Adoption of low-dose CT lung cancer screening is growing, and that’s a good thing. But rising LDCT utilization is also creating new challenges, illustrated by a new study in JACR that found high variability in radiologist interpretations.
CT lung screening participation rates are rising, from single digits a decade ago to 19% in a recent survey.
- But there’s still a long way to go to get lung screening to the roughly 75% participation rate seen with breast cancer screening.
One barrier to getting more people screened is false positives, which can cause patient anxiety and trigger a cascade of additional exams and procedures to work up suspicious findings.
- A major contributor to false positives may be variability in how radiologists read CT scans, so the JACR researchers investigated how interpretations for the same images changed between imaging specialists.
Researchers looked at interpretations made by 1.4k radiologists, each of whom read a minimum of 50 LDCT screening exams (a total of 200k exams), then calculated false-positive rates for both baseline and follow-up exams. Researchers found…
- High variability in false-positive rates, from 11% to 22%.
- In random pairs of radiologists, on average, one radiologist would have twice the odds of making a false-positive call as the other.
- Radiologists with false-positive rates higher than the median rate also unsurprisingly had higher sensitivity than those below the median (97% vs. 92%).
What’s causing the variation? It could be because some radiologists use image interpretation patterns that increase false positives but also catch more cancers.
- In statistical terms, researchers found that for baseline screens read by radiologists who were below the median false-positive rate, they would have about 100 fewer false-positive screens for one fewer true-positive cancer.
Radiologist variability could be reduced through programs to standardize screening interpretation, such as those implemented in mammography screening.
- These could include support tools, automated feedback systems, educational interventions, and targeted double reading.
The Takeaway
The new findings on radiologist variability in reading CT lung cancer screening exams illustrate the delicate balance between detecting more cancer at the risk of producing additional false positives. Is the tradeoff worth it? We think we know how most patients would feel.
