Do DMIST results underestimate FFDM's impact?

2005 09 27 16 01 15 706

The results of the landmark Digital Mammographic Imaging Screening Trial (DMIST) likely came as no surprise to radiologists who interpret digital mammograms. The formal findings that digital mammography is a superior imaging technology to use for women younger than 50 years of age, women with dense breasts, and premenopausal or perimenopausal women verified opinions these radiologists had already formed based upon their daily work experiences.

But those findings may even underestimate the impact of full-field digital mammography (FFDM), according to some mammography experts.

The DMIST trial, published in the New England Journal of Medicine (September 16, 2005) was the largest clinical trial for mammography performed in the world, enrolling 49,528 women, at a cost of $26 million to the U.S. government. Beginning in October 2001, participants with no signs or history of breast cancer had an annual mammogram performed at one of 33 sites, followed a year later by a second mammogram.

Both a traditional analog film-screen mammogram and a digital mammogram were performed on the same day for each visit, and two different radiologists interpreted the conventional and digital mammograms for each patient.

While the study began just as digital mammography equipment was becoming commercially available in the U.S., DMIST leader Dr. Etta Pisano and other distinguished mammographers persuaded the U.S. government to fund the trial of the technology because if it proved to be superior, its high cost could be better justified in the fight against breast cancer. An estimated 8% of all diagnostic imaging facilities proceeded to purchase and install at least one FFDM unit prior to the first published data. asked some radiologists involved in mammography for their professional opinion on the impact that the DMIST results would have on women's health, women's imaging services, and the impact on adoption trends of digital mammography by radiology facilities.

Increased confidence

"I'm pleased as punch that the results came out as positive as they did," said Dr. Steve Parker, a radiologist affiliated with the Sally Jobe Comprehensive Breast Centre in Denver. "It's been our experience almost from the beginning that FFDM is substantially better than film-screen."

The Denver practice became a beta site for digital mammography in the mid-1990's, evaluating a prototype for Fischer Imaging and later purchasing one of its very first commercial systems in 2001.

Dr. James Ruez, a radiologist with Woman's Hospital in Baton Rouge, LA, said he and his five colleagues were pleased and relieved by the decision.

"We are delighted to be among the top 10% of early adopters," Ruez said. "In our early experience, without having real data to back up our opinions, information density is higher than what we were seeing in an analog image, particularly with respect to calcifications."

When Hurricane Katrina struck, the department was beginning the process of deinstalling seven analog mammography rooms and replacing them with FFDM systems. He said that the DMIST findings further justified the hospital's sizeable investment and faith in the superiority of this new technology with respect to workflow improvement and for image quality.

"The results only confirm what our we thought," said Dr. Lawrence Bassett, Iris Cantor professor of breast imaging at the University of California, Los Angeles, and one of the 12 authors of the DMIST study. "We believe that digital images allow us to better evaluate the dense breast, because we are not limited by the film IH & D curve and have higher contrast."

Dr. Bruce Schroeder, director of breast imaging at Eastern Radiologists of Greenville, NC, said his practice made "a leap of faith 1.5 years ago." In addition to interpreting its own patients' studies using digital systems, Eastern Radiologists also reads mammograms for a large number of rural medical centers in eastern North Carolina that perform five to 30 analog studies daily. The mammographers had reached the conclusion that digital provided more information than film, but as a working practice, had no formal way to prove this, Schroeder said.

Underestimated impact?

Despite the positive findings, the nature of the trials may underestimate the impact of the technology, some experts feel.

Bassett pointed out that first-generation equipment was used at all the investigation sites.

"What we use today is far better than what our department used in the study, and it continues to improve," he said. "It is my opinion that the results should be multiplied by an unknown percentage in view of the improvements that have been made in the past five years with the newest systems. The image quality and image processing capabilities of new systems are much better, and it stands to reason that we can see more."

Mammography has constantly been evolving in its ability to improve image quality and the ability to see things, Bassett said.

"Digital mammography is one more step in this process," he said.

Schroeder felt that the DMIST study was "fairly heavily biased against digital."

"This had to be, because it is important to err on the side of a proven technology," he said. "Nevertheless, the trial took brand-new technology and doctors who were not experienced using it against a mature technology where both the equipment and the doctors were good at what they did. The fact that digital was slightly better but not statistically significant different for the overall population, and measurably better on the young and dense-breast subgroups of the population says to me that digital mammography technology is much better, especially with the improvements in equipment and software that have been made in the past four years."

An evolution, not a revolution

However, some radiologists are expressing concerns that patients are interpreting the DMIST results to believe digital technology itself is a real advance in breast cancer detection. This overlooks the impact of experienced readers, some say.

"One of the regretful side effects will be that the marketing of the technology will overshadow the experience of the interpreting physician in the process of breast cancer detection," said Dr. Rebecca Zuurbier, chief of breast imaging at the Sibley Center for Breast Health in Washington, DC, and a proponent of digital mammography. "It's the 'driver,' not the 'car' that impacts cancer detection. Radiology groups with little or no commitment to experienced professional interpretation of these studies may shortcut and market the 'car' to gain a market advantage."

Zuurbier also pointed out that patients may unfortunately defer their annual mammography exam until a "digital" appointment is available. The National Cancer Institute of the U.S. National Institutes of Health (NIH) estimates that 40% of women undergoing screening mammography have dense breasts, and only 8% of facilities have digital systems.

Dr. Cheryl Perkins, senior clinical advisor of the Susan G. Komen Breast Cancer Foundation, points out that "film mammography remains an excellent screening tool and should be continued on schedule rather than delaying an exam to get a digital mammogram."

Another cancer-fighting tool

Digital mammography adds a new option for women who traditionally had problems with the accuracy of mammograms because they have dense breasts, Perkins said.

"The DMIST results show that digital mammography can better assist radiologists in their detection of breast cancer in women at a much younger age," she said.

The DMIST study results may get enough attention to make gynecologists more aware of the fact that they should be referring more women for baseline mammograms at age 35, said Janet Sterritt, vice president of mammography women's healthcare solutions at computer-aided detection software developer iCAD.

"We are at the point where the five-year rate of survival of breast cancer patients is 98%," Sterritt said. "The 2% who die are young, because their cancers tend to be more aggressive. The excitement surrounding digital mammography may get a higher percentage of these women walking through the door and having access to the technology that can detect a higher percentage of their cancers."

Unfortunately, as Schroeder pointed out, North Carolina does not currently reimburse for the use of digital mammography for Medicaid patients. So an issue with digital mammography adoption facing a North Carolina-based hospital radiology department or radiology practice is what to do with the high-risk young Medicaid patient who might be an ideal candidate for digital mammography based on the DMIST results.

Paying for digital

Another issue with digital mammography is its price tag. FFDM systems cost almost four times more than analog systems. Not all radiology facilities can afford to purchase these, nor should they, the experts say. Although Ruez' hospital went totally digital, he said that the workflow advantages of eliminating paper and film and the related cost-savings for 45,000 exams per year justified the expense. Not many hospitals have this kind of volume, he said.

Low-volume mammography centers cannot justify the cost of FFDM systems, according to Schroeder and Zuurbier. Rural hospitals and clinics that do four to 10 mammograms per day provide a great community service by offering mammograms to a patient population who would not be willing to travel outside their community, Schroeder said. He pointed out that the DMIST results showed that film-screen and digital mammography are equivalent for the majority of the population.

Zuurbier expressed concern that hospitals might be pressured by a "digital is better" refrain, and purchase digital systems to their financial and workflow detriment.

Breast Health Management, a national consulting firm headed by Gerald R. Kolb, stresses that high volumes pay for digital mammography systems. Kolb counsels clients not to invest in a digital mammography system unless it can displace two analog units, or capacity can be significantly increased. High volume is not just needed to generate revenue to pay for the system; it also forces the radiologist reading the images to become proficient with digital.

"A two-unit facility with aging equipment should not trade these systems for a single digital unit," Kolb cautions. It is important to have screening mammograms and diagnostic mammograms scheduled for different rooms to achieve maximum workflow productivity.

Equipment vendors of computed radiography (CR), computer-aided detection (CAD), and FFDM believe that Food and Drug Administration (FDA) clearance computed radiography mammography systems will serve as a catalyst to adoption.

Andrew Vandergrift, national program manager of women's health imaging systems at Fujifilm Medical Systems USA, expressed hope that the DMIST results would enhance the validity of CR mammography to the FDA. Twenty percent of the digital mammograms performed during the trial used the equipment currently being evaluated by the FDA, Vandergrift said.

Kolb and other industry experts believe that the lower cost for acquisition of CR mammography may be an incentive to convert to this technology, even by radiology departments and imaging centers with low volume. Digital transmission of images in off hours will be more secure than sending films by courier, and the practice of outsourcing to experts may increase.

"Mammographers are like hen's teeth," Schroeder said. "You just can't find them."

Based on his observations, Kolb said that a highly experienced mammographer doing digital reporting and interpreting digital images could easily read 100 screening studies in less than two hours. If reimbursement is $40 per exam, an expert mammographer can generate between $2,000 to $2,400 per hour.

"Everybody wins in this scenario," Kolb said. "The radiologist makes a good dollar. The community hospital provides its patients with a level of expertise in interpretation that its own staff does not have, and the patient has greater confidence in her diagnosis."

By Cynthia Keen contributing writer
October 24, 2005

Related Reading

Harnessing technology, training to make the most of FFDM, October 3, 2005

DMIST study: Younger women may benefit most from digital mammo, September 16, 2005

Copyright © 2005

Page 1 of 570
Next Page