Digital Mammography Workflow: Lessons Learned at The Elizabeth Wende Breast Clinic

The Elizabeth Wende Breast Clinic (EWBC) in Rochester, N.Y., was an early adopter of full field digital mammography (FFDM), CAD, breast imaging workstations and PACS to manage its comprehensive breast care practice that images 350 women every day. The facility has fine-tuned its workflow through trial and error - and is leading the call for better connectivity solutions to seek to unite valuable diagnostic information from disparate vendors' FFDM systems, workstations and PACS - to enable better patient care.

The challenges of mammography have been well documented. The combined stressors produced by insufficient reimbursement, ever-increasing workloads, innate flaws in the means of detection and the precision required in reading these critically important exams increase the intensity of these endeavors. As recent studies have indicated, breast cancer deaths have declined thanks to earlier detection enabled by excellence in breast imaging.

Yet while earlier and better detection of breast cancer is, for sure, saving lives, many breast imaging centers are "killing" themselves in integrating all of the current elements of the analog and digital imaging world to continue to improve breast cancer detection and advance care.

When they entered the realm of digital mammography in November 2002, little did the staff of The Elizabeth Wende Breast Clinic anticipate how dramatic a shift would occur in their work patterns or how demanding the problems with connectivity between systems would become. Three years after the installation of their first digital mammography unit, they have realized that their experience is invaluable to share in hopes of reducing the headaches for colleagues who are following in their digital footsteps. As leaders in this field, they are dedicated to participating with vendors, governmental agencies such as the FDA, professional organizations and fellow clinicians to improve and enhance the technical aspects of digital mammography.

EWBC was founded in 1976 as the nation's first free-standing mammography and breast imaging center devoted to the detection of breast disease. Typically this busy radiology practice performs 350 mammographic studies per day, including both screening and diagnostic exams. The clinical patient work pattern involves "double reads" by two physicians for each exam coupled with computer-assisted detection (CAD) software. Some 70 percent of their patients prefer to wait in the Clinic for the results of their exams which can include ultrasound, additional mammographic views and core biopsies that are performed during the same appointment rather than requiring patients to return at a later date. This approach is designed to reduce their patients' apprehension that comes from waiting for additional testing to be performed.

Stamatia Destounis, MD, staff radiologist at EWBC and a clinical associate professor at the University of Rochester Medical Center, explains that for patients who are being worked up for an abnormal mammogram, the physician will follow that patient throughout the process. Therefore, as a physician, she doesn't necessarily sit all day in a darkened room to read images in batches. She may go to do an ultrasound scan and a core biopsy and then return to read other mammograms.

Besides the issues produced by their patient-centric work pattern, EWBC radiologists straddle analog and digital worlds with some full-screen analog and digital imaging units. They are attempting to maximize the number of patients who are scanned on the digital units, and they don't plan to purchase more analog machines. This situation means that with any given patient, they may need to review both films and soft-copy versions of newer images. Comparison may be hampered because the image size may be different, the level of image contrast does not match and there can be annotation challenges that make it more difficult for physicians to communicate needed additional views to the technologist.

Destounis says that it takes longer to read a digital image than film screen because there are more ways to manipulate the image data produced by FFDM. Digital images can be windowed and leveled, zoomed in on, and physicians can examine the images in several different ways. She relates that all the radiologists experienced a learning curve and that they have increased their efficiency as they became more comfortable with the technology.

"In all honesty, I have to say that digital mammography has helped us see through dense breast tissue like nothing else," Destounis says. "I'm impressed every day when I see microcalcifications and things that were very difficult to see on regular film screen."

The Clinic participated in the American College of Radiology Imaging Network (ACRIN) multicenter Digital Mammographic Screening Trial (DMIST) using two digital units, each from a different vendor. In June 2003, they installed a third unit from a third vendor.

Overall, they've experienced formidable challenges as early adopters of digital mammography, and their success to date has been a result of a team effort.

EWBC radiologist Margarita Zuley, MD, reports that the most frequent question she receives from colleagues relates to their reasoning in going with three different FFDM vendors. She explains that just as with analog machines that provide different clinical information and prove more advantageous for certain patient profiles, the same is true with digital systems. With a large patient pool and a variety of imaging needs, having the flexibility of more than one vendor's system has proven valuable to their diagnostic approach.


Entering the digital era

Theresa Wade, MPA, ACMPE, who serves as administrator of EWBC, suggests that most radiology offices have no idea of the problems they will encounter when they decide to install digital mammography equipment. She describes a cultural change in the way work is completed for physicians, technologists and all other staff involved. From an administrative standpoint, new staff was required - including a PACS administrator, a person to back-up that position and a coordinator for the technologists to improve their workflow.

"When you consider bringing digital into your office, you need a clear vision as to your plans, and you must communicate that well from the top down," says Wade. She believes that all employees truly must understand the goals of the new approach to reduce resistance to the change. "The payoff won't be right away with digital. In fact, you'll be less efficient for awhile."

Judith LaBella, RT(R)(M), lead technologist for digital imaging, describes the change in their clinical work patterns. With analog imaging, their typical pattern was for the technologist to select the patient's chart, review it against prior images, find an imaging room with a machine that was not in use, call the patient, perform the exam, process the film, check it and place it for reading with the previous study. Now that they have an R2 ImageChecker CAD, the films are processed through that system.

Using FFDM, technologists remain in the room with the imager, and a digital coordinator selects the chart to look for patients who have had previous digital images or those with dense breast tissue or who have calcium deposits, or those who have had a previous diagnosis of DCIS (ductal carcinoma in situ), and she selects digital machines for them as a priority. Once the images are captured, the technologist places the chart outside the room and picks up the next chart for the following patient. The difference in clinical work pattern is that the technologists remain in the imaging room, and the coordinator does all of the "running."

One of the challenges that arises with digital systems is bilateral communication between the technologists and physicians who are reading the studies. Currently, they use a paper-based method, but when they go to a paperless workflow, they anticipate new challenges.


Basic connectivity issues

Zuley explains the history of connectivity problems involved FDA requirements that yielded "silo systems" in a "soup to nuts" configuration where all portions of the imaging chain were self-contained. The unfortunate outcome is that one vendor's acquisition device is incapable of communicating with a system or workstation of another. Given the mobility of our society, that means that they are unable to take their images to another center. Another ramification is that if a patient has a scan performed on Vendor A's system, either she must have scans performed on that machine for the foreseeable future, or they must print a digital image to be able to display it on the alternator close to the workstation for Vendor B's scanner. And Zuley notes that hard-copy images do not provide the same information as soft-copy reads.

"The long and short of it is that radiologists must choose one of two paradigms," Zuley says. Either they have to purchase a vendor specific solution which includes acquisition unit, diagnostic workstation and a mini-PACS to archive mammography. This is a separate PACS from a pre-existing PACS, which is an expensive solution. The second option is to follow an open architecture solution. The downside of that approach is the DICOM incompatibilities. The upside of that approach is that the radiologist can use all of the functionality of their pre-existing PACS and can read off a single workstation for all modalities.

Because they have found that different systems are capable of providing the best images in certain cases, they believe that the first option is untenable. Having the flexibility of a multi-vendor approach provides the best possible patient care to her way of thinking.

While they currently use a GE Senographe, Lorad Selenia by Hologic and a Fischer SenoScan FFDM unit, and plan to install a Siemens FFDM system in the near future with their Sectra Confirma workstation for their PACS, Zuley stresses that they appreciate the capabilities of all of these systems equally to complete their critical tasks. To this end, she has become quite involved in new initiatives that are expected to improve connectivity.


Making connections

The workflow challenges created by these disconnects in system and vendor connectivity have made EWBC a driver for change. At the June Society for Computer Applications in Radiology meeting in Orlando, Zuley moderated a session designed to raise awareness to connectivity issues among vendors, regulators and breast imaging centers. The group has evolved into a multidisciplinary IHE (Integrating the Healthcare Enterprise) Digital Mammography Technical Working Group that was scheduled to meet at the recent RSNA meeting. Comprised of practicing radiologists, members from the mammography system manufacturing sector, the FDA, professional organizations such as the American College of Radiology and experts from the IT community, the group is  charged with producing a technical white paper that provides detailed technical parameters to be included in IHE mammography profiles.

David Channin and David Clunie who have written most of the DICOM standards will edit the white paper. The draft technical document will provide a roadmap for vendors to follow. Zuley applauds this collaborative effort that should yield standards that will address incompatibility issues.

Finally, there is a significant problem caused within the context of FDA regulation of mammography that is not found in other imaging modality segments.

"Because these [FFDM systems] are PMA [pre-market approval] products, they are required by law that any changes they make must become a PMA supplement," explains Zuley. This prerequisite means that vendors must conduct large clinical studies, and pay thousands of dollars in fees to make modifications to improve their products. The FDA is participating in the working group and clinicians and vendors are hopeful that the FDA will change their process so that future upgrades will require 510(k) approval only. Zuley notes that in designing the technical working group, they sought members purposefully from all sectors because they considered that approach would be critical to their ultimate success.


Conclusion

At the SCAR Digital Breast Imaging Forum in June, Zuley observed that in her experience she believes that all professionals who are involved in the production, regulation and clinical use of digital mammography are truly interested in full development this technology. At the end of the day, high-quality patient care serves as the focus for every professional's interest and efforts to improve the current state of affairs with the connectivity issues that occur in clinical practice. Rising to the challenges assumes critical importance in the diagnosis and disease management for patients with breast cancer. While great strides have already been made, there is much more to do. And the EWBC crew believes that dedicated professionals will ultimately solve the issues that are so perplexing today. Connectivity will be a reality - that will ultimately connect breast cancer patients with better diagnosis and care.