Erasing scanned body parts raises questions

August 1, 2005

Technological developments have so expanded the scope of imaging that many traditional components of diagnosis-history, physical exam, old records, even laboratory data-have been deemphasized, neglected, or bypassed. The ascendancy of imaging has correspondingly increased the scope, prestige, and income of radiologists, who are generally regarded as the proprietors of new, advanced techniques and recognized as interpreters of the information they provide.

Technological developments have so expanded the scope of imaging that many traditional components of diagnosis-history, physical exam, old records, even laboratory data-have been deemphasized, neglected, or bypassed. The ascendancy of imaging has correspondingly increased the scope, prestige, and income of radiologists, who are generally regarded as the proprietors of new, advanced techniques and recognized as interpreters of the information they provide.

The secure position of the radiologist as manager of diagnostic imaging is being challenged on many fronts, however. The increasing capture of extremity MR by orthopedists and widespread adoption of imaging modalities by cardiologists are just two of the serious issues confronting not just radiologists but the U.S. healthcare system as it seeks to reduce costly self-referral practices.

Contests for control between specialists are not new, but could take on new dimensions as imaging equipment-especially multislice CT scanners-becomes more sophisticated and acquires new capabilities.

The power and versatility of computer-aided imaging can bring into focus pathologic conditions hitherto revealed only through surgical inspection or interventional techniques. The advent of 64-channel CT offers the prospect of visualizing soft plaques in the coronary arteries, the lesions most apt to ulcerate and block an artery. Both radiologists and cardiologists are eager to direct 64-slice studies, which promise to make diagnostic catheterization of the coronary arteries obsolete.

Who will eventually wrest control? Will it be radiologists, masters of the intricacies of CT interpretation elsewhere in the body but not fully knowledgeable about the heart and its vessels? Or will it be cardiologists, students of the heart in all its dimensions but untrained in the assessment of imaging findings in the lung and thorax?

As you think about this question, consider the following: Even as it is technically possible to review the heart in detail, it has become possible to erase the lungs and thorax from an image slice. An image from a multislice cardiac CT scanner can be made to reveal a picture of only the heart, even though the entire thorax has been irradiated. Such a visually restrictive organ-specific MSCT image will be attractive to cardiologists, who would not have to worry about missing lung pathology because they will not be able to see it. A vendor can offer such specialized equipment solely to cardiologists, who may use it in part to recoup income they might expect to lose from the decline of diagnostic catherization.

I imagine most radiologists would be aghast at such a modification of the CT image protocol. They would be loath to interpret images with such a device, knowing that diagnostic information is being suppressed to the possible detriment of patient care. Moreover, radiologists would be alarmed at the potential of malpractice considerations engendered by not being able to spot a treatable lung lesion.

RISKING MALPRACTICE

It is unknown whether such a circumstance really places a cardiologist or CT operator in jeopardy. Many patients, smokers in particular, are at heightened risk for both coronary artery disease and bronchogenic carcinoma. But a successful malpractice action rests not only on the existence of a diagnostic mistake but also on the fact that the mistake harms patient outcome. We still don't know whether very small nodules in the lungs seen on CT but indistinguishable on plain films will grow and metastasize. It has not been proven that detecting minuscule lesions early and removing them would yield a cure.

For many patients with concurrent lung pathology, a digital scout film with all intrathoracic information included will reveal most lesions. Furthermore, judgments against a physician in lung cancer cases are not generally imposed near the time of a missed diagnosis but tend to occur several years later, when the consequences of neglect or error produce an adverse outcome. Thus, the user of a visually restricted cardiac-only CT could practice for several years before a malpractice case arises.

What about the manufacturers of such devices? Here again, their risk will probably not be apparent immediately. A class action suit against them would take years to initiate and complete, and recent legislation has redirected most class action suits away from state courts to federal courts, where it is more difficult to define a valid class and achieve success.

OUTSIDE THE HEART

The implications of a visually restrictive organ-specific CT are not confined to heart views. If it is possible to exclude the rest of the thorax when viewing the heart, then it would be equally possible to exclude the rest of the abdomen when viewing the colon. Hence, virtual colonoscopy (or CT colonography) could be accomplished by filtering out everything except the large intestine.

Gastroenterologists are exempt from Stark regulations when they open a freestanding diagnostic and treatment center. In that setting, they can self-refer with impunity. The additional opportunity accorded to CT vendors is obvious. They would offer gastroenterologists a visually restrictive colon-specific CT scanner. The advantages to a gastroenterologist are equally obvious. Having this device means control of both CT colonography and optical colonography. The patient need no longer visit a radiologist for colon cancer detection by virtual colonoscopy but would go to the gastroenterologist for a complete colonoscopic evaluation. If anything suspicious is seen on a virtual study, even only retained fecal fragments simulating a polyp, then an optical colonoscopy would be ordered and performed at the same center.

The gastroenterologist would be able to corner the market in diagnostic procedures for the detection of large bowel cancer, as a separate CT colonoscopy performed by a radiologist would no longer be attractive to patients. The gastroenterologist could hire a radiologist to read the CT colonography studies. Since the real income is in the technical fee and not the professional fee, the radiologist's portion would be reduced while the gastroenterologist's would be enriched.

Would the gastroenterologist be placed at risk by controlling both organ-specific virtual colonoscopy and optical colonoscopy? Might he or she be sued for missing an abdominal lesion within the scope of a CT examination that would have been observed in the subtracted-out information beyond the colon? In otherwise asymptomatic individuals, the disappointing results of full-body CT screening should give gastroenterologists some cheer. A major problem with CT screening has been the frequency of false positives, which promote further workup and the eventual ruling out of disease.

Renal tumors are the only malignant lesions occurring in asymptomatic individuals detected by this technique that can be treated. Even then, the prevalence of these neoplasms is so low in otherwise normal individuals that false positives greatly outweigh true positives, so the malpractice risk for gastroenterologists is probably not very high.

Another effect of eliminating information is its reinforcement of the fashionable notion that every specialty requires access to imaging. Acceptance of this concept has led some institutions to offer prospective candidates rewards of bed services at the expense of an existing radiology department. Fulfillment of this promise means carving out territory from radiology. The deployment of visually restrictive organ-specific CT makes it easier for a dean and nonradiologist chair to play the access-to-imaging card. This device makes it possible to claim there is no need for radiology to serve patients who have diseases in the chair's specialty.

Manufacturers that seek to pursue this new route to enhanced profits must be stopped immediately. The leaders of radiology should raise concern, both in public and before legislators and regulators. A consensus should be articulated that the deposition of gratuitous radiation, shorn of its diagnostic capability, is a public health outrage. The radiology community should vociferously support physician self-referral restrictions.

The American College of Radiology's proposal of a designated physician imager could serve to limit the profitability of these devices. To be effective, quality issues and radiation risks must be joined together to counter claims of legitimacy or efficacy for visually restrictive organ-specific CT machines.

Dr. Baker is chair of radiology at the University of Medicine and Dentistry of New Jersey.