Winds of change still buffeting radiologists and manufacturers

Winds of change still buffeting radiologists and manufacturers – impact of managed care on radiology

The revolution wrought by managed care continues to toil radiology. Managed care has re-introduced consideration of cost into medical decision-making, a development with far-reaching consequences. It has spurred rethinking of the role of radiology, reductions in reimbursement for radiological services, reductions in requests for radiological procedures, reorganization of radiology facilities, re-engineering of equipment suppliers, revision of radiologists’ training, and a renaissance of interest in computerized systems to manage processes and data.

Radiologists feel somewhat beleaguered. To a large extent, their large incomes were a product of the stream of brilliant technical advances typified by the inventions of computed tomography (CT) and magnetic resonance (MR). Because these new technologies were undeniably effective but expensive to purchase and install, it was easy to justify high technical fees and correspondingly high professional fees. As long as everyone was being paid independently by a medical insurance system that made little attempt to control costs, hardly anyone noticed or cared that radiologists climbed almost to the top of the medical profession in terms of income.

But now that managed care has brought cost considerations to the fore, many managers feel they pay too much for radiological services and want concessions in both technical and professional charges. Managers, including physicians who care directly for patients, have come to see radiology as merely a service – an essential and important service to be sure, but still only a piece of the whole diagnostic puzzle. Especially under global capitation, radiology services are just another expense to be minimized, consistent with reasonable standards of quality.

In these circumstances, radiologists are groping for a more essential role. Dr. Ernest Ferris, outgoing president of the Radiological Society of North America (RSNA; Oak Brook, Illinois), warned in his final address during the society’s annual meeting this past December that radiologists would become mere providers of a commodity service unless they took steps to re-establish their role in medicine. He observed that there are no new technologies today with the promise that CT and MR once held for radiologists.

Radiologists must provide consistently superior service and demonstrate that their performance is superior to that of other physicians not specifically trained in the nuances of imaging, he said. In addition, radiologists should position themselves as consultants to referring physicians, providing advice about the most effective and efficient imaging work-up for difficult diagnoses, not simply performing whatever tests were ordered. They should become more directly involved in patient care in the manner that interventional radiologists do. They should develop expertise in information technology, showing all of medicine how to employ computers to maximize efficiency and and control costs.

The degree to which radiologists will be successful in carrying out this program remains to be seen, but the implications are far-reaching because what happens to radiologists will influence what happens to radiology departments and to the $3.8 billion-a-year industry that provides radiological equipment in the U.S.

There are some 25,000 practicing radiologists in the U.S. If the nation were staffed with radiologists in the same manner as the Kaiser plans and other well-organized HMOs, the United States would need only 13,000. This vast oversupply might be expected to lead to a diminution of the number of radiologists in training, but that has not occurred. While such a development is in the interest of the profession, it is not in the interest of each academic center, which gains both funding and prestige from graduating residents.

Nevertheless, pressures on radiology as a profession continue to mount. Radiologists’ incomes have declined on average by about 10%, and by more than 30% in some areas, California among them. With revenues falling, radiology groups choose to work harder rather than to add new members. Radiologists are gradually acceding to demands that their services be routinely available around the clock. Recognizing that the hospitals where they practice are under equivalent financial pressure, radiologists are increasingly taking responsibility not only for interpreting images but also for assuring low-cost, efficient operation of their departments.

Modest market growth expected

An important consequence of falling revenues to providers has been weakness in the markets for medical imaging equipment, as shown in Table 1. The high point was reached in 1992, when total U.S. sales exceeded $4 billion. Except for Picture Archiving and Communication Systems (PACS), the name given to electronic image management systems, sales of all kinds of medical imaging equipment fell over the next two years to a total of $3.5 billion in 1994. Sales stabilized in 1995 and have made some recovery in 1996, though, with the exception of ultrasound, not to the levels of 1992. Magnetic resonance equipment sales were especially hard-hit as the installed base approached saturation. Some growth is expected during the next four years, so “that the total market, including PACS, will reach almost $4.5 billion in the year 2000. Reasons for these patterns are discussed below.

Fundamentally, there are two countervailing trends. On the one hand, there are the cost pressures widely attributed to the effects of managed care, which militate against purchase of new equipment. On the other hand, medical imaging plays an essential role in diagnosis, and even expensive modalities like CT and MR have become routine for a variety of diagnoses because they are usually definitive. Though there are no new modalities with great potential, there are many technical developments that promise to enhance the effectiveness of existing modalities. Furthermore, in a number of instances new equipment offers the prospect of future savings, thus furnishing a motivation to buy even in times of strained financial circumstances.

X-ray is a steady replacement market

X-ray imaging embraces several distinctly different kinds of equipment, including radiography and fluoroscopy (R&F) systems, mechanical tomography systems, angiography and cardiac catheterization systems, a variety of specialized systems for mammography, urography and chest examinations, as well as portable radiographic and fluoroscopic equipment. Most of this equipment is technically mature, and most sales are replacement sales of existing equipment, limiting this market’s potential for growth. Still, the large installed base provides a steady market for replacing obsolete equipment.

There are new developments on the horizon, however – notably in digital capturing of X-rays. Until now, film has played an essential role in static X-ray imaging, where it serves to capture the X-ray image, display it for interpretation and store it for archival purposes. While electronic alternatives for displaying and archiving medical images have made considerable progress in the digital modalities, X-ray imaging has remained largely confined to film because there has been no entirely satisfactory electronic means for capturing the images. Computed radiography (CR), originally introduced by Fuji (Stamford, Connecticut) in 1981 and now available also from Eastman Kodak (Rochester, New York) and Agfa Gevaert (Richfield Park, New Jersey) as well as Fuji, substitutes a reusable phosphor plate for film, but the plate requires transfer to a reading device in a manner similar to transfer of film to a fill processor. CR has found favor as an X-ray detector for portables because digital processing can restore overexposed or underexposed images, but there are still only about 500 such systems installed in the U.S.

Table 1

U.S. Sales of Medical Imaging Equipment (millions of dollars)

1992 1996 2000(E)

X-ray 1,100 1,100 1,200

Ultrasound 950 1,050 1,200

Computed Tomography 690 590 700

Magnetic Resonance 850 450 600

Nuclear Medicine 350 280 300

Picture Archiving 150 300 470

and Communication

Total 4,090 3,820 4,470

Source: Concord Consulting Group

But direct-capture digital systems are now reaching the market. A Swiss company, Swiss Ray (Hochdorf, Switzerland), has applied for FDA approval of the “Add-on Bucky,” a device consisting of a scintillating screen that converts the X-ray image to a light image, followed by some optics that focus the four quadrants of the image on four CCDs (charge-coupled devices like those used in video cameras). A similar approach is being pursued by Fischer Imaging (Denver, Colorado) and Trex Medical (Danbury, Connecticut), a subsidiary of Thermo-Electron, which formed Trex from two mammography companies, Bennett (Copiague, New York) and Lorad (Danbury, Connecticut), to create full-field digital mammography systems. However, these systems suffer from low efficiency in transferring light from the screen to the optics. In addition, because of the space taken up by the optics, they do not fit into existing equipment, which must be modified to accept the new receptor. For these reasons these systems are not expected to make significant inroads in the market, despite the advantages of digital processing following digital capture of X-ray images.

Philips Medical Systems (Shelton, Connecticut) sells Thoravision, which captures an X-ray image of the chest as a pattern of trapped electrons on a rotating selenium drum in the same way that a Light image is captured in a copier. Instead of spreading toner, the Philips system reads out the trapped electrons through a line of electrometers, thus producing the array of digital image data. Although this system has been available for two years, only a handful have been placed.

Of greater promise are systems that use amorphous silicon or amorphous selenium to convert radiant energy to electrons, which are then read out through a large-area array of transistors in layers below the silicon or selenium. Sterling Diagnostic Imaging (Newark, Delaware), which bought DuPont’s (Wilmington, Delaware) medical imaging business in 1995, promises delivery of a selenium-based digital system in 1998. The European companies Philips (Best, the Netherlands), Siemens (Erlangen, Germany) and Thomson (Velizy, France) recently announced a consortium – Trixell (Moirans, France) – to pursue development of a similar system. Two companies – dpiX (Sunnyvale, California; a subsidiary of Xerox) and EG&G Reticon (Palo Alto, California) – have developed prototypes of large-area detectors using silicon, and this technology has been licensed by Varian (Palo Alto, California). Depending on details of the design, these systems can be adapted either for static imaging or for fluoroscopy.

The advantage of these systems is that the receptor is a fiat panel which should be thin enough to fit into existing X-ray machines without modification. Although they could be moved from one machine to another, they are expected to remain permanently in place. In use, they provide the advantages of digital capture, including digital processing and the ability to display the image almost immediately after exposure, so that the technologist can know whether the image is satisfactory before the patient leaves the room. They also have the ability to send the image electronically anywhere for interpretation or other functions.

These systems still present some difficult technical challenges, particularly reliable manufacture of large transistor arrays without significant blemishes. At the moment, manufacturers speak of prices above $100,000, which is far too high to expect that such devices will find widespread use. However, manufacture of large arrays of electronic components is gradually being improved, as the growing availability of flat-panel displays attests. Thus, it seems improbable that flat-panel X-ray detectors will sweep the market in the next few years. In the longer run, they hold the promise of bringing the last nondigital medical imaging technique into the digital realm.

Technical advances boost ultrasound

Ultrasound presents a different picture. Having declined and stagnated in 1994 and 1995 for the reasons noted above, ultrasound grew some 10% in 1996. Unlike X-ray, nuclear medicine, MR and CT, where the principal suppliers are large full-line companies, ultrasound is dominated by companies that have no other medical imaging products, like Acuson (Mountain View, California), Advanced Technology Laboratories (ATL; Bothell, Washington) and Hewlett-Packard (Palo Alto, California). Also unlike the other modalities, ultrasound is widely dispersed, installed not only in hospitals, but also in doctors’ offices – notably obstetricians/gynecologists, cardiologists, urologists and ophthalmologists.

Ultrasound systems long have made use of digital components, but in recent years advances in technology have allowed these systems to become wholly digital, providing much more flexibility and accuracy in beam-forming and signal processing than has heretofore been available. As a result, ultrasound images are better than ever. Even less-expensive machines produce high-quality images, keeping the market alive by making older equipment noticeably inferior.

In addition to technical advances in the equipment, ultrasound contrast agents promise to give the field a boost, particularly in vascular imaging. Molecular Biosystems’ (MBI; San Diego, California) Albunex has been a disappointment, but newer agents from MBI and others, like Sonus Pharmaceuticals (Bothell, Washington) and Schering AG (Berlin, Germany), promise to enhance ultrasound’s ability to visualize and measure blood flow.

Spiral CT boosts the market

Computed tomography was introduced in 1973 and is properly regarded today as a mature technology. Still, there have been some new developments that have kept the market vigorous. Spiral CT, in which the X-ray source rotates continuously around the patient, allows collection of sufficient data for a single slice in a second or less and for a complete study in less than 15 or 20 seconds, a short enough time for a breath-hold. Because spiral CT is continuous, there is more flexibility in division of data into slices; it is possible to construct slices of different thicknesses or to construct intermediate slices between those in an original reconstruction.

All of the major CT manufacturers – GE Medical (Milwaukee, Wisconsin), Siemens (Iselin, New Jersey), Picker (Cleveland, Ohio), Hitachi (Tarrytown, New York), Toshiba (Tustin, California), Yokogawa (Tokyo), and Elscint (Haifa, Israel) – have introduced spiral CT. Elscint has a unique design employing two rows of detectors and a slightly lengthened X-ray beam so that its machines collect data in two intertwined spirals, thus cutting total scan time in half. The special capabilities of spiral CT have served to keep the CT market from declining appreciably despite the pressures discussed above, and this market should be sustained as older equipment in the installed base is gradually replaced.

Magnetic resonance holds a secure place

Of all the modalities, MR suffered most from economic pressures, declining in 1994 to only half of its maximum sales in 1992. With some 4,500 MR machines installed in the U.S. (compared with 5,800 CT machines), this market is beginning to feel the effects of approaching saturation. At the same time, MR provides definitive diagnoses for common diseases and injuries of the brain and spine, as well as the knee and shoulder, giving it a secure place in the diagnostic arsenal.

Because of its technical richness, researchers continue to develop novel ways to use MR. Functional imaging of brain activity is now a routine technique used for pre-operative planning in university settings. MR shows promise as an alternative to biopsy for women with mammography-detected breast lesions. Some researchers think MR can become a one-stop analyzer of heart function, providing all the data that now comes from ultrasound, nuclear medicine and coronary angiography. Although MR does not create real-time images as do ultrasound and fluoroscopy, it can be adapted for interventional procedures, where its ability to visualize lesions invisible by other techniques gives it a unique role. These are all reasons to expect that use of MR equipment will grow and with it the market for MR equipment.

Manufacturers have not been laggards in producing systems to meet various market needs. All of the large manufacturers now include systems with fields of 0.5, 1.0, and 1.5 tesla, giving customers a complete range of choices with respect to field strength. General Electric, Siemens, Picker, and Hitachi offer systems with fields around 0.2 tesla using permanent or resistive magnets and open designs to avoid claustrophobia and facilitate interventional procedures or operation of life-support apparatus. Toshiba introduced a remarkable open magnet system called OPART, which has a 0.35-tesla superconducting magnet that requires no externally supplied cryogens, keeping the coils at superconducting temperatures (below 10 Kelvin) by refrigeration, a technical tour de force.

Nuclear medicine’s diagnosis role declining

Nuclear medicine seems to be the stepchild of medical imaging. Having lost its claim as the only modality capable of functional imaging, nuclear medicine has seen its role in diagnosis has steadily eroded. Nevertheless, it remains the best way to identify bone metastases and to assess defects in cardiac perfusion, and it has a number of less-frequently employed roles as well. The problem is that it has no important new roles. Designers are focusing their hopes on new radiopharmaceuticals that may expand nuclear medicine’s roles and on adaptation of standard cameras for positron emission tomography (PET).

Ten years ago, monoclonal antibodies provided hope that nuclear medicine would expand, since these easily manufactured “magic bullets” gravitate rapidly and consistently to occult or incipient cancers and infections, and are most easily localized by nuclear medicine. And the amount of radioactive material necessary to make them detectable is much smaller than the amount of material needed for detection by MR or CT. Monoclonal antibodies, antibody fragments and pep-tides in these roles have all been somewhat disappointing, but hope springs eternal, and current research may yet pay off.

Also providing hope is use of SPECT (single photon emission computed tomography) cameras for examinations using a positron-emitting isotope of fluorine in FDG (fluoro-deoxyglucose). Glucose consumption is a sensitive indicator of cellular metabolic activity, and tracing the path of injected FDG provides information about the presence or absence of activity in the brain and heart. SPECT cameras with two heads are designed to create tomographic images of the distribution of gamma-emitting radioisotopes such as technetium, iodine, gallium, and thallium, where two heads double the capture of emitted gamma rays.

The same hardware can also be used to capture the two gamma rays emitted by positrons from radioactive fluorine when they are annihilated after encountering electrons. These rays follow paths 180 [degrees] apart; construction of an image depends upon the fact that simultaneous arrival of gamma rays in both heads implies that the annihilation took place on the line joining the two points where the arrivals were detected.

The technique is called coincidence detection, and it is the basis of PET imaging. However, PET cameras are large and expensive, costing well over $1 million, whereas dual-head SPECT cameras cost about half this amount and are useful for other examinations using conventional radiopharmaceuticals. But, like PET, coincidence detection using SPECT cameras must contend with the relatively short half-lives of the positron emitters. Fluorine has the longest half-life – about two hours – but its use places hard-to-maintain requirements for prompt delivery and use of FDG. Thus, coincidence detection provides some potential for nuclear medicine applications, but it is hard to visualize widespread use outside of research institutions. Both ADAC (Milpitas, California) and Picker have FDA-approved coincidence detection machines on the market, and Elscint, Siemens and SMV America (Twinsburg, Ohio) have systems in development. Park Medical (Toronto, Ontario, Canada), which pioneered these developments, has placed its bets on a proprietary coded-aperture technique instead of simple coincidence detection.

Growing interest, market for PACS

Contrary to the cautious optimism surrounding existing imaging modalities, picture archiving communications systems enjoy heightened interest and a growing market. These systems embrace all configurations of image management systems where images are captured, stored, transmitted and displayed in electronic form. They include not only full-scale systems that do away with film almost entirely but also less complete systems, which can be called mini-PACS, and teleradiology systems, which transmit images from one place to another, either bringing images to the radiology department for interpretation or sending images away for other users to interpret.

The reason for growing interest lies in the same forces that have retarded growth elsewhere in the health care system. Computer systems promise more efficient operations, thus addressing the need to do more with less. They allow elements of care to be dispersed, thus supporting the growth of regionalized healthcare. They make it easier to assemble and process data of various kinds, thus making billing and collection more efficient and supporting outcome studies and other measures of quality.

Only a handful of filmless radiology departments exist in the U.S., mainly at military and VA hospitals, Those hospitals have plans to extend filmless operations ever more widely. Beth Israel Hospital (New York) and the Medical University of South Carolina (Charleston, South Carolina) also have installed full-scale PACS and commenced filmless operations. Suppliers include Lockheed Martin (formerly Loral; Hoffman Estates, Illinois), Agfa, and Siemens. While other hospitals have ambitions for filmless operations, most are daunted by the cost of full-scale PACS (typically $5 million to $12 million and more) and resistance from radiologists who worry about the technical limitations of PACS, especially those of displays and workstations.

For this reason, most installations are mini-PACS, which are smaller, less expensive and easier to implement. Mini-PACS typically serve only part of the radiology department, like ultrasound, nuclear medicine or the cross-sectional modalities CT and MR, or they address only part of the department’s operations, like archiving or furnishing images and reports to other users, such as physicians in emergency rooms or intensive care units. Mini-PACS for ultrasound are provided by Acuson and ALl Technologies (Vancouver, British Columbia, Canada); ATL recently sold its ultrasound mini-PACS business to Kodak. Nuclear medicine mini-PACS are available from ADAC, GE, Siemens, and Sudbury Systems (Sudbury, Massachusetts).

Numerous other suppliers provide systems ranging from simple point-to-point teleradiology to full-scale PACS. They include the film companies Agfa, Kodak, Sterling, Imation (St. Paul, Minnesota), Fuji, and Konica (Wayne, New Jersey); the imaging equipment companies GE, Siemens, Philips, Picker, and Elscint; the defense contractors E-Systems (San Antonio, Texas) and Lockheed Martin; the information systems companies IDX (Burlington, Vermont) and SMS (Malvern, Pennsylvania); and specialized companies like Cemax-Icon (Fremont, California), Olicon (San Clemente, California), CompuRad (Tucson, Arizona), Access (Natick, Massachusetts), Images on Call (Dallas, Texas), Line Imaging (Atlanta, Georgia), Rogan (Grafton, Wisconsin) and others.

It is hard to see how all these companies can survive in this business, and analysts predict a shake-out each year. While small companies (and even some large ones like ATL) leave the business each year, new ones keep coming along to keep the number of suppliers growing. For this reason the business is highly competitive and not very profitable, and there is little reason to expect an abrupt change.

It does seem clear, however, that the current situation cannot persist indefinitely. Who will emerge as winners remains to be seen, but the front-runners seem to be the film companies, who see in PACS the natural evolution of their film business; the imaging equipment companies, who see PACS as another “modality,” and the information system companies, who see PACS as a way to handle another form of data. However this competition works out, PACS will have an increasingly important role to play in medical imaging.

COPYRIGHT 1997 A Thomson Healthcare Company

COPYRIGHT 2004 Gale Group