Zaidat et al.¹ understated the accelerating crisis of physician oversupply in our field. With 80–100 US training programs currently graduating 100 new neurointerventionalists each year,² the number of practicing neurointerventionalists will easily double by 2020.

The authors reported that 95% of the US population is now “adequately covered” by neurointerventional services. Suzuki et al.³ demonstrated that 99% of the US population was within 200 miles of a neurointerventionalist, based on The American Society of Interventional & Therapeutic Neuroradiology membership rolls from 2002. Thus, new graduates will continue to overpopulate areas already adequately covered by neurointerventional services.

This growth in the number of physicians and programs decentralizes care and reduces volume at centers of excellence. Care could worsen because it has been shown that outcomes are better with increased case volumes and operator experience.^4,5 Furthermore, competition for cases places undue pressure on new and low-volume operators to treat patients with marginal indications. This same pressure is a major disincentive for inexperienced operators to transfer complex cases to regional centers of excellence that they view as direct competitors. The continued overtraining of neurointerventionalists represents an impending disaster for our field and our patients. We have created this problem ourselves, so we need to recognize it and stop perpetuating it. Until systematic measures can be enacted at a societal level to standardize training and appropriately match the number of trainees to the demand for services, all neurointerventional fellowship training should be stopped.

Author Response

Drs. Fiorella and Cloft raised important points about neurointerventionalist manpower in the United States in their WriteClick submission and recent article.² We agree with them.¹ Very few US hospitals have adequate neurointerventional procedural volume criteria as recommended by professional societies.⁶ However, there are no strong manpower data on neurointerventionalists or procedure numbers. In general, this has been a limitation of manpower studies in medicine.⁷

To address the concern of oversupply, urgent solutions may be considered:

1. Neurointerventional training should be increased to 3 years, including diagnostics and neurointerventions. Currently, training is inconsistent (1–2 years). A minimum volume should also be required (e.g., 200 angiograms and 200 neurointerventions).

2. For the annual Fellowship Match Program, the number of fellows per year for any given program should be based on case volume and number of faculty.

3. Board certification should be required from the American Board of Radiology or American Board of Neurosurgery, multisociety (Society of NeuroInterventional Surgery [SNIS]/Society for Vascular and Interventional Neurology [SVIN]/American Society of Neuroradiology/American Association of Neurological Surgeons/CNS Cerebrovascular Section), or United Council of Neurological Subspecialties.

4. Multisociety guidelines (SNIS/SVIN) should establish annual numbers of neurointerventional procedures for skill maintenance. For example, the guideline for interventional cardiologists is 75 percutaneous coronary interventions per year per operator.⁸

5. Experience from cardiology showed that mandated interventional cardiology training programs’ accreditation by ACGME and mandated board certification of the trainees reduced the number of graduates by 50%.⁸

An immediate call to action should consider the above recommendations not only to address manpower but also to meet our societal responsibility for future, high-quality neurointerventionalists.