The Disaggregation of Radiology

Article Outline

• Abstract
• Radiology's Changing Trajectory
• Teleradiology: Enabling, Disruptive
• Radiology's Business Model
• From Departments to Service Lines
• Implications for Radiology
• Conclusions
• References
• Copyright

The authors discuss certain market and political forces buffeting the traditional structure of radiology, both in practice and in the academic setting. These forces can be, to a certain degree, disruptive and produce fragmentation of what are now integrated radiology services and specialties. The potential fallout from the current rapidly changing environment of health care, including strategies for delivering care along service lines or within discrete episodes of care, may have a profound impact on the future of radiology. Understanding the dynamics of the current environment may help plan strategies for dealing with the potential impact on our specialty.

Key Words: Health care policy, radiology practice, radiology training

On August 22, 2007, the Advisory Board Company (a prominent hospital administration consulting group) [1] sent out the following message to its subscribers:

Faced with turf battles, St. Luke's separated neuroradiology group

In light of its growth, St. Luke's has struggled to staff an adequate number of neuroradiologists and neurointerventionalists for its acute stroke program. Since 1993, St. Luke's grew its number of neurologists from three to twelve and its neurosurgeons grew from three to seven, but maintained only one neuroradiologist. The decision stemmed from St. Luke's private practice radiology group's refusal to hire additional neuroradiologists to keep pace with the rapidly expanding volumes of the stroke institute, forcing St. Luke's neurologists and neurosurgeons to read their own patients' brain images after hours.

While Rymer [a hospital administrator quoted throughout the communication] notes that St. Luke's “crisis situation” of relying on one neuroradiologist was not unique, she says the hospital directly addressed its concerns by strategically moving its neurointerventionalists and neuroimagers out of the radiology group. As a result, the current, independent neuroradiology group allows neuroradiologists and neurointerventionalists more control over the development of services, while also ensuring a more equitable call schedule and distribution of revenue. In addition, this new group resolves neuroradiologists' dissatisfaction over taking more call than St. Luke's other radiologists, but not being compensated at a higher rate.…Based in part on the strength of its regional stroke network, St. Luke's has become the de facto neurosciences center in the region, with physicians from regional hospitals frequently referring any complex neurovascular case to St. Luke's.

The Advisory Board Company went on to point out that St. Luke's has not only gone on to improve clinical outcomes but has reaped an average 10% annual increase in stroke volumes. The hospital boasts an intervention rate of almost 30%, 10 times greater than the national average, for acute stroke admissions to the emergency room.

The implications of this vignette are significant for radiology. One of the traditional competitive advantages of radiology as a specialty has been the ability to gather multiple physicians with subspecialization interest into one business aggregate. However, as medical “service lines” are beginning to compete at the patient value level in teams that do not necessarily conform to a traditional (if not medieval), academically derived departmental structure, this competitive advantage is eroding.

A compelling argument for medical service integration and competition of integrated providers directed at the continuum of care of specific disease processes, such as neurovascular disorders (service lines), was made in the recent book Redefining Health Care: Creating Value-Based Competition on Results by Porter and Teisberg [2]. This text has become required reading for hospital administrators and has influenced health care policy planners. One of its main themes is that integrated practice units can optimize quality while reducing costs by promoting competition at the patient value level. A basic assumption made is that patients often require services around relatively discrete disorders. Given this, integrated practice units can provide and compete most efficiently for such services. Reimbursement structure, argue the authors, would be transformed to reward this strategy. Obviously, the historical, economic, and cultural underpinnings of our existing health care system present many roadblocks to the authors' vision. The practicality of their vision has been debated, and that debate is illuminating (and can be found online for any interested reader [3]). The arguments notwithstanding, there is a clear trend in medicine toward the greater integration of health care services into more efficiently organized teams, and this trend affects radiology significantly.

Back to Article Outline

Radiology's Changing Trajectory 

It is important to appreciate the health care industry's changing paradigms. Writing in the Harvard Business Review in 2004, McGahan [4] postulated that an organization's strategy cannot succeed unless it is aligned with its industry's change trajectory. That change trajectory is, to a significant degree, determined by two threats of obsolescence. The first category of threat is one aimed at the organization's core activities, a threat usually posed by new outside alternatives. The second category of threat is aimed at the organization's core assets, that is, its resources, knowledge, or brand, because of changes that diminish value.

It can be argued that the core activities of radiology as an organization include the production, interpretation, and distribution of quality imaging studies of patients. Because the production of quality imaging studies has become almost automatic through advances in technology, and because the interpretation and distribution of the information has been enabled by teleradiology, the need for point-of-service core activity by radiologists is diminishing. The threats to radiology's core activities include competition from cardiologists, orthopedists, and others who want to produce and distribute, if not interpret, images.

The core assets of radiology as a specialty are an independent consultative service based on whole-body knowledge, a brand name, early and deep knowledge of state-of-the-art technology, and, most important, visual skill and experience with in vivo pathology. However, the value proposition of these core assets is diminishing when integrated service units take over the strategy of patient care, as the independence of the subspecialized radiologists begins to wane and specific organ system knowledge begins to dominate the whole-body knowledge asset. Additional threats to the core assets of radiology include anything that diminishes the need for radiologists, imaging studies themselves, or specific imaging equipment or that supplants visual diagnostic capability. In addition to teleradiology, computer-aided diagnosis, genomics, and others can be added to the list.

Not surprisingly, the change trajectory in radiology is being experienced not only with the neuroradiology services at St. Luke's. Indeed, certain integrated neurologic institutes, such as Barrow's in Phoenix, Arizona, have sliced the pie even thinner. Interventional neuroradiologists, once part of an independent radiology group that had split off from its general radiology group a long time ago, are being supplanted by endovascular neurosurgeons, with only the diagnostic aspects of neuroradiology being left in the disaggregated, independent-contractor radiology sphere. Elsewhere, interventional radiology is joining vascular surgery and, in some instances, cardiology, in a separate integrated approach to cardiovascular and peripheral vascular disorders. Dedicated musculoskeletal imaging groups have developed over the past 2 decades, and some are being assimilated into orthopedic practices. Mammography, boosted by recent increases in reimbursement for digital diagnostic mammography and the upsurge in applications for breast magnetic resonance imaging and magnetic resonance-guided biopsy, likewise has begun to splinter from the traditional core of general radiology.

Historically, radiologists have aggregated for the purposes of covering numerous specialties, allowing efficiency of deployment within hospitals, particularly at night. After hours, in particular, the value of subspecialized radiology waned in favor of lifestyle factors (lesser frequency of night call). In a more recent trend, night call has begun to be outsourced. At first hastened by the same lifestyle decisions, and not threatened by the healthy margins in reimbursement, the practice of outsourcing night call has spread to approximately 25% of all US hospitals [5]. The development of “dayhawk” outsourcing was predictably not far behind. Indeed, a number of hospitals around the country have replaced on-site radiologists with a combination of outsourced services and locum tenens interventionalists, including Long Beach Memorial Hospital, one of the foremost institutions in southern California (and ironically the spawning ground of NightHawk Radiology Services). Another example, El Camino Hospital, was one of the “plum” practices for Stanford University and University of California, San Francisco, radiology trainees entering private practice in years past, but it recently outsourced its radiology department to a combination of a remote service, in-house cardiologists for endovascular intervention, and a hodgepodge of locum tenens coverage.

Back to Article Outline

Teleradiology: Enabling, Disruptive 

The threat of teleradiology to nonspecialized radiology groups has not gone unnoticed by the radiology establishment. A resolution presented at this year's ACR annual meeting reflects this:

that the ACR is concerned about a model of care where the radiologists are removed from (point of service)…the ACR regards care by on-site radiologists preferable to that of teleradiology, the latter being most useful as a supplement to on-site care for purposes such as subspecialty consultation, and to provide coverage for underserved areas where the physical presence of radiologists is not feasible.

This language still begs the question of sites where radiologists are present but underserving. The requisite service level at any facility may be in the eyes of the beholder. The authors of the resolution seem unaware that the train has left the station.

The enabling aspect of teleradiology, in fact, allows for 24/7 subspecialization of diagnostic radiology interpretation, more so than can be mustered by most aggregated, even partly subspecialized groups. Academic departments, such as those of the Cleveland Clinic and Massachusetts General Hospital, already provide this service, as do some private entities through wide-area distributive networks. This same concept can be used by clinicians, or the integrated practice units envisioned by Porter and Teisberg [2], to tie in distant subspecialty diagnostic radiology expertise to the point of service. Hence, the last line of point-of-service absolute need for radiologists is in the interventional sphere. The hands-on aspects of this subspecialty and its relatively broad bandwidth, from central line insertions to highly specialized procedures such as transjugular intrahepatic portosystemic shunting and aneurysm coiling, has been the loss leader for aggregated radiology groups. However, as reimbursement is lowered for diagnostic services, the generalist group's tolerance for less remunerative (in terms of reimbursement and relative value units) interventionalist partners has come under scrutiny. With the training of surgeons and cardiologists now including endovascular procedures and even some drainage procedures, the threat to these interventional core assets of radiology is even greater.

Back to Article Outline

Radiology's Business Model 

As daunting as this change trajectory is to clinical practice, the impact on academic training programs may be even more profound, because such programs really have nurtured and defined clinical practice. Indeed, the resulting challenges are perceived somewhat differently because of the additional responsibility of supporting the research and teaching missions on which our specialty relies. A traditionally strong competitive advantage of radiology has flowed from its academic fountainhead, providing it flexibility in adapting to new technology and shifting market dynamics. Radiology will again need to draw on those strengths. Experiments in organizational restructuring to deal with new realities ushered in by the major trends of digitization, globalization, patient-centric care, and the decentralization of organizations are well under way. These major trends are, of course, intertwined and are nudging radiology away from its previous, basic business model. If one subscribes to the axiom “structure follows function,” the disaggregation of radiology's organizational patterns follows the disaggregation of many of radiology's functions. A short summary of the traditional radiology business model, which academic departments fostered, provides some context.

The traditional business model was based on a geographically centralized radiology department (or imaging center) with the equipment, technicians, nurses and radiologists, and a data repository (the film library) all centralized. The first wave of change triggered by the microprocessor (early digitization) actually fortified this model, because computed tomographic and magnetic resonance machines were expensive, and centralizing them made financial sense. Combined with film, which needs a central archive, this business model was bolstered. Subsequent stages of digitization, however, began to disrupt this model by allowing more decentralization, although clustering expensive capital equipment still made economic sense. The digital communication revolution, in the form of the Internet, has brought globalization to the doorstep of radiology, and thus, the current phase of the digital revolution has begun to noticeably erode the old radiology business model. Moore's law,¹ Kryder's law,² and Metcalf's law³ continue to make imaging equipment and communication more powerful and less costly per level of performance. As the performance curve flattens out in relationship to clinical needs, prices will begin to drop. This phenomenon is quite evident in ultrasound, as illustrated by the dispersion of that equipment throughout the medical enterprise and the partial (so far) dismantling of the centralized business model. What will happen when we have $150,000 computed tomographic scanners and $300,000 magnetic resonance scanners?

Managing a large-scale imaging operation even with less costly machines still takes time, effort, and skill and favors at least some clustering of machines, even in a less central, more networked imaging service. Although there are several trends and countertrends, the centralized radiology department as a business model is morphing and is capable of supporting service-line approaches. The physical decentralization into a network architecture facilitates horizontal integration, which, however, can drive traditional radiology organizational disaggregation. With more standardized interfaces, radiology is already well on its way to greater horizontal integration.

Back to Article Outline

From Departments to Service Lines 

The service-line approach to patient care has been around for quite some time, although it is gaining new followers. It is gaining momentum in health care as it aims to redesign professional-oriented workflow to a more patient-centered one. This reflects more consumerism in health care, a trend likely to continue. The issues raised by organizing around service lines compared with around subspecialty silos are not new. The business literature has discussed this dilemma in its jargon between a functional and a product (service) line structure for decades; that discussion is well summarized in Porter and Teisberg's [2] Redefining Health Care. The product (service) line approach has become dominant except in areas with rapid technology changes. The MBA answer on how to organize is along dimensions that are most important to the customer. In health care, that is not always the case, because the customer, the consumer, and the payer may be 3 different entities, and because professionals, as defined by their academic sociology, have dominated structural decisions, some would argue selfishly so. Physicians, particularly, have been historically endowed with a mantle of cultural authority, which at its extremes runs counter to the service-line model. Today, however, we are entering the mode of constraining the convenience of professionals to improve the convenience of patients.

Although not ideal, the compromise solution for balancing the need of subspecialty skill and the need for customer-oriented service is a matrix organization. This balance is one of the paradoxes in managing an organization, and paradoxes have no solutions. As Charles Handy [6] noted, paradoxes are not managed in the sense of “controlling” but in the sense of its original meaning of “coping with.” This paradox is particularly acute for academic radiology departments, which are expected to train subspecialists for other organizations' service lines, while such training requires deep subspecialty silos. This will pose a conundrum and a strategic challenge as their parent organizations adopt service lines. Because medical care is not getting simpler, the need for subspecialty knowledge and care will likely increase. This echoes Adam Smith's fundamental observation of increased productivity resulting from the division of labor, (ie, specialization), as more recently amplified by Peter Drucker [7]; Management paradoxes are here to stay.

Back to Article Outline

Implications for Radiology 

So where do all these trends and countertrends leave us? Fundamentally, radiology needs to reflect on its real role in health care. Diagnostic radiology's real role is the information business. The current value proposition is the provision of information for medical and therapeutic decisions on the basis of human imaging data. This information facilitates decision making on the part of physicians and patients just in time at the point of service, or in real time for interventionalists and surgeons. The processes in this information business are disaggregating.

To better understand this disaggregation, a simplistic model of radiology processes is helpful: image acquisition → image processing → image interpretation → information transmittal → medical decision. The first step, of course, is managing and understanding the acquisition device, and as noted above, that was one of the key features (high price, high maintenance, and high operating costs) of the centralized business model. Equipment is getting more sophisticated and simpler simultaneously. The trend is not uniform. The next step is processing raw data into relevant images. This may occur in the machine, in a workstation, or in India. It is clearly disaggregating. The next step of interpreting these image data relies on the radiologist's visual system and knowledge base. It takes training to interpret visual data, and that will remain a radiology core competency with computer-aided diagnosis as an adjunct. It will be some time before a radiologist's trained visual system is replaced by a computer, but it won't be impossible (see The Singularity Is Near, by Kurzweil [8]). The nearer term reality is that it can be outsourced, and the other possibility is that other subspecialists' visual-cognitive system may do it quicker, for lower cost, higher quality, or all of the above, improving on patients' current value proposition. The reader would do well to grasp the basic message of Clayton Christensen in his books The Innovator's Dilemma [9] and The Innovator's Solution [10].

For image data to be converted into valued information for decision making requires an increasingly broad fund of knowledge to be integrated during image interpretation. This is an opportunity and a threat for radiologists as this fund of knowledge grows in bandwidth and importance. The “whole-body knowledge asset” remains important but will not be sufficient going forward. It is here where radiologists will need to build multidisciplinary bridges to access and understand growing, relevant, but unfamiliar bodies of knowledge (genomics, proteinomics, molecular medicine, computational biology, nanotechnology, etc). Service lines provide the opportunity for radiologists and other subspecialists to acquire accretive knowledge bases to make their information business more valuable. However, a radiologist totally isolated in the service line without stimulation by or access to deeper levels of technical imaging expertise may slowly lose a differentiating edge that adds value to the information. There is no reason that a neuroradiologist cannot be as useful to an oncology service line as well as to a stroke service line. Granularization is not a given. Thus, subspecialty academic training of radiologists may still run parallel, not counter, to the strict service-line mentality. This dilemma of being part of a service line while maintaining subspecialty expertise will not be unique to the academic training of radiologists. The disaggregation of many processes in the radiology information business contributes to the organizational disaggregation, as does the dispersion of knowledge bases necessary for making interpretations truly informative. A challenge will be to maintain social capital in any new network structure for radiology.

Another major challenge to training radiologists is the closer association or even fusion of imaging with therapeutic procedures. This crosses many traditional subspecialty lines. Although interventional radiology may have had an unsteady home in radiology departments, the importance of image-guided therapy will increase. Although not as profitable as large-scale magnetic resonance and computed tomographic operations, interventional radiology provides very important direct links to patient care. This is critical in a patient-centric environment. A core competency in this therapeutic approach is an understanding of image data (especially in real time) in conjunction with minimally invasive therapy techniques. This can play to the strengths of radiology, even though there is certainly competition from other subspecialists. Success in this arena, however, requires a modification of radiology culture to include effective, direct patient care. Reimbursement models will strongly affect the organizational place of radiology-trained interventional radiology specialists (note, however, that such specialists will require academic radiology training, even if they are surgeons or cardiologists).

Back to Article Outline

Conclusions 

In short, the digital revolution, globalization, consumerism, and market and government forces have changed the landscape of radiology forever. This landscape remains unsettled, and hence, radiology, though enjoying unprecedented success, may be at a tipping point. Although unsettling, it is time for radiology to reflect on its future contributions to the medical care system: the provision of information for medical and therapeutic decisions on the basis of data in a broad array of imaging formats combined with funds of knowledge that now must extend far beyond its previous whole-body knowledge asset. Information is made valuable by being fast, accurate, accessible, and relevant at the right time.

Radiology will learn to manage in the sense of coping. But how? Here are some general directions. Radiologists must assume greater responsibility for managing the entire information business and dispose of the “film reader” mind-set. As a subspecialty, radiology will need to be innovative in integrating imaging data and new funds of knowledge in a fashion similar to the transition of molecular biology to systems biology (ie, develop systems radiology). Another tack to be considered, again analogous to a branching in biology to computational biology, is consideration of computational radiology. This refers to the ability to “mine” the huge and growing image databases and integrate them with genetic, biologic, and drug databases, among others [11]. This integration makes the information more relevant. The analogy to biology is appropriate because diseases (and health) are being redefined on the basis of greater biologic granularity. The information business follows the daunting laws of Moore, Kryder, and Metcalf and so presents opportunities and threats, as it does in all other informational businesses [12]. Finally, interventional radiology needs to be fully integrated in the service-line model and be seen as pivotal to direct patient care via frontline therapy, be it with radiologists or other specialists trained in the academic underpinnings of imaging technology.

Although it sounds a bit nebulous, it will behoove the specialty to seek an overarching mission or vision that could serve as an organizational adhesive. Attractive compensation levels without night call are unlikely to serve that purpose. Academic radiology does have the advantage of being part of institutions with stated missions that provide an organizational glue. However, even glue can harden to the point of cracking. New adhesive formulations may be necessary to transform medieval guild-based departmental systems into ones less likely to crack under the stresses of complex architectural transformation. In the competitive information business, isolated stragglers will lose out. On the other hand, a robust information business can support service lines and even become one itself. One thing is clear: radiology, in both the academic and the community practice worlds, will need to explore and invest in innovative organizational forms.

Back to Article Outline

References 

1. Advisory Board Company. Content to subscription members only. http://www.advisoryboardcompany.comAccessed June 4, 2008
   • View In Article
2. Porter ME, Teisberg EO. Redefining health care: creating value-based competition on results. Cambridge, Mass: Harvard Business School Press; 2006;
   • View In Article
3. Porter ME, Teisberg EO. Health reform: redefining health care. http://healthaffairs.org/blog/2007/03/23/health-reform-redefining-health-careAccessed June 4, 2008
   • View In Article
4. McGahan A. How industries change. Harv Bus Rev. 2004;82:98–106
   • View In Article
   • MEDLINE
5. NightHawk Radiology Services. Home page. http://www.nighthawkrad.netAccessed June 4, 2008
   • View In Article
6. Handy C. The age of paradox. Boston: Harvard Business School Press; 1994;
   • View In Article
7. Drucker P. Post-capitalist society. New York: HarperBusiness; 1993;
   • View In Article
8. Kurzweil R. The singularity is near: when humans transcend biology. New York: Viking; 2005;
   • View In Article
9. Christensen CM. The innovator's dilemma: when new technologies cause great firms to fail. Boston: Harvard Business School Press; 1997;
   • View In Article
10. Christensen CM, Raynor M. The innovator's solution: creating and sustaining successful growth. Boston: Harvard Business School Press; 2003;
    • View In Article
11. Ayres I. Super crunchers: why thinking-by-numbers is the new way to be smart. New York: Bantam; 2007;
    • View In Article
12. Kessler A. The end of medicine: how Silicon Valley (and naked mice) will reboot your doctor. New York: HarperCollins; 2006;
    • View In Article

• 1 Microprocessor power (in transistors per square inch) doubles every 18 months.
• 2 The storage capacity of hard drives doubles every 2 years.
• 3 The value (performance) of a telecommunication network is proportional to the square of the number of nodes on the network.