An article for physicians by physicians
Originally published April 2009
Clear communication among patients, referring physicians, technicians and radiologists, as well as attention to details in follow-up, will permit optimal outcomes from screening programs for breast cancer.
The CMPA has conducted a review of the legal actions and regulatory authority (College) complaint cases involving breast cancer that closed between 2003 and 2007. The results will be presented in two parts. Part I addresses some of the risk management issues for physicians when screening patients for breast cancer. Part II, to be published in an upcoming Information Letter, will consider several risk management issues when diagnosing patients with symptoms and signs that might be related to breast cancer.
Mammography is used both for screening for breast cancer (screening mammography) and to contribute to the diagnosis of patients with clinical signs or symptoms of breast disease or patients with a suspicious finding on a screening mammogram (diagnostic mammography). The diagnostic workup may include additional mammographic views such as coned or magnification views that focus on areas of concern, and/or guided needle or trocar biopsy. Other diagnostic imaging approaches may be useful in specific cases.
Part 1 - Screening for breast cancer
A woman may be screened by her physician for breast cancer because of a recommendation related to the periodic health exam for her age group, or because of her individual risk factors for developing breast cancer.
All provinces and territories in Canada have established screening programs for targeted populations, with criteria for ongoing quality control. For breast cancer screening, women may enter these programs in various ways; the entrance criteria and process may differ from one province/territory to another. Most provincial/territorial screening programs will accept patients without a referral from a physician. Screening mammography is also performed in hospitals and clinics on patients who choose not to participate in a provincial/territorial screening program or who do not meet its enrolment criteria. Patients and physicians should be aware of the process used in their region.
The following two cases illustrate several risk management considerations.
An asymptomatic 53-year-old patient was sent by her family physician for a screening mammogram. This revealed an asymmetric density in the upper outer quadrant of the right breast. The radiologist investigated further with coned compression views and interpreted the asymmetric density was likely due to overlap of otherwise normal fibroglandular tissue.
The radiologist's report was sent to the family physician's office with a recommendation for a follow up mammogram in six months. The family physician did not see the report and the office did not have a system for following up on ordered investigations.
Eleven months later, the patient returned for a routine periodic health examination. Not recalling the previous request for the already reported mammogram and not reviewing the medical record where the report was filed, the family physician sent the patient for a second screening mammogram. The second mammogram now revealed a progression of the asymmetric density noted the previous year to a 1.5 cm density with associated microcalcifications. The radiologist recommended an ultrasound examination and commented that "if no definite cyst is found in this location, an excisional biopsy would be recommended to exclude a neoplastic lesion."
The ultrasound examination, performed later the same day by the same radiologist, failed to show a cyst. The ultrasound report did not refer to the report of the mammogram nor did it recommend an excisional biopsy based on the failure to detect a cyst.
The ultrasound report was attached to the mammogram report and both were received at the family physician's office. The physician saw only the mammogram result. Unaware that the complementary ultrasound result had arrived, the physician continued to wait for it.
The patient was seen by another family physician two months later for an unrelated issue. The latter physician requested the patient's medical record. That record included the mammogram reports of the previous two years as well as the ultrasound report, and the radiologist's recommendation for an excisional biopsy. A subsequent investigation confirmed malignancy and the patient underwent a partial mastectomy and axillary node dissection for T2N0M0 infiltrating ductal carcinoma Grade III with extensive lymphovascular invasion. Further oncologic treatment followed.
The patient complained to the College that the first family physician did not inform her of the abnormal mammograms and the need for a follow up biopsy which resulted in a delay in the diagnosis of her cancer.
The College was critical of the family physician for the lack of an office system to follow up test results. It also expressed concern about the reporting of the imaging studies. The College acknowledged that the second mammogram and ultrasound reports were somewhat confusing: the recommendation on the mammogram report was unclear as to the timing of the ultrasound, and the latter report did not make reference to the recommended course of action. Nevertheless, the College required the family physician to attend the College committee in person to be cautioned, following which the physician implemented an office system to track the results of investigations.
A 50-year-old patient experiencing bilateral throbbing breast pain attended her family physician after noticing a change in the texture of her left breast. On physical inspection the physician noted a flat appearance of the left nipple and palpation revealed a slight difference in the left breast parenchymal texture. The physician referred the patient for a mammogram but did not indicate the patient was symptomatic. The mammogram was therefore booked as a screening mammogram rather than a diagnostic mammogram.
The patient completed an entry questionnaire used at the clinic. She noted some bilateral breast tenderness, but responded "no" to questions regarding the presence of breast masses or nipple abnormalities. The mammography technologist noted at the bottom of the patient's questionnaire that the left nipple was abnormally hard and had a tendency to retract. The mammogram, which was performed, processed and read as a screening mammogram, was reported as normal by the radiologist.
Despite the normal mammogram report, the patient's family physician maintained a high index of suspicion based on the unexplained clinical findings on breast exam, and referred the patient to a surgeon. The patient was seen by the surgeon five weeks after the mammogram. He arranged to have the mammogram reviewed by a second radiologist, who noted a suspicious 15 mm nodular density behind the left nipple. Further investigation followed, and a biopsy done later that month confirmed a cancer. The patient underwent appropriate surgical and oncologic treatment.
The patient lodged a complaint with the College against the first radiologist alleging a delay in diagnosis. When asked by the College to review the films, the radiologist admitted that he would likely have identified the abnormality if he had reviewed the information available to him at the time.
A peer radiology expert for the College, although noting the patient's mammogram should have been scheduled and evaluated under the diagnostic program and not the screening program, was critical of the first radiologist for not taking into consideration the information provided by the technologist and available when reading the films – regardless of whether it was classified as a "screening" or a "diagnostic" mammogram.
This case highlights the importance of providing complete and accurate information on the requisition for an imaging study, thus improving communication and other processes in the system for the provision of care. It also emphasizes the importance of clinical judgment where a suspicious and unexplained clinical finding may warrant further assessment in spite of a normal report of a mammogram or breast ultrasound.
It is worthwhile for radiologists to be aware that some physicians provide patients with a requisition for a screening mammogram at the time of an earlier appointment such as a periodic health exam, requesting that the patient obtain this screening mammogram prior to her next appointment. In the interval, the patient may develop a symptom or sign of breast disease and provide this new information to the technologist at the time of the "screening" mammogram. This important new information would not appear on the physician's requisition for the mammogram and may only be recorded in the patient questionnaire completed with the technologist.
Appreciating the limitations of mammography is important for those who have ordered the study and those reporting on this type of imaging. It is widely accepted that not all breast cancers can be identified on mammograms. Some mammograms show very subtle or even no detectable signs of malignancy, for example a mass, microcalcifications, architectural distortion or an asymmetric or developing density. Some malignancies are completely obscured by overlying highly dense glandular and fibrous tissue.
It is common for radiologists to compare the current mammogram with one or multiple prior studies. It is common on such retrospective assessments to now be able to identify early changes in the breast architecture that reflect the development of cancer. Knowing where to look, one may in hindsight identify these subtle changes which at the time were below the threshold of detection of a competent radiologist. For example, in referring to a prior study, the radiologist as appropriate may choose to use language such as "There is a developing density or cluster of microcalcifications" or "Knowing where to look, I now see…." or "With the benefit of the knowledge I now have, I note that…."
In speaking to your patient about such retrospective findings, it is important to reassure the patient that this is a recognized limitation of mammography, indeed of many diagnostic imaging studies, and does not likely represent poor interpretation or poor clinical care.
The CMPA experience related to screening for breast cancer
Breast cancer screening programs exist in all Canadian provinces and territories. Given the large number of screening mammograms performed in Canada each year, there are very few medico-legal problems related to screening mammography. This speaks to the high quality of these valuable programs.
Nevertheless, this CMPA analysis identified a small number of cases, eight between 2003 and 2007, of which three were College complaints and five were legal actions.
The College, in part based on expert opinions, expressed concerns in all three of the cases they investigated. In four of the legal files, the experts' opinions were not supportive of the care provided by the physicians involved, resulting in settlements. One legal action was dismissed.
Managing medico-legal risks - Breast cancer screening
The following risk management considerations are based on the expert opinions:
For referring physicians:
Have you provided sufficient clinical information to the radiologist?
Do you have a tracking system to facilitate the follow up of investigations?
Have you correctly interpreted the meaning of imaging reports?
Have you made the patient aware of any abnormality and the need and process for follow up?
Have you made the patient aware of any limitations of mammography studies?
Have you considered the information provided by the referring physician, patient and technologist when interpreting screening mammograms?
Have you recommended follow up or further investigations as appropriate?
If recommending further investigations or other follow up, is your report clear as to who would be responsible for arranging these?
The role of medical experts
In these cases, medical experts were asked to comment on the care provided. These were physicians of similar training and experience working in the same kind of practices. The experts help establish the applicable standard of care.