In this month’s AJNR, Dr. Loevner and her colleagues at the University of Pennsylvania have potentially stirred themselves up some controversy. They have called attention to a problem that is very well known to us in head and neck radiology, and indeed all radiologic subspecialties in tertiary oncologic centers: the high frequency of significant reinterpretations of outside imaging studies. This is such a problem that nearly every outside head and neck imaging study (that are not otherwise inadequate) performed elsewhere on a patient referred to our institution is submitted for reinterpretation. We now see that these errors can have a significant impact on staging, treatment, and prognosis. And this work from the University of Pennsylvania is the product of a single institution; one can only imagine the numbers throughout the nation’s institutions, as well as those cases in which the imaging is never reinterpreted by someone of Dr. Loevner’s skill and experience. Clearly, there is a major problem in the interpretation of head and neck studies by radiologists in this country.
What are the root causes of this problem and are there any solutions? First, to the root. Head and neck radiology is a relatively young (and difficult) sub-specialty. Resident training in head and neck radiology is limited, and many programs have no dedicated head and neck radiologist on staff. As a result, residents and fellows in these programs may see very little head and neck imaging or receive inadequate training. This may help explain the fact that in the Loevner study, misinterpretations were as or more prevalent from referring academic centers than the private practice setting. No formal head and neck radiology fellowships currently exist, although head and neck training is a required component of ACGME approved neuroradiology fellowships. Interested radiologists may arrange private, but usually limited visiting fellowships with several senior members of the ASHNR, but these are generally informal. I know several now-prominent head and neck radiologists who started this way; many others simply trained themselves or established their qualifications through experience and publishing.
Another issue is that very few radiology practices see enough head and neck imaging to allow any one individual to gain enough experience to be fully comfortable with it. Also, as mentioned in the Loevner article, the radiologist designing or interpreting a head and neck scan often lacks sufficient clinical history to acquire the appropriate images, much less interpret them accurately. To properly protocol and interpret a head and neck imaging study, one requires not only a basic knowledge of the disease process and physical findings, but also an understanding of the post-treatment appearance of any surgical, chemotherapy and/or radiation therapy the patient might have previously undergone. This makes the radiologist a useful member of the treating team. To gain this knowledge takes commitment and study, which is facilitated by involvement in a multi-disciplinary head and neck tumor board.
Finally, though not mentioned in the Loevner paper, interpretation is inextricably linked to the issue of image quality. For many of the reasons already mentioned, the quality of head and neck imaging studies referred to tertiary care facilities is often poor (1). Although a study of image quality is necessarily more subjective than studies based on histology and other objective measures, I once attempted to quantify the shortcomings of outside head and neck CT examinations referred to our institution (1). This study showed that the overwhelming majority of outside scans on patients referred to our Head and Neck Surgery clinic were deficient in at least one and usually several critical aspects (windowing, contrast bolus, gantry angulation, etc.). If a head and neck scan is improperly performed, correct interpretation is much more difficult, if not impossible. Furthermore, poor image quality directly results in increased health care costs, due to the frequent need to re-image patients whose initial scans prove to be inadequate. This is a daily occurrence at our facility.
Misinterpretation of head and neck imaging studies is so prevalent, that in my experience, some radiologic diagnoses such as the perineural spread of head and neck cancer, are far more commonly missed than made. It may even be that the standard of care is actually to miss perineural tumor spread radiologically.
Are there solutions to this problem? It is unlikely that there will be any sudden increase in the prevalence of head and neck disease to provide radiologists with the necessary experience in interpreting these studies. One move made by the American Board of Radiology, to make Head and Neck an equal part of the Neuroradiology CAQ (certificate of added qualification) examination, was an excellent step toward encouraging radiologists to become more proficient at head and neck imaging. Apparently, however, this is not enough.
In regard to insufficient clinical information, one solution is to defer reading any case for which history is unavailable, until the ordering physician can be contacted. In the rare case that our computer system lacks relevant history, I routinely hold off protocolling or interpreting a case until I can speak to the head and neck surgeon. Another solution is education, something to which the ASNR and ASHNR are committed. The ASHNR annual meeting is primarily focused on education, with review talks on virtually every aspect of head and neck imaging. This is an open meeting to which all radiologists are invited. The ASNR annual meeting always has a large amount of head and neck programming. In addition, the RSNA and ARRS always have refresher courses in head and neck imaging. For the past four years, there has been a refresher course at the RSNA on techniques and pitfalls in head and neck imaging (2); the ARRS more recently began such a course (3). Finally, at the ASNR.org web site, an online CME module “ASNR eCME” (http://www.asnr.org) sponsored by the Neuroradiology Education & Research Foundation, includes two fine introductory presentations on staging of head and neck neoplasms. Available free to all members of the ASNR, these are an easy and available means to increase one’s competence in this field. Obviously, such educational efforts are successful only to the extent that practicing radiologists avail themselves of the opportunity.
It remains to be seen if and how the advent of teleradiology and rapid dissemination of images will allow better sharing of cases between radiologists in practice and those with greater interest and expertise in head and neck. Would such an over-read service be overwhelmed by those desiring help, or would it sit idle because radiologists will not recognize or acknowledge a need to send these tough cases (and, presumably some revenue) to others?
What of the lack of formal training programs in head and neck radiology? Perhaps the time has come for a concerted effort in this direction. The ASHNR should directly address this issue, with an eye toward the establishment of at least some dedicated head and neck radiology fellowship programs. Perhaps head and neck could be emphasized in the second neuroradiology fellowship year. Dr. Loevner has reminded us all of a serious deficiency in radiology, one that will continue unless addressed strenuously. Rather than opening a can of worms, let this serve as a wake-up call, a call to action, one that all of us in head and neck radiology should act on, in the interest of the patients and physicians we serve.
- Copyright © American Society of Neuroradiology