Challenges in Mammography Interpretation and Over-reliance on Imaging

Key Insight

Interpreting mammograms to detect breast cancer is complex due to the varied appearances of potential indicators. Cancer can manifest as lumps or calcium deposits, but distinguishing between benign and malignant forms is not clear-cut. Benign calcium deposits come in various forms, such as 'eggshell,' 'railway track,' 'popcorn,' or 'milk of calcium,' while 'irregular' calcifications can be either benign or cancerous. Similarly, while cancerous lumps typically have ragged walls, benign lumps can mimic tumors, and vice versa.

Unlike other internal organs with consistent anatomical patterns, breast tissue lacks a standardized appearance, making it difficult to define what is 'normal' from one individual to another, or even between a woman's left and right breast. A study involving ten board-certified radiologists reviewing 150 mammograms revealed significant discrepancies: cancer detection rates varied from 37% to 85%, and one perplexing mammogram yielded five different interpretations, including normal, benign, and cancerous, for the same patient. This variability is influenced by both radiologist skill and individual temperament.

An overly cautious approach in mammography, while seemingly beneficial, poses a significant problem. In the aforementioned study, the radiologist who identified the most cancers also recommended unnecessary follow-up workups—such as biopsies, ultrasounds, or additional X-rays—for 64% of healthy women. As mammography is a screening tool designed to exclude healthy individuals, such high rates of false positives cause undue anxiety, discomfort, and expense, diminishing its effectiveness. The reality is that the vast majority of ambiguous findings on mammograms are benign, and radiologists must account for this, much like airport baggage screeners who cannot flag every ambiguous item without causing systemic gridlock.