Lower screening mammography recall rates are associated with higher rates of breast cancers detected between screenings, according to a new study published in Radiology. Does this relationship between recall rates and interval cancers mean breast cancer screening programs should establish a minimum recall rate?

The authors studied more than 5 million screening episodes from 84 facilities taking part in the U.K. National Health Service Breast Screening Program (NHSBSP). All data was from a three-year timeframe with a three-year follow-up in women ages 50 to 70. Overall, the recall rate was 4.56 percent with an interval cancer rate of 3.1 per 1,000 patients screened. Author Elizabeth S. Burnside, MD, MPH, from the University of Wisconsin School of Medicine and Public Health in Madison, Wisconsin, and colleagues noted that lower recall rates were associated with higher interval cancer rates. According to their estimates, it would take an additional 80 to 84 recalls to “avoid one interval cancer.”

In addition, the authors added, cancer detection rates and interval cancer rates were lowest in younger age groups.

“Recall rate had more of an impact on interval cancer in patients who are older,” Burnside said in a prepared statement. “The lower number of recalls required per interval cancer avoided in older women and incident screens as compared to younger women and prevalent screens, respectively, demonstrates a slightly different ‘value’ in terms of the trade-off between recall rate and interval cancers.”

Since lower recall rates lead to more interval cancers, and interval cancers are associated with a worse prognosis than cancers detected during screening, Burnside noted that screening programs have reason to consider implementing a minimum recall rate.

“A big-picture lesson in our study is the power of rigorous quality assurance infrastructure to help breast cancer screening programs learn from actual practice and use that information to make informed programmatic decisions for the future,” she said in the statement.