations. Consider the situation in which the radiologist performs satisfactory mammographic and ultrasound examinations of a focal palpable lesion. The imaging contribution to patient management may be dispositive. If cancer is found, a biopsy is prompted. If a simple cyst is found, either aspiration or conservative management may be pursued. If a benign-appearing solid mass is found, options for surveillance or biopsy are presented.
If no lesion is imaged, however, then management is dependant upon clinical examination alone; the recommendation of clinical evaluation is both appropriate and useful because a small number of palpable cancers escape both mammographic and sonographic detection. Of tantamount importance is the observation that where no lesion is characterized, an image-based recommendation for biopsy is neither appropriate nor legally required. Note that such a recommendation of directed clinical management of a palpable lesion is distinguishable from global disclaimers or universal recommendations for clinical management. In this case, a deliberate attempt to evaluate the abnormality has been made and has not contributed to diagnosis; it is a pertinent negative, albeit inconclusive, study. Should cancer be discovered subsequently, the standard of care has been met (recall that there is no warranty of certainty). On the other hand, failure to address the clinical issue with appropriate images and a report (bearing in mind that not all symptoms require additional images) may subject the radiologist to claims that the standard of care has been breached. Imagers, therefore, need to recognize potential restrictions regarding the imaging component of breast evaluation. Properly chosen positions remain both medical and legally defensible. In this context, it should be noted that mammographic lesions for which there are no clinical findings need to be diagnosed using imaging parameters.
In like manner, an interpretation of mammographic abnormalities should account for clinical findings. Several studies 4,9,10 have shown the low likelihood that focal areas of asymmetric tissue represent forseeable cancer, although they account for a relatively high proportion of retrospectively assessed false-negative mammograms, and perhaps a disproportionate number of cases that come to legal attention when this is the antecedent mammographic finding. The association of a focal asymmetric density with a palpable lesion may, however, increase the likelihood of cancer and prompt an imaging diagnosis of suspicion. 11 This concept reinforces the need for proper direction of the patient to screening or diagnostic studies, where additional imaging may be required to meet the requisite standard of care in a symptomatic patient. More ambiguous, and often controversial, is the proper care of the patient who has either had an inadequate physical examination or for whom vague symptoms such as general thickening have been reported. Facilities need to develop deliberate approaches to such cases, based on the population of physicians and patients they serve.
In the same regard, the use of film markers for areas of focal physical findings and problem-oriented reports reflecting appropriate imaging evaluation often serves as notice that an area of clinical concern has been reasonably evaluated. Should a similar complaint arise the next year, and be documented in the same area, extensive repeat imaging may not necessarily be required. This kind of deliberate approach with documentation has been validated in precedentsetting appellate decisions.
Although attention to optimal radiographic technique is a guideline for all facilities, and the ability to produce reasonable images is a requirement for Mammography Quality Standards Act (MQSA) certification, any given case incurs possible legal redress when images are not satisfactory. Unsatisfactory image quality may occur because of a multitude of factors, especially when images are blurred, poorly exposed or compressed, or obscure adequate tissue from view. All of these circumstances can be remedied using repeat films; the radiologist who interprets a study is considered the supervising physician and is responsible for determining whether such an effort needs to be made. Such matters lend themselves to judgment based on experience and expertise because not all mammographic images in any practice are optimal. Not infrequently, magnification images are obtained to provide additional information for a given anatomic problem, using a technique that often requires longer exposure times and special patient-positioning strategies. If decisions are made (in screening or problem solving) based on images of poor quality, then both detection and diagnostic goals will be compromised. Such films are presented at a trial as evidence, coupled with subsequent demonstration of more suspicious mammographic findings, to impugn a defendant who might otherwise have avoided this situation.
The actual images obtained by a radiologist provide a basis for imaging evaluation. These images-subject to MQSA provisions- belong to the facility, not the patient. The formal medical record of the evaluation is the interpretative report. Reports that do not integrate the reportable information in a logical manner and report it as such invite legal redress. There are several circumstances in which this may occur. While it is understandable that additional studies (such as MRI) or prior comparison studies will provide, on occasion, a stronger basis for management recommendations, federal law (MQSA) establishes a statutory standard of care stating that a definite management recommendation must be issued, using the American College of Radiology Breast Imaging Reporting and Data System (BIRADS ? ) terminology, following a diagnostic mammographic examination. Employing an indeterminate reading (such as BI-RADS 0) in such a report is a breach of the standard of care, and it requires no expert testimony to substantiate this position. If such ambiguity causes a substantial delay, for any reason, in the diagnosis of cancer, it becomes difficult to defend the report in a negligence case. Given the time limits involved in reporting, there may be some latitude available for access to additional information. Attempts to retrieve prior studies are often unsuccessful, however, so a deliberate plan should be in place to account for such exigencies. Another approach, in an equivocal case, is to categorize an abnormality reasonably according to the BI-RADS format and note in the report that the conclusion may be subsequently amended as more information is obtained.
While often considered routine, the intraoperative evaluation of specimen radiographs presents a situation in which unanticipated legal exposure occurs. Sometimes, especially in practices with multiple practice sites and manpower schedules, the radiologist who performs a preoperative localization procedure before surgical excision does not review the specimen radiograph. In such situations, which can usually be anticipated, the preoperative images must be available for comparison during assessment of the specimen radiograph. For example, the recovery of calcifications usually implies that the targeted lesion has been removed, but discrepancies between incidental and targeted calcifications can occur; recognition of this is facilitated by having preoperative images for direct comparison. In addition, some cancers that manifest themselves mammographically as low-density masses or architectural distortions are better seen using ultrasound. Consequently, preoperative localization using sonographic guidance is employed. Radiographic specimens reviewed under such circumstances may not definitively demonstrate the lesion (specimen ultrasonography has met with limited success, and indirect specimen compression may be of use on occasion). In an effort to reassure the surgeon that the lesion has been recovered, radiologists have sometimes extended the limits of their ability to define the target lesion, creating legal problems in a predictably finite number of cases of failed surgery. Reasonable interpretation, noting the parameters regarding the interpretation, and direct communication with the surgeon should provide a basis for avoiding this predicament.
As in any aspect of radiology, reasonable conduct is the benchmark by which both medical and legal accountability are measured. Public expectations, difficulty in performance, and the natural history of breast cancer present special challenges to the breast imager. By appreciating the legal atmosphere in which such conduct is judged, and by recognizing common themes for mismanagement and legal redress, many of the perceived perils may be avoided with relative ease.
R. James Brenner, MD, JD, is director, Breast Imaging Services, Eisenberg Keefer Breast Center, John Wayne Cancer Institute, St Johns Health Center, Tower-St Johns Imaging, Santa Monica, Calif; and clinical professor of radiology, Geffen School of Medicine at the Univeristy of California, Los Angeles.
- Physicians Insurers Association of America. Breast Cancer Report. Rockville, MD: PIAA; 2003.
- Bassett LW, Monsees BS, Smith RA, et al. Survey of radiology residents: breast imaging training and attitudes. Radiology. 2003;227:862-869.
- Daubert v Merrel Dow Pharmaceuticals, 509 US 579,1993.
- Ikeda DM, Birdwell RL, O’Shaughnessy KF, Brenner RJ, Sickles EA. Analysis of 172 subtle findings on prior “negative” mammograms in women with screening detected cancer. Radiology. 2003;226:494-503.
- Yankaskas BC, Cleveland RJ, Schnell MJ, et al. Association of recall rates with sensitivity and positive predictive values of screening mammography. AJR Am J Roentgenol. 2001;177:533-540.
- Rosen EL, Baker JA, Soo MS. Malignant lesions initially subjected to short-term mammographic follow-up. Radiology. 2002;223:221-228.
- Brenner RJ, Sickles EA. Acceptability of periodic follow up as an alternative to biopsy for mammographically detected lesions interpreted as “probably benign.” Radiology. 1989;171:645-646.
- Brenner RJ. False negative mammograms: medical, legal, and risk management implications. Radiol Clin North Am. 2000;38:741-757.
- Harvey JA, Fajardo LL, Innis CA. Previous mammograms in patients with impalpable breast carcinoma: retrospective vs blinded interpretation. AJR Am J Roentgenol. 1993;161:1167-1172.
- Sickles EA. Periodic mammographic follow-up of probably benign lesions: results in 3,184 consecutive cases. Radiology. 1991;179:463-468.
- Moskowitz M. The predictive value of certain mammographic signs in screening for breast cancer. Cancer. 1983;51:1007-1011