A child was brought to the Magnetic Resonance Imaging (MRI) room for a brain scan. Accompanied by an anesthesiologist, the child was receiving sedation for the MRI via an infusion pump with a long IV tube. The anesthesiologist was aware that the pump needed to be kept away from the magnet. The pump was placed 10 to 15 feet away from the MRI magnet on top of a garbage can, as is the practice at the hospital—no bracket is used to secure the pump.

When the scan was completed and the patient was to be wheeled out from inside the scanner, the anesthesiologist brought the pump to the foot of the bed to secure it to a bracket there. However, the child made an unusual noise, which caused the anesthesiologist to turn around suddenly. As he did so, the pump flew out of his hand and hit the magnet, which is always on. The impact damaged the pump, but the child was unharmed.

Root Cause Analysis later revealed the following background information: Metal items are kept outside of the MRI room, but an exception is made for infusion pumps, which are allowed inside the room at a safe distance from the MRI magnet. However, at this hospital, no bracket is used to secure the pump, and no markings are present on the floor or elsewhere to indicate what is considered a safe distance. Following this event, the hospital has committed to purchasing MRI-safe pumps and installing brackets in MRI rooms to secure the pumps. Additional staff education and posted warnings have also been put in place. Other solutions such as using metal detectors and double-checking people before they enter the MRI room are also being considered.

by John Gosbee, MD, MS, and Laura Lin Gosbee, MASc

This incident was a serious “close call” that reveals a number of vulnerabilities associated with the use of MRIs. Although the child was unharmed, this incident might easily have produced significant injury. As such, the health care organization needs to analyze and act upon this incident to avoid harmful outcomes in the future.

Clinicians are well acquainted with the risks that MR scanners pose to patients with implanted metal devices, such as pacemakers and aneurysm clips. However, awareness of this hazard has not completely prevented patient deaths.(1) Although clinicians may regard the magnetized projectile described in this case as a bizarre and unlikely hazard and thereby dismiss its potential threat, by talking with almost any MRI technician, they will learn that projectile incidents are far from rare. Three illustrative cases from the Food and Drug Administration’s Medical Device Reporting system are listed below:(1)

• “Two steel tines (parts of a forklift) weighing 80 pounds each were accelerated by the magnet striking a technician and knocking him over 15 feet resulting in serious injury.”
• “A pair of scissors was pulled out of the nurses hand as she entered the magnet room. The scissors hit a patient causing a cut on the patient’s head.”
• “A patient was struck by an oxygen bottle while being placed in the magnet bore. The patient received injuries requiring sutures.”

Moreover, in a widely reported case, a child died in July 2001 at a New York hospital when a ferrous oxygen cylinder struck the child’s head after the cylinder was inadvertently brought into the MR room. Other databases document additional cases (2), and one report in the literature describes five cases at a single institution.(3)

Before any discussion of systems approaches to this and all cases involving projectiles, one needs to know about other hazards caused by the permanent (always on) strong magnetic field.(2,4) In addition to turning ferromagnetic objects into projectiles, the MRI magnet may also cause injuries by allowing an item to encircle or crush the patient. For example, a so-called “sand bag” that encircles the ankle to hold a leg in place may contain ferrous material. Also, the magnetic field can cause certain devices to malfunction, including IV pumps (by reversing flow after the magnet causes the pump to rotate in the opposite direction). Some implants, clips, or shrapnel (ie, in veterans) within the patient’s or caregiver’s body can “twist” to align with the magnetic field. This twisting can be disastrous, if the forceful movement and position of these objects is such that they crush or pierce vital tissues or organs. The magnet can also introduce a current into innocent-looking loops containing heat-conducting material, which can cause full-thickness burns in unconscious patients (eg, ECG leads). In addition, several items found in health care delivery systems can cause artifacts in the scanned image. The manufacturers of medical devices are required to test and provide data as to which items are incompatible with MRI activities.

Compounding these hazards is a set of characteristics unique to MRI and the difficulty in identifying safe and unsafe items. Within an MRI room, a large, invisible magnetic field is present. Without obvious cues regarding the location and intensity of this permanent magnetic field, it is relatively easy to carry in items that appear safe but in fact contain ferromagnetic material (such as the iron-containing sand bags mentioned earlier).(5) In other words, equipment and consumables that are “safe” 99% of the time may become “unsafe” near MRI. When this happens, it is neither fast nor simple to emergently shut down (quench) the magnet. Replacing the liquid helium and providing maintenance to the quenched magnet can cost $20,000 or more per episode. FDA-sanctioned labeling of items with “MR safe” and “MR compatible” can be confusing. MR safe items are those that will not be attracted to the magnet. “Compatible” means the item is both safe and will not cause artifacts in the scanned image. However, both labels apply only to certain strength magnets. Given so many hazards, it is impossible to create simple lists of safe and unsafe materials.

To prevent further projectile hazards in the MRI room, this hospital and others like it must implement several changes. Interventions that might eliminate the vulnerabilities associated with MRI have yet to be discovered (or are currently not technologically feasible). With this in mind, all interventions should be viewed as a single piece of the puzzle, and most will be rendered more effective when applied in organizations with strong cultures of safety. Some remedies (such as checklists and double-checks) are necessary but insufficient. More effective is a philosophical shift in which the MRI suite is thought of as an “isolation” room, akin to other rooms in which one takes precautions upon entering due to the presence of hazardous organisms (such as Ebola virus) or radioactivity.

There is little evidence regarding the most effective interventions to prevent incidents such as the one described. We believe some moderately useful interventions for this particular case would include:

• Restrict the items and people entering and exiting to those that absolutely need to be in the MRI room.
• Develop and implement training and videos that explicitly illustrate the hazards and highlight the potential catastrophic consequences of violating protocol.
• Place warning signs on doors.
• Place labels on acceptable equipment.
• Require special clothing for caregivers and patients (eg, pocketless pants and shirts for caregivers and gowns for patients).

In addition to these interventions, professional societies have developed lists and procedures that are well worth considering.(2,6,7) These include naming certain zones and areas, checking qualifications of operators, and restricting access.

A hospital might consider restricting purchases of items to enter the MRI suite to only MR-compatible ones. Implementing this commonsensical solution is surprisingly complex. Since hospital equipment tends to “migrate,” hospitals might need to expand this purchase pattern to include the entire radiology department or adjacent areas, which would multiply the expense of this solution. Even using metal detectors or magnetometers leaves subtle hazards unchecked. Because many metal items are not ferromagnetic, metal detectors may identify too many items (false positives), ultimately creating an environment in which providers ignore the warnings (“routine rule violations”). On the other hand, using magnetometers to screen for ferromagnetic material will miss many items because some heat-conducting materials are not ferromagnetic (false negatives).

MR hazards and possible countermeasures have been known for years by specialists within hospitals and safety personnel at regulatory organizations. Take a moment to speak to your institution’s radiology technicians or MR radiologists, and you will likely find that your own institution has had a close call in the past year. This case and discussion illustrate the importance of extreme caution before entering the MRI room with potentially hazardous items such as ferromagnetic medical equipment, coiled wires, and patient implants. Along with this concern, we must be prepared to support a broad set of measures to safeguard people who enter the MRI suite. Without everyone’s help, we will hear about cases like this one again.

John Gosbee, MD, MS

Laura Lin Gosbee, MASc
Human Factors Engineering and Health Care, LLP

1. FDA/CDRH resources page. U.S. Food and Drug Administration Web site. A primer on medical device interactions with magnetic resonance imaging systems.
[ go to related site ]

2. The safe use of equipment in the magnetic resonance environment. Health Devices. 2001;30:421-44.
[ go to pubmed ]

3. Chaljub G, Kramer LA, Johnson RF, et al. Projectile cylinder accidents resulting from the presence of ferromagnetic nitrous oxide or oxygen tanks in the MR suite. AJR Am J Roentgenol. 2001;177:27-30.
[ go to pubmed ]

4. Shellock FG. Magnetic resonance procedures: health effects and safety. Boca Raton, FL: CRC Press; 2001.

5. Kerr JD. MRI safety: everyone's job. Radiol Manage. 2001;23:36-9.
[ go to pubmed ]

6. Kanal E, Borgstede JP, Barkovich AJ, et al. American College of Radiology White Paper on MR Safety. AJR Am J Roentgenol. 2002;178:1335-47.
[ go to pubmed ]

7. American College of Radiology. MRI monograph: Safety and sedation. Reston, VA: American College of Radiology; 1996.