Guidelines for Diagnostic Neuroangiography: A Model to Emulate from a Neuroradiologist's Perspective

The Quality Improvement Guidelines for Adult Diagnostic Neuroangiography published in this issue (page 146) is an extremely important document for three reasons. First, it is the result of the joint efforts of the three radiologic subspecialties whose members perform the overwhelming majority of these procedures. Second, the criteria focus on quality of procedural performance by any physician undertaking the procedure, rather than credentials based upon training. Third, the standards are based upon data derived from the literature and evaluated by a group of experts in procedural performance to reach a consensus. Each of these points is significant and deserves further explanation.

Members of the three societies who have authored this document perform the most neuroangiography, so the experience and expertise needed to review and evaluate the literature is represented. The members of the American Society of Neuroradiology (ASNR) and the American Society of Interventional and Therapeutic Neuroradiology (ASITN) deal with problems of the brain and the spinal cord, and the blood supply to these organs, on a daily basis. Using a variety of procedures, including neuroangiography, their professional careers are devoted to imaging these organs. The members of the Society of Cardiovascular and Interventional Radiology (SCVIR) also perform catheter-based procedures on a daily basis, including a large number of neuroangiographic studies, especially in the nonacademic world. Members of the SCVIR have had extensive experience developing practice standards and guidelines, and have pioneered the quality improvement (QI) approach in radiology. Rather than one society producing a parochial standard by excluding the input and experience of the others, these three societies have joined forces to produce a document that has one goal: the assurance that this procedure is performed at the highest level possible, for the benefit of patients.

The training of the radiologist to perform neuroangiography properly is highly variable, whether or not the individual is a member of one of the three mentioned societies. In addition, members of a number of nonradiologic specialists currently perform, or would like to perform, neuroangiography (possibly to accompany more complex therapeutic procedures), including cardiologists, vascular surgeons, neurologists, and neurosurgeons. Each group feels its members have a “right” to perform neuroangiography, based upon often poorly documented “training”. Some have even composed self-serving documents to substantiate this “right”. The QI approach changes the “game” from one of “training” and credentialing to one of actual performance.

Most institutional credentialing committees will surely require some documentation that a practitioner has been trained to perform neuroangiography. As the procedures are performed, however, data will be collected. Each practitioner must adhere to specified thresholds of indications, success rates, and complication rates. If these thresholds are exceeded, a review will be initiated. Improvement is the goal, but loss of privileges is the risk. A practitioner no longer will be able, whatever the specialty, to advocate his/her “right” to perform neuroangiography. The practitioner will be forced to quote his/her own complication rates to a patient, instead of those from the literature, so that the patient may make a meaningful decision regarding risk and benefit. Institutional credentialing committees will not be faced with debilitating “turf wars”; the numbers will speak for themselves.

We hope this methodology will become a model for other procedures and interventions, both within and outside of radiology. The Joint Commission on the Accreditation of Health Care Organization (JCAHO) has already established that the QI methodology will become the “measuring stick” of the future. Therefore, it is incumbent upon us to become familiar with the techniques and to begin producing our documentation. This methodology will also protect our specialties from those who would attempt to undertake procedures that we practice so diligently. We should not to be afraid of displaying and documenting our expertise; it is to our ultimate benefit.

What about the methodology used and the numbers advocated as thresholds? The authors have used a modified Delphi approach, requiring them to review the literature in order for experts to reach a consensus on reasonable standards. This is not pure science; it is not a meta-analysis only. It is based upon experts examining the literature, evaluating its quality, and reaching an agreement. The thresholds for indications and success rates are straightforward and need no explanation. The definitions for complications and their thresholds, however, require discussion.

The non-neurologic major complication rates are rather generous, in my opinion. The amount of contrast medium used for the average cerebral angiogram is below that used for an enhanced CT of the brain. Most angiographers use a nonionic agent, so renal failure is probably very rare after neuroangiography. Arterial injuries requiring intervention, or persistent bleeding requiring transfusion, are exceedingly rare in my experience.

The neurologic complication rates require the most scrutiny. Some may consider the definition for a transient ischemic attack (TIA) rather broad. Imaging studies, such as diffusion MR imaging, have changed our concepts of TIAs and small strokes as well as of reversibility and irreversibility of injury to tissue. A TIA is probably a rather short event, lasting minutes to a few hours. Nonetheless, the authors have grouped TIAs and reversible strokes together, so the issue of definition is moot. The allowed rate of 2.5% is generous.

The most feared complication of neuroangiography is a permanent neurologic deficit. A rather wide range is cited from the literature, and the authors have wisely chosen a value on the lower end of this range. This is the complication that must be the most closely evaluated, because the consequences are the greatest. A rate of 1% is generous; a good practitioner of this admittedly complicated procedure should never exceed such a number. How “generous” is this rate? A recent article by Cloft et al (1) describes a meta-analysis of three recent prospective studies of complications among patients with subarachnoid hemorrhage, cerebral aneurysm, and arteriovenous malformation who underwent cerebral angiography. There was a permanent neurologic complication rate of 0.07%. This series did not include patients undergoing angiography for cerebrovascular occlusive disease, which is the group, in my experience, with the highest neurologic complication rate. The low rate of permanent neurologic complications, however, supports the generosity of the rate in this QI document, a rate that any experienced angiographer should meet easily. It should be emphasized that the authors list thresholds only as guides for an institution. Case mix and other factors may require specific QI rates to be altered.

In summary, the authors are to be congratulated for presenting a number of models to us. Neuroangiographers from multiple societies, in a model of cooperation, were willing to focus their efforts to produce an excellent document that has quality patient care at its heart. It is a model for the future development of guidelines for other image-guided procedures. We hope it will be a model for the rest of medicine.