The Case

A 60-year-old woman was seen for a routine visit by a physician assistant (PA) at a family medicine practice. A right breast mass was palpated and felt likely to be benign. However, the PA ordered a diagnostic mammogram and ultrasound examination. The radiologist reviewed the images, noting focal asymmetry in the right breast. The assessment was challenging due to dense breast tissue. Ultimately, the films were interpreted as “probably benign” findings (BI-RADS Category 3) and follow-up imaging at 6 months was recommended to ensure stability. The report noted that a biopsy should not be delayed if a “suspicious mass” is present on physical exam. Reassured by the report, the PA did not order a biopsy or refer the patient to a breast surgeon.

When the patient returned 5 months later for a follow-up appointment, the mass was found to have increased in size. She was referred to a surgeon for a biopsy, but there was confusion about whether the patient herself was to schedule the appointment or whether the surgeon’s office would contact her. After an 8-week delay, the patient contacted the family medicine office. The PA was surprised that she had not yet seen the surgeon and arranged for an urgent surgical appointment and biopsy of the enlarging right breast mass. The biopsy showed invasive breast cancer, which, now 7 months after the initial presentation, was found to be metastatic to the axillary nodes and spine. It is unknown if an earlier diagnosis would have affected the outcome of this aggressive malignancy.

The Commentary

By Saul N. Weingart, MD, MPP, PhD, Gordon D. Schiff MD, and Ted James, MD, FACS

Breast cancer is an exceedingly common and potentially fatal disease. The most frequently diagnosed non-skin cancer among women in the US, breast cancer accounted for 250,520 new diagnoses and 42,000 deaths in the US in 2017.¹ Early detection and appropriate treatment have been associated with improved survival, with the majority of patients diagnosed with early-stage disease experiencing long-term survival.

Against this backdrop, the present case illustrates several challenges in the timely diagnosis of this feared cancer diagnosis in women. How to explain to the patient how a diagnosis so widespread and worrisome could have been delayed for more than a half a year?

Breast cancer mortality has decreased internationally, concurrent with the widespread use of screening mammography and the introduction of more sophisticated cancer-specific therapies.² Although there is debate among cancer epidemiologists and professional society panels about the appropriate use of screening mammography for selected patient populations,^3,4 the favorable trend in breast cancer survival is impressive, especially in resource-rich countries.^1,3

Nevertheless, cases of delayed breast cancer diagnosis are a common cause of medical malpractice claims. According to a study by The Doctor’s Company and CRICO, the captive malpractice insurer of the Harvard-affiliated hospitals and practice groups, missed and delayed breast cancer malpractice claims are a regular occurrence.⁵ In a study of 562 breast cancer malpractice claims from 2009-2014, the most common contributing factors to diagnostic delays were:

• misinterpretation of diagnostic studies (49%),
• delay or failure in ordering diagnostic tests (27%),
• failure or delay in obtaining a consultation (17%),
• miscommunication between patient/family and providers (16%), and
• miscommunication among providers (12%).

There is, however, evidence that greater awareness of diagnostic pitfalls and adopting best practices for protocol-driven communication and workup may be effective.^6,7 Despite improvements in the workup of breast complaints and follow-up of abnormal mammograms, radiologists cannot always distinguish benign from malignant lesions. In addition, delayed diagnosis due to access barriers continues to happen frequently. Adding further to the urgency of improved access and diagnosis is the fact that wide disparities in care lead to excess morbidity and mortality among women of color, those with poor health literacy, and those who are resource-constrained.⁸

The present case illustrates some of the challenges that can lead to delayed diagnosis of breast cancer and provides insight into the complexity of the diagnostic process. Notwithstanding the 7-month diagnostic delay in this case, many aspects of the patient’s care went well. The patient sought care. The PA discovered an abnormal physical exam finding. Initial imaging studies were ordered, performed, reviewed, and communicated. Moreover, a follow-up appointment (in primary care) was successfully arranged and completed. While many components of the diagnostic process went well,⁹ the process is sufficiently complex, and fraught with uncertainties and interdependencies, that potentially preventable delays occurred.¹⁰

The first challenging element in the case was the radiology report of a BI-RADS 3 lesion. This is an intermediate category that breast imaging radiologists use to designate lesions that are “probably benign,” based on defined radiologic criteria. These findings are associated with a low risk of malignancy, usually estimated as <2%, although figures from 0.9% and 7.9% are reported in the literature.¹¹ The American College of Radiology says that “a probably benign finding is not expected to change over the suggested period of imaging surveillance, but the interpreting physician prefers to establish stability of the finding before recommending management limited to routine mammography screening.”¹² BI-RADS 3 falls between the lower risk categories of BI-RADS 1 (normal) and BI-RADS 2 (benign) and the high risk categories of BI-RADS 4 (suspicious) and 5 (highly suspicious). BI-RADS 3 has been described as “perhaps the most difficult of the assessment categories for the breast imager to properly use... there is considerable interobserver variability in the assessments” of BI-RADS 3 findings. There can also be variability in the ultrasonography technician’s and/or supervising physician’s performance and interpretation of the ultrasound.

Similarly, there are challenges and opportunities for misinterpretation at the receiving end of a BI-RADS 3 report. In this case, the report suggested the need for a more timely biopsy “if a suspicious mass is present on physical exam.” In its guidelines, the American College of Radiology notes that “the use of a probably benign assessment for a palpable lesion is not supported by robust scientific data, although there are two single-institution studies that do report successful outcomes for palpable lesions” (i.e., <3.2% malignancy).^13-15 One might argue that the finding of any new palpable breast mass in a 60-year-old woman should be considered “suspicious,” certainly more so than in younger age groups, in which fluctuating hormone levels and benign palpable findings are common. Current best practice guidelines suggest that breast radiologists should palpate any identified mass and consider the pre-test probability of cancer and the corresponding sonographic findings before assigning BI-RADS category 3.^16,17 As a general principle, any woman aged 40 years or more with a palpable mass, regardless of the mammographic interpretation, requires careful evaluation and follow up.¹⁸

The physician assistant in this case perhaps assigned excess weight to the radiologist’s interpretation of the imaging findings, forestalling additional consideration or help-seeking. Although radiologists may consider themselves to be “essential and central... in the appropriate care of women with palpable masses, because most of the workup and diagnosis of a palpable breast lump is routinely completed within the imaging department,”¹⁹ primary care providers cannot rely on radiologists’ authority to dismiss worrisome findings on the history and physical exam. People are susceptible to what Nobel psychologist Daniel Kahneman describes as “WYSIATI” – what you see is all there is.²⁰ We assume that what is reported is all there is, rather than searching for missing information or additional guidance. We are inclined to defer to expertise without assessing the scope of the expertise.

The case provides limited background needed to understand important information about the family practice work environment. In this case, the provider was a PA; however, it could well have been any primary care provider, particularly if they were relatively inexperienced in diagnosing and managing breast lumps. Important questions to be considered are:

• Did the provider assess the patient’s cancer risk based on factors such as family history of malignancy, or personal history of early menarche or prior breast biopsy?
• What was the training and skill level of the provider in breast examination?
• If the provider was new or had minimal experience, was there a more experienced clinician available to consult?
• Did the culture in the organization encourage consultation and provide support for new clinicians?^21,22
• Did the provider seek advice about the need and timing of follow up? Many primary care clinicians arrange a 1-2 month follow up of a breast lump or other unexpected exam findings, notwithstanding the BI-RADS category 3 assessment from imaging.

Equally important, we know little about the instructions that the clinician provided to the patient herself. Was she advised to perform breast self-exams to monitor for any changes in the breast lump, and shown how to do so? Although at this time, research has concluded that breast self exam is not effective for breast cancer screening, there is likely benefit for patients monitoring known symptoms (e.g., following breast lumps or other findings in “watch and wait” diagnostic strategies) to determine if the symptom is worsening or resolving.²³ Was she advised who and how to call with questions, or what to do if she observed or felt a change in the lump? We know that patient behavior may contribute to diagnostic delays. In the Doctors Company-CRICO study cited above, patients’ failure to complete recommended appointments or treatments occurred in 19% of cases. In another study of delayed breast cancer cases, patient-related factors accounted for one in four delays.²⁴ Contributing factors include fear of receiving a malignant diagnosis and misunderstanding or ambivalence about a clinician’s recommendation for follow-up. Linguistic, health literacy, financial, and logistical barriers (e.g., arranging time off work, childcare, transportation) can adversely affect patients’ ability to navigate the diagnostic process.²⁵

Both the system and the patient may have played a role in the second delay in this case: delayed surgical evaluation of a growing breast mass. The narrative attributes this delay to ambiguity about who was to arrange the appointment, a phenomenon that is common in clinical care and under-studied. While closed-loop test result management is a robust intervention embedded in many commercial electronic health records (EHRs),²⁶ closed-loop referral management is a less mature intervention.^27-29 Few healthcare organizations have a robust mechanism to track completed referrals or to ensure that intended consultations are completed. Appointments might be initiated by the referring MD, their office staff, or the patient directly. It is often not clear to the patient who is responsible for scheduling an appointment. We have little information on the variety of ways that these systems fail or how to prevent such failures. It is often impossible to distinguish patients’ ambivalence to seek recommended care from the myriad other obstacles that get in their way (e.g., lack of communication, poor education, failure of care coordination). This can be a perplexing problem, given that completing a referral for testing or consultation is in the medical-legal interest of the referring clinician, the financial interest of the intended recipient, and the best interest of the patient.

Diagnosis is a complex process that relies on cognitive processing by health professionals and the design of effective organizational processes. Viewed through the lens of ‘diagnostic excellence’, the case illustrates multiple process-related challenges. While cognitive failures may have played a role, we can do more to ensure that systems support clinicians and patients and make it easier for them to do the right thing. Better decision support systems with enhanced mechanisms to reduce variability and enhance reliability in both primary care and radiology practices, as well as enhanced patient education, may effectively reduce the risk of a delayed breast cancer diagnosis.

Take-Home Points

1. Despite the frequency of breast complaints, evaluating a breast lump is a beguilingly complex process, prone to cognitive error and system failures.
2. Potentially discordant findings between imaging and exam require meticulous, timely follow up.
3. Relying on official test reports may convey a false sense of security, under-estimating risk, and over-estimating confidence.
4. Patient education, access, alignment, and navigation are important strategies to promote patient partnership in diagnostic excellence.
5. Ensuring structured inter-professional handoffs and closed-loop referrals are under-appreciated opportunities to reduce the risk of diagnostic delays and associated harm.

Saul N. Weingart, MD, MPP, PhD
Chief Medical Officer
Tufts Medical Center

Gordon D. Schiff, MD
Brigham and Women’s Hospital
Harvard Medical School

Ted James, MD
Chief, Breast Surgical Oncology
Beth Israel Deaconess Medical Center

Acknowledgements:

The long-standing process for submitting PSNet WebM&M case submissions is anonymous. Users may contribute by submitting a case at the following link: https://psnet.ahrq.gov/webmm/submit-case

Periodically, the Primary-Care Research in Diagnosis Errors (PRIDE) Learning Network, a collaborative project convened by the Brigham and Women’s Hospital Center for Patient Safety Research and Practice, and the State of Massachusetts Betsy Lehman Center for Patient Safety, contributes cases and commentaries from their monthly discussions of diagnosis error cases to PSNet. PRIDE is funded by a grant from the Gordon and Betty Moore Foundation. This case was produced in cooperation with the PRIDE Learning Network. We acknowledge the assistance of the PRIDE project director Maria Mirica, PhD, in preparing this case discussion.

We acknowledge the assistance of Dr. Paula Grabler (Assistant Professor, Department of Diagnostic Radiology and Nuclear Medicine, Rush Medical College). Editorial support and review for this case was provided by Gordon Schiff, MD.

References

1. Center for Disease Control and Prevention. United States cancer statistics: data visualization. Available at: https://gis.cdc.gov/Cancer/USCS/DataViz.html. Accessed 11 Sept 2020.

2. Bleyer B, Baines C, Miller AB. Impact of screening mammography on breast cancer mortality. Int J Cancer 2016; 138: 2003–12.

3. Oeffinger KC, Fontham ETH, Etzioni R, et al. Breast cancer screening for women at average risk: 2015 guideline update from the American Cancer Society. JAMA 2015; 314:1599–1614.

4. US Preventive Services Task Force. Breast cancer: screening, 2016. Available at: https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/breast-cancer-screening#fullrecommendationstart. Accessed 11 Sept 2020.

5. Greenberg P, Ranum D, Diegal D. Navigating risks in breast cancer diagnosis and treatment. Patient Safety and Quality Healthcare; 2015. Available at: https://www.rmf.harvard.edu/Clinician-Resources/Article/2015/Navigating-Risks-in-Breast-Cancer-Diagnosis-and-Treatment. Accessed 12 Sept 2020.

6. CRICO breast cancer management algorithm: a decision support tool. Boston, MA: Risk Management Foundation of the Harvard Medical Institutions; 2019. Available at: https://www.rmf.harvard.edu/Clinician-Resources/Guidelines-Algorithms/2019/BCA-Guideline-Contents. Accessed 11 Sept 2020.

7. Poon EG, Haas JS, Louise Puopolo A, et al. Communication factors in the follow-up of abnormal mammograms. J Gen Intern Med 2004; 19:316-23.

8. Freund KM, Battaglia TA, Calhoun E, et al. Impact of patient navigation on timely cancer care: the Patient Navigation Research Program. J Natl Cancer Inst 2014; 106:dju115.

9. National Academies of Sciences, Engineering, and Medicine. Improving Diagnosis in Health Care. Washington, DC: National Academies Press; 2015.

10. Perrow C. Normal Accidents: Living with High-Risk Technologies. Princeton, NJ: Princeton University Press; 1984.

11. Lee KA, Talati N, Oudsema R, Steinberger S, Margolies LR. BI-RADS 3: Current and Future Use of Probably Benign. Curr Radiol Rep. 2018;6(2):5. doi: 10.1007/s40134-018-0266-8. Epub 2018 Jan 27. PMID: 29399419; PMCID: PMC5787219.

12. American College of Radiology (ACR). BI-RADS Atlas – Mammography Reporting Systems. 2013. Available at: https://www.acr.org/-/media/ACR/Files/RADS/BI-RADS/Mammography-Reporting.pdf. Accessed December 11, 2020.

13. Graf O, Helbich TH, Fuchsjaeger MH, et al. Follow-up of palpable circumscribed noncalcified solid breast masses at mammography and US: can biopsy be averted?. Radiology. 2004;233(3):850-856. doi:10.1148/radiol.2333031845

14. Harvey JA, Nicholson BT, Lorusso AP, Cohen MA, Bovbjerg VE. Short-term follow-up of palpable breast lesions with benign imaging features: evaluation of 375 lesions in 320 women. AJR Am J Roentgenol. 2009;193(6):1723-1730. doi:10.2214/AJR.09.2811

15. Shin JH, Han BK, Ko EY, Choe YH, Nam SJ. Probably benign breast masses diagnosed by sonography: is there a difference in the cancer rate according to palpability?. AJR Am J Roentgenol. 2009;192(4):W187-W191. doi:10.2214/AJR.08.1404

16. Berg WA, Berg JM, Sickles EA et al. Cancer yield and patterns of follow-up for BI-RADS category 3 after screening mammography recall in the National Mammography Database. Radiology 2020; 296:32-41.

17. Expert Panel on Breast Imaging:, Moy L, Heller SL, et al. ACR Appropriateness Criteria® Palpable Breast Masses. J Am Coll Radiol. 2017;14(5S):S203-S224. doi:10.1016/j.jacr.2017.02.033

18. CRICO Breast care management algorithm. Available at: https://www.rmf.harvard.edu/Clinician-Resources/Guidelines-Algorithms/2019/~/link.aspx?_id=19E2C8000CC542B7B2D103B98357B97C&_z=z. Accessed 15 October 2020.

19. Lehman CD, Lee AY, Lee CI. Imaging management of palpable breast abnormalities. AJR Am J Roentgenol. 2014;203(5):1142-1153. doi:10.2214/AJR.14.12725

20. Kahneman D. Thinking, Fast and Slow. New York: Farrar, Straus and Giroux; 2011.

21. Barton MB, Harris R, Fletcher SW. Does his patient have breast cancer? The screening clinical breast examination: should it be done? How? JAMA 1999; 282:1270–80.

22. Palakshappa D, Carter LP, El Saleeby CM. Discrepancies in after-hours communication attitudes between pediatric residents and supervising physicians. J Pediatr 2015; 167:1429-35.e2.

23. Nekhlyudov L, Fletcher SW. Is it time to stop teaching breast self-examination?. CMAJ 2001; 164:1851-2.

24. Weingart SN, Saadeh MG, Simchowitz B, et al. Process of care failures in breast cancer diagnosis. J Gen Intern Med 2009; 24:702-9.

25. Gerend MA, Pai M. Social determinants of black-white disparities in breast cancer mortality: a review. Cancer Epidemiol Biomarkers Prev 2008; 7:2913-23.

26. Poon EG, Wang SJ, Gandhi TK, Bates DW, Kuperman GJ. Design and implementation of a comprehensive outpatient Results Manager. J Biomed Inform 2003; 36:80-91.

27. Ramelson H, Nederlof A, Karmiy S, et al. Closing the loop with an enhanced referral management system. JAMIA 2018; 25:1715-21.

28. Emani S, Sequist TD, Lacson R, et al. Ambulatory safety nets to reduce missed and delayed diagnoses of cancer. Jt Comm J Qual Patient Saf 2019; 45:552-7.

29. Institute for Healthcare Improvement / National Patient Safety Foundation. Closing the Loop: A Guide to Safer Ambulatory Referrals in the EHR Era. Cambridge, Massachusetts: Institute for Healthcare Improvement; 2017. Available at: http://www.ihi.org/resources/Pages/Publications/Closing-the-Loop-A-Guide-to-Safer-Ambulatory-Referrals.aspx. Accessed 29 Sept 2020.