Is the gambler’s fallacy impacting radiologists as they interpret screening mammograms? The authors of a new case study published in the Journal of the American College of Radiology looked to answer that very question.
“It has been documented that in multiple settings, the sequencing of decisions affects decision making,” wrote Andrew L. Callen, MD, department of radiology and biomedical imaging at the University of California, San Francisco, and colleagues. “Colloquially, this idea is known as the gambler’s fallacy: the tendency for people to underestimate the likelihood of ‘streaks’ occurring by chance. For example, someone taking a multiple-choice test may be vexed by the chance occurrence of multiple sequential ‘C’ answers and may second-guess the next answer based on this pattern alone.”
The authors explored data from all patients who underwent screening mammography at a single institution throughout 2014. The BI-RADS assessment of more than 8,500 consecutive examinations were reviewed, with the team noting the order in which they occurred.
The team determined that it would be proof of the gambler’s fallacy in action “if a greater proportion of false-positive assessment occurred after a preceding streak of negative assessments.” They also assessed the proportion of true-positives and false-positives among examinations preceded by “increasing numbers of BI-RADS 1 or 2 assessments.”
Overall, the team could not detect any evidence of his phenomenon impacting the institution’s radiologists.
“Our study found no effect of the gambler’s fallacy on decision making in screening mammography,” the authors wrote. “That is, radiologists’ screening mammography decision making was not affected by the number of preceding BI-RADS 0 or BI-RADS 1 and 2 assessments in the readout session.”
Callen and colleagues also noted that their study did not cover radiologists without subspecialty training in breast imaging or those who spend much less time reading screening mammograms. To see if such specialists are impacted at all by the gambler’s fallacy, they wrote, would be “an area ripe for future research.”