Digital mammography:
Cost and workflow issues

Gillian M. Newstead, MD, FACR


Dr. Newstead is a Professor of Radiology and the Clinical Director of Breast Imaging at the University of Chicago, Chicago, IL.

More than 276,000 new cases of breast cancer will be diagnosed this year.1 Of those, roughly 212,000 will be invasive cancers and 62,000 will be ductal carcinoma in situ (DCIS). Approximately 41,000 American women and 460 men die of breast cancer annually. At present, there are >2 million U.S. women living with a diagnosis of breast cancer.

In recent years, the death rate for breast cancer has steadily decreased, although the incidence has continued to increase. This trend toward earlier diagnosis and increased survival is often attributed to the effectiveness of screening mammography programs.2-4

Each year, approximately 1 million additional women enter the screening age group. In 1999, approximately 42,000 mammograms were performed in the United States. The following year, 44.5 million mammograms were done, and by 2004, that number jumped to 49 mil-lion.2-4 As the baby boom generation continues to age, this number will only increase. It is estimated that there were approximately 70 million women over the age of 40 in the United States in 2005 and that, by 2025, the number will increase to >88 million.5

The cost for screening all of these women can be significant. Approximately 49 million women will undergo screening mammography this year. Of those 49 million, roughly 10% will be called back for additional workup, such as diagnostic mammography, ultrasound, or magnetic resonance (MR) imaging. Of the women who go on to diagnosis, many of them (65%) will be found to be cancer free. They will all have undergone diagnostic mammography, some will have ultrasound, and a small percentage will have MRI. Approximately 36 million of the 49 million women screened will be truly healthy, but there will still be approximately 40,000 cancers not detected at mammography. The direct cost of screening alone is estimated to be $2 billion a year, with the follow-up examinations adding another $1 billion.

As the demand for breast imaging services continues to increase, the resources to provide these services continues to decrease. A 1995 survey found that only 3% of all radiologists actually specialized in breast imaging.4,6 In addition, there is a shortage of breast imaging fellowships nationally and a decreasing number of applicants to these fellowships, meaning that we are not training enough breast specialists for the future. There is also a shortage of breast imaging technologists in various parts of the country, and burn-out (due to the increasing workload) among the present staff can be a problem. If the average volume of mammograms that are currently read by interpreting physicians were to remain constant, then a 38% increase in the number of radiologists would be needed by the year 2025 in order to meet the demand for screening mammography.

The economics of breast imaging

What is the basic economic situation in the typical breast imaging practice? The American College of Radiology (ACR) conducted a survey that studied what it costs to perform a mammogram either in a hospital or an outpatient setting, not including the physician’s fee for interpretation. For the hospital practice, they found that the actual cost to perform a screening mammogram was $93.98. With typical reimbursement of approximately $80, it is clear that the hospital lost money for each screening exam performed.

In the outpatient setting, where the costs can be more controlled, the survey found the cost per mammogram to be approximately $59.00. Nonetheless, it is clear that there are no large profits to be made from screening mammography.

Another economic study analyzed professional income at 7 geographically representative academic programs in the United States.7 For this study, the researcher took the revenue and subtracted from it the direct costs (salaries of the physicians, fellows, malpractice insurance, and secretarial and other assistance) and indirect expenses (such as billing expenses) to determine mammography’s contribution margin. The study found that in all practices mammography had a negative profit margin. The loss for physician full-time equivalent (FTE) was variable between the practices studied but was between $50,000 and $100,000 per year. The author also found that diagnostic mammography was the driver of this loss because of the increased physician time and personal involvement in such cases. He concluded that a 3-fold increase in reimbursement for diagnostic mammography was needed just for breast imaging practices to break even.

Making mammography more efficient

As part of the previously mentioned study, an activity-based cost analysis was performed using a time-motion study.7 In this part of the study, the radiologist was followed by an observer with a stopwatch who would clock exactly how much time the physician spent doing each task. The researchers calculated that for each screening examination, a total of 5 minutes of physician time was involved for all activities, including not just diagnostic reading but quality assurance (QA), comparisons to prior images, and all other activities performed by the radiologist. The study also found that diagnostic mammography and ultrasound each required 25 minutes of physician time, as did consultations. Interventional procedures took roughly 60 minutes.

Compliance with unfunded federal mandates can also affect a practice’s bottom line. One study at a large East Coast practice found that it cost $7.82 per patient to comply with federal mandates such as MQSA regulations, etc. (Destouet JM, personal communication). This can be a significant burden, particularly when the reimbursement rate is only approximately $80.

Digital mammography

Will converting to digital mammography help the bottom line? I believe that whether it helps or not, we are going to have to do it. It is essential—especially for practices that are based in multimodality, multisite facilities—for mammography to move into the digital arena. Therefore, the question really is how we can convert to digital and be cost-effective?

Digital mammography can be beneficial to routine practice. It provides for the rapid review of images, it decreases the time required for the technologists to acquire and process the images, and it allows for immediate repositioning and image retakes when necessary. The real-time image display capabilities of digital mammography also increase patient throughput by eliminating film processing and its associated costs, and digital technology expedites all work-ups and interventional procedures.

The teleradiology applications of digital mammography can also be beneficial. Centralization of screening interpretations can address both cost and personnel shortage issues. Second-opinion services and conferences, as well as computer-aided detection (CAD) services, can also be facilitated using digital technology. Given the shortage of radiologists, it would be very helpful if screening images were acquired at remote sites and then electronically sent to a central location for reading by a radiologist who specializes in breast imaging. It is possible that diagnostic examinations could be performed remotely as well, with the radiologist at the interpretation workstation site and the technologist at the location where the image is being acquired. With real-time synchronization between the technologist and the radiologist, diagnostic procedures could, theoretically, be performed remotely.

What about the physician interpretation time? At New York University, we investigated that when we first began using digital mammography roughly 5 years ago.8 We evaluated the time spent by both the physician and the technologist when performing a digital mammography examination compared with the time spent on an analog study. The technologist imaged 100 patients with a 2-view digital screening exam. An additional 100 patients underwent a standard 2-view analog exam. A stopwatch was used to monitor the time spent by the technologist performing the examination, interacting with the patient, and checking the images. The physician time for viewing and decision making was recorded with the stopwatch, but not the reporting time.

We found that the technologist time was reduced from 10 minutes, 29 seconds for an analog examination to 6 minutes, 12 seconds for a digital examination, saving more than 4 minutes of technologist time per patient. The physicians, however, took longer with the digital images, going from 25.8 seconds per analog study to 76 seconds per digital study. One of the reasons for this increase in reading time was the need to compare the new digital images with the prior Şlm mammograms and the need for panning and other image manipulations with the digital technology.

One of the concerns with the conversion to digital technology is the large initial expenditure required to make the switch. It is estimated that it takes 5 years to net the present value of the costs associated with converting to digital.

As for reimbursement, private insurers vary in adopting Medicare reimbursement rates for mammography. There are still some states in which getting paid for digital mammography is a problem and others in which the reimbursement for digital is higher than for analog. It is not uniform across the country. This can affect the bottom line. It has been estimated that full-Şeld digital mammography (FFDM) must have a 50% adoption rate by insurers and be used at 90% capacity in order to be proŞtable

Strategies for survival

As the demand for services increases and the available personnel and resources diminish, radiologists in general and breast imagers in particular are going to have to increase productivity. Teleradiology with image reading performed overseas is helpful, but the bottom line is that radiologists are going to have to work harder and be more efŞcient. If we are unable to keep up with the imaging demands of our patients, we risk losing some of the market to nonradiologists.

Increased productivity will not work, however, if a practiceÕs proŞt margin is negative. One way to help address this issue is to separate screening and diagnostic cost centers. Online evaluations and procedure scheduling should all be part of the diagnostic arena. Fast interpretation of screening examinations is helpful, and automated reporting and reduced paperwork are both very important.

It is also important that the technologists, many of whom are already assuming a significant responsibility in helping us care for our patients, have a pathway for increased responsibility. It is certainly possible for technologists who wish to proceed up the ladder to perform tasks that are currently performed by physicians and to aid physicians in conducting diagnostic workups. The specifics of how this can be done are beyond the scope of this article, but I certainly think that it is important for the breast imaging community to retain motivated, excellent technologists.

The technologist must be able to focus on patient care. The presence of a technologist’s aide who can perform such duties as greeting the patient and escorting her to the changing area can be very helpful. Technologists should not be spending valuable time assembling the charts, processing and hanging films, setting up rooms, and folding laundry. These are all things that can be done by a technologist’s aide, leaving the technologist free to focus on patient care and the imaging aspect of the process only.

Information integration

If breast imaging is going to survive and be economically viable, it must become Şlmless and paperless. Seamless integration is an important concern when attempting to achieve this goal. Currently, we all have separate systems; we have multiple information systems from multiple vendors. We do not have one ÒwholeÓ that is integrated and organized. We need to drive the standards toward problem solving and toward integration proŞles in radiology and across the healthcare institutions.

The Integrating the Healthcare Enterprise (IHE) initiative is working toward the goal of uniform “plug-and-play” operations for all clinical information systems from all vendors and across all modalities. The purpose of this multiyear, multistakeholder, inter-operability initiative is to move beyond simple Digital Imaging and Communications in Medicine (DICOM) standards and harness all existing healthcare and information technology (IT) standards to produce technical recipes, called “integration profiles” for how these standards should be implemented. To say a system is DICOM-compatible is only a first step; it does not necessarily mean that all of the information is being integrated into a form that the radiologist can use, thus becoming more efficient. That is the purpose of IHE.

What can an individual radiology department do to improve information workflow? First, it’s important to identify the core functions of each system within the department. Then determine what is needed to communicate between these different systems. How does the mammography workstation talk to the MR workstation? What images are on the MR report and how can they be pulled up and viewed at the workstation?

For efficient workflow, we need efficient modality acquisition and a scheduled workflow integration profile. Most departments have a radiology information system (RIS), a hospital information system (HIS), a picture archiving communication system (PACS), and more, but these systems don’t always communicate with each other efficiently. In addition to managing and archiving images, we need to be able to efficiently and accurately move all of the relevant data (including patient registration information, order placement, and filing) throughout the entire enterprise.

Issues regarding CAD results and information storage still need to be addressed as well. Should CAD marks be stored? Should the modality automatically push the CAD results to the workstation? Should both the “for processing” and “for presentation” image data be archived? Are there liability issues that need to be considered with regard to information storage? These issues still need to be addressed as mammography services migrate to digital technology.

The key to success in all of these issues lies in the design of the workstation. Workstations must become vendor-neutral and multimodality-capable. Computer-aided detection must also become an essential part of any workstation. As we deal with ever increasing data-sets from mammographic, ultrasound, and MR images, CAD will be a vital part of this interpretation.

Patients expect access to high-quality mammography; they want to be educated concerning screening and abnormal findings. They also want the latest equipment. They want digital mammography, and they want high-tech MR studies. Coordination of all imaging studies must be a priority for all breast imagers.

Conclusion

We are at the crossroads in the transition from analog to digital breast imaging. Ultimately, a fully integrated digital breast imaging center will allow improved efficiency of operation and enhanced imaging techniques. The rest of radiology has moved to a digital environment, and we must as well. Negotiating the jungle of electronic information in order to achieve an integrated workflow and viewing environment is a major challenge. The mammography and MR data sets are large. A standard MR examination may provide >2500 images, and transmission of these large data sets can be problematic. Development of a single, integrated workstation, importing disparate information from various modalities—mammography, ultrasound, and MR—is a high priority. A multivendor, multimodality viewing environment, with adequate monitor space, and a fully integrated computer-aided multimodality detection and diagnostic system will allow the breast radiologist to function efficiently in the new digital environment. Let’s hope that we do not have to wait too long!

REFERENCES

  1. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006.CA Cancer J Clin. 2006;56:106-130.
  2. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics. Health United States, 2005. Available at http://www.cdc.gov/nchs/hus.htm. Accessed May 19, 2006.
  3. American Cancer Society. Breast Cancer Facts & Figures 2005-2006. Available online at: http://www. cancer.org/docroot/MED/content/MED_2_1x_Breast_Cancer_Facts__Figures_2005-2006.asp. Accessed May 19, 2006.
  4. Wing P.IOM Mammography Projects and Related Background Information.Rensellaer, NY: State University of NY, School of Public Health, Center for Health Workforce Studies. 2005.
  5. US Census Bureau. U.S. Interim Projections by Age, Sex, Race, and Hispanic Origin. Available at http://www.census.gov/prod/2004pubs/04statab/health.pdf. Accessed May 19, 2006.
  6. Sunshine JH, Maynard CD, Paros J, Forman HP. Update on the diagnostic radiologist shortage. AJR Am J Roentgenol. 2004;182:301-305.
  7. Enzmann DR, Anglada PM, Haviley C, Venta LA. Providing professional mammography services: Financial analysis. Radiology. 2001;219:467-473.
  8. Newstead GM, Schmidt RA. Integration of Digital Mammography into Clinical Practice: Initial Experience. Presented at the 6th International Workshop on Digital Mammography. Bremen, Germany; June 22-25, 2002.