Cross-matching mammo exams with treatment data sets yields large-scale insight

January 07, 2019
by John W. Mitchell , Senior Correspondent
In the long-term conversation about the pros and cons of screening mammography, a key data point has always been lacking. Conclusions about the link between mammography and outcomes have been based on extrapolation modeling from a small sample size.

“Some of the most critical issues include debate regarding the mortality reduction that women diagnosed with breast cancer receive from screening versus that from improved treatment, the decrease in interval cancers because of screening, and the benefits of screening to younger women,” Dr. Margarita Zuley, vice chair, quality and strategic development, and chief, division of breast imaging, Department of Radiology, University of Pittsburgh Medical Center, told HCB News.

Such debate has led to often conflicting breast imaging screening recommendations. Several national medical associations and leadership groups including NCCN, ACOG, and ACR continue to recommend annual screening beginning at age 40, according to Zuley. Others, such as USPTFS and ACS, recommend age 50 and 45.

Zuley said this creates “chaos” for patients.

However, a proof of concept evaluation just published in the Journal of the American College of Radiology might soon offer better insight based on two large data sets with the power to create agreement.

“This study is a proof of concept trial to link the American College of Radiology National Mammography Database (NMD) – which collects all breast imaging history for participating facilities – to the network of required state cancer registries, which collects all details of the cancer diagnosed and treatment until patient death,” said Zuley, who served as lead author. “Such linkage would allow direct analysis … this linkage would resolve the current debate and inform best practice.”

She said that it is feasible for any clinical provider to reproduce the study. Every MQSA-approved mammography facility must perform annual audits, and as part of the audit must collect detailed data on imaging performed, interpretations, and cancers diagnosed. Also, facilities are required to provide cancer information to their state registry so that every facility collects the needed treatment and outcomes data for patients primarily treated within their site(s).

Such linkage, Zuley added, would allow actual patient outcomes to inform medical decision-making regarding the benefits of the many elements of screening and treatment for cancer. The idea can be extended to other cancers as well.

“A national linkage will require collaboration by several entities including CDC, ACR and potentially NCI as well as others," said Zuley. “Linked data and resultant analytics will require funding, storage, security, and oversight.”

More Women's Health

New study finds AI breast screening interpretations on par with those of radiologists

South Dakota passes breast density law

FDA warns against thermography alone for breast cancer detection

Mammography reports nationwide to include patient breast density

Joseph J. Cappello

Three decades of data clearly show mammography saves lives

Hologic launches Unifi Analytics to curb mammo downtime

Sarah Silverman calls out radiologist after bad breast screening experience

Self-compression for breast cancer screening yields clinical advantages

Mammo prior to breast reduction surgery may do more harm than good