MRI is not the safe modality. (There… it needed saying and I said it.)

A friend of mine recently posted something on LinkedIn with the opening line that MRI is not safe. Boy did that stir up some discussion about MRI safety and risk!

I offer that preamble for two reasons… First, I shouldn’t be recognized as the first person who has said this. Second, conversations prompted by my friend’s posts have really been rattling around in my head since then. I’ve been thinking a lot about the nature of risk in radiology, and why MRI is so qualitatively different.

Let’s look at CT as a point of comparison…

• There’s no risk to people in the environment from CT if it’s not actively scanning a patient.
• The risks from CT, even from a reportable-level error in ionizing radiation exposure, are almost entirely stochastic. [Meaning -short of catastrophic levels of exposure- we have no idea beyond the tiny probability whether it will produce any harm to the person.]
• *If* radiation exposure from a CT exam did wind up producing a cancer, it would likely take decades to appear. Even then it may be eminently treatable (or it may be fatal… that, too, is a roll of the dice).

Now let’s look at MRI…

• Magnetism from clinical MRIs doesn’t (as far as we know, and we’ve done over a billion MRI studies by now) produce any physical harm to people. But it can (and has) interfered with medical devices (e.g., pacemakers, insulin pumps, etc…), the malfunctions of which have killed scores of people. This risk exists 24 / 7 / 365 in the room that houses the MRI, and sometimes even in adjacent spaces.
• Where the risks from CT are almost immeasurably-small likelihood of cancer decades later, in MRI it’s instantaneous. We crush people with flying magnetic objects with frightening regularity.
• In addition to the most ‘photogenic’ MRI accidents, projectiles, we also deafen people with extremely loud sounds. We burn patients, and sometimes trick patients’ active sensing devices – like pacemakers or neurostimulators – into delivering inappropriate therapy, which can sometimes be harmful. These other injury types, while limited to when the MRI exam is happening, are also well known risks.

To be clear, risks to any individual patient from any single exam is extraordinarily low… for both CT and MRI.

So how do we protect people from the risks of harm for each of these two modalities?

For CT (for nearly all ionizing radiation modalities), providers are typically required, by state law, to have radiation safety committees (with participation requirements, minimum meeting requirements, QA/QC requirements), named radiation safety officers with explicit statutory obligations, licensure requirements for who is permitted to administer ionizing radiation. If you including nuclear medicine / molecular medicine and their use of radioisotopes, add in federal-level legal requirements and oversight for a radioactive materials (RAM) license.

For MRI in the United States, if you don’t live in one of about a half-dozen of the US states that actually have minimum requirements for MRI operators (some of which – and I’m looking at you, Massachusetts – only require training in ionizing radiation modalities to operate an MRI), there are no state laws or licensure requirements for who operates MRIs. There are no safety committee requirements. There are no safety officer requirements. And there are no federal oversight requirements. Two states that I’ve recently been looking into their state laws governing licensure (California and Ohio) don’t even acknowledge that MRI exists as a medical imaging modality to be regulated by their state’s Department of Health!!

And if you have internalized the marketing messages of the accreditation organizations… that signing on with them, in the words of one of them, ‘assures the highest levels of quality and safety’ for your MRI service, I hate to be the bearer of bad news but this is unequivocally untrue. The accreditation organizations, in order to be meet the minimums to be ‘deemed’ by Center for Medicare / Medicaid Services (CMS), only need to parrot the incredibly vague criteria in the CMS conditions of participation and interpretive guidelines. They don’t even routinely QC their own MRI safety standards by looking at how MRI accidents happen, how patients or staff are harmed, and how their standards might be made more effective at actually providing meaningful improvements in safety.

So if regulation, licensure, state law, federal oversight, and accreditation standards don’t provide meaningful support for MRI safety practices, who does?

If the individual institution (hospital, imaging center, research institution) isn’t taking it upon themselves to identify best practices, codify them in written policy, provide training to all appropriate stakeholders, and internally verify compliance with their policies, nobody else is going to do it for you! This isn’t CT where meeting state licensure requirements is actually a meaningful demonstration of complying with safety best practices. For MRI you’re on your own!

The simultaneously great and exasperating news is that virtually 100% of MRI injury accidents are preventable through existing best practices! This is great because MRI can live up to the promise of being the safe modality with the existing tools and practices. This is exasperating because the structures that are supposed to assure safety (licensure, regulation, and accreditation) have almost completely ignored those tools and practices, leaving providers to ‘decide for themselves’ what they want to do (or not do) for MRI safety.

So, yes, MRI is dangerous… but it doesn’t have to be that way. It only is because the safety structures we have in place for the rest of radiology ignore (and have for ~40 years) the unique risks in MRI, and the safety standard practices that have been published for more than 20 years.

# # #

If you want to get an objective look at your site’s MRI safety, or get assistance strengthening it, please contact Gilk Radiology Consultants.