Description:
Who was responsible for MRI safety at Redwood City Hospital when an ICU nurse was nearly killed by an MRI scanner?

In this episode of The Invisible Force, we take a deep dive into the CMS investigation of the 2023 Redwood City Hospital MRI accident—and uncover a disturbing question about who was actually responsible for keeping the hospital’s MRI service safe.

The CMS report found problems ranging from the physical design of the MRI suite, to failures to follow established safety procedures, to serious questions about MRI safety training and emergency response. But as we follow the investigator’s questions, another problem begins to emerge: MRI safety governance.

Who was the hospital’s MRI Medical Director? The answer changed during the investigation. When investigators asked for documentation of the appointment, a department organizational chart appeared—with the MRMD and MRSO designations handwritten onto it. When they asked for documentation of MRI-specific training, the hospital could not provide it.

And then there’s the decision that became critical during the accident itself: who had the authority to order a quench of the MRI magnet?

Using the CMS Statement of Deficiencies and Plan of Correction, we examine what the investigation revealed about MRI safety policies, ferromagnetic detection, Level 2 MRI safety training, MRMD and MRSO responsibilities, emergency response, and the systems intended to protect patients and healthcare workers from MRI accidents.

The story raises a larger question: Was this simply a series of individual mistakes—or evidence of a much deeper failure of MRI safety systems?

And importantly, the CMS report also provides something hopeful: a roadmap of changes that could make MRI accidents like this preventable.

Next: the OSHA investigation reveals an even deeper documentary record of what Redwood City Hospital taught its employees—and what they actually knew.

Show Notes:
CMS-2567 “Statement of Deficiencies | Plan of Correction” Report for Redwood City Hospital (8.4 MB)

Transcript:
“So let’s switch now to the question of MRI safety governance. Kaiser Policy states, quote, the MRI safety committee will meet quarterly. End quote. who’s the radiologist on that committee? Who’s the site’s MRI medical director?”

“Our MRI Medical Director role is filled by our chief of radiology.”

“And what documentation do you have for the chief’s training to serve in the MRMD role?”

Hello and welcome back to the third series of the Invisible Force Podcast. Invisible Force is an investigative documentary podcast about MRI accidents, patient safety failures, healthcare mysteries, and the hidden risks of modern medicine. Before we pick up our story from the prior episode, this is a great time to reintroduce you to the co-hosts for our podcast.

John Posh, an MRI technologist, an MRI educator, and longtime advocate for MRI safety practices. I’m Chief Academic Officer and MRI program director at one university and adjunct MRI faculty at a second university.

And I’m Toby Gilk, an MRI facility architect and planner, safety consultant, and co author of the new MRI safety textbook, The Technologist’s MRI Safety Handbook.

In our previous episodes, we recreated the accident for you, described how we inadvertently turned out to be the people who first reported it to the California Department of Public Health, and how the fact that it was an injured employee as opposed to a patient that triggered a separate independent investigation by OSHA. When we last left you, we had the CMS report in hand and we were just getting into it while we were waiting for the OSHA report to be finished and released.

In this episode, we’re doing a deeper dive into what the CMS report found as deficiencies in the Redwood City Hospital’s MRI safety practices and what the hospital agreed to in order to correct or satisfy CMS. At its most basic, hospitals have to meet CMS’s conditions of participation for radiology services, including MRI, hospitals “must meet professionally approved standards for safety and personnel qualifications,” and they “must be free from hazards for patients and personnel.” CMS’s underlying finding for this was that the hospital failed to do this.

For accidents like this, there’s often a tendency to point at the last act in the chain of events and pin responsibility for the whole event on that one thing. That’s direct. We have information that supports that as a conclusion, and it quickly satisfies the ‘torches-and-pitchfork” crowd that clamors for someone to blame.

Those early conclusions, they tend to be based on only those details which bubbled up to the surface first. The easy answer that’s tempting to jump straight towards also makes it more difficult to identify faulty systems or the other contributing factors. When done well, these CMS reports strive to get deeper than who is the first person we can blame this on.

In a nutshell, the core finding from the CMS report was “the many safety failures by the MRI and non-MRI personnel created a culture of unsafe practices leading to the severe injury of one hospital personnel and the potential for injury to one patient.” Now we think that that summary finding actually reverses the cause and effect, but we’ll get to that later. There were a bunch of different types of deficiencies that we’ve grouped together into four different buckets the suite, the processes for training, the emergency response, and the governance structures. Pretty much everything.

Let’s start with one of my favorite subjects, the bricks and mortar of the MRI suite and how it contributed to the accident. If you remember our last series on the Colombini accident, we said that the MRI addition to the hospital in that case hadn’t followed the building codes, and that that played a key role. Here, as near as we can tell, there were no failures of the design to meet the applicable codes. Rather, here it’s a failure of the codes and standards to protect MRI patients and healthcare workers.

For nearly 20 years, most MRI suites have been designed and built following a model called the 4-Zone model. In that model, the key part is that you’re supposed to pass through two intermediate spaces between the public realm, zone 1, and the MRI scan room, zone 4. One space is for screening and clearing people, zone two, and the next space, zone three, is essentially an airlock entrance. It usually serves other functions like being the control room where the technologists work, etc. Redwood City had zero separate zone two for screening and preparation area. Both zones two and three were in that ante room.

That same industry standard model, as well as Kaiser Permanente’s own policies, called for the ability of the MRI Tech to directly supervise anyone in that zone three area. As we learned in our earlier recreation of this accident, that was functionally impossible because the control room was walled off from that airlock anteroom. As soon as the technologist dipped into the control room. She was functionally incapable of directly supervising the people who were in that airlock or zone three space.

The size and arrangement of that tiny Zone 3 anteroom airlock space made safe patient handling into a type of tile puzzle exercise, which also played a key role in this incident.

You might be thinking that all of these problems are an indicator that the architects or the builder or the code inspectors that they messed up and planned and built an MRI suite that violated all sorts of standards. If so, you’d be wrong. while the Redwood City Hospital MRI suite design fell far short of that four zone model, it still complied with the California Building Code for MRI facilities. And it met their hospital accreditation organization, the Joint Commission. It met their requirements for zone three, zone four layouts. Joint commission standards don’t actually describe or require a zone two type of space in their minimum standards. But these are really subjects that we’re going to get into in a later episode.

Then there’s the gap between what was documented policy and what happened in real life. While there’s no indication that the ICU patient or nurse or nurse’s aide were injured because they had a pacemaker or a piece of metal shrapnel got pulled by the scanner, apparently none of these three were screened before they walked into the MRI room. It could have been that they all would have been properly screened if the nurse hadn’t directly proceeded into the MRI scanner, but in theory at least, both the nurse and the nurse’s aide were MRI safety trained and were supposed to know that the screening needed to be completed before anyone went into that MRI scanner room.

The hospital did have a special type of metal detector called a ferromagnetic detector, or FMD. They had that mounted on the doorway of the MRI scanner room. Now, these products are different than the ones you pass through in the TSA line at the airport. These specifically alarm these specifically alarm only on the presence of magnetic or magnetizable materials… things that can go flying at the MRI scanner. The mystery here? None of the staff present for the accident reported having heard the FMD alarm go off. So was the FMD broken on the day of the accident? The suggestion that the FMD malfunctioned and failed to alarm is particularly odd because in the CMS report materials, it describes the radiology director as having tested it on the same day after the accident and found it to be working. If it was turned on and the volume turned up, it would have squawked at the ICU bed passing through the opening.

Both Toby and I have worked with FMD companies before, and most of these products are pretty bulletproof. Now, unfortunately, it’s also often the case that FMD scanners sometimes get put into MRI suites, and the hospital or imaging center doesn’t change anything else about how they operate. You’d be surprised by just how much steel there is in our clothing or shoes. Anesthesiologists got some bad news about those Dansco clogs. Metal is everywhere in the hospital equipment that wasn’t specifically purchased to be non magnetic and used in MRI.

When an FMD just gets dropped into an MRI department without any integration work, it can sometimes lead to alarm fatigue to the point where the MRI staff don’t even hear the alarm while it’s happening. Or they’ll just turn it off when the boss isn’t around. No one there that day reported hearing the FMD go off, but they also insist that it wasn’t turned off or the volume turned down either.

The hospital had a policy that the MRI’s scanner room door was not only to be kept closed, but kept locked shut. The MRI tech blocking the door open with the MRI table when she set the suite up for the pit stop style rapid transfer was a direct violation of the site’s policy, requiring the door to be kept closed.

Then there was something that when I heard it, I think I scared my dog when I screamed out in reaction. So in the MRI safety literature, there are two formally recognized levels of MRI safety training, level one and level two. Level one, which is often for tech aides and transporters, and sometimes respiratory or anesthesia staff, every once in a while, housekeeping staff who service the MRI, and every once in a while, nursing staff from patient floors will also wind up getting this lower level, this level one training. This hospital, however, had a system that a newly hired ICU nurse, after going through a PowerPoint instruction module, had equal MRI safety training, level two MRI safety training, as an MRI technologist.

Now, level two designation, according to the ACR who invented the designations, carries with it some pretty powerful responsibilities. If you’re level two, you can be given the keys to the kingdom. You have no you have independent access to the MRI suite. You can come and go as needed. As a level two tech, as level two, you’re also supposed to know about the required practices like screening people and equipment before going into the scanner room and following them religiously. You’re supposed to have enough knowledge and competency in MRI safety to not only look after yourself, but to know enough about MRI safety practices that you can be responsible for the safety of others.

The facts of what happened on that day all suggest that the nurse didn’t demonstrate the knowledge or competencies that are all supposed to be essential parts of that level two designation. So if that’s true, how was the level two designation earned? How does an ICU nurse who worked in the hospital for less than two months and may have never been to the MRI suite before? How does that person get the same MRI safety designation as an MRI tech?

So if the MRI’s suite design was problematic, and if the safety practices were written down but not followed, and if the MRI’s safety training credentials were sometimes handed out without assessing someone’s competence, who was actually responsible for safety at Redwood City Hospital?

The sponsor of Invisible Force Podcast is CAIREreporting.org. That’s CAIRE spelled C A I R E. CAIREreporting.org is a confidential MRI adverse incident reporting system available to us the public as well as offering an enterprise level solution to hospitals and imaging centers for secure and confidential reporting of MRI accidents, incidents, and near misses. If you have direct knowledge of an MRI accident that may have been swept under the rug, or if you want information about how CAIRE could be set up as a private internal error reporting tool for your hospital or imaging center, CAIREreporting.org can help you with both of those. CAIRE has an assembled panel of international experts in MRI safety and accidents. Reports submitted, either through the public website or through an enterprise system, get reviewed by their experts, who then deliver insights into the contributing causes of how accidents or near miss events happened and, more importantly, steps that could prevent similar incidents in the future. For more information, please visit CAIREreporting.org. That’s care spelled C-A-I-R-E.

In the first half of this episode, we detailed a number of things that went wrong, from the initial build out of the MRI suite to existing policies that weren’t followed, to some pretty grave failures of MRI safety training. But each of these conditions didn’t just fall from the sky, they all came to be because of decisions and actions of the people who were responsible for the safety of Redwood City Hospital’s MRI service. Who were they?

MRI exams, in the USA at least, require a prescription. Your regular doctor might refer you for an MRI exam, but it’s almost always the radiologist who legally prescribes the exam for you. This means that there is at least one radiologist with responsibility for the safety of MRI exams. Who was the MRI services supervising physician for Redwood City Hospital? In the opening radio play, you heard us dramatize a piece of the CMS investigation that explores that.

In those CMS investigation interviews, the investigator apparently didn’t have any documentation as to who the main radiologist for MRI safety was. But in the interview, that investigator asked the radiology director for the hospital initially identified the chief of radiology, the head radiologist, as the MRI medical director or MRMD responsible for the clinical aspects of MRI safety.

But when the CMS investigator asked the follow up question what specific training the chief of radiology had to make them qualified to serve in the role of MRMD, the only thing the radiology director could offer was that the chief of radiology had completed their radiology residency.

If, as the radiology director argued, a radiology residency was enough, then that would mean inherently every radiologist could be an MRMD, because every radiologist completes a radiology residency. There’s a serious problem with that assertion, however. You see, the ACGME, the group that establishes the minimum residency training requirements for all the medical specialties including radiology, in the current year, in 2026, the ACGME does not require any MRI safety training for radiology residents. None.

Now there are some residency programs that do teach MRI safety. We know this because both Toby and I have done some resident training before. But the idea that completion of a radiology residency in and of itself is proof that a radiologist is trained in MRI safety is embarrassingly off the mark. We’d love it if radiology residency was some sort of guarantee of sufficient MRI safety knowledge, but today that’s just not the case. The same is true for radiologist board certifications.

The CMS report itself doesn’t list out the duties of a person who’s supposed to serve as the MRI medical director, but the investigator makes repeated references to the MRI safety standard document, the ACR Manual on MR Safety. Incorporated by reference, the ACR manual describes the important role of the MRMD in concert with the MRI safety officer or MRSO.

After the investigator pressed for the chief of radiology’s MRI safety or MRMD role specific training, the hospital made what, to us at least, was a very odd pivot. Essentially, later, the radiology director said she misspoke and that the chief of radiology wasn’t the MRMD as they previously asserted. Instead, there were two radiologists, subdivision chiefs, one for neuroradiology and one for musculoskeletal radiology, and these two shared the role for MRMD. In addition to sharing the role of MRMD, these two radiologists also shared the role of the site’s MRI safety officer or MRSO.

You might think that this would help clear up the org chart of responsibility, but in our minds it only added suspicion. When the CMS investigator asked the same follow up question as before, asking to see confirmation of the appointments of the sub chiefs to these combined MRMD / MRSO roles, the radiology director produced a printed out org chart that showed the relationships among the radiologists. On this org chart, someone had handwritten the MRMD and MRSO designations under the neuro and MSK sub chiefs positions.

To their credit, the CMS investigator didn’t give up on their prior line of questioning about the specific training for the MRMD roles and now the MRSO roles too, for the sub chief radiologists pressing to learn more about what MR safety specific training these two individual MRMD slash MRSO radiologists had received. And again, the CMS investigator asked for substantiation that they had received appropriate MRI safety specific training. The CMS report indicates that the hospital never provided any information about the chief or the sub-chief’s MRI-specific safety training before the accident. The shifting story about responsibilities, the handwritten roles in the org chart, the failure to provide MRI specific, the failure to provide any MRI-specific safety training documentation, failure to provide any MOI safety specific documentation, none of this is inspiring a lot of confidence.

We also found it very odd that the responsibilities for the MRMD and MRSO roles were both assigned to these two radiologists. There are at least a couple of the key elements in the original definitions of the roles which in a typical arrangement would make it difficult, if not outright impossible for a radiologist to act as MRSO. The standard definition describes the person in the MRSO role as someone who is “readily accessible and available at all times that the MRI facility is accessible,” or “ensures that proper policies and procedures for day-to-day MRI safety are enforced.” In many facilities, the radiologists perform much of their service remotely, if not in a distant part of the hospital, from somewhere that’s not in the hospital at all.

It’s not physically impossible for a radiologist to do the tasks of an MRSO, but the most frequent work patterns and workload for most radiologists would make it difficult for a radiologist to fulfill those defined roles of MRSO. Working away from the MRI suite at all times makes meeting obligations of the MRSO role essentially impossible. These are just a couple of reasons that we almost never see radiologists serving in MRSO roles.

Since we’re into the realm of responsibilities, what about the MRI tech’s decision to not push the quench button? We got a copy of the Kaiser MRI safety policies that were in effect at the time. And while it does have a policy section about quenches, it actually didn’t explicitly say that the MRI technologist is empowered to make the call to push the button.

There may have been dueling perceptions of what the documented policy said, with the director and MRMD or MRMDs thinking it gave green light to the tech, but the tech thinking there was still some additional permissions needed. From what we’ve learned from both official and unofficial sources, it seems that the tech understood her authority to be the person who physically pushed the quench button, but not the person who makes the call that it’s the correct course of action. We will learn later that, immediately after the accident, before pulling the nurse out, the tech did reach out for permission, but didn’t actually connect with anyone who gave it to her, at least not immediately when she actually needed it.

This decision to not quench the MRI scanner is inexorably a part of the decision to tear the nurse out from between the bed and the MRI scanner. Specific to that, from the CMS report, they said the nurse was “forcibly removed from the magnet while still pinned by the ICU bed. Review of the nurse’s inpatient record, the treating physician indicated that this could have caused more damage to the areas where she sustained lacerations.” The choice to tear the nurse out and all of the consequences of that decision are really rooted in the question of the Quench Authority, an authority that was ambiguous on the day of the incident.

If you take all of these together, what seems to be emerging isn’t just a collection of oopsie moments that compound into a tragedy, but instead it suggests a pattern. In big accidents like this, a typical first instinct may be to find a single person or single action to blame, but the CMS investigation uncovered weakness after weakness, each of which was the product of decisions or lack of decisions throughout the enterprise. This wasn’t appearing to simply be a smorgasbord of several failures of policy, practice, or training, although all those are included, this was looking like systemic collapse.

The silver lining to this story is that the CMS report prompted a plan to fix MRI safety elements that were identified as having been weak. Now, either John or I have had a chance to go to Redwood City Hospital and see how much improvement has actually come from the hospital’s implementation of their plan of correction response. But the documents suggest that the plan was fairly comprehensive. From revised MRI safety policies, policies that were significantly expanded, to the improved MRI safety training for radiologists and MRI department staff, including new MRMD training programs, a replacement of that ferromagnetic detector, and retraining of all radiology staff on the updated site policies and more.

The good news is that each of the discrete problems that the CMS report identified are fixable. There are solutions or tools to strengthen knowledge, policies, and practices. What the CMS report is less clear about is whether Kaiser or Redwood City Hospitals saw the laundry list of problems as only a checklist, or if they saw it as a grave warning of what had appeared to be a major systemic blind spot.

These sorts of MRI accidents aren’t inevitable. They aren’t a baked in or inescapable risk or cost associated with MRI services. In fact, it’s almost exactly the opposite. MRI accidents are almost one hundred percent anticipatable and preventable. While this one took a CMS investigation shining a light on the weaknesses of the MRI safety practices of Redwood City Hospital, there’s nothing stopping existing facilities from doing the same prospectively to prevent their own accidents.

We began this episode reading you the core finding of the CMS report. “The many safety failures by the MRI and non-MRI personnel created a culture of unsafe practices leading to the severe injury of one hospital personnel and the potential for injury to one patient.” After a deep review of the report, we see it differently. The pre-existing weak systems of prevention, the unclear structures of safety responsibilities, and the weaknesses in MRI safety training were the products of a culture that didn’t take MRI safety as seriously as it should have. The plan of corrections, however, was something of a roadmap to fix the discrete problems and hopefully rehabilitate the culture of safety at this facility. Three years on, we don’t have any measurements of success, so we can only hope that there have been meaningful shifts in the safety culture at Redwood City Hospital.

Right about the time that we had finished digesting this CMS report, there was a ping in our inbox. Turned out it was the long-awaited OSHA report, and it came with something that CMS hadn’t given us much deeper documentary record of the training and the internal perceptions of Redwood City Hospital’s MRI safety protections.

“So let’s switch now to the question of MRI safety governance. Kaiser Policy states, quote, the MRI safety committee will meet quarterly. End quote. who’s the radiologist on that committee? Who’s the site’s MRI medical director?”

“Our MRI Medical Director role is filled by our chief of radiology.”

“And what documentation do you have for the chief’s training to serve in the MRMD role?”

“I’d like to amend our prior statement. We actually have two radiologists, the subchief for neuro and the subchief for musculoskeletal, who serve as co-MRMDs.”

“So not the chief, as previously stated. And okay, the MRMD specific training for the subchiefs.

“Well, they’re radiologists. They completed a radiology residency.”

“But what about MRI specific training?”

“Okay, we’ll circle back to that one later. And the documentation of these appointments as co MRMDs?”

“Here’s a department org chart.”

“Okay, I see the subchiefs on the org chart, but it doesn’t indicate that they’re the facility’s MRMDs.”

“I can fix that. [writing noises] There, now it does.”

“You… you literally just hand wrote that in.”

In our next episode, we’ll look specifically into what we learned from the OSHA investigative report, which turned out in some ways to be much more detailed than the CMS report. For this week’s show, our sources were the CMS Statement of Deficiencies Report, as well as a whole bunch of confidential sources, both from within Kaiser and from outside.

Our opening radio play dramatizing the investigator’s uncovering of a few concerns about the hospital’s MRI safety governance, well the voice actors for that dramatic reenactment were Brandon Grugle and Julia Shafini.

If you have any information about this incident or any other MRI accidents, please reach out to us through our website, InvisibleForcePodcast.com. There you’ll find episodes, show notes, a tip line. You can also leave us a voicemail with information about this or any MRI incident. Just call area code 631 MRI-tips. The number again is 631 MRI-tips or 631-674-8477.

If you work in hospital safety, healthcare regulation, CMS, healthcare policy, or MRI, and something in this story just doesn’t make sense or sit right with you, we want to hear from you. Please reach out to us.

Lastly, we ask you to like and share our podcast with your friends and colleagues. With your help, we will unravel the mystery of what happened in this incident and with a little bit of luck will help make sure that accidents like this don’t ever occur again.