One trend, in medicine, is for primary-care and specialty physicians to use ultrasound—and, perhaps, other imaging modalities—in their practices. From my perspective (and with 20 years of experience as an imaging and hospital administrator and consultant, US medicine is evolving away from fee-for-service reimbursement and toward outcomes-based payment for patient care. Both imaging and the way we use imaging are rapidly changing, and we should think about how to use imaging in the most effective manner.
Only 25 years ago, both CT and MRI were rare and in their infancy, ultrasound was crude, and the vast majority of radiology involved plain radiography or some variant of radiographic technology. It took the development of inexpensive computing power to drive the radiology technologies that we have today. Because of the relatively limited scope of radiology in 1980, however, the primary-care physicians and specialists of that time could have a reasonable command of which imaging study to order in a particular episode of clinical care.
Advances in imaging are largely the result of digital imaging, and they have roughly paralleled (and been facilitated by) increases in computing speed. Digital images have the unique ability to be analyzed using algorithms, and they also provide for manipulation to improve the accuracy of interpretation. An aspect of increasing imaging capability is, however, the need to recognize when a specific modality is appropriate for use. New technologies do not always make a material difference in the diagnosis of a condition, and this implies a role for the radiologist in helping referring physicians to determine the appropriate imaging for a given condition.
We can no longer assume that referring physicians are fully conversant with which imaging study or procedure will accomplish their diagnostic goals. Moreover, we cannot assume that referring physicians will accomplish those goals in the most economical manner; this is becoming increasingly important, as medicine moves toward accountable care.
The increasing complexity of imaging dictates that radiologists assume the more important role of diagnostic consultants. In this role, they should be expected to both understand the technology armamentarium and to be prepared to assist referring physicians with the selection of the appropriate imaging tool to use for a given patient and diagnostic goal.
Unfortunately, economic, regulatory, and medicolegal constraints in the current fee-for-service environment might make this new role counterintuitive for practicing physicians. We cannot, however, make progress toward a shared goal of providing care that is more effective (both clinically and from a cost perspective) unless we make this change.
A common thread running through all of imaging, in its relationship with clinical medicine, is the shift to a more active role for radiologists in the development of evidence-based imaging pathways that can be integrated into medicine. Augmenting this process is the need for radiologists to be more available to listen to clinicians’ diagnostic problems and to recommend an appropriate imaging course.
Protocols are often viewed with a good deal of suspicion by physicians—in particular, by physicians who have a concerned patient in front of them. It is easiest, in many instances, to begin at the end and throw the most advanced imaging available at the diagnostic problem, in the name of applying the so-called art of medicine to the totality of the patient’s circumstances.
The art of medicine is a very valuable concept, but it is possible both to use science (as expressed by evidence-based protocols) and to allow nuancing of the evidence by the physician. Process control is found in documentation, measurement, and some system of accountability that provides periodic review of outcomes. Physicians should be allowed to depart freely from the protocol on an individual-patient basis, but should provide documentation of the reasons for such departures, and the results of such actions should be reviewed in the context of patient outcomes.
A key aspect of any protocol is buy-in by those who are charged with applying the protocol, and this is only achieved by allowing the physicians who are affected on the referring side to have a voice in the process. I advise selecting only physicians with clinical credibility among their peers, but they must also be physicians