During the meeting, an analysis of the FDA's medical device accident database, MAUDE (manufacturer and user facility device experience), was presented that showed that reported MRI accidents in the US have risen over 500% from 2000 to 2009. During that same time period, according to market research firm IMV, the overall number of administered MRI exams increased by approximately 90%. So if accident rates grew at more than five-times the utilization rate, in what ways do MR safety protections need to be strengthened?

It turned out that there is disagreement as to what the growth numbers actually represent. Some presenters at the meeting provided their individual or company experience showing that internally reported accident rates have remained steady over the time period in question. They hypothesized that the growth in the FDA's database was more a function of increased reporting, and not an actual increase in the number of accidents. (It is widely acknowledged, even by the FDA, that reporting rates are abysmally low, leaving little room for any change but increased reporting.) However, even among those who suggested increased rates of reporting, no examples could be identified of any advocacy, either within or towards the MRI community, to improve reporting rates.

The animated debate about the accuracy of the FDA's accident numbers was a pointless exercise, however, since there is no alternative source of nationwide MR accident data. We can acknowledge the database's failings, we can try and improve the utility of data collection going forward, but for the time being, it's the only data source that transcends states, vendors and provider type. It isn't perfect – by a long shot – but it's the best we have. To ignore what it offers would be foolish.

Key causes

Delving into the FDA's data, even one year's worth of adverse event reports included a wide range of events, from the absurd to the catastrophic. The overwhelming majority of reported injuries from the last couple of years' data fell into one of three categories: burns, projectiles and hearing damage.

Burns Burn events mostly fell into one of three etiologies: burns caused by electrically conductive materials in the bore with the patient; burns caused by proximity to/contact with active RF elements; and resistance to electrical currents generated in the patient's own body as a result of large calibre tissue loops.

Projectiles Similarly, projectile events were the product of several conditions. These included: not being aware of the persistence of nearly all MRI systems' magnetic fields; forgetting about metallic materials that were brought or carried into the magnet room; mistakenly assuming that a metallic object was nonferromagnetic and therefore not a projectile risk; and even knowing that an object was a projectile risk, yet assuming that the subject could maintain a critical distance from the scanner to prevent an accident.

Hearing Damage Temporary and even permanent hearing damage have been claimed in several reports filed with the FDA. In 2009 and 2010, the majority of reported hearing damage events included confirmation that the MRI provider had given hearing protection to the injured patient. It is not clear as to whether the patients were given the foam ear inserts and ignored them, or whether they were inserted incorrectly in the patients' ears. For individuals who don't know how to properly insert these plugs, or can't differentiate proper versus improper fitting, simply being "provided" with hearing protection is insufficient.

This suggests an area in addition to patient screening in which compliance needs to be monitored. In my opinion, MR staff should instruct patients on the proper way to place these plugs, explain how to tell if they're seated correctly, and offer over-the-ear muffs for patients who struggle with the plugs.

Avoidable accidents

Unlike the current row in the European Union over the Physical Agents Directive, the US debate on MRI safety is based upon actual accidents and injuries directly attributable to MR environments. Instead of debating the merits of limiting exposure to magnetic fields based on dubious claims of hazard, the US conversation is looking at the MRI injuries that are clearly documented, and accidents that could clearly produce injury.

The interesting thing, to me, about these reported injuries is that they each represent known risks of MR imaging, each with established means of mitigating the hazard. These aren't device malfunctions. These aren't surprises (or ought not to be). And while newer MR system designs are advancing sound dampening or active noise cancellation to reduce the likelihood of hearing damage, the physical forces that can produce burns and projectiles appear to be inescapable elements of MR imaging. But saying that the forces are inescapable is not the same as saying that accidents are inevitable.

The FDA conference also saw data presented from a research project that I completed with Emanuel Kanal, chair of the American College of Radiology's (ACR) MR Safety Committee. In that project, we compared every clinical MRI accident recorded in the FDA's database in 2009 and 2010 against specific, explicit criteria for accident prevention, as described in both the "ACR Guidance Document for Safe MR Practices: 2007", and the Joint Commission's Sentinel Event Alert #38: "Preventing accidents and injuries in the MRI suite".

Our analysis found that the overwhelming majority of accidents (65% from Sentinel Event Alert #38 and 85% from the ACR Guidance Document) would have been mitigated by explicit performance criteria contained within each standard.

The maddening fact is that we know what causes accident and injury in the MRI environment, and we have distilled the worst offenders down to a short list responsible for some of the most common and most severe accidents. For each of these, we have mitigating standards that could dramatically reduce the injuries. Yet all of these safety standards are almost entirely unenforced by the accreditation organizations and licensure officials.

The one repeated refrain that came from the FDA meeting, offered by professional societies, companies and individual speakers, was a call for a uniform set of MRI safety standards to apply to all providers, such that a minimal level of safety in the MRI physical environment is assured to both patients and caregivers.