Mammogram

If you’ve had a mammogram in the past few years, chances are it involved computer-aided detection (CAD).

CAD is a type of computer software that helps radiologists interpret results by automatically marking suspicious spots.

Then it’s up to radiologists to review those spots. It’s often described as a “second set of eyes.”

Mammogram Device

The technology increases the cost of a mammogram, but is it worth it?

A new study published in the Journal of the American Medical Association Internal Medicine found that CAD doesn’t improve diagnosis accuracy. It may even result in missed cancers.

Dr. Constance D. Lehman, Ph.D., director of breast imaging and co-director of the Avon Comprehensive Breast Evaluation Center at Massachusetts General Hospital in Boston, led the study. She did the research while at the University of Washington in Seattle.

The senior author was Diana Miglioretti, Ph.D., dean’s professor in biostatistics at the University of California at Davis.

Learn More: Mammogram: Purpose, Procedure, and Risks »

Data Reveal Radiologists Do Better Without CAD

The research team used data from the Breast Cancer Surveillance Consortium. The study involved 323,973 women who had digital mammograms from 2003 through 2009.

There were 271 radiologists from 66 facilities who read the results. Tumor registry data identified 3,159 breast cancers within one year of screening.

The researchers analyzed cancers that were positively detected. They also looked at accurate negative findings. Then they tracked cancers found by mammogram and those found within a year of a negative result.

CAD did not improve radiologists’ accuracy in any of the circumstances, researchers said.

We're concerned that using CAD for screening mammography does not benefit women. And [it] may even increase the chances a radiologist misses a breast cancer.
Diana Miglioretti, Ph.D., University of California at Davis

There were 107 radiologists who read mammogram results with CAD and again without CAD. They missed more cancers when using CAD.

Lehman theorized that radiologists using CAD may be overly dependent on it.

“We're concerned that using CAD for screening mammography does not benefit women,” said Miglioretti in a press release. “And [it] may even increase the chances a radiologist misses a breast cancer.”

CAD did better at finding more noninvasive ductal carcinoma in situ (DCIS). That’s stage 0 breast cancer. It’s not clear if that affects outcomes.

Related Reading: Why Do We Still Not Know Who Needs a Mammogram? »

Why Do We Keep Paying for This Technology?

CAD has been around for almost 20 years.

Lehman estimates that CAD systems cost the United States more than $400 million per year. That’s $1 of every $10,000 we spend on healthcare.

Dr. Joshua Fenton, associate professor of family and community medicine at U.C. Davis, wrote the editorial response that was published with the study and feels Medicare should stop paying for CAD.

“Congress decided to establish Medicare reimbursement for CAD in 2000, when there was limited evidence of its effectiveness,” Fenton told Healthline. “Many private insurers follow Medicare's lead in coverage decisions. CAD is widely used in the United States because providers are reimbursed for using it.”  

CAD is widely used in the United States because providers are reimbursed for using it.
Dr. Joshua Fenton, University of California at Davis

Fenton has been talking about CAD for a long time.

He led a 2007 study that found that women who got screened at centers using CAD devices were more likely to have an abnormal result. That increases the need for a biopsy to rule out breast cancer. Additional testing can contribute to stress and anxiety.

Digital mammography may improve the reading anyway, he noted.

For a mammogram screening read with CAD, Medicare pays about $8 to $10 per screen, said Fenton. Many private insurers pay up to $20.

“We need to make sure there is stronger evidence for new technologies before we start paying for them,” Miglioretti said. “Before people are charged extra money for them.”

More than 90 percent of mammograms in this country are interpreted using CAD, according to Fenton.

If CAD isn’t helpful and it’s adding to healthcare costs, why do we still use it?

“There are many with significant financial interests in CAD who publicize their own facts to support the status quo,” said Fenton. “This is a problem in many areas of the U.S. healthcare system, where what we do is not necessarily based on the high-quality science, but on politics, influence, and sometimes force of habit.”

Read More: New Breast Imaging Techniques Reduce False Positives and Increase Breast Cancer Diagnoses »