The fully digital breast center holds great promise for streamlining care, but can its inherent efficiencies be captured without detracting from the quality of the patient experience? In a search for answers to these and other questions, the author traveled to a filmless breast center in Sweden that utilizes both computed radiography mammography (CRM) and digital radiography mammography (DRM), to observe, question, and compare the operations of that breast center with counterpart US facilities.
Most medical personnel who are familiar with mammography are aware of the work of Laszlo Tabar, MD, and others, studying mammography in large Swedish screening populations over the past 25 years, and demonstrating that early detection of and intervention for breast cancer not only prolongs life, it saves lives. 1,3 What is less well known is the economic struggle that Sweden has gone through to maintain its breast cancer screening program over the past decade and a half. A complete discussion of Sweden’s economic challenges is well beyond the scope of this article; however, the relative impact may be illustrated by the fact that Sweden’s ranking for expenditure on health care as a percentage of GDP has gone from number two in the world (behind only the United States) at 13.6% early in the 1990s, to number 24 in the world at 8.7% in 2000 (compared with 13.9% for the United States). 4
Achieving this kind of reduction in the face of rapidly escalating drug and new technology costs within the framework of a national health system would normally require severe rationing of health care, as well as substantial cutbacks in screening and other preventative health care programs. Sweden took a different tack. Rather than reduce the amount of health care delivered by the system, Sweden’s hospitals, which direct or deliver most health care in the nation, have been reengineering how they deliver care, to optimize the use of all human and technology resources in the health care system. While this has not been an entirely painless process, the overarching goal of improving the delivery economics without compromising the quality of patient care, or the patient care experience, is being achieved.
Reengineering, a word that was overutilized in US health care a decade ago, has become a fact of Swedish health care. Each process was evaluated to determine what was important, and how each component could be optimized. The fundamental outcome of the reengineering process that affects the breast center is, surprisingly, not the use of digital mammography; rather it is the almost total elimination of the conventional patient file from the process of rendering care. This would not be possible without digital mammography; however, the creation, electronic transmission, manipulation, and storage of the digital image are only part of the story. Taken together, the digital image and its corresponding electronic medical record, however, create an extremely efficient delivery mechanism that is translatable to the United States with very little in the way of modification.
THE HELSINGBORG EXPERIENCE
With CRM on the horizon, and to learn about how digital mammography really works in a clinical breast center, with CRM and DRM operating side by side, the author traveled to a hospital in Helsingborg, Sweden. Helsingborg is a seaside community with a population of approximately 120,000, located in southern Sweden across the strait from Denmark, and about an hour by train from the Copenhagen airport. In the past, Helsingborg’s economy was based on fishing and ocean commerce, but now it has become more of a seacoast resort and regional information technology center.
In all of Sweden, the delivery of mammography is centralized in major hospitals that serve the Swedish counties, which might be considered somewhat analogous to American states. At the Helsingborgs Lasarett (hospital), approximately 14,000 women are screened each year, with an additional 3,500 diagnostic mammograms and 1,000 ultrasound examinations being performed annually. It should be noted that the diagnostic mammograms and ultrasound examinations include patients who are referred for workup from the other mammography facility in the county.
The Helsingborg breast center is located within the hospital, a short walk from the front entrance, past the hair salon. While it did not have the lush decoration of many US breast centers, the center was a comfortable, dedicated facility, and included breast surgery as well as breast radiology. On the radiology side, the center was equipped with a direct digital mammography unit for screening mammography; a CR system is utilized for diagnostic mammography and also functions as an upright stereotactic unit for core biopsy and fine needle aspiration biopsy. A dedicated ultrasound unit is also utilized for diagnostic ultrasound and biopsies.
For interpretation of all breast images, two multi-modality workstations are utilized and are served by a breast imaging mini-PACS. The mini-PACS is networked to the hospital’s enterprise PACS, allowing other patient images to be reviewed at the breast center (eg, CT, MRI), as well as providing long-term archival storage for digital breast images.
Without considering how screening mammography patients are scheduled, which is a function of the national health system, the process of care in Sweden is quite similar to that in US breast centers. Patients report to the center, register, sit in a sub-waiting area, receive the procedure or procedures that they were scheduled for, and are released. If the patient has come for a diagnostic mammogram, it is reviewed prior to her leaving the center, and, if she requires any further studies or even a biopsy, she is offered that study or procedure before she leaves the center.
It is how information is created and moved within and through this process that distinguishes the Helsingborg system from its US counterparts. Beyond the fact that all mammograms are digital, the entire information process is automated. When patients register, their histories are updated in the mammography information system. The patient is given a sheet of paper with a bar code and name and patient ID on it. This “ticket” is the only paper generated up until the report is printed. The patient then proceeds back to the sub-waiting area adjacent to the mammography rooms. There are changing rooms available; however, these are used only for removal and storage of outer garments. Patients are called to the mammography rooms by appointment order. When the patient enters the room, she undresses to the waist, as the technologist uses the bar code on the patient’s “ticket” to verify the proper patient identification in the daily worklist and to bring up the patient’s history on the computer screen. The technologist verifies certain information (eg, confirms lack of a clinical finding) with the patient while the patient is preparing for her mammogram, then moves the patient to the mammography unit for positioning. It is important at this point to note that with digital mammography, exposure, while important, is less important than with film screen, because of the availability of contrast and leveling and other tools for use by the radiologist to adjust the view of the mammogram in the interpretive process. For screening, using the DRM unit, mammograms were observed as taking 4 to 6 minutes from the time the patient entered the mammography room until she left the room. This time included the time required by the patient to undress to the waist and redress.
Diagnostic mammography patients are handled in much the same manner, except that the time allocated for the patient to be in the mammography room is 20 minutes. The length of the scheduled time is not determined by the technologya diagnostic series can typically be acquired in well under 10 minutesrather these examinations are often complex, with the radiologist adding views after reviewing the initial diagnostic images. It is important to note, however, that the diagnostic images are moved electronically to the physician’s workstation, and that the technologist is able to stay in the mammography suite with the patient. Relieving the technologist of the need to physically develop the mammograms and then take them to the radiologist for her review, waiting until the radiologist either releases the patient or orders more views or an ultrasound examination, greatly speeds the process of delivering the diagnostic mammogram.
Further explanation of this process is considered to be important because diagnostic mammograms typically take 30 to 45 minutes when delivered with film screen technology. Diagnostic mammograms are performed on the CRM system in the conventional manner, but using imaging plates rather than conventional film cassettes. Once the plates have been “exposed,” the technologist inserts them into the four mammography slots in the imaging plate reader, whereupon the processing time is approximately 70 seconds from the time that an individual imaging plate is placed in a slot. (It should be noted that the image processor that was observed was also capable of processing all standard CR images, up to and including 11 x 14.) The total processing time for four views was less than 90 seconds. As soon as an image is processed, it is available to the technologist on her workstation, and she can immediately send it on to the physician’s interpreting workstation. The physician has the priors available and can quickly review the new digital images to render an interpretation and provide real-time guidance to the technologist for further studies, or to release the patient. Note that the processing time, while less than the time required to process film, has often been used as an argument to depreciate the value of CRM as compared with DRM. In an operational context, it is clear that CRM’s effectiveness is not dependent upon processing time; rather it is determined by the speed with which the images can be put in front of the interpreting physician for her review.
While CRM is utilized in Helsingborg for diagnostic mammography, there is no reason why it could not be utilized for screening mammography as well. In fact, it was utilized exclusively before the MDM was acquired. Throughput times were longer because an image processor with only a single plate capacity was used, but with the upgrade to the current four-slot unit, throughput for mammography is optimized. Throughput times would have to be increased slightly over DRM to accommodate processing times; however, an 8- to 10-minute screening mammogram (six to seven mammograms per hour) would easily be achievableif facilitated by good information management.
To round out the Helsingborg breast center, diagnostic ultrasound, fine needle aspiration biopsy, and core biopsy are offered on the radiology “side.” Patients requiring an ultrasound examination or a biopsy receive them immediately following the diagnostic mammogram. The intent is to provide each diagnostic patient with a diagnosis on the day of her visit for the diagnostic mammogram. In fact, the patient’s appointment is for “diagnosis” rather than for a specific procedure. Only those patients who are unsuited for percutaneous biopsy or who may have an uncertain diagnosis (eg, radial scar) will go on to a surgical biopsy, and the Helsingborg program reports that well in excess of 90% of patients who go to the operating room have a preoperative diagnosis of breast cancer. All breast cancers are reviewed in a twice-per-week breast conference, attended by all physicians in the diagnostic and treatment specialties, as well as by the technologists and nurses in the program.
THE RADIOLOGIST PERSPECTIVE
Gerald R. Kolb, JD |
From the interpreting physician’s perspective, the Helsingborg breast center is not entirely film-free. Digital mammography is relatively new, and an archive of prior digital mammograms has not yet been developed for comparison with the current studies. The breast radiologists initiated digital mammography using standard rotating viewers on which to view the prior film studies, but found that the extraneous light was just too overwhelming in its reflected effect upon the soft-copy workstations. Current practice is to use an 11- x 14-inch dental image light box for the review of prior films. This viewbox is angled and placed beside the workstation. Reviewing prior images in this manner is cumbersome, as the physician must remove the patient’s jacket from the cart, then remove the patient’s prior images from the jacket, place them on the viewbox, review the images, and return them to the jacket. Observed times for this process were less than expected, probably because the jackets were placed on the cart in the same order that the patients appeared on the workstation worklist. In any event, while the radiologists reported that they would like to be able to dispense with film entirely, they were able to move through the interpretive process quickly, even with the film-hanging requirement.
CONCLUSION
So, what is the impact of all of this efficiency? The Helsingborg breast center sees approximately 120+ patients per day on the radiology “side.” Staffing of the center includes two receptionist clerks, four technologists, and 1.3 FTE breast radiologists. All screening patients also receive a clinical breast examination (performed by a technologist), and all screening mammograms receive a blinded double reading. From the author’s experience, even an efficiently run film screen breast center with these volumes would require four to six mammography units, compared with two units at Helsingborg, and six technologistswithout clinical breast examinations. Two breast radiologists would be the standard professional complement for this volume.
In the context of US practice, there is always a tendency to approach care that is delivered in another country, especially a country with a national health system, as being “different” and not reproducible in the United States. The danger in this approach is to ignore the workflow and efficiency lessons that these systems have had to learn the hard waythere is no additional reimbursement in Sweden for digital mammography. Efficiency does not mean lining patients up in the hallway awaiting their mammograms. It does mean using highly effective technologies to acquire images, and supporting those images with well-integrated information systems that can eliminate the cost, and the delay, of people moving paper and film around.
The author would like to thank Boel Heddson, MD, head of the Mammography Department at Helsingborgs Lasarett, and her staff.
Gerald R. Kolb, JD, is president of Breast Health Management Inc in Bend, Ore, [email protected]. In addition to consulting on workflow issues in breast centers, he consults for several vendors of digital imaging equipment, including computer-aided detection.
References:
- Tabar L, Smith RA, Duffy SW. Update on effects of screening mammography. Lancet. 2002;360(9329):337,339-40.
- Duffy SG, Tabar L, Chen H-H, et al. The impact of organized mammography service screening on breast carcinoma mortality in seven Swedish counties: a collaborative evaluation. Cancer. 2002;95:458-469.
- Tabar L, Fagerberg G, Duffy SG, et al. Update of the Swedish two county program of mammographic screening for breast cancer. Radiol Clin North Am. 1992;23:187-210.
- World Health Organization, World Health Report, Table 5, Selected national health accounts indicators: measured levels of expenditure on health 1997-2001. Available at: www.who.int. Accessed May 19, 2004.