To the Editors:
We are delighted that the topic of musculoskeletal ultrasound in rheumatology is being discussed again in the editorial by radiologists Roemer, van Holsbeeck, and Genant (Roemer FW, van Holsbeeck M, Genant HK. Musculoskeletal ultrasound in rheumatology: a radiologic perspective. Arthritis Rheum 2005;53:491–3). Not surprisingly, they recommend that ultrasound is best left to radiologists. What is surprising though, is the fact that their comments about rheumatologists performing ultrasound studies are published in a rheumatology journal.
The authors' criticism is that rheumatologists may gain only “superficial and one-sided knowledge” of sonographic anatomy since they lack “insight into MRI and CT anatomy.” We agree that radiologists who specialize in musculoskeletal cross-sectional imaging are likely to have more exposure to these imaging modalities. We believe, however, that this does not automatically make them more proficient in detecting and interpreting the sonographic findings that are mentioned such as early erosions, synovitis, tenosynovitis, or synovial and entheseal inflammation. Current textbooks on diagnostic ultrasound written for radiologists do not contain chapters on these rheumatologic indications. Rheumatologists, who in North America have at least 3 years of training in internal medicine and at least 2 years of training in rheumatology, are likely to have a deeper understanding of the underlying pathophysiology of the plethora of conditions that lead to inflammatory arthritis. Therefore, rheumatologists who perform musculoskeletal ultrasound will know where to look for the conditions they treat, and how to interpret their findings. Moreover, they may have the advantage of knowing their patient and target their examination accordingly. An office-based rheumatologist-sonographer can integrate patient history, clinical findings, serologic studies, and sonographic findings.
Roemer and colleagues raise the point that ultrasound studies that are self-referred by rheumatologists may lead to a probable “4.4–7.5 times” increase in the cost of diagnostic imaging. In our opinion, this inexpensive point of care diagnostic modality is more likely to decrease overall health care costs dramatically. In many instances, in particular for the above mentioned indications, conventional radiography and magnetic resonance imaging are unnecessary if ultrasonography is available. Since treatment options in rheumatology include disease-modifying agents that may average $15,000 per year, early identification of patients who need these medications and the exclusion of patients without inflammatory arthropathies will help with cost effective management.
We believe that musculoskeletal ultrasound belongs in the hands of the rheumatologist. We agree with Roemer et al that standardization of training and examination is needed in the near future. The different foci of expertise of radiologists and rheumatologists combined may facilitate this undertaking.