New medical technology diffusion key element in health cost increases.

Magnetic resonance angiography (MRA) is a noninvasive imaging. method of vasculature based on MRI technology. MRA produces images that are strikingly similar to conventional angiograms but can do so at a fraction of the cost and without the patient discomfort and side-effects associated with the use of contrast media. There is great anticipation that, as MRA matures, many angiograms will become obsolete. Similar to many new technologies, MRA has been presented as cost-effective.

However, it is a simple truth that no technology is cost-effective unless used cost-effectively. If the MRA is done in addition to the angiogram, it may in many cases. contribute unnecessarily to the rising costs of health care. A common example is the work-up of carotid artery disease. At present, a common diagnostic approach is to perform an initial screening test with duplex ultrasonography, followed by an angiogram for surgical planning. MRA now represents a third test, and where it fits into the imaging hierarchy is still uncertain. Should MRA replace the screening ultrasonography, or should it replace the angiogram? Perhaps it could replace both, thus streamlining the diagnostic work-up of carotid artery disease to one imaging test. Maybe the neurologist consultant is comfortable with the MRA, but the vascular surgeon still requires an MRA for surgical planning. Cost-effective use of MRA requires a clear idea of the appropriate imaging hierarchy and agreement among the different physicians caring for the patient. Although the established literature does not suggest that all three diagnostic tests should be routinely performed, the uncertainty of the exact role of MRA creates the definite potential for just that scenario.

The cost-effectiveness of MRA depends on its limited use as an alternative to an angiogram. However, there are myriad other opportunities for MRA to be used as an additional diagnostic test. For example, the decreased morbidity associated with MRA changes the risk-benefit ratio of the test, thus favoring its application in situations where the potential information from an angiogram was not perceived to be worth the risk. The most common example is the use of intracranial MRAs. Because an MRA can be easily appended to an intracranial MRI, it is now routine to see these tests done in tandem--for the routine work-up of the patient with dizziness or nonspecific headaches. The diagnostic yield of such a practice, or how the information gained from the MRA would be used to manage the patient, is unknown. In many cases, this type of testing is done to rule out diagnoses that were considered so unlikely that an angiogram was not warranted. There is an additional financial incentive to perform the tests in tandem. Although an intracranial MRI and MRA are done at the same time in the same machine, suggesting some economy of scale, current CPT codes permit separate billing of the two procedures, thus enhancing reimbursement.

The basic principle at work here is that any favorable alteration of a risk-benefit ratio will inexorably lead to broadening the patient selection criteria. This principle was dramatically illustrated during the diffusion of laparoscopic cholecystectomy. As with MRA, laparoscopic cholecystectomy was primarily represented as a costeffective technology because of the reduction in hospital stay. However, during the early 1990s, when the conversion from open to laparoscopic cholecystectomies was evolving, Aetna witnessed an approximate 50 percent increase in the volume of cholecystectomies. The reasons for this increase are not entirely clear. Some have suggested a reservoir effect, i.e., legitimate candidates who may have deferred a cholecystectomy because of contraindications to an open surgical procedure. Others have hypothesized a "woodworking" effect, i.e., the emergence of new candidates. These patients may have pain that is not entirely typical of gall bladder pain or may have a presumptive diagnosis of acalulous cholecystitis. These patients "coming out of the woodwork" likely reflect a broadening of patient selection criteria.

Experience with other medical technologies has shown that only a small window of opportunity exists to appropriately control diffusion until adequate clinical studies (going beyond the demonstration of technical feasibility) can better define the appropriate niche for a technology such as MRA. Frequently, that window of oppor tunity is sharply defined by thecoverage policy of third-party payers. Widespread coverage encourages rapid diffusion, which, in turn, diminishes incentives to do clinical studies. Widespread diffusion also leads to the creation of premature standards of care. Coverage policies are extremely difficult to alter once something is considered a standard of care, even if the standard is unsubstantiated by outcomes studies in the medical literature.

Over the coming months, payers will face increasing pressure to provide coverage for a wide range of applications of MRA, going far beyond its original niche as a low-cost, safe alternative to an angiogram. These initial coverage decisions will potentially influence tne use of MRA lor years to come. Medicare nas already taken a strong stand on this issue, recognizing MRA only as a substitute for an angiogram in limited circumstances. In October 1995, Medicare issued a national policy statement that limited coverage to the evaluation of carotid arteries in patients who could not tolerate an angiogram.

Elizabeth Brown, MD, is Director, Technology Assessment and Clinical Guidelines, Aetna Health Plans, Aetna Life and Casualty, Chicago, Ill. The opinions expressed in the article are those of the author and not necessarily those of Aetna Health Plans.