“Why — now — ten years later would you post these documents?”

Excellent question! Here’s why I didn’t publish these long ago…

I didn’t have them.

Yes, the civil suit had been underway for years. Yes, individual documents had been filed and made public during the course of the civil litigation, but the civil suit was only resolved a year ago and it took several months for the last of the documents to be made public through the Westchester County Clerk’s Office (who, by the way, were profoundly helpful in accessing these public records).

Here’s why I am publishing them now… Despite the fact that this is the watershed event in MR safety, the degree to which the industry has really dissected this event and identified the causative factors has been wanting. Desperately wanting.

I’m currently working with a colleague on a root-cause-analysis of this event, drilling down through the simple (don’t have ferrous oxygen tanks in the MR suite) to get at more meaningful elements of this accident that we can work to prevent similar accidents. It promises to be unlike anything you’ve learned about why this accident happened.

Given the trajectory of MR accidents and adverse events, this sort of analysis appears to be desperately needed.

So, what are the documents? They are transcripts of the depositions of many of the key people involved in the accident and couple of ‘official’ reviews. These are the source materials. The news accounts you’ve previously read are all synthesized from these (or from others’  interpretations of these). If you’re so inclined, you can download and read these for yourself.

The essential elements of the sequence of events for the accident are these:

• Michael Colombini, a young boy, was injured from a playground accident
• The ER had a head CT run, which revealed an unknown / asymptomatic brain tumor
• The boy had surgery very shortly thereafter to remove the tumor
• Prior to discharge, the boy was sent for a baseline MRI as a reference for future monitoring
• The boy was sedated prior to the exam and placed in the MR with a cannula to deliver oxygen
• Before the exam began, the anesthesiologist observed a decline in O2 saturation, and realized that the oxygen from the wall outlet was not flowing, despite his attempts to turn it up
• The anesthesiologist called the technologist who was to administer the exam to the door of the MR room, instructing her to find and fix the source of the problem with the oxygen flow
• This technologist was not familiar with the oxygen supply system, which — in apparent violation of codes — was fed to only the MR exam room from a bulk cylinder without any pressure or flow alarms
• The technologist sought her colleague who she believed knew the oxygen system and together they entered the MR equipment room to try and fix the supply problem
• The anesthesiologist cried out for help, though the technologists in the MR equipment room could not hear this
• A nurse (who had accompanied an earlier patient to the MR suite was returning to retrieve an item she had left) heard the anesthesiologist’s cries for help and handed him a portable cylinder near the door to the MR exam room
• The anesthesiologist turned to approach the boy with the oxygen tank when the magnetic attractive force of the MRI pulled the cylinder from the doctor’s grasp
• The tank flew into the MRI where it struck the boy in the face and head, inflicting fatal wounds

The following PDF documents vary in size from 1 MB to 25 MB, and will take a few minutes to download, depending on your connection speed.

Deposition of Patricia Lauria, technologist who was to have administered the Colombini scan
Deposition of Paul Daniels, other technologist on duty who assisted in the repair of the oxygen supply
Deposition of Jian Hou, MD, anesthesiologist who sedated / monitored Colombini for the MR exam
Deposition of Terrence Matalon, MD, Radiologist who was simultaneously the hospital’s Director of Radiology and president of the private company subcontracted by the hospital to provide operations for the MRI service
New York State Department of Health incident report
Westchester Medical Center incident review

As you might suspect, these documents are but the tip of the iceberg of the body of the court filings in this civil suit. However, for those interested in what happened and why (as opposed to the legal maneuvering), these documents are the most illuminating.

In the months ahead, the 10th anniversary of the 2001 Colombini fatality will include a deeper look into this accident and the changes that have taken place (and those that are still needed if we wish to avoid repeating this accident). This has begun, slowly, with the new building code requirements that are being adopted by various US states and the Joint Commission, but may pick up steam with federal government intervention.

Please check back periodically for the latest information on MRI safety… both as it relates to specific preventions, such as ferromagnetic detection systems, and broader awareness such as knowledge of the factors in the Colombini fatality.

Click here for Tobias’ Twitter Profile

That’s right, the FDA has updated it’s MRI accident figures available online through the MAUDE database. We were alarmed and astonished when we thought that the rate of increases in MRI accidents was only 270% (from 2004 to 2008). Turns out that the FDA must have found additional accident reports that were in a stack of junk-mail, or got lost between the sofa cushions, which means that the rate if adverse events went up, significantly, in 2008 from the prior calculation.

Somehow, when I did the analysis last year (in 2009) of the 2008 numbers, it was apparently 11% shy of the final total. When we add the (previously uncounted) adverse events, the actual rate of accident growth is 310%!!!

Between 2004 and 2008, MRI Accident Rates Increased 310%

That’s right, in 2008 we were more than 4 times as likely to injure someone during an MR exam than we were just four years earlier!

What would happen in your town if:

• Traffic accidents quadrupled in 4 years?
• Rates of violence in schools quadrupled?
• Divorce rates increased 4x in 4 years?

Alarm bells, that’s what! People for certain would not be complacent.

There would be efforts to figure out why, and fix whatever was going wrong. Reduced speed limits or more traffic enforcement? You bet! Counselors in the schools and demands for greater teacher and parent involvement? Darn right! Lay and religious leaders reassessing the very nature of the marital institution in our society? Abso-friggin-lutely!

So, with an exploding rate of MRI injuries and adverse events, what is being done to identify and curb the source of these incidents? [cue cricket sounds]

NOTHING!

Apart from the continuous efforts of a small cadre of MR safety advocates, whose cries have (apparently) fallen on deaf ears, there are no substantive accreditation, licensure, or regulatory actions that have reversed the trend of the last several years.

The silver-lining may be that the increase from 2008 to 2009 was very modest. Perhaps we’re leveling-off, or perhaps, like 2004, this is just a momentary pause before we skyrocket upwards again. And given the FDA’s marked upward adjustment of the 2008 numbers, it may wind up being another year before we can feel confident about the 2009 accident report numbers.

Click here for Tobias’ Twitter Profile

Just a few months ago a service engineer was replacing a fan-blower assembly in an MRI unit (a part that is notoriously ferromagnetic). Working alone in the suite in the evening, after the regular staff had left, the engineer had finished early… or that’s what the security guard thought when he called to her and got no reply.

Turns out that she had been struck and pinned to the MR scanner by the blower assembly, and was unconscious, if not already dead, when the guard checked to see if she was still there.

This tragic story is something of a departure from my typical mantra of patient and staff safety. Yes, this was a trained individual who knew about the risks of the MR environment and materials she was working with. And yes, this was a vendor, and not a hospital worker or patient. But this is a repeatable condition, and an accident which, because there have been MRI accidents involving such a tremendous variety of ferromagnetic materials, deserves a little analysis for a ‘lessons-learned’ output.

One of the (theorized) main contributing factors to this accident is the design of the magnet room. Since the advent of active shielding, we’ve seen MRI rooms go from the size of racquetball courts to office cubicles. In this case, the clearances around the magnet were uncomfortably tight, and what space there was between the magnet and the walls of the suite was purportedly infringed by shelves, storage and clutter.

By failing to provide an appropriately-sized room to accommodate not only the MRI unit itself, but also the service and storage needs, the layout may have substantially increased the likelihood of an accident.

And while conventional screening methodologies wouldn’t have helped in this particular scenario (the object already in the MRI room), it’s not like this is the only strange thing that has been brought into a MRI room to be ‘sucked’ into the scanner. Yes, we all know about oxygen tanks (well, apparently we don’t, as there was another one reported recently, here), but they aren’t all!

Personal computers, iPods, filing cabinets, desk chairs, anesthesia machines, cribs, gurneys, wheelchairs, dollies, staplers, power tools, axes, roller skates, ‘sand’ bags, hampers, mop-buckets, and the list goes on, and on, and on… All of these, and many, many more objects have found their way into MRI scanner rooms. Sometimes the people involved, like in the circumstances surrounding the recent fatality, know that they’re taking a risk. But at least as often the accident occurs because the person is unaware of what they’re doing.

The magnets don’t take vacations. They’re not on just when ‘taking the picture’. They’re not turned off for the night when the last patient is done for the day. The risks are omnipresent, which demands that we are equally vigilant about providing the appropriate protections for everyone and everything that approaches the MRI room.

In the weeks ahead, I hope to have information for you about some of the efforts in the works that may help codify some of these expectations at the point of care. Suffice it to say that right now, for the first time in the U.S., substantive consideration is being given to explicit MRI safety requirements at the point of care. This is still all in the formative stages, and lots of work remains to be done. But perhaps when it is, there’s an umbrella drink and sandy beach with my name on them.

Click here for Tobias’ Twitter Profile

Every so often there is an MRI accident description that is so stunning that it sends a jolt through me, reminding me why I do what I do. This is the entry that I came across just two weeks ago…

“Event Type Injury   Patient Outcome Life Threatening; Hospitalization Required Intervention”

‘Wait a minute… that may border on interesting (and I’m being very generous there), but that’s hardly a ‘jolt.’ What gives?’

Well, two weeks ago that exact same incident report described an accident in which an MR technologist had to be rushed to surgery to extract the pair of scissors from their forehead! Turns out that the tech was between the magnet and another staff person who had brought a pair of ferromagnetic scissors into the magnet room. The magnet ‘grabbed’ the scissors and started drawing them into the bore, but for the fact that the technologist’s head was in the way!

My longstanding complaint with the FDA and their MAUDE data has been that they seem completely disinterested in actually collecting it with regard to MRI accidents. The overwhelming majority of MRI accidents are a product of the inherent risks, and not a result of equipment malfunction and it seems to me as if the very existence of the MAUDE database is NOT as a patient safety tool, but as a CYA tool to validate that the approval of a medical device was warranted.

Now the hobbled MAUDE data (built on the FDA’s Medwatch program) has gone mute and the actual substance of an accident report is no longer there. Today there’s more information about the postal address of the company that manufactured the MRI than there is about the accident that is being reported, at least that’s the contemporary reality via the online database.

So I implore the FDA to please restore the narratives associated with MRI accidents. Removing this information does a grave disservice to patients and those who advocate for MRI patient safety.

It took a while, but slowly I realized that tellin‘ and getting in trouble were two, very different things. This is a lesson that we in the MRI community would do well to learn regarding accidents.

MRI accidents – in which materials, devices and implants which have no business in the MRI scanner room – occur all the time. Mostly, these are caught in advance, before anyone is injured or anything is damaged. Fairly frequently though, the MRI’s magnet ‘finds’ cell phones, money clips, pocket-knives, unsafe wheelchairs, etc…. And fairly frequently large ‘black hole’ artifacts on initial scans expose ferromagnetic devices or implants that the patient disavowed in their clinical pre-screen. (A veritable photo album of projectile accidents is available in one of my recent posts, which you can access here.)

Because these sorts of events are often rationalized to be statistical inevitabilities, brushed-off as events that occur at all MR facilities (and, therefore, are certainly not news to anyone), little – if anything – changes as a result.

Now, because these accidents and ‘near-events’ are typically accepted as unavoidable, they are very, very rarely reported outside of mandatory reporting structures, such as the State of Pennsylvania’s Patient Safety Authority. And even within mandatory structures such as Pennsylvania’s, many projectile events never get a 2nd thought, much less an event report. The upshot is that others, who might have learned from these surrogate mistakes, are more likely to have the same mishap because neither the nature or the frequency are shared with the community at large.

‘But wait, Mr. MRI-safety man. There is a national MRI accident database kept by the FDA and reporting to the FDA’s database is mandatory.’

Yes, there is a FDA accident database. And yes, technically, reporting is mandatory. The accident reporting system is called MDR (Medical Device Reporting) and the data is published in the FDA’s MAUDE database (which anyone with an internet connection can access and search). The three criteria for mandatory accident reporting are:

• Death (pretty unequivocal)
• Serious Injury
• Unreasonable Risk of Death or Serious Injury

What constitutes serious injury? Well, if you’re a healthcare provider, you get to make that call as there is no standard that the FDA uses. Want to call a traumatic amputation a ‘slightly-less-than-serious’ injury? OK. Want to not report it because you’ve determined that it isn’t a ‘serious’ injury? The FDA says that’s OK, too.

We don’t even need to touch on unreasonable risk unless, of course, we’re talking about the risk of death, but then we open the question of what makes something ‘unreasonable.’ I suspect that the FDA leaves that determination up to the individual as well.

So given the two indisputable truths above (1) MRI accidents are reported only a mere fraction of the time and (2) even the reporting requirements have loopholes that you could drive a truck through, what is the point of this article?

Ignore precedent. Ignore lawyerly readings of the reporting language. Report every MRI accident and near-event that occurs.

‘What? OK, Mr. MRI-safety man, you’ve really gone off the deep end on this one!’

Hear me out. Remember my intro to this article… the difference between being ‘told on’ and getting in trouble? That wasn’t just a cute story to draw you into an unrelated article, it’s an analogy for what happens when MRI providers report accidents and near misses to the FDA.

First, accident reports go through your MRI manufacturer and do not need to go directly from you to the FDA (except in cases of death). Second, accident data is anonymized. There is nothing that needs to identify you, your facility, your location, anything other than the scanner involved in the accident.

But the most important reason to have your MRI manufacturer report these events is that, by sharing the information about the types and frequencies of accidents, our entire industry will be able to better respond, as a whole, to the [wink, wink] events that are too often taken for granted.

With a more accurate picture of what goes wrong and, even more importantly, what steps we can all take to make things right, we can each do our small part in helping to improve the safety and effectiveness of the entire MRI industry.

There is no 7-foot tall troll waiting for you at the other end of your accident report to the FDA. There is no enhanced regulatory scrutiny that follows. Heck, nobody external to your organization will even ask you if you’ve done anything to increase your chances of catching that same mistake again. In short, there is no real downside, though the information you provide may ultimately help another MR provider prevent a costly or dangerous accident.

No, I’m not talking about the year of record sales of ferromagnetic detection systems or the publication of no less than three MRI safety best practice guidance papers… What I’m talking about is the numbers of MRI accident reports to the FDA.

Yes, we’re on track to see roughly a 20% jump in reported MRI accidents from 2007 to 2008… and 2007 was already a record year!

‘But, hey, MRI utilization is growing… wouldn’t you expect accidents to grow, too?‘

Well, yes, if we weren’t making any progress towards reducing MRI accidents, I would expect the numbers of accidents to increase about 3%, the approximate rate of MRI utilization growth over the last several years. But to tack on an additional 17% growth in accidents above the number we’d expect to see, that, in a word, is STUNNING!

And if the anecdotal information about providers seeing patient volumes drying up in the wake of the financial crisis is statistically significant, we may wind up seeing little, if any, net growth in MRI utilization this year!

The technical, clinical and financial factors in MRI are all aligned to increase the risks to patients and caregivers. Quite simply, it isn’t sufficient to rely on the ‘status quo’ safety engine. If it were effective, we wouldn’t be seeing such dramatic growth in accidents, year after year, after year.

As we approach New Year’s Eve, I propose an industry-wide resolution. I propose that we unify behind the goal of making 2009 a zero-growth MRI accident year.

This sort of resolution will require more of each of us, more than what we thought was ‘adequate’ last year or the year before. This will require a serious commitment to ACR 4-zone protocols, labeling everything within zones III and IV with the current (and correct) ASTM nomenclature, careful identification of all implants and devices, code drills and, yes… my favorite… ferromagnetic detection systems.

So while we may enjoy the forthcoming New Year’s revelry, I urge you to take a moment to consider not simply your personal resolutions, but how you can be an active participant in ‘No More in ’09′ initiative to turn the tide of MRI accidents.

The population of the United States is approximately 300,000,000 (data).
The number of MRI’s performed in the US in 2007 was approximately 27,500,000 (data).
The rate of increasing utilization has been fairly consistent at around 3% per year (data).
If this growth-rate is consistent for 2008, this year there will be approximately 28,300,000 MRI exams in the US.

Based on these recent trends, this means that the odds of a person in the US getting an MRI in 2008 was approximately 1 in 10.6, or a number nearly equivalent to 10% of the US population received MR exams in one calendar year.

The rate of officially-reported MRI accidents has grown from 40, in 2004, to 114, in 2007. (data). This represents an increase of 185% in three years of reporting. This means that the risk of an adverse outcome during an MRI exam has nearly tripled in a three-year period!

And while the gross numbers of officially-reported MRI accidents appear low, it is widely believed by MR safety experts that something on the order of 1% of the actual number of accidents are reported (source), perhaps fewer. The FDA’s numbers are so low, in fact, that the extrapolating numbers reported by those hospitals participating in the Pennsylvania Patient Safety Authority reporting (not all MRI providers in the state, by the way), Pennsylvania’s reports would appear to represent nearly twice of the total number of MRI accidents in the FDA database (source).

So, if we multiply the FDA’s reports of accidents by two orders of magnitude, a more accurate estimate of the actual number of annual MRI accidents (based on 2007 calendar year reports) would appear to be 11,400. Assuming an equal number of MRI exams (and accidents) each day of the year, this suggests that every day there are, on average, just over 31 MRI accidents per day! This is enough for most states in the Union to have a daily MRI accident.

The upshot of all of this is that if you (or a loved one) are told to get an MRI next year, it’s truly a game of Russian roulette. Has the MRI accident occurred in your state for that day already?

This is the state of things today, but it doesn’t need to remain that way.

There are a number of improvements to be made… but none of them are mysteries. Work with referring physicians to inform them of contraindication risks. Educate patients regarding screenings before they show up at your doorstep. Screen for clinical and physical risk factors. Use ferromagnetic detection. Provide hearing protection. Position patients correctly.

If we do these for each and every MR patient and visitor, we can slash MR accident rates and reduce the aerobatic growth trend, making MRI safer for every patient, including ourselves for our 1 in 10 chance of a 2009 MRI.