Diagnostic ultrasound is ideal for many patients. No other imaging option gives them a real-time look at their own anatomy or so openly encourages discussion with the radiologist about their symptoms and the evidence of disease on the monitor.
Diagnostic ultrasound is ideal for many patients. No other imaging option gives them a real-time look at their own anatomy or so openly encourages discussion with the radiologist about their symptoms and the evidence of disease on the monitor. Ultrasound is safe, patient-friendly, and inexpensive. But while patients welcome it, many radiologists in the U.S. want it to go away.
This contradiction in the application of diagnostic ultrasound was the subject of an October 2006 consensus conference of the Society of Radiologists in Ultrasound (SRU) that brought ultrasound's leadership together in San Francisco. Two days of lectures and debate led to a strategy statement defining a plan to maintain radiology's leadership over the modality (see related article).
U.S. radiologists are not so much turning away from ultrasound, as they are turning toward MRI and CT, said Dr. Harvey Neiman, executive director of the American College of Radiology. Facing an unprecedented demand for their services, they increasingly choose these two modalities because of their superior speed, convenience, and profitability.
Ultrasound remains immensely popular elsewhere in the world, where it fits the needs of countries less willing or able to spend as much as residents of the U.S. on healthcare. In the U.S., radiology is embracing its tradition of darkroom image interpretation with the radiologist sitting at a PACS workstation reading CT or MR studies performed by a technologist, said Dr. Beverly G. Coleman, SRU president. This efficient arrangement is possible because the two modalities are performed according to well-established protocols.
"You put the patient on the table and zip them through. If it is a renal stone, you do this, this, and that, and you are done. Patients rarely see a physician face to face unless it is an interventional procedure," Coleman said.
Ultrasound, however, is not a spectator sport. It requires hands-on involvement to confirm the sonographer's findings and to survey the region of interest for disease that may have been missed, according to Dr. John J. Cronan, a professor of radiology at Brown University School of Medicine.
"You can't just sit there and look at the images without talking to the patient and still do good ultrasound," he said.
Reimbursement rates do not compensate radiologists for their involvement with the procedures. They can earn $700 by reading an abdominal CT or MR, while fees for reading an ultrasound study for a patient with identical symptoms may earn them $125, Cronan said.
A well-known teaching hospital charges $749 globally for abdominal ultrasound compared with $2186 for contrast-enhanced CT and $3177 for MRI. For those services, the radiologist's professional fee is $245 for the ultrasound, $358 for CT, and $629 for MRI.
Radiologists are responding to ultrasound's financial disincentives by doing their best to ignore the modality, Cronan said. This attitude is reflected in the stacks of ultrasound studies, studded with sonographers' notes, that routinely pile up during the day. They are the studies read last, and they often are directed to the radiologist with the least departmental seniority.
"That is the state of ultrasound in a large portion of the community," Cronan said.
The lack of innovation in clinical applications is making a bad situation worse. U.S. radiologists blame the FDA for long delays in the regulatory approval of ultrasound contrast media, which is important for applications development. None of at least nine microbubble contrast agents submitted for an FDA new drug application in the early 1990s has been approved for clinical use, despite widespread adoption of these agents after regulatory clearance in Europe, Canada, and, most recently, Japan.
After numerous rebuffs, no pharmaceutical company is actively seeking approval for an ultrasound contrast medium in the U.S. The FDA approved two agents for cardiac applications in the mid-1990s, but inability to gain reimbursement for applications outside the heart has discouraged radiologists from their off-label use.
FDA resistance has revolved around safety concerns stemming from several deaths linked to ultrasound contrast media administration in Europe, said Dr. Lennard D. Greenbaum, president of the American Institute of Ultrasound in Medicine. The European Union affirmed the agent's safety and attributed the fatalities to the patients' underlying medical conditions.
The FDA has also been concerned about the relative efficacy of microbubble-enhanced ultrasound compared with other modalities, said Dr. Barry Goldberg, director of ultrasound at Thomas Jefferson University Medical School and the Jefferson Ultrasound Research and Education Institute. For the detection of liver lesions, a high priority for advocates of contrast-enhanced ultrasound, the FDA required applicants to use contrast-enhanced CT as the gold standard in a clinical trial. This requirement created reproducibility problems when coronal CT cross-sections were compared with angled planes acquired between the ribs with ultrasound. Lesions in the ultrasound views appeared in different positions than in the corresponding CT images, leading FDA officials to question whether ultrasound correctly characterized the size and location of the tumors.
After years of inaction, the AIUM joined Bracco Diagnostics, Bristol-Myers Squibb, and GE Amersham in 2005 in seeking a way to break through the logjam. At the invitation of Dr. George Mills, director of the division of medical imaging at the FDA, AIUM representatives proposed in March 2006 a framework for a model clinical trial protocol. More progress was made at a November 2006 meeting. Supporters worry, however, about additional delays following Mills' departure from the agency in January.
"That has to be a setback," Neiman said.
Regulatory problems have left ultrasound as the only major imaging modality practiced in the U.S. without an enhancing pharmaceutical component. Ultrasound research initiatives have been handicapped because the largest market for new services is prohibited from using contrast media, according to consensus panelists. The lack of a contrast agent added to arguments that favor a shift to other clinical modalities.
"If you'd ask a radiologist to do CT or MR without contrast, they'd say you're crazy, but that is exactly what we do with ultrasound," Greenbaum said.
Ultrasound's reputation has suffered among radiology residents, who are turned off by the modality's low reimbursement and practice inefficiencies. The slow pace of technological advancement compared with other modalities adds another disincentive to specialize in the field. Fellows seek out new technology as a gateway to discovery and academic achievement, and such opportunities are lacking with ultrasound.
"In the last year or two, younger people have spoken with their feet," Cronan said. "Residents and people coming out of training and going into academic careers are not pursuing ultrasound because it is considered peripheral."
The problem carries over to resident instruction, where fewer experts in an aging workforce are available to teach ultrasound techniques. Some resident training programs no longer employ instructors who have a primary interest in ultrasound. Obstetrical and gynecological ultrasound has been absorbed into the ob/gyn program in other institutions, he said.
Interest in breast ultrasound remains strong, however. Utilization is growing, with more practitioners accepting it as essential for young women with palpable abnormalities, regardless of their mammographic findings, said Dr. Robert Schmidt, director of breast imaging research at the University of Chicago. About 40% of the patients who received diagnostic mammography at the facility in 2006 also underwent breast ultrasound. The number of applications for breast ultrasound fellowships is increasing.
The Brown University program remains committed to general radiological ultrasound, Cronan said. Its residents receive eight months of ultrasound instruction during four years of training.
Individual programs are afforded considerable flexibility in their ultrasound training, Neiman said. In the 1990s, residents generally received two to three months of in-depth exposure. Now, ultrasound is frequently folded into abdominal or breast imaging, where it is taught with several other modalities.
"I'm afraid that residents today are being exposed to smaller amounts of ultrasound than 20 years ago," he said.
The American Board of Radiology serves as a moderating force against this trend. To earn board certification, residents must demonstrate competence in ultrasound physics and in use of the modality for interventional guidance as well as a comprehensive range of diagnostic applications.
While radiologists are losing interest, miniaturization is spreading ultrasound throughout the healthcare enterprise. Compact scanners weighing less than eight pounds generated $10 million in revenue when first introduced by SonoSite in 1999. Sales rose to $300 million in 2005 and are expected to reach $1 billion by 2010. Nonradiological markets for hand-carried ultrasound were credited for most of the 12% growth in ultrasound utilization in 2005.
The trend is irrepressible, said Dr. Jonathan M. Rubin, a professor of radiology at the University of Michigan.
"The technology, with ever-expanding features and image quality, is destined to become ubiquitous," he said.
SonoSite is attempting to realize that prediction by serving point-of-care markets in anesthesiology, cardiology, endocrinology, emergency medicine, ob/gyn, primary care, and surgery in addition to radiology. Some hospitals are equipped with up to a dozen systems, said David Willis, vice president of competitive development.
A pilot program sponsored by GE Healthcare is exposing first-year medical students to hand-carried ultrasound scanners at the University of South Carolina and Wayne State University in Detroit. GE donated between 30 and 50 Logiq-E laptop scanners to each program. The initiative is based on the theory that ultrasound will become an everyday tool for clinical practice, said Omar Ishrak, CEO of GE Clinical Systems.
"To drive that philosophy, we felt that ultrasound should be embedded in the broad curriculum, not just special courses but topics such as anatomy, physical examinations, and other medical school courses covering each of the four years of the curriculum," he said.
Hand-carried devices are becoming a mainstay of emergency medicine, where they bring rapid access to ultrasound to the bedside in crowded emergency rooms. Many emergency physicians are learning to perform diagnostic ultrasound and ultrasound-guided procedures. Educational opportunities enabling emergency physicians to perform ultrasound have grown from a single ER ultrasound fellowship program in 1998 to more than 20 in 2006, according to Dr. Paul Sierzenski, director of emergency ultrasound medicine at Christiana Health Care Services in Newark, DE.
Emergency physicians are learning to perform ultrasound to improve patient care, he said.
"The crux of emergency medicine is taking care of the sickest of the sick and making sure they don't die," Sierzenski said. "This technology helps them do that, so it is imperative that emergency physicians learn how to use this technology and get credentialed."
Strong evidence-based medical literature covering more than 30 years of scientific inquiry into ultrasound's appropriate application could help radiologists establish the respect among nonradiologists needed to maintain their professional primacy over the field. Speaking at the SRU consensus conference, however, Dr. Bruce Hillman, chair of the American College of Radiology Imaging Network, described problems with the ultrasound literature.
Hillman's evaluation of six consecutive issues of the Journal of Ultrasound in Medicine found that case studies and case series of little scientific merit dominated coverage. Few of 55 clinical trials published in those six months were scientifically disciplined enough to provide valid, unbiased, reproducible data, he said. Though Hillman has also criticized the general radiological literature in his writing and lectures, his critique of ultrasound research methods was particularly harsh. Few studies were hypothesis-driven. None was multi-institutional. Selection bias was common, blinded readings were rare, and statistical analysis was faulty.
Hillman helped establish ACRIN as a mechanism for radiologists to organize and direct funding for sophisticated evidence-based clinical trials financed by the National Cancer Institute. He noted that only one of 24 ACRIN projects-a multicenter trial measuring the value of breast ultrasound screening-relates to diagnostic ultrasound. Ultrasound ranked last among imaging modalities supported by the NCI-funded studies in fiscal 2006, receiving only 10% of the $170 million devoted to cancer imaging research.
Hillman recommended improving research training for fellows and faculty members. Better research facilities are needed, with a shift in research emphasis from anatomic to molecular imaging research questions, he said. He urged the ultrasound societies to sponsor prizes and grants to reward quality work and invited researchers to develop multicenter trial proposals for possible ACRIN or NCI R01 grant funding.
While executing Hillman's recommendations would require considerable time and attention, technological solutions to ultrasound workflow problems may soon come to the rescue. Three-D ultrasound enables radiologists to apply a protocol-driven strategy for the interpretation of exams.
The conventional image-freeze-print process of ultrasound interpretation will be set aside for a volumetric approach to diagnosis, according to Jim Brown, senior director for clinical marketing at Philips Ultrasound. From that volume, radiologists can select and view up to 25 specific slices for interrogation. Examination times can be cut by at least a quarter.
Improved PACS capabilities also factor into implementing volumetric image interpretation, said Terri Bresenham, vice president of diagnostic ultrasound and information technology at GE. DICOM standards have yet to address true volumetric ultrasound imaging, and not all PACS give the user all the necessary options to manipulate ultrasound volumes. Recent innovations make information gathering shorter and information flow more seamless and accessible.
Automated ultrasound promises to reduce dependence on individual operators for image quality and to boost the efficiency of image interpretation. SomoVu, an FDA-cleared whole-breast system, features a 14.7-cm transducer at the end of an articulated arm. Preset imaging parameters reduce the need for direct monitoring by the radiologist during imaging. Studies are stored digitally, permitting efficient workstation interpretation.
Two studies testing SomoVu's performance were presented at the 2006 RSNA meeting. One involving 165 patients produced a 94% agreement with handheld ultrasound. The other identified lesions were missed with screening mammography.
If new technology is the mother of discovery, ultrasound researchers have access to new ultrasound technologies that could give birth to novel applications for earlier diagnosis and improved treatment. Elastography, initially developed for MRI, adds a new dimension to ultrasound by displaying differences in the elastic characteristics of tissues. The technique is helping improve the differentiation of cysts from metastatic breast masses and the calculation of the age of deep venous thromboses.
Silicon ultrasound could revolutionize transducer design and raise ultrasound resolution to new heights. Ultrasound image resolution has always been associated with bandwidth and the number of channels etched into the piezoelectric material that has been at the core of transducers for decades. The technical limit of that technology was reached with 50 x 50-element 2D arrays, said Richard Chiao, Ph.D., vice president of innovation at Siemens Medical Ultrasound.
Silicon has been adopted as the material of choice for the next wave of innovation. It will help produce 2D transducers with 300 x 300-element arrays.
Additional resolution improvements have been realized with the initial use of phase aberration correction, a data acquisition strategy that compensates for differences in the speed of light through tissue. Zonare Medical Systems pioneered this approach and introduced it on a commercial scanner, Rubin said. Ultrasound has numerous applications in image-guided ablative therapies.
Molecular imaging researchers are using ultrasound microbubbles to transport drugs to targeted anatomy, where the agents are released with the help of focused ultrasound. In other experiments, ligands targeted to P-selectin have been attached to microbubbles for imaging inflammation or peptide-based ligands targeting angiogenic processes, according to Dr. Katherine Ferrerra, a professor of radiology at the University of California, Davis.
Ultrasound advocates will continue to work with the FDA to chip away at the regulatory obstacles to the clinical adoption of ultrasound contrast media in the U.S., Neiman said. A strategy statement, written by Neiman in January on the SRU's behalf, emphasized the need for better research, including more involvement in ACRIN trials, and support for facility accreditation and the aggressive recruitment of ultrasound trainees and sonographer assistants to serve as physician extenders.
Local outreach is also required, according to Dr. James Borgstede, an ultrasound specialist at the University of Colorado Health Services Center in Denver and president of the ACR. Radiologists need to become more service-oriented in their approach to ultrasound by communicating with referring physicians to build good relations and by emphasizing prompt service and rapid reporting of results.
"Rather than taking a 'woe is me' approach to ultrasound, we need to go out and market our skills," he said.
Mr. Brice is senior editor of Diagnostic Imaging.