• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Medical Devices

  • Print
  • Share
  • E-mail

Transcript for Public Workshop - Magnetic Resonance Imaging Safety, October 25, 2011

UNITED STATES OF AMERICA

DEPARTMENT OF HEALTH AND HUMAN SERVICES

FOOD AND DRUG ADMINISTRATION

+ + +

CENTER FOR DEVICES AND RADIOLOGICAL HEALTH

+ + +

MRI SAFETY PUBLIC WORKSHOP

+ + +

October 25, 2011

8:30 a.m.

FDA White Oak Campus

10903 New Hampshire Avenue

The Great Room (Room 1503)

White Oak Conference Center , Building 31

Silver Spring , Maryland 20903

FDA: SIMON CHOI, Ph.D., M.P.H.

Senior Science Health Advisor, Office of Center Director, CDRH

DIANA RIVI, M.P.H.

Public Health Analyst, Office of Surveillance and Biometrics, CDRH

SUNDER RAJAN, Ph.D.

Office of Science and Engineering Labs, CDRH

TERRY O. WOODS, Ph.D.

Office of Science and Engineering Labs, CDRH

JANA DELFINO, Ph.D

Office of In Vitro Diagnostic Devices

MITCHELL SHEIN, M.S.

Chief, Pacemaker, Defibrillation and Leads Branch, Office of Device Evaluation, CDRH

ERICA JEFFERSON

Press Contact, Office of Public Affairs


 

SESSION 1 : MRI SAFETY, FERROMAGNETIC DETECTORS, INNOVATION IN MR IMAGING

JOHN V. CRUES, III, M.D., Medical Director, Radnet, Inc.

STANLEY FRICKE, Ph.D., MR Physicist, Children's National Medical Center

VERA KIMBRELL, Society of Magnetic Resonance Technologists

ELLEN LIPMAN, M.S., R.T.(R)(MR), CAE, ASRT

MANUEL MURBACH, M.Sc., IT'IS Foundation

CAROLYN KAUT ROTH, RT(R)(MR)(CT)(M)(CV) FSMRT, CEO, Imaging Education

Associates

ANNE MARIE SAWYER, B.S., RT(R)(MR), FSMRT, Stanford University School

of Medicine

CHARLES STANLEY, RT(R)(CT)(MR), CIIP, President, Section for Magnetic

Resonance Technologists

SESSION 2 : MRI SAFETY, FERROMAGNETIC DETECTORS, INNOVATION IN MR IMAGING

RICHARD EATON, Industry Manager, Medical Imaging and Technology

Alliance

MICHAEL STECKNER, Ph.D., M.B.A., Toshiba Medical Research Institute

USA/MITA

CHARLIE McINTYRE, Chickasaw Nation Medical Center

JOSEPH R. BARWICK, Metrasens

KEMP MASSENGILL, M.D., CEO & Quality Assurance Director, Mednovus

TOBIAS GILK, M.Arch, RAD-Planning & Mednovus SAFESCAN

ASHOK SARASWAT, M.S., B.Ed., RT(R)(MR), MRI Education Program Director,

Ohio State University Medical Center

SESSION 3 : MRI SAFETY, FERROMAGNETIC DETECTORS, INNOVATION IN MR IMAGING

EMANUEL KANAL, M.D., FACR, FISMRM, American College of Radiology, University of Pittsburgh Medical Center

FRANK G. SHELLOCK, Ph.D., FACC, FACSM, University of Southern California

JAMES MEYER, M.D., Children's Hospital of Philadelphia

SONJA ROBB-BELVILLE, MSRS, RT(R)(MR), The University of Texas MD

Anderson Cancer Center

DEVASHISH SHRIVASTAVA, Ph.D., University of Minnesota

FRASER J.L. ROBB, Ph.D., GE Healthcare, Inc.


 

ROUNDTABLE 1 : MRI SAFETY, FERROMAGNETIC DETECTORS, INNOVATION IN MR IMAGING

TERRY O. WOODS, Ph.D., Moderator, Office of Science Engineering Labs, FDA

JOHN V. CRUES, III, M.D., Medical Director, Radnet, Inc.

GEORG FRESE, Siemens AG

STANLEY FRICKE, Ph.D., MR Physicist, Children's National Medical Center

TOBIAS GILK, M.Arch, RAD-Planning & Mednovus SAFESCAN

SIMON GOODYEAR, M.D., CEO, Metrasens Ltd.

EMANUEL KANAL, M.D., FACR, FISMRM, American College of Radiology, University of Pittsburgh Medical Center

KEITH KOPP, President, Kopp Development

ROEE LAZEBNIK, M.D., Ph.D., GE Healthcare

DIANA RIVI, M.P.H., FDA

ANNE MARIE SAWYER, B.S., RT(R)(MR), FSMRT, Stanford University School

of Medicine

CHARLES STANLEY, RT(R)(CT)(MR), CIIP, President, Section for Magnetic Resonance Technologists (SMRT)

OTHER PARTICIPANTS :

LEON , Georgetown University Hospital

JASON LAUNDERS, ECRI

MAX AMURAO, Georgetown University Hospital

TERRI, Johns Hopkins Hospital

BRAD STAGER, Biomet

PAT DEVLIN, University Hospital of Cleveland

JOHAN VAN DEN BRINK, Ph.D., Philips Healthcare MRI/MITA

AMY SHOTMAN, ASRT

JOEL FELMLEE, Ph.D., Mayo Foundation

MAUREEN HOOD

CHERYL SHOATS, Johns Hopkins Hospital

HEIDI EDMONSON, Ph.D., Medical (MRI) Physicist, Mayo Clinic Rochester


 

INDEX

PAGE

STRUCTURE OF THE DAY AND LOGISTICS - Simon Choi, Ph.D., M.P.H. 6

INTRODUCTION: MRI SAFETY WORKSHOP - Diana Rivi, M.P.H. 8

DAY 1 INTRODUCTION: GENERAL MRI SAFETY, FERROMAGNETIC

DETECTORS AND INNOVATION

Diana Rivi, M.P.H. 8

Sunder Rajan, Ph.D. 16

SESSION 1 : PUBLIC COMMENT ON MRI SAFETY, FERROMAGNETIC 19

DETECTORS, AND IMPACT OF INNOVATION -

Terry O. Woods, Ph.D., Moderator

John V. Crues, III, M.D. 20

Stanley Fricke, Ph.D. 22

Vera Kimbrell 26

Ellen Lipman, M.S., R.T.(R)(MR), CAE 28

Manuel Murbach, M.Sc. 32

Carolyn Roth, RT(R)(MR)(CT)(M)(CV) 35

Anne Marie Sawyer, B.S., RT(R)(MR) 39

Charles Stanley, RT(R)(CT)(MR) 44

SESSION 2 : PUBLIC COMMENT ON MRI SAFETY, FERROMAGNETIC 48

DETECTORS, AND IMPACT OF INNOVATION -

Terry O. Woods, Ph.D., Moderator

Richard Eaton 48

Michael Steckner, Ph.D., M.B.A. 52

Charlie McIntyre 56

Joseph R. Barwick 59 Tobias Gilk, M.Arch 63

Kemp Massengill, M.D. 67

Ashok Saraswat, M.S., B.Ed., RT(R)(MR) 70

OPEN PUBLIC COMMENT AND Q&A SESSION 74

SESSION 3 : PUBLIC COMMENT ON MRI SAFETY, FERROMAGNETIC 91

DETECTORS, AND IMPACT OF INNOVATION -

Terry O. Woods, Ph.D., Moderator

Emanuel Kanal, M.D., FACR, FISMRM 91

Frank G. Shellock, Ph.D., FACC, FACSM 96

James Meyer, M.D. 98

Sonja Robb-Belville, MSRS, RT(R)(MR) 103

Devashish Shrivastava, Ph.D. 107

Fraser J.L. Robb, Ph.D. 113

INDEX

PAGE

ROUNDTABLE 1 : MRI SAFETY, FERROMAGNETIC DETECTORS, 118

AND IMPACT OF INNOVATION ON MR SAFETY CONCERNS -

Terry O. Woods, Ph.D., Moderator

SUMMARY OF DAY ONE - Sunder Rajan, Ph.D. 225

ADJOURN 225


 

M E E T I N G

(8:30 a.m.)

DR. CHOI: Good morning and welcome to the FDA Campus. For those of you joining us by webcast, welcome to about 200 of you joining us through the Internet.

I'm Simon Choi from the Office of Center Director, Center for Devices and Radiological Health.

This is FDA's public meeting designed to inform CDRH on MRI safety. The format for this public meeting will be to hear presentations from the public on the following areas: general MRI safety, ferromagnetic detectors, innovation, implants, and accessory devices. During each public comment session, each speaker will have six minutes to address one or more of the topics posted in the Federal Register.

Following the open public sessions, we'll have a roundtable discussion of experts. Each of the roundtable sessions will be moderated by a FDA moderator.

Unfortunately, we will not have time for questions from the audience following each individual presentation. We do encourage you to comment or ask questions during the roundtable sessions, and there should be index cards available at the registration table for you to fill out if you want to do it that way or you can stand up to one of the audience microphones during the roundtable sessions.

Written comments may be submitted to the public docket for this meeting by searching for the docket number FDA-2011-N-0640. That's FDA-2011-N-0640.

If you're a member of the press or media, you can contact Erica Jefferson. Is Erica here today? She's in the back there in the white. If you're from the press and you want to speak to a FDA employee, please refer to Erica.

This meeting is being recorded for transcription purposes. Each presenter and roundtable participant should identify themselves each time they speak so when they transcribe, we can get an accurate recording of who said what during this meeting.

For those of you during the breaks that would like to access your business or personal e-mail, we do have Wi-Fi available, and the password for the Wi-Fi is guestaccess, one word, all lower case, guestaccess.

This is a public meeting. I won't read all of this, but we encourage a scientific debate, and there is a distinction between dissemination of scientific, technical, regulatory information versus commercial promotion of a consultant, commercial product, or service. So during your discussion or your PowerPoints, please refrain from those kinds of comments.

At this time, I would like to invite Diana Rivi from the Office of Surveillance and Biometrics to give a brief introduction.

MS. RIVI: Thank you. Thanks for joining us. Again, my name is Diana Rivi, and I'm a Public Health Analyst in the Office of Surveillance and Biometrics as Simon said, and it's just a fancy way of saying a postmarket office.

So on behalf of FDA, we welcome you all to this two-day workshop. We're all here to improve the safety profile of what we already know as a safe technology, but we want to make it even safer. And this workshop is our way of using a proactive approach to do that. We want to hear from you about the ways that we can all work together to ensure the continued safety of patients and providers who are affected by the growing use of MRI technology.

FDA wanted to have a workshop to discuss MRI safety issues with all the appropriate stakeholders, with industry as well as with the healthcare and professional community, with you all.

We find it very important that we work together to increase the awareness of safety issues that may incur in the MRI environment and then identify appropriate solutions to reduce the risk. FDA wants to hear from and collaborate with all of you.

So Day 1 will focus on general MRI safety and the use of ferromagnetic detectors as well as the impact of innovation or the future of MRI. Day 2 is going to focus on scanning patients with medical implants and the use of accessory devices in the MRI environment and all the issues surrounding that.

So why are we here? In the next few slides, I will talk about the MRI landscape, and then describe some of the safety issues that have been reported to FDA and then explain why we need your input going forward.

So why MRI and why now? Due to its diagnostic benefits, the use of MRI has continued to increase over time. In fact, there are approximately 30 million scans performed in the United States per year. MRI use is becoming more pervasive across the care spectrum, and there's a marked increase in interventional radiology procedures as well.

Additionally, the substitution of MR for CT to reduce population radiation exposure will most likely increase the demand for MRI procedures in the future, and also the technology has evolved. There's been a shift to higher tesla strengths, and likewise medical device technology has evolved over time.

We have now lots of sophisticated medical devices maybe in, on, or with a patient at any given time. So the increased demand for scanning patients with implanted medical devices assumes additional risks, and as the amount of accessory medical equipment used in the MRI suite increases, patient and staff awareness of MRI hazards becomes very important.

So while MRI is a safe modality, there are known risks inherent to the technology and to the environment. So we know about the static field and what happens when you introduce a ferromagnetic device into that environment, and then you run the risk of a projectile event. We know about the RF field and the concern we share with undesired heating effects or electromagnetic interference. We know about the gradient magnetic field, which may result in interference with active implants or nerve stimulation, and we know about the acoustic noise that MRIs impart and, of course, here we're concerned with the patient and the risk of hearing issues if not provided with adequate hearing protection.

Other safety concerns exist beyond the MR technology. Discussions around these topics will occur during tomorrow's sessions.

So now I'm going to briefly outline the types of events that have been reported to FDA, but first it's very helpful to describe FDA's Center for Devices and Radiological Health postmarket surveillance system.

So we all know that FDA CDRH evaluates premarket submissions that come in from the medical device industry, but we also have a postmarket office, which is the one I'm in, which monitors device performance after a device has been approved or cleared for market.

MedWatch is FDA's nationwide adverse event reporting system which serves to monitor all medical device performance once they're on the market. Manufacturers, user facilities such as hospitals, and consumers can all report under MedWatch. So this is both a mandatory and voluntary reporting program.

We also have an alternate hospital-based reporting mechanism known as the Medical Product Safety Network, or MedSun, which is a group of hospitals who all report to FDA, and this mechanism provides FDA with the clinical perspective.

Mandatory and voluntary reports are entered into the Manufacturer and User Device Experience database, also known as MAUDE, and you can search this online. I've provided the link here. So medical device reports that are stored in MAUDE provide FDA with a qualitative snapshot of adverse events for a specific device or device class.

And I want to plug the importance of reporting these issues to MedWatch because anyone can file a report, and we encourage that because this is how FDA learns about what's going on with marketed devices. Of note, there's some limitations to our data. For example, there's known underreporting of medical device events, and of the reports we receive, they vary in quality. For example, not all events indicate a root cause. So we don't have the ability to establish causality between the device and the event.

And also MAUDE data is not intended to be used to evaluate the rates of adverse events. So any apparent trend in numbers should be interpreted with caution.

So there are many variables to consider when we talk about the reports that have been submitted to FDA. So please keep these points in mind as I outline the types of events that we've seen.

So we conducted a broad search of MAUDE data over the past five years, from 2006 to 2010, and we wanted to capture reports that involved all devices and products likely to enter the MRI suite, even items that are not considered a medical device but might pose a risk, like buffers and mops and things of that nature. We didn't limit the search to any one product or product code. It's simply a search of all reports that FDA has received each year, and of these reports, we conducted a text search of MR and MRI and then, of course, a manual review was conducted for applicability. So that's kind of the framework that we were using when we were reviewing the reports.

So after reviewing all the reports received by FDA in this time period, we categorized the most frequent reported device safety concerns. Thermal related issues were the most commonly reported, followed by compatibility issues, then hearing issues, projectile events, as well as nerve stimulation or shocking.

But what's most interesting about many of these reported events is that they can be prevented, and that's what makes this topic so interesting to us. Safety information about MR hazards in addition to safety guidelines and recommendations, they all exist and they're out there, but not all are followed. As I describe the reported problems, I'll also go into how we think they can be prevented.

So we're defining thermal-related issues to include patient heating, burn, blister, redness, or an indication that the patient experienced warmth during the scan. While not all reports specified the severity of the burn, we received quite a few second-degree burns, some blisters and third-degree burns as well. Please note, there are multiple reasons for thermal-related injury reports.

Several reports do attribute the cause, and sometimes there were a combination of factors that came into play. For example, some reports indicate that the patient was not padded or, when they were, inadequate padding was used, like sheets for example. Some reports indicate that a patient was not positioned properly, and they formed a closed loop. Other reports indicate that the patient received a burn because they came into contact with the bore or the coil during their scan. And others indicate that a patient may have been burned because a device was in, on, or with the patient at that time.

And what's most interesting about these is that they're likely preventable if certain clinical practices are followed, such as thorough patient screening practices or gowning practices, adequate padding and positioning practice, all of which stem from education.

So when we reviewed the compatibility-related issues, many reports indicate that a device, such as a pain pump, for example, became nonfunctional after being exposed to a MRI. Motor stalls are routinely reported to MedWatch. And we see issues with compatibility labeling, notably the safety labeling was missing or the labeling stated the device was not MR safe, yet a scan was performed anyway. And, again, compatibility related concerns and labeling concerns are all going to be discussed in depth during tomorrow's session.

So much of what we're seeing in the reports that describe hearing issues may be attributed to staff practices as well. Of these reports that indicated cause, the reasons for hearing issues are associated with no hearing protection is provided to the patient or faulty hearing protecting was provided. So there was an incorrect fit of what was provided to the patient, and it fell out during the scan.

Similarly, hearing issues do not always specify the patient outcome, but of the reports that do, the most common reported outcomes include ringing of the ears, hearing loss, or the patient had to follow up with an audiologist after their MRI scan.

The most reported projectile incidents are attributed to site access issues where staff brought in ferrous devices or products, labeling issues where compatibility labeling was not available as well, as screening oversights. So patients had metal on their bodies or staff had brought in objects with metal.

While these projectile events are low probability, they carry very high risk to patients, providers, and hospital staff, as we all know. And, again, these events are preventable with patient and provider screening practices and facility protocols that are in place, site access restrictions and site markings, as well as facility-wide education.

Furthermore, the recommendations to reduce the occurrence of these events are outlined in ACR Guidance Document and other published literature, and we'll be discussing the role of ferromagnetic detectors as one solution, or not later today.

Nerve stimulation is a very complex issue, and unfortunately the reported events don't provide much explanation as to why a patient experienced shocking or nerve stimulation, but half of these events implied that it might have been associated with an accessory or implanted device.

So as discussed, preventing harm is likely possible if the users are aware of the known hazards and follow existing recommendations pertaining to MRI safety.

We want to hear from you about ways that FDA can collaborate with stakeholders to increase awareness and learn ways to address preventable events that may be addressed through clinical practice solutions, things like: facility protocols, i.e. patient screening procedures and protections, staff safety awareness and education, MR site restriction and what have you.

And also what can FDA do within our regulatory framework? How can FDA partner with professional organizations and accrediting bodies to ensure safety?

So now my colleague, Dr. Rajan, will introduce today's sessions before we dive into the meat of it. Thank you.

(Applause.)

DR. RAJAN: Okay. Good morning. My name is Sunder Rajan. I'm from the Office of Science and Engineering Labs, and I just wanted to go over a few more points to set the stage for today's events.

Thank you, Diana. Diana has done a good introduction to the general areas of MRI safety. I just want to say a few words about ferromagnetic detectors and innovation and how they can impact safety.

Okay. I don't think this audience needs an introduction definition to what is a ferromagnetic detector. So we at FDA know that this is not a regulated device, but it's an integral part of MRI practice these days, and so we really wanted to find out a little bit more about how to improve safety, what are the pros and cons of their use, and other considerations.

So we are in the process of conducting a small survey, at about a half a dozen sites, to get some feedback more or less in preparation for this meeting, to see if there was anything interesting or surprising to learn, and the survey questions mostly focused on general experience at the sites, what kinds of quality control practices, what kinds of experiences they have with projectile events and the other related safety practices, and the sites included basically the handheld detectors and the pole types and the archway types that are being sold out there today.

And what we heard was that basically the archway and pole types are the expensive types. They rely on the vendors to be installed and to be adjusted for sensitivity. They remain on all the time, and these devices commonly cause alerts. We call it false positives, not that we did a scientific statistical analysis; basically these are just nuisance alerts that tend to annoy the customers, nonetheless probably useful.

About the handheld systems again, there's no formal training required. These are generally turned off. There's a battery issue. These have to be turned on and off all the time, and they're used in Zone 2 or 3 by the techs. The alerts provided to the items that are on the patient, items such as jewelry and so on, and they are a little bit invasive because the operator has to bring one of these devices close to the body to detect something. Sensitivity, therefore, is not clear, and these, what we heard, serve as a good adjunct for patient screening.

So really there's no major surprises here. These ferromagnetic detectors are quite effective in preventing harm, it appears. However, there are probably some room for improvement, and because these are only as good as the installations and the people and the policies, there appears to be need for better standardization. These can provide a false sense of security. The handheld units serve as good adjuncts, and we really need to understand a little bit more about the sensitivity of these devices. Perhaps they're too sensitive.

And installation is an important part. In other words, if these are installed during the installation of the site, they appear to have less practical concerns as opposed to installing them after the fact, and then you have to place them someplace where it's not very convenient or causes false positives.

And, finally, I think that going forward, MRI is a fast changing field. There are lots of new and exciting things happening, and I'm sure some of these new things, innovations, can impact safety, and we just wanted to at least have this topic heard, hear from you, to see what sorts of new technologies are out there and what kind of potential safety concerns we should be looking at and be aware of, such as the magnet designs, coil types, you know, multi-transmit coils, new procedures, biopsies, interventional procedures, and finally, of course, the hybrid PET-MR systems. So soon you're going to have a bunch of PET technologists running around in the MRI suite as well. So we have to be careful about these innovations.

So really to summarize, we request your feedback on these various aspects, and indeed I am personally very excited that this meeting is happening today. A number of us have worked very hard over the year. We wanted to have this last year, and it's taken a year and a half to make this happen. And this is really your meeting, and we want to hear from you. We want to hopefully take back some useful information so that next time a person dies in a magnet, it should be of natural causes, and we want to make sure that we make MR, which is a tremendously safe modality, even safer.

So thank you all for coming here, and hope we have a very exciting session today. Thank you.

(Applause.)

DR. WOODS: Good morning. My name is Terry Woods. I'm also from the Office of Science and Engineering Labs at the FDA. I'm going to moderate the public comment sessions today. So, again, this morning we'll begin with the first comment session. If the speakers for that session aren't already in the front first three rows, if you could begin moving up to the first three rows. The way this is going to work is you each have six minutes for your presentation. Because we do have a lot of presentations today, we're going to have to ask you to stick to your six minutes. I have a timer that I'll set, and when we get to about 30 seconds from the end, I will raise my hand and let you know that you're reaching the end of your time.

Let's see. Our first speaker today, and I guess, again the themes today are MRI safety in general, ferromagnetic detectors, and innovation. Our first speaker is Dr. John Crues. He's the Medical Director of Radnet.

DR. CRUES: Thank you very much. I appreciate being asked to give a comment today. Most of you don't know who I am, but I am a diagnostic radiologist. In the past, I was Chairman of the ACR Committee on Magnetic and Biologic Effects. I've co-authored 18 papers primarily with Frank Shellock, who you'll hear more about later today. I'm also a former member of both the ACR MR Accreditation Commission and ICAMRL. I was on the ICAMRL Board of Directors up until last year and former Chairman of the ISMRM Safety Workshop.

First I'd like to emphasize that MR is very safe. We looked into the last year and half of data that we have from our company. We had 785,584 MR examinations over that 18-month period, and these show the different types of adverse effects that we had that were significant, 10 burns, 4 shocks, 4 metal abnormalities, and I'll talk a little bit more about those in a minute, 2 hearing, 1 pacemaker, 57 contrast reactions. So by far, the biggest problem that we have that's significant is really contrast reactions. Of those, 44 we ended up calling 911. We had no deaths. A total of 122 adverse effects.

If you look at those that required 911, it turns out 20 percent were either anger management or panic attack, 7 percent syncope, 7 seizures, 1 abdominal pain, hearing we had implant problems but only 10 percent of the time, 10 burns, and 38 contrast reactions. We had no projectile problems in the 785,000 studies.

The other thing I wanted to point out, I think there was a question about what the people use as policies and procedures. Policies and procedures which are written down really include that all MR personnel are screened prior to hiring and after any procedures. All patients and non-MRI personnel must complete screening forms and be screened by MR technologists. All security personnel must be screened personally and for ferromagnetic devices before entering into the magnet room, and all restraint devices must be plastic. That's really what we have written down. Really the book of safety is actually the training of the technologists who run the center, and the safety really covers more than just what's in the policy and procedures.

And I just have a few comments that I'd like to end with. I think guidelines have a lot of value when supported by objective evidence, but we really have little cost efficacy analysis that any of the potential guidelines have been recommended out in the community. So I think we really do need more cost efficacy analysis.

I think the ACR zoning is a very valuable concept, but in the outpatient setting, which all of this is outpatient MRI, no inpatient MRI, the concept would be expensive and impractical to actually design the suites around that concept, and again, there's very little evidence in the outpatient setting that it would be cost effective.

I agree with the ACR White Paper that people are really the most important factor, and I really think training is by far the most important thing.

Ferromagnetic detectors, we don't believe are necessary in outpatient screening. We really had no significant adverse effects that they would have inhibited. So we really don't think that they would be cost effective.

And then finally, I believe that MRI safety is a rapidly changing area. There are new devices all of the time, and I think that the existing certification board with respect to the ACR and ICAMRL are able to rapidly change certification requirements in the marketplace and put those in their criteria, and they also have mechanisms to enforce those on an ongoing basis. So I think the new certification agencies will really be the key point and the best way for us to go forward in enforcement safety. Thank you.

(Applause.)

DR. WOODS: The next speaker this morning is Stanley Fricke. He's with the Children's National Medical Center in Washington. Dr. Fricke, welcome.

DR. FRICKE: Well, thank you. We're anything but an outpatient center at Children's National. One of our biggest challenges and quickly growing is that we anesthetize a significant number of our patients, and because of that, we have to have a lot of support equipment in the room, many times physiological monitoring which probably most of you do. We have ventilators, full anesthesia setups, and it can get complicated.

Our patients come from several departments, our neonatal intensive care unit, our pediatric intensive care unit, and several other units in the hospital. Each unit brings their own support set of personnel, and because of this reason, part of my job as the MR safety officer is to train, I don't know, the whole hospital, 5,000 employees, and it's difficult.

Part of the difficulty is, and as we all know, the monitor should be no metal, no metal, no metal. And then the first question that comes up is, okay, but how much metal can we bring in the room? So I'm going to start out in the theme of Children's National Medical Center by saying big metal, little metal, small metal, ferrous, nonferrous, and this may sound like Dr. Seuss, and that's the way it feels sometimes during the day. I'm not quite sure how to operate this, but -- okay.

So I decided to show some clips from our video surveillance system, and if you take a look at this system, I wonder if anybody can point out the problem. It certainly is not obvious, but -- is there any way to do the pointer here or the laser? Okay.

If you look up in the center frame, you'll notice two people entering what is our Zone 3. Neither of these people are technologists. One person is a respiratory therapist, and the other person is a nurse, both trained in MR safety. Both of these people have been specifically trained in the device that you see coming through the door, and both of these people understand no earth metal objects are supposed to go into the room, and what do you expect happens?

Well, let's back up in time a little bit. Here we see our MR technologists coming into the room and blocking open the largest door that we have to Zone 3.

Now, why do we have two doors on Zone 3? The primary reason we have two doors is because the large door is to bring in the support equipment and the entire village that will come with a complicated case. The small door is a pedestrian door that is supposed to only have access with a metal detector on the inside to non-MR technologists. This would be everyone who is not a MRI-trained and certified technologist. This has been put in place primarily to not allow anyone that isn't trained to bring equipment into the room.

You can notice on the bottom right-hand corner, one of our induction bays for anesthesia, and you can probably count four or five people, some people ducking in and out of the picture there.

Now, in this frame, we have a situation that takes the MR safety officer aback, and -- it out here, because we have the Zone 3 door blocked open, and now the Zone 4 door. And I would also ask you to examine these pictures and see if we see anything else. It won't be obvious, but in the bottom right-hand slide again, you'll notice somebody coming through a set of doorways in the hallway. Well, that set of doorways is actually the Zone 1 and 2 barrier. All of these are card keyed access. So basically we have a complete path that goes from Zone 1 into Zone 4. All the doors are open, and the worst part about this picture is nobody's looking.

Well, a few things happened, but you can see the village is in preparation to bring the child in. We have a respiratory therapist who is checking out his machine, the ventilator, and I will point out at this time that if you read the ACR guidelines and most other guidelines, everybody is familiar with some sort of form that is used to try to identify metal that's in the patient, but that's not the only metal we're worried about. So at Children's National, we've decided that we have to include an equipment check, and this check is to be done just before metal comes into the room, and needs to be signed by a doctor, by a nurse, and by our radiology technologist.

Well, what happens? Well, the obvious happens. We go into the room and we pass the 200 gauss line with this machine. By the way, this is a ventilator that is MR conditional, which shouldn't hurt anyone. If you notice in the top left-hand corner, tubes are behind the respiratory therapist, and the machine is stuck to the magnet. Two people cannot pull it off of the magnet, and what we have is now the trained nurse leaving and abandoning the therapist to his own and, of course, we have a ventilator, $75,000, stuck to the magnet.

Could a metal detector prevent this? No, because the metal detector would, of course, beep when this went through.

This was held up to the magnet by myself when we first purchased it. It didn't stick. It's MR conditional. That's our metal detection challenge for the day.

(Applause.)

DR. WOODS: Our next speaker is Vera Kimbrell from the Society of Magnetic Resonance Technologists.

MS. KIMBRELL: Good morning, everyone. I don't have a PowerPoint presentation. So listen real close, okay.

I have two roles today. First of all, I am on the Executive Committee of the SMRT. So we, of course, have a statement. Also, I'm a MRI technologist and have been a MRI technologist for 20-something years, pretty much since the beginning. So just a few things that I'd like to bring up.

First and foremost, as a Society, we support the ACR Guidance Document as a basic starting point for MRI safety, but in addition to the ACR Guidance Document, we have just a few things.

First of all, a plea to the manufacturers of implants and devices for better documentation and more useable, easy to read and understand documentation. We also have a plea to the manufacturers and vendors of equipment to give us better guidelines and better safety information.

We also believe that since MRI is the number one and foremost role of a MRI technologist, that you recognize that we are the gatekeeper of the door, of the door to Zone 4, of the door for the patient to enter the room and everything else to enter the room. The training that we get and the licensure that we hold is very, very important. Perhaps this will be addressed by the CARE Bill, but if not, then we have a long way to go towards certification and licensure of a MRI technologist and staff.

We also request that the FDA focus more on education, not only education of professionals, but education of the public, education of the referring community, education of anybody that's going to use the MRI scanner.

MRI scans are more prevalent as we age. As we, the baby boomers, need medical care, almost everyone needs a MRI. It's important that we don't withhold or deny or delay care to patients, and to do that, we need concise, easy to understand information. We need to be able to get that information quickly. We need to get these patients scanned safely, and in order to do that, we need cooperation from everybody, from the manufacturers of each of these devices, to the radiologists, to the referring community, and to the patients themselves to have a greater awareness of just exactly what they have in their bodies.

So thank you. I appreciate the invitation to be here and have a good day.

(Applause.)

DR. WOODS: Thank you. Our next speaker is Ellen Lipman who is from the ASRT.

MS. LIPMAN: Good morning. I'm Ellen Lipman and have been a registered radiological technologist for 33 years and a registered MR technologist for 14. I'm currently the Director of Professional Development with the American Society of Radiologic Technologists. ASRT is pleased to participate in today's discussions on magnetic resonance safety and appreciates the opportunity to share with the FDA some of our suggestions on how to make MR safer for patients and technologists.

ASRT is the world's largest radiologic science organization with a membership of nearly 143,000. Ninety-five percent of our members are technologists who are certified by the American Registry of Radiologic Technologists. ASRT members practice in all medical imaging disciplines using both ionizing and nonionizing radiation. ASRT has a robust magnetic resonance chapter with more than 11,000 members. We also represent around 40 percent of the ARRT's 27,000 registered MR technologists.

ASRT publishes and maintains the ASRT Practice Standards for Medical Imaging and Radiation Therapy, which includes practice standards for the magnetic resonance community. The ASRT also produces the MR Educational Curriculum.

Completion of an educational program accredited by the Joint Review Committee on Education and Radiologic Technology and primary or post-primary certification in MR by the ARRT is the minimum standard recognized by the ASRT House of Delegates. The ASRT also believes there should be consistent technical personnel standards. The ASRT advocates for this approach in the CARE Bill.

MR safety extends beyond the patient. I'm sure most of you have seen the pictures of maintaining and cleaning equipment stuck in the MR scanner. It's crucial for facilities to understand MR environment safety measures.

Implanted devices also pose a problem. The MR's strong magnetic field can affect devices. In addition, many patients don't know what type of implants they have. This can lead to inaccurate information during the MR screening process.

Finally, there's a lack of clear understanding of what sort of MR related medical mishaps and accidents are reportable and to whom they should be reported.

It's intuitive that an individual with specialized certification may contribute to a culture of safety in the MR suite more so than someone who has not completed a similar rigorous process. However, this is not the case in imaging, particularly in magnetic resonance where states have stronger and more stringent requirements over hairdressers than they do over who may perform MR imaging.

It's also intuitive that educationally prepared and clinically competent MR technologists who follow ASRT's MR standards of practice will apply safety principles in the clinical settings.

Also, a collaborative team approach that involves radiologists and referring physicians, medical physicists, MR technologists, facility administrators, architects and engineers can enhance patient safety. Appropriate metrics and data collection methods should be developed and implemented.

While not directly involved in the development of the ACR Guidance Document for Safe MR Practices, ASRT supports the use of this document. However, it's important to note that the standards in this document regarding technologists' qualifications are inconsistent with ASRT recommendations that all MR technologists be either primary or post-primary certified by the ARRT.

Stakeholders should develop standardized educational materials in multiple languages focusing on MR safety for patients and for direct patient care staff such as nurses, technologists, physicians, and anesthesiologists. Safety materials should also be developed for other facility staff such as maintenance, housekeeping, and materials management.

Education can help prevent magnetic projectile accidents caused by staff who may not be familiar with magnetic resonance imaging principles and hazards.

ASRT recommends that manufacturers provide patients with the name and model of the implanted devices and whether or not it is considered safe for MR imaging. Patients should receive this information in writing, and it should be incorporated into the patient's electronic health record if possible. In addition, device manufacturers should make implantable devices that are safe for MR imaging.

At this time, there is still question as to what constitutes a reportable event into the FDA database system, and there's no clear understanding of who and how should near misses be reported. Many technologists are not aware of reporting systems and what type of incidents should be reported. It is important that all reporting on mishaps and near misses must be nonpunitive for objective data to be collected and solutions created to improve the quality and safety of MR imaging procedures.

ASRT supports MIPPA and believes that the valid accreditation programs that focus on the imaging process used by the technologists and the interpretative skills of the radiologists along with quality assurance measures should be an integral part of imaging facility operation. MIPPA should be expanded to include all advanced diagnostic imaging settings.

ASRT believes that it is imperative that all accrediting organizations require modality-specific advanced certification for technologists performing imaging exams.

ASRT appreciates the opportunity to present today with many of our organizational colleagues as we work together to create a safe and productive MR imaging environment. Thank you.

(Applause.)

DR. WOODS: Our next speaker is Manuel Murbach from the IT'IS Foundation.

MR. MURBACH: Good morning. I'm Manuel Murbach from IT'IS Foundation in Zurich, Switzerland, and we're dedicated to thorough exposure assessments in MRI and other radiofrequency assessments. So today I'm going to share some recent results of studies in the safety evaluation of local exposure in MRI.

So as you may know, the whole body exposure in the MRI environment can be assessed easily by the power budget, but the local SAR or the local temperature distribution is known to be assessed numerically.

So for this study, I performed a lot of simulations at 1.5T with a generic birdcage model and circular polarization using 9 anatomical human models, which I show in the next slide, and in 10 different set access positions. So from head imaging position to calf imaging position and the patient was mainly lying on the back and the effect of RF loops were investigated separately.

So this is the positioning configuration. You see here our pregnant model in the generic birdcage and the positioning. So said axis position of 0 is defined here as the heart at the Iso-Center, minus 250 millimeters will be the head and plus 825 millimeters the calves in the field of view of the MR birdcage.

These are the models used for this study. We have sort of normal models, like Duke and Ella, and then also obese and Japanese and children model and also one pregnant and fetus model shown here.

So for the results, this is the 10 gram peak spatial SAR value shown. The solid line is the average and the color shaded area is the minimum/maximum spread across all these different anatomical models, as a function of landmark position and normalized to first level control operating mode according to the standard.

So you see here for all models, we can reach up to 100 W/kg load for 10 gram SAR for legs in the field of view and where the limit is the partial body SAR, but also for trunk imaging positions, you can get up to 80 W/kg in certain models, and children are exposed to somewhat lower local SAR, and the fetus, that's the dotted line on the right, to even lower local SAR since there also the exposure of the mother is limited to normal operating mode.

This is a comparison. The red line is usual standard occupational SAR limit of 10 W/kg. So we are way above this. These are the same numbers. So I'm not sure whether the movie's working, but now to translate, the SAR into temperature has some uncertainties. On the left you see the 10 gram peak spatial SAR distribution for upper sternum position exposure scenario and on the right -- I believe this is probably not working or click on it maybe. Okay. There you will see on the next slide, I just have a still image, you see a 30-minute heating with this exposure, and you see the 39 degrees centigrade iso-surface view after 30 minutes heating.

So, and in this -- we performed a volunteer heating experiment, putting two temperature probes. Here everything got shifted a little bit. That T1 is on the left shoulder, T2 on the right shoulder, and beneath it was roughly first set of controlled operating mode and got temperature increase in the shoulder of about 4.5 degrees kelvin, and this can only be explained with a detailed local perfusion, local block perfusion model.

So to conclude, so we have heat spatial 10 gram values in adults that may exceed 100 W/kg and potentially even much more when considering RF loops. So general RF safety consideration are not applicable here.

The temperature evaluation, if we don't use a temperature dependent local perfusion model, this would result in temperature increases of more than 10 degrees kelvin in steady state. So we need the detailed temperature modeling.

Local SAR hotspots in children, the good news is generally lower than adults due to the smaller cross-section and therefore smaller induced eddy currents. So we don't need special safety considerations in children for the local SAR.

However, for the pregnant model, we had a very average high SAR in the amniotic fluid of about 4 W/kg average even with the mother exposed to normal operating mode of 2 W/kg. So we need thorough additional thermal evaluation there because there might be some limited heat loss capability of the fetus.

And last, so that the tools for this comprehensive in silico evaluation are available.

Thank you very much. That study was supported by the MRI+ Framework Project.

(Applause.)

DR. WOODS: Thank you very much. Let's see. Our next speaker is Carolyn Roth who is the CEO of Imaging Education Associates.

MS. ROTH: Thank you for having me to present today. My name is Carolyn Roth. Most of you know me as Candi. I am a MR technologist. I'm a technologist by trade. I'm also an educator, and I'm here speaking to you today on behalf of myself and of MR technologists, and also my colleague, Bill Faulkner. Bill was unable to attend this meeting, as he's running an educational course this week, but Bill and I kind of went back and forth to discuss what we felt was urgent to be discussed in a meeting of this type.

We understand technologists want a safe working environment. They want to know what the rules are. They don't know what the rules are at this time. The White Paper on MR Safety is a great starting point, but it needs to be a living document and technologists want to be involved in that.

They will follow the rules. They may bend the rules on occasion. One thing we need to understand well is that techs will find shortcuts. I had a technologist at one of my sites with 12-plus years experience who went into the scan room with scissors to open her saline bags, and she knew how far from the border scissors could be and still be okay until the day she left the scissors in the room and a new tech the next day brought them and they nearly went through a patient's head. So they will bend the rules.

Essentially I think that training is key. I mean any device, like a ferromagnetic detector, is only as good as the person who uses that device. So a device in and of itself, I can buy the best golf clubs in the world, and I can still only drive 50 yards.

One of the things that I have encountered, Bill and I have encountered over the years in imaging, is imaging has become more automated. From a quality perspective, there's photo timing, there's protocoled scanners, and what has happened now with automation is that skills are going away on occasion, and so we feel that education is going to help keep these skills current and fluent within the technologist world. I've watched technologists now use automated technique settings and not even know what TR and TE they're using. So if that's happened, from an image quality perspective, my fear, our fear is that that may also happen if we start to automate the screening process. So we can't do that. We have to continue to make clear and concise training for everyone.

So what kind of training do we need and who needs to be trained? Essentially we need to train not only the healthcare workers, the ancillary healthcare workers. Yes, the technologist is the gatekeeper and, yes, the onus almost always falls on the technologist to block the scan room door, but we need to be training the radiologists. The patient needs to be trained as well. Right now the patient takes information from Hollywood. They watch an episode of House, and then they believe they understand MRI safety.

The patients also need to understand how serious this is because patients lie. We've had patients who have literally told us I don't have a pacemaker, I don't have a pacemaker, got to the scan room door and say you know what? I do have a pacemaker. I deliberately told you I didn't because my neighbor told me you wouldn't scan me if I told you I had a pacemaker. So they need to understand as well.

In fact, radiologists also need to be trained. It's interesting, Bill was at an imaging center recently, and the safety manager of that site was a radiologist, didn't come to the safety workshop because he had seen a video once and Philips wouldn't sell him an unsafe piece of equipment. So it's very scary that the safety training is very lacking and very loose and inconsistent site to site.

So what kind of training do we need to provide for technologists? We need to provide safety associated with the static field. Yes, projectiles are an issue, but they're not the only issue.

The incident that you see here was one that happened overseas but recently something similar happened in the United States, and essentially a ferromagnetic detector, yes, might have stopped that had the patient or had the individuals known, but the individual to whom an incident similar to this occurred essentially said I thought the magnet was off. So a detector could beep all day long. If the individual bringing the metal into the scan room doesn't understand that the magnetic field is always on, then the detector is of no use in that case. So, yeah, a detector would be helpful with the proper training.

Who needs training? Again, everybody needs training. They need training on the static field. They need training on the RF field. Not too long ago there was a report of a patient who wasn't changed into a hospital gown and their blouse literally caught on fire. So these are things that people need to understand.

People need training on gradient fields. I can't tell you how many lawyers I've gotten phone calls from who have told me a patient has tinnitus one year out from a MRI. The patient has hearing loss one year out from a MRI. So we not only need to understand how these different aspects of MR work, we need to understand how to prevent incidents in these areas.

Not only do we have to train the in-house persons, the technologists and the radiologists, we need to train ancillary persons. Recently there was smoke in a MR imaging facility with whom I work, and the firefighter came in and literally pushed the technologist out of the way and said, once you dial 911, I own this center. Luckily, the tech could reach the clinch button, reached in and clinched the magnet. That was his only recourse, essentially, potentially saved the life of the firefighter.

So at the end of the day, my goal is training. I don't want us to rely on Hollywood for that training. They don't get it right. Is it the onus of the center? Is it the responsibility of the vendor? Is the Government responsible? I think we need to get some kind of documentation together, some kind of safety standards such that our safety training and our safety understanding is consistent site to site to site. Thank you for your time.

(Applause.)

DR. WOODS: Thank you. Our next speaker is Anne Marie Sawyer who is from the Stanford University Department of Radiology.

MS. SAWYER: Thank you, Terry. Good morning, everyone. So I'm here to talk to you today about technologists, not technologists like myself because I consider myself very fortunate that I'm at Stanford, and I have a safety committee made up of two very brilliant individuals and myself. One is a Ph.D. who has been in MR for hundreds of years, and one is an M.D. who has also been in MR for hundreds of years. So I don't want to talk to you about that.

What I'd like to talk to you about is something following up what Candi was saying about the lack of education, lack of required annual, ongoing education by technologists for technologists.

There's very rigorous testing that goes on to get these MR scanners approved by the FDA, but then they're out there, and they're out there and there's no required annual, ongoing training for the people, the technologists like myself who are operating these, and it's a whole list that we can see there, policies and procedures, so on and so on, and I'm really preaching to the choir today because I think probably you all know this and you support all of this.

There's also a lack of focus in my opinion on patient burns. Is it radiofrequency? Is it the varying gradient magnetic fields? And how do we as technologists differentiate? And so how do we protect and prevent these things from happening so we can protect our patients? This is, after all, why we're there is healthcare, for the patients.

I know many of you think that this may be the main problem in MR, the main safety issue. I don't think so. I think these are our main problems, and these are where we should be focusing, and the only way that we can focus on this and learn, and I'm not doing this right, again, I'm going to go back again and again and again, like Candi was saying about education and training, knowledge and support, and support by the administration, support by whoever manages your facility, whether it's M.D.'s or someone else. It takes money. It takes time, and people aren't doing it now, and I'm sure they're blaming it on the economy like everything else.

So here's a long list, and as technologists, we don't understand all these things and why they happen, and you can pad and you can do all these things, which I think are very important, but we need to understand why these things happen in each and every situation, because if we don't understand, we can't develop that critical eye that we so desperately need to protect our patients.

Risk again, varying gradient magnetic fields. We don't understand these things either as technologists. Many of you do and you're experts in this field, but the large majority of technologists out in the field, they don't understand these things. They don't understand the difference between a transmit and receive head coil versus a receive only, and this is why some of the burns have occurred because they don't know. Either they want to know and they don't get support, they don't get the funds for training and education, or they just don't care because they've just not been inspired to care, and I really blame this on the people that run the facilities, the administrators and so forth. They don't understand that when you're transmitting with a body coil, you're transmitting over a much longer length than you are with a transmit/receive head coil. They don't understand these things, and this I will use from Frank Shellock, they don't understand this either, that some conductors may be more dangerous at 1.5T than they are at 3T. Everybody just naturally assumed as we go up in field, that they become more dangerous, but that's not the case. Again, we as technologists, we don't understand how these things work, and we need to understand, through training, through regulation, through standards, through policy, something.

This is our problem, and now I know you can look at many of these and say, oh, yeah, that happened because that happened, and we always say hindsight is 20/20 but, you know, we need to understand these things before we go into it. We need to understand padding, a quarter inch of air guaranteed. I learned this from Joe Schaefer a long time ago. He's sitting out in the audience today, and if you can't pad and if you can't guarantee a quarter inch of air, you don't scan that patient, and you have no idea how many facilities are out there today that don't pad their patients when they're transmitting with the RF body coil, and they're just lucky that they've gotten through all these years without a burn. So they don't think it's a problem until it is a problem.

So how do we protect our patients from all the accessory devices that we put into the magnet? And I can tell you, I work at a research facility where about 80 percent of what I do is functional brain MRI, and we can't stuff enough devices in there, EEG, eye trackers, this and that and everything. So this sort of thing is just going to get worse.

How do we protect our patients from things that break down such as the body coil in this instance? These things aren't perfect. They're like toasters. They break down sometimes. So we have to protect our patients from this.

How do we protect our patients from themselves? They conduct just like one of the conductors.

So what is the potential value of ferromagnetic detectors in MR screening? I think we know what these all are. I'm concerned that perhaps they don't get to the internal metal. I'm concerned that people get into the false security of these false positives. I know I was just reading on the plane here about do car alarms prevent theft? Well, no, they don't because everybody has just, they ignore these things because they go off all the time. So it doesn't matter if you're the owner of the car or the person stealing the car, people just ignore them.

So what is the potential value? Again, if you're using these things and they work at your facility, that's great because you probably have really good training for your technologists and they really have a critical eye and so they truly are using these things as an adjunct, but I don't think they'll be used like that at all facilities.

And you do really have to change, and if you don't have the facilities to change your patients into scrubs or gowns, I really don't think you should be operating because I think it's critical that you take the street clothes off so you can focus on more important things like biomedical devices and implants.

And, again, other people brought this up. It's not my idea, but again you hear these alarms all the time, and it really doesn't tell you what does that mean? It's not going to stop you, like Candi showed the table going into the magnet.

Thank you.

(Applause.)

DR. WOODS: Thank you very much. Our speaker for this section is Charles Stanley who is the President of the Section for Magnetic Resonance Technologists.

MR. STANLEY: Thank you for the invitation to be here. I'd like to start by reading the vision and mission statement of the SMRT, and I think that will go a long ways towards telling you why we are glad to be a part of this conversation today.

First of all, our vision statement says that the SMRT strives to be the leading professional organization in the medical imaging community, providing education, professional advice, and support for MRI technologists and radiographers throughout the world.

Our mission is to enable MR technologists and radiographers throughout the world to achieve professional excellence through the provision of high quality education. We support and promote worldwide communication of current and emerging clinical and research information in the field of MRI, and we provide a forum for its dissemination. Most probably everyone in this room is part of the MRI LISTSERV.

We develop relevant and accessible member services and resources to enhance professional development, and finally we provide informed advice and support to our global membership to enable the SMRT to work with local, regional, and federal agencies and organizations to accomplish the above objectives.

So I think that goes a long ways towards telling you why the SMRT is excited about this conversation and why we're thankful that we had the invitation to be here today.

So we would like to stand in support of the ACR Guidance Document as a minimum standard to start the conversation around MRI safety. We do that with the following caveats though. We feel the technologists working in MRI facilities must be educated and licensed and certified. We feel that the passage of the CARE Bill will provide a regulatory framework for this, but even that bill does not adequately address the specific level of safety training and retraining needed by each and every technologist working in MRI.

We also think that clear directives are needed to manufacturers of implants and devices to more clearly define what MR conditional labeling means, and we also call upon the vendors to provide us spatial gradient maps for each and every system that we operate.

So, again, we're excited that this conversation is taking place, and that could and should lead to greater clarity around some these issues, and we ask that the frontline defenders of safety, the technologists, be kept a significant part of all discussions and decisions regarding this.

For the remainder of my time, I'd like to make a couple of personal comments outside of the SMRT.

So my personal statement is that I have been a technologist for 25 years and a certified MRI technologist for 13 or 14 years, I'm told; now I don't remember. I've managed technologists at two major academic medical institutions in the U.S. I've had literally tens upon tens of technologists work for me. I can say with 100 percent certainty, at those academic medical institutions, that M.D.'s do not review every screening form for patients. Okay. And, again, that's not a dig at the M.D.'s. I'm just telling you happens in the real world. So instead they rely on the screening practices of technologists to bring to their attention potential issues. So essentially they don't know what they don't know, okay. So what I'm saying is, is that they're only as good as the weakest screener in the technologist pool.

So with that, I would again from a personal standpoint recommend licensed, certified technologists with standardized, mandatory training and retraining every year. Every major industry or profession that operates potentially dangerous equipment requires this type of training, whether we're talking about airline personnel, bus drivers with CDL licenses, or nuclear power workers. Our field should require no less level of training. Thank you.

(Applause.)

DR. WOODS: Well, thank you to all the speakers for this first section. We really appreciate your input. This is your opportunity to talk to us. You know, at the FDA, we deal with device manufacturers a lot, but we don't get feedback from clinical people who are in the sites every day very often, and again we thank you very much.

If you have additional information about how things are working, what's not working, and even more important if you have specific ideas about things that should be changed, please go to that docket that Simon gave you the link to before. It's also that the website that you used to register.

Right now, why don't we take a 20-minute break or so. Come back at about 10:00, and we'll get started with the next section.

(Off the record.)

(On the record.)

DR. WOODS: We're ready to get started with the second comment session, and because everyone's doing such a wonderful job of staying on time, it looks like we may have a little bit of extra time after this session, and if we do, what we will do is, there are microphones in the room. So if people have comments or questions for specific presenters from the public session, after all of these talks are over, you can come to the microphone, and we will use our time for additional discussion.

Our first speaker for the second session is Richard Eaton from the Medical Imaging Technology Alliance.

MR. EATON: Good morning. My name is Richard Eaton. I'm the Industry Manager for the Medical Imaging and Technology Alliance, MITA. We're the medical division of the National Electrical Manufacturers Association or NEMA in Arlington, Virginia. We represent manufacturers of MR scanners and MR equipment, and we're the largest U.S. imaging equipment trade association.

What I'd like to do today is talk a little bit about the safety aspects of the 510(k) applications and how they're intimately involved with safety, talk a little bit about how safety-related standards are important, and also close on some of the activities that MITA members are involved in, in terms of safety.

Now, as I just alluded, safety is a key part of the 510(k) process, and you see a number of different types of examples of how that is so. As part of the 510(k) content, you see that the MR system safety characteristics include everything from static field strength, peak and A-weighted acoustic noise, operating mode access requirements, operating mode limits, max SAR for each transmit coil, and you can read the remainder. You see it's involved in a great many different ways.

The 510(k) also calls for the applicant to provide recommended user instructions for a MR diagnostic device, and these include a wide array of areas which are relevant to the user operation of MR, including screening of patients and shutdown due to emergencies, recommended training, level of supervision, emergency procedures, excessive noise, liquid cryogens, controlled access areas. So there's a great many areas that are involved in the user instruction which is all part of the 510(k) process.

Instructions on safe operation are also provided by a manufacturer in the 510(k), and these include everything from potential energy conditions and means provided for normal emergency shutdown, emergency procedures, controlling for excessive noise, and the safe handling and potential hazards of liquid cryogens.

Now, in addition to the 510(k) process itself, in terms of its components, conformity to international MR safety standards is also very important, and one of the key standards in that area is IEC 60601-2-33, which are the particular requirements for the safety of magnetic resonance equipment for medical diagnosis. Its main components include safety information provided in the user manuals, EMF and acoustic exposure limits for patients, and measurement procedures to demonstrate compliance to EMF and acoustic exposure limits.

Safety information is also part of the MR safety standard, and it discusses in that standard, potential hazards caused by static, radiant RF fields, acoustic noise, cryogenic liquids and gases, and measures to control these hazards including, as mentioned before, controlled access area, prescreening and supervision of the patient, and it touches on medical emergencies, the concept of operating mode, compatibility of MR with other medical devices, and technical emergency procedures.

Now, MITA, as part of NEMA, has been engaged in developing MR and other imaging equipment safety and performance measurement standards for many, many years. Our MR technical committee writes and maintains safety-related standards, and just to give you a couple of examples, safety-related standards including measurement of acoustic noise and SAR, which are represented respectively by NEMA Standard MS 4 and NEMA Standard MS 8, and these standards are incorporated into the IEC 60601-2-33 standard for MR safety.

Now, we are involved in a number of activities which relate to safety, one of which involves international standards for implantable devices, and as an example of that, the Joint Working Group, ISO/TC 150, Implants for Surgery, and MT 40, Magnetic Resonance Equipment For Medical Diagnosis, which is working on publication of ISO/TS 10974, which covers the requirements for the safety of MR imaging for patients with an active implantable medical device.

There's a number of MR activities. Here's just a few examples that manufacturers of MR are actively engaged in that deal with safety. We mentioned IEC MT 40. There's a Joint Working Group which is writing technical specs on implants and MR. There are a number of committees that interact with the American Society of Testing of Materials, labeling standard, making that an international standard. That's being handled by IEC Working Group 45. We also work with the ASTM Committee, writing standards to determine safety related quantities like force, torque, artifact, and heating, and the European Society of Magnetic Resonance in Medicine Biology is engaged in assessing new safety issues including NSF and introduction of a MR safety officer in facilities.

So, again, I want to thank FDA for inviting us here today to discuss MR safety, and MITA stands ready and willing to work with all stakeholders to ensure MR safety. Thank you.

(Applause.)

DR. WOODS: Our next speaker is Michael Steckner from Toshiba Medical Research Institute. Welcome, Michael.

DR. STECKNER: I'd like to thank the FDA for this opportunity to speak here today.

This is a presentation discussing the use of the MAUDE database. Key points that I'd like to make in this presentation are MRI has a low rate of incidence per million procedures performed. Last year's has already been mentioned. IMV estimates there are over 30 million procedures performed in the United States.

Per the FDA, MAUDE data is not intended to evaluate rates of adverse events. Thus, inferences regarding MRI incident rates based on MAUDE are potentially unreliable.

Education and training are an excellent way to build the safety culture to ensure MRI safety.

For those of you who use the MAUDE database, I direct your attention to a comment on the website there. MAUDE data is not intended to be used either to evaluate rates of adverse events or compare adverse event occurrence rates across devices.

I know personally that using the MAUDE data can be a little bit challenging, and I'll just list some of the possible things that you should be considering. First of all, there are duplicate and triplicate entries. There are follow-up reports being filed on certain incidents, and if you do any rate analysis, that can skew your results.

As already mentioned today, there are blank narratives. So you really don't know what's going on. As well, I have found duplicate MDR numbers, but perhaps the FDA purges those out periodically; I don't know.

As well, there's an alternative summary reporting system by which MR vendors can report incidents, and it's not consistently updated into the MAUDE database. So there could be holes, of incidents that have been reported but are not found in MAUDE.

So if you're using the MAUDE data, caution is required for reasons already specified at the FDA website.

In addition, rate analysis could be unreliable if you do not consider the growth rate in procedures as well. There's 20 years of information in there, and it's imperative to really analyze all the data, not pick selected years out of the database.

As well, if you haven't removed the duplicate entries, there's potential for incorrect rate analysis.

As well, if your analysis doesn't include the ASR data, you may be potentially missing some information.

There's various possible reasons for which there's increasing incident reports. First off, the installed base is growing. In the United States, it's estimated there's around 10,000 scanners, and that number's probably growing slowly at this point in time. However, if you look at the worldwide situation, it's different. There's about 30,000 scanners, and it's growing fairly quickly in some countries. So clearly there's an increased number of MRI procedures ongoing.

There's also increasing awareness amongst user facilities to report incidents. As well, and it's not very well known, there are foreign incident reports in the MAUDE database. If there's a device in a country that's also being sold in the United States, if there's an incident there, and vendors are aware of the incident, they must report and do report it to the MAUDE database. So there will be an increasing number of foreign reports as the international install base increases.

Here's a graph of the procedure growth in the United States only based on IMV data from 1995 onwards, and since then, it's been estimated that there have been 328 million MRI procedures performed, a significant number.

So let's do a first assessment based only on MAUDE data of the attractive or projectile type forces. Roughly about 7 percent of the MAUDE database seems to fall into that category. So that works out to a bulk rate of about 1 incident per 4 million procedures performed. So one does a linear analysis, slope analysis of that, one winds up with a constant rate of incidence since the late 1990s when these incidents first started being reported, and what I mean constant, the slope of the line is .003 incidents per 1 million procedures per year. So that's flat.

So how do we ensure MRI safety? It's vital to improve education and training. The people who use the scanners, do they completely understand the warnings? Do they understand the contraindications for use? Are they following the labeling instructions? Are they reading the manuals? There are zones around the magnets. Are those zones being enforced? Is the training for people going into these zones appropriate and adequate? Are they following best operational procedures and practices? There's a wealth of information out there which is excellent. I hope it's being well used.

Are patients being routinely monitored during scans? Is there a distinct effort to build an appropriate safety culture? This last point is particularly critical because if you have a safety culture and adequate training, people will start recognizing the series of events leading up to an incident or an accident, and hopefully we can break that chain before something actually happens.

There's actually scientific basis behind wanting to do that. If you reach back into the 1930s accident investigation literature, there's this concept of this safety pyramid introduced by Heinrich, and it was substantially enlarged in the sixties by Frank Bird by analyzing industrial accidents. Analysis showed that there's a lot more incidents than there are accidents.

Furthermore, there seems to be coupled rates of occurrence for a certain number of incidents. There's a smaller and expected ratio of accidents. It varies across industries, but that concept seems to be fairly universal.

So the safety literature concludes that if you educate, if you do safety awareness, not on specific types of incidents, but broad range, make people aware of what can happen and why, then they can start to capture things before they happen. You then reduce all types of incidents. If you reduce all types of incidents, the accident rate will also come down because of that coupled ratio. So that's why education is so critical in building a safety culture.

So, in conclusion, MRI has a low incident rate per number of procedures performed. The number of MRI procedures is growing. Per FDA recommendations, caution about using a MAUDE database for rate analysis. There's also a growing awareness of the importance to report incidents. Lastly, an improved safety culture through education and training is the best way to reduce incidents and accidents. Thank you very much for your time.

(Applause.)

DR. WOODS: Thank you. Our next speaker is Shumin Wang from Auburn University. Are you here, Shumin Wang?

Then we'll move onto Charlie McIntyre from the Chickasaw Nation Medical Center.

MR. McINTYRE: Great workshop so far. We've already learned a lot.

My name is Charlie McIntyre. I'm the MR manager at the Chickasaw Nation Medical Center in Ada, Oklahoma. We service a lot of tribes in the state of Oklahoma, and we recently, well, I shouldn't say recently, about a year and a couple of months, purchased a new magnet which is a big feat for the Native people. It was a high field magnet. It's a 1.0. We wanted good quality. It was also an open magnet so we can accommodate large folks like me and any claustrophobia patients that we may encounter.

Now, as we were going through out our applications, on our second visit on our application visit from the vendor, the vendor made a passing remark that we should be aware of what's called a spatial gradient strength, and they quickly showed the diagram of it. I didn't quite know what to make of it, but the more investigating I did, the more I realized I didn't know, and that may play a part in having yearly training as has been recommended, and that was my fault for not understanding the new terminology, conditional versus compatible. I'm sure we'll talk about that tomorrow.

I found out I wasn't alone, however. Our radiologist that reads didn't understand, and I think that the vendor that sold us the applications really didn't have a good grasp on it.

What it is, is that our vertical high field introduces extra forces that aren't assessable on a traditional horizontal field. So little testing has been done I've found out. In fact, maybe very little.

These implants affect stents, heart valves, shunts, mediports. Now, keep in mind, our radiologists are not on site with this. They're offsite, and so I'm the only safety officer acting in behalf of them. I feel like I'm alone on an island. So I can't buy a surgeon saying it's okay, because when I question them, even they don't know the extra forces that are placed on this vertical high field magnet.

In conclusion, I would please encourage the vendors and MR experts to have universal testing preferably with the same language, specifically addressing the gradient, spatial gradient, as in my case. I know it's a small MR aspect in the whole world, but it does affect a lot of people, or at least have the vendors be upfront and show them that this limitation is there until further testing can come about.

I e-mailed various device implant makers, and the local reps didn't understand what I was talking about, and they went further up in research, in R&D, and they knew. Either they knew about it and they didn't know when they were going to test for these higher gradient fields or these higher forces in general, or they were going to do it and then there was some disagreements on how the test was going to be performed by the MR experts. So I was given the okay one week to do a test, and then the next week they yanked it because again there was some disagreements again on how the testing was done.

So I encourage universal testing, and as a technologist, we would appreciate the clear, concise direction, and I appreciate your time.

(Applause.)

DR. WOODS: Thank you very much. Our next speaker is Joseph Barwick from Metrasens.

MR. BARWICK: Good morning. We also appreciate the time to present to the FDA as part of this workshop. My name is Joseph Barwick. I'm with Metrasens. We're manufacturers of ferromagnetic detectors called Ferroguard.

I wanted to share our experience in being in a position to be able to provide and implement this technology to over several hundred successful deployments, and our approach is to carefully consider existing practices and procedures for this technology.

The technology has you have heard and will see in future presentations is really a tool that needs to be utilized appropriately in order to allow for the proper effectiveness of increased safety.

We've optimized the technology to account for real world environments and applications as well, and that's part of the interaction with the users of the technology and feedback that we've gotten along the way.

There are some important considerations and conditions that go into the proper implementation of the technology in general. I think to start with, you need a ferromagnetic detection system that has been engineered, that has a robust technology offered and capable of consistent ferrous detection. The most important part is to a specified, identified standard of performance. The alerts that are provided by this technology need to be communicated in a useful manner. We heard stories of people saying the alarms go off, false positives, see such things. It's important and critical that the technology produces alerts in a useful manner, and at a time in which corrective actions can be performed in advance of any opportunity to present an object into the MRI suite.

The deployment of technology needs to be flexible and discrete as to not create additional hazards. If technology has been deployed inappropriately or in a position that creates additional hazards too late in the process or create adverse alarms, it could hinder the existing safety procedure. So great care needs to go into the application of this technology and the implementation as well.

Probably the most important part that was touched on earlier by several of the other presenters here is the conscientious interaction that needs to displayed or utilized by staff. The technology again is just a tool, and it requires staff use it and respond to it appropriately, and that comes into play with training. It comes into play with deployment. It's comes as part of the entire process of integrating the technology.

So, in summary, FMDS is a tool. It's not a magic bullet. It's not a replacement for existing safety procedures or practices, and it should be integrated into existing, proven screening procedures such as the ACR Safety Guidelines as well as some of the other current procedures and practices.

We performed an in-house study at Kaiser Permanente in which we screened patients post-traditional screening methods or practices. We set up our technology in a way that allowed for advanced screening through traditional practices, and we came up with the results of about 28 percent of the patients who were screened post-traditional screening practices, still possessing various materials that were detected by the technology. So, again, further evidence that although there are great existing screening practices and procedures in place, it is a subjective methodology and it does require the individual user to make sure that they're doing everything as appropriate as possible.

I have a couple of user comments that we received back from several of our users when asked specifically about providing commentary for this forum.

We have a user, Craig Satori at Bellin Health in Green Bay, and just to quote some of this response here, he's indicated that, "After double-screening and changing patients into hospital garments, they still find new ways to try to get foreign materials into the MRI environment, but their magnetic detection system finds the cell phone, the ankle bracelets, and remaining hairpin among other things. For others walking into the MRI area, it gives them reason to pause and ask if they can proceed into Zone III, and that very pause is a win for MRI safety."

Another one of our users, Christopher Vineyard at the University Health System in San Antonio, said, "Even though we have plenty of warning signs posted near the MRI suite, adding an active visual and audio ferrous warning system gives us tremendous peace of mind."

And, lastly, Kara Mallon from Seattle Children's indicates that, "Ferromagnetic detection plays an essential role in keep our patients, families, and staff safe at Seattle's Children's."

Essentially ferromagnetic detectors do detect items that have been missed by current screening procedures. The benefits are obviously the reduction of potential projectile effect incidences, artifacts, and workflow issues. The presence of ferromagnetic detection also increases staff confidence in overall safety procedures and implementation and also increases patient confidence and comfort before and during the MRI scan.

In our experience again, the technology is effective when it performs to exacting standards governing sensitivity, consistency, and reliability of detection, when the technology is deployed in a manner that enhances existing, proven safety methods, again not as a replacement, not as a substitute, but rather an enhancement or additional tool; also, when users are properly trained and utilize the technology within the proper context and modify existing behavior or practices to enhance the efficacy of the technology.

Thank you.

(Applause.)

DR. WOODS: Thank you very much. Let's see. Simon, do you have any suggestions about this other projector?

Okay. And in the words of the title of the next presentation, I guess, no one's perfect but we're giving it our best shot. Why don't we take the chance and go onto our next speaker, Kemp Massengill from Mednovus SAFESCAN.

DR. CHOI: (Off microphone.)

DR. WOODS: I think that will work.

UNIDENTIFIED SPEAKER: (Off microphone.)

DR. WOODS: Yes, Simon. Can you go to Toby Gilk's presentation? So Toby Gilk will be our next speaker.

MR. GILK: Good morning, ladies and gentlemen. My name is Tobias Gilk. I work with and for Mednovus, a manufacturer of ferromagnetic detection systems, and with Radiology Planning, an architectural design and consultancy firm.

My remarks today are really not going to be related to specifically either one of those. I'm going to speak more broadly to MRI safety. In addition to my role with those companies, on the screen are a bunch of my participations over the years in various MR safety efforts and initiatives.

Many of you may have seen versions of this slide or this graphic previously. I've been publishing iterations of this for several years. This is the reports year over year into the FDA's MAUDE database, and Mr. Steckner's comments about the failings of the MAUDE database are absolutely correct, but I do feel as though, number one, since it's the only available resource out there that helps us quantify MR accidents, it would be foolish of us to ignore its utility and, number two, whatever errors there are in the system, I think at least we can look at it year over year and presumably, you know, whatever was wrong with the system in 2005 was probably equally wrong with it in 2007. So we can take a look at it year over year.

I'd like to present this data in a slightly different format today. I'd like to take a look at MR safety, the rates of reported accidents, and using the year 2000 as a baseline. If we look at the years from 2000 to 2009, MAUDE records a 523 percent increase in MR accidents. If we look over the same period of time, data from IMV in terms of change in total procedure volume, over the same timeframe, there's been a 90 percent increase, it's actually slightly less than 90 percent in total MR procedure volume. So the gap between the red curve and the blue curve represents the degree to which we have failed to manage safely with the growth of MR.

Now, you notice in this illustration, there's a noticeable kink that occurred very shortly after the 2001 Colombini fatality and the 2002 release of the initial ACR White Paper on MR Safety. It is my opinion that following an accident that grabbed the industry's attention and corrective measures that were released the following year, we had a brief period in the recent history of MR safety in which the industry was fine and, in fact, we bent that accident curve below the growth curve for the one and only spot on this graph in the year 2004.

What happened after the year 2004? Well, it wasn't for a lack of additional guidance. We had two additional releases of the ACR White Paper and then renamed the ACT Guidance document, and also released Joint Commission Sentinel Event Alert 38. None of these additional pieces of recommendation from the industry have really done anything to bend that curve in the direction we would like to see it.

Why is it that the best practice recommendations haven't responded to MR safety? The key word is recommendations. There is not a single regulatory, accreditation, or state licensure body that has physical safety requirements at the point of care for MR safety that would prevent the 2001 Colombini accident from occurring. Not one state licensure, not one accrediting body, nobody has physical safety requirements at the point of care that would prevent that accident for the thousands of others that occur.

So what do we actually need to provide appropriate protections for MR safety? And first let me state, that the question of MR safety is not a question of the safety of the device. Virtually all of the accidents that occur are point of care related accidents and not directly attributable to the MR device.

What do we need to protect these patients? We need objective, measurable MR safety standards. We need two faults, pass/fail, yes/no, do you have, do you not, measures to assure some fundamental foundational level of MR safety at the point of care.

This slide, Dr. Emanuel Kanal and I did an evaluation of two years of each and every MR adverse event in the MAUDE database from the years 2009 and 2010. For those that we could identify causation, which was slightly less than half of those in the cohort, for those that we were able to identify causation, we tested them against explicit true/false, yes/no, pass/fail kinds of criteria in both the Joint Commission Sentinel Event Alert 38 and the ACR Guidance Document.

The Sentinel Event Alert mitigated about 65 percent of all of those accidents. The ACR Guidance Document mitigated about 85 percent of all of those accidents in the two years of MAUDE data. Eighty-five percent mitigation from this one document.

So why do we not mitigate 85 percent of these accidents? Because nobody has actually required, not even the ACR requires that its own accredited facilities to utilize the Guidance Document and the safety precepts therein to protect patients.

So my request to the FDA, to the professional community, is to adopt objective, measurable MR safety standards, and if you don't know where to start, if you don't have a jumping off place, I would very strongly recommend that you use the ACR Guidance Document and the precepts therein as the starting point for point of care MR safety requirements.

And with that, thank you very much.

(Applause.)

DR. WOODS: Is Kemp Massengill here?

MR. MASSENGILL: All right. First of all, thank you very much. My name is Kemp Massengill. I'm with Mednovus, and we make ferromagnetic detection products.

This is actually a major problem. It's underreported. Many near misses are swept under the rug. In fact, some accidents are also swept under the rug. Underreporting, and that's documented. Toby Gilk has VA, I guests from the VA and also elsewhere.

One of the problems that we're dealing with, in fact, the problem is human nature. Certainly from the patients' point of view, they see this big instrument, they're going to be going into this tube, they're usually frightened and patients in general have a tendency to kind of forget something. For instance, I had a gallbladder operation about 30 years ago, and somebody asked me if I had any surgery, and I said, none at all. They said, well, what's that scar right there? Oh, that. Well, that's the gallbladder.

Dealing with the general public is always difficult. Some people are a little bit hostile when they're having medical tests. Others want the test but in general people are nervous. So you have two different situations. You have human nature on the side of the patient and you have human nature on the side on the operator. These are trained professionals. Nevertheless, these accidents happen and they happen fast.

I've got a video, I don't know if the video will play, it shows you how fast it is, but if the video doesn't play, if you fire a gun, there's a de-acceleration. On the other hand, a rocket that's going faster and faster. So this is far more similar to a rocket than it would be to a gun.

Okay. Should we trust human nature? That's the big question. These are various accidents. The lights are on quite a little bit. Is there any way to dim those slightly or just keep on going?

DR. WOODS: Just keep going.

MR. MASSENGILL: Okay. I'll just keep on going.

There were slides that were very easy to come by actually. It didn't take long at all, keeping in mind that whatever went into the bore of the magnet went there fast, fast, fast, and you just see various objects.

Not a single one of these centers had ferromagnetic detection. Obviously if there were a patient in the bore at that time, it would be absolute catastrophe.

So the question is, should we trust human nature? No ferromagnetic detection. Would you want your 16-year-old daughter to have a MRI at this center and have this happen or your 4-year-old granddaughter? It's just absolutely amazing, and many of these people are in the field, trained professionals for years and years. Now, I should have a little audiovisual effect with a huge bam, every time like bam.

There's a lot of MRIs done, and certainly the vast majority of them, probably 99 percent, are done absolutely safely, but there's the other 1 percent.

It's very hard to pull things off the magnet, and do not do this, but if you take a magnet in one hand and one in the other hand, and you get them and you get them and you move them closer and closer, all of a sudden, bam. Nothing happens and then.

(Playing video) File cabinet.

So the question is, do we trust luck? And in general, luck is really dumb. Dumb luck or do we do something else? As a vendor, I won't talk about products per se, but all the slides that I've shown did not have ferromagnetic detection in place.

What is cost? You know, I'm speaking just generally. Ferromagnetic detection, you can do a good job for $3,000. In a $3 million suite, that's 1/1000th, 1/1000th. People are saying, well, gee, it's so expensive. It's so expensive. It's so expensive. 1/1000th.

Is increased patient safety worth it? Nothing is infallible, and ferromagnetic detection is certainly not infallible. It's the second line of defense. It's not a substitute for conscientious, caring MR technologists. It's a second line of defense, but is it worth it? That's up to you.

Thank you very much for caring, and it is certainly a real privilege being here. Thank you very much.

(Applause.)

DR. WOODS: Our last speaker this morning is Ashok Saraswat from Ohio State University.

MR. SARASWAT: I'd like to thank FDA for inviting me to talk about MR, and we're all experts on MR. So I'm just going to save some of the discussion on that.

As we know that we all have our best interest, FDA, I have done a lot of research on FDA's website, what they have done, and I want to thank them for the work they have done in the MRI field for the patient safety in general.

And the recent example I saw was how quickly FDA approved that Revo or Revo Medtronic pacemaker. So again, you know, I have been in the MR business since 1985. I'm an educator. I'm a researcher. I'm a scientist. So I'm just going to stick with the facts. I'm not here to teach you about MR safety because you all are experts on that.

One thing I noticed is that all MR accidents, most of them are preventable. You know it, and I may be preaching to the choir, but they are preventable.

The objective is that how do we reduce these accidents and have the best outcome for the patients. So, again, a lot of speakers have very eloquently talked about the need of training, and I fully, in my opinion, I fully support that. The training, knowledge, and also one of the slides said nobody's perfect, I agree with that, but also the practice makes you perfect. The more you do, the better you get at that.

Who can help? Who can guide the questions? Regulations, institutional, governmental, you know, who needs to regulate what? Some are for it. Some are against it. So I'm not getting into that discussion either.

But I do believe that, as they said, that all accidents are preventable. So just like any strategy, I believe in, you know, act, plan, do, check, some of you know that.

Who needs the training? MRI worker, technologists, and there are a lot of known technologists in this field. Yes, they all need recurrent training. Physicians, I come from a time where the physician sat next to me, and they approved each and every scan, MRI scan, and they looked at the screening form. Everybody's getting busy. So, yeah, the training for physicians is also necessary. I'm not sure who's going to do that, but out of let's say 50 or 60 radiologists at -- University Medical Center, maybe 10 are good in MRI safety practices. They feel comfortable with that. So, again, cardiologists, anesthesiologists, and other people, they all need MR training, internal customers, external customers. Training is the key definitely.

The future trends, we look at the new innovations. MRI does have positive impact but, yes, there are some negative outcomes as well, accidents, but I still think that MRI is a relatively much safer modality. When you look at the total number of accidents, some people say accidents are on the rise. Yes, and I also believe that some are duplicate entries, but on the same token, there are a lot of accidents that are not even reported. So the reality is for all of us to think about, you know, where is the truth?

So who is responsible for MR safety? Everybody who is as a team.

Is FDA taking active role by providing essential guidance and standards? Yes, I have researched FDA's websites and spend a lot of time. Yes, they have.

Is there more they can do? I'm glad they are willing. They're listening to us. So I'm sure that they're going to do what they, you know, have invited to listen to our opinion. So FDA is taking active role and, yes, in my opinion, there's more that can be done.

Do we need more regulations? Yes, no. I am in favor of, yeah, we need some kind of regulations, but do they need to be very strict? No. They can be internal, like radiology community can have their own regulations. Then some state guidelines, federal guidelines, they need to be there as well. So a mixed approach, a team approach will work in my opinion.

I have couple MR system manufacturers that, you know, maximum spatial gradient, we need that map. I'm not sure, you know, I'm not accusing anybody. Why are we not getting the accurate maps? Yes, we're getting maps, but with the shielding and all that, the MSG, the points are not as straightforward. So we need better disclosure and some transparency on that as well.

So CDRH has done a good job on medical device safety, and I'm hoping that more work needs to be done, and it will be done.

Implant manufacturers, by the same token, we need some accurate information, detailed information. I realize they're not required to, but from business point of view, it's the best practice that if they disclose it fully and improve testing. I've looked at the testing procedures. Some of them are complicated. So make it simple. Make it standard. I need, you know, I need to hear a little bit more on that.

I'm sure I'm going to learn in a couple days, both today and tomorrow, a lot of resources out there. I know a lot of people are sitting here, FDA, CDRH, NEMA, MR Safety, Dr. Frank Shellock is here, and detectors, you know, ferromagnetic detectors. All these people are here.

So again, in the conclusion, I believe that, you know, this is a great start, and we have work to do and, yes, I am proposing that we look at ACR guidelines. I know there is a revised version that is about to review, don't have much information on that, but ACR is working. ASRT, SMRT, they're doing a great job. I'm a former policy board member of SMRT. So we support, we discuss these things on day-to-day basis, and again, you know, I want to thank FDA and the ferromagnetic detectors, I look at them as a tool. It is a tool. It is not to replace the normal screening process, but again, it's a tool, and that is just my opinion. So thank you so much.

(Applause.)

DR. WOODS: Thank you, and thanks to all of our speakers. Has Shumin Wang from Auburn University arrived?

Okay. Then again we have some time before our lunch break. Is there anyone in the audience who has an additional comment on the things we've been talking about this morning, general aspects of MRI safety, ferromagnetic detectors, the impact of innovation in MRI technology on safety, on the existing safety information that we provide?

Could you come to the microphone please? Thank you. You have to come to the microphone please.

UNIDENTIFIED SPEAKER: (Microphone not working.)

DR. WOODS: Thank you. Also, if you could identify yourself for the transcription please.

UNIDENTIFIED SPEAKER: (Microphone not working.)

DR. WOODS: Brad, does that microphone in the center work? Did I hear something?

UNIDENTIFIED SPEAKER: (Off microphone.)

DR. WOODS: Okay. Anyone else? Go ahead. Please feel free to come to microphone in the center.

UNIDENTIFIED SPEAKER: I'm from University of Minnesota, Center for Magnetic Resonance Research. We are a high field system, and this comment is regarding the earlier comment on firefighters. Our Center and our University recognizes that in this kind of incident, you know, firefighters will come, and in our Center we have annual trainings with firefighters where they are made aware of where they can and cannot go, and now they can rescue someone if they need to be rescued from the system.

DR. WOODS: Thank you.

DR. CRUES: John Crues from Radnet. I'd just like to caution people a little bit in using the MAUDE data to suggest that there's a marked increase of incidents, of events. I think there's an increased reporting of events, but I think in large part, that's because people are aware of it and they're reporting more.

In the data that we have at Radnet, which is on the average of about 500,000 studies a year over the last 15 years, we have not really seen much of a change in our organization. I just showed the last 18 months, but we have data going back 15 years, and it's been very stable.

MS. ROTH: Candi Roth again, Imaging Ed, former SMRT President, former educator, University of Penn.

Two comments. One on the firefighter comment, just in response, we did train our firefighters. So those firefighters did have MR safety training, and that was all well and good in the classroom, but when they got to the site and smoke was in the room, all bets were off. All the training didn't apply. That's just my experience from that one site.

And, Chris, to touch on your comment about standardized training, I agree that training needs to standardized, but I don't think that cookie cutter training is going to work in all sites. So maybe what you would need, for example, for a pediatric site, there may be a little diversion or diverted training that might work better in an outpatient site or university site. So maybe there's something like a CPR course for lack of a better example to start with, with diversions or with alterations that fit. So maybe we start with one cookie cutter and then go from there. It's just a suggestion.

DR. KANAL: Manny Kanal, University of Pittsburgh Medical Center. I'm not sure what belongs at this point in time versus the roundtable discussion, but since so many people brought it up, a few points to bring up that are perhaps more objective. The number of legal cases being brought to my attention that are involving MR safety has continued to increase on a steady basis in the past 25 years, way beyond the 90 percent or so increase that we expect in the last 5 years and based on utilization, but far more importantly, so many people are talking about, and I know John has mentioned several times, the number of cases he's had are steady over the years, and out of a few hundred thousand, there were 0 ferromagnetic incidents.

This goes to underscore that the last person to ask is the leader of the site. The people to ask are the technologists. They run the site. They live and breathe this stuff, and they're the only ones that really know, and if I'm a technologist and I have an accident, or I shouldn't say an accident, quite the opposite, an incident where somebody has their toenail clippers in their pocket, and it flies in and nobody gets hurt, the very last thing I'm encouraged to do with my system is to report to my boss that I didn't do my job well. So zero incidents reveals how little we know about what's going on in our site for reasons of our own structure and our own construct.

And, finally, almost 100 percent of the serious injuries or the injuries that were serious enough to go to court are not on the MAUDE database, including my own institution, University of Pittsburgh Medical Center. Many institutions have a gag policy that as soon as something serious occurs, by my contract, I am precluded from reporting it to the FDA or to anyone else or speaking about it in any public setting.

So to assume that the MAUDE database is representative is also an error from the point of view that even the serious injuries are not necessarily captured.

(Applause.)

DR. WOODS: Thank you.

LEON: Hi, my name is Leon. I'm from Georgetown University Hospital. One of the things I'll be expecting, that will hopefully come out of this workshop, is some discussion about stroke patients. MRI is being used more frequently now for stroke patients. So maybe there will be some discussion about the safety of stroke patients because of the time that we have to scan these patients.

DR. WOODS: Thank you.

MR. LAUNDERS: My name is Jason Launders, and I'm from ECRI Institute. Like a few other people in this room, I've had the pleasure of reading through lots of MAUDE reports trying to understand why accidents happened and taking into account all the shortcomings of the reporting process.

One of the things that I've noticed in a lot of those reports is that they all say that the technologist was trained, but the reason the accident happened was they were distracted. So we count training is absolutely critical and we need to have training standards, but we also need to understand that training isn't everything, just as much as the ferromagnetic detectors aren't everything.

Distractions happen. That's why most car accidents happen. That's why most accidents happen, and we have to understand that in how we go forward and make this safer. Thank you.

(Applause.)

DR. WOODS: Thank you.

MR. GILK: Tobias Gilk with Mednovus and Radiology Planning again.

I just wanted to add one thing that's sort of on the principal agenda for this first day's talks, that hasn't been addressed in the first two rounds, and perhaps it will be in the next set of speakers or during the panel discussion. One of the questions was about the future of MR safety and changing MR technology, and there are a couple of things that I think are important to underscore the importance of this meeting and the initiative, and that is that if we take a look at the traditional, historic model, when a MR patient is, you know, walkie-talkie, knees or head, is sort of what MR practice was built upon, and as was just mentioned, you know, as we're dealing with emergent and trauma imaging, stroke patients, cord compressions, as we're looking at MR guided interventional procedures, as we are looking at higher and higher magnetic field strengths, for static magnetic fields, faster and stronger gradients, if we look at just the U.S. demographics in the absence of changing technology or changing clinical utilization, we are entering the bolus of the baby boomers crossing the 65-year threshold, and the rate of utilization for MR, once an individual is 65 or older, the utilization rate is 2.4 times the rate of everyone 64 and younger. So even if we changed nothing in terms of the technology for MR which we know is moving at light speed, if we changed nothing in terms of the clinical utilization, we are going to have a greater proportion, greater numbers of patients in the 65 and older who are going to come to MR with greater comorbidities, a much greater likelihood of implants and devices.

So as we sit here today and gaze into our crystal ball, in terms of what we expect, technology, clinical practice, and utilization for MR, no one of those general aspects gives us hope that next year, 3 years, 5 years, 10 years from now, the MR safety condition is going to be better.

In fact, each and every one of those suggests exactly the opposite, that the risk factors associated with a patient population, what we're doing to them and the magnets we're sticking them in, the risk factors for all three of those are going upwards, and when we have a population base, an annual exam population base of 30 plus million, small changes in risk factors are going to result in significant numbers of additional near misses, incidents, injuries, and potentially fatalities. Thank you.

DR. WOODS: Thank you.

MR. AMURAO: Max Amurao from Georgetown University Hospital. I know this is the choir, and we've heard over and over again that, yes, MR technologists are at the frontline, and as mentioned earlier, people get distracted. So I would encourage people in the choir to expand the training opportunities for a larger community because it's not just the technologists and the physicians and probably the MR scientists, but there's a whole group of other people that interact with the MR environment as you well know. We have transporters. We have housekeepers. It's almost like taking the analogy of a flu shot. You have a more effective inoculation or a more effective program if you have a broader base of people getting trained. So perhaps maybe we could consider enlarging our scope of training, not just to the technologists, but also to the ancillary staff that we work with.

DR. WOODS: Thank you.

TERRI: Hi, my name is Terri. I'm a nurse from the Johns Hopkins Hospital. I was surprised to hear throughout all of the presentations that no one really talked about the new technology or somewhat new technology of intraoperative MRI. This concept itself and this treatment provides many challenges to work through.

We are currently building a facility that will have an intraoperative MRI, and just the process of going through the safety and the regulations and the education is truly inspiring, but I'm just very surprised that we haven't heard anything here.

So perhaps as we move along with MRI safety, this is really a new technology that we really do need to concentrate on because again it does provide its own challenges, and as the gentleman before me said, we not only just look at the MRI technologists now or the RNs in the room or the physicians, but again it's a challenge for all the staff because it is an open area, and even though the magnet itself sits in a bay, the opportunity for anyone to go in that room without a proper screening can happen. So the safety component is really essential.

DR. WOODS: Thank you.

MR. STANLEY: Hi, Charles Stanley, Stanford University. Again, I'm not sure how much of this should be for the roundtable, but since we're having comments about it, I wanted to make a comment or two about the training and the distractions. I appreciate the comments of others, but again to me it underscores exactly what we're talking about this morning. We've heard this recurring thing, that there's not standardized training amongst different sites. We have one site that has a best practice of 0 percent. Unfortunately, that's not representative of the entire MR community. And I do agree with the gentleman who was talking about these people are trained, but there's distractions.

But the point that I made at the podium is that other industries have figured out how to do this. The guy who landed the plane on the Hudson River, he didn't do that every day. What he did was he trained for the fact that that may actually happen. That's what we need in our training. We need to train for those distractions, because the reality is, yeah, there are people at anesthesia who are constantly trying to sneak in your room. There are housekeepers. There are firefighters. So the point is you incorporate that in your training. You certify people on it. You ask them to be retest on it every year. You don't get to continue to work if you don't go through it again. I mean to me it seems pretty simple. We need to incorporate those things in our training instead of putting devices in that again there's nothing wrong with the device, but the point is, is if you're distracted enough for the device to go off, you're also distracted enough that it's too late. I mean if you hear the beep, it's too late. It's already done. So the point is you need to incorporate those distractions into the training.

DR. WOODS: Thank you. Let's go to the microphone on the left, on my left.

MR. STAGER: All three of these mics should be working now, too. So --

I'm Brad Stager (ph.) with Biomet, of course, an orthopedic manufacturer. A follow-up comment, I think a perfect follow-on here, is we have procedures in place that require us as an employee to immediately report any events that occur or anytime we hear of any event with regards to our products. This is a procedure that's in place. We're retrained on it every year. It's pretty simple. I don't know for sure that we're required to do this training, but it's a great practice to have in place. It could be very simple for anyone to implement as long as management is behind it. In our case, we're able to look up those procedures anytime we need to on a central database of our procedures. If we don't know or remember who we're supposed to contact, we can easily go up and look and find out where we're supposed to let people know, the appropriate level of management, or we can go directly to our management who elevates the complaint. So the idea is for the technologists and the hospital administrators, or the center administrators, it would be an easy fix to just put procedures in place that say you can either go to this website, go to your management, there are no repercussions in saying the patient didn't tell me they had those nail clippers. I guess that's -- and just the point is, it's an easy fix. It doesn't necessarily have to be mandated, but there could be written procedures that could be a simple fix as far as that goes.

DR. WOODS: Thank you. All right. To the center.

DR. FRICKE : Well, I'd just like to say a few words about the interventional suite. It's true. It's not just interventional suites today. So much equipment's coming into the MR room that any guidelines would tell you it's probably not a good idea. And then we say we have to have a procedure.

One of the things that's coming out especially with the higher fields though is that if you are in charge of testing of that equipment, and you test it appropriately, as I did with a given ventilator, to show that it should be safe in the environment, conditional but safe, in other words, it won't kill somebody by itself, and then later on you find out months later that it's become magnetized, who knows why? Maybe the batteries were replaced or something. Part of this is metal detection.

The ACR guidelines do not spend a lot of time discussing equipment that you have to bring into the room and how it should be treated, tested, annual maintenance, preventative maintenance, et cetera. I think this is reality. It's the future. I think we need to pay attention not to just screening the patient. That's a big, big issue, and we've seen many examples today of what that means, but the equipment that goes in there, we need to think about that better. It's not just processing. It's also equipment tube tested, from a day-to-day, week-to-week, month-to-month basis, to make sure that that equipment is still MR conditional and safe to operate. If it's a hazard, we have to put a hazard label on there, and then really think about the process.

DR. WOODS: Thank you.

UNIDENTIFIED SPEAKER: Thank you. I know we're all under a tremendous amount of pressure especially from administrators and I think even more outpatient centers to continue to stay profitable and stay afloat. I think that the urge to be multifaceted may be hurting us a little bit. Do ECT. Do x-ray. Do MRI. And maybe that's part of the distraction and what may potentially happen. Maybe we need to stay in MR because it's a whole field in itself and maybe not worry about multi-training. That's just a suggestion. Maybe we don't have a transporter because the hospital can't afford it. Again, that's another -- we don't have a secretary to help us. Maybe that's another distraction. So if we could all put that in our papers, that that may be a factor in us not paying 100 percent attention, which we should be to the patient at hand.

DR. WOODS: Thank you.

MS. SAWYER: Anne Sawyer from Stanford University. I just want to make a brief comment, and then ask a question.

I spend a great deal of my time as a technologist and as the manager of our magnets, correcting people and then trying to educate them, and this includes surgeons and clinicians and radiologists and scientists, about why the word compatible is not used anymore, and why conditional is so much more appropriate in the acceptable terminology. So I'm asking -- and this question was really asked at an ISMRM safety meeting in Portugal, I think it was two years ago. So I'm just going to repeat the question. Why do we still see the word MR compatible on slides from the FDA?

DR. WOODS: We tried to use the term MR compatibility, and you're right. Unfortunately back in the '90s when we came up with those definitions, we wrote definitions for MR safe and MR compatible and realized that we needed to change those definitions, and we're trying to encourage the use of the MR conditional. I guess we try not to use that word. I can't say that we didn't, but again, it's hard to avoid the use of the word compatible. I really don't think we use the term MR compatible anymore. We try to say, you know, MR compatibility, but you're right. It was an unfortunate choice. We're trying to get rid of that term, and we ask everybody to try to use the new terms, MR conditional, MR safe, and MR unsafe.

MS. SAWYER: I'll just make one more comment. If you study to write for newspapers, they teach you that you need to write at a sixth grade reading level. So to me, when people see compatible and compatibility, it's essentially the same word.

DR. WOODS: Thank you.

MS. DEVLIN: Good morning. This has been an interesting conversation for me. My name is Pat Devlin. I'm from the University Hospital of Cleveland. I am not a MRI technologist. I'm a radiology administrator. My background has been radiation safety from nuclear medicine training, and I became a MRI manager five years ago at which point when I started looking into our program, I was very distraught by the fact that very little had been done in terms of protecting patients and employees in very scary magnets.

When I look at my radiation safety program and background, these programs are administered by Departments of Radiation Safety, and we're expecting MRI technologists to basically manage safety for their department. This needs to be an institutional level, not managed by a medical director that's a radiologist and very busy, as well as MRI technologist. It needs to be mandatory reporting for all incidents so that you can quantify the damage that's being done to everyone with this.

This equipment is an amazing diagnostic tool, but there are distinct safety concerns with it. We have five magnets in our main hospital. We have one MR unit that we just brought live about a year ago, thankfully have had no incidents. We have a -- coming up in two months, and really my only ability to move our program forward was with the help of Tobias Gilk and Manny Kanal, who I contacted when I became a manager.

I also think that the new legislation that's coming up is going to force institutions to do dry throughput for the magnets that's going to cause more and more problems as people start seeing MR as a CT and not recognize that they need more than one technologist and more than 20 minutes to turn over every patient, and I think if you don't really start documenting what's happening, it's only going to get worse. Thank you.

DR. WOODS: Thank you.

(Applause.)

DR. VAN DEN BRINK: I'm Johan van den Brink from Philips Healthcare in Netherlands. I'm listening to what's being presented here as the emphatic calls for a safety culture as also conveyed by Michael Steckner. I wanted to make one comment here on that aspect. If you go to, from my training background to chemical industry, we know there's plans where you come in. There is a big screen saying 5,400,000 hours without an accident. Listening to what's being said here, there is not really a culture at multiple places to report incidents. This probably is the very, very first thing FDA needs to focus on to make MR more safe.

DR. WOODS: Thank you. On the left.

MS. SHOTMAN: My name is Amy Shotman (ph.). I'm with the ASRT, and I just wanted to elaborate upon a comment made earlier regarding the CARE Bill. It is HR 2104 in the House of Representatives, and the intent of the legislation is to establish federal minimum standards for individuals performing diagnostic radiation and planning radiation therapy. We think that this will improve patient safety as well as save money in the Medicare system through reducing the incidents of repeat images and also just to sort of standardize excellence within the profession.

I think it's safe to say that this Congress is slightly more dysfunctional than most. So if you individually or through your organizations can reach out to legislators, I know ASRT and the Alliance working on the House CARE Bill would appreciate it.

DR. WOODS: Thank you.

MR. KOPP: Keith Kopp with Kopp Development. Following up on Anne Sawyer's comment a little earlier, our company makes ferromagnetic detectors, and one of the biggest issues we run into is improper labeling of devices in the MR environment. This ambiguity between conditional and compatible, what it means to be safe, there are two definitions for that, the old and the new, and we find many, many times even from the people who supply the MR equipment, they'll have things labeled nonmagnetic which indeed are somewhat magnet and actually are risky.

So education is very important, but part of the education process is to say follow the labeling, and yet the labeling cannot be relied on, and I think this is a very critical issue that is within the purview of the FDA, is encouraging if not enforcing proper labeling. Thank you.

DR. WOODS: Thank you.

(Applause.)

DR. WOODS: Anyone else?

Then why don't we adjourn for lunch, come back at, let's see --

(Whereupon, at 11:20 a.m., a luncheon recess was taken.)

A F T E R N O O N S E S S I O N

(1:00 p.m.)

DR. WOODS: Welcome back, everyone. If you'll take your seats, we can get started on the next session of public comment.

Our first speaker this afternoon is Dr. Emanuel Kanal from the University of Pittsburgh Medical Center and the American College of Radiology.

DR. KANAL: Thank you very much. I very much appreciate the FDA allowing us this opportunity. This meeting was held to discuss factors that can minimize patient risk and staff risk, and along those lines, I have two recommendations from the American College of Radiology and from the University of Pittsburgh Medical Center.

The recommendations are as follows. We would like, number one, first and foremost, to encourage the FDA to standardize MR safe practices and protocols for all MR sites and to define and adopt minimal, acceptable practice patterns. And, number two, in this regard, we feel that they should adopt the American College of Radiology MR Safe Practice Guidelines as their template from which to build such recommendations.

The MR Safety Committee of the American College of Radiology was established in 2001, right after the Michael Colombini tragedy, and the Committee was charged with reviewing all MR safety incidents and finding out, using root cause analysis, what was their commonality, if any, of cause and finding ways to break the chain to see if we could stop them from occurring in the future.

The initial Committee and the ongoing Committee, now its 15-member panel, and we've found unique expertise from every member of the panel. We have radiologists in private practice, academicians, researchers. We have pediatric radiologists. American Society of Anesthesiology is represented there. We have MR technologists, MR nurses, National Electrical Manufacturers Association representation. We have representation from legal counsel. The FDA sits on that panel. In short, we tried to get as much expertise together from different walks of life to provide input as to what could make this a safer environment.

We've very actively pursued input also from over 100 outside sources, from ECRI, from other MR technologists and radiologists. We've actually formed a MR technologist safety leadership group and got them to provide a rather extensive document which served as one of the foundations for our whole document, and these are some of the people that were on that committee, and these are some of the most well-educated, most-experienced, and well-known technologists in the field.

The first publication was in 2002, and it was then the White Paper on MR safety. It was updated again in 2004 as the Safe Practice
Guidelines, re-updated again in 2007 with a massive rewrite, and all of these were peer reviewed and published, and then in 2011, actually it's on my desk right now, we're finalizing and correcting some of the references. It's being submitted now for peer review and publication for yet another update.

This document is the first to define or recommend that there be safety protocols specific for magnetic resonance at each site, that there be a MR safety officer or medical director specifically in charge of those issues. This is the document that created the now universally applied four zones of MR safety. This is the document that defines different levels of MR personnel based on their responsibilities and their duties. It recommends mandatory and recurrent magnetic resonance safety training and education for all those involved. It recommends ferromagnetic device testing and labeling. It was the first of the organizations that recommended the usage of ferromagnetic detectors as an adjunct. In fact, the very terms, MR safe, MR conditional, and MR unsafe, which are today ubiquitous and accepted by the FDA, the ASTM, and the entire industry, were actually first recommended by individuals from the MR Safety Committee.

But the problem is that the American College of Radiology is not a regulatory body. It's a source of knowledge and expertise, and it's advisory in its capacity. The Food and Drug Administration, on the other hand, is a regulatory agency and does have, not only the ability but, pardon me, the charge to do so, and as such, to quote the main purpose of the FDA, is to protect citizens from products that are inherently unsafe, and regulation of medical devices is certainly within their purview.

So return to them specifically, and why should we specifically be recommending the ACR MR Safe Practice Guidelines? Well, there are multiple reasons, but let's just use some practical ones. In the well over 100 legal cases that I've been involved with since 2002, that have gone to court or have threatened to do so, the vast, vast majority have quoted this document and have held it up as what they have defined as the national standard of care, which it has literally and de facto become.

The Joint Commission published the Sentinel Event Alert Number 38 in 2008, and in there they provided some risk reduction strategies, all of which come from this document. They have "Joint Commission recommendations." Of the 10 recommendations that are outlined in red, 9 of them are directly coming from this document.

You already heard from Mr. Gilk and myself, we have performed this review of 2009-2010 MAUDE database. We found that of the 154 magnetic resonance-related adverse events from the database, 84 percent would have been mitigated or would not have happened if the recommendations of this document would have been implemented. These are simple, common sense. They work. They're direct, straightforward, and inexpensive to implement.

Since 2004, we know that there has been an almost doubling of the number of MR sites, but we know that the number of reports, not just to MAUDE, which as we all know is exquisitely limited, but the number of cases that have come to my attention at least from legal actions have markedly continued to increase.

Even if they don't increase, one thing is absolutely certain. They're not going away, which means we're not doing a good enough job, and voluntary compliance isn't sufficient. I'm doing a complete 180. I always wanted us to police ourselves. It is my belief that we have documented that we don't do a good enough job.

In summary, the American College of Radiology recommends that the FDA define and establish standardized minimum MR safety-related protocols and practices to be followed by all MR sites, and we are confident that the first such template for such documented standardization should be the ACR MR Safe Practice Guidelines.

It's important that you recognize that these were created by specifically groups of experts in the various areas of MR safety, MR physics in its clinical implementation as well as input with well over 100 other sources and that it is constantly updated, maintained, and reviewed as the industry continues to evolve.

And, in closing, I'd like to say that I think these are the ultimate and low hanging fruit. The point of this meeting was "The overall goal is to discuss strategies to minimize patient and staff risk in the MRI environment." Here we have a document where if we follow it, we know that we'll be able to increase safety with minimal expense and with minimal difficulty. This should be continued to be maintained and updated as our industry continues to evolve.

We thank you for the opportunity to present. Thank you very much.

(Applause.)

DR. WOODS: Our next speaker is Dr. Frank Shellock from the University of Southern California.

DR. SHELLOCK: Good afternoon. I welcome the opportunity to present to you today. I want to thank the Food and Drug Administration for organizing this very important workshop.

It looks like you have my -- Dr.  Troy. Anyway, I'll deal with this.

I want to talk about two different topics with regard to MRI, general MRI safety issues, and one of the questions is where are the resources? Where's the information? We need education on an ongoing basis, and this is vital to ensure safety in the MR setting.

A lot of different types of resources exist. I want to go over some of the websites where this information is readily available to everybody that wants to access it, and it is there for the taking. So with regard to different type of websites, the Food and Drug Administration has MRI information on their website, and it covers all the important aspects of safety, bioeffects, patient management. Same thing with American College of Radiology, acr.org, where, as Dr. Kanal pointed out, the White Paper is there, and they also have links to important MRI safety resources.

The International Society for Magnetic Resonance in Medicine, which is the premier society for MRI healthcare workers, has comprehensive information for safety, bioeffects, and patient management as does the Institute for Magnet Resonance Safety, Education, and Research.

We now have several training videos. These are programs that are designed to train MRI healthcare workers, also ancillary staff and MRI workers with regard to radiologists and technologists, and that information is available for free, and so those programs are there and can be accessed on an ongoing basis, and they're updated on a regular basis as new information comes to light. And, of course, mrisafety.com, we have information for thousands of implants and devices. We have many aspects of safety covered on mrisafety.com and readily available to answer questions as needed.

And importantly, and I think this is sometimes overlooked, we have over 150 topics covered, and these topics are covered or updated on an ongoing basis, and this information can also serve to supplement the ACR recommendations, their White Paper, and also to help flesh out some of the policies and procedures that may be needed, for example, for pediatric settings, outpatient facilities, or research facilities.

And I think that one of the important roles of the FDA in this entire matter is to help increase awareness of MRI safety issues and keep their site updated and add information and important links to best serve the MRI community and the MRI healthcare workers in general.

Thank you very much for your attention.

(Applause.)

DR. WOODS: Thank you. Our next speaker is James Meyer from Children's Hospital of Philadelphia.

DR. MEYER: Thank you. Hi. I'm happy to be here to speak today to you. I'm a practicing radiologist. I'm also the Associate Chief of my department at Children's Hospital of Philadelphia.

Today I really want to just discuss two areas of concern. First is the increasing presence of MR scanners, which has already been alluded to in the comments afterwards, after the last two sessions, in what are predominantly and primarily clinical environments rather than imaging environments.

And second is how the present push that's going on to increase the implementation of the electronic health record in the United States can be used to improve MR safety in particular in regard to implanted devices.

So as was said earlier, you know, there are more scanners now that are in a clinical setting. Some of them are going to ORs. We have one that is right next to a CAT lab, and there are challenges that come with this. One is that, and certainly it is what we faced, was that we could no longer have that classic four zone MRI design. And also, and this is probably the bigger importance, is that the clinicians who are now closely integrated into this environment have really different priorities as well as different understandings of the MRI risks and safety risks than the imagers do.

And then there's this other point which is that, you know, if you have something that's right next to an OR, it's not very hard to go and grab some ferromagnetic scissors in the OR, stick them in your pocket, and then run over to the MR scanner and not think about the fact that they're in your pocket. It's much more difficult to do that when the MR scanner is three floors away from the clinical environment. And as a result of this, I believe and certainly I think it's proven that there's an increased risk of safety incidents.

So we've kind of dealt with already, and what we've done is we've come up with a few different approaches. It coincides with an overall hospital effort of increasing patient safety and awareness of patient safety, and the first, a key element, is deference to expertise, and this is the expertise of the MR technologists. We have made it very, very clear. It took us, we started about two years ago, that it is the MR technologists who lets people go in that MR scanner. It is up to them. We don't care if you are the CEO of our hospital. If the MR technologist tells you to stop, you stop.

Now, it says may require a culture change. It does require a culture change. It is, you know, you have a physician and you have a physician who is, especially like one of, you know, our sedation docs, an anesthesiologist, if the patient starts to de-SAP, a lot of kids with, you know, a lot of sedation going on with us, they start to de-SAP. What does that person want to do? They want to run in the scanner and take care of that patient. Understandable but they don't go in unless the technologist says they can go in, and maybe it takes five seconds longer to do that, okay, it's a whole lot better than having some device or some object fly out of their pocket and kill that patient.

So we've also added some physical barriers. We now have a chain that goes across the outside of the scanner threshold. We do have metal detectors. They are helpful. They obviously are sometimes, when you have somebody running, it's buzzing when the person's already passed it.

We've also come up with a very standardized approach to screening at the MR scanner, and the best way to be standardized and consistent is to use a checklist. We talk back about, you know, what the nuclear reactor industry has done, what the airline industry has done. They rely on checklists. So we've put in a checklist.

This is our checklist, and if you're going to do anything that comes up to try to improve patient safety, one of the things you're supposed to do, it's a good idea, is to track how well you're doing, and you can track it by your number of events, but events are relatively rare hopefully. So what we've done is we do randomized checks of how well our technologists are doing and following this checklist.

So you can see when we started it, 56 percent. Now, this is compliance with the overall list, which includes not only screening the patient and the other employees who might be going in but also checking that we're doing the right study and on the right patient. And it took us a few months to get up to about 90 percent and now up to about 95 percent. Obviously our goal is to get to 100 percent. The big push is going to be to get to 98 percent. Those last few percent are tough.

So that's number one. That's the way we've dealt with it. I think it's an increasing issue, but I think that the big thing, the way you're going to deal with it, is by taking care of the way your people deal with it in your situation.

Next is the electronic health record. Lots of implanted devices. There are more implanted devices. We all know this. They cause a significant risk. Patients don't know what they have in them, and what we usually do is we have to chase them down. I'm sure you do the same thing. Your technologists, in our case, it is our technologists, not our physicians, who are chasing down these implants. What are they made of? You know, what information do we have from the manufacturer? So I'd like the manufacturers to be required to provide this information. They should be clear and concise as they've already been asked for, and it should be easily available on a manufacturer's website.

The other thing is, and that's what the push for the electronic health record is, is that if you're going to put in electronic health record, take this opportunity to include MR safety. Make it so that you have to include an implant device section. It will improve safety. It will improve efficiency which, of course, will save some money and, you know, the meaningful use criteria that have been developed, if we could have it that one of the meaningful use criteria is implant device sections in electronic health records, that gives a carrot to those who are developing electronic health records as well as the hospitals that will be implementing them.

The device information that I'd like to have required, as I'm sure everybody else would like to have required, maybe more than I have listed, but I think this is what we feel would be necessary and would be very helpful. So I would push, a huge push, to see if we can get to have this be a criteria for meaningful use.

So, in summary, there are an increasing prevalence of MR scanners in the clinical environment, and if you're dealing with that situation, or will be dealing with it, be prepared for some changes. Be very careful, as best you can as far as the physical environment, but also work on your culture, work on your culture early. If you can do it now, do it now. And then as far as the implant devices, again I would just like to push for having the electronic health record, including the presence of implant device information as a criterion for meaningful use. Thank you very much.

(Applause.)

DR. WOODS: Thank you. Our next speaker is Sonja Robb-Belville from the University of Texas MD Anderson Cancer Center.

MS. ROBB-BELVILLE: Thank you. Good afternoon, ladies and gentlemen. And, again, my name is Sonja Robb-Belville, and I am the MRI Education Coordinator for the Diagnostic Imaging Program at the University of Texas MD Anderson Cancer Center.

I'd like to acknowledge the FDA for recognizing the importance of MRI safety by holding this workshop, and also the members of the MRI community who stood for and requested that this workshop be held.

My background includes 20 years in the field of medical imaging, 14 of which have been as a scanning MRI technologist and 11 of which have included educating both students new to the field and technologists already working in the field of MRI and obtaining their certification.

So today I want to contribute that perspective to the workshop. And, of course, I'd like you to consider that the time tests or track record with millions of MR procedures performed throughout the world, we know that MRI is one of the safest imaging modalities used properly. It's also the most dangerous and deadliest, the fastest, used improperly.

So the key points I'd like for everyone attending this workshop to take away with them today are that the safety guidelines and recommendations already exist. We're clear on that. But there's a cavernous gap in between their existence and their implementation, and the barriers that are present to that, they include the lack of standardized regulation, the subsequent lack of enforcement, okay. So there are multiple organizations worldwide as we've seen, ISMRM, the ACR, NEMA, the Medical Devices Agency, the IEC, just to name a few that have recommendations and guidelines.

Most, if not all, of these guidelines are supported by scientifically valid research. Yet one look at the MAUDE database shows that the reporting is significantly increasing over the last five to eight years. With all these guidelines and recommendations already in existence, one must ask why the rate of accidents continues increasing at such a dramatic rate.

It's the ethical duty and privilege of all healthcare providers to take every necessary step to minimize the risk versus benefit ratio of patients undergoing any medical procedure.

As an educator, in an effort to reduce the risk versus benefit ratio for patients, I teach my students as well as practicing MR technologists about the existing safety guidelines, such as those generated by the ACR Guidance Document for Safe MRI Practices and the labeling standards for implantable and accessory medical devices developed by the ASTM which were adopted by the FDA in 2005.

In practice, healthcare providers are often not aware of these existing guidelines. For example, poor dissemination of the updated implantable and accessory medical device terminology, many healthcare providers refer to the devices labeled after 2005 as MR compatible devices. Furthermore, technologists frequently deliver healthcare in an environment that is not managed by healthcare professionals.

While there's nothing inherently wrong with making a profit by providing healthcare services, these managers are not bound to the same code of ethics as healthcare providers. Management therefore uses very different criteria in measuring successful performance. All too often managers tell technologists that existing safety guidelines and recommendations are not regulations, and moreover, since their implementation is not required or enforced, that MR safety guidelines aren't important.

Couple this with the perfect storm we have brewing on the horizon. Multiple factors including declining reimbursement rate, increased demands on the healthcare system, and the advancements in technology that are on the horizon, we are going to be asked to do less with more and to do it faster, and that is going to provide a confluence of those factors that will force those technologists to push those patients through faster and is going to create a potential for a spike of even more safety accidents with MRI.

Further, as an educator, standardization of the training is necessary. There are varying degrees of programs, varying program lengths, et cetera, in terms of people being able to obtain their certification as MRI technologists. So we must standardize that as well and provide robust simulation environments for students to practice performing MR procedures without putting patients at risk.

I've had the honor and privilege of serving on the ARRT's Practice Analysis Committee for the MR certification exam, and I met with them just last week in Minneapolis, and they were taken aback by the dramatic revisions that were needed to our certification process. We spent every bit of that meeting hammering out everything new that needed to be put into it, and they were shocked.

This being said, MR technologists depend upon the scientific community and the FDA to determine which guidelines and recommendations are scientifically valid and should be implemented. The existing guidelines and recommendations must be carefully examined for scientific validity, then standardized, regulated, and enforced. Additionally, it is imperative that they are effectively disseminated to the members of the medical community.

In closing, we must build a safety culture as we heard several say today, and it does not require reinventing the wheel. There are several industries that we could look to and use as a map to provide a model that we could adapt to MRI. As technologists, we're privileged to be the gatekeepers, but we need the powers that be have our backs. We are just one spoke in that wheel.

Again, MRI is one of the safest modalities used properly, and it is the deadliest and fastest used improperly. Thank you.

(Applause.)

DR. WOODS: Thank you. Our next speaker is Devashish Shrivastava from the University of Minnesota.

DR. SHRIVASTAVA: Thank you. I'm going to talk about in vivo radiofrequency heating during magnetic resonance imaging of high and ultra-high fields. The goal of this investigation was to improve RF safety at high and ultra-high fields by, one, developing a fundamental understanding of the RF heating, second, studying thermo-physiological effects of the RF heating to determine harmful temperature-time threshold and, three, correlate global and local SAR to in vivo temperatures versus time to ensure safety.

The motivation came from the fact that RF heating and its thermo-physiological effects are not well studied in high and ultra-high field imaging. Safety in high and ultra-high field imaging depends on the determining special and temporal distribution of temperature in vivo and not SAR, and that is because of the blood flow effects.

No validated engineering tool existed to determine heating in vivo. Maximum local SAR and hot spots may not always correlate, and that is again because of the blood flow. Blood can cool the tissue but also can transport its energy to the rest of the body where it can heat where it is traveling to.

So in materials and methods, we use anesthetized pigs. I'm going to show you four different results with four different coils. We have a 3T head coil. We have two 7T head coils. We have a 9.4T head coil. I'll show you how RF heating takes place during head imaging in the head in different places. We use fluoroptic temperature probes to measure this heating, and we use bioheat transfer models to predict this heating.

We use swine as animal model because of thermal mass. We can get swine as big as humans. That is important because that's how we store energy in our body. Surface area, the swine surface area is very close to human surface area. So the amount of energy that a swine can get rid of on its surface area is very close to humans.

Thermoregulatory control in a swine is very close to humans. For example, cardiac output, sweating, water loss, metabolic energy, and critical hot temperature. Thermal and electrical properties of the swine are also very close to humans. One thing that is good in our favor is that swine do not actively sweat. Therefore, this is kind of a worst case scenario because we as humans would be able to get rid of this excess energy using sweating while swine would not be able to, which would be relevant for older patients.

This is a setup. You can see that this here is the head coil around the pig's head. We have temperature probes, in the image, where we have several temperature probes. We have five specifically at different depths in the brain, and we use these head coils to put power in the head while we measure temperature in the brain and the body of the swine.

This is a typical temperature profile. You will see that in the anesthetized pig, during the pre-RF period, temperature keeps going down preliminarily, and then during the RF period, temperature will go up, and then post-RF, the temperature will come down.

Now, these are long experiments, and that is because body takes roughly three hours to stabilize. If you want to know how much heating you are going to cause during the RF period, you need to know how the temperature is dropping in your body, and therefore, you have to wait roughly three hours for them to give you a pattern so that you can subtract that RF heating curve and know what your RF heating is going to be.

Similarly here in this fourth RF region, you need to again wait for 3, 4 hours because in the first 45 minutes, body will just cool down and you won't know if your heating has done anything bad or adverse to change the chemical physiology of this animal.

This is again a typical temperature during anesthesia. During anesthesia, temperature drops linearly in the brain, under the skin, and under rectal body, and that was the justification for subtracting three RF slopes from the other RF region to get to the temperature distribution.

To model the temperature profile, you use Pennes' bioheat transfer equation. This is an empirical gold standard bioheat transfer equation that is used all over the field. This is an empirical equation, and therefore a relationship between equation parameters and physiology is unknown. Blood is treated as an infinite heat sink, and therefore, in all practical cases you are going to underestimate heating inside the body.

This is the generic bioheat transfer model. We had to, given the limitations of the Pennes' bioheat transfer equation, we had to devise our own bioheat transfer model by conserving energy in the finite tissue volume, finite heated tissue volume embedded with the blood vessels, and this is the form. We can talk about this more if you're interested, but this basically gives us two coupled equations. One is for the tissue temperature, and another is for the blood temperature, and you can see that they are coupled by the term that seems to be a permanent energy transfer between the blood and the surrounding tissue.

Going to 3T RF hearing results, we took this 11.2 inch ID coil. It had 15 rungs, volume head coil, and we deposited roughly 50 watts of net RF power to the coil. The electrical Q measurement based efficiency of this coil was 80 percent, and then after the fact, we calculated also the thermal efficiency based on the measurement which came out to be 55 percent.

This is the RF heating results in the skull. You can see that the skull temperature was varying all over the place. However, our model predicted these temperatures much more closely compared to the Pennes' bioheat transfer equation. This is the temperature heating inside the brain. Again, you'll see that, one, heating is unique in the brain for a given position. Second, there's no steady state in our heating. It just keep going up, but Pennes' bioheat transfer equation predicts a steady state, and it underestimates the heating.

This is rectal heating, same thing.

7T RF -- I have only one minute. So I'll go quickly now. I'll just show you the 8 channel of 9 inch ID coil, again we put roughly 50 watts of power in this case, and the coil coupling efficiency was 33 percent and coil thermal coupling efficiency was 55 percent.

What I want you to see is that in the 3T result and in the 7T result, the amount of heating in the brain is same even though our measurement, electrical coupling measurement which are typically done to measure how much power goes from coil to the head, is very different. Eighty percent in one case and 33 percent in another case. The heating in the brain was same, two degrees of RF heating inside the brain of this pig. So there's a discrepancy between what we measure and what we predict in terms of how much power is going to the head and how we calculate our SAR values.

And then -- results. So, okay.

In summary, RF heating is coil dependent, but especially and temporally unique, we saw that. We also showed that generated by a heat transfer model, similarly RF heating more accurately compared to the Pennes' bioheat transfer equation because of the inappropriately modeled thermal blood to tissue heat transfer, and we also found out that the electrical coupling efficiency may not be a good indicator to determine power deposited in vivo and therefore calculate SAR values, and that heating on the skin is unreliable predictor of in vivo heating because we as mammals have tremendous ability to regulate our blood temperature on the skin, sorry, blood flow to the skin which will change the temperature profile. Thank you.

(Applause.)

DR. WOODS: Okay. I guess now to put Niels' co-authors on the spot -- he was not able to be here today. Is there one of his co-authors who is here who is prepared to give his presentation? Feel free to say no, but I also don't want to deprive you of that opportunity if you'd like to take that opportunity.

Okay. Seeing no takers, then I guess we'll go on to our last presentation in this session, Fraser Robb from GE Healthcare.

DR. ROBB: I'm Fraser Robb from GE Healthcare speaking on behalf of MITA. As human beings, we all suffer from the human condition, and we all get sick and die at some stage, as well as our families and friends, and it's been a primary motivation for innovation in medicine throughout the centuries. If you look at this drawing from the Middle Ages, we see clearly what the artist imagines for the future is that one day there'll be miraculous machines that look like kind of a big hole in the wall. You know, there will be some kind of efficient transport mechanism, but largely there will be miraculous machines for the future that can do wonderful things for healthcare.

You can see at the top left-hand corner they're following good clinical practice, doing volunteer scanning procedures. In the event of a -- see all the evil spirits go out the top -- fast forward 500 years. This is our picture from the University of Aberdeen, 1979. This is the first clinical imager that was ever developed, and even at that time, you know, the researchers were very conscious of the aspects of safety related to these new miraculous machines, and as our static field gradients -- although the biggest concern at the time was banging your head off the face of the magnet if you woke up with a start. And although many things have changed, the pillow seems to have stayed the same throughout the decades ever since.

So, you know, in terms of trends in the industry, I think we've all been aware that clinical MRI is now a real clinical reality, and at the same time there's over 50 7T systems in the world producing extraordinary image quality, wonderful resolution, and in many ways, this is nothing really to be scared of. You know, there's 7T. Researchers have been working very hard on all aspects of 7T, including MR safety. We saw a nice presentation from Minnesota just earlier. And 7T really has been 15 years or more in incubation. So there's a lot of question perhaps originally -- ever be used clinically, but I think we need to do a little research base and see what's been done over the last 15 years to reassure ourselves that, you know, this is something that could be a reality in hospitals one day.

At the same time, systems have got wider bores. You've probably seen that recently, and the industry has really moved from a 60 centimeter bore to 70 centimeters. At the same time we've moved from quadra-drive to multipoint drive. You know, in my own mind I think potentially this should make MR safer in the long run in that we all now have far greater control over the RF P1. Of course, it requires a lot of calculation and good understanding. And potentially this is something that could make MR safer.

At the same time, -- coils that all the MR technologists here just love MR coils because they're sometimes -- that contact the patient almost immediately, and so in terms of complexity, they've really developed from perhaps 4 channel from the original -- 21, 22 years ago, the 4 channel in the '90, to 16 channel in the last decade, and there's a lot of development in higher channel type coils, and in many ways the coils are becoming more complex, but we also have tools now and more carefully understand them. There's numerical tools available now, now that allow us to design to much higher standards. And also element sizes are getting smaller, so instead of very large elements in four channel systems, we're now getting elements about six, seven centimeters in diameter, so smaller voltages are just in those coils, and in general, they are getting more complex, but I think our ability to understand how to, you know, to develop them has also increased equally.

We're looking forward to the future. You know, could transceiver arrays become a reality? You know, is parallel transfer, parallel transmit around the corner beyond what we've done already? Here we really need to look to these wonderful new sound modeling tools that can give us really good understanding of local side effects.

And, again, my point of view, the increased beyond control should lead to, you know, more careful understanding and planning of local SAR, and so as a means to the future, say, lead to a safer future for MR.

How will MR evolve? Well, here's another picture from the University of Aberdeen. This is actually a head only scanner that was developed more than 10 years ago. It's a low-field device. It's purely research. But, for instance, you can imagine, you know, the future where not all MR systems look like they do right now in that the whole body needs to -- the whole patient needs to go in the scanner; perhaps you'll only need to scan the region of the body you're interested in. So I think by definition, through the specialization of MR, we go to a safer future.

Here's a couple of examples of that. On the left-hand side, there's a dedicated extremity scanner. So only the knee or the elbow or the arm goes in the scanner, and the rest of the patient sits comfortably outside. Here's -- on the frontiers of research, we see a neonatal scanner. And that is a research project at the University of Cincinnati --

So here we have a cuddly soft lamb from Toys R Us. That kind of gives you the idea of one day you might be able see specialized MR scanners in NICU, so that you save the traumatic experience of taking a baby, a 24-week baby out of the NICU, across the hospital, to a remote radiology department. So by definition, that's a safer future for MR.

You know, there are other aspects that's in research at the minute, also going towards mainstream MR, and I think industry understands both the risks of MR and PET separately. So I don't think we should be necessarily any more afraid of combining modalities like we did with PET and CT.

Hyperpolarized carbon imaging, some wonderful research is going on at the University of California San Francisco. And, you know, looking at new types of contrast agents to show off new types of disease, now provided that we work closely with the regulatory agencies, there's no reason why there should be any -- We will be able to get wonderful new diagnostic power, you know, working with regulatory industries, won't necessarily have to be any less safe than the conventional methods today.

Combining ultrasounds and MR, they're a wonderful new technique, certainly much improved or hopefully provides improved outcome for the patient but, you know, it's like again we understand ultrasound, we understand MR, so let's just work sensibly together to understand what the future is from a safety point of view.

Going into, looking into the future, you know, perhaps MR cardiac EP could become reality one day. So, for instance, instead of using the x-ray technique. So these will very complicated procedures from a workflow point of view, but certainly better than using ionizing radiation for the same -- you know, in principle, better than ionizing radiation for the same technique.

And then someone mentioned interventional MR earlier, and see here's just an image of what a suite might be. Again perhaps as MR proves itself, you know, as new technology for the future, we can move away from more direct ionizing techniques. We just have to work, you know, and understand workflow and the safety of both systems to make sure that we provide what the regulatory agencies are looking for.

Beyond that, that's kind of -- there's plenty other areas that MR could move into, that I think in many ways innovation can provide an extra safety margin as we move forward as we provide greater tools for understanding.

(Applause.)

DR. WOODS: Thank you very much. I'd like to again thank all of our speakers for being so prompt and ready that we are once again ahead of schedule. So what we'll do is take a 15-minute break, come back at 2:00, and then we'll have our roundtable.

(Off the record.)

(On the record.)

DR. WOODS: If our roundtable speakers could come up and take your seat behind your name cards cleverly arranged in alphabetical order.

I'm going to try to use this microphone. To the roundtable speakers, the microphones on the table, you have to hold the button down the whole time you speak, and I guess to avoid holding the button down, just to start, first of all, I'd like to let you know that the roundtable speakers, we chose to try to give a representation from many of our stakeholders, from different groups. So we have clinicians, MR vendors, implant manufacturers, ferromagnetic detector manufacturers, and a good representation from the clinical community as well as one FDA moderator.

And the format of this roundtable, what we'd like to do is have people from the audience come up to one of the microphones and ask a question of the roundtable and again, to get the diversity of opinions, we'll just give everyone an opportunity to answer all of the questions. You know, if you have nothing to add, you can pass to the person to your left.

So is there anyone who has any questions in general for the group on the subject of today, which was MRI safety in general, ferromagnetic detectors, and then the impact of innovation on MR safety?

DR. MEYER: Hi, I'm Jim Meyer. I have a really basic question, but it's something that we're dealing with right now, is that when you have a -- basically how do you guys deal with an implanted device that has had a change in its MR safety rating, has become a more -- is now unsafe from conditional previously. Do you, you know, I'd just like to know if -- because we're actually dealing with an issue like this right now, and it is obviously very frustrating, but we just wonder if it's, you know, how we should deal with it, and since we have a great panel in front of us, I'd love to hear some opinions.

DR. WOODS: Can I ask for you to clarify that? Is that something that the labeling for the device has actually changed or --

DR. MEYER: Yes.

DR. WOODS: Okay.

DR. MEYER: The labeling for the device has changed. It went from a device that was felt to be safe and now is unsafe.

DR. CRUES: This is John Crues from Radnet. If it's labeled unsafe, then we do not allow it in the magnet.

MR. FRESE: Georg Frese, Siemens Healthcare. I think I'll pass. It has nothing to do with the MR system.

DR. FRICKE: I'm Stanley Fricke, Children's National Medical Center, and we see this as well. Typically we're not allowing them in the scanner, although if we have scanned the same patient under the same conditions in the past, we evaluate, of course, the benefit and if true benefit is there and it outweighs the risk, and at the same time we have a surgical opinion that's going to be prepared to intervene if necessary.

MR. GILK: This is Toby Gilk. I don't think that this is the exact context of your particular problem, but I know that one issue that comes up is when we test for MR conditionality on a particular scanner, that the test conditions and the information available to the clinical providers do not necessarily match up 100 percent and, in fact, what we often find are facilities, there's the St. Jude heart valve replacement that's sort of a poster child of this condition, where it gets tested on a MR system, and then clinical providers who have the exact same MR system, the information they're presented with suggests to them that they cannot safely scan that implant based on the information that they get (a) from the implant manufacturer and the test conditions and (b) from the MR system manufacturer, and the disconnect between the way that that information is presented prevents a lot of MR providers from scanning patients utilizing the magnet that was used to demonstrate the safety of the implanted device, and I think that there's some inherent problems in the test criteria, test and labeling criteria that we're not actually testing safety. We're testing one set of parameters, and whether or not it's safe, and the fact that a manufacturer of an implant decides they want to change their mind with respect to what the criteria are for minimum safety, I think that that goes to a larger question of, you know, we're not actually defining the safety threshold or the safety tolerance of an implanted device. We're just establishing one set of conditions in the test parameter.

MR. GOODYEAR: Simon Goodyear, Managing Director of Metrasens, a ferromagnetic detection system manufacturer. As such, I don't have any experience in making those sorts of decisions, but just to talk about our research and development team are looking into potential technology that could help with making decisions like that and recognizing when implants have changed, and that may be something that would be available in the future.

DR. KANAL: Manny Kanal, ACR. I'm not quite sure I understand the question. What is it that has changed if I may ask of you? Is it that someone has taken a device, took the information that they knew in the past, in the past labeled it X and now labeled it Y but no information has changed about it?

DR. MEYER: Correct. As a matter of fact, this goes back to the -- this is a patient who we've actually scanned twice before. At that time, it was felt to be MR conditional. The manufacturer changed it. Now, we need to do a follow-up scan, and now it's MR unsafe. So the device, you know, it's been in this child for probably eight years.

DR. KANAL: Thank you for the clarification. I can't begin to understand that unless there was a change in the information obtained. I believe you're trying to suggest the science is unchanged. It's just how someone chose to label the same science. If that's the case, I'd change nothing in how I would -- I no longer represent the ACR. I only represent myself on this response. I would change nothing. If I thought it was safe based on the information and testing that was provided before, I would do so again. If it was unsafe, regardless of how they chose to label it.

We have stent manufacturers who say that you have to wait six or eight weeks, and Frank Shellock and I have been laughing with tears as to how silly that is if there's no significant ferromagnetic properties. How they chose to label it isn't really the concern. The concern is, is it safe or not? So I guess I'd need to know for the specific device you're requesting.

The generic answer would be if the data suggests it can be safely scanned under certain conditions, that's how we would proceed regardless of what labels they chose to attach to it.

The last thing I'd like to add is I've heard several times from you and from others in the panel, I think it is an extremely dangerous precedent to assume that a previous scan went without any difficulty and therefore it adds to my confidence that I can get away with it next time. I've crossed this street several times with my eyes closed, and therefore it's safe to do so in the future is not based on science.

DR. MEYER: So in response to that, I mean I think you're absolutely right. So one of the things is that we don't know why the manufacturer changed their designation. Maybe they changed it because there have been issues and maybe we have gotten away with something. Maybe we had a sedated child who, you know, something was happening but we didn't know about it. So that's, you know, your point goes to the side of saying, well, it's MR unsafe. I mean it hasn't changed.

DR. KANAL: I would therefore have to, since the information being provided seems to have changed, then I have to change my response. If you're saying that there is a possibility that the information available has indeed changed about the device, then I would err on the side of safety and suggest that it may not be safe.

DR. MEYER: I don't know.

DR. KANAL: I'm not sure I really understand the specific case.

DR. MEYER: Yeah. Okay. That's why we're having this --

DR. KANAL: I would certainly contact the manufacturer of the device for clarification.

MR. KOPP: Keith Kopp of Kopp Development, makers of ferromagnetic detectors. I really don't have any specific knowledge I can give you on that, so I will pass.

DR. LAZEBNIK: Roee Lazebnik, Chief Medical Officer for GE Healthcare, MRI. I also don't have anything specific to add on this issue.

MS. RIVI: Diana Rivi, FDA. I also would second going back to the manufacturer and speaking with them, but I think also this topic might also be addressed in tomorrow's sessions as well.

MS. SAWYER: Anne Sawyer from Stanford. One thing I just wanted to say about safety policies and procedures and why you have a safety team is that your policies and procedures should be a working document that you revisit often, including the screening form, and obviously to hopefully address things similar to this, and I'll also repeat what other people have said. Hopefully you have talked to the manufacturer, and I know how difficult personally that can be to get accurate information now, but it raises a lot of questions at least in my mind. Is it the old devices? Is it new devices going forward? Did they change the materials? Was any of this reviewed by the FDA? If they did, I don't know if that's required. Sometimes it is. Sometimes it's not. So it just raises a lot of questions in my mind. Do you need to change your policies and procedures? But again, there's a lot of homework you need to do it sounds like.

MR. STANLEY: I'm Charles Stanley, SMRT. I would just kind of reiterate what Dr. Kanal said and just say that if you're the M.D. and you had information specific about that device that led you to make a medical decision to say that it was safe, and that information hadn't changed, then you would proceed. But if you're basing your decision, risk versus benefit, on the information that you're getting from the manufacturer, and the manufacturer has changed their position, then you would have no choice but to change your position as well if that's what you were using to establish your criteria for scanning.

DR. MEYER: Thank you all.

DR. FELMLEE: Hello, everyone. My name is Joel Felmlee. I'm a physicist from Mayo Foundation. I'm interested in your thoughts on I guess the following. I can see where there's a class of patients that ferromagnetic detectors can work very well, but I can also see some classes where they might be difficult. So I'm interested in your assessment of those two and specifically patients that are on a table, that the table has been brought outside the room but now needs to go into the room, such as anesthesia where the patient's been induced, present a real challenge for us. I'm interested in your opinions on that.

DR. WOODS: Joel, could you clarify the two. Say you have the one patient that's on a table versus someone on foot. Is that --

DR. FELMLEE: Yeah. The patients on the table will set off the detectors, and it's really a challenge for us because they've been outside the room. There's lots of equipment associated with anesthesia, and sometimes it will get left in the folds of the blankets and stuff, really complicated cases. I'm challenged with how to handle that from the safety aspect, and if these detectors could be helpful in that, I'd like to understand it better. We're frankly challenged by this as an industry.

DR. CRUES: Well, I think the issue is that the table itself sets off the alarm, and therefore every time you bring a patient in, there's an alarm, in which case the technologist would have to just start ignoring the alarm, which might give them a false sense of safety, and in the outpatient setting, we really don't have that issue. So I'll pass to other members of the panel.

MR. FRESE: I can only make a brief statement. A patient table is really difficult to design without ferromagnetic parts. For example, the needle bearings on something like that which are not creating some kind of hazards might be ferromagnetic parts, and that could cause the signal, the ferromagnetic detector to be activated.

DR. FRICKE: While I appreciate the question, because I live it every day as well, so I sympathize with you, I think this is a challenge first of all to the vendors. Please, come on, if we can put a man on the moon, let's make a table that doesn't set off the detector. It's doable.

Secondly, you're right. If you're sending something through the detector that's going to say I have metal and you have to ask yourself a question, do I already know that this metal is there? That's the wrong answer. That detector is not a yes or no machine. That detector is, you better check it again. When it goes off, I don't care if the table's going through, you're going to check.

Well, since you know you're going to check in this particular case every time, what we have implemented because of the same issues and we've been quite successful with this, is a specific equipment time-out. That's where you stop, you check the equipment that will be setting off the metal detector, you check it for everything from the charge on the battery or in your case, the patient themselves, physically with one person responsible for the entire process, keeping your eye on the patient.

The only way to do this is to have an equipment time-out, and this is a person-oriented thing. The metal detector won't get it. You've already said the metal's there. So you've posed the question that theoretically can't be solved. Protocol in this place is the only place to solve it.

The only stop point that you can do that is immediately before the patient's going into the room, not 5 minutes before or 10 minutes before. It has to be immediate because if it's anything but immediately before the patient's going into the room, things change.

MR. GILK: This is Toby Gilk. I take your point. You know, we can put a man on the moon. Why can't we make these tables nonferromagnetic, and I would certainly hope that the MR system vendors work diligently towards that end.

There are accessory manufacturers that make transport devices, not the, you know, dockable MR system tables, but depending upon your operations, if there's a way to transport your patient on the nonsystem patient transport, there may be alternatives that are available to you where the transport itself doesn't set off the detector.

I think that there's tremendous room for improvement in this in terms of reducing extraneous ferromagnetic material to improve the specificity of ferromagnetic detection alarms for nonambulatory patients. My comments are directed specifically to the nonambulatory patients. The issue that you raised really generally doesn't apply to our walkie-talkies. So I think that there's a great deal of improvement that could be done in the MR system tables. There are third party products that have various levels of performance in terms of that, and then there are also, you know, handheld ferromagnetic detection systems where you can screen in a highly localized manner without the pass through. It does not provide the sort of bulk mass screening, but it can be used in conjunction with system tables and other components, other transports that are ferromagnetic or have ferromagnetic components.

MR. GOODYEAR: Simon Goodyear from Metrasens. We agree that for nonambulatory patients, this is one of the key things to overcome to get effective use out of a ferromagnetic detection system, so much so that we, as a company, and I'm going to be careful not to be too commercial, but we invested in making sure that there was ferromagnetic free transfer equipment. This is not the docking stations which dock to the MRI, but there are ferromagnetic free transfer equipment available already to make the system effective.

DR. KANAL: Manny Kanal. Joel, I think your question actually answered it perfectly. I don't even know if I could say it more succinctly than you did. There are situations for which ferromagnetic detection is ideally suited and others for which it's not, and I think the answer to this is relatively straightforward and easy, not the clinical situation but just the answer.

In 2011, it's my opinion that, number one, pass-through ferromagnetic detection systems are essentially useless in the setting of someone who's going to be on a transport that contains sufficient ferromagnetic material to alarm. So what we use for that in our institution is two things. First and foremost we use our technologists. That's why they earn the massive salaries they earn, and this is why they're worth their weight in gold. This is what we need to ensure that there's somebody knowledgeable doing the appropriate check. I used to love to say when we were talking about nonferromagnetic detector, but just plain metal detectors years ago, I used to say if you need a metal detector to detect the oxygen tank between the legs of your patient under that blanket, you're not doing a good enough screen.

So first and foremost, no kidding, it's the technologists and, number two, in our institution, we actually do use handheld detection in that format, that if we're looking for something that we think is potentially problematic, we can use that for screening focal regions of the patient. It is absolutely not used for total patient screening, just for answering questions in a specific focal region.

MR. KOPP: Keith Kopp for Kopp Development. I'll deal with the ferromagnetic detection part of this. Again, it's related to a comment I made earlier. Many sites want to have as best they can an overall comprehensive MR safety program which includes a lot of elements, sometimes ferromagnetic detectors. And in the same discussions, they'll buy what is labeled nonmagnetic gurneys and wheelchairs, et cetera, and the terminology nonmagnetic is very loosely defined in that many times they are magnetic.

Our field engineers, for example, always carry a hand magnet with them because they'll be confronted by whoever purchased this gurney, for example, and say, well, they said it here. The labeling says that it's nonmagnetic, and our demonstration is we'll stick the magnet to the side and it sticks to the unit. So they know, but we're caught kind of in the middle, and more importantly safety is compromised.

So this is another reason that all of the people, the manufacturers of the MRs that offer accessories, the other purveyors label this equipment properly so that somebody like yourself that's trying to do the right thing for patients has a fighting chance of getting it done and doesn't get these extraneous alarms.

DR. LAZEBNIK: The use of ferromagnetic detectors, at least from my perspective, is not specifically advocated for or against. What I would urge the room to consider is that the recommendations that come from the MR vendor regarding safely admitting patients into the room and so forth are independent of your choice to use or not use a ferromagnetic detector. So I would urge everybody to consider that the guidelines should still be adhered to again independent of whatever the reading was.

MS. RIVI: Diana Rivi, FDA. I think this is a good question, and since it's a practice-related question, I'm going to let those who practice best answer it.

MS. SAWYER: Anne Sawyer. I've been working in MR for 26 years, a little over, and all I have seen is that the patient tables for the MR systems have gotten heavier, gotten heavier, added more metal. In fact, they've gotten so heavy and have so much metal that I can barely pull it away from the front of the shroud, and we've measured this. And now we see a trend for putting RF coils inside the table. So I can only see that they're going to get heavier. I don't see that they're ever going to have a table that's not going to set off a detector.

So I have to agree with Dr. Kanal on this. I mean it is the technologists. They have to be knowledgeable. They have to be strong. They can't bend. And those are the gatekeepers as has been said by several people here earlier today. So that's what we need.

And we also need support from whoever runs your facility, whether it's M.D. or administration. You have to have their support because if you don't have absolute buy-in, you have this gray area, well, if Dr. so and so pushes, it didn't solve anything.

MR. STANLEY: I'm Charles Stanley, and this is going to be not the SMRT position. This is going to be my position. Since Bill Faulkner couldn't join us today, I'm going to do my best to do my Bill Faulkner impersonation, and I'm going to say, well, if even the manufacturers of these devices are saying that there's nothing that replaces 100 percent effective screening by technologists, my question would be is, why would you not just use that? Why would you even take the risk on something that's admittedly 95 percent effective and go to the 100 percent effective solution? Why not start with the 100 percent effective solution and make absolutely sure that your practice is that everyone is effectively screened?

DR. KANAL: Joel, another thing I'd like to suggest as a generalization. This is not just for somebody in that position but somebody unconscious or somebody in whom a procedure was just performed or anesthesia. A practice that does not appear in any guidance document to my knowledge yet but just as a recommendation that people consider is that immediately prior to entering a room, anyone with a procedure, anyone in whom anesthesia or a deep sedation has been initiated right before going into the room, my recommendation is that you consider remaking the bed the patient is on. It's at that time that you find needles that were stuck in the side or whatever. Just last minute. We're typically used to things from working in that kind of environment, the scissors, the tweezers, the needles, whatever, that might be left on the sheets. They may be difficult to find. So just conceptually remaking the bed before the patient is brought in might be a consideration.

DR. FELMLEE: Thank you.

DR. FRICKE: I have just one last comment. One of the things we implemented, and I was lucky enough to be on the design architectural phase of our radiology renovation for MRI, we actually have two doors going into Zone 3. One is a small door, that has a metal detector in the doorway, and the other is a large door that is meant for use in exactly your situation. The difference is this. The large door has no metal detector in it. The point is if the metal detector cries wolf all day, who's going to look and check to see if there's a wolf?

So if you have a way to design a process that cannot have the metal detector on when you know full well it's going to set it off anyway, to keep that thing from crying wolf, you're going to be a little bit ahead.

And secondly, I agree. Our time-out before we go into the room is done and exclusively done by the technologist who is going to be responsible for that scan. It's done immediately before taking the patient into the zone.

Now, there's a check-off that each person's responsible. For example, if it's a complicated case where we need a ventilator connected to the patient, our respiratory technologist has to say, my piece of equipment is okay. If anesthesiology is involved, the anesthesiologist has to say, my equipment is okay. So each person is responsible for their own equipment on their own time, in their own way, but immediately before going into the zone.

DR. WOODS: All right. Next question please.

DR. STECKNER: Michael Steckner, Toshiba. A comment that's come up in several of the presentations this morning and this afternoon is with regards to the use of checklists. Casting that question in the context of both passive and active implants, what would be the panel's thoughts about having such a critical question posed on the workstation prior to or before the scanning starts say as part of the registration process? Thoughts?

DR. WOODS: Michael, just to clarify, a question as to whether the patient has been screened or --

DR. STECKNER: Yeah, so to specifically ask have you asked about passive or active implants?

DR. CRUES: We certainly ask for all implants, and then I think it would be very useful to get, once somebody has implants, to ask specific information about the implant the person has, and passive versus active, I think is an important question.

DR. STECKNER: Maybe I'll just clarify what I meant is literally a question on the workstation that would then be captured and then perhaps part of the patient record for that examination.

DR. CRUES: Yes, I think that that information should be documented because I do think it's important information as to the specifics of the implants, especially where they're active or not, and that should be documented.

MR. FRESE: Michael, I'm not really sure if I understood your question correctly. Do you mean what we are already discussing in MT 40 about the using the pace when --

DR. STECKNER: Yes.

MR. FRESE: Okay.

DR. STECKNER: That's exactly what I'm asking.

MR. FRESE: Then I can briefly speak as -- of MT 40. MT 40 is the maintenance team responsible for the IEC Standard 60601-2-33 for safety, and we are meeting here in Washington around this safety workshop to work on the first amendment to the third edition of the standard. The third edition was published in 2010, and exactly as the implant issue is a very important thing, we are currently discussing how to create a user interface which is easy to handle and which doesn't provoke any errors and while we are in the middle of the discussion, and so far I can't come up with any results of the discussion.

DR. FRICKE: Well, I don't have too much input except to say that I wish every implanted device had an RFID on it so that we could immediately get the information we needed.

MR. GILK: And as long as we're wishing, I think the idea of having a question pop up on the workstation, most of the time the techs that I observe working, the first time they sit down at the workstation is when the patient is positioned on the table in the bore, and that is probably the last time that we want to find out for the first time that the patient has an implanted device.

If we wanted to do something at the system level, that did that, I absolutely loved the suggestion that was posited earlier today of having meaningful use criteria for recording the specifics of an implanted device. If there was a red flag in the patient's electronic medical record indicating that they have an implanted device, and if the system software simply identified the presence or absence of that red flag and highlighted that for the technologist as they're preparing for the exam, and not as they're about ready to sit down and run the first sequence, that I think could be tremendously useful.

I think having the question pop up on the screen for the purpose of having it pop up on the screen is of little benefit to anyone, but if we can integrate it into a more effective workflow, then I think you've got something.

MR. GOODYEAR: Yeah, I can't really add too much to that apart from the problem from listening today is more about getting the information into that, rather than how you record it --

DR. KANAL: Manny Kanal. I would consider it useless or hindrance on the operator console, and hopefully we'll work as an industry towards having it entirely integrated into the electronic health record that not only is obtained and entered electronically but that for all return visits at any site for that patient, that previous screened data will also be in front of us so we can say, I see that you're on dialysis or I see that you've had a pacemaker or what have you, and we can work from that for the future.

MR. KOPP: Keith Kopp. I can only approach this from the aftereffects that we've heard from people who are using the ferromagnetic detectors, and we've received many comments saying that we've picked up this, that, and the other. A large population of the units are used in the Veterans Administration, for example. Many of these people have implants and various other types of devices in them that frankly they don't remember and our elderly population.

My point is, is that ferromagnetic detectors are not designed to pick up objects inside the person, but there's nothing in the body that shields it. So I would encourage anybody who has these to not overlook and go through whatever your normal secondary screening process is, plain-film x-rays or whatever, because rarely if you repeat a test and it shows nothing on the outside, it's worth pursuing inside, and that's what many people do, although none of our devices are tested specifically for this type of application. One should not ignore a false positive in this regard.

DR. LAZEBNIK: Well, I can't comment on the specific efficacy of adding the language to the check in. I do think you raised an important point in terms of the terminology. So even the terms active and passive can be loaded, and should we move to a mode where such a question is asked at check in? It's very important that both the industry and the clinical side both agree on the meaning of those terms, and that it is very carefully defined and those terms are well understood.

As the technology for implants becomes more complex, you can certainly imagine that our definition of active today, for instance, is something that contains a battery, may change. Similarly, the term passive, you know, although sometimes it is associated with things like just very simple orthopedic implants, could also change. So I think importantly whether or not we choose to modify this language, standardization of terminology is key.

MS. RIVI: Diana Rivi, FDA. Another clinical practice question. So I'm going to refrain, but again, we are going to be talking about implants at the workshop tomorrow as well.

MR. STANLEY: Charles Stanley. The wrong time to check the flaps is when the plane's already in the air.

I would say, though I know where the question can come from, I know in the IMRIS systems, there's a checklist that you work through prior to the magnet being introduced in the room, but remember that's because that's a whole different paradigm. The patient doesn't enter the room with the magnet in there. What you do is you work through that checklist, and then at that point, the magnet's introduced into the room.

MS. SAWYER: Yes, Anne Sawyer, Stanford. I agree with what Charles said.

MS. ROTH: Candi Roth again, Imaging Ed. I want to step back to --

DR. WOODS: One minute, Candi.

DR. CRUES: I'd just like to comment. I think all the screening has to be done before you get to level four. So maybe I misunderstood the question. The screening has to be done before the patient ever enters the scanner.

DR. WOODS: Okay. Thank you.

MS. ROTH: Sorry, John.

DR. WOODS: Go ahead.

MS. ROTH: I didn't mean to cut you off there. I want to step back to this notion of a nonmetallic patient transport system. So to my mind, we've got this plastic stretcher. So now we rely solely on the ferromagnetic detector. Like that bothers me a little bit that we would be more worried about moving to nonmetallic patient transport systems than improving our screening, our patient screening. Like I guess that's what freaks out many of the technologists for whom I represent at this meeting is that they're going to rely solely on this one -- automation worries me because then skills tend to go away. So just comments. Thank you.

DR. CRUES: Well, I'd just like to reiterate the comment that I made earlier today, that we have a lot of risks in MR scanning and especially outpatient centers which is what I'm involved with.

Ferromagnetic objects is not a major issue. We've all seen ferromagnetic objects, even small ones, in scanners. When they're there, you know they're there. They change the imaging, they change the signal to noise. So you have to go in and take them out, and even small ones often require quenching of the magnet. We had several of those incidents back in the 1980s. So I've been through that.

But that's not a significant problem in our company where we've done several millions of these scans over the last 10 years. This isn't a major issue in the outpatient area.

What I would like to just reiterate, which I think many of the speakers earlier today reiterated is that over and over again, the key point that everybody emphasized is the training of the technologists. It's really the technologists which are the focal point. It's the technologists that really have to be responsible for the screening because they are with the patient.

On the other hand, I think everybody makes errors. We're humans, and if you have a good fail-safe system that really operates with few false positives, that can always be beneficial especially if it's low cost. But I think you really need good cost efficacy analyses before you jump into that sort of requirement.

DR. FRICKE: Well, I don't think anybody's advocating a fully automated process. I think that a fully automated process -- actually I don't even like that term. I think you can build architectural barriers. We have stoplights at intersections that don't always work and accidents happen, but imagine if no intersection had a stoplight. Stoplights are automated. That doesn't mean traffic's fully automated. You have to know the rules of the road, and you have to do your own checks, check your equipment, have your car inspected in some states. It's the whole process that makes it safe.

So if we can make a big part of that process that in some cases is humanly impossible or improbable that humans alone would be able to detect it at 100 percent, then I think we win, and that's what we're looking for. The question is today everything's getting so complicated, what is the process?

And ACR guidelines do a wonderful job, I think, outlining the elements of that process. Today we need to perfect those, flesh them out, and at some point the real question is yes, and then who's the compliance officer.

MR. GILK: Candi, I'll add a couple of things. One is that I think ferromagnetic detection is a dismal failure when facilities buy it thinking that it is their safety net, that it is their catchall, that if they're --

MS. ROTH: I suspect that's where I'm coming from because there are people who do believe it. I'm not against them. I think they're fantastic, but I agree with what you're saying there.

MR. GILK: For facilities to buy it in lieu of, you know, a competent, qualified, aware and attentive technologist, they deserve what they are inevitably going to receive.

Now, for those facilities that utilize ferromagnetic detection as a feedback mechanism, as an opportunity to learn just how poorly patients actually conform to screening instructions, and how they, the technologists, can refine and improve screening protocols and identify where things are getting through, whether it's, you know, new fashions that have, you know, I don't know, aircraft cable in the hems of your jeans, or I'm making stuff up here, but our patient population, what they wear, what they carry, what they have in their pockets, what they have in their hair, is going to change, you know, next year to the next to the next. We have a dynamic screening condition, and having a feedback tool, having a device that helps to alert technologists to when patient compliance has broken down is a tremendous tool, and those who buy it in lieu of, you know, doing anything else in safety, God help them because I'm not going to, but those who utilize the products as a part of an integrated system, those are the folks who really see benefit from the new technology.

MR. GOODYEAR: Simon Goodyear, Metrasens. Back to Toby's words, and I totally agree with that, but just bring the three observations from the discussions we've had today, that the three topics that have come out strongly. One is the training. You say we need better training for this. The second is there are people that are saying that ferromagnetic detection systems can help out here, and this afternoon, Dr. Kanal talked about guidelines that can help increase safety.

The key thing for me that I hope comes out of today's discussion is actually the answer is we need to bring all those elements together. So the trainers need to train how to use effectively with ferromagnetic detection systems under the guidelines of the ACR, and it's only when you bring those three elements together that you get an increased level of safety, and that to me is the big message I'd say.

DR. KANAL: Candi, to reiterate what you I'm sure doubtless have seen has been going on the last few days on the LISTSERV, the question, playing devil's advocate, the question says I'm a little afraid of completely automating the process. So let's take my response to LISTSERV and reiterate it here.

There has been no one at this meeting that I've heard state that they wished to have this as a completely automated process, but your statement does indeed represent what so many technologists were stating, which I find fascinating because if I were a professional in the field, I wouldn't be afraid of something helping me do my job. I would be very thankful. So there seems to be some kind of defensive push, if I can be so bold and so frank, that the professionals called MR technologists, many seem to have found a hesitation as if someone's trying to replace them.

Not only is no one trying to replace you, it's a recognition that you're invaluable but that you're also human. You've made an interesting observation. I'm afraid of complete automation.

Candi, I'm afraid of humans. I'm afraid of human error, and I think that both of them have to be recognized as realistic fears.

MS. ROTH: The more that we stop doing things manually, the more mistakes we make, the less knowledge we have. I mean that's a fact. You've probably noticed the more automated the protocols have gotten, the less many technologists know about what they're setting, I mean what they're running. That's a fact. You know that. So essentially --

DR. KANAL: If you end up coming in on an approach with an airplane and you're coming in too steeply, an automated voice will come out, by the way, typically female with a British accent --

MS. ROTH: That's on your plane.

DR. KANAL:  -- and they say -- no, on my plane, the female comes out. (Laughter.) And the voice will say pull up, pull up. Now, I'm not going to sit there and get into an intellectual argument saying I'm a darn good pilot, I don't want you telling me what to do. I appreciate the assistance. Even if the rate of descent was acceptable, and it was a false alarm, I'd much rather have five false alarms rather than the one time I was a little preoccupied and didn't realize I am descending right into a tower.

So frankly, there's no conflict here. No one's expecting to replace the tech. Having or not having, if the ferromagnetic detector is broken that day, that technologist's job isn't the slightest bit easier or the slightest bit harder. It's unaffected. The only thing it does is gives us another safety net, and frankly I think we should all be appreciating it. If it would have been free, I would want it everywhere. It should be a discussion of the cost.

MR. KOPP: Keith Kopp. I hear where you're coming from, and I've run into this issue in a previous career I had with airbags. I was involved with the development of the airbag, and there was a big discussion at the time in the halls of Government here of should we mandate it? Should we have them at all? It will make drivers sloppier. They'll rely on the airbags to save them.

And indeed, you know, the thing that's interesting is there's enormous statistics on highway accidents, and the bottom line is, most people still realize they still had to do their job. They had to still watch that intersection, watch traffic lights and so on, but it did provide an additional tool in helping save lives, and that's really what ferromagnetic detectors are.

They're meant to be an additional thing you do and a supplement. No more than a supplement. No more than a tool that gives you information, and again maybe sometimes it'll be false information, but it's meant to say, hey, this patient said I didn't have anything. I have found something. It actually, our own experience has found it's made people more aware. You'd never want to automate this because there's so many other things you need to screen for and make judgments on. So this is just another tool in the toolkit, and it has to be used that way.

MS. SAWYER: I'd like to say, I think I know where Candi's coming from. This is Anne Sawyer. But I think part of my fear is not that I'm going to be replaced -- or I wish I could be replaced. I'd get a day off. But my fear is I've worked with a lot of technologists over 26 years, and let me tell you, if there's one thing technologists like to do is they like to find shortcuts.

MS. ROTH: You've got it.

MS. SAWYER: And so we're not talking about us. We're talking about the other technologists, and I think we all know there's bad doctors and there's good doctors and the same thing for MR technologists. There's good technologists. I see them in the room. And there's bad technologists. And I can tell you one thing for sure, they ain't here today because they're not interested in learning about this, and those are the ones that scare me.

I also know about human nature, and I'm afraid to use one of these because I'm afraid I will get lazy and get in a rush and then I will try subconsciously for it to do my job for me, and that's what I'm afraid of, and I think that's what I hear in the tech server, is that that's what they're afraid of. They're not afraid of losing their job. They're afraid of making a mistake or becoming reliant on a piece of equipment, and I really don't think you can compare it to airbags in a car because airbags are just sort of there, and at least for me, I forget they're there, but this is an actual tool that you pick up and you're using every day.

MR. STANLEY: I'm afraid of humans, too, just different ones. Not the technologists. I'm afraid of the administrators. I'm afraid of the people that don't have anything to do with MR who think they understand it, and they make decisions that affect our workflow every day. Of all those technologists that have been up on the podium today, 20 plus years, 15 years, remember when there was always two technologists at the scanner.

MS. ROTH: I remember when we had metal detectors.

MR. STANLEY: Do we have two technologists at the scanner today? Literally at lunchtime, I got a text message from a former employee of mine, who knew that I was going to be here today, and they were saying, okay, check this out. I work in an interventional magnet that's surrounded by an interventional suite. We always had two technologists there. They've made a decision to go to one technologist, and this person said, well, what if somebody codes? Well, you can run out in the hallway and get someone to come help you.

See it's people that are making decisions like that that are affecting us, and that's the issue that I have. I can't remember who it was, maybe it was Simon, who had said it, but whomever said it, talking about all three of the things coming together. I think if we mandated that all three of those things came together and stuck to that, I absolutely 100 percent agree that ferromagnetic metal detectors belong in our suite, belong as part of our screening practice, but the problem is, is if we don't do and mandate all three of those things, then what's going to happen is, is the people who make the decisions about the technologist daily workflow are going to almost force them to rely more on those metal detectors because they're going to be everything.

And as Anne said, it's human nature. I've got five things to do. I'm just going to wait and listen for the bell to ring. That's not the techs in the room today. The techs in the room today are all freaks, and we're going to do it right every time.

But as we heard earlier today, there's 40,000 MR technologists in the U.S., and I guarantee you there's not 40,000 techs in this room today that are interested in doing it right.

MS. ROTH: Thank you all.

DR. WOODS: Thank you. I'd like to, for the assistance of the transcription, to remind everyone including the questioners to please give your name before you speak, and I'd like to I guess take the Moderator's prerogative to ask a question of the roundtable.

FDA does not currently regulate ferromagnetic detectors, and do you see a need for FDA to regulate ferromagnetic detectors as medical devices?

DR. CRUES: I'll defer on that one. John Crues.

MR. FRESE: Georg Frese. While we have discussion within our MT 40 meeting, not only one time but several times, whether to stipulate a requirement to introduce metal detectors, but we're of the opinion that we are responsible for the MR system and not for the operation of the MR system.

So what we are recommending is to think about the possibility to improve the result of screening and to control the access area by thinking about such equipment. However, I think there's an agreement we will not make a requirement to equip a MR system with a metal detector. That is speaking for ISE, and if national organizations will set such requirements, okay.

DR. WOODS: I guess maybe to add one more piece of information. Terry Woods. A medical device is something that you need to provide some evidence that it performs its designated task safely and effectively, and I've heard some comments about the utility of ferromagnetic detectors. So that might be one motivation for asking the FDA to regulate them as medical devices. Again, I'm not saying this is a good or bad idea, but just curious to hear your opinions.

DR. FRICKE: Stanley Fricke. Thank you, Terry, for the question because in the training that I have to do, a lot of times I get the question, well, I've used a metal detector. It doesn't really seem to work. It picks up everything. It doesn't pick up anything. Yeah, it's because they all work differently, the wand, the arched wave, the different vendors, and I think it needs to be -- there's time for some regulation in the detectors themselves. The main reason is that if we're going to rely on them to any extent whatsoever, in the medical field, and I believe that it is to some extent recommended in the ACR Guidelines or at least suggested, and I think it also depends on which year of the guidelines you read. We're looking at 2007 and hopefully 2011, and I don't know what that says.

But this issue is, if I have a metal detector that I'm in charge of setting the sensitivity for and somebody else can change it tomorrow, that the maintenance crew in my hospital can change the environment around because of maintenance issues or things like that, we now have a questionably working medical device, and with no regulations on it whatsoever, some institutions will do an excellent job of controlling their own detectors and making sure and ensuring that they work properly, but I don't know too many institutions, and I'll take this from a different radiological field, that calibrate their own Geiger counters and just sort change the settings on them at will depending on what they want to detect. It's not a Geiger counter, but it is detecting metal, and I think to straighten out this industry and get it on track, I think you need to tighten up what's going on with these metal detectors.

MR. GILK: This is Toby Gilk with Mednovus. I think there are two components sort of implicit in your question. One of them is sort of a basic degree of performance testing and, you know, effectiveness of the tool, and then the second one is approval as a medical device. I'll let the other manufacturers of ferromagnetic detection systems speak for themselves, but to my knowledge, as of the current time, not one of us proposes using these devices for things internal to the body of the patient. We are looking for things in the clothing, in the pockets, you know, concealed in their hair, in their footwear, what have you.

My understanding is that screening for things external to the patient sort of is outside the traditional definition of a medical device. I am in no way opposed to the idea of some kind of performance criteria or performance test or that we establish some kind of benchmark in terms of what these items are capable of doing, and to that end, some of that work has already been done by ECRI Institute who has done a study. The study is a year old at this point, but it was all of the ferromagnetic detection products on the market as of a year ago.

So I think in terms of performance testing or performance assurances, some sort of minimum standards, much of that work currently exists.

Now, if we are going to say that as an element of MR pre-screening, ferromagnetic detection systems should be subject to FDA review and approval, I think that also begs the question, well, what about the conventional metal detectors that are provided by or sold by the AMR system manufacturers, and do we want to only sell conventional metal detectors that are similarly reviewed and approved by FDA for MR pre-screening?

We don't have an answer for that. I think from the quality assurance or the functionality assurance of the ferromagnetic detectors, that information currently exists and is available. From the regulatory, how are we going to classify this. I think we establish a pretty significant double standard if we only apply this to ferromagnetic detection and don't apply it to other pre-screening detection tools that are currently marketed.

MR. GOODYEAR: Simon Goodyear from Metrasens. If through regulation you can achieve what I was asking in my last comment, about bringing together better training, the guidelines and the technology and, of course, I'll be very supportive of that. If it does go through to regulation, then I would say that prerequisite levels of performance need to be addressed.

So having come from an industry in security metal detectors, where there are specific standards for airline metal detection, I propose that there should be in this industry similar standards for ferromagnetic detection systems for that regulation to work because if you don't do that, then you're going to get other systems coming in which may not be achieving what you want it to do.

DR. KANAL: Manny Kanal. I'd like to see TSA responsible for all MR safety in the United States. While I don't feel strongly about the question about regulation, I think we do need to keep in mind what Candi was saying before, is we don't intend, to the best of my knowledge, no one in this room has proposed intended on relying on this data. It's an adjunct only.

As such, if we go back to what I believe I understand is the charge of the FDA, it would be considered at best a low risk device, and as such, it should be exempt, and I see no reason -- whether or not I even have one, the technologist job is unchanged. So from a nonemotional, purely definition point of view, if it is, in fact, a low risk device, that the information could only help me, but not hurt me, I don't see why the FDA should be at this stage considering it.

MR. KOPP: Keith Kopp. It is useful to have some performance criteria so that a customer would know what they're buying, but our goal here at this meeting is, what can we do as a collection of different interests to improve MR safety? And in truth, a ferromagnetic detector, be it FDA approved or otherwise, does no useful function, performs no useful function if the tool is not used properly and is not part of what several earlier speakers have already stated, is that it really -- its Achilles heal is not so much in what it detects. All of the products available now do detect reasonable sized objects.

The real issue is how do you implement this intelligently so it doesn't degrade the quality of screening but actually truly adds another layer of Swiss cheese in the screening process that helps it, and I think the other issues are much more important to the overall process.

Relating to FDA approval, in discussions that we've had with the other vendors, we all know that our devices, as I mentioned a little earlier, can detect implants. We don't advertise, promote, or any other way do it, and the reason is, the cost of ever going through the approval process is so prohibitive that we just keep it off the table as a discussion. And what happens is, does this help the overall safety of the screening process or does it hurt it?

The say it is now, it would be nice if there were some standards. NIST, the National Institute of Science and Technology, in this country sets up, as Simon alluded to, a procedure for here is the criteria we use to meet at least a minimal requirement, and so that would be useful. But to make it a medical device, I don't think it fits in the overall FDA, and I don't think it adds to the fundamental goal of this meeting of improving MR safety.

DR. LAZEBNIK: Roee Lazebnik. I don't have a specific comment on that question.

MS. RIVI: Diana Rivi, FDA. I just wanted to thank everyone for their feedback on that question because that's exactly the type of feedback we wanted to hear. If ferromagnetic detectors did become a regulated medical device, which we don't even know at this point, there would be a lot of work that would have to ensue about its classification. But for now, we just wanted to hear your feedback on that question.

MS. SAWYER: Anne Sawyer. And in my humble opinion, not knowing all the ins and outs of the FDA, I would say, no, it shouldn't be regulated by the FDA.

MR. STANLEY: Charles Stanley, no comments.

MS. HOOD: Maureen Hood from Bethesda, Maryland. I'm a MR technologist, nurse, and scientist. My comment is going back to where Candi was coming from. I want to point out to the FDA especially that there are certain MR vendors that are now kind of marketing their products to niche types of clinics or small or remote areas, saying their equipment is so easy to use now, you don't need to hire a MR technologist anymore. And the scary thing about it is they're also now saying all you need to do is buy a ferromagnetic detection system and hire one of the people in your clinic or use one of your people in the existing clinic to push the buttons on the scanner, which I think is very scary, and I'd like to hear some comments especially from the FDA.

DR. CRUES: John Crues from Radnet. I think there's some real issues here especially concerning the very specialty scanners like some of the extremity scanners where you basically have everything preprogrammed and you literally just push a button. I think invariably what we've found, having worked with a lot of those, is we see a lot of variation in image quality, and I really believe over and over, it comes down to the quality of the technologist. There really is a lot that the technologist has to do in terms of image quality. If they don't get a good scan for whatever reason, they need to recognize that, and I think that there is a real risk as people try to cut more and more costs in our current system, that maintaining quality control is a major issue, and one of the major costs of maintaining quality control right now is manpower, and I think it's got to be really emphasized that people are the primary source of safety in MRI.

I'm not opposed to devices like ferromagnetic detectors if there's a real risk that's been shown, and they're really shown to be cost effective to supplement that, but I think we're all agreeing and it really has to come out in whatever document comes from this meeting, that the first line of defense has to be the people involved, and there cannot be shortsightedness in decreasing the level of people who are involved in protecting patient safety.

MR. FRESE: Georg Frese. I think that is our -- target to make scanning of patients as easy as possible and to improve the workload, but that doesn't mean that a MR tech should be replaced by someone else. Both together will improve workflow and quality of MR examinations. The effort of the MR tech and, of course, the development of MR systems in terms of using.

DR. FRICKE: Well, I think that the market might be going in one direction, but the insurance companies are going in another direction because they don't like rescans any more than anybody else.

So I believe that part of the ACR accreditation process, that asks several questions about techs and doctors and certificates, et cetera, will also be part of the process that won't allow this to happen except for in the most obvious cases where it financially is doable.

Do we want to automate everything and take the people out? I don't think so. It's not going to work.

Is there a world where you can have your big toe scanned at Wal-Mart instead of having anybody around to take a look at it and get an automatic diagnosis? Yeah. If it says the toe's on the wrong foot, you're probably going to go to your doctor anyway.

MR. GILK: This is Toby Gilk, Mednovus. I think that one of the things, and this is sort of an undercurrent throughout many of the presentations and much of the conversation, what we as an industry are experiencing tremendous contractions in terms of reimbursement rates for technical revenue in particular, and MR systems aren't really getting any less expensive.

So the only options that we have are to either try and squeeze these variable costs and/or increase the throughput for our system, and I don't see either of those two directional shifts changing anytime in the near future. Reimbursement is just going to continue to get squeezed, and we're going to have to figure out over and over again how to image more patients while spending less money on overhead, and I think one of the overall objectives of this meeting is to, you know, let's stake out safety, patient safety, as one of the fundamental, you know, we will not drop below this level aspects of providing imaging, in providing MR services, and I think when we come to the position of, you know, it's either this tech or it's that tool, in some ways that says to me that we've passed the point at which we already made some pretty poor decisions to let ourselves get to that point.

I don't have an easy answer for you. I know that these are the kinds of artificial battles that are set up with the administration financiers of MR services. I think we just simply need to say if you want to provide this service, there's a minimum acceptable level of safety, and this is necessary to protect not only the patients but also the technologists who are disproportionately represented and the folks who are injured in MR accidents outside of the burn. So this is to protect the patients. This is to protect the technologists, and ultimately safety is also in place to protect a $1.5 million or $2 million scanner and the useful life of it. Great scanners have lousy throughput when there is a bed stuck to the side of them.

MR. GOODYEAR: Simon Goodyear, Metrasens. I've visited many hundreds of facilities in Europe and in the U.S., and there's one recurring theme to me having spoken to all the techs and superintendents around the world, and with or without FM being in place. The level of safety is always directly linked to the quality of the tech that's in charge of the area, and to me that answers your question from my viewpoint.

DR. KANAL: It's actually a pleasure to put faces on the names we've been seeing for so many years. Maureen, you actually bring up two fascinating points, and I'm afraid I'm going to differ a little bit on at least one of my answers. You did bring up the point that I think that it's fascinating that the FDA needs to hear, and we spoke about this during the break. So thank you for the opportunity to bring this up.

We have systems being marketed, I'll try to be as politically as improper as possible, to cardiologists and orthopedic surgeons and others who frankly don't have two functioning neurons to rub together who will be asked to implement not only the running of these sites, but the safe implementation of these devices.

The FDA needs to understand loud and clear that this is potentially a catastrophe waiting to happen with the perfect storm of less and less funding being allocated towards the appropriate employees, in this case, the technologists, let alone their continued educational funding as well. So recognize that more and more we're having individuals that are not specialized in this area, not trained, not educated.

In the University of Pittsburgh Medical Center, the cardiologists obtained, independent from the Department of Radiology, their own magnet run by them, organized by them, and frankly operated by them, and it took quite a few weeks before the first buffing machine entered their system, and then the hospital calls to radiology and say, why did this happen? And our response is because of you permitting it to function in that fashion, not under the auspices of those who know what they're doing. Please don't get upset at me. You don't have to be a radiologist. You have to know what you're doing.

So the first issue that you raised, I completely agree with, and that is that it is being marketed as a push button operation, press here, dollar bills come out here, and you're only laughing because I'm putting in an interesting fashion, but you all recognize that that is exactly how it is being marketed, and conceptually the FDA needs to see that problems that we're seeing now, if you think we had incidents in the past, you watch what happens as more and more of these get in place.

And now the time to disagree. Perhaps I'll be the only one in the room who feels this way, but I do not believe it's appropriate to paint all magnetic resonance the same. I actually am a believer in Wal-Mart's approach, and I do think that there will be and have no reason not to be specifically designed extremity scanners where you can't even find or measure the 5 gauss line outside the field, and we take patients with aneurysm clips and pacemakers, stick their legs right in there, if you can restrict the power deposition, I don't actually see that there's any problem whatsoever to having an outpatient setting where they push a button, don't know what they're doing and prospectively define, you can't touch TRT, flip angle TI. It's all automated. It doesn't bother me in the least.

So just the fact that it's conceptually magnetic resonance doesn't make it necessarily dangerous or potentially dangerous, and I'd like to see the FDA intelligently apply their regulatory authority accordingly.

MS. HOOD: I want to comment on that real quick. I agree with that scenario. What I'm talking about are these 1.5T scanners that are out there and even the ACR allows some of the loopholes for MR technologist personnel. You don't have a technologist for these scanners, and some of these 1.5 and 3T scanners are going in remote areas in the United States where maybe they can't hire an MR tech out there, and the radiologist isn't even on site. So this is getting a little scary.

DR. KANAL: If I may, the American College of Radiology Safety Committee has a magnificent group of individuals on that committee, and it's intentionally chosen to have no more than one represent each of their niche areas. So there's a MR safety person or a MR physicist or somebody from NEMA or somebody that's a MR tech, not more than one. However, the way it has been run until now has been that every decision has to be unanimous or it's not included in the publication. I think I'll say that again, unanimous for 15 human beings.

MS. HOOD: I think the ACR needs to relook at the MR breast imaging.

DR. KANAL: Well, that's a different organization. I'm talking now only about the Safety Committee. Having said that, that means if anyone on the Safety Committee feels that it is not in my best interest or those that I represent to have a MR sub-certified technologist running the scan, I will not vote to have that as part of the final document. And if you'd just use your own imagination, there are several of us on that committee that were not able to vote that way despite the fact that it has been brought up multiple times. And for reasons of conflict of interest and what have you, it might be appropriate to have a regulatory agency like the FDA look into those matters and perhaps put some sort of standardization as to what's required.

MR. KOPP: We hear much the same thing. This is Keith Kopp. We see the downward pressure on reimbursements and throughput and so on, and in situations like this that I've been in, in the past, and in this one, sometimes you have to force safety regulations to be in place, a minimum standard which is what's been talked about a lot today.

There is a start in this direction, that is actually effective in several months. CMS, the Medicare-Medicaid people, through Health and Human Services, created a thing called MIPPA, which is a reimbursement arrangement which requires outpatient centers to meet a minimum standard by one of the three accrediting organizations, and part of that accreditation is image quality, but then another equally important part called out is assuring the safety of both personnel and patients.

So if this truly is enforced, it's a very powerful tool to get what I call the bottom feeders to come up to standard in that they will not get reimbursed by CMS, which is the main cashbook stream of Medicare, unless they comply with this minimum accreditation process. The hope is the accrediting organizations will indeed enforce the safety part of it. It is part of the legislation, but there's competition among these three to get as many people signed up as possible, but it does show some hope of having some teeth in a regulation that actually affects on the other side, the bottom line, for not being safe.

DR. LAZEBNIK: Roee Lazebnik. There are two issues that I believe you brought up, the automation related to safety and automation related workflow. Now, there's no question that there's a variety of automated safety mechanisms that are built into the MR scanner, and hopefully that'll continue to evolve. Simultaneously, the MR scanner is a very complex piece of equipment, and certainly I would never advocate that without a skilled, trained, and very knowledgeable operator, the MR equipment should be used. In fact, we promote very specific guidelines for safe MR operation in our documentation, and it's certainly in alliance with the technologists, with the radiologists, and quite a lot of human effort is required to operate this system appropriately.

Now, on the workflow end, I think it's very legitimate for industry to continue to innovate and continue to hone the complexity, the increasing complexity of the MR scan so that, for instance, the time required for an exam doesn't balloon from 30 minutes to 10 hours. We've seen phenomenal and fantastic evolution in terms of what a MRI exam provide, and I'm certain that we will continue to be on that trajectory.

The increased information that's associated with a MR exam needs to be processed with increasingly complex and, you know, sometimes automated tools, whether it's to handle the number of pulse sequences or the amount of information contained in one sequence, visualization, post-processing. I'm sure you're very familiar with this. So that type of automation, I would say, is tremendously useful to the clinical world because again it does let you extract more information from the same MR exam in a shorter period of time, and that's not just a market demand. I think that's helpful clinically. Certainly that should not come any way into conflict with maintaining a very high level of safety associated with the MR exam. So automation is not a substitute for safety. It should always compliment safety.

MS. RIVI: Diana Rivi, FDA. So what I'm going to respond to are the marketing claims that you had spoken about, and I just wanted to let you know at CDRH, our Office of Compliance has a group that looks at the advertising materials that manufacturers make, and they determine if false safety claims are being made on their products. This doesn't seem to fit in my mind as I understood your comment, but I'm happy to talk to you afterwards and get more specifics and if we need to explore that.

MS. SAWYER: Anne Sawyer. I really have to agree after being in this field for so many years, that there is a trend by vendors, maybe not all vendors of MR systems, but certainly some, to make it a one size fits all, and it only takes one person to run the system and that also that person doesn't have to be that well educated. And this may not be something that you've seen in brochures, but it certainly is something that's communicated by the salespeople because, you know, everybody wants to spend less money, and especially they want to spend less money on personnel.

And I can tell you the more knowledgeable you become and the longer that you're around, you become more expensive and that you're not the one they want to hire. They want to hire somebody who is just going to come in there and push the buttons. And if you don't possess education, if you don't possess the knowledge of even basic physics and principles and how things work on the system, there isn't a chance that you're ever going to understand MR safety and how to maintain it, and that may be just my opinion, but if you're not knowledgeable, you cannot maintain MR safety.

MR. STANLEY: Charles Stanley, SMRT. The SMRT supports only licensed, credentialed technologists performing MRI scans, and this actually articulates some of the fear around talking about the metal detectors earlier.

I'll just divert to that for just a moment, and that is that human nature again replaces things. We've heard several people on this panel who felt that ACR accreditation would make sure that we had qualified technologists doing these exams. Unfortunately, that's not correct. If you look at their own document, it simply says there has to be at least one person on there. I know sites, not too far from here, that have one part-time MR certified technologist on their staff who may work one Saturday a month, and guess what? That's the only credentialed technologist that they have working for them, and they meet the guidelines to get the sticker. And so my point is, is getting the sticker didn't make sure that the practice was there.

So it's the same thing with the metal detectors. Putting the metal detector in doesn't make sure that the screening processes are in place, but again, the SMRT supports licensed, credentialed technologists performing MRI scans.

DR. FRICKE: Stanley Fricke from Children's National. I'd like to follow up on one of the statements I made earlier. I think that part of the issue lies in our own hands for these automation questions. I find it very frustrating these days to get good image quality off of some of the newest and latest imaging devices because they've taken the control out of our hands, and they've put it into research only. You understand what I'm saying.

Well, if we don't complain about that, we're going to end up with a push here down button that's called the Ph.D. machine, and the machine will be the doctor, and that's wrong.

The issue in automation, whether it's detection or image quality, which both I am convinced reflect on patient safety and well-being as well as all the workers in the environment, is in the control of people that decide it's not working the way they need it to work, whether it's a metal detector or your MR scanner. If it's not doing the job, write it down. Make a complaint. Voice it.

I'm not real happy with all the metal detectors I have. I've been extremely active in trying to change that technology in speaking to the vendors and also the FDA. I'm not equally as happy with the equipment that I have right now, but if I don't have the group of people that I work with excited about changing those things as well, the voice is going to be one in a million, and nobody's going to listen.

DR. VAN DEN BRINK: Johan van den Brink, Philips. I would like to come back to the electronic health records -- through for a second. The question is that let's say implants prescreening would be documented in the electronic health records. I would like to go to another condition that might go unnoticed for a while, pregnancy. This is being reported in the risk interface, and comes to the MR system, and the question I have is in our instructions for use, we tell you to not exceed first level control modes, normal modes I mean, not go into first level control modes in case of pregnancy. Now, early pregnancy is probably not told to you or hard to notice, and in the risk interface it says the known pregnancy state. How would you feel if the MR system which prevents you from going into first level control modes, if the risk interface was going to tell you the known pregnancy state, and that then also relates to the electronic health records type of thing of prescreening. Suppose there was information on an implant. How would you feel that the MR system would prevent you from doing things because there isn't enough information from the electronic health record coming to you?

DR. CRUES: John Crues from Radnet. I really like the idea of electronic health records, and I like the idea with integrating into all the functionality in healthcare. What you're asking right now I think is a very interesting, long-term goal, but I think right now we don't have the sophistication in the systems to have it directly interface with the scanner to tell the techs what to do or what not to do.

UNIDENTIFIED SPEAKER: The risk interface does provide you with that information for sure.

DR. CRUES: I guess I would have to defer on this. I would have to have more experience with it. At the current time, I would really want the technologist to make that decision based up on our current models, but if I was shown that it was really effective and accurate, I would eventually be probably willing to change that decision.

MR. FRESE: I think this is a responsible user -- opportunity to summarize the history of the scanning of pregnant patients. In the IEC standard, up to edition 2, there was no restriction at all for the examination of pregnant people, women, and I think there was a warning to be cautious and have risk/benefit assessment, especially for the first trimester. And then with edition 3, we have introduced the requirement that pregnant women should only be scanned in the normal mode.

DR. FRICKE: Well, this is clearly a topic that nobody's an expert on at the moment. There's very little information about what is doable and what isn't doable. We have indications on both sides of the line of what would be a good idea to do. If today my scanner inhibited me from scanning above level one, I'd be pretty upset. I'd get another scanner. We scan pregnant women every day. Babies move. You think it's hard to scan a heart, you try to scan a baby. If you're trying to scan the baby's heart, that's even more of a challenge. You need the speed. You need to go. The question is, is it clinically indicated?

I would also --

UNIDENTIFIED SPEAKER: No, the question is, is it safe?

DR. KANAL: Could you say that again please?

UNIDENTIFIED SPEAKER: I would say the question is, is it safe?

DR. FRICKE: The question in safety is always a question. Also is it clinically indicated? That's always been the case. You can't really debate that. I mean if you're going to burn something out, that's just a problem, but the question here is what is the need? What is the outcome and what's the need?

So putting the brakes on in a general mode is a good idea. I don't know if I'd like my car to put the brakes on in the middle of the highway just because it probably needed to maybe slow me down a little bit, but if I need to override it, I need to override it.

And so the answer given about having the technologist make the decision, I think this needs to be an informed decision that the doctor also needs to be involved with and maybe other people in the case.

At Children's National Medical Center, we have the involvement of two medical doctors, two researchers that are very much aware of the MR physics and heating issues, and the technologists. It's not a simple question right now, and I think the real problem is, is that the actual knowledge of what is the effect on the fetus, there is some, but it is not a complete and informed decision. Right now I think it's a little premature.

DR. KANAL: We're approaching it from the point of view of the physician, and the answer to the question is relatively clearly defined. Leaving all emotions and ethics and morals out of it, it's a risk/benefit question, and I would actually personally appreciate the checklist, meaning it says this patient has a pacemaker. Are you sure you wish to proceed? But it's not even an override. It's a yes or no, and we're done. There's no such thing as taking away that choice unless it's overtly going to clearly damage the patient, but the way society will look at it is a risk/benefit assessment and as for the decision, we have tremendous legal precedent by now that the decision to scan or not scan a patient is considered, in every state that I'm aware of in which this has come up, a patient care issue, and it is the considered the practice of medicine.

And so I would like to reiterate exquisitely strong that although you may hear us using terminology such as the technologists making a decision, no, of course, not. The technologist isn't licensed to practice medicine, and I don't think any of us would condone that the technologist would make that decision. The technologist is enforcing or acting out upon the orders of the physician.

MR. KOPP: The only comment I would have is that, as has been stated earlier, this checklist appeared on the console, for example, of the MR. The assumption is the patient is already in the bore, and that in many cases is way too late to even deal with a checklist. Checklists are very useful, very important as with other safety protocols, but the timing is everything. In many cases, the timing has to be before the patient goes into Zone 4.

DR. LAZEBNIK: Roee Lazebnik. The question that you asked I think raises the important issue of how many limitations should a vendor build into the equipment, and my answer to that would be it's very difficult to build limitations unless there's an agreed upon standard.

So, for instance, we certainly limit the maximum SAR value or the maximum RF energy that can be deposited in a patient because that is a well agreed upon standard. It's very difficult, if not impossible, in many cases, to anticipate the exact clinical situation and all the complexities of using the MR scanner with a particular patient in a particular clinical situation so that the general philosophy has been we trusted the clinician as a knowledgeable user, and we provide them with quite a few options to scan their patients.

So the most specific answer I can give you is we need well thought out and agreed upon standards in order to build these type of limitations into the system. It should not be up to one specific vendor or a small group necessarily to say this is a good idea to build this limitation here or there. It needs to be something that is well agreed upon between the clinical community and the vendor community.

UNIDENTIFIED SPEAKER:  -- speaking. As -- pointed out, indeed we do have these consensus standards in place. So in this case, the recommendation or even the agreement is there. I do agree with you that, of course, we should agree as an industry together. The question here I had was more towards -- information.

MS. SAWYER: Anne Sawyer. If the vendors want to include something to help make MR examinations more safe, what they should include, what they used to many, many years ago, is automatic included in the price of the scanner, is education for the technologists, on site applications and a didactic course, and not have it be an option because when the administration is going down the checklist of what they're going to buy to go with their scanner and they see tech education, the great majority of the time, that's going to go unchecked. So if you want to help make MR examinations more safe, the vendors who sell the MR systems need to include ongoing regular education of the technologists as part of the purchase price of the scanner.

MR. STANLEY: Amen, Anne. But I would just say that it's too early without, with a lack of standards about what those clinical decisions would be based on, taking that clinical decision more completely out of the hands of the technologists who is a representative of the M.D., I think at this time is too early, and so I would not be in favor of that.

DR. WOODS: Just as a matter of housekeeping, I'm looking at the time. I think with our time remaining, we probably have time for the questioners that are standing now. No, don't, don't leave. We have time for your questions, but I propose that so that you don't have to continue to stand there. We go in the order that I think I saw you stand, two on this side, Jana in the center, and then the two people on this side.

UNIDENTIFIED SPEAKER: (Off microphone.)

DR. WOODS: Frank, you can have the last word of the day.

So again, questioners, you can continue to stand if you'd like to or find a seat, but we'll go through the rest of you, and I think that will probably take the rest of the time we have for this afternoon.

MS. SHOATS: My name is Cheryl Shoats, and I am the MRI manager at Johns Hopkins Hospital. This is a MRI safety question, of course, but I'm really looking for your opinion or what is your practice at your facility.

So we have 12 scanners currently, and in April of 2012, we'll be acquiring 7 more scanners. So we will have a fleet of 19 scanners, and of those 7 additional scanners, 4 of them are going to be in very challenging areas of us to zone out, but what we're doing with our MRI policies is we're updating what we currently have, and we're also putting our best practices into policies, and the question that I have revolves around screening patients, MRI screening of patients.

So if you have an unconscious patient, because we have a scanner that's going into the ED, is actually going to be embedded in the ED -- It's a 1.5 scanner. If you have an unconscious patient or a patient that's alert and not oriented, that's unidentifiable, how do you go about screening those patients or in other words, how do you go about -- who's responsible for signing that screening form?

DR. FRICKE: Well, at least from what I understand in the ACR guidelines, first of all you need two doctors to agree that absolutely MR is indicated. It's the only thing that can be useful or it is absolutely needed.

Secondly, the patient needs to be basically made with no clothes on so that you can try to identify scars or entry points for any type of metal that may have come into their body at any point, and if such a scar or site is identified, either two points of x-rays or a CT is to be taken in that area. But again, the scanning under these conditions has to be very well justified.

DR. KANAL: Manny Kanal. For the past three and a half years, I have specialized in performing and only performing emergent neuroradiology, and further it's been divided only into CT and MR emergently, and so the first thing that was said, I completely agree with what you were saying. The first point you brought up is critical. In retrospect, there seem to be extremely few indications for true emergent MR.

For the unconscious patient for whom it is required, the ACR guidelines did list specific indications as to how that would be pursued especially for patients in whom you're saying who would be the one responsible for signing the form? Well, I don't understand the question. Whoever the physician is that has said I'm going to accept this patient for scanning in the United States. One of the most common problems we find when a problem occurs is that this physician says, I'm not responsible for the patient, the technologist is, and the technologist is hired by the hospital. And the hospital says, but you're the radiologist, and the radiologist said, but I'm only responsible for the interpretation or to make the third level of complexity, I didn't order the contrast. The gastroenterologist did. And everything I've said to you now are direct quotes and they're constantly repeated in lawsuits time and time and time again.

So the answer to the question, the courts have always found it an extremely easy answer. Who is the physician executing the study? It's not the one who ordered it. You're not a physician. I guess it's Manny. So 100 percent of the time the radiologist has been held responsible. Even though they said, I'm only here to interpret the study, the law doesn't seem to agree with them.

So that leaves us with an interesting situation which is where, unless there's a contract which specifically states I'm only interpreting and not doing anything else, and I have since been informed by attorneys that even that's not necessarily going to cover the radiologist, the implicit contract between the patient and that physician, the site that's executing the study, is that there is somebody watching over the best interest of that patient, and that has to be a physician.

So I would just point that out, and in closing on this response is that (a) I believe it's the radiologist who, not always alone, you can also find other people responsible, but certainly the radiologist will be found responsible for the safe execution of the study, not just its interpretation. Number two, at the RSNA this year, there's a special session that will be provided specifically on medical legal implications of MR safety, and for the first time in my 33 years in medicine, I invited a plaintiff's medical malpractice attorney to share the stage with me, and so we'll be addressing those issues pretty directly.

MS. SHOATS: Thank you.

DR. EDMONSON: Heidi Edmonson from Mayo Clinic.

MR. STANLEY: Sorry. I just wanted to comment. First of all, Manny, how are you going to get that lawyer past 100,000 radiologists into the building? That's going to be a feat. I would like to say --

DR. KANAL: It's a safety issue.

MR. STANLEY: I'm going to try my Bill Faulkner impression one more time, and that is that in this case, the technologist is smartest, and they play the dumbest. They need to absolutely take themselves out of that equation. It's the radiologist who makes that decision, and we follow his orders or her orders.

DR. FRICKE: I would recommend one more thing. Since you have planned on doing just ED imaging with this or primarily, that you come up with a specific screening form that has the dual consent of your doctors that are requesting it in some sense, and I believe a process for a radiologist that understands what they're getting themselves into; sometimes at 2:00 in the morning, you're not always going to have your top MR radiologist there. It also begs the question of the medical safety officer and the availability of this person, somebody that's extremely knowledgeable.

So I think you'd probably internally want to look at process of liabilities and issues like that, and then define the form, that people have to sign before it goes to imaging, maybe even right on a specific screen form for that type of screening situation.

MR. STANLEY: This is Charles. Actually we did something very similar to that at the University of Virginia where if it went to the level that the radiologist needed to have informed consent, then guess what? Signing the screening form wasn't good enough. You filled out an informed consent form, and that actually would stop a number of MRIs. You know, it's one thing to just say sign the bottom of this form. It's another thing to say, no, it's got to be an informed consent, and obviously you know what an informed consent form looks like. We do it on, you know, most of our interventional procedures, but that is very effective in screening out those nonemergent MRs.

DR. EDMONSON: Hi, Heidi Edmonson from Mayo Clinic. I'd like to revisit the easy button MRI, and just I think most of us here are from highly populated areas, and something that I struggle with intellectually I guess is just for these remote areas, having a safe practice and maybe limiting the services they provide to make sure that they're providing safe, well-employed services versus potentially keeping that patient population from having access to the diagnostic modality. And so I'd like to hear what people's thoughts are on making sure that patients have access to MRI.

So the Joint Commission has a special accreditation program for critical access hospitals where there's potentially small hospitals that don't have all of the services but are 35 miles away from, you know, a larger hospital. Would there be a role potentially for an accreditation of a critical access MRI scanner?

DR. CRUES: John Crues from Radnet. I think when it comes to safety, if you're going to have patients go to a MR scanner, then I think any MR scanner, you need to have the same standard safety precautions as anywhere else, and I don't see where that would be a difficult thing to do even in a rural area. If you're going to have a technologist with a scanner, they needed to be trained just as well as technologists elsewhere, especially in the area of safety, and all of the other safety criteria I think have to be the same as if they were in an urban area.

DR. FRICKE: I think that special circumstances always require a certain amount of customization, although you have to know where to make the cuts. The automotive industry has been trying to cut costs here, there, and everywhere, and I think Toyota with its brakes probably went a little bit too far. But I think you're bringing up a good point about special circumstances.

In terms of the technologists, I'm going to assume if you have a MRI, you absolutely must have at least an x-ray technology, so maybe cross-training. In terms of the entire site, it really depends. It's too complicated of an issue talking about 1.5 or 3T magnet, the same as in the clinic. It's a multimillion dollar investment. I think the personnel really shouldn't be cut out of that budget. Cross-training to fit the need of throughput which might be reduced maybe, but why else would you cut somebody.

DR. EDMONSON: And I'm not necessarily talking about cutting the MRI technologists. I guess I'm saying that with the ACR requirements as they sit, they're the same requirements to get a site accredited for an MRI mobile unit that maybe only visits a particular site three hours a week, as they are for major medical center that has, you know, a scanner running 24/7. And so it's very difficult for these small sites to participate in a quality program, and since they might be associated with that critical access hospital, would not be affected by MIPPA.

DR. KANAL: Manny Kanal. I think we're dealing with two different issues. When you're talking about the site accreditation committee, you're talking about, and for today only, talking about the quality of the equipment, the quality of the individuals, in a sense, that's the site. It is not yet today married to the safety requirements. I have to make sure you understand that because it's not necessarily what everybody would like, but it is the fact today, the requirements for ACR site accreditation do not mandate that they satisfy any safety criteria that the Safety Committee may have put out. Those are distinct. So I'd rather not discuss the site quality at this date, but only maintain a response to your question as far as the safety is concerned.

And for that, I think I'm going to slightly disagree with John. I think we're going to actually say the same, from the point of view of safety it should be the same, but there is one area in which I think, I'll either disagree or that -- well, let's just see what he feels. In the Safety Committee there is a separate individual I mentioned to you about before. That's the representation, Dr. Bell, the representation from the American Society of Anesthesiologists. They are way ahead of us in this area, and if it's critical access that you're referring to, they often require sedation or anesthesia, and special monitoring, and for those kinds of patients, no, we don't believe all sites are the same. Anesthesiology has, already way ahead of us, started to sub-certify different sites based on what level of anesthesia capability they're able to implement in an emergent or nonemergent fashion. We are attempting to incorporate that into the ACR Guidelines, although I can tell you now that it's not yet accomplished for the 2011 version. It is very much on the table for subsequent ones. So at least as far as critical care pathways and what those patients require, I do believe that they are different from the average site.

DR. CRUES: Manny, in particular, what safety recommendations would you not require them to satisfy that you would require a regular urban to satisfy?

DR. KANAL: Actually the answer would be the exact reverse, that sites that wish to be able to be critical access pathways are going to need to satisfy that which the other sites do not, the availability within 30 minutes response, the availability of certain personnel for monitoring certain personnel for sedation and anesthesia. That's an adjunct, if you will, to the MR study itself. It's not take a picture. It may take us three minutes to get the sequence, four minutes for that sequence, and so the ancillary requirements that sedation or anesthesia are monitoring for an acute care patient, that may be required, you may not hold that for the average site, but we would hold that only for the site that wishes to be further in a sense certified or accredited towards those kinds of acute care patients.

DR. CRUES: Well, let's say we're talking about two sites, one rural and one urban, that are dealing with the same medical situation. Would you have the same standards for both, or would you have different standards for the two?

DR. KANAL: Since it's not yet passed, I can't speak for what the rest of the Committee would do. I would absolutely advocate the same standard for both.

DR. CRUES: And that's really what I was saying. I think they should have the same safety standards.

MR. STANLEY: I would just reiterate again, I know your question wasn't just about the technologist portion, but I'm going to be a broken record. The SMRT stands in favor of licensed, credentialed technologists performing MRI scans wherever that happens, on a bus, on a plane, on a train, it doesn't matter. We support -- and at Wal-Mart. We support licensed, credentialed technologists.

DR. DELFINO: Hello. I'm Jana Delfino from the FDA, and I'd like to ask a question about technologist training. Continued and standardized technologist training has been brought up multiple times as a way of improving MRI safety. Given that FDA regulates medical devices and until Congress tells us that we can do so, FDA can't impose training requirements on MR technologists, how can FDA help get to that goal, and specifically since we're standing in a FDA building, what can FDA do to help get to that goal?

DR. CRUES: John Crues from Radnet again. To answer that question, there's actually something that I don't know that I would need to understand from you. Right now I think the credentialing agencies that CMS is requiring are the best way to proceed in that direction. I just don't know what the relationship is between the FDA and the ACR, ICAMRL and Joint Commission.

DR. DELFINO: Well, all separate agencies.

DR. CRUES: Personally right now I think the best mechanism to try to ensure that there's proper education of the technologists is through the accreditation agencies, those three that we've been talking about, and I think they would certainly like guidance from the FDA where appropriate, but personally instead of creating other secondary or maybe competing regulations, I would like to work within the organizations that are currently in place.

MR. FRESE: We heard quite a lot about training and training issues. What I'm missing is the user manual. I've never heard something about the user manual, and I think that's really a problem. I'm the head of the designated complaint unit. So I get all the customer complaints and safety issues on my desk and my team is operating that, and I'm really astonished about the ignorance of the user manuals. Maybe that is something which FDA should support, that they are more frequently right and consult it.

DR. DELFINO: Sorry. I'm not sure I understand your statement about the ignorance of the user manuals. Did that have to do with the content of the user manuals or the fact that nobody was reading them, because it would seem like two different issues?

MR. FRESE: Well, the user manual contains a lot of safety information and recommendations and which is not really designed -- safety which cannot be designed into the MR system because it's how to operate the system. We can only address these issues which we've learned and which we have assessed through our risk assessment, by writing a section in the user manual and recommend, for example, creating emergency plans or something like that, and maybe that could be subject to some kind of auditing.

DR. WOODS: I guess, Georg, to clarify again is it that you're surprised that the users aren't aware of the information that's in the manual? Is that what you're trying to say?

MR. FRESE: Exactly.

DR. FRICKE: That's an interesting statement, and it's somewhat, a little like the user manual is the first thing that should be read, as a matter of fact, at every level, but in this day of evolving equipment and upgrades, typically I believe people are relying on their past knowledge on how to proceed. Although read the manual is a strong message, I think to some extent, the manual is part of the equipment that is regulated. To some extent, I believe that the manual needs to be readable as well.

I remember a very famous phrase in the beginning of one of the manuals for MRI that said to turn the machine on, turn the machine on. Well, you know, where's the button?

So good information is important, but I think when we're talking about devices, really the information on the devices, the testing, the changing of the information, there's a lot of information out there right now. I'd like to be able to have my institution subscribe to a set of devices that it cares about and get regular updates. I'd like the patients themselves to have devices to be identified in any clinic on the planet.

I think that somewhere along the line, the regulation of how the devices are used and the information on these devices can be controlled. I don't have a really good concept of how that would actually be implemented, but if it's education, that's the education we're asking for. What is it the patient has? Where's it going? Can it be used in the MRI? Yes or no. Has that changed since five minutes ago? And that's really what I think the problem is because if a patient comes into the clinic of any age and you ask if you have an implanted device to the patient, parents or anybody around, the answer may be yes or no. And so I was right. It's information. It's about the person, the implants.

What can the FDA do to regulate this, I'm not sure. I'm sure there's plenty of ideas in the room, but I think that's the direction that we need to think about.

MR. GILK: This is Toby Gilk. I think the FDA has a few tools in its arsenal in terms of how to see the good ideas that are collected in this meeting turn into something meaningful at the point of care. I think one of them is simply to state your expectations, the FDA's expectations with respect to MR safety. I think that simple, explicit act helps to establish a standard of expectation, legal or otherwise, for healthcare providers. I think that there is a special relationship between the FDA and CMS.

While I think, and I agree with John in the respect that I think the accreditation organizations are probably the boots on the ground for anything that has to happen, but if FDA can talk with CMS and establish some fundamental patient safety, environmental safety, foundational standards, for MR, with CMS, then CMS in turn turns around and takes those to IEC, Joint Commission, ACR as fundamental elements of their accreditation process.

I think that we need to be careful about distributing the responsibility for the actual implementation too far and wide. I live in Kansas City, and my residence is on the Kansas side of the state line. My work is on the Missouri side of the state line. If we delegate the responsibility of this too far down to the state level, it's very likely that there may be safety requirements where I live, but two miles away where I work, there may not be any governing safety requirements.

So I think the FDA is bully pulpit stating what the FDA expects. I think that's the FDA's relationship with CMS and ultimately CMS' relationship with the accreditation organizations, but I think it needs to be stated that whatever minimum safety criteria come out are ones that are not limited as MIPPA is just on the outpatient imaging side, not limited just to the hospital side, not limited to radiology owned MRs, but applied to all magnets, all settings, for all patients in the U.S.

MR. GOODYEAR: Simon Goodyear from Metrasens. So I'm not sure how the FDA can inference this, but let me just talk from my viewpoint as managing director of a manufacturer of ferromagnetic detection systems.

We feel as a company that it is our responsibility to train anyone who purchases one of these systems, and it's no one else's responsibility. For that system to operate correctly, we have to provide as part of the FMD, the whole package of which, as everyone has agreed today, training is one of the most important elements to that, and so to me, whether all manufacturers behave like that, from the magnets or the FMDs, to me that is how important we think it is.

And going beyond that actually, as a MR safety provider, sponsoring other elements of MRI safety programs to go around ferromagnetic detection systems, again we feel as a manufacturer in this area, it our responsibility to be involved in those, and that's what we're doing.

DR. KANAL: Manny Kanal. Two points. Georg and I just as recently as the last break discussed this. So I can at least address the question he asked as to why don't people read the safety manuals.

There's a very simple reason why some people don't read them, and that's because they're written by attorneys, and not to denigrate them, but to have you recognize, as long as the vendors that create MR systems will continue to have statements such as, patients with implants should not be scanned, well, that means it's a worthless document and it's built to protect the vendor, not my patient and not me. So, of course, we're going to ignore it because it's not written for me. It's written to protect you in a court of law. So that's not a place that any of us will reasonably expect to find useful information.

Now, as you've shown me, people are attempting to address that and correct that, a dismal situation, and when it is, then there will be something that we might want to turn to.

The question is how should the FDA or how can the FDA address some of these issues of technologist education? It's an amazing situation because we could have an eight-month pregnant female getting into a motor vehicle accident and showing up in a MR scanner in a rural or urban situation, and the person who's scanning them is the person who drove the truck on that mobile MR that afternoon because the tech quit, and what I just described is a literal situation. So I'm not quite sure why it's so unbelievably unregulated and why a piece of equipment that has the ability to provide tremendous diagnostic potential or to kill you is unregulated in that fashion as to who operates it.

I do have respect for what I have been taught by you, which the FDA has gone out of their way to teach us, that they're not in a position to regulate minimum educational requirements or what have you, but on the other hand, I very much agree with John. We have organizations, accrediting organizations that we're turning to, but the problem is there are three of them, and as long as there are three of them that CMS has turned to, they may come out with different standards that again is not going to help. In my opinion, it'll muddy the waters even further.

For this to work, there has to be a single benevolent dictator, telling everyone what we think is in everyone's best interest, and the bully pulpit as you've already heard that the FDA has can be used in conjunction with CMS to create what you consider minimum acceptable standards. You don't even have to mandate them. Have a friendly conversation with CMS and tell them this is what we think is reasonable. They don't know anything about MR safety. They're looking to others with expertise to help them define it, but they've looked to multiple locations.

Everyone respects the FDA. Everyone understands that CDRH has input into these issues. It would be relatively straightforward to be able to come out with a statement saying these are what we think would work. We'd like to see these implemented by other organization or organizations that are in a position to enforce it, and that wouldn't be that difficult to do. The best way to do it perhaps might be in your labeling. You can't mandate the labeling, but I think it would be difficult to anticipate that the vendors of the country would take issue if you equally apply to all MR vendors a statement that said something to the effect of appropriate education and/or perhaps even, God forbid, certification would be required for the anticipated operators of these complex pieces of equipment.

MR. KOPP: Keith Kopp. I realize the limitations the FDA has in the regulatory elements of the operation of a MR, but there are examples, the CMS being one, where you can encourage an environment, and by that I mean the crash testing done for cars. The FTC basically publishes a NRC style rating, and what it helps is, yes, the cars that are safer are a little more expensive.

If you're going to have a MR done, and you know MR Suite 1 has a part-time tech that shows up once a month and MR Suite 2 has a full-time, trained, certified technologist, and you had an energy, no, not an energy style, but a safety style, a quality style rating system that you could promulgate, then an educated consumer, and we are all consumers when we're on the receiving end of this, could choose the more expensive facility, knowing full well that the likelihood of a safe outcome might be higher.

So you don't necessarily always have to have regulatory authority. You can use indirectly a bully pulpit saying this facility has these capabilities that will contribute to a safe operation. The facility has fewer. This facility over here is a bottom feeder, and if the consumer basically can decide, I'd rather go to someplace where I have a better chance of getting out without injury, or if I'm a technologist, I'd rather work at this place because I'm less likely to get injured, you may be able to get this operation going in a direction using something like the ACR safety recommendations as a basis to say, okay, these facilities have these capabilities and you as a user or as a hospital or whatever, you decide which is best.

MR. LAZEBNIK: Roee Lazebnik. In my opinion at least, safety really is a triangle of documentation, training, and experience. From the vendor perspective, certainly we put a tremendous effort in providing a large volume of documentation to the best of our ability, and I think it's certainly tempting to say, you know, the solution is more training and that would be phenomenal, but I do think it's, you know, to second the opinion, it's very important to always consult specific devices documentation specifically because devices are different. A three-piece scanner from one vendor is not necessarily the same in many important respects as a three-piece scanner from another vendor. We all know that different pulse sequences, for instance, are offered by different vendors. That information would be very difficult to capture in one standardized training across the entire country and certainly worldwide because it would not necessarily apply comprehensively to every device manufacturer's specific system, and thus the triangle.

It's, yes, training is extremely important. A technologist's and radiologist's experience is very important, but adherence to the documentation in consultation with the specific documentation for a given device is also extremely important.

MS. RIVI: Diana, FDA. I'm just very thankful to hear all the comments about the specific things that FDA can take on. That is a tricky issue because there are so many players in this arena, but definitely in collaboration with the accrediting bodies is one direction that I feel we would explore after this workshop, and I'm sure there's going to be more coming in the next day that we can all address together.

MS. SAWYER: Anne Sawyer. I'd like to say a long list of requirements, but I won't go there, but I'll just say about documentation. One of the reasons it's not read is because there's no time. It gets back to the money thing, right. There's no time given to the technologists to read the documentation. There's no encouragement to read the documentation. So, you know, that's one of the reasons why it's not read.

MR. STANLEY: Charles Stanley. Help me out. I'm naive. Is it the FDA that requires the sticker on the scanner that says must be used under the guidance of a licensed physician?

DR. DELFINO: So, yes, FDA approves medical devices --

MR. STANLEY: Okay.

DR. DELFINO:  -- for either prescription or over-the-counter use.

MR. STANLEY: Okay. Then what I would like for them to do is add a second sentence to that sticker that says that it also must be a licensed certified technologist that's running that scanner, and that would solve that. I think that's what Manny was saying earlier, and I would like to again correct, I've heard again from this panel, that we think that the accreditation agencies are addressing this right now, and if we would just adhere to what they've already put in place, that we would be okay.

And I've again reviewed, as I was waiting my turn, each one of the pdf documents from the accreditation agencies, and if you look at qualifications for technologists, there's about 17 ors, or, or, or. There are plenty of ways that you can get an accreditation sticker on your scanner and not have licensed, certified technologists running that scanner, plenty of ways. So unless we change those rules, the accreditation bodies are not the answer. But, yeah, I'm in favor of another sentence on the sticker.

DR. FRICKE: Stanley Fricke. Just one more comment. I think that the FDA role here, we're talking about metal detection in generic MR safety, and tomorrow we're talking about implanted devices. But I find it exceedingly difficult again when a patient comes with maybe or not an implanted device, and somewhere along the line, this maybe or not is ridiculous today. It needs to be straightened up. We need to know what is it you have, make, model, serial number, lot, and all that. So that somewhere along the line the regulation on how that device is used and how it's tracked is something that I don't think we have a uniform consensus on at the moment.

DR. WOODS: We have two more questions and not a lot of time.

MS. ROBB-BELVILLE: Hi, Sonja Robb-Belville with the University of Texas MD Anderson Cancer Center.

A couple of comments and then a question that I have for the panel.

It is distinct. ACR MRI accreditation is completely distinct from the ACR safe practices document and, in fact, ACR accreditation, MR accreditation says when it comes to safety, see the safe practices document.

As Charles just pointed out, there's about 15 or, or, ors that will allow someone to get behind that console and operate that scanner and not be credentialed whatsoever. They may have any allied health associate's degree and have six months of qualified supervision, that is not specified as to what qualified supervision is, and it's a huge loophole in MRI being conducted safely on patients. Further, when we say licensed, certified technologists, we're talking about licensed, certified, MR-certified technologists.

Cross-training went out in the '80s, okay. In the '80s, it was researched to do a MRI of the knee. It's no longer researched to do a MRI of the knee. I do not envy my students that are with me for one year to get their entire MR curriculum that the ASRT and the SMRT developed together. One year, and that is a long time. There are programs out there that are six weeks, et cetera.

That being said, I'd like to know what the panel's thoughts are on that.

DR. CRUES: John Crues from Radnet. I was also on the ARRT MR test committee for MR accreditation for six years, and I've discussed this at length. Ideally we would all like for all the technologists everywhere to be ARRT certified. The problem with that is that that's only about 15 percent of all the technologists that the United States needs. So there would be an enormous shortage, and at current rates, there are not enough being accredited each year to really fill in for even the current growth, which is much lower than it was before.

So the problem is that it's impossible to do that mandate right now because there is just not enough manpower to satisfy the need in the marketplace, and there's been a big push to try to get more and more technologists interested in getting accredited, but it has led to limited success.

So the ACR really was pushing to have ARRT accredited MRs at all of the ACR accredited sites, but they also realized that that was impossible. The sites just would not be able to get ACR accreditation because there are just not enough trained technologists to fill the slots. So that's the reason why they've backed off of that requirement and they're one of the supporters. They have four of the nine board members on the ARRT. So they're very much behind the ARRT, but in the marketplace, it would be impossible to do that.

So that's why they have a number of other pathways where people can get reasonably good training to be MR techs, but I think everyone would agree with you, in the ideal world, everyone would like the techs to be ARRT certified. It's just not possible.

MR. FRESE: I don't know enough about the accreditation program in the United States to comment on that, but I'm pretty happy that we don't have these problems in Germany because the operation of MR systems are much more regulated than here.

DR. FRICKE: Well, I don't think anybody really can truly say that they're done being trained in MRI. I don't know anybody that knows everything about the equipment. I don't know anybody that ever will know everything about the equipment. And the minute they do know everything about the equipment, one second later they suddenly don't. So continued training I think is protocol. I'm just curious as to how the FDA can do anything exactly about that.

MR. GILK: This is Toby Gilk, and thank you for asking your question because you give me the opportunity to give the additional answer that Terry cut me off on last time.

Dealing specifically with the accreditation organizations, and you gave the example, the ACR and how their safety guidance is not an integral part of their accreditation program. I think that in the interest of not spending our time, energy, and effort overlaying an additional layer of bureaucracy on imaging, because at some point, all of that effort winds up taking away from the productive things that we can do in terms of delivering patient care and delivering it safely, I think that the accreditation organizations are probably the best implementation tool, as Manny pointed out earlier, because they are in competition with one another. In effect, there's a financial disincentive for these membership organizations to impose any kind of additional safety requirements that their membership might perceive as a burden.

So when the accreditation organizations talk about quality and safety, they have a direct incentive to run screening the opposite direction from either of those if they place any burden on the accredited provider.

So what I think, if we agree that they're the boots on the ground that can effectively monitor the compliance, I think we also need to agree that they in and of themselves cannot be the arbiters of what minimum acceptable practices are. That needs to be kicked up the food chain a little bit.

And as a part of my answer to Eana earlier, about what can the FDA do, if we work towards CMS establishing the minimum acceptable standards, the accreditation organizations are essentially the frontline troops for CMS standards. If we can get CMS to acknowledge, you know, four zone minimum standards, technologists' training and credentialing, whatever winds up being the end product of this, if that becomes part of CMS' conditions of participation, automatically the accreditation organizations have to observe that and enforce that at the point of care.

So for technologists' qualification and training issues, for ferromagnetic detection, for ACR four zone, for labeling and testing and all of the products, information and ideas that come out of this meeting, we need to distill those and ultimately I think get them to CMS.

DR. KANAL: Everything that John said, I completely agree with. I actually was the head of the committee at the time, and still am, for the ACR document, and several times, as you may perhaps be aware, I tried to push through that they should be MR sub-certified. This last, or the second to last iteration in 2007, I had difficulty even getting it to pass that there had to be RTs let alone MR sub-certified just because of the real or perceived shortage, but the fact remains, as long as it can't be implemented, then no matter how good it sounds, you're destined to failure, and you can't hurt the system.

Perhaps the only thing I might suggest, differently from what John said, is that he ended with I think because of the shortage it can't be done, and I agree. But I wouldn't want to apply it quite as broadly. Because of the shortage, it can't be done today. There is a possibility of suggesting, which I'm not sure if it would be successful, but suggesting that by the year 2000-and-blank, you will be MR sub-certified or you will not be running a system, and that won't happen because of an ACR. It's going to have to come because of a regulatory agency or some sort that has the teeth to enforce it, and it has to be aimed to the future because today we don't.

If it does come out, I think the problem that John and I both lived through, a shortage, will go away because by the time that deadline hits, we're going to be opening up the job market. Do you want to be a MR tech? This is what you're going to have to do, and people will scramble to have that happen.

MS. ROBB-BELVILLE: I have some graduates that would love to talk to you all.

MR. KOPP: I agree with what was said earlier in the sense that it cannot be the safety standards, cannot be done at the accreditation level. They're in a competitive situation, and in reviewing the documents I've seen from each of the organizations, they put the least burdensome requirements on the facilities they're accrediting.

Even though the legislation, the MIPPA legislation, for example, through CMS calls out this safety of the patients and staff, I have not seen any documents from any of the three accrediting organizations that have any safety standards built into them. I may be wrong, but from what I've seen, there is no such thing. It's strictly, what's the minimum we can get away with to meet the accreditation requirements? So it has to come from higher up, either through CMS, FDA, or some other organization.

MS. RIVI: Diana Rivi, FDA. So just to clarify, the FDA regulates the medical device industry. It isn't our jurisdiction to mandate certification and training or make clinical practice recommendations, but we do have the ability to work with those organizations that do, and that's what our role is in this.

MR. STANLEY: Okay. Well, I'd just like to say that I'm not sure when was the last time the ACR people met and went over the discussions about the technologist shortage. If it was related to the 2007 document, I can tell you that the times have changed. I can't offer you numbers right now today, but I can give you all kinds of anecdotal information. Every technologist in this room understands what Advance Magazine is, right?

Okay. Advance Magazine in 2007 was about as thick as a phone book with technologist jobs. The thing's barely in publication today, okay, because there's no jobs anywhere. Sonja's graduating techs out of her school but can't find MR jobs, and they're willing to move anywhere and do it. So we understand that there's pockets of shortages that exist, sometimes in large urban areas where you can't afford to take the job that's there because you can't afford to live there, but there is not a technologist shortage right now. That's point number one.

Point number two, we do recognize that the ARRT is the premier organization that's always been out there for credentialing of technologists, but the SMRT and the ACR doesn't recognize them as the only organization that could credential technologists. So I'm not necessarily advocating that it has to be the ARRT.

Now, I do agree with Manny's suggestion that maybe there's a way that we can kind of work towards that. There are tremendous amounts of online education that's provided. There are providers of that online education in this room today who have helped certify, credential, whatever you want to call it, literally hundreds of technologists over the years. So there's ways it can be done, but I just think that it's very shortsighted to say that we're going to continue a substandard practice because to raise the bar, it's going to hurt us. I mean I can't see us applying that to any other industry. We wouldn't let someone fly an Airbus 320 that was only certified on a Cessna because it was going to be too arduous of a process to get them there. We wouldn't fly the plane. We simply would not fly the plane, and so I just think that we should take it upon ourselves to figure out how to do it instead of just saying it's too hard to be done.

UNIDENTIFIED SPEAKER: I'll add one thing to that, Charles. According to the ARRT, 80 percent of technologists practicing MR are MR certified.

MR. STANLEY: I understand, but I think I discussed that with you, of all places, on Facebook. God, you love social media, but the ARRT only surveys their own technologists. So what they're saying is 80 percent of their technologists who list MRI as their primary discipline are certified in MRI or post-certified in MRI, but that is not all inclusive of every technologist that's practicing MR by any stretch of the imagination. Again, as Manny alluded earlier, there's the truck driver. There's the nurse. There's the secretary, so on and so on, and those people don't get ARRT questionnaires sent to them.

DR. CRUES: I think you have to realize, however, that the ACR and ICAMRL believe that their criteria are criteria for minimum standard for technologists, and they're not just truck drivers. The truck driver who came off the street would not satisfy either of their requirements.

MR. STANLEY: I understand --

DR. CRUES: If you're talking about other MR certifications besides ARRT, we tried to make it only ARRT, but they have just recently, just a few years ago, the ACR finally agreed to accept other certifications, and you're correct. If you widen it to other certifications, you may get up to 40 to 50 percent of MR people out there, but if you fired 50 percent of the MR technologists right now, there are not enough certified people out there to fill those positions.

MR. STANLEY: I don't disagree.

DR. CRUES: I agree with Manny. I was one who really pushed for ARRT certification period, but I think there are ways that people could put pressure to try to get more education, and I really believe the ACR would like to do that.

MR. STANLEY: But my question to you, and please correct me if I'm wrong, does the ACR document state that all technologists that work at the accredited facility have to have one of those criteria in place?

DR. CRUES: I haven't read the latest document. Maybe, Manny, you have. ICAMRL requires all of the technologists to have some level of training.

MR. STANLEY: And I may be incorrect, but I just read it again, and the way I read it is, is that you must have one of those people on staff, but that doesn't necessarily have to be the one who's performing the MRI scan. And so what happens is there are places that have one of those people who meet that criteria on staff, but that is not the person who is doing all the scans.

DR. WOODS: Okay. And I think I'd like to table this question now and --

MR. SHELLOCK: Thank you, Terry, now that the blood is pooled in my lower extremities.

DR. WOODS: There we go. Everybody's --

DR. SHELLOCK: Frank Shellock.

DR. WOODS:  -- fired up and ready for --

UNIDENTIFIED SPEAKER: Frank, weren't you over there before?

MR. SHELLOCK: Yeah, I had to walk around. Anyway, University of Southern California School of Medicine. One comment and then a question.

So I'm actually getting a call from CNN right now. Obviously they know that this workshop is going on, and they want a comment from me because the statement that I got by e-mail is that they want to know why there's such a huge increase in the MRI accidents and what, you know, they've obviously seen is this blog and/or a nice graph with the line going up and so on and so forth that says it increased in rate of accidents, you know, 400 plus percent, what have you.

And I think it's important, you know, as the MRI community members that we are to give the right message out.

We know that there's an increase in the reporting of accidents, but we don't know why that is, and I know the FDA has commented on this in their publication in 2010 when they reviewed the thermal injury information. So is the increase in accidents related to an increased number of MRI procedures and/or because we're more aware that we need to make commentaries and we need to report to the MAUDE database?

So we need to get a handle on that, and I think as the FDA is here and listening to our comments on this topic, that they should be proactive and try to get good data about this so we can really know what's going on, and hopefully, you know, there is not an increase in the rate of accidents.

In talking to the vendors, it appears that according to their data and their analysis of their in-house data, not only is there a stabilization or a consistent level of the number of accidents but it actually seems to be dropping. So what really is going on?

Now, with regard to my question, I know the 2011 ACR document is coming out, and I'm glad to see that this continues to evolve and what have you, and I'm wondering, you know, based on Manny's comment that all MRI facilities should adhere to that, has the document changed dramatically to the extent where it will encompass all the different MRI types of facilities? So outpatient, academic centers, research centers. We've got a MRI facility at USC, on the USC campus, away from the medical school. UCLA has the same thing, in their neurology department. So there's a lot of those types of situations and, of course, the dedicated extremity systems and the systems that are in other specialized areas. Has the document evolved to the extent where it will incorporate those types of situations?

And as far as the commentary, I'd like to hear from the panel, what are the challenges that exist to actually follow the ACR recommendations in the White Paper? Because I know that there are many challenges particularly in an outpatient facility versus an academic center or a hospital based facility in order to adhere to that, you know, verbatim. And maybe I'll sit down for this, but I'll stand up as far as I can go. So go ahead.

DR. CRUES: Wow. Well, the first thing I'd like to say is I think mandating accurate reporting of adverse effects in a way that's reliable would be very valuable because I actually have more faith in our internal recording of adverse effects than Manny does, and certainly our data would support the data that you hear from the manufacturers, in that over a 10 year period of really looking carefully, and I think we have very good internal reporting of adverse effects, we certainly have seen no increased rate.

So part of that could be that it's outpatient and a lot of the risks that are associated with MR are much more of a problem in inpatient especially as we're getting much more complex environments like in the operating room, and when you're actually dealing with critical patients like Manny was referring to earlier which you don't have in an outpatient setting.

So it could be that there's a real increase, but my guess is, it's probably in the hospital or it could be, which I think is more likely, there's been a lot more press about adverse effects of MR and injuries, and I think because of that increased awareness in the medical community, more people are reporting those incidents than in the past. But the fact that some sites may be inhibited from reporting, I think, is something that really should be addressed because I think we need accurate data, and I think that's very important.

As far as restrictions of implementation, I don't know what's coming out in the new document. I think some of the architectural designs concerning zoning for safety I think is very relevant to the inpatient setting, but I just really don't think it's been shown to be a significant issue in the outpatient setting, and that would be a major hurdle in the design of outpatient imaging centers and markedly increase the cost.

And another thing I'd like to state here is that the cost of actually putting in a device is not that important. What's really important is the cost to the day-to-day function. Let me explain. If it takes 5 more minutes to do each patient because you get false positive alarms, it would have to go off and you'd have to then spend 5 or 10 minutes going through the patient, that will lose about a patient a day, which would end up losing about $100,000 a year, which over 10 years is $1 million. So that's much more of a concern to the operator of imaging centers than putting in $3,000 for a device. So it's really the operational inefficiencies, and if it's worth it because it actually saves lives, then it's worth it, but in the setting where it doesn't save lives, we're much better putting those kinds of resources into areas where we get much more safety for the dollar.

MR. FRESE: Georg Frese, Siemens. I would like to confirm the statement that the incident rate is remaining constant over the last 10 years. I don't understand the figure that the increase of reported incidents in the MAUDE database is fivefold. Maybe that statement, of course, is correct regarding MAUDE, but we internally, of course, have the duty for product monitoring, and as I said before, I'm the head of the designated complaint unit. That means any safety issue related to our systems which are coming to our attention will come on my desk, and we are, of course, having statistics and the statistics prove there is no increase in the incident rate during the last 10 years. I must admit I don't know if the figure now related to the States only or that is a figure worldwide.

DR. KANAL: Well, 30 plus years ago when MRI was just getting rolled out into some of the clinics and access to the PC by the average person was much more limited as well, but in 2011, I think everybody has three PCs at their disposal, and if you add that to the fact that online reporting has become actually available and to some extent, strangely enough, user friendly, I think reporting certainly increased.

But when it says increase in incidents, and this is what I haven't heard here, is that per magnet, per number of people scanned, or just total incidents?

DR. KANAL: Well, as the person who's sort of put those numbers up loud and proud --

DR. SHELLOCK: Yeah, I'll give them your phone number after I talk to them. So, all right.

DR. KANAL: My slide anyway was the total number of incident reports in the MAUDE database. Now, with respect to question of, you know, why are they going up when there are individual providers or vendors who, you know, have a long track record and they don't see these changes in their installed base, where the numbers that I presented are presumably nationwide and they do include, you know, a minority at least that are identified in the narrative of international reports. I have been, you know, as intimately involved in sort of general MR safety culture as anyone over almost the last 10 years. I am not aware of a single body, a single institution, from the FDA to the ACR to the ARRT to any organization that has done any kind of awareness campaign in terms of reporting either how to, where to, when to, you know, or what the mechanisms are for reporting. I certainly, given what we all acknowledge are dismal reporting rates, I cannot rule out the notion that more people simply are reporting than had been. It's much easier to go that direction than the other direction because, you know, estimates of reporting are vetted around, you know, one to two percent rate.

So is there room for that number to move up and not reflect actual increases in accidents? Absolutely. I'm not aware of anything that anyone has done in the community in any kind of broad-based effort to increase awareness of reporting or facilitate the ease of reporting. So I --

DR. SHELLOCK: Just one comment, too. Isn't it true that the method of reporting and the means of reporting on the MAUDE database has changed over the years as well, Terri?

DR. WOODS: I think we should let Diana from the Office of Surveillance and Biometrics respond to that.

MS. RIVI: We have a paper form that can be sent in, and we also do have online reporting. Does that answer you question?

DR. KANAL: So while we can, you know, have a debate over the validity of the numbers given, that they do represent a very small sample size of the total number of accidents, I think -- nobody has demonstrated to me any substantiation for the idea that the numbers are growing because more people are reporting in any way, shape, or form. That coupled with things like Manny's anecdotal, receiving more inquires regarding legal proceedings resulting from MR accidents, the indications that I see, nothing supports the notion that the industry became more effective at reporting their mistakes.

With respect to your question about the hurdles of implementation of the ACR Guidance Document, I think that the key principles of the ACR Guidance Document appointing a safety officer, you know, you want to go get 500 free business cards printed with somebody's name as safety officer printed on it? That's eminently doable, zero dollar implementation cost.

John made mention of the four zone principle. I think that that has been grossly overcomplicated in the industry. The key tenet of the four zone principle is we are in Zone 1 right now. Zone 4 is your magnet. Do you have a facility that cannot provide two spaces between the front door and the magnet?

DR. SHELLOCK: A lot of places don't have that luxury and they're already in place, and so how do you handle those situations? I mean at the University of Southern California we've got eight magnets. We have an active hallway that goes into the magnet room. So perfect example, and we tried to put a ferromagnetic detection system there. It was set off by the cars moving in the car lot beneath that. So --

DR. KANAL: When was your last magnet put in?

DR. SHELLOCK: A couple of years ago.

DR. KANAL: And it's on an active hallway in contravention of --

DR. SHELLOCK: Same thing at Cedar Sinai Medical Center. So high profile medical centers, it's not uncommon to see that.

DR. KANAL: Well, it may not be uncommon --

DR. WOODS: Maybe I can interrupt and we can continue answering the question about the ACR guidelines.

DR. KANAL: I would say not uncommon but I would say unconscionable.

MR. GOODYEAR: I think there's always been some controversy over the numbers, and I don't think we'll ever get to the bottom of that for commercial reasons, but the one thing that I always found is I've never been to a MRI site when people get down and honestly talk to you. They can't talk about a near miss situation, and for me, one near miss is too many and one accident is too many, and even if numbers aren't substantially increasing, it doesn't mean that you shouldn't be looking to improve safety through whatever means, whether it's training, whether it's technology, or some other aspects. That's my first point.

On the challenges ahead, one of the points I'll maybe bring up that hasn't come in today is on implementing design for safety. Okay. This will let out the facilities you're talking about existing already, but let's look to the future here, and let's get in the very early stage of designing the layout of hospitals and MR facilities to take account of workflow safety practice and the availability of technology and designing that in the very early stage, in 10, 15 years time, you'll be sitting down with a very different invitation. So that's just one challenge today.

DR. KANAL: Manny Kanal. You've provided me with so much material.

DR. SHELLOCK: I know. I know.

DR. KANAL: Let's try to stay focused and brief, both difficult considering the questions.

Number one, the perceived or real increase in adverse events, what a sidetrack we're finding ourselves on. Thirty years. We have almost 100 percent preventable problems that we're still dealing with as if it was 1984. Time's up. We're not doing a good job of regulating ourselves. It's time to have daddy tell us what to do and have some positive and some negative for doing what we're supposed to and not doing what we're supposed to. Just de facto, we're not even close to zero.

I think it's fascinating that we use statistics, statistics. I have one serious adverse event for, and then someone will make up a number, a million cases. Patient comes in, he has in his pocket, again real case, unfortunately legal, comes in, patient has in his pocket nail clippers. The technologist says, oh, he sees the pocket bulging out with a sharp point to it and says, oh, wait a minute. You have something in your pocket? Oh, shoot, I forgot, my nail clippers. Here they disagree. She claims she told the patient, hold it, I'm taking you out. He claims he heard hold onto it, I'm taking you out. Either way, he puts his hand in his pocket to hold it, opens the pocket, flies out and hits him in the eye, and he loses an eye. It's completely preventable, 100 percent so, that the ferromagnetic detector could have helped. It's not even to be discussed. Of course, it could have helped. But it only happened once.

You know, I always like to quote, one loss of life or limb can ruin your whole afternoon. Conceptually the fact is we're not doing a good enough job, and to say that we're accepting a 1 out of 1,000, no, the objective is just like the FAA. The objective is zero. We are not approaching zero. We're arguing as to whether it's really a fivefold increase or is it steady or is only twice as much, and they're only seeing the ones that call them. Our attempt to conjecture, that more people are reporting. It's almost -- I'd love to hear. Who is the person who is doing the reporting other than the ticked off person who got injured because I don't see any more people that are coming to me saying now I know. I need to report somewhere. If they do report, they don't always report to the FDA in any case. Who they do report it to is their attorneys.

So conceptually the question isn't are there fivefold or twofold or what. The question is why aren't we going down to zero? Why isn't it markedly decreasing with 30 years of experience as to what it takes to prevent almost 100 percent preventable adverse events? That's the real issue, and that's what CNN needs to hear, is that the industry is focusing on zero as our ultimate goal, nothing else.

Having said that, the question about the difficult implementation, I think you sort of answered that one yourself by calling it, we can't afford the luxury of four zones. I think other sites would tend to differ with the choice of words of whether that would be considered a luxury or minimum requirement entrance fee.

The idea of calling four zones a luxury is something I wouldn't expect would be an acceptable part of this discussion. How about sites that don't have it? We can maybe try to work out getting them in there, implementing them, or at the very least, let's talk about future magnets.

I am opposed to the idea of grandfathering in old sites, but at least include that, at least acknowledge that for all future implementations, a design for safety is an awful lot better than an afterthought considering it a luxury because it's not. It's basic, and do we need a scientific study to prove if I can't -- what is Zone 3? The definition of Zone 3 is the area where if I leave you unaccompanied, you may intentionally or accidentally access areas that can be harmful to the uninitiated. Well, to have somebody tell you that I need a control study to show that if I restrict them from such an area, it'll be better for them. No, I don't really think that's where my research funding should be going. So conceptually I'm not sure I see what the major implementation limitations are.

Having said that, the last thing you asked was what about anticipating the ACR document? What about anticipating outpatient centers or inpatient centers or research centers or I think you said also extremity magnets? One of the nice things about that is they were anticipated from the 2002 draft on. I'm not actually sure what the question is asking. Perhaps you can clarify it.

For example, it defines those areas that are potentially harmful. So by definition, if there are no harmful -- for example, Manny has played devil's advocate. What if it's an extremity scanner, the 5 gauss line, for example, is inside the faceplate of the magnet. You can't even reach it unless you stick your arm in there. Well, if it's not, then it wouldn't meet the qualification for being Zone 3, would it? It's not an issue. It's literally from day one prospectively designed to incorporate research centers, outpatient centers, inpatient centers. We're up to the fourth version of it now, and every single one of them had those in mind.

So I guess the question is, does it anticipate their needs? Yes, I actually very much hope that we did and hope it continues to do so with all subsequent versions as well.

UNIDENTIFIED SPEAKER: Well, I think the FDA needs to recognize that those involved with those specialized facilities are not radiologists. They're not ACR members. They have no idea what the ACR White Paper for safety entails, and so there's a disconnect there between what goes on at those facilities and, you know, what the ACR safety paper presents.

DR. KANAL: We're in agreement on that.

UNIDENTIFIED SPEAKER: Pardon me.

DR. KANAL: We are completely in agreement on that.

UNIDENTIFIED SPEAKER: But I wanted to get it on the public record, Manny.

MR. KOPP: Keith Kopp. Dr. Kanal is a very hard person to follow, but I'll make a couple of comments.

We can debate the merits of the MAUDE database or not, but there are two databases here in the U.S. that provide very credible information that supports what Toby's data initiates. The VA, for example, keeps track, the control of the expenditures of incidents, when they happen, near misses and so on. Their data shows increasing. As a matter of fact, their data is detailed enough, they say on average, every MR site within their system has at least a near miss every month. So it's a fairly compelling document. They have a list of reasons that they've gone through because they've done a root cause analysis and come up with a list which actually have been aggravated by the newer magnet designs. The higher spatial gradients, the higher fields, et cetera. And the types of procedures, many things we have talked about.

The other group that has done a very detailed study is the State of Pennsylvania. Part of their legislation is misses and near misses have to be reported. Well, interestingly enough, just that database for one state shows almost as many incidents as the FDA database does for the whole country, and it shows the same trend, increasing with each ensuing year. So, yes, we can all debate. That's been pushed around, but there are sources, even today, that show the occurrences of incidents are underreported and the trend is in the wrong direction, and as Manny points out, we have within our means the ability to reduce this to near zero.

DR. LAZEBNIK: Roee Lazebnik. A comment I would like to add is that I think it's useful in addition to looking at one big global number, really dive down and understand the complexities of what drives that number. So, for instance, we know what's been expressed by members of the panel, that overwhelmingly misuse issues are greater than use issues, and by that I mean I know very often when an issue occurs, it is preventable, that is to say, that if you were going to improve safety somewhere, it does have to do with things like how the equipment is operated.

The other component of doing this type of sub-analysis is the MR environment has been changing in extraordinary ways over the last 10, 20 years. As you said, we now have extremity scanners that serve one type of population, which probably tells us about the case mix, that those types of scanners are going to see, is different than say a whole body scanner inside a hospital. Even the ratio of inpatient to outpatient exams has changed. Outpatients typically don't show up with a ventilator, unlike some inpatients, and that does represent an additional hazard. Should that affect the general principle of how you screen people into the MR environment? You know, maybe not, but the reality is that there are some hazards, for instance, in the inpatient environment that are very atypical in the outpatient environment. And as we talk about how can we improve safety, I think those are all factors to consider.

Obviously as MR technology evolved, MR equipment is able to provide more information and more diverse pathology than ever before. So there were the patients who were "too sick" to get a MR exam because it was too long in the past, but now we will routinely take them down to the radiology department because a more rapid MR is possible.

So any analysis of the hazards I think shouldn't focus necessarily on one number but really should involve at least some layer of detail as to what's driving those numbers and seen in the context of how MR has evolved.

MS. RIVI: Diana Rivi, FDA. I want to try to respond to a bunch of things I've heard in this conversation. First of all, we've all talked about MAUDE limitations. We all know they're there, and as said before, the data is not intended to evaluate the rates of adverse events.

I also heard mandate accurate reporting. I don't know what accurate reporting means, but our regulations do mandate that manufacturers report deaths and serious injuries and malfunctions to the FDA. The regulations also mandate that user facilities, such as hospitals, report serious injuries and deaths to the manufacturer as well as to the FDA.

And I mentioned in my presentation this morning, we have a network of hospitals that have been trained to report to FDA as part of the Medical Product Safety Network, known as MedSun, and that's been our attempt at some sort of reporting awareness campaign. We've done a lot of outreach to them to let them know about FDA's regulations. A lot of times it's just the manufacturers that are reporting to FDA. There's just a fraction of user facilities or hospitals that do report, and many of them are through this program, and I think there's just an awareness gap about FDA's regulations of what needs to be reported.

But as part of the MedSun program, we're fortunate to get not just the serious injuries and deaths, but all those close calls and potential for harm events that we've been talking about, and that's kind of been our eyes and ears to the clinical community to see what's going on with these products.

And when we developed this MedSun program, we heard, you know, there's a lot of barriers to their reporting to FDA. A lot of times they're scared of FDA or their hospital doesn't support it, that they don't have that culture of safety, and basically there's just that knowledge gap about reporting requirements. So I hope that helps answer some questions about what FDA's mandate is for adverse event reporting.

MS. SAWYER: Anne Sawyer. I just wanted to bring up something Frank was asking about, these other different kinds of facilities that get MR, and this goes back I believe to something that Manny said earlier about, I can't use the word you did about other cardiologists, neurosurgeons. What was it you said again?

DR. KANAL: Two functioning neurons.

MS. SAWYER: That's right, that they didn't have two functioning neurons. So anyway --

UNIDENTIFIED SPEAKER: To rub together I think he said.

DR. KANAL: Well, to constitute one intact synapse.

UNIDENTIFIED SPEAKER: Yeah, we get the gist.

MS. SAWYER: We have this problem at Stanford, and I manage a research facility, and one of the things I'm always nagging our scientists and researchers and M.D.'s and Ph.D.'s and technologists about is that we can't just worry about us and our facility because what we do reflects on all of Stanford. So if there's an oxygen tank that goes in and injures somebody in my facility, you think people are going to want to take their child or their wife or their husband or whomever to Stanford Hospital? Well, no, they don't. So I'm trying to get people to understand this responsibility they have.

So along those lines, recently cardiology has put in a research magnet and psychology has put in a research magnet, and when they were going through the designs of the cardiology, environmental health and safety would come to me and beg me to come over there and look at their designs, begged me to come over there and tell cardiology how they were doing all of this wrong. Well, I work for radiology. First of all, I don't have time, but secondly, they don't want my input. They didn't ask for it.

And then when psychology was designing their facility, they asked if their architects could come over and see our facility. So I said, fine, okay. You know, I owe the MR community whatever little expertise I have in MR safety. So I spent two hours going through all our magnet suites and our entire facility explaining each and every item, why we do it that way, how it has to do with screening, and when we got to the very end, and they were all writing madly. There must have been five or six of them, and they looked at me, and they said we just want you to know that we understand every item that you've brought up with regards to MR safety and patient confidentiality and all these things, but I just want to tell you, each and everything that's on this list are things that they've already told us that they don't want and they don't think is important. So these people have a magnet, and again I think, and maybe I'm really naive, that the FDA should regulate who gets to buy these things and who gets to use them.

MR. STANLEY: If it's not the FDA, whomever. We need one ring to bind them all, and that's the point. If somebody needs to bring it all together because they're not going to go away and, you know, reporting is just the tip of the iceberg. If we're saying right now the only people who are halfway reporting are the radiology people, you can guarantee that the people with the two functioning neurons or nonfunctioning neurons are not even aware of the database, want to be aware of the database, are going to report into the database, and they are the ones that are more likely to cause the issues than the diligent efforts of the radiology department.

So again, if it's not the FDA, it has to be somebody. I think that's the takeaway message today is, okay, tag, you're not it. Okay. Then who's going to be it, but we are now at a point that we need an it.

DR. WOODS: And with that, I would like to sincerely thank everyone who has served on this panel and the roundtable for a marathon session, and we really appreciate you all sharing your diverse expertise with us. I will dismiss you to the audience, and we will have a brief summary of the day, and then let you go.

(Applause.)

DR. RAJAN: Okay. This is Sunder Rajan again. I guess in retrospect I realize that we forgot to invite the cardiologists specifically to this meeting. (Laughter.)

Well, I've been told that I should summarize in a few minutes the entire day's events, which was a lot, and I've tried to simplify things for myself. Basically I thought I picked up information on four broad areas.

We did hear the concept or the topic of training and education and accreditation and so on over and over again. It's clear that this is something that we have to pursue as a community, keeping in mind that FDA has its limits.

Our primary role at FDA, of course, is safeguarding public health, but the way we do this is by, at least from a MR perspective, we clear MRI devices, looking at the safety and effectiveness. Training is a little bit outside our scope, but we do have the ability to perhaps team up with people that -- perhaps CMS and so on, and we should continue to pursue this and continue to see what we can do in terms of accreditation, in terms of improving safety and effectiveness.

So also really most of us are more or less aware of this, but the whole idea behind this workshop is not so much that we find a way for FDA to come and provide the training and the education, but raise the awareness in meetings like this so our stakeholders can go back and continue to stress this issue. And those of you that really believe that many of these issues can be rolled into the concept of safety for the machine, please write to us and let us know, and maybe there's a way to roll it into what we call the safety and effectiveness bundle for what we require for our scanners. That's something that we look to you for input.

The second major area is MAUDE data. It seemed a little contentious. There were different views on the trends, whether it was stronger than the procedural trend or whether there's 500 percent, but I think we need to go back and look at the basis, look at the assumptions and look at the data, and try to make some sense out of exactly what's going on with those numbers. No matter whether it's 100 or 200 or 500, we still have a responsibility to bring that down to 0, and that's something that's understood. And most of the incidents are preventable. So we have a responsibility to see how we can bring that down.

Furthermore, it is disturbing that at least from Dr. Kanal's presentation, that large groups are somehow not reporting the information. So the whole area of what is reportable and how we can ensure that everybody reports in will probably improve our database and our knowledge of what's exactly happening so that we can take some action about it.

Ferromagnetic detectors, again a lot of discussion about the practice aspects, and what we heard was sort of consistent with what we are finding in our small survey. Everyone seems to support use as an adjunct for these devices as an integrated part of the safety. There was no clear consensus that this device needs to be regulated at this time, but it appears that we could benefit from some practice standardization.

And lastly, but not least, last but not least, technology. Technology is marching ahead. We heard a couple of science talks on modeling, computer modeling for SAR, and one thing interesting was the amniotic fluid region has 4 W/kg, and also in the pig studies -- show that continuous wave, 50 watts increases the brain temperature slightly, very interesting.

And then we heard about the recent developments in the MR technology and all the various areas of use, which reminds us again that we have to be vigilant about safety in the face of technological innovation.

Okay. So with that, I'd like to close today's session and thank you all for sitting through this long session and your valuable input, and I'm looking forward to tomorrow's excitement. Thank you. (Whereupon, at 5:30 p.m., the meeting was adjourned.)

C E R T I F I C A T E

This is to certify that the attached proceedings in the matter of:

MRI SAFETY PUBLIC WORKSHOP

October 25, 2011

Silver Spring, Maryland

were held as herein appears, and that this is the original transcription thereof for the files of the Food and Drug Administration, Center for Devices and Radiological Health, Medical Devices Advisory Committee.

____________________________

SUSSY MOREHOUSE

Official Reporter