Imaging of shoulder opens new turf fight for radiologists

Although musculoskeletal radiologists have long debated the relative roles of MRI and ultrasound in imaging shoulder injuries, they have reached consensus on a number of diagnostic algorithms. But the proliferation of inexpensive, low-end compact ultrasound systems has introduced a new controversy.

"The issue isn't MRI versus ultrasound. The issue is nonradiologists versus radiologists," said Dr. David W. Stoller, director of musculoskeletal MRI at the California Pacific Medical Center in San Francisco.

As a speaker at the International Skeletal Society meeting in Singapore last October, Stoller was expected to talk up the merits of MRI versus ultrasound for shoulder imaging. But Stoller's presentation was not what the audience expected to hear. Instead he lectured on the threat posed by nonradiologists performing diagnostic ultrasound exams.

"Ultrasound and MRI are complementary, so I wasn't there to prove that one modality is superior to the other. The concern is about the misuse of ultrasound by nonradiologists," Stoller said.

Although MSK radiologists have been raising eyebrows lately over the increasing use of ultrasound technology by nonradiologists, they don't necessarily view the encroachment as a menace. Some even feel reassured by MRI's mounting prominence in the field under their aegis (see "Radiologists look over their shoulders, knees, and hips," April 2005, page 67). Perhaps they should take another look.

Orthopedic surgeons and their brethren are increasingly wielding ultrasound probes for the workup of shoulder conditions. Nor is the trend in decline. A recent study from Norway, for instance, encourages orthopedic surgeons to start using ultrasound as a diagnostic technique for full-thickness tears of the rotator cuff and for pathology in the long head of the biceps (Acta Orthop 2005;76(4):503-508).

For musculoskeletal imaging specialists like Stoller, the possibility that these nonradiologists might be performing diagnostic ultrasound scans without proper knowledge or equipment is haunting. Now they must ponder the prospect of competition from nonphysicians as well.

In addition to internists, rheumatologists, and sports medicine physicians, chiropractors and physical therapists are also beginning to acquire the latest generation of affordable, compact ultrasound equipment. These machines can be purchased for as little as $20,000, a far cry from the $200,000-plus high-end ultrasound equipment usually found in radiology or ob/gyn departments. Moreover, some of these practitioners have a history of forays into diagnostic ultrasound thwarted by clinical imaging watchdogs.

For example, in June 2002, the American Institute of Ultrasound in Medicine published a statement discouraging adult patients from seeking diagnostic ultrasound exams of the spine. Chiropractors had been performing these procedures despite the lack of evidence in the medical literature to support their value. One chiropractor even patented an "ultrasound spine scan" and waved it at ultrasound manufacturers, but the invention never caught on.

Not all chiropractors condone such practices. The Council on Chiropractic Practice Clinical Guidelines, an evidence-based manual, recognizes ultrasound's appeal but deems its chiropractic applications to be investigational.

Chiropractors, physical therapists, and other nonradiologists cannot bill the Centers for Medicare and Medicaid Services for diagnostic procedures. But gray areas exist regarding who can perform these tests. CMS allows certain diagnostic exams to be performed by nonphysician healthcare professionals-such as physical therapists-even without physician supervision. In addition, some physicians are using loopholes in federal antireferral regulations to bill for scans in interpretations (see "Specialists garner a bigger share of medical imaging," November 2005, page 68).

"Ultrasound is being misused as both a sword and shield," Stoller said.

Low-quality ultrasound that is performed, interpreted, and billed by nonradiologists has become like a sword pointed at the core of quality musculoskeletal imaging, he said. And the volume of clinical literature that is flawed or taken out of context to extol ultrasound's alleged superior virtues has become the shield used to propagate this trend.

In his presentation at the Singapore meeting, Stoller reviewed clinical literature on ultrasound imaging of the shoulder and concluded that most of it contains flaws.

He found anomalies in study design and research methodologies, errors that have been used to support mistaken conclusions, multiple biases, and significant interobserver variability for ultrasound interpretation between radiologists and nonradiologists. He came across several articles advancing the view that ultrasound performs at a similar or slightly better level than MRI, particularly in regards to the diagnosis of partial and full tears of the rotator cuff (see accompanying article).

To test ultrasound's value first-hand, Stoller and colleagues conducted a blinded prospective study. They enrolled 100 consecutive patients presenting with shoulder pain and focused on partial and full-thickness tears of the rotator cuff and lesions of the labrum. They found that almost half of these patients had labral tears undiagnosed by ultrasound. In addition, they discovered that 40% of patients presenting with shoulder pain actually suffered from conditions completely unrelated to the rotator cuff.

The team also checked the accuracy of ultrasound and MRI on seven human cadavers to validate results. A sonologist and a sonographer with extensive ultrasound experience correlated findings with a high-end ultrasound scanner against those of a 1.5T MRI system. They found that MRI did not miss any relevant pathology. Conversely, they found that the experienced sonographer missed a significant number of findings, despite the availability of top-of-the-line ultrasound equipment.

The results sustain the view that diagnostic ultrasound performed by nonradiologists with subpar equipment puts patients at risk, Stoller said.

"I support the use of diagnostic ultrasound when it's the appropriate resource or when there is limited MR access," he said. "But quality care standards suffer when complex injuries get managed on the basis of incomplete diagnostic information."

SOCIOECONOMIC IMPACT

Nonradiologists see diagnostic ultrasound as a potential profit center, particularly in applications seen as money losers by established imaging centers. With more handsome reimbursement provided by MRI, ultrasound is viewed as the disposable modality. Scores of referring physicians see this as an opportunity to extend the scope of their clinical practice and strengthen their patient base. One physician's small loss turns into another's big gain.

"Many physiatrists, rheumatologists, primary-care physicians, ER physicians, and sports medicine physicians are getting into ultrasound for injection (guidance) and diagnosis. This is partly based on financial incentives. Reimbursement for ultrasound can be lucrative for these practices," said Dr. Lynne S. Steinbach, chief of musculoskeletal imaging at the University of California, San Francisco.

However, these nonradiologists-physicians and nonphysicians alike-may not realize their lack of expertise. When combined with the reduced diagnostic performance of the equipment they are likely to use, that lack could have a negative impact on the quality of shoulder imaging exams. The implications for radiology in particular and healthcare as a whole are daunting, according to MSK experts.

About four million people in the U.S. will seek medical attention for shoulder injuries each year, according to the American Academy of Orthopaedic Surgeons. Shoulder disorders account for about 5% of all consultations with family physicians. Despite treatment, 80% of patients remain symptomatic after six months, and 50% are still symptomatic 18 months post-treatment. More than half of these patients report recurrent pain so severe that it interferes with their work, even three years after their first medical claim. Nearly two-thirds of patients reduce their involvement in recreational activities; almost half report sleep problems; and about one-third suffer from depression because of their shoulder ailment. More than one-third of all patients with some type of shoulder injury will lose their jobs.

An accurate diagnosis is critical in order to minimize the economic and social impact of shoulder injuries. A delayed or incorrect diagnosis exposes patients and the healthcare system to the burden of unnecessary imaging studies, expensive treatment, and extended recovery. A disability rating provided by the RAND Institute for Civil Justice confirms that patients with shoulder injuries confront larger income losses than those with other types of musculoskeletal conditions (J Occup Rehabil 2002;12(3):205-221).

"This is a huge problem," said Dr. Bruce Forster, an associate professor of radiology at the University of British Columbia in Vancouver. "In Canada, many patients wait for up to a year to see a specialist, and it can take months before they get an MR. In countries with socialized healthcare like Canada, it can take years. The negative impact on quality of life is considerable."

Moreover, poor diagnostic quality could also undermine therapy. Healthcare dollars will hardly be saved if patients still walk in pain after diagnosis, undergo invasive arthroscopy, or receive an MRI several months late, Stoller said.

DAMAGE CONTROL

A plain x-ray should be the starting point for patients with shoulder injuries, said Dr. Levon Nazarian, a professor of radiology at Thomas Jefferson University in Philadelphia. The complexity of the lesion should eventually call for either ultrasound or MRI scanning.

"Ultrasound should be used for rotator cuff or dynamic impingement. If you want to investigate labrum pathology, bone tumors, or fractures, you should refer the patients to MRI. That's the basic algorithm," Nazarian said.

Nothing is written in stone, however, and there are other considerations. Imagers should always opt for an MRI of the rotator cuff if they lack ultrasound expertise. On the other hand, MRI is not for every patient. Besides the obvious contraindications, such as pace makers, metal implants, and claustrophobia, MRI is a much more expensive diagnostic test.

A number of sonologists are troubled by what they interpret as an overzealous attempt to defend the MRI business at the expense of ultrasound.

"Reimbursement issues have been driving some radiologists to stick to their MRI-which I think is a great technique-but they don't realize that there is a wealth of information out there about other modalities that they are not willing to embrace," said Dr. Marnix T. van Holsbeeck, a professor of radiology at the Henry Ford Hospital in Detroit.

Radiologists cannot protect their turf by knowing less or by embracing less. They can only protect their turf by knowing and embracing more. As ultrasound is appropriated by other specialties, radiologists are failing to see that their referral bases are eroding, van Holsbeeck said.

"We are at the crossroads. You can only protect your MRI by knowing what's out there that's non-MRI. At this point, some radiologists are blinded to what is happening outside the realm of MRI," he said.

A host of other issues includes a fear by MSK radiologists of burning bridges to referring providers. Even though radiologists may lose the ultrasound business eventually, they are reluctant to alienate chiropractors and other nonradiology practitioners who provide steady referrals. While not a pressing concern to MRI specialists at present, loss of referrals looms large as a potential financial issue in the future. There are already indications of MRI self-referral by nonradiologists.

Radiologists need to view ultrasound and MRI as complementary modalities, said Dr. Douglas P. Beall, chief of musculoskeletal ultrasound at the University of Oklahoma. Both have strengths and weaknesses that only well-trained imagers can appreciate and use to their advantage.

While MRI is probably the single best test for an overall anatomic survey, ultrasound could decrease healthcare costs by providing an effective, inexpensive exam for evaluating certain musculoskeletal disorders affecting tendons, muscles, peripheral ligaments, and for processes that require real-time observation. MRI, on the other hand, has made gains in evaluating the osseous structures, articular cartilage, and fibrocartilage structures and in assessing deep-seated ligaments.

"Ultrasound is the flashlight and MRI is the floodlight in the search for pathology," Beall said.

Mr. Abella is an assistant editor of Diagnostic Imaging.