P-R-O-C-E-E-D-I-N-G-S

9:01 a.m.

CHAIR HARVEY: Good morning. The meeting of the National Mammography Quality Assurance Advisory Committee has begun. First of all, Dr. Finder will discuss conflict of interest statements.

DR. FINDER: The following matter addresses conflict of interest issues associated with this meeting and is made a part of the record to preclude even appearance of any impropriety. To determine if any conflict exists, the agency reviewed the submitted agenda and all financial interests supported by the committee participants. The conflict of interest statute prohibits special Government employees from participating in matters that affect their employer's financial interests.

However, the agency has determined that the participation of certain members, the need for the services outweighs the potential conflict of interest involved is in the best interest of the Government. Therefore, waivers from the informed participation in general matters coming before the committee have been granted to certain participants, this is their financial involvement in facilities that will be subject to FDA's regulations on the mammography quality standards with the accrediting and certifying respecting bodies with many factors of mammography equipment or with their professional affiliations, since these organizations get back to the Department of Agency Deliberations.

These individuals are Nancy Ellingson, Alisa Gilbert, Maryanne Harvey, Linda Pura and Dr. Carolyn Hendricks, Debra Ikeda, Andrew Karellas, Etta Pisano and Donald young. The waivers are currently on file for Melissa Martin, James Camburn, Amy Rigsby, Drs. Catalina Ramos-Hernandez and Miles Harrison. Copies of the waivers may be obtained from the agency's Freedom of Information Office, Room 12A-15 at the Parklawn Building.

Several of our members also reported that they received compensation for lectures they have given or will give on mammography related topics. However, they have affirmed that these lectures were offered because of their expertise in the subject matter and not because of their membership on the committee.

We would like to note for the record that if any discussion of states or certifying bodies wants to take place in any meetings of the committee, it would be a general discussion only. No vote would be taken and consensus sought. In the interest of getting as many new points as possible, all SGEs including state employees would be allowed to participate in the general discussion, so that all view points would be heard.

In the event that the discussions involve any other matters not originally on the agenda in which an FDA participant has a financial interest, the participant should excuse him or herself from such involvement, and the exclusion will be noted for the record. With respect to all other participants, we ask, in the interest of fairness, that all persons making statements or presentations to disclose any current or previous financial involvement with accreditation bodies, stated so in your concessions under contract to FDA, certifying bodies, mobile units addressing implant imaging, consumer complaints and mammography equipment.

CHAIR HARVEY: Thank you. For our first item of business this morning, we would like to take a moment to introduce ourselves to each other and to Dr. Hendricks, who is joining us for her first day. So, Andrew, would you start, please?

DR. KARELLAS: I'm Andrew Karellas from Emory University, Department of Radiology. I am a medical physicist specializing in x-ray imaging and mammography.

DR. IKEDA: I'm Debra Ikeda from Stanford University Medical Center. I'm Director of Breast Imaging. I'm a radiologist.

MS. GILBERT: My name is Alisa Gilbert. I'm a seven year breast cancer survivor. I'm a community representative. I'm also the co-founder of the Unbroken Circle, advancing cancer care and services for the American Indian and Alaska native populations.

CHAIR HARVEY: I'm Miles Harrison. I'm a general surgeon from Baltimore Sinai Hospital, Hopkins system. I am involved in breast cancer care and my practice was involved in starting a non-profit Sisters Surviving Breast Cancer support group.

MS. PURA: Good morning. I'm Linda Pura. I'm a clinical coordinator with the California Department of Health Services, Cancer Detection Program, and the president and co-founder of the Los Angeles County Komen Foundation.

DR. PISANO: Hello, I'm Etta Pisano. I'm a radiologist in breast imaging and professor of radiology and biomedical engineering at the University of North Carolina in Chapel Hill.

DR. FINDER: I'm Charles Finder. I'm the executive secretary of this committee. I'm also a radiologist.

CHAIR HARVEY: I'm Maryanne Harvey and I recently retired from 32 years of service in Government, and now I'm working part-time for the New York State Department of Health.

DR. HENDRICKS: I'm Carolyn Hendricks. I'm a medical oncologist. I practice in Bethesda and I specialize in breast disease.

MS. MARTIN: I'm Melissa Martin. I'm a consulting medical physicist in the southern California area, and we currently provide the medical physic services for around 200 mammography facilities.

DR. YOUNG: I'm Don Young from the University of Iowa. I'm a medical professor of clinical radiology and have been the director of the Breast Imaging Diagnostic Center.

DR. RAMOS-HERNANDEZ: Catalina Ramos. I am a community representative with the Y-Me National Breast Cancer Organization and also at the University of Illinois, the Midwest Latino Health Research Training and Policy Center.

MS. RIGSBY: I'm Amy Rigsby. I'm a radiological technologist specializing in mammography, and I'm the technical director of The Rose in Houston, Texas.

MR. CAMBURN: I'm Jim Camburn. I'm chief of the Radiation Safety Section in the State of Michigan, and we regulate mammography facilities across the state, many of them on behalf of FDA.

CHAIR HARVEY: Thank you and welcome. For our first order of business, we'll have the study of FDA's Advisory Committee meetings, and our presenter is Dr. Katherine McComas. Good morning.

DR. MCCOMAS: Thank you and good morning. For the record, I don't have any conflict of interest really. I'm here before you today in collaboration with the FDA to seek out public understanding and perceptions on conflict of interest procedures that the FDA wishes to monitor and manage potential conflicts of interest of its Advisory Committee members. So I'm responsible for all the questionnaires which were seen on your chairs, and I also distributed a questionnaire to the Advisory Committee members as well.

This is a voluntary study and it is also anonymous. It's being conducted across multiple centers at the FDA, and we're really seeking to understand what people know and understand in order to offer recommendations for improvement when it's done. It is being funded by the Joint Institute for Food Safety and Applied Nutrition, which is a joint FDA University of Maryland Institute, but it's not relative to just escape. Obviously, we are collecting at CDRH, CDER, CBER and CFSAN, to throw in a few acronyms in the morning.

So I will be around today, as well as my research assistant. We will be here this afternoon. If you've got any questions, I will be happy to answer them. If you have a chance to fill it out today, that would be terrific. There's a box out at the registration desk so you can drop it in. Otherwise, there's a business reply envelope that you can drop it in as you leave here as well. And again, thank you very much for your time and, please, respond. The more responses we get, the more accurate an assessment of what people know. Thank you very much for your time.

CHAIR HARVEY: Does anyone have any questions, at this time, for Dr. McComas?

DR. MCCOMAS: No?

CHAIR HARVEY: Thank you. Next on our agenda is approved alternative standards.

DR. FINDER: For those not familiar with this section of the regulations, which is 900.18, FDA may approve an alternative for quality standard under section 900.12 when the agency determines that the proposed alternative standard will be at least as effective ensuring quality mammography as the standard proposes to replace, and the proposed alternative standard is still limited in its ability to justify and amendment to the standard or it offers an expected benefit to human health that's so great that the time required for amending the standard would present an unjustifiable risk to human health, and the granting of the alternative is in keeping with the purpose of statute 263(b).

Since the last August meeting, the division has approved one alternative standard dealing with combined mammography medical outcomes audit for multiple mobile mammography units. FDA approved this alternative standard on November 4, 2002, and amended it later on December 3, 2002. The amended alternative became effective on a latter date. Some accreditation bodies credit each mobile unit separately, even if two or more units are under the same ownership. This approach leads each mobile unit in certified separate facilities.

Therefore, the alternative requirement became effective and separate mobile medical outcomes audit had to be performed for each unit. This alternative allows owners of multiple mobile mammography units to perform a combined mammography medical outcomes audit for all units if the specified conditions are met. The original standard states that each facility shall establish a system to collect and review outcome data for all mammograms performed, including follow-up on the disposition of all positive mammograms in correlation with pathology results with the interpreting physician's mammography report.

Analysis of these outcome data shall be made individually and collectively for all interpreting physicians at the facility. In addition, any cases of breast cancer among women in the facility that subsequently become known to the facility shall prompt the facility to initiate follow-up on surgical and/or pathology results and with review of the mammograms taken prior to the diagnosis of a malignancy.

The approved alternative, as amended, is each facility shall establish a system to collect and review outcome data for all mammograms performed, including follow-up on the disposition of all positive mammograms in correlation with the pathology results with the interpreting physician's mammography report. Analysis of these outcome data shall be made individually and collectively for all interpreting physicians at the facility.

In addition, any case of breast cancer among women imaged at the facility that subsequently are known to the facility shall prompt the facility to initiate follow-up on surgical and pathology results and reviewing mammograms taken prior to the diagnosis of malignancy. This is the new part here. In situations where multiple mobile mammography facilities are under the same ownership, they may be treated collectively as a single facility for the purposes of meeting these requirements if all of the following conditions are met.

Each facility must consist of a single mobile mammography unit. The same entity or group administers the operation of all the included mobile facilities. The same lead interpreting physician has the responsibility for assuring that all will be included and all of the facilities meet the requirements of 21 C.F.R. 900(b) correctly. The same group of radiologists read all of the images from all of the included mobile facilities and all of the included mobile facilities provide services to the same patient population. That was the only alternative that we approved since the last meeting.

Anybody have any questions? Yes?

DR. PISANO: How do you define patient population? And that's with the QUS, isn't it?

DR. FINDER: In this case, all these mobile facilities go from the same site. They rotate through the same sites.

DR. PISANO: So you couldn't have like, let's say, I wanted to do it at USC and I had one going to eastern North Carolina and another going to western North Carolina that wouldn't qualify that?

DR. FINDER: Probably not.

DR. PISANO: That's a shame. Because I mean, it seems like those populations are really not significantly different. I mean, there are racial and then differences between the two geographic areas, but it meets every other criteria. It seems like it would be a shame that we had to not meet this criteria.

DR. FINDER: I think you need to show that they are the same populations or similar populations and that would apply. I was going on the assumption that you were talking of different areas than what you submitted due to population, but if it's similar, then it will.

CHAIR HARVEY: Any other questions? Thank you. This is the period where we'll open the public hearing.

DR. FINDER: We have two public speakers. The first one sent in written materials and asked they be read into the record. The next is a representative from NEMA. So let me start with the first one and read into the record the letter. It's a three page letter, so this is going to take a little bit of time. Okay.

"My name is Kathy McKinney-Tovar and I am a 44 year-old woman. In November 2000, I was diagnosed with breast cancer after being misdiagnosed for two years. I would like to share the events that led to my diagnosis, because they are pertinent to the agenda of the April 28, 2003 meeting of the National Mammography Quality Assurance Advisory Committee. I appreciate you sharing my letter with the committee, as I am unable to personally attend the meeting.

My story begins with my first mammogram which was made in 1998; I turned 40 that year. Radiologist A interpreted my 1998 and 1999 mammograms. A second radiologist, Radiologist B, also appears to have reviewed my 1999 mammogram as his name is on the report. In both years the radiologist(s) made observations about microcalcifications in the left breast and considered them to be benign and recommended no follow-up other than routine annual screening. A benign cluster of macrocalcifications was also noted in the upper, outer left breast in the 1998 report.

The 1999 report refers to the same feature as a loose grouping of small benign microcalcifications and states that the calcifications are stable. Any radiologist who reads mammograms should know there is a huge difference in the significance of macro versus microcalcifications. In 2000, my mammogram was made at a different mammography clinic, because my insurance had changed.

The new radiologist, Radiologist C, indicated that there was a dominant mass with associated microcalcifications in the upper, outer quadrant of my left breast, and that the lesion was now more easily recognized than on the two previous mammograms, especially the 1998 films where no discrete mass was identifiable, although some of the calcifications were clearly present. He wrote that a small component of the nodule was evident in 1999. His findings led to a biopsy and my breast cancer diagnosis. From my story, you can see that the problem in detection was not with the quality of the films, but with the interpretive skills of Radiologists A and B.

After my first mammogram, I discovered a pea-sized lump in the upper, outer quadrant of my left breast where Radiologists A and B had noted a cluster of calcifications. I saw a surgeon who examined the lump and told me it was a benign cyst and ordered no further follow-up or evaluation, because I had recently had a mammogram. After the second mammogram, I returned to the surgeon again, because the lump had grown. He assured me that it was benign and once again recommended no further follow-up or evaluation.

Although he passed away before I was diagnosed, I believe that he relied on the mammography reports of Radiologists A and B, who misread my films, when he concluded there was nothing suspicious about my lump. On June 24, 2002, the Archives of Internal Medicine published an article[1] stating that surgeons who relied on benign mammography reports when deciding not to biopsy a mass were responsible for 30 percent of physician-caused delayed diagnoses. You can see from this article and my own experience that the problem of having a misread mammogram becomes an even bigger issue when surgeons rely on mammography reports when forming their opinion about whether or not a biopsy is warranted.

By the time I was finally diagnosed, my cancer was Stage IIB, not an early catch. If my cancer had been caught two years earlier when the evidence appeared on my first mammogram, it would have been caught at Stage 1 or possibly Stage 0 when the treatment may have been limited to a lumpectomy and my chance of being cured would have exceeded 90 percent. As it was, I followed the suggestions of my doctors, who recommended the following treatments in an effort to save my life: Mastectomy, chemotherapy, radiation therapy, Tamoxifen and Lupron. I hope and pray that I won't die of breast cancer because of my delayed diagnosis and that there will be no long-term side effects of the treatment that will injure my health or cause my death.

Prior to my diagnosis, I, like all women I know, assumed that if I went for an annual mammogram and promptly reported breast lumps or other abnormalities, I could rest assured that breast cancer would be diagnosed in the earliest stages when the survival rate is very high. I now know that I was wrong. People in the medical profession have told me that a misdiagnosed breast cancer is all too common a story.

Because breast cancer is so frequently misdiagnosed, I began to question the value of mammography. I researched the issue and learned that mammograms are very difficult films to interpret, but they are still the best screening tool we have. I also learned that the biggest obstacle to getting an accurate reading is the skill of the radiologist. The New York Times ran an article[2] on June 27, 2002 regarding that fact. The article reported that while the best radiologists will miss as many as 10 percent of the cancers, some radiologists are missing as many as 40 percent!

While researching the issue, I learned that the skills of radiologists reviewing mammograms could be improved if the following actions are taken:

1. Test radiologists at regular intervals to ensure that they are qualified to read mammograms. Require additional training for any radiologist who does not pass the skills test and prohibit those who cannot pass the test from reading mammograms. Look at the training that Swedish radiologists receive, because they have the highest breast cancer detection rates in the world.

2. Devise a tracking system that will keep accurate records of how many cancers a radiologist misses and provide concrete feedback to radiologists who misinterpret a mammogram, so they can learn from their mistakes. Any radiologist who is found to miss more than a "reasonable" number of cancers should be prohibited from continuing to read mammograms.

3. Require that mammographers read enough mammograms on a yearly basis to remain as accurate as humanly possible. Studies suggest that 2,500 is the minimum number of mammograms that need to be read annually. Under current mammography regulations, physicians must review only 960 mammograms every two years, a number that studies have shown to be too low to be effective. Look to Canada and other countries that require their radiologists to read at least 2,500 films a year to learn how they provide mammograms to women living in poor and rural areas.

4. For each radiologist maintain records of the percentage of cancers they have missed and the number of mammograms they read annually. These records should be made available to all women who wish to have access to that information before scheduling a mammogram.

I urge the committee to implement the improvements listed above to the Mammography Quality Standards Act to ensure that only qualified radiologists are allowed to interpret mammograms. Don't let your mother, sister, daughter or wife be misdiagnosed by a radiologist who isn't qualified to interpret mammograms. Sincerely yours."

Since I will not be able to take any questions for this person, I suggest we move on to the next public speaker, unless somebody wants to make a comment.

CHAIR HARVEY: Any comments on the letter? I know it's always distressing to hear when the system doesn't work. It works for many women and works well. It doesn't always work, and there's some good points in that letter to be considered. Dr. Harrison?

DR. HARRISON: Reluctantly, I'll comment. In the patient's absence I'm not sure that my comment is going to fall on the ears that I would like to hear it, but in defense of the radiologists involved here, I think that at the point at which there was a palpable mass, I'm a general surgeon who receives 65 percent of my referrals are for breast lumps and evaluation of breast disease, and I always, when there is a palpable finding, especially in the face of associated calcifications, will do some kind of tissue acquisition right there at my office, be it a skinny needle biopsy, skinny needle aspiration biopsy or core, depending on how large the palpable finding is.

Again, I'm kind of preaching to the choir here.

CHAIR HARVEY: Yes.

DR. HARRISON: I wish the person were here. So in defense of the radiologists, I mean, you can't diagnose something until it's there to be diagnosed.

CHAIR HARVEY: Yes.

DR. HARRISON: And I think there was another opportunity to make that diagnosis that was not attributable to the radiologist not seeing something that had not evolved completely. I just felt the need to make that comment.

CHAIR HARVEY: Sure. Thank you.

MS. PURA: I, too, would like to comment on that. I think --

CHAIR HARVEY: Your name?

MS. PURA: Sorry. Linda Pura. When there is a palpable mass, I don't think you stop at mammography, and I think that the triple test comes into play here when you're looking at the kind of breast exam, you're looking at radiology and you're looking at pathology, and the only way to get the pathology is to stick a needle in there and get some tissue. So I, too, concur with you, doctor, and I tend to feel that the primary care clinician kind of let the ball drop here.

CHAIR HARVEY: Yes.

MS. PURA: Not only the radiologist.

CHAIR HARVEY: Thank you. Any other comments? You want to say something, Dr. Ikeda?

DR. IKEDA: Yes, Debra Ikeda. I wish we could address the person themselves, and it's distressing to hear her thoughts on mammography. As you know, mammography is not a perfect test, and that we will be discussing some of the issues she wishes to address in her letter, and it's hard to make any comment about calcification micro versus macro calcifications are not really as important as the shape and change over time, so it is hard to address this person's concerns.

I do have to say though we will be discussing some of the issues later on, and I think that maybe that discussion can be held, and my comments can be held until that discussion comes up.

CHAIR HARVEY: Thank you. Are we ready, Dr. Finder? Okay. Our next guest NEMA. Please, introduce yourself.

DR. SANDRIK: Hello, I'm John Sandrik. I am employed by GE Medical Systems. We manufacture, sell and service medical equipment, including mammography units. I'm also the chairman of NEMA Mammography Subcommittee of the X-ray Imaging Section of NEMA, and I'm speaking on behalf of the members of that subcommittee this morning, and I thank you for the opportunity to do that.

DR. FINDER: Do you want to state what NEMA is?

DR. SANDRIK: Oh, pardon me. The National Electrical Manufacturers Association, a group of manufacturers making equipment, in this case for medical purposes.

The comments generally address what has been included in the Draft Guidance Document #6 on AEC testing. We have broader and more detailed comments that will be submitted before the comment period ends. But today, I wanted to address two specific issues. One, the definition of configuration, and a somewhat related issue regarding the intent of the AEC performance test.

Question 1 of the Draft Guidance is intended to address what is a configuration. However, what it does is quote the rule, which somewhat identifies configurations in terms of example, but never really defines configuration. As mentioned, it says you have to do testing in all combinations of equipment configuration provided, for example, grid, nongrid magnification, nonmagnification and various target-filter combinations. However, there is no actual definition of what a configuration is outside of this list of examples.

What I'm suggesting here is that these examples are not consistent. Most of them involve clinical choices intended to achieve a specific medical imaging goal, while the others, the target- filter combination is really an element of the Automatic Exposure Control System, which spans a range of clinical applications. What we suggest here is a definition for configuration. The collection of system elements and their geometric arrangement selected by the operator to achieve a specific clinical imaging purpose.

Under this, we would include the contact of grid configuration, the magnification of nongrid. We would also like to include image receptor size, that is either the use of the 18 X 24 or 24 X 30 image receptor. We feel that these are all common in the element that it involves conscious choice by the operator to set up the equipment to solve a particular clinical imaging problem.

This is not the case generally with the choice of target-filter combination. The target- filter is usually at least proposed or in some cases automatically selected by the Automatic Exposure Control System of the unit, in some cases without any operator intervention whatsoever. Another point to consider is when performing the AEC performance test, the rules state that the kVp should be adjusted appropriately for the thickness of the object being tested.

We would suggest that the target-filter falls in the same rule as a means of controlling the x-ray spectrum just as kVp does. Hence, you could have a very plausible scenario where the molybdenum-molybdenum target-filter combination could be chosen for the 2 centimeter thickness. Molybdenum-rhodium combination for the floor, and a rhodium-rhodium combination for the 6 centimeter thickness. Hence, each thickness would be its own configuration. The mean optical density and the range would be identified by 1 point at each thickness. We would find it's not a very practical or meaningful measurement.

The consequences of identifying something as a configuration indicated here and from the guidance that's been published or proposed, within a configuration an optical density range of plus or minus .15 is set as the limit. We would view this as a challenge for the Automatic Exposure Control System, and the unit would be calibrated to try to achieve this goal.

On the other hand, between configurations a density range of plus or minus .3 is applied. We would suggest that, in this case, we're dealing with differences that are outside the control of the Automatic Exposure Control System, that is the difference in speeds of the cassette and the film associated with the different images. We suggest that this is consistent with much of what's in the answers to questions 4 and 7 in the Draft Guidance #6.

The other item that we would like to address and it is, as you'll see in a bit, related is what is the intent of this performance test in the first place. Is the MQSA Agency Performance Test an evaluation of the x-ray unit or the total facility capability, that is x-ray unit, screen unit, screens, film and processing? There is a suggestion of this intent published in the original preamble to the rules back in 1997 when, in fact, the issue of film variability was raised, and in that context of the film variability, would it be possible to see AEC performance at a plus or minus .15 level? And FDA's response to this comment was because film variability can be eliminated as a source of bias in the AEC Performance Test. There is no justification for increase in the AEC action limits. So there is a suggestion that FDA intended this performance test to be for the unit only and not a facility wide test.

Further suggestion of this is included in the answer to question 6 in the proposed Guidance. FDA states because the AEC Performance Test involves many parts of the imaging chain, the medical physicist needs to make sure that the AEC is a part responsible for the failure. Problems with the processor film emulsion or the use of different cassettes may lead to a failure that is not the fault of the AEC. This does further suggest in the AEC's performance as part of the mammography unit still falls short of an explicit statement.

Looking now at the context of the variability we're dealing with, probably many of you have seen this quote before, and, in fact, it was mentioned in the preamble from the CDC, ACR recommended equipment specifications for mammography equipment. A density difference of .3, and this is at a density of about 1.25, between any two films of the same type, and the same manufacturer exposed and processed together is a reasonable maximum to be expected from manufacturing variability for films of the same age and storage conditions.

This statement has been amplified somewhat in a later publication and the Kodak Technical Bulletin released a note that a difference of .3 at a density of 1.25 translates into a bigger difference for clinical films exposed at a great OD. For example, high contrast mammography film, such as KODAK MIN-R 2000 film are frequently exposed in an OD of between 1.5 to 1.7 in order to maximize contrast. The density difference of this OD level may be greater due to the increased contrast.

Another element of this context is a screen variability. In fact, by regulation it states uniformity of screen speed of all the cassettes in the facility shall be tested and the difference between the maximum and the minimum optical density shall not exceed .3. A case study, based on some recent discussions, let's call them discussions, between a field engineer and a physicist, was based on looking at some of these mixed things of screen-film, cassette size, film types, whatever.

The top two rows identify the elements used by the physicist in testing the AEC system. Indicated here is that he mixed both the 18 X 24 and the 24 X 30 image receptor sizes. The screens, the cassettes he chose for the test indicated by his uniformity study provided an optical density of 1.69 for the small cassette and 1.81 for the large cassette. So in a regime where we are required to be within plus or minus .15 optical density for the AEC, we've already introduced a bias of .12 in the density between the two image receptor sizes.

The field engineer did a recalibration and then tested the system using two matched cassettes and found that the system passed, but there was some discussion of whether this was really what was intended by the rule or not. Basically, the question should the bias in optical density introduced by the difference in screening speed be eliminated when testing the AEC?

So reviewing and again questioning what is the intent to have the screen variability density? Optical density, which is regulated and tested, there is a film variability of .3 density or more not regulated, not monitored. There's a processor variability. The requirement is the mid-density not move more than plus or minus .15 OD checked by once a day monitoring, but not indicated that it should be part of the mammography performance test when evaluating the AEC.

Within that view, we have the AEC performance required at plus or minus .15 OD. Again raising the question is a mammography unit expected to hold plus or minus .15 without control of other variables? So our request to FDA would be to develop a definition of a configuration, including equipment or image receptor size, excluding target-filter combinations and applying the plus or minus .3 OD limit between configurations. Also, then clarifying the intent of the AEC Performance Test and our indication being that it should apply to the mammography unit alone, control of or correction for other variables should be expected, and then to develop guides of good practice in performing the test.

We note that in questions 3 and 4 of the Draft Guidance rather detailed procedures have been provided on calculating the mean optical density, checking the action limits and evaluating the performance within and between the configurations. We would suggest perhaps that in question 6 would be a convenient location for putting in an explicit statement of what the intent of the AEC Performance Test is, and to provide additional detail on how to control or correct for the various affects that could finally effect the final result of the AEC Performance Test.

Thank you, and I'll be glad to answer any questions if you have some.

CHAIR HARVEY: Any questions? Any comments?

DR. YOUNG: Don Young from Iowa, and I want to thank you for your constructive and critic comments.

DR. SANDRIK: Thank you.

DR. FINDER: One thing I just want to say. You will be around later in the afternoon?

DR. SANDRIK: I will.

DR. FINDER: In case there are any questions that come up during the guidance discussion, I presume?

DR. SANDRIK: I will.

CHAIR HARVEY: Dr. Karellas?

DR. KARELLAS: This is Andrew Karellas. What John presented makes a lot of sense, and we encounter some of these problems in the survey of mammographic units. Sometimes we find these problems and we try to go around, especially on the issue of non uniformity of the screens. You want to ensure that this system behaves very well on the nonconformity part, because other subsequent tests will fail. But it is true that it would be very difficult to make the system comply if you just go randomly and you just test and you let it go, then you will not be able to use the specifications.

CHAIR HARVEY: Thank you. Ms. Martin?

MS. MARTIN: Melissa Martin. I'm the other medical physicist on this panel, other than Dr. Karellas, and yes, I would reiterate what has just been said. Reality is a lot of this may be education of the medical physicist on how to perform this test, and that's what I think John was showing there. If the physicist does not eliminate the variability in the cassettes, then they are introducing factors into the AEC testing that really are not the responsibility of the manufacturer. It's really not a test of performance of the unit. It's a test of cassette uniformity within the department, and that's a different test.

CHAIR HARVEY: Yes.

MS. MARTIN: So I would agree. Using the suggested factors, I have not found that a problem with using the .3 between configurations if you eliminate the variability between the cassettes.

CHAIR HARVEY: Those are good points to remember also for the inspector that often the problems that are found are caused by variations and how the testing was set up, rather than the equipment's performance. Thank you. Any other comments? Thank you. All right.

So that completes our section on open public hearing, that's correct. We're scheduled for a break right now, but I think we're just getting into the meat of the matters, so I think we'll sort of postpone that until people look like they are in need of a break.

So now we move into the period of open committee and we will begin with an overview of MQSA inspection findings and current inspection follow-up actions. Our presenter is Michael Divine, who is in the Inspection and Compliance Branch of the Division of Mammography Quality and Radiation Programs. Michael, are you ready? It's the last thing you do.

MR. DIVINE: My name is Michael Divine. I'm with the Inspection and Compliance Branch of the Division of Mammography Quality and Radiation Programs. If you were here last year, and many of you were, you will see that this presentation is very similar to the one last year. This is basically an update on what's been happening since then, and some of it will actually be almost identical to last year, since it's an overview of some of our actions.

There are basically two main parts of this discussion. One will be a set of graphs that go over the inspection data, what we find in facilities as far as problems as it has changed over time, and the other will be things that we do after the inspection, which is certain types of actions in terms of dealing with mammography problems.

Now, the difference between this presentation and last year is the last presentation was in August, and the Federal Government operates on a fiscal year that starts October 1st, so last year's data for the portion of the fiscal year that we were discussing was pretty much complete, so we had mostly a complete year to look at. Right now, we're about halfway through the fiscal year, so when we get into the discussion of the most recent data, I would suggest that you just consider that that's about half of what we normally see, and what I would like to also point out is that our data is showing that the problems have been decreasing over time, so if you sort of multiply this fiscal year's data by 2, I would think that that's a little over estimate, because we expect that since the data has been decreasing and the problems have been decreasing, that it won't actually be double what you can see on the chart.

Now, this is an overview of basically what we do with inspections in terms of classification of the types of problems. Level 1 is the ones we consider to be the most serious, that theoretically could result in a facility getting what is called a warning letter, that basically tells them that we find this to be a serious problem, and if they don't correct it, they could face certain types of actions. Level 2 is less significant. We don't send warning letters usually for these types of problems, but we do ask that the facility respond to us within 30 days explaining how they are going to correct the problem, and then we evaluate that response.

With Level 3, this is the highest level of finding or observation that we see during an inspection. We generally don't expect the facility to respond. The facility can respond, and many do, and we will evaluate that response, but we usually don't require it in this particular case.

Now, this is a slide that's pretty much identical from last year and it also shows the trend has been continuing in terms of improvement in facility compliance with our regulations. If you look at the blue line at the top, those are the inspections where no problems were found and would be put on the inspection report. The yellow line is Level 2, that is decreasing, Level 3, and the best one of all is the one at the bottom, which is very low, which is the Level 1 and more serious problems. That has also been decreasing, but because of the scale of this chart, it's hard to see that change, but I can assure you that it's there.

Okay. The first thing I would like to talk about is personnel issues at Level 1, and as you can see on this graph the number of facilities, we do roughly about 9,000 inspections a year. So if you look at the actual numbers here, and almost all the charts you're going to see from now on is the actual number of facilities that got this type. The numbers are not very high compared to just the 1 percent or 2 percent of the total number of inspections.

The first one there is interpreting physician. This is the requirement that they have board certification or the alternative of two months, if they had qualified before our final regs went into effect, or three months if they qualified after the final regs. And as you can see, the numbers are going down, as are the ones for physician license, which has dropped dramatically. The physician license issues are mostly regarding expiration of the license that is really a technical problem, rather than being any substantial problems.

As I mentioned last year for the medical physicist requirements, those are almost non existent and don't even really show up much on this graph. The technologist problems have also been decreasing over time. Another Level 1 problem, disregarding the Quality Assurance Program for the facility and specifically the Quality Control Testing, we are also seeing a drop in the numbers over time. The first one is if they have a string of days that are consecutive where they failed to do the processor test. The second one is if they did the processor test, the data shows that it was out of the actual limits, but they didn't have any documented corrective action about what they did about that.

The third one is just a number of days that don't necessarily have to be consecutive, but there was a large number. Basically, when we look at the processor data for missing data, what we're generally looking at is we look at the entire year, we pick the worst month out of the year, and then we count the number of days missing in that particular month to arrive at these numbers.

The last one is for Phantom QC missing, and as you can see that's dropped quite a bit, but if you were to multiply it by 2, even if that was the final year data, it's still going to show a decrease. This is our phantom image data. These tests here are actual tests that are done during the inspection. Now, this is Level 2 data. We do have Level 1 for phantom testing and for under processing, but those numbers are so small they're not even really worth putting on the graph, and as you can see even at Level 2 almost all the data, at least right now, is going to be below 20 facilities out of about 9,000 inspections.

So this is a very good picture of how the facility is performing. At the day of the inspection when we go in, our inspectors evaluate the phantom, and they also do a test called STEP which is an evaluation of the processor performance. And as you can see, the under processing problems that we found with that STEP test have pretty much gone away altogether, which is good news.

We do find some fog problems, but we think that this number will continue to decrease over time. This is also Level 2 problems, and these relate to the annual survey and also an equipment evaluation, which is required whenever they bring in a new x-ray unit or processor or they perform a major repair. Now, a little explanation may be needed between these two graphs here, because they sort of look the same, and it really is the consequence of when we do the inspection.

The first bar, which has more of a problem, is basically because we look at the annual survey that they have, and we compare that date from the previous survey, and if it has gone for more than 14 months, which we use as our window to decide whether they have met the annual requirement, then the facility will get cited for that. The second set of bars here is actually when we go in to the facility on the day of the inspection, the last survey they have actually done is more than 14 months, so the reason this is lower is because it depends on the day of the inspection and when they would have that survey done.

The middle columns relate to when we look at the survey, we're looking to see if all the tests that are required to be done are there, and in some cases some of the tests are missing, but over time this has been dropping. The fourth one here, this is the unit evaluation not done. This would be if they had installed a new processor, actually, I'm sorry. If they installed a new x-ray unit since the last inspection, but they hadn't had an equipment evaluation done before they started using the unit.

The last one here is relating to processor, and that jump that you see between 2001 and 2002 was the consequence of a lot of our inspectors not knowing that this had to be checked during the inspection. That was corrected. Then there was, obviously, a big jump. That actually has dropped down, and we expect that to continue to drop over time.

Switching over back into personnel requirements at Level 2, this is for the interpreting physicians. The CME data is continuing to decline, less than 100 facilities right now. Initial experience also very low. These are things that are checked. Usually not checked every year, unless there's been a change in personnel. If you go back year after year, you see the same interpreting physicians, who generally don't ask to see those same documents if they've been checked previously.

However, for the last two sets of bars when we're talking about containment requirements, these are checked every year. The continuing experience this is the 960 mammograms that have to be read every 24 months, that has been less of a problem and has been declining. CME is a little worse, but that's also gotten better over time.

With the technologists, the first column here is an initial requirement related to whether they had -- right now, if they were qualifying that, it would be 40 hours of training of mammography. Under our interim regs it would have been just that they had to have training, but we didn't specify the number of hours. As you can see, this is very, very small compared to the number of inspections we do.

With the continuing experience, you see this jump here, and that's a consequence of the fact that these requirements took effect actually on this day in 1999, and we had to wait at least 24 months before we checked that. And, of course, when these started to get checked on a frequent basis, the numbers jumped up quite a bit. This number here was reflected by the kicking in during the fiscal year, and, as you can see, the numbers have been declining. Though, you know, it's hard to say what we're going to have, because if you double that, it might actually show an increase, but I expect that when we get the final year data it should show a decrease.

Continuing education similar to the interpreting physician. It's been showing a decline. Medical physicist, initial experience very similar to the data we have in the Level 1. Also, it's pretty much gone away altogether. Continuing experience, just like the technologist, this kicked in on the same time frame, so we did see an increase, but, as you can see, it has dropped quite a bit, and with the continuing education it has also been declining over time.

This is a mixture of Level 1 and Level 2 problems, and I want to point out that only the first set of bars pertain to Level 1. This is relating to the medical records issue, and most of what we're talking about here right where we say communication system is inadequate, is we're talking about they have to have an adequate system to send mammography reports out within 30 days and patient letters within 30 days. And this was a problem when this first took effect. Fortunately, by the time we see this data, things have dropped down to where it's just a very, very small number of facilities. And some of these are corrected very quickly.

We still see a certain number of problems with the assessment categories, and that, I think, was just a lot of facilities needed to change over the way that they communicated. Once again, that has improved quite a bit. And it's almost nonexistent the number of facilities where the interpreting physician is identified on the mammography report. This has never been a real big problem, but right now it has shrunk to almost nothing.

These are unit requirements. The x-ray unit specifically, and a couple of procedures that the facility must have as part of the quad insurance program. Unit missing features, this is the most prominent thing that we find with this problem, which is actually, as you can see, pretty much gone away, was that a lot of facilities only had one image receptor size. And that has decreased because a lot of facilities fix those problems. I think a lot of it is that a lot of the units have worn out and a lot of facilities have had to replace them, and almost every unit you buy these days comes with both of those as part of the package. So but I think even with the numbers that we did see, the facility basically was able to rectify the problem.

With the other two items, which are procedures that they must have, the first one is the tumor complaint procedure that they have to have in case of facility complaints about an issue regarding mammography. We find that that has been pretty much fixed, and it's only in a very small percentage of facilities, and the same is true for having an infection control procedure if the unit is contaminated with body fluids.

The next thing, and this is also a consequence of requirements kicking in years after the regulations took effect, because all of these bars, and there are six different inspection questions that we're talking about to deal with the medical outcomes audit and how well the facility is able to collect the data regarding biopsies and being able to factor them into an analysis. The first one, which has really been in effect since the start, which is basically they have to either obtain all the biopsy results for all their positive mammograms or at least document that they made an attempt to get all the results, and most facilities generally are able to get the results if they ask for them. This is, like I said, a very small problem right now. It has pretty much gone away.

The next one is similar. Actually, let me correct this. This one is if they were not entering all the positives into their system. The next one is if they actually had no biopsy results at all, and couldn't show that they made much of an attempt to get them. The third item here relates to the audit interpreting physician. They have to identify who is going to do the analysis. And this is simply a matter of identifying that person that's going to look at the data. As you can see, it was a problem at the beginning, because this was something that kicked in after the final regs took effect, but it has dropped down a bit where it's not much of a problem.

The last three are also requirements that took effect, and we found a lot of facilities. One is once they have collected data, they have to do an annual analysis, so a lot of what you see in this middle bar was the first time that they were supposed to do an analysis, and a lot of this has to do with facilities sort of getting used to the new regs and having to do what we're specifying what they have to do. And in most cases it's a matter of documenting that they have done this and being able to show the inspector.

This one here has to do with they have to break it down by each interpreting physician. Once again, this is something that they hadn't traditionally done, though a lot of facilities do, but we expect that to decline over time, because the data is usually available and it's not difficult to separate it out like that. And this is that they were not doing it for the entire facility. Once again, this is related to the first question, but it's actually a separate question.

Okay. This is sort of a catch-all, and the reason this number is so high, we're up into almost 1,000 facilities, is there are numerous things that a facility has to do and that we check during the inspection regarding personnel. We check all their personnel for the continuing requirements. We also check the new personnel for the initial requirements. So there's a lot of different things, and this is a Level 3, which is our lowest level, that if they had any problem at all with something missing at the time of the inspection, they will get cited for this then. So that's why this number is so high, but it is declining over time as are all the other personnel requirements. So we don't expect that that's going to be a problem in the future.

Okay. Now, I'm going to switch gears here and to a large extent. If you heard my presentation last year, you will find that this is pretty much an overview of the same thing that I talked about last year. The types of things that we can do, and this is beyond let's say just sending letters to the facility, such as a warning letter. These are actually, I would say, positive actions that we can take.

One is if the facility has had serious problems, under MQSA we have the authority to do a follow-up inspection, and we usually don't do these for every inspection when we find problems, but if we believe that we need to verify that the fixing of the problems, we have the option of doing a follow-up inspection.

Another type of action we can take, which actually correlates with the third item, which is something in our regulations called Additional Mammography Review, if we believe that the quality of mammography has been seriously compromised, we can request that the facility have mammograms and possibly mammography reports evaluated usually by their accreditation body or that we have the option of selecting a different reviewer. But the vast majority are done by the accreditation bodies, and that basically is to determine whether the quality of the mammography may represent a serious risk to human health, and so there are criteria for deciding, each accreditation body has a criteria, what that is and they are required to notify us if the Additional Mammography Review does rise to that level.

If it does, then we will go back to the facility and require patient and physician notification. And the physician we're talking about, the patient's referring physician, so the facility would have to notify each patient and physician regarding this problem with the mammography with recommendations about what they need to do.

The fourth item which we consider a regulator action, which is Additional Mammography Review. It's more less a public health type action. Whereas, the directed plan of correction we consider to be a regulatory action. Like it's actually classified under the law as a sanction. And basically a directed plan of correction is we notify the facility by letter that they have been notified in the past that they had serious problems, they did not correct the problems, and now we lay out a plan for things that they must do to correct their problems in the future.

And a lot of that has to do with sending to FDA films, such as phantom images, or records to show that they are continuing to comply with the requirements, and we can set a time frame under which they have to follow this plan. And they also have to send in a lot of detailed information about their procedures. We might actually require them to have strict management oversight of things that may be not necessarily having that kind of oversight at the lowest level in the facility. And usually, we impose this for a year. Though if a facility improves dramatically during the year, we may cut it off sooner.

Going a little stronger is we have the authority to charge civil money penalties against facilities for violations. This can be $10,000 for performing a mammography without an MQSA certificate. It also could be $10,000 per day or $10,000 per violation for violating the act. If we believe that the problems are so severe that the facility should no longer be doing mammography, we have two options. One is suspension of their certificate, and the other one is revocation.

Now, the differences of suspension and revocation is suspension will shut them down, but there's a possibility that if they correct their problems, we could actually lift the suspension. And there are two types of suspensions that we can do. One is where we notify the facility they have the right to a hearing before we take the action, and then if we decide after the hearing that we still want to suspend their certificate, that we would do it at that time or not if we decided not to.

With revocation it's different. I should mention the other part of suspension, excuse me. If we believe that there's a serious risk to human health based on the findings that we have obtained or we believe that the violations were intentional, we have the option of suspending the facility's certificate without holding a hearing, and that's more or less an emergency option that we have. But the criteria in the law is much stronger for doing that. Usually the option is that we give the facility a hearing before we take the suspension action.

With revocation we have to offer the facility a hearing prior to taking the action. The big difference with revocation is that it's such a severe action that the owner/operator of the facility cannot own or operate a mammography facility for two years. So it's somewhat of a permanent action that we think is needed if the violations are of that severity.

Injunction is the last resort action and hopefully with all these other options, we would never get to this point, but if everything else fails or we can't get the facility, for instance, to stop performing a mammography without a certificate, we can go to federal court and get an order that shuts the facility down, which is called an injunction.

With follow-up inspections, what we do is, let's say if a facility gets a warning about serious problems, we go in and check for corrective actions for the serious problems. Usually we wouldn't do a follow-up inspection unless there were Level 1 problems, though we have the option of doing it. If the other problems were sufficient in number and there were other issues that would not necessarily rise to the level of Level 1, we could do it. But the vast majority of cases where we've done follow-up inspections has been where we found Level 1 during the previous inspection.

This is usually limited in time and scope to looking at the specific problems. If we went into a facility and we found, let's say, that they hadn't been doing the phantom test and the processor test, but in general everything else checked out, we would-- if we did a follow-up inspection, we would focus on those problems that had been found, and we wouldn't necessarily have to do an entire inspection. The only time we would do an entire inspection with a follow-up inspection is if the problems were so widespread that we felt we had to check everything again. So usually the time to do these is much less than an annual inspection.

As I mentioned Additional Mammography Review, it's usually done by the accreditation body, though we have the option of going with another party, though that would be rare that we would do that. If that shows a serious risk to human health, there could be patient and physician notification, as I mentioned earlier.

One of the things we routinely do if we find a Level 1 phantom image during an inspection, we generally do a very limited Additional Mammography Review to see if there has been an impact on the mammograms. We could also do one if the interpreting physician turned out to be unqualified, though most of the time when we have had these Level 1 problems of interpreting physician, those turn out to be more documentation related, and we don't need to do the Additional Mammography Review.

Most of the Additional Mammography Review, is a consequence of the fact that they're not doing it or the interpreting physician problems, is we're basically looking at clinical image problems, which is basically looking at whether the mammograms have been effected. We also have the option, and we've done this in a few cases, where if we find that there's a major problem with the Quality Assurance Program, that they haven't done some of the tests for quite a long time, more than just a simple Level 1.

We sometimes will do Additional Mammography Review in those cases. And we've also done it in cases where we find that there was fraudulent record keeping. When we see falsified records, it really calls into question a lot of things, and even if you find that there's areas where you don't have problems, it's hard to really trust the records if some of them have been falsified. So we've done several AMRs in situations where there has been fraudulent record keeping.

Patient and physician notification, the trigger is that we almost always do an AMR before we do patient or physician notification, because we want to make sure that before we take this drastic step that we know that there is a serious risk that we have to deal with. This provides the patients and physicians with follow-up options. In many cases the patient may decide to have a new mammogram at a different facility or they may consult with their referring physician as to what they might want to do.

One thing we try to do with these letters is just like the patient results letter that's sent to the patient informing her of the results of the mammogram, this is a letter that we try to put in plain language, try to get away from any technical jargon, so that she clearly understands what happened and what she needs to do.

Directed plan of correction, I pretty much covered a lot of these items during my previous slide. But basically we impose additional requirements in the facility, and we do a monitoring program. Basically, they have to send in records, usually, with the directed plan of corrections. And if we find that they're still not complying with the plan, we have the option of going in periodically to check on things, so there will be additional inspections during the course of this directed plan of correction.

Suspension, as I mentioned, this is for serious violations, and usually this is where we may actually have warned the facility. We may actually have put them under a directed plan of correction. We've done a lot of things to try to work with the facility to try to get them to correct problems, but everything we've done up to that points seems to be ineffective, and we really think that the facility needs to shut down until they can correct their problems, and as I mentioned, health hazard, and there's also the option if we can show that the violations are intentional, then we have the option of suspending without holding a hearing in advance.

This is just a summary of the data to date for the entire program, Additional Mammography Review. If you had been here last year, there was just a few slide increases and a number of Additional Mammography Reviews since the last meeting. And with directed plans of correction, we've imposed one of those since the last meeting. And that concludes my talk.

CHAIR HARVEY: Thank you. Mike, can you tell us how many facilities that there are currently certified?

MR. DIVINE: I believe there is about 9,200.

CHAIR HARVEY: So we've lost about 800, 700? I remember 10,000 as a number sticks in my mind, so would that be a good estimate?

MR. DIVINE: Yes, it was close to 10,000 back in '94 when we first started certifying facilities. I'm not exactly sure what the decline has been, but obviously it has been quite a few.

DR. PISANO: There is a letter or there's a statement by Leonard Berlin in the file that we received. I'm Etta Pisano. And in here it says we've lost 700 facilities in the last two years. 700 units, I guess, no facilities. Because when we talk about facilities you mean units, right, not facilities?

MR. DIVINE: Well, usually we're talking about facilities.

DR. PISANO: We're talking about facilities.

CHAIR HARVEY: Facilities. Okay. So we've lost 700 units. I don't know if it's accurate, but that's what he said.

MR. DIVINE: Right.

CHAIR HARVEY: Right.

DR. PISANO: Facilities, okay.

MR. CAMBURN: Could I ask a question of Mike?

MR. DIVINE: Yes.

CHAIR HARVEY: Jim?

MR. CAMBURN: Jam Camburn. On your slide about interpreting physician qualifications?

MR. DIVINE: Yes?

MR. CAMBURN: We've noticed a problem in our state for a while, and I don't know if this is just unique to what we're seeing or whether it is in other states as well. Dealing with physicians who are locum tenants.

MR. DIVINE: Yes.

MR. CAMBURN: We had a hospital just last week that a year ago had a clean MQSA inspection, no violations at all. We went in last week for an annual inspection, 14 citations, 12 of which were Level 2, all pertaining to locum tenants.

MR. DIVINE: Yes, I think a lot of facilities don't necessarily look at them the same, though we do during the inspection, and I think the problem is is that they come in and then they leave, and then somebody says okay, we didn't cover that, and in some cases it may be difficult to get back in touch with the physician, he has moved on to another facility.

There have been cases where we've actually made an attempt to contact the firm that places locum tenants, because I think working with them is also important to try to correct these problems, because once they understand that when this person gets placed at a facility, even if it is for a week or two, they have to have a package of information going along. And hopefully that will work in some of these cases, because once a particular locum tenants firm understands what they need, they can make sure that all the physicians that they are going to place have this package of information.

But we find that that same problem is probably more prevalent than the established physicians, who have been on the staff at the facility for some time. Especially with the continuing requirements, which have to be updated continually.

CHAIR HARVEY: Thank you. Yes, Dr. Pisano?

DR. PISANO: This is Etta Pisano again. I'm curious, you know. The thing that strikes me from these data are that there aren't that many violations. I mean, it's very rare, so I'm wondering, and I don't know if you are the appropriate person to ask this of, but I'll ask it any way. Is there any thought about making this inspection process less onerous on the facilities and perhaps eliminating some of the things that are rarely violations?

Just speaking for my own facility, we basically have to shut down for two days to get inspected. So we run our facility, but we run it at under strength while different machines are being tested. So I'm wondering if we never have violations or very rarely, maybe random selection, random inspections instead of everybody getting everything inspected every year, has there been any thought about that?

DR. FINDER: It's Dr. Finder. Let me jump in and save Mike from answering the question. There are a couple of initiatives that are involved with that. One is the Inspection Demonstration Program in which looking at and inspecting "good" facilities, and I use good in quotations, that meet certain criteria every other year. We have, in the past, looked at the idea of shortening the inspection. And one of the disadvantages to that is that the major cost for the inspection really isn't the total amount of time there, but it's actually sending out the inspector.

So while it might save the facility a little bit, it probably wouldn't save them that much in both time and money. So I think that the Inspection Demonstration Program, which will allow good facilities to be inspected every other year, probably will accomplish the same type of goal. And we do realize, obviously, that the inspections impact patient care on those days, say nothing of the people preparing for it in the weeks before hand.

So we think that by allowing those types of facilities to be inspected every other year will get the most bang for the buck.

DR. PISANO: The other thing, I really wish we could get those numbers on the number of facilities, because I remember the same way that Maryanne does that we had about 10,000 last year, and now we have 9,200.

MR. DIVINE: No.

DR. PISANO: So my question is what has happened? Have we really dropped that many? I remember the year before last we had gone up or stayed about the same and this year seems to be a difference.

DR. BARR: This is Dr. Helen Barr. I have emailed our data people to try and get you those numbers as quickly as I can.

DR. PISANO: Great. Thank you.

CHAIR HARVEY: Dr. Karellas?

DR. KARELLAS: There are some situations, although not all of them, where there has been some consolidation, so it would be interesting to have not only the number of total facilities, but the total number of mammographic units.

CHAIR HARVEY: Units, right.

DR. KARELLAS: Now, I'm not implying that we do not have a few. I believe it is a fact that in many areas it is becoming more difficult to schedule a mammogram. So things have not been easier than a few years ago for the patient, but there is a combination of pure closing and some consolidation, so more detailed statistics would be good, especially if we could find some volumes of mammograms read by radiologists, and I think also the number of active units would be somewhat of a reasonable indication.

May I comment on the facilitation of the inspections? I believe that --

DR. BARR: Dr. Karellas, can I answer your data questions first?

DR. KARELLAS: Oh, sure.

DR. BARR: This is Dr. Helen Barr again. You're asking for a lot of information. I don't know how quickly I can get all of it. I can certainly supply it to you. It may not -- you know, it may be after this meeting. I can tell you related to the number of units whereas the average mammography facility had approximately 1.2 units in the past, that is now up to 1.7 units, so the average facility size is getting larger.

CHAIR HARVEY: Is that good stereotactic? Dr. Barr, does that include stereotactic and other -- no, this is just diagnoses. Okay.

DR. BARR: No, that's strictly --

CHAIR HARVEY: Okay. Thank you.

DR. BARR: -- mammography units.

DR. PISANO: Over what time was that change?

DR. BARR: Over the last couple of years. During the same time period which we've been seeing the decline of facilities, we've been seeing the average number of units per facility go up.

DR. PISANO: Okay.

CHAIR HARVEY: Dr. Karellas?

DR. KARELLAS: Thank you, Dr. Barr. On the audits, I think that facilities could report, perhaps annually or every two years, could submit a report in place of an inspection, that the report could be an audit of their operations. Corporations do that for other purposes of reporting, whether it's environmental or tax. And hopefully not a too detailed report could be done as a self audit on certain critical issues that would have participation of radiologists, the technologists or other interested parties, and that would give them the opportunity to sit down and assess what they are doing. Of course, random inspections should always be an option of the Agency.

DR. FINDER: This is Dr. Finder. I just want to point out that the requirement for on-sight inspections, annual, is in the statute itself. It's not a regulation. It's the law. And the Inspection Demonstration Program actually could not go on without them changing the law, which they did during the last three authorizations to allow that to occur.

CHAIR HARVEY: Yes, Dr. Pisano?

DR. PISANO: This is Etta Pisano. I want to clarify what I was suggesting. I'm not suggesting -- what I was suggesting was to look at the date that has already been collected over the last 10 years or so and figure out which tests are hardly ever failed, and then eliminate those as a routine inspection subject. And I think that actually might help more than the "good performing sites," because -- and I know it's very good what you're doing with the "good performing sites," but I think that it is quite a high cost to the facilities to prepare and to perform.

These inspections do cost a lot of money. The facilities both in terms of the time that they take for people, but also in missed opportunity costs, you have to shut your machine down for the inspection, so if we can even for a couple of hours, that's several mammograms. And I think if we could do our best to limit as much as possible, reduce as much as possible those costs, given the amount of data we have, at this point, I'm not suggesting the things that people really do continually fail should be dropped. I'm talking about things that rarely have ever failed, reducing that.

Not making it so it never happens, but randomly selecting sites and saying you must inspect this at this inspection. I'm suggesting -- the only reason I'm coming up with this suggestion is because we have just been through this with the digital QC program that we've imposed on the ACRIN trial in the DMIST study. We started out with a huge laundry list of inspection criteria that we decided every facility should do, because we didn't know what they should do, and now we've collected data for about two years or a year and a half as part of the trial, and we found that a lot of things never have failed or have rarely failed, and so we've cut back our requirements.

You have even better data than we have over many more years. So I'm just trying to make it less onerous on the facilities if things that rarely have ever failed, it would be nice to have some things cut back, and that's my suggestion.

DR. BARR: Yes, this is Dr. Helen Barr again. Dr. Pisano if we hear a clear indication from the committee that that's a direction you would like us to look in, certainly I would entertain that idea. What I would like to do first is get through the Inspection Demonstration Program, because even if we can present evidence that violation free facilities can go longer than a year and keep their violation free status without us being in their facility, I don't know whether Congress is going to let us go there.

If we get an indication from Congress that that's a way, if we can present that evidence that that's what the program shows, and Congress gives us an indication that we can go that way, I would certainly be willing to entertain looking at the actual inspection elements themselves.

MR. CAMBURN: Jim, do you have any comments from Michigan on this?

MR. CAMBURN: Well, in terms of the number of facilities, I can comment on that. We have seen a slow decrease in the number of mammography facilities over the years, as I think FDA has. We topped out at about 350 facilities. Now, we're down to about 330. However, the number of mammography machines has never been higher.

CHAIR HARVEY: Yes.

MR. CAMBURN: So we're seeing more machines per facility. More two and three machine facilities, some 10, 12 machine facilities. We're also seeing an increase in the number of mammography exams every year. So the facilities that still exist and are doing mammography on average are doing more exams. So these all seem to be kind of good trends in a way.

CHAIR HARVEY: Right. Looking for those facilities to do the most volume.

MR. CAMBURN: Yes.

CHAIR HARVEY: Right.

MR. CAMBURN: Yes.

CHAIR HARVEY: Do you have any other comment about the idea of cutting back on any aspects of the inspection?

MR. CAMBURN: Well, I agree with the comment that Dr. Finder made that a lot of our costs are getting to the facility, getting set up, doing some tests, and to do a few more come almost free for the inspector. And in terms of -- at least the way we schedule inspections in Michigan, they are all scheduled in advance, and in some cases far in advance. We'll have some facilities during one annual inspection actually say can we plan on you next April 14th for the inspection? We don't do too many a year in advance, but we do a lot of them several months in advance. So they are not really losing patients. They can schedule our inspections around the patients. And from what we've seen, we haven't really interfered too much with the day to day operations of the mammography facility.

CHAIR HARVEY: Thank you.

DR. IKEDA: This is Debra Ikeda. I have to tell you that to have the inspectors, we do shut down for inspection, and it's just a fact of life, and we cannot do patients on those days. Similarly, I must tell you that I employ a full-time quality assurance person just to deal with the regulations, just to deal with the paperwork, and to get ready for the inspection, and so it takes not only the time at the time the inspector shows up before our actual inspection, but it also takes radiologists time, quality assurance time, technologists time, so there is a lot of hidden costs, I think, that may not be accounted for by just accounting how long it costs to send an inspector out.

MR. CAMBURN: Yes, could I comment on that? Jim Camburn again. Yes, I agree there are costs like that, but in many of those cases, for example, the quality control time that's spent and having a quality control technologist, that's probably something that your facility should have any way of doing those tests and looking how the quality control is being handled, even if an inspector never walks in the door, because it's an essential part of high quality mammography to do those tests.

I know our inspectors, too, when they go into a mammography facility, they try, especially on the facilities that have more than one machine, they work around the mammography facility's schedule as much as possible. It interrupts to some extent, but I don't think we have any facilities that really feel they are closed down for a day.

DR. IKEDA: Dr. Ikeda again. My point was only that I would like the committee to recognize that there are costs associated with personnel to do the inspections and to maintain high quality, and that those costs are not insignificant. For example, they could be $40,000 a year plus benefits, etcetera, on top of that. And so it could be another reason why some facilities are not able to keep up, especially, if they are small.

For example, the cost per unit will decrease as each facility increases the number of mammography units. If they have one person already doing the QA, but if you've only got one machine out in some little rural place, you can't afford a $40,000 person. It might lost less in that rural facility, but it still costs time and money for the facility.

MS. PURA: Linda Pura. As a consumer advocate, I know we walk a very fine line between stringent and availability, but I would have to say that I have great concerns about minimizing inspections, even for those good facilities, and I realize that perhaps there are things we can cut back on, but I certainly say that closing down for one day may be a lot cheaper than the litigation that we had seen, and perhaps the decrease in reimbursement would be better off if we could increase it and perhaps make up for some of that. But I cannot see reducing the amount of inspections.

DR. BARR: This is Dr. Helen Barr. Just so that you know, Congress asked us to do that. That was a Congressional thing that we are required to be doing. I also wanted to say that some of the things that we've never seen a problem with are in the areas say dose. You know, we never see a dose problem. I think if the time comes when we're going to look at this, it's like a clear indication from the committee of what kinds of things, you know, would you be willing to give up does, you know, and I think we need a clear indication from the committee of what types of things you were thinking about if we got to the point where we were looking at this.

CHAIR HARVEY: Okay.

DR. PISANO: Again, Etta Pisano. Again, I want to emphasize though I'm not suggesting that you never inspect it. What I'm suggesting is that you select a random sample to inspect it at. So in other words, it would be less risky, so the facilities would still be monitoring it, presumably. I mean, we wouldn't want facilities to stop monitoring dose. And we would want them to be able to show you their dose. But you may not have to do measure the dose yourself at the inspections or look at their records of dose. That's a good example, I think, of something that perhaps it does add time to the inspection.

Again, I'm not advocating the inspections be eliminated. I actually think what I'm suggesting is a little more conservative than what you are actually doing, which is to continue doing the annual inspection, but figure out a way to make them shorter for facilities. Because it does cost money to close even for -- even though they plan for your visit a year in advance, they still have to stop taking images for a few hours, which does cost them, that's a cost.

So I agree you have made it minimally onerous or as little onerous as possible, but it's still onerous on the facility. It does cost them. It might make it less expensive. I agree it's worthwhile doing, but I would like us to see it move in the direction the way that we have with DMIST's QC Program for digital, which is when we realized the test never fails, rarely impose it or make it less, you know, do it in the subset of the patients -- of the facilities.

DR. BARR: Thank you.

CHAIR HARVEY: Thank you. I believe we'll have a break at this point in time, about 15 minutes. We'll return at 10:45. Thank you.

(Whereupon, at 10:31 a.m. a recess until 10:49 a.m.)

CHAIR HARVEY: Welcome back to our meeting. We've been requested, members of the committee, to, please, speak within four to six inches of the microphone, so that our words can be retained forever. It's easy to look around and lose track of the fact that you are no longer speaking into the microphone. All right. Dr. Finder is ready, I'm sure, to talk to us about good guidance practices and directions for discussion of the MQSA Guidance under the final regulations. Dr. Finder?

DR. FINDER: Okay. Before we begin our discussion of the proposed final regulation guidance, I would like to briefly explain the procedures that FDA is following as it develops new guidance.

In response to public comment regarding the use of guidance documents, FDA held an open public meeting on April 26, 1996 and again on February 27, 1997. They published a Federal Register notice outlining the steps the Agency needed to take prior to issuing guidance. In brief, it stated the following: "Guidance had to be developed in an open manner that permitted input from the general public and the regulated industry. In most cases, new or controversial guidance had to allow for such input prior to its implementation. While statutes and their associated regulations were binding and enforceable, guidance was to represent a way or ways of meeting the regulations, but other ways would be acceptable as long as they met the requirements of the regulations or statute."

Before we begin our discussions, I would like to emphasize the following: We are here to discuss the guidance, not the underlying regulations. Regulations have already gone through their own extensive approval process, and while they are subject to future change, the purpose of today's meeting is to address the proposed guidance. When you hear or see the word "shall, require or must," they refer to the underlying regulation. Whereas, the word "should, may or recommend" refer to guidance.

The guidance documents and many of today's presentations may be found at FDA's mammography website, and I can give you the website. It's www.fda.gov/cdrh/mammography/nmqaac.html. I also want to make mention of the fact that the committee is going to be looking at two different guidance documents, #6 and #7.

#6 was published as proposed guidance, and is actually open for public comment right now. It will be open until, I believe, the middle of May, so we are looking for your comments on that. That document mainly deals with Automatic Exposure Control. We also have document #7, which was published as a final guidance document, and by final I don't mean that it can never be changed. We certainly are interested in the committee's comments on that document. But it has gone out as final already. Certainly, it's open for change in a future revision.

So with that stated, why don't we begin with document #6, which is the AEC document. And I do want to make mention of the fact that as brought up by our public speaker earlier in the morning, this is a fairly complex topic involving this portion of the equipment, component of the equipment, and we have been writing guidance on this for many, many years, and the guidance has been changing and evolving over time. Some of it deals with the reality of how our guidance has been handled by the physicists out in the field by the facilities, and we're trying to make it clear and clarify some of these issues.

Other issues have come up as uses of AEC have changed over time and as the equipment has changed over time. You have to remember that the original regulations were written or they began to be written in 1996, so we're basically talking about regulations that may be as much as six or seven years old, and they may not always apply in every single case. So with that said, go ahead.

CHAIR HARVEY: We're open for comment. Ms. Martin?

MS. MARTIN: Melissa Martin, one of the medical physicists on the committee. I would like to bring to the attention of the committee one of the tests that the physicists do and it's the Automatic Exposure Control. I think it's on Page 3, actually, but it's section 900.12(e)(5)(i)(A), (B) and (C).

The request or comment is that it says that the physicist and basically the engineers, also, will test these Automatic Exposure Control Systems with "a homogenous material." For those that may not be aware, one of the common homogenous materials used, at least previously, Lucite, has been an accepted material. What has been a fairly common experience among many of the medical physicists is if the installation people set these units up with Lucite, when the medical physicists go in to test these units using breast-equivalent plastic, either 50/50 or 55/45 dense-to-fatty, the AEC systems fail.

I would like to recommend that the homogenous material not include Lucite, because if we're bringing this up to a standard, the breast- equivalent plastic, the purpose of these machines is to duplicate breasts, not Lucite, and Lucite really doesn't duplicate the performance of the breast- equivalent plastic, which is readily available any more for testing.

CHAIR HARVEY: Okay. You go ahead.

DR. KARELLAS: This is Andrew Karellas. I'm afraid that the only downside of that is the availability. It is widely available, but only from a couple of manufacturers, and it's probably more expensive, so I agree with you that the so-called BR12 or breast-equivalent material is perhaps more appropriate. But I would imagine that some manufacturers might find the poly methyl methacrylate is the generic name for Lucite.

One other problem, though, even more than the use of poly methyl methacrylate is that some people may think they are buying Lucite where they may be buying some other form of an acrylic plastic, and then they have two clear looking materials that they are really not the same. At least when you buy the breast-equivalent material, you know that it is somewhat more expensive, but you know what it is.

MS. MARTIN: Yes, I do not negate the fact it is much expensive. It is also a much better representation of what happens in the clinic. And again, what is happening is for the major vendors I know are using the breast-equivalent plastic when they are calibrating their units, and you do not see problems with these units. It's the vendors that are saving money, that are setting it up with the PMMA, that the physicist then comes in with the breast-equivalent plastic and finds out that it won't pass.

It's just a problem that is developing, and it's developing because we're getting better. But yes, definitely, it is a more expensive item. You cannot go with that option. I mean, definitely, I'm just making the recommendation.

DR. KARELLAS: This is Andrew Karellas again. I have no disagreement with Ms. Martin's suggestion. I think it makes very good technical and scientific sense.

CHAIR HARVEY: Any other comments? Any other sections? Ms. Martin?

MS. MARTIN: This is Melissa Martin. You shouldn't turn me loose. The other major comment I would like to put in is I find it interesting that, and this may be bound by regulation, we only have to test it for -- the physicist is only required to test 2 to 6 centimeters. I really like the "recommendation" that's in there, because at least in southern California, most of our ladies or a large number of our ladies have more than 6 centimeters of breast. And if I were a facility, I would find it very inadequate if my machine was only good for 6 centimeters.

And the other question I have is what is the recommendation for those facilities where you get a unit that simply cannot make it through 8 centimeters of breast? I have a hard time, as a medical physicist, saying that unit is adequate to perform when it simply doesn't have enough power to even penetrate an 8 centimeter breast. I think we cut short this 6 centimeters. I really think we should have made it through 8 centimeters at least, because there's a lot of women with at least 8 centimeters of breast tissue out there.

DR. KARELLAS: Andrew Karellas again. Melissa, isn't that taken care of by the tube output requirements? I agree with you about the tracking issue. The tracking issue, many physicists tested beyond 6, but you don't have to comply beyond 6 cm. But as far as the ability to penetrate for the x-ray beam, the breast and produce an image without a system backup, can that be addressed by the requirement of the output of the tube?

MS. MARTIN: It wasn't only units that I basically failed, and I took the initiative to tell them that it wasn't an adequate unit. I'm just bringing up things that are happening with the AEC testing that's going on out there.

DR. FINDER: It's Dr. Finder. I agree with Dr. Karellas. The issue about penetrating the larger breast is handled by the outpoint measure. The issue about the AEC and the tracking of that, again, is handled under the regulation. The regulations, when they were written, dealt with 2 to 6 centimeters as the requirement. Yet, the best we can do at this point is the recommendation. We can't require something that's beyond the regulation. And that goes into the whole issue of what is guidance and what is regulation.

There certainly is the potential when we amend the regulations again to consider that. But at this stage, the best we can do, as we have done already, is to put in their recommendation for thicknesses outside the 2 to 6 centimeter range. There must be more comments, because this has been an issue that keeps popping up. We have heard comments about what is a configuration. Any discussion about that?

Should various target-filter combinations be considered part of the configuration, should different receptor sizes be considered configurations, and what's the consequence of making those decisions?

MS. MARTIN: Well, if you go back to the way the guidance is written, though, that it's in a clinically used mode, that really takes out most of your questions. Because you wouldn't have like a magnification mode with a grid. I guess I didn't have a problem interpreting them, as long as it's a clinically used configuration, that really answered a lot of the configuration questions.

DR. FINDER: Well, let me raise this question. Suppose target-filter is a configuration, a specific configuration, how do you test it over 2, 4 and 6 centimeters if you don't use them clinically over, a target-filter combination, 2, 4 and 6? How would you test it?

MS. MARTIN: I wouldn't test it, because it's not clinically used.

DR. FINDER: Right. But those are the issues that are being discussed. And obviously, it raises some questions and people are confused about what they have to test, and is there a way in our guidance to make it clearer what we're trying to get at. Whether we're looking at this as an AEC test or a full system test or how these all impact. And I think many of the comments that were made earlier in the morning deal with the fact that some of this is not well understood in the community out there, and people are trying to apply standards that may not be the ones that we were initially trying to get at in the regulations or even in this guidance. And are there ways to improve what we're saying here to make it clearer to everybody what we want?

I will say that these guidance questions have been written and rewritten and rewritten many times, and I'm hoping that this is going to be "the final" version of this, so that everybody when they read this will have a clear understanding of what we want until the next time we have to bring this up and have to revise them all over again. But that's my goal here is to try and get a coherent document that gets what we want in terms of the guidance, so that everybody understands where they have to go.

I can't say that many of these questions were written at various times over the last seven or eight years, and again the purpose of this document is to try and make one cohesive, coherent document. That said, there must be other comments, I hope, otherwise we have plenty of time for lunch.

DR. PISANO: This is Etta Pisano. The only reason I'm commenting is because you've asked for comments. I will say that I think it's a well-written document, and I understood it.

CHAIR HARVEY: That's a great radiologist.

DR. FINDER: So now we're going to go ahead and revise it.

CHAIR HARVEY: Yes, doctor?

MS. MARTIN: This is Melissa Martin again. The one thing I would like to encourage, and I mean again, I had no problems understanding it as written. And again, I would just really like to follow through with the you've got it in as a recommendation that the physicists test the magnification mode if it's clinically used every year, and I realize that's a recommendation, but I would really like to make sure that that is kept in there as part of the annual physicists evaluation. I probably find more problems with the magnification modes than I do many of the screening modes.

CHAIR HARVEY: Okay. Dr. Karellas, do you have any other comments about this document, any aspect of it?

DR. KARELLAS: This is Andrew Karellas. No, I don't think I have any unless somebody has a specific question for me.

CHAIR HARVEY: Yes. So you felt that physicists would find it straightforward?

DR. KARELLAS: I wouldn't go that far. Sometimes even the best most detailed written instructions could raise questions.

CHAIR HARVEY: Right.

DR. KARELLAS: And these systems are becoming increasingly complex.

CHAIR HARVEY: Complex.

DR. KARELLAS: And I am sure there will be questions in the future, but for the time being, I believe that we have done the best we can.

CHAIR HARVEY: Well, Dr. Finder, that's a positive response from your physicist.

DR. FINDER: I was hoping for more. I was expecting more.

CHAIR HARVEY: I know. I sent this out. Maryanne Harvey. And I didn't get any comments back, people commenting or carrying on about any particular section of it.

DR. PISANO: Just this point of clarification. Were you expecting us to email you our comments in advance of this meeting?

DR. FINDER: No.

DR. PISANO: Because I didn't realize that.

DR. FINDER: No.

DR. PISANO: Okay.

DR. FINDER: No, you didn't have to email comments. And again, this document, itself, #6 is open for public comment. We are expecting to receive comments through the formalized process at the end of the open public comment period, which again is in the middle of May. And after that, we're going to look at all the public comments that we receive and consider revising the document, at that point. I just didn't think we had done such a, you know, good job as it appears we had in this, and as long as everybody seems to understand this.

CHAIR HARVEY: Ms. Martin has a comment.

DR. FINDER: Yes?

MS. MARTIN: Okay. Just to go back to what John Sandrik presented this morning, because I think we got a good presentation, and I don't think we ought to gloss over it. Was the intent of the FDA to include this AEC testing as a performance of the entire system at the facility or was it geared strictly to the particular unit? Because we have basically -- we are putting an onerous on the manufacturers, because they are dealing with all of the film-screen combinations that are in use out there. And I think we just need it on the record as to what we're really looking for.

DR. FINDER: Right. And I think the presentation earlier this morning raised some good points that maybe we should include additional questions in here, and additional guidance about what we're trying to get at. I think it's even a little bit more complex than what you said, because we're not only looking at -- I guess there are three possibilities.

One is are we looking at the AEC as an individual item, individual component or are we looking at the unit? And by unit, you can't look at it just as a unit, because the film, obviously, is involved in the screens, the processing. So when somebody says unit, I think of that entire system.

CHAIR HARVEY: System.

DR. PISANO: System.

DR. FINDER: And this is more of a system test. Now, is this supposed to be a test that encompasses every possible variable at a facility? And I think it was brought up earlier today, and it has been well-known, I think, by most people who look at this issue, that if you try and put in all the variables that you possibly can, there's no way that anybody can pass this test.

CHAIR HARVEY: Yes.

DR. FINDER: It has to be done under specific controlled conditions. If you're going to start slipping in different cassettes that have a requirement of only a .3, and try and get a .15 action limit out of it, you're bound to fail. If you start including film from different film batches or different ages or whatever, of course, you're going to end up with trouble.

I guess the recommendation would be to use a single cassette, to use film from the same box one right after the other. What you're trying to do is isolate the AEC as much as you can, but you have to realize that in order to do the test, you're dealing with a number of other components that have their own variability. So I think the idea here is to try and isolate the AEC as much as possible, with the assumption that if you got an AEC that's functioning and all the other tests are within their action limits, that in general practice when a facility takes a film, whether it uses a different cassette or changes its film slightly or whatever or the processing conditions aren't exactly the same, they are still going to be within the correct ballpark so you can read that film.

But if you try and impose this .15 requirement by sticking in different films, different sized cassettes that have different properties, different processing conditions, like you process the films on different days you do the test on different days, I would say that you would have a lot of trouble trying to meet our test results. And I think that it is important that that concept be emphasized in this document, because if people are going to try -- and we do know that there are people who do these tests that way, and then go back to the manufacturer and say that the AEC is broken, fix it. And that is not correct. It's not trying to follow the regulations. It's not following the guidance.

We try to address that in our question about if the AEC Performance Test fails, does that mean that the AEC is the cause? And the answer we said was no, and try to give an explanation. We certainly can expand upon that. One of the things that we have been told in the past not to do is to get too specific with our recommendations for testing requirements, because the physicists come back and say that's job. We know what we're doing.

And at other times, we get people who run into these situations. So we're trying to get a balance between those two. And anything that you people can suggest that would make it better, this document better, because we know that this is the perfect document, right? There's no problems with it. But we certainly can make it better, and that's the purpose of this.

CHAIR HARVEY: A comment from the audience. Mr. Mourad, come up to the microphone, please.

DR. MOURAD: Wally Mourad, FDA. I just want to emphasize one other point here in addition to what Dr. Finder said, and that is the following: Initially, in our initial guidance on how to test the AEC, we tried to stick to the regulations to see that the AEC is tested like it should perform, whether the different combinations are being used at the facility or not. That was one approach.

The other approach was to test it as it is used at the facility. So in a sense, it became do we test the technique charts that are being used at the facility or do we use a test that specifically is geared to determine whether or not the AEC is functioning as it is designed. So we took the second, the earlier approach that is what is reflected in the document now is to test it as it is used by the physicist, because this is based on comments that we have received from several physicists in the field.

Now, if you folks are comfortable with that, let's make sure that this is the way we're going to go. That's all I want to say.

CHAIR HARVEY: Thank you.

DR. FINDER: Yes. This is Dr. Finder. I think that is a very important aspect, because as I said our guidance has evolved over time, and in the beginning, we were trying to make this more of an equipment type test in the sense that if an, for example, AEC mode was used anywhere in the 2 to 6 centimeter range, it would have to be tested at 2, 4 and 6.

The problem is if it was only used at, let's say, 6 centimeters and not used at 2 and 4, we were getting comments back that it couldn't meet those requirements as stated in the regs. So we have evolved over time to make this more of a clinically used AEC test. So that you use a specific mode only in the 6 centimeter range, that's where you would test it. And the guidance that we've come out with here tries to specify how that would be tested. And I hope we have made it clear enough in the guidance that that's what is going to happen.

MS. MARTIN: I would just like to reiterate, yes, that is the purpose of this test, and I would speak highly in favor of the document as is.

DR. FINDER: Okay.

CHAIR HARVEY: Are there any comments from the audience? No?

DR. FINDER: You've got all your questions answered? All your hopes and dreams have been handled.

CHAIR HARVEY: Or squashed, one or the other.

DR. SANDRIK: John Sandrik for an encore performance. I would not hesitate to expect that when I get back to my office there will be another call about another variation of interpretation on how this works. Actually, I think, our view is in many ways it's a good document. I say, I think, the need of an explicit statement is a unit test or a facility test. I think that needs to be put in there.

DR. FINDER: Let me just clarify that.

DR. SANDRIK: Yes.

DR. FINDER: I would say a system test versus a unit test, because again you can't test this -- you can't perform this test without the entire system.

DR. SANDRIK: Right, I agree. In fact, the AEC system is calibrated based on the response of the cassette and the film, so it is all part of it. However, the AEC system cannot detect that different cassettes have different speeds or different films are from different boxes. It only knows basically the transmission and the response of a sample screen and a response of a sample film. So to whatever extent that changes or varies say between the 24 X 30 box and the 18 X 24 box of film, the system doesn't know that.

DR. FINDER: No.

DR. SANDRIK: So we feel that there should be some explicit mention that it is not intended to include those kind of variables that are not part of the calibration. So, you know, I think clarifying that would help. Whether or not you view 24 X 30 or 18 X 24 as different configurations may be resolved from the point of view of if you really correct the differences between the cassettes or the different films, that may resolve some of the issue of observing some of the variation systems.

But I still think in reading some of the answers to the questions, FDA seems to be kind of going back and forth a little bit. I think this was maybe question 7. It had to do with is a different image receptor a different configuration? And the first sentence is no, it's not a different configuration, but the answer concludes that the different receptor size that will be compared at the plus or minus .3 level, which if you go back to question 4, is what FDA applied between configurations.

So in a way it's saying no, it's not a configuration, but we'll test it like it is a configuration. And I'm saying well, if you just call it a configuration, then you're testing it and evaluating it under the same conditions, and it may just simplify life, at least one aspect of life.

DR. FINDER: Dr. Finder. Nothing we do ever simplifies this.

CHAIR HARVEY: Interesting.

DR. FINDER: Because --

DR. SANDRIK: They're not here to help.

DR. FINDER: Let me just say that there are implications whether you call it a configuration or not.

DR. SANDRIK: Yes, I agree, because if it's not, if it's a different configuration, it doesn't have to be tested annually, and that's part of the issue is if we call -- then it only has to be tested. If it's a different configuration, it only has to be tested during the equipment evaluation, which can occur basically when you have a new piece of equipment or if you have a major repair to this. So there are implications to this, and that's why we kind of wrote it the way we did to give the recommendation that even though it's not a different configuration, it's part of the same configuration if you're going to test it. You can use the .3 action limit on this.

So we're trying to go a fine line between what the regulations say and what the guidance says.

DR. FINDER: Right.

DR. SANDRIK: And what we can recommend.

DR. FINDER: Yes, so I can understand there are other implications involved here. I think, you know, similarly if you call the target-filter combination, the different configuration with the implication that you didn't have to test it annually, I think you would end up with a strange situation though. If you weren't ever testing your molybdenum- rhodium combination or rhodium-rhodium combination, anything but an equipment evaluation, because it is viewed as a different configuration, and you only had to test say what you use for the average breast, you're probably leaving a fair amount of your Automatic Exposure Control System untested.

So it is a bit complicated and it probably needs more thought put into it than just from our point of view. I guess one other comment. I probably shouldn't speak too much off the top of my head, but Ms. Martin brings up an interesting question that we've also been dealing with is this acrylic versus breast-equivalent plastic issue, and I guess one comment I have on that as far as what is breast equivalent.

I think if you look at some of the recent and not so recent publications about what really is the average breast and what's its composition, the consensus has been coming down that the average breast is more like a 30 percent glandular composition rather than a 50 percent composition. Particularly, when you get to the large breast, it's probably nowhere near the 50 percent that we're talking about with breast- equivalent tissue. Maybe it's in the low 30, 20 percent glandular kind of content.

So this question does 50/50 glandular or 55 percent glandular really represent the clinical situation any better than acrylic, I would hesitate to guess off the top of my head, no. I don't think that either one is all that representative of the full range of what breast conditions are, and probably both are getting away from what they should be at the larger breast, which I think also goes to the issue of testing at 8 centimeters. I think definitely 8 centimeters acrylic or 8 centimeters of tissue equivalent plastic are both well beyond the range of what might be 8 centimeters of a real breast.

And I think in the paper by Kruger, et al from a couple of years back in Medical Physics, they indicated that something like 8 centimeters of tissue equivalent plastic was at the 99.9 percentile range of their calculation of 24,000 women that they studied. So is it really necessary to demonstrate absolute tracking out to 8 centimeters of 50/50 plastic when that might represent 1/10th of the percent of your population, as opposed to maybe say film density control plus 1 under that condition in those cases.

So I think there are questions to be answered here and maybe looked into. I think, you know, a system could be tuned up to track breast-equivalent plastic. It could be tuned up to track acrylic. Does either one really represent any better tracking for the breast? It's definitely a given that breast-equivalent plastic is really representative of breast over this whole range. Amen.

CHAIR HARVEY: All right. Any further comments? No? Thank you. Dr. Barr?

DR. BARR: I just have some of those numbers here if you want them.

CHAIR HARVEY: Oh, excellent. Certainly.

DR. BARR: I just have the number of certified facilities. I have asked Ms. Harvey before we adjourn today to collect from you what other types of data you would like to see that I can send you when I pull this up. Sorry, I pulled up the wrong one. All right. These numbers are on October 1st of each year that I'm going to give you.

2000, our data base shows 9,933 certified facilities; 2001, 9,558; 2002, 9,306; and today our data base shows 9,189.

CHAIR HARVEY: Okay. Thank you.

DR. FINDER: Helen, you don't have the number of units though? Is that correct?

DR. BARR: Not at this time. If that's a data piece that people would like to see, please, pass that on to Ms. Harvey.

DR. FINDER: All right.

CHAIR HARVEY: Yes.

DR. FINDER: They would like to see that.

DR. BARR: Okay.

CHAIR HARVEY: Thank you. Yes, this is Maryanne Harvey. The number of actual systems within the facility is probably more important total than the total number of facilities, depending on where they come from.

DR. FINDER: It's Dr. Finder. To some extent, yes, but another extent, no. Obviously, we're looking at access, and access isn't only related to the number of units. It's location facilities that accept certain insurances and things like that. So I guess one could say that what's better to have, you know. You could take it to the extreme and have one facility that had 10,000 units. I don't think anybody would consider that good access. There has to be some kind of happy medium there.

But I do think that the idea of consolidation to have facilities that have more units doing more cases, more emphasis on mammography probably is not a bad thing if it's not taken to the extreme.

CHAIR HARVEY: Too far, right. Thank you. All right. We have one more guidance document.

DR. FINDER: Well, before we do that, I just want to make sure.

CHAIR HARVEY: Okay.

DR. FINDER: Nobody has any other comment about this document?

CHAIR HARVEY: Speak now or forever hold your peace.

DR. FINDER: Exactly.

CHAIR HARVEY: He doesn't want to hear about it if you come up after.

DR. FINDER: Because this is a topic, one of my favorite topics is AEC. It is the issue that --

CHAIR HARVEY: His new novel will have this in.

DR. FINDER: Yes. It is the issue that never goes away. It keeps coming back in different forms, and we are hoping, and I do mean the word hope, that this document after it has been revised, taking into account whatever comments we do receive from the public, will be the definitive word on AEC testing, at least for MQSA purposes. So nobody has any other comments, let's go to the other document.

And again, just to kind of clarify. This document went out as final. For those people who don't know, we have been issuing guidance documents since the beginning of the program, all the final documents have been included into what is called the Policy Guidance Help System. It's a type of a search engine in which one can look and find virtually all the guidance that we have issued, and we are now in the process of taking all the guidance in there, and if you print it out, it's about 700 pages of guidance.

We are in the process of reviewing it from top to bottom, and as we go through we are trying to update, modify and make this as consistent with the regulations and other guidance that we have issued over time as possible. So this is one of the steps.

Guidance Document #7, I can assure you that there is Guidance Document #8, #9, #10, and I think #11 in the works. So as we go through each of the sections of the Policy Guidance Help System, we will be revising and updating those things. So this is just a work in progress. And if anybody has any comments about this, and I hope they do, otherwise, we're going to go to lunch early. Yes?

DR. PISANO: I would like to make a comment. I want to talk about page 10 of 23. I don't know if everybody has the same pages that I do, but I printed this off the web, so under question 5.

DR. FINDER: No, It's going to be difficult to see where.

DR. PISANO: Okay.

CHAIR HARVEY: Yes, that's page?

DR. FINDER: I will try and get the question.

DR. PISANO: The question wording is what should a facility do if it closes or decides they will no longer provide mammography services. So whatever page that is on yours. 13?

DR. FINDER: 13, okay.

DR. KARELLAS: 13 on mine.

DR. FINDER: 13.

CHAIR HARVEY: Yes, you and I printed it this morning.

DR. FINDER: All right. It's page 13, line 28.

DR. PISANO: Okay. I wanted to know, it doesn't directly address digital mammography, because the hard-copy images are just a subset of the data available, obviously, and so are the facilities who do digital mammography required to somehow provide? How are they supposed to -- it doesn't describe that issue, so I just was curious about that or is there -- I couldn't find another question that seemed to address that.

DR. FINDER: Okay. In another area in the guidance, we talk about what is a medical record for digital. And we have basically said that it can either be the electronic data or a hard-copy. The difference here, I think, is that it has to be in some type of format that it's available to the patient. So to have an electronic data base with all the information in there and no way to get it out, doesn't serve the patient's purpose. There has to be some mechanism available so that if a patient needs a copy of her mammogram, it has to be retrievable.

Now that may be electronically, if they can get it into another machine so that they can produce a hard-copy some place else if they need it or if the other facility can accept that electronic transfer, that would be fine. But it is going to be more complicated in terms of the digital units for that. Does that answer the question?

CHAIR HARVEY: I think Andrew has a point to make about this, I think, maybe.

DR. KARELLAS: This is Andrew Karellas. I find this challenging. Let's assume the scenario that we find the electronic version of the mammogram, we would give the patient perhaps a CD with the cases. I'm afraid that frequently the patient may not be happy, because when they go to display the data at some facility, they may not be able to read them unless they have the exact system that we use. And I would suspect that five, six, seven years from now, it will be even more difficult.

Alternatively, we could provide a print. However, if a facility closed and they just have the data stored, clearly, some facility would not be able to provide the print. So the facility would normally provide data, but sometimes, ultimately, these data can be retrieved, I'm sure, but certainly not conveniently, and this is my concern.

DR. FINDER: Okay. And your suggestion for correcting that concern?

MS. GILBERT: This is Alisa Gilbert.

DR. FINDER: Yes.

MS. GILBERT: I have a question. Further down in the documentation where it says under the same question when attention, closed facility notification, records retention, wouldn't that same documentation be able to be transmitted to these two mentioned, State of Iowa and Illinois facilities, and if a patient would like to retrieve her records, couldn't that -- no, okay. It seemed like they would be able to be the hub of that and then be able to provide that instead of a CD for them.

DR. FINDER: Well, these agencies don't act as repository for records. The facilities are responsible for maintaining the records in some manner that the patients can use them. The issue here is for the film-screen facility, you generally have the hard-copy, the film itself, and that can widely be read pretty much anywhere, radiology offices, even doctor's offices at least have a viewbox.

The problem with digital is that many of these machines have specific type of formats that aren't compatible with other machines. Most places don't have these units altogether, and if you give somebody a disk, they have no way of being able to regenerate that into an image. So it is an important question. I guess the point I'm trying to think of is what's the answer, and I was hoping that Dr. Karellas was just about to give us the answer.

DR. KARELLAS: This is Andrew Karellas again. Of course, technically, the answer is easy. You have some place that you go to and perhaps for some fee, somebody can take the data and extract the data. However, the reality is that a CD may contain three, four, five, six cases of patients.

If a facility has closed, they are not likely to be able to have anybody to do anything, other than pull a CD and copy. They may not even be able to copy very easily, but let's assume that they can copy the CD, and that's perhaps the easiest thing to do, that will contain other patients' information and you cannot do that, of course.

So you really have to extract that case out and put the data in the proper form, and this is fairly straightforward in any good hospital environment today that they have computer oriented people. They have to save, to save the data, either in the original, either in the DICOM format or in our machine, the so-called raw format that almost anybody can use.

So technically, it can be done, no question about it. The problem is that the facility may be arguing with the patient. The patient may be arguing with the facility, and there is nobody to do it and basically, the cost is the huge issue here. The facility will not be able to do that because of the cost.

CHAIR HARVEY: Dr. Pisano?

DR. PISANO: Charlie, just clarify for me again. The facility can provide just a hard-copy and that's met their obligation according to the current regulations?

DR. FINDER: Correct.

DR. PISANO: So it's just if the patient wants a soft-copy. So in terms of regulations, the facility doesn't have to provide the soft-copy? In my opinion, this is not as complicated as it is. I mean, it doesn't have to be as complicated as it really is, because DICOM readers are widely available in, as you mentioned, most hospitals, but on top of that, they are available on the Internet.

So I am not sure that women could just plug it into a computer, download one of these DICOM readers and see it. But it's not that much more complicated than that. The manufacturers have created a barrier, in my judgment, to reading across devices, so that if I have three different systems at my place, I can't read my Fuji images on my GE system. I can't read my Fischer images on my GE system, and I wish that would go away, because then that would at least help.

You know, you don't have to find a place that has the same device to look at your images. You get what I'm saying? If you have a GE image from a place that's closed down, you could carry a disk. I think you are getting at a lot of the real issues in terms of who is going to copy it for you and all that kind of stuff.

But it sounds like that's not a regulatory issue. That's more of a practical issue, because you can use this hard-copy. The facility can fulfill its obligation by printing out hard-copies and handing them to the patient, but it's really not that complicated to look at images with a DICOM viewer. We can't do it right now, but it's really not that hard, because it's just software unavailability.

DR. KARELLAS: Yes. This is Andrew Karellas again. Dr. Pisano is absolutely correct about the reading across platforms. However, now, imagine a facility. This facility has closed. This is not an open facility. It's closed. That means that things are not very good there in mammography at least.

Now, all these data have been placed at some safe place, and there are all these nice CDs that nobody knows why they are there for. So they have closed completely or has closed its mammography practice and now, there is nobody that will do anything. Even if they pull that cassette, they will say there are the six people and there is where the problem starts.

If it's a fortunate facility that they are thriving otherwise, they can do the data transfer and all that, that will be very, very costly and they will really mind that. Now, the printing part, you cannot take a CD out and print it, because now, in order to print it, you have to pull out the whole thing on a work station, which is an integral part of an advanced digital mammography system, display it properly, print it on a printer that has been through quality assurance. So you cannot print, because you don't have a printer if you're doing mammography. So it's a big problem.

DR. FINDER: This is Dr. Finder. I just want to make mention, this is an issue that isn't just a problem for FFDM, for Full Field Digital. It's the same type of problem, slightly different when we're dealing with film-screen, when a facility closes, and if they follow our recommendations here, if they follow our guidance, then there is usually not much of a problem. But if they don't, if they just go out of business and go bankrupt, lock their doors, we have had to deal with those types of situations.

I know that Michigan has dealt with this, and I believe New York has also. It becomes very problematic to try and get those films available to the patients that need them. The issue of FFDM is even more complicated, because at least with film-screen, if you can find the films, you can hand them over. There is no real manipulation that has to be done.

FFDM adds several steps to it and other complications. Obviously, if all you have are the CDs and the hardware has been sold as part of the bankruptcy, you have nothing to put that CD into. You have to find another facility that has that capability and is willing to do it. It is even more problematic than the film-screen, so again, your solution was --

DR. KARELLAS: This is Andrew Karellas again. I think the only real solution is not very practical, but when you have a digital mammogram and you print images, the good thing is that you can have multiple originals, because a printed mammogram is not like a duplicate. So you can have as many originals as you want.

I do not recommend that this is the best thing to do, to hand people their mammograms, but in some countries they do that. In fact, in some countries, the medical institution doesn't even hold onto your films and that will increase the cost. But if I really have any concern whatsoever about that mammogram and for some reason I think that I may need it five years from now, printing an extra copy and keeping it is the only way to know that I will have one.

However, that will add to the cost, the complexity and I know that many facilities would absolutely not favor that, because that will really defeat the whole purpose of doing digital mammography and we will be printing more films than ever before. But, of course, this is an option, but, of course, not an easy option financially for facilities to do.

DR. PISANO: I think I'm not sure how to solve this problem either, but I think you need another question that deals specifically with Full Field Digital Mammography and whatever you decide to do with it. Ideally, every patient would have a printed copy before the facility closed or they would have instructions on how to print off the disks that are being stored somewhere.

I mean, I think a printed copy is fine. If you've got to have a copy, that's the best you can do. That's fine, but facilities shouldn't just be able to close and leave a stack of disks with no way for the patients to access those images. That's not fair to the patients and they need some way. So I think you need some guidance here, specifically on this issue.

DR. IKEDA: This is Debra Ikeda. I am not sure how you could force a facility to print out all the mammograms if they are bankrupt. Where are they going to get the money for all the printed copies?

DR. PISANO: You know what I'm saying?

DR. IKEDA: But I'm not really sure about the right answer.

DR. PISANO: Maybe you could require the facility to notify the patients, so that the patients could -- and give them names of places where they can get them printed for a fee or something like that. I mean, it's very complex, because, as Andrew said, they are stored, you know, 10 mammograms per disk or whatever.

DR. IKEDA: Dr. Finder, this is Debra Ikeda again. Don't the regulations say that the patient should be able to have either an electronic copy of their mammogram or a hard-copy and first hard-copy is printed at no cost to the patient. Isn't that correct?

DR. FINDER: That is correct. I think there are a couple of issue here. One is we can issue guidance and we have issued guidance for film-screen and we certainly can add a few things for FFDM. The problem isn't with the people that follow the guidance.

DR. IKEDA: Right.

DR. FINDER: It's the ones that don't and what do we do? And that is a whole other issue and, obviously, there are problems with the facilities that have gone bankrupt, and there are limits to what we can do in those cases. Sometimes, we have actually worked with the courts that are involved with some of these cases to try and get access to these films.

We now have additional issues, if not problems, with the HIPAA requirements in terms of privacy, and one of the issues, obviously, is it's one thing to go and pull an individual patient's records and transmit those to the patient. It's another thing when you have got six or 10 or whatever on a CD. You cannot release that to the patient and let her go around and make copies herself, because it would have information that she is not allowed to have, so it does become more complicated.

CHAIR HARVEY: Maryanne Harvey. Before people get permission to get a certification, can't we ask them what they might do if they had to close down the path? I mean, perhaps when we look at large hospitals, we have reason to believe they may stay in business, but a lot of places don't know what will happen to them.

They could have a terrible, you know, flood or hurricane or something that might destroy the practice above and beyond the fact that maybe some people just are not as conscientious, but up front, maybe there is some way we could ask them, a security bond?

DR. FINDER: This is Dr. Finder. This issue has actually been brought up before at this committee in the past, if I remember, not too long ago, either at the last meeting or the one before, where these issues were talked about. Obviously, if you ask a facility that's just starting out what are you going to do when you go bankrupt, it doesn't send the exact right signal that, you know, you are counting on them to be there for quite awhile.

The other thing is if you ask them, chances are that whatever they tell you won't be the situation when that event actually occurs, which may be five, 10 weeks down the road. No, I mean years down the road. So asking them is fine and I will say that the facilities that are trying to do the right thing usually can and they do it. It's the ones that end up in financial problems and close because of bankruptcy or they are abandoned where we have the most problem.

I will tell you that just in the last few weeks, we have had the issue of HIPAA, that's the Health Insurance Portability and Accountability Act, brought up in that a facility closed. There were other facilities in the area that wanted to take over the records and they were unsure if under HIPAA they were allowed to do that if there were some privacy issues.

We are in the process of getting them a formal legal opinion about that, but my general feeling from what I have heard from the lawyers I have talked to so far, who are not HIPAA lawyers, but more lawyers nonetheless, is that HIPAA does not preclude that. The idea, obviously, for HIPAA is to protect the patient's privacy, but not to worsen their medical care, and not allowing them access to their films certainly would be.

So I think that will eventually be resolved, but there is a lot of confusion about what you can release and what you can't release and what you have to do before you release something. So those questions need to be answered, and we are trying to get answers for facilities with respect to that.

The real problem comes down to though a facility that can't do anything, because they are out of business, and one of the comments that I believe was made at the last meeting was the question of posting a bond ahead of time, so that money could be used in case the facility is closed, and that also raised a whole number of issues and a fair amount of discussion, at that time, about what that would do to new facilities trying to start in mammography and whether it would force additional facilities to close, because they wouldn't be able to put up those bonds.

So I will say this. The vast majority of the facilities that close do take some action to allow patients to get their films, but there are a number of facilities that don't, and we hear about them and we try and work through the situation as best as possible, but it is certainly not optimal.

DR. RAMOS-HERNANDEZ: This is Catalina Ramos. I know it's a regulation saying that a patient may have or have the right to have one, the first mammography printed. What about asking for all people who are dealing with digital mammography to have printed out one copy either way, to keep in the records or to give it to the patient.

So anyhow, that should be done, should be there in case that anything happens and there is only one copy. It can be used by the hospital or can be used by the patient in case that this closes. That will avoid having the burden of printing all of them together when there are no more funding.

And I understand that might be a huge expense, but little by little, if there is required that anywhere, someplace, is one printed copy, it will ensure that if the facility closes either the patient has already one copy or is one copy in place to be handed to patients.

DR. FINDER: I would be interested to hear what the people who have FFDM units would say.

DR. KARELLAS: This is Andrew Karellas again. Two problems with handing patients copies. Number one is the cost, that these printed films are not cheap. I don't know the exact pricing, but I would image that four, five films may be easily $6, $7, $8 or $10, perhaps even more, and that's a huge part of the whole if you look at the technical fee, especially in the hospital situation. That's just a huge part.

Number two is that you cannot guarantee that patients will come back with their mammograms intact. I would think that if you gave me my x-rays, five, six years later, I am not sure that I would know where they would be, and I think most people are like that. So the yield will be relatively low, I think, relatively -- well, maybe not small, but you definitely would lose a fraction. People will come back and they just would not have them, so that's the number two problem.

DR. IKEDA: This is Debra Ikeda. That is a good thought, to print every single film. If you do film-screen mammography, part of the advantage of doing this is to get a good electronic copy that we can manipulate. The cost is, I would say, well, maybe 10 years ago it was $2.17 per film. I'm sure it's up now. It takes about four minutes to print a film, and if you are running 40 patients having their, you know, good mammograms through a good facility, that's a lot of time.

Secondly, those films must be labeled, put in a film copy jacket, labeled with the patient's name with the medical record number. You need a film handler or film librarian to take those films, collate them, put them in the right place, make sure they get put in the patient's jacket and then have a facility to store them. And so while it seems like a great idea, actually, the hidden costs of storing film are in the film library, the pack system that we use, the transporters, cost of making the film jacket, the labels and the storage facilities, film librarians at those facilities to retrieve the films.

So those are the other issues that would be involved in printing every single film. We do that for film-screen mammography. There are those facilities available, but that is why most places don't print out every single digital film.

CHAIR HARVEY: This is the last comment. Go ahead, Dr. Pisano.

DR. PISANO: I have another idea that actually is based on a demonstration project that is going on right now, kind of in response to your point. There is something that has been funded by the National Library of Medicine, which is called the National Digital Mammography Archive, which is headed by Mitchell Schnall at the University of Pennsylvania.

He is actually a radiologist who does mainly MR, but he has coordinated a group of research facilities to store mammograms at a centralized facility at the University of Pennsylvania. And I think this gets back to Andrew's comment earlier where in some countries, the patient's record is owned by the patient. I mean, we have this model in this country where the hospitals own the record and patients access their data when they access that hospital.

This model is the patient stores her records at an off site place. Presumably, she would pay those costs, and then when she goes to another facility, she can access those images. So you don't have to have a hard-copy. The images are stored in their digital format and you give permission. I mean, there is HIPAA privacy measures in place. This is a demonstration project. It's not huge, at this point, but it's a model for how medical records could be stored without it being at a set facility.

Like today, we are all traveling to Washington. If one of us got sick, I doubt we would need our mammogram, but we might need other parts of our medical record to be accessed by the facility where we visit. So it occurs to me that the way we store data right now is not very flexible, and that perhaps we, as a country, especially will HIPAA privacy concerns, should look more toward a different model of storing information. This, obviously, works for mammography very well, because it is being modeled already.

And there are some companies, I know Fuji has a system where you can store off site and access from different facilities also, so there are different packs. You know, right now, UNC's films are UNC's films and it would, you know, take a huge effort to get them to Duke, which is nine miles away, you know. So I think there are ways that we could change the system.

And, obviously, that's beyond the purview of this committee, but I would hate to see us going to printing every film when I think this, the way that it's being modeled for this new system that I just described, is much more cost effective and much better for patients. Because let's face it, we all lose things, and I have handed mammograms to patients many times. There are people who come to my practice who want their films to store in their garage or wherever, in their attic or something, and then they show up the following year and they have forgotten their films. Then you have to wait a few days.

So this is a way to have them accessible from anywhere, any time, and they are less likely to get lost, they are not zero probability. So anyway, I would like to see us moving in that direction, rather than printing for patients.

CHAIR HARVEY: Okay. Ms. Pura, did you have a comment? Okay. Excellent. I had one comment to make before lunch. We have conflict of interest forms that people are asked to complete. Perhaps during the lunch hour, you might have a moment to do that. And if we have any other data requests that we want from Dr. Barr, you should think of those and have those ready for when we leave.

Dr. Finder, did you have anything?

DR. FINDER: I just wanted to mention that we're just beginning on this document, and I am not even sure that the current discussion on film retention is finished. So when we come back after lunch, we certainly can start back on this. I would caution the committee though not to discuss anything during lunch about this. Everything has to be discussed in the open meeting. So you will have to come up with some other topic for lunch. Okay.

MS. PURA: We can't talk about it though? Is that right?

CHAIR HARVEY: One moment, please. We will return at 1:15.

DR. PISANO: 1:15?

CHAIR HARVEY: Right.

DR. PISANO: What are we supposed to do with those things, these things?

(Whereupon, at 11:56 a.m. a recess until 1:00 p.m.)

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

1:01 p.m.

CHAIR HARVEY: The afternoon session of the National Mammography Quality Assurance Advisory Committee has begun. Dr. Finder has passed around a sheet of paper for the members to put down what time they would like to leave as a result of flight or other transportation, pressures, and we would like to be sure that we have a full counting on all the discussions this afternoon, so we're going to move forward and be organized. Okay.

Returning to the discussion on the Guidance Document #7, Dr. Finder?

DR. FINDER: I guess we left off where we were talking about adding some additional guidance for FFDM in terms of record retention, and we certainly can take a look at that. I do want to make mention of the fact that the record of retention from mammography is really just a small part of overall medical record retention, and that digital mammography is probably the latest of the other modalities that have gone digital.

Presumably, some of these problems, hopefully, have already been dealt with in other digital modalities, and we'll take a look and see what we can find there and try and come up with some kind of guidance to deal with that question.

CHAIR HARVEY: Any other comments? Any other concerns raised in this document?

DR. PISANO: I have one.

CHAIR HARVEY: Yes, Dr. Pisano?

DR. PISANO: Okay. Etta Pisano. On page 15 on my document, which I know is not the same pages as yours, it says mammographic modality, question 4, is Full Field Digital considered a mammographic modality? I just wanted to --

DR. FINDER: It's page 23, line 16.

DR. PISANO: I don't know if this is explicitly direct in this particular question, but I didn't know where else to bring this up. I think that the continuing education requirements for digital mammography are excessive, and I don't know where this fits in this document or any time we're going to discuss today, but I am supportive of the initial requirements, but I think that the ongoing requirement of, I forget what it is, eight hours every -- how much is it?

DR. FINDER: For interpreting physicians, it's six hours every three years in each mammographic modality used.

DR. PISANO: I would like to see us address whether that's really necessary or not. I mean, I would like to hear other people's opinions, too, but as someone who has been in digital mammography for a long time, I can say that I think once you learn how to do it and if you are practicing it, it's not really that different. The main thing that you have to do differently is the soft-copy interpretation and once you know how to do that, you know how to do it.

So I can't see really why people need six hours every three years on that, so I just want to make that point. I think it's excessive. I think I am supportive overall of the CME requirements in breast imaging, but specifically in digital mammography, I think it creates a burden on the practitioners that's really not justified. It's my opinion.

CHAIR HARVEY: Maryanne Harvey. Dr. Dershaw raised the same issue.

DR. PISANO: Right.

CHAIR HARVEY: In his comments to Congress.

DR. PISANO: Anyway, I don't know where it fits.

DR. YOUNG: Maryanne?

CHAIR HARVEY: Yes.

DR. YOUNG: Don Young. It has been my experience that the learning curve is pretty steep in digital mammography, and once you have got up at the top, not too much more is going to contribute to that. On page 23, I had one question, too, just a point of information, line 11. It says the term xeromammography is in there. I wonder if we could delete that. I am not aware of any xero units in operation in the world, let alone in the United States yet.

DR. FINDER: Yes, I think that's true. I don't think there are any xero units still left. That was in there originally when the guidance was written years ago. We can consider taking it out.

This is Doctor Finder. Dr. Pisano, I just want to make mention of the issue about the CME. We were planning on rolling that into the reauthorization.

DR. PISANO: Okay.

DR. FINDER: I guess we can wait and talk about it then.

DR. PISANO: Okay. That's fine. I wasn't sure when to bring it up.

CHAIR HARVEY: Okay. All right. Any further comments or concerns?

DR. FINDER: So this document is perfect also? Okay. Not quite.

CHAIR HARVEY: It's a work in progress, right? I think that means that we're ready for Dr. Barr to talk on the status of MQSA reauthorization? Dr. Barr is the Deputy Director of the Division of Mammography Quality and Radiation Program. Welcome.

DR. BARR: Thank you, Maryanne. Before I start, I wanted to give you some more numbers that you asked for. I have your unit numbers for FY '99 through 2002. In '99, there were 12,655 units. In 2000, there were 12,956 and for '01 and '02, and my data person put a message here that yes, it's not a mistake, the numbers are the same. For '01 and '02 it's 13,173. Charlie, we have a question?

MS. GILBERT: '99 there were 12,655?

DR. BARR: Yes.

MS. GILBERT: And then what was the next year you mentioned, 2000?

DR. BARR: 2000, 12,956 and then in '01 and '02, 13,173. So that goes along with what we had said earlier. I am going to talk a little bit about MQSA reauthorization. The last reauthorization of MQSA expired on September 30, 2002. Many of you are probably wondering how we're continuing to operate. We have been conducting business as usual since that time, and the reason is that our authority to inspect and certify facilities does not sunset with the expiration of the last reauthorization.

The reauthorization is actually a reauthorization to appropriate funds to MQSA, and while we were on a continuing resolution, that was no problem. My best understanding of the issue since we came off the continuing resolution is that with or without authorization to do so, we got some funds dedicated to MQSA until such time as there is another reauthorization.

A delay in reauthorization is not unprecedented and was, in fact, experienced the first time MQSA was up for reauthorization. When the original authorization expired in the fall of 1997, it was, approximately, a year until MQSA was again reauthorized in the fall of 1998.

MQ authorized during the last Congressional session. I would characterize that as primarily due to controversies over issues surrounding the skills of physicians interpreting mammograms, as well as, presumably, other more pressing Congressional matters, such as a war.

On April 8th, a full committee hearing of the Committee on Health, Education, Labor and Pensions was held. Three witnesses were called to testify in that hearing, two for the majority and one for the minority. Drs. David Dershaw and Leonard Berlin, two prominent radiologists, were majority witnesses and Ms. Diana Rowden of the Susan Komen Foundation was the minority witness. I believe the committee has transcripts from all those witnesses, and those are also available out on the front table for anyone else who would like to see them.

The three witnesses wholeheartedly endorsed reauthorizing MQSA citing its contribution to the declining mortality from breast cancer. The focus of the testimony centered around the controversy that Congress is struggling with, that of physician interpretive skills. While the number of mammograms that a physician reads was briefly touched upon as a possible factor in interpretative ability, the focus of the statements addressed a proposal to include physician self-assessment as part of the continuing medical education already required for physicians under MQSA.

The ideas that physicians would likely attest to the fact that five of the required 15 CMEs in a three year period would be of a self-assessment or interactive kind. There is no extreme hard evidence to suggest that self-assessment improves physician interpretive skills, but it seems like it might be intuitive that physicians assessing their skills might lead to some improvement. Experts ranged from this is a feel good requirement, that really won't do any good, to that this is a good and necessary requirement that would indeed improve the way physicians interpret mammograms.

The downsides of this approach are primarily seen as twofold. First, that there will be further regulatory burden on an already heavily regulated profession, and that more requirements would tend to fuel the trend of physicians leaving the practice of mammography and new physicians choosing not to specialize in mammography, eventually decreasing patient access to mammography as facilities continue to close.

Secondly and related, many experts fear that unless the results of such self-assessment are made non-discoverable, such results will not only have the effects of driving physicians out of mammography and contributing to the declining new physicians going into the profession, but may also contribute to what many consider the medical litigation crisis in mammography and medicine as a whole in this country.

In the Komen witness testimony, FDA was cited for not requiring results of the required medical outcome audit under MQSA to be reported either to the agency itself or to any cancer registries in addition to the facility using the data itself. The facility based use of the data was a decision that this advisory committee, at the time, was intimately involved in.

That witness called for FDA to enhance the medical audit by requiring facilities and FDA to collect data on the number of mammograms performed, the percentage of cases reported out in each BI-RADS category, number of biopsies performed, follow-up results of all further imaging studies and data on retrospective reviews of all mammograms of patients diagnosed with breast cancer, as well as patient age and ethnicity. The data would then, in the witness' view, be used to determine how particular facilities are performing and, presumably, make that information available.

At some point in this discussion, Dr. Finder is going to lead. Hopefully, Mr. Camburn, a panel member, will comment a little bit on this matter with some experience that Michigan has had in this realm. Mention was made in one witness testimony of dropping the modality specific continuing education requirement, specifically citing digital mammography, but as Dr. Finder will explain when he leads the discussions, not only pertains to digital, but to all mammographic modalities currently beyond the initial training requirements and the new modality CMEs required with differing number of hours for physicians, technologists and physicists. The witness felt that the requirements are often difficult to meet and does not contribute to improved quality or patient safety.

Recommendations for removing the current exemption for stereotactic and other interventional mammography procedures appeared in two of the testimonies. One witness called for a two year reauthorization period rather than the five year time frame that the last reauthorization was for.

That is a flavor of the witness testimonies, which, of course, you can read in detail. I would love to turn back to Dr. Finder who will lead a discussion on some of the issues mentioned and address several specific questions to the committee for your input. I will be available during this discussion if you have any additional questions.

CHAIR HARVEY: Thank you. Questions, concerns, impact?

DR. FINDER: All right. This is Dr. Finder. I just wanted to mention that when this meeting was originally scheduled, we had hoped to use this meeting to try and further define what had been already reauthorized. Due to timing and other circumstances, what we now have to deal with is what we think possibly might actually come out of Congress. We don't have anything firm.

So I guess what we would ask you to do is to discuss the issues that were kind of laid out here, and try and give us your sense of when, presumably when the reauthorization actually occurs, how we can proceed, because there may not be enough time to have another meeting before we actually have to go ahead and start amending the regulations or implementing new procedures.

So if we can kind of get a general discussion of some of the topics that you have heard here, that would help us when and if the time comes that we actually have to implement something.

CHAIR HARVEY: Do we have any concept, this is Maryanne Harvey, of when it might happen? Are they talking as though this could be in the near future?

DR. BARR: This is Helen Barr. It's really difficult to tell, Maryanne. They are talking. They are planning to convene a meeting in the next couple of weeks, and having not met since the hearing, the principals involved are going to meet within the next few weeks to discuss what came out in the hearing. All I can say at this time is that it should be in this Congressional session, how quickly, I don't know.

CHAIR HARVEY: Debra is waving her hand.

DR. BARR: Oh, yes.

DR. IKEDA: This is Dr. Ikeda. I have a lot to have comments on, but I think that before I make my comment, I understood there was some other information we are going to hear from Michigan?

DR. BARR: This is Helen Barr. Actually, Dr. Finder has a couple of specific questions to address to the committee, and then we could also have Mr. Camburn comment on his experience in Michigan about sort of a rating system for facilities, so however Dr. Finder would like to have that happen.

CHAIR HARVEY: Okay. Well, do you have specific questions that you would like to throw out to the committee, Dr. Finder?

DR. FINDER: Yes. They were under a different topic.

CHAIR HARVEY: Category?

DR. FINDER: You know, different categories.

CHAIR HARVEY: Okay.

DR. FINDER: And I was hoping that in the general discussion, these would come out and if they weren't answered, then I could raise them. Actually, I would be interested to hear what Michigan has to say.

MR. CAMBURN: Okay. This is Jim Camburn from Michigan, and we did try our rating system of a number of years ago for mammography facilities, actually, quite a number of years ago. It was in 1992. And at that time, we had a Mammography Quality Control statute in Michigan. We had one beginning in 1989, but 1991 was the first year we really were able to implement part of it in terms of making inspections of 100 percent of all of the mammography facilities in the state.

Prior to that time, we were spot checking, but 1991 was our first year of annual inspections. And what we found was that with phantom imaging, there was a wide range of performance. Many facilities would just blow the phantom image away. They would image more of the test objects, many more than required for ACR accreditation. Others would struggle, but they barely meet the minimum number of test objects that needed to be imaged for accreditation, and still others would fail.

And at that time, we had about a 20 percent failure rate, and it seemed to us that mammography facilities were not necessarily taking quality control issues to heart that much, especially compared to the way mammography is practiced today, and we thought it would be helpful to facilities if we could, somehow, share with them the results of our phantom image tests, show the problems and encourage them to make some improvements.

So we came up with a grading system. If a facility just barely met the imaging standards for meeting the ACR standards, we gave them a C. If they did a little better than the minimum required to pass, a B, if they did a lot better, an A, and if they failed, they got an E or an F. The F was if they failed terribly.

As it turned out, 20 percent failed. 50 percent got As and Bs. I guess that leaves about 30 percent for Cs. And we mailed these grades out to the mammography facilities in the state, and it wasn't two days, I don't think, before one of the major newspapers in the state caught wind of this and requested from us under the Freedom of Information Act a list of all the facilities in the state and the grades that they got, and we were by law required to release that information, which was then promptly published in the newspaper.

A lot of facilities were irate that we had done this. They found that some patients were shopping facilities. You know, maybe they had been scheduled for a mammogram at a facility that got a C, and they wanted to cancel that appointment and go to a facility that got an A or a B, and it was never our intent to suggest that that one single criterion of phantom image quality was the sole criterion to use in selecting a mammography facility.

I mean, it didn't rank the qualifications of the radiologist or the technologist or how new a mammography machine they had or film processing materials or anything else, but that is how it was interpreted. And there was a cry throughout the state by many mammography facilities for us to immediately stop that ranking system and go to a pass-fail, either the facility imaged adequately or it didn't, and that's what we did and we have been doing pass-fail ever since.

CHAIR HARVEY: Maryanne Harvey. Sometimes your best efforts to try to be helpful don't always work out the way you want it. Many years ago, before we got into MQSA, New York State actually had an internal Quality Assurance Program that we put into affect back in about '87, and out of that came a phantom score, so we knew what the phantom scores were, and then the newspapers wanted to know what the phantom scores were and I didn't want to give the phantom scores, because my theory was the same as yours. It only represents a small portion of what is going on.

Well, under the Freedom of Information Act, I lost, and the worst part about it was it was in our local area and one of the chief techs at a facility who had actually gone out and bought a new piece of equipment raised his score by 40 percent. The old inspection numbers came out and were printed in the paper and, of course, they never show up, you know, in the same font, bold, when they make, you know, the error correction that this facility had gone through all this trouble to try to increase their mammography quality assurance and I felt terrible about it.

Not that our system was inaccurate. It wasn't, but everyone who called would get a 10 minute lecture on how you pick a mammography facility and it's not based solely on the phantom image. But sometimes what you say and what people want to hear, everybody wants one number, and so sometimes you end up -- but I think that's all part of the growth curve, not just for the inspectors, but for everybody to realize that these things are important and that image quality is important, and it's a way then to get a lot of people's attention.

Unfortunately, it wasn't. This one facility lost, at least temporarily, I don't think for very long, because they are a good quality facility. Sometimes, that's what does happen. So then we went to MQSA. Right. Dr. Karellas?

DR. KARELLAS: Every expert knows that phantom scores represent only a fraction of the total image quality, even in the phantom itself, because frequently you will see a phantom that all the features are seen and the score itself, even after corrections for artifacts, will be acceptable.

However, when you look into the phantom and all the other artifacts that are objectionable, you may find, in fact, very frequently we find phantom images that they are not acceptable. Now, although technically what is to be seen there is seen, and you can receive a passing score.

On the other hand, the image quality has so many other components, that people often disregard, that have to do with positioning and all the other factors that represent the image quality that we do not see in the phantom. So I think we all know from the experts point of view, but perhaps from the public's point of view just representing just that score is difficult or providing any score on the facility is definitely possible, but I don't know who is willing to undertake this task and do a good job on that.

CHAIR HARVEY: I don't think anyone, at this point. Any other? Yes, Dr. Ikeda?

DR. IKEDA: I think I would like to. Can I address a question that was raised by Dr. Barr?

CHAIR HARVEY: Sure.

DR. IKEDA: I would like to address the question of mammography self-assessment by radiologists, because I think that was an important issue. First of all, I wanted to recognize the considerable efforts and dedication of Diana Rowden and others at the Susan Komen Foundation and other groups in trying to improve patient care and prove quality for mammography, and a lot of that had to do with implementation of MQSA and those ratings.

I was in Michigan at the time those ratings came out in 1991 and, of course, we got an A. Just kidding. But that was before MQSA, and MQSA has really turned around quality in mammography to make it equal all over the United States.

I am very concerned, however, about testing radiologists' ability. They call it self-assessment, but I am pretty worried about that. And some of that is reflected in that letter that you wrote, you read earlier during this session. And there is a perception, and there is a public perception that mammography is 100 percent accurate, that's a black and white and a cut and dry test, and it is not, and I think everybody knows that.

There are many technical limitations of finding breast cancer against a dense background. I see many radiologists and mammographers struggle with four films on a viewbox or on a monitor to try and find the dense cancer against a dense background that looks like a polar bear in a snowstorm, and you cannot see it, or just trying to find the most subtle findings of cancer.

My concern about trying to test radiologists' abilities is that, again, there is a spectrum and some people say that trying to read a mammogram, as one of my male colleagues said, is trying to like read your wife's face when you come home late from a golf tournament and you're late for dinner or when you're late for something, and you are trying to read the expression.

There are some things that are cut and dry and other things that are not. I am concerned about testing of radiologists, because I am afraid that it is going to turn radiologists away from mammography. There are concerns about this just as was raised with the issue in Michigan, for example, that these results will become discoverable. I think that was also addressed in Dr. Berlin's address to the committee on Health, Education, Labor and Pensions on April 8th.

First of all, radiologists are in short supply and breast imagers are in even shorter supply. I think that was one of the reasons that we got this article about "Can Radiographers Read Screening Mammograms?"

Dr. Lettie Bassette had just done a survey of radiologists and residents who were looking to figure out what specialty they want to go into, and they were not as interested in breast imaging for number one, the stress of it, because you want to find the cancer. I mean, there is some woman depending on you to find her tumor. Second, they are worried about the litigation and malpractice issues, and third, all the regulation that is involved in it, and when running a center doing all the quality control, which is necessary to run a good center.

But those are some of the reasons that people do not want to go into it, and it's always a fine line to try and balance quality against turning people away and turning people off of it. I am concerned that if you put another test in this arena, that there will be a further disincentive for radiologists to go into mammography.

First of all, as Dr. Barr said, there is a perceived crisis in malpractice litigation for mammography, and one of the unintended consequences of this may be that these self-assessments will then say well, Dr. A is worse than Dr. B when it's actually an evaluation of a certain set of films. So the liability of that could be tremendous from a malpractice standpoint.

Other issues that radiologists face are the high costs of malpractice insurance. What happens to the radiologist that gets a B or any other thing in such a test? Is their malpractice going to go higher, and then will they end up going out of business? The workload is quite a bit in any practice and the stress levels are pretty high, and thirdly, for all this effort, there is a rather low reimbursement for doing mammograms compared to other imaging modalities in which reimbursement is much higher.

There is a marked decline in the availability of radiology faculties at universities or in our ability to recruit and our ability to recruit technologists or have people actually go into the health field. So those are my things that I would like to discuss and that I am concerned about in putting in a self-assessment and then having such findings discoverable.

CHAIR HARVEY: Thank you. Maryanne Harvey. One of the things I understood that Dr. Barr had said, that there would be a substitution, so that five hours of the 15 hours that are required now would be for some form of a self-assessment. So it would seem as though the doctors need to get some type of continuing education anyhow, and this is shifting the focus into a different direction in which you would perhaps either add a course or maybe in your office go through a CD or some other set of exercises in which you could test yourself.

There must be some way we can make this, so that this is a personal test and the information isn't necessarily available for other individuals to see. I remember when we first started our mammography program back in 1986, we told people this was coming and we let them get a hold of phantoms, and everybody who had a general purpose piece of equipment that they were using for mammography disappeared, didn't do it anymore.

They took a look and saw what they didn't see and realized they weren't playing in the game. So sometimes, I think if you let people go through some of these self-assessments, they may realize that they don't have the skills or they have lost the skills, and maybe they would self remove themselves from doing mammography. I think part of it is an opportunity for people to assess themselves against some program that has been set up.

Up in Syracuse, I hear that there is a program in which there is a merry-go-round of films that come up, a set of four. Each set of four has one film in it in which there is some kind of a change that needs to be identified, and as the individual goes through and reads these, you receive some score about your proficiency in being able to pick up as you go, so you can sort of self test. And they are working with this at the hospital and seem to be happy with what they are finding out about it.

I haven't had a chance to talk to the researchers on it, but sort of secondarily I heard that people who have ADHD, people who are actually better readers, they can't focus for a long time at anything, I guess, but actually were able to focus very well at reading this. So, you know, you learn other things, too, I guess, as you do some of these pieces of information. Dr. Barr, I can see that, yes.

DR. BARR: This is Dr. Barr. I think Dr. Ikeda and I, as well as the witness testimonies, outlined the myriad hazards that could be associated with this information being discoverable. I wonder if Congress made the information non-discoverable by law how the committee would feel about such self-assessment.

CHAIR HARVEY: We have a guest from the audience. Charlie?

DR. SHOWALTER: I'm Charlie Showalter, American College of Radiology. We have been talking with Congress about this quite a bit over the last year or so. Last fall when the Democrats were in control of the Senate, the discoverability issues was one that they really couldn't deal with and didn't want to deal with. They were unwilling to agree that any results of this self-assessment would not be discoverable, and we were unwilling to agree that it was a good idea unless it is non-discoverable.

So I just wanted to be sure that everyone on the committee understood where the ACR was coming from on this, that we have a product. We think that probably it's not a bad idea or maybe a great idea either, but it's certainly not a bad idea to have it as a part of the 15 hours, but only under the condition that it is non-discoverable under any circumstance.

CHAIR HARVEY: Yes, Dr. Pisano?

DR. PISANO: I just want to endorse some of the points that were made by Dr. Ikeda and the last speaker from the ACR. I don't think I can emphasize enough to this committee and to the public sitting here how beleaguered mammographers feel right now, I mean, breast imaging radiologists.

As was mentioned, we have the highest malpractice rate in terms of numbers of suits per reader of any area in radiology already, and it's difficult to recruit people right now into our discipline. There are many unfilled fellowships that were just mentioned, and I must say in general, I am very concerned about increasing requirements period, because I am afraid no matter what it is, no matter how innocuous it is, it will chase people away from our field.

You know, people who go into radiology have lots of choices. They can choose to do neuro MR, which one could argue has equally strong effects on patient lives as mammography, in fact, maybe more so. And yet, they never have anyone looking over their shoulders if they do that. There are no regulations and there is a much lower risk of malpractice and they get paid better on top of it.

So every single requirement that is added to the MQSA requirements is in that context, and basically what we're doing is making it more and more difficult for us to recruit radiologists to our field, speaking as someone who trains residents, and I know Debi does, too, trains residents and trains breast imaging experts as a fellowship director.

So I just want to caution everyone. It's not just the issue of malpractice and discoverability. It's really the issue of being able to train people in the future and really reducing access. I am concerned that we are going to reduce access in the long term to breast imaging. There are a few exams.

I actually think, and I am not expert on this area, but I know that at UNC, we have a conference where we review all of our cases and it's a peer review conference. And because it's a peer review conference, it's not discoverable. Everything we say there is not discoverable. So I am wondering if that's kind of a back door way out of it being non-discoverable is to call it peer review, and if it's peer review, then it's not discoverable.

So no matter what the Congress does, if you carry out peer review activities, if you review your own performance in a way that you can call it peer review. In other words, I am reviewing my colleagues. They review me by my assessment of myself. I must say I don't think on its face, there is anything really bad about this, except for the fact that people may use it incorrectly and the fact that it may scare people away from our field.

I mean, I really think we're at the point where incremental changes may make a huge difference, and I want to echo what the third person who spoke to Congress, the minority person said, was this may have a big affect on practice patterns. We may really see people leave in droves. This may be the straw that breaks the camel's back, and we better not be authorized for five years if that happens. I have major concerns about this mainly because I feel like we're already on the edge of a precipice and we may be pushed over with this.

CHAIR HARVEY: Yes, Dr. Showalter?

DR. SHOWALTER: I just wanted to observe there is another thought that was put out by Dr. Dershaw at the hearing, and that was basically just what Dr. Pisano was saying, that we have got all these requirements and to put this additional requirement on us is unreasonable. Something needs to go, and so as a follow-up to what her suggestion was this morning about looking at the inspection, one could also take another look at the current requirements and see.

They were written many years ago and I know when they were written. I was there and it would be fair to, at some point, now or some later time, take a look at the whole set of requirements and see after some years of experience, and this is the thought that Dr. Dershaw was making, are we where we need to be here?

DR. PISANO: I just have one other point about the other suggestions that were made about keeping track of how many mammograms are read and having the FDA have access to that data. I find the most onerous thing for me personally as a radiologist of the inspections is presenting information on the audit.

I also find the whole audit extremely onerous. It's not that we don't do what we're supposed to do for patients. We do, but having to put together a report for each of my radiologists every year when I know that we review -- I will just give you what we do at UNC on top of what is required.

We actually review as a group, all five breast imaging radiologists with the residents and fellows, we review every single patient who has had a biopsy and every problematic patient. As a group, we do that every week. We spend two hours per week doing it, and I think that is what is needed for patient care purposes. And to have to sit down at the end of the year and tally it up and give a report is sort of beside the point, and so I find that an extremely onerous requirement without much benefit at my practice.

Now, if I were in setting where I never reviewed the mammograms, maybe it had some benefit, but I am not sure it does. I am not sure that a private practice that doesn't ever look at its mammograms after the biopsies are done is doing what it needs to do for patients. I think that it would make much more sense for us to require the extras that we are doing, which is that is to have somebody look at every mammogram after a biopsy than to tally it all up and give a report at the end.

We are not epidemiologists and I don't think we're doing a very good job with these exercises. So I really wish we, if anything, get rid of that requirement not because patients care. I obviously want patients to get good care, but I don't believe that knowing that I have a 10 percent versus a 4 percent call back rate really affects people.

I don't think those abstract numbers make a difference. What I think that makes a difference is seeing the cases that you call back that were false positives, seeing the cases that prove to be cancers. That makes a difference. That is good quality. That is how you improve your mammography performance, not knowing what your absolute rate is at the end of the year. So I have a lot of problems with the current auditing requirement, so I don't want it increased is my point.

DR. YOUNG: Maryanne?

CHAIR HARVEY: Yes.

DR. YOUNG: Yes. Dr. Young speaking. A couple of comments and I am going to echo some of the recent comments made by Dr. Pisano and Dr. Ikeda. But when I train residents and fellows, when I look at self-assessment skills or assess skills, they fall into two categories within mammography.

The first is their perceptive skills. There are some people who simple don't see abnormalities. They don't appreciate them as an abnormality. The second aspect of skills assessment is the way they render the report. In other words, their interpretive skills, and I separate those two.

There are some people who have very excellent perceptive skills, but poor interpretive skills. Hopefully, in mammography we'll find that group of radiologists that has good perceptive skills and good interpretive skills, because once you have the information at hand, it's how you portray that back to the patient and to the referring physician, so that you map subsequent patient care episodes. It's training and experience that hones mammographic skills. It's not tests.

CHAIR HARVEY: Yes, Dr. Ikeda?

DR. IKEDA: I want to just reiterate that it always sounds really good to do a self-assessment test and have it to be that much of your continuing medical education, but it really does have, I think, unintended consequences, and those consequences, I think, may be severe and may impact the health care of women all across the United States, but limiting access because of a shortage of qualified, trained radiologists, because people are not going to want to go into this field if they are going to get tested.

And there is one little paragraph from Dr. Dershaw that said "While self-assessment testing may be of value, it should also be recognized that there are no data to indicate that such tests provide feedback that accurately determines competence. There is also no science to indicate that such tests result in improvement and quality of medical care."

I do agree with Dr. Pisano that getting feedback on your own patients is probably the best thing that radiologists can get to become better, and it is the law that we must obtain biopsy data or histology or follow-up on every single mammogram that we call either a 4, suspicious, or 5, highly suspicious for malignancy, and so that data is available to every facility.

You said you need to get your own follow-up data and it is there in every facility. So again, I want to strongly recommend, number one, that I don't think it's a good idea, number one, because I don't think you are going to be able to recruit anybody else, and I think older radiologists, I am not trying to be discriminatory, but radiologists that are just in edge of wanting to retire, they are thinking about it, I can see that some of them are not going to want to.

Although, they want to help women, they just don't want to go through one more thing and will just drop out. I am sorry to say that. I want good radiologists reading mammograms. That was the full intent, I think, of this discussion, but I just don't think it's a good idea.

DR. HENDRICKS: Carolyn Hendricks. I am going to speak to this issue from the standpoint of being board-certified in internal medicine and board-certified in medical oncology, and there has already been a tremendous level of acceptance of some very stringent and time consuming and expensive testing required for recertification in the medicine specialties and I accept that, and I think that recertification, including self-assessment modules, which are now required by the ABIM, are actually going to improve the quality of care that the medicine internists and the medicine subspecialists provide.

So I am concerned to hear a reluctance on the part of a specialty for self-assessment. I think it is coming. I think that especially with the trade off in the terms of number of hours, because I think everyone agrees that we really want high quality mammographers and that radiologists who do breast imaging know that there is a difference and that you see poor quality imaging in your community.

And so when I am looking around and see my subspecialist colleagues, I know that this will be a test, and I think that knowing that the subspecialists in my community who are able to meet those stringent requirements for recertification, which has a big emphasis on self-assessment, I think that is going to improve the quality of subspecialist care in the community across the United States.

DR. IKEDA: This is Dr. Ikeda again. I understand your concern. I would just like to find out, since I am unaware of what the requirements are for internal medicine, what exactly are the requirements. Is each medical practice or oncology practice required to report to the Government about how many patients they say, the mortality data and then last, are those tests discoverable and is the data on which patients are seen, the number they are seeing and the mortality data or adverse outcomes reportable to the Government on a yearly basis, because that is basically the burden that we are under right now. You know, it's the law, so we have to do is and probably it is helpful for quality overall.

DR. HENDRICKS: Carolyn Hendricks. And I just want to speak to the self-assessment component. These are standardized tests of recertification. It's boards. There is a grandfather clause, which might be appropriate in this context whereby, for example, older radiologists nearing the end of their career would not be required to participate in self-assessment. But, you know, they are standardized tests, so they are modules that you get that you submit for scoring, and the information is available and, of course, hospitals require recertification for participation.

Now, JCAHCO, for example, would require that the medical staff be board-certified to practicing their specialties. I think it's something that is not foreign to radiologists. This is the process of board certification, which now includes a big component of self-assessment. And like I said, I was very discouraged that I would need to undergo these self-assessment modules, but I am young enough to be vulnerable and to be affected by it, so I'm doing it and I really think that it will improve the quality of care.

DR. IKEDA: This is Dr. Ikeda again. I guess I asked also a second -- well, first of all, I think that there is --

CHAIR HARVEY: We can't hear you.

DR. FINDER: What did you say?

CHAIR HARVEY: We can't hear you.

DR. IKEDA: For the ABR, there is a reassessment for radiology, so there is a reassessment test for radiology and radiology residents who go into practice. That is for radiology. Correct me if I'm wrong.

DR. PISANO: No, you're right, you're right. You're correct.

DR. IKEDA: That's the first point, so there is a reassessment, and there is reaccreditation for the ABR. I think I am correct.

CHAIR HARVEY: Dr. Harrison had a question.

DR. HARRISON: Yes. Dr. Miles Harrison. My board certification actually is in general surgery and we, indeed, have requirements for self-assessment exams and for recertification. I think, however, you are comparing apples to oranges in this particular case, because when we look at our recertification, the information that we are being tested on is pretty objective as we do our recertification.

We are talking about a process that has some inherent subjectivity to it as one radiologist or let's take 10 radiologists looking at a film who can come up with a different interpretation. That's maybe too broad. Three or four radiologists coming with different interpretations.

So to compare our recertification and self-assessment to what has been discussed here, I think it's not a fair analogy. We as general surgeons do not have to report case loads. We don't have to report outcomes as we are doing our recertification. I don't think an analogy exists here at all to what they are saying with other subspecialties, self-assessment exams and recertification.

DR. PISANO: I just want to second that. This is Etta Pisano again. We are talking about federal requirements, as opposed to board requirements.

DR. HARRISON: We don't have any.

DR. PISANO: There is a big difference. If we were sitting here and talking about the American Board of Radiology requiring this, I don't think you would have -- and if it was, essentially, that is definitely peer review once you get -- the board exam stuff is your private information within your specialty. You can say you are board-certified or not, but they can't find out your score or anything like that. So that would be a whole different conversation.

If the board were talking about this, I don't think you would have the same concerns from us. It's the fact that the federal Government is requiring, is talking about requiring it, that it's making people uneasy, especially in the context where, you know, nothing else is being regulated so stringently. Cardiology isn't regulated anywhere near mammography and, as we all know, heart disease kills many more women than breast cancer.

So why aren't we regulating cardia catheterizations? Why aren't we making sure that cardiologists know how to read cardio caths? I mean, those are the kinds of questions I think people have, and neuroradiology isn't regulated at all compared to mammography, and we compete directly with neuroradiology in trying to get people into the field. So I think those are the issues that really -- I applaud your board's, you know, assessing your competence through self-assessment. The issue is really the board versus the Government. That's the issue.

CHAIR HARVEY: Dr. Harrison, did you have another comment?

DR. HARRISON: Well, I don't know whether I'm beating. I have made the comment.

CHAIR HARVEY: Okay. Dr. Karellas?

DR. KARELLAS: Dr. Hendricks, I think she made a very good point of a practice that is practiced by many fields, including radiologists. The board does have certificate of added qualifications in several specialties. I am not aware about mammography and I will not speak about it, because it is a special case. But the trend is very similar among the different specialties, including radiologists.

The big question here is as Dr. Pisano pointed out, is the question not about self-assessment. In the letter by the two radiologists who testified to Congress, it has been made very clear that most radiologists would be willing to do some sort of a self-assessment test provided that this is done under the conditions of non-discoverable, peer review and all of that.

So the question is really whether it is mandated by the federal Government versus a peer review like certification or institutional requirement, and there is a huge difference between the two. So I think this is where the problem is.

DR. HARRISON: Harrison. I think I will go on and state it. When we look at our ability to sit for our recertification examinations, there is no requirement that we have had a certain low level of untoward events. So you can sit for that exam, take that exam and pass the exam. Yet, you may not be a safe surgeon. I hate to say that and I don't want to frighten anybody there, but generally, the exams help weed those kinds of people out, but there is not the onus to report as we're speaking here. So I just -- rubs you in the wrong way.

And I have got a direct stake in this as a general surgeon who does 65 percent of my practice with breast and breast diseases. I am finding that the experienced radiologists who read for me for years, indeed, as they get close to that retirement age are just not wanting to deal with all the regulation. So consequently, I am, I am 54 now, so I can say this, standing with a couple of kids trying to make a decision about a mammogram and their cumulative knowledge and experience doesn't exceed mine.

And so, therefore, it makes it difficult for me to do what I do if the experienced radiologists reading mammograms are not going to stay in the profession, and I think when you are looking at regulations like we're talking about, indeed, folks are going to leave as soon as they can.

DR. HENDRICKS: Hendricks. If I could respond? I don't want there to be too much emphasis on the component of my comments that related to board certification. It was more the self-assessment component, which is a new component of board certification since board certification has been around for a long time.

I think we agree that quality is something that is hard to measure and that radiology and breast imaging is unique in many ways, but the consumerism of medicine and the fact that cardiac catheterization is not far down the block from being thoroughly investigated with a consumer report card sort of a thing, I think we have to acknowledge that this is coming and I have to say that I understand that self-assessment might be somewhat intimidating or a little bit of a turnoff in terms of a way to assess quality, but I think it is coming, and I think if we're specifically talking about a trade off in a certain number of hours that are currently required and not necessarily more hours, we really have to weigh the value of that information with the discoverability issues aside.

DR. IKEDA: This is Debra Ikeda. Dr. Hendricks, I think we agree that self-assessment is good and that people learn very much so from self-assessment, but again, the point, I think, is that this would be a federally regulated self-assessment and discoverable and, of course, for boards, if it is peer reviewed, it is non-discoverable and, of course, this is a good thing. But if it is discoverable, it is going to drive radiologists out of this field and I have no question about it. Radiologists, I think, who want to be better.

Everybody wants to be good or the good people want to be good. Everybody wants to be good, and most people want to do the best they can, and self-assessment is a key component. But again, we're talking about federal regulations and I think at this level and from here, what the ACR has said, I do not think that a federally regulated mandate to do a discoverable five hour self-assessment is going to be an act that is going to help women's health across the United States, because I am concerned it will drive more radiologists out of the field.

CHAIR HARVEY: Dr. Barr?

DR. BARR: This is Helen Barr. Again, I raise my question and I think I have only heard Dr. Pisano address it. Leaving the discoverability issues aside, because I think, you know, we have heard tons of evidence that there are problems with discoverability. But if I go back to Congress with what this committee feels, I would really like to know what they feel about self-assessment if it were made non-discoverable, and so far I have heard Dr. Pisano say that whether it is discoverable or not, she has some concerns about there being another regulatory requirement, and I am wondering if anybody else can address the issue, so when I talk to Congress I can give them some sense of that.

CHAIR HARVEY: Dr. Barr, I had another question, too.

DR. BARR: I get the discoverability issue loud and clear and, believe me, so do they. I don't know what they are going to do about it, but they get the message loud and clear.

CHAIR HARVEY: I was wondering. Has there been any discussion in the reauthorization about increasing the number of films that radiologists or radiological technologists perform or read in a year, that perhaps 960 is really too low?

DR. BARR: Early on in the reauthorization process, I did hear some discussions of that, but that seems to have gone by the wayside in the wake of this physician self-assessment discussion.

CHAIR HARVEY: Was it either one or the other? Does it feel like it's going to --

DR. BARR: That's what it feels like to me right now. I can't say that's what it's going to be. I think there is some feeling that we don't know --

CHAIR HARVEY: Right.

DR. BARR: -- the number of mammograms that physicians have to read and that without that knowledge, it's hard to increase the requirement. Plus there are some studies, Beam studies, that volume actually dis-equates with how well a physician reads. So I think the answers there are too nebulous. I think the focus on self-assessment is this is tangible.

CHAIR HARVEY: Right. People are looking for --

DR. BARR: It feels right whether there is evidence or not and, as I said, if I could hear from anyone else, it might be helpful or else I will just take Dr. Pisano's opinion that any regulatory requirement, discoverable or not, seems burdensome.

DR. HARRISON: Well, Etta, was that an accurate statement of your position? Okay.

DR. PISANO: Yes, I really do want to repeat. Let's, if anything, take regulations away and not add. Originally, 10 years ago, I wouldn't have believed that I was at this point.

DR. HARRISON: The analogy to general surgery, we are actually in our 13th revision of our self-assessment examinations, all of which is voluntary. You can take that if you would like. It is clearly not discoverable. It goes directly back to the American Board of Surgeons, the American Board of Surgery, and maybe some combination and permutation of that taking the discoverability away and making it a voluntary self-assessment may be something to look at.

CHAIR HARVEY: I mean, would the ACR make this as a requirement as part of the certification process? And I would also like to hear from the non physicians on our committee, because I think there are other opinions perhaps about this subject also that need to be expressed. Do we have anyone?

DR. YOUNG: Maryanne, Dr. Young, I'm not a non physician, but a point of information, Dr. Showalter might be able to answer this, doesn't the ACR have a couple of CD-based self-assessment courses available? I was at a meeting on Friday where one of the discussants in an interactive session pulled a couple of disks out of his pocket and recommended them.

DR. SHOWALTER: Yes, we do. We have two CD-ROMs that are available currently. They are for self-assessment, continuing education. A physician can go through and answer a series of questions, send the results in. They will get a score. They will be issued the continuing education credit, and then shortly thereafter, their score will be disassociated with their name, so that --

DR. YOUNG: And so there is no --

DR. HARRISON: That's exactly what our board does.

DR. YOUNG: No discoverabilities?

DR. HARRISON: There is a number, but it's not associated.

DR. YOUNG: It's an excellent series of programs.

MS. PURA: Well, I'll stick my neck out. Linda Pura. Being a Komenite and also being definitely a consumer advocate, I do believe strongly in self-assessment and I do believe in a skills based hands-on type of assessment. I think it's vital in all professions that are medically oriented, and I feel strongly that it should be inclusive.

As to the question as to voluntary, involuntary, I would like to hear more discussion about that, but I think five CMEs is certainly not difficult to achieve and a CD-ROM type of situation is very comfortable in your home, in your office, etcetera, and I do know that this is available to everyone who is practicing radiology.

I am afraid I do believe very strongly in medical outcomes, and especially with Category IVs, so that I feel that self-assessment should be included as a requirement. I would like to see self-assessment required in more of ours. I mean, I have to get 30 units continuing education every two years within my field, and none of it is required to be hands-on self skilled, etcetera, but I do feel strongly about that. I would like to see it. I have been doing that since I have been a nurse for centuries now. So I have to disagree with those who do not feel that it should be another inclusion, but I do.

DR. HARRISON: Inclusion or requirement?

MS. PURA: Requirement.

DR. HARRISON: A different word.

CHAIR HARVEY: Dr. Karellas? Dr. Ikeda?

DR. IKEDA: Dr. Ikeda. I just wanted to restate my position for Dr. Barr. I support Dr. Pisano's statement. I do not think additional regulations would be helpful. I discourage it and I think it will drive radiologists out of the field. I also agree IVs are important and to audit is very important, but it is required by law, and so every radiologist in every facility has to look over IVs and Vs and report them to the Government yearly by physician and by the facility.

DR. PISANO: Actually, I don't know if this is state -- maybe it's different in your state than my state, but in my state, we don't have to tell the Government how many IVs and Vs we have. We just have to show that we have done it ourselves.

DR. IKEDA: We have to show that we follow-up on them and that we get the biopsy results, but we have to report that we have done it and we have to see --

DR. PISANO: Exactly.

DR. IKEDA: -- all of our IVs and Vs, so I probably misstated that.

DR. PISANO: Okay. I just was curious if it was different in California.

CHAIR HARVEY: Dr. Karellas?

DR. KARELLAS: Andrew Karellas. I believe that many of us, if not most of us, feel positive about overall self-assessment continuing education of some sort. The problem is that number one, we do not know what kind of self-assessment. How would that be done? It's very difficult to approve something or promote something or being very strongly in favor where you don't really know what the details are.

Certainly, additional continuing education may not yield many results. You can just sit there even doing a self-assessment. Besides, what is self-assessment? That self part is somewhat -- is that a computerized test? Is it something as it was mentioned before that you take the test and then your name is disassociated with the test, and that was a good experience for you? And do we have any proof that this kind of testing really helps?

So it's an interesting exercise, but I am also concerned about making radiologists just putting more time into what we think is a scientific way of improving their skills and 10 years later, we conduct a study and find out that that whole thing was a waste, because people don't really learn anything new. So we have to be very careful about that kind of a self-assessment.

First of all, we need to know what is proposed. There is no question in my mind that if there is a federally mandated and discoverable process for this kind of the self-assessment, that many radiologists who specialize in mammography will no longer specialize, and the sad thing is that some of them will be the good ones, and perhaps some of them will be perhaps the not so good ones, but then that result will not be positive.

DR. FINDER: This is Dr. Finder. I would like to try and focus on this issue and address some of the questions, because I had a number of questions and Dr. Karellas actually touched on many of them. I think we are all at a disadvantage, because again, this meeting was supposed to be scheduled after the reauthorization, so we would have known exactly what Congress had come up with and we're not at that stage.

I would pose, however, the following question. Assuming that they do pass in the law, it no longer is now voluntary, it's mandatory, a requirement that five of the 15 CMEs be in self-assessment, and that these be non-discoverable. Let's assume that. I can't guarantee that, but let's assume it.

Then I have a number of questions that I think we need to deal with, because if they do put that through, we're going to have to take some type of action probably before the next meeting. Let me kind of put it into some background. Right now, we require 15 CME every 36 months for the physicians. It all has to be Category I. And as I understand it, this self-assessment only applies to the interpreting physicians. It does not apply to the mammography technologists. It does not apply to the medical physicists.

So assuming that they are going to put a new requirement in for self-assessment, five of the 15 have to be in self-assessment, non-discoverable. Then I would like to kind of deal with some of the questions that you just raised. First of all, what does it mean to be self-assessment? And I have a number of questions there.

Does there have to be a test involved? If there is a test, is the test, you know, graded by the person? Does it have to be graded by the person putting on the CME? Is there a passing score? Do they have to be somehow be able to compare themselves to some type of local or national average to see how they are doing, those types of things? Does it have to be limited? And we're not sure what kind of language they might come up with, but this self-assessment thing may have to be limited to interpretation, so that a general CME type course that we would accept, let's say, on various aspects of mammography might not be acceptable depending on how they phrase this.

So these are the types of things that we could use some guidance on. I'm not saying that we're going to come up with anything definitive, but assuming that the law is changed, we would like to try to get some detail on that.

DR. HARRISON: Miles Harrison. The specificity within radiology and the onus of that subjective reading, the interpretation, if you will, I think does make it slightly different from other subspecialties, but we don't need to reinvent the wheel here. We don't need to question whether self-assessment examinations work or not. I mean, we have exit polls prior to receiving your recertification. I have recertified two times now.

The exit poll asked whether you did see self-assessment examination for surgeons or not. So they know that up front, and the people who did that, the passage rates directly correlate with those of us who did CSAP and you find a much greater population of surgeons who did not do CSAP. So we don't need to reinvent the wheel. Do self-assessment examinations work? Yes, they do. We don't need to ponder that question.

I do believe that the difficulty here is because there is this subjective component that we just -- of course, we have to case manage, but in a real sense when they put it in an objective test, you are taking a subjective subject of clinical acumen and putting it in an objective form, and I think we could object. I don't know. I am not a radiologist. I suspect that the intellect that we have around this table and in this United States, we could objectivize that in some way, shape or form to come up with this answer. We don't need to reinvent the wheel. It works.

Now, Deb, I saw in your eyes. In no way am I trying to put another -- well, no. I am not trying to put an onus on radiologists to have another requirement. I am in agreement with the general gestalt that this kind of requirement, certainly if discoverable, creates a problem. I think we are all in agreement with that.

I also would agree with Etta who says that we need another thing to do like we need a hole in the head. Okay. So then with that as my sentiment, do you understand that I am not talking against what you are feeling, but I am saying we don't need to reinvent the wheel. In fact, that tool does work, and I think that if sober minds, if we're going to head in that direction, then I really think that a subjective thing like interpretation can be objectivized in some way. My apologies for that long answer.

CHAIR HARVEY: No, that was fine. Thank you.

MS. PURA: Linda Pura. There are all sorts of types of self-assessment tests and there are self-assessment tests where you compare your interpretations or your answers to the given answers or the correct or right answers, and that is probably the best way to do a voluntary self-assessment test.

DR. HARRISON: That term is called standard of care.

MS. PURA: Right.

DR. HARRISON: And that is exactly what we do, we assess against a standard.

CHAIR HARVEY: Yes, Dr. Pisano?

DR. PISANO: Etta Pisano. Having said that I don't want us to do this, let's assume that Congress tells us that we have to do it. Okay. So I am going to assume now that the Congress has told us that we have to do it. I have actually done a fair amount of work in this issue myself and it is extremely complex, creating a test that is valid and reproducible and actually reflects performance characteristics.

And unfortunately, we have a test already that the ACR had -- that's fortunate that we have that, but I think there aren't that many tools out there to do that, and I am a little concerned about having a test that is valid, reproducible, accurate and all those things that you need a good test to be.

Actually, this is the first time I heard anyone ask does it have to be a test? If it weren't a test, that would be outstanding, because for example, I consider what we do, what I described earlier, of sitting there and looking at every mammograms two hours a week as a self-assessment test, I mean, a tool. We are, essentially, sitting there collectively and talking to each other about maybe you shouldn't have recommended a biopsy on that case. You know, I think it's benign. I thought it was benign up front. You know what I'm saying?

That is as much an assessment of our performance as any test would give us. It might not give us a nice grade at the end of it, but in my opinion, it's more useful. So if there is some flexibility in what Congress sends you in terms of if it has to be a test, then you're stuck, but if you have some flexibility, then practices might actually start doing more of what I just described, which I think is extremely important for patients, not just to figure out what your statistics are, but to actually look at the images and learn from the images. So I like that idea.

MS. PURA: I'm taking this microphone. Linda Pura. I think, Dr. Pisano, you are in, you know, a rarefied situation. You are in a university setting, and I don't deal with those types of mammography units, fortunately. And so there isn't peer review for many of the facilities that I work with, and they need peer review, but we don't have peer review.

So I think that the self-assessment would be very helpful in that kind of a situation where the standards are set, the answers are known, and they can compare themselves to those answers and maybe learn something from them. I think that is really vital, because not everybody is at Stanford or at the University of North Carolina, and at the majority of practice in Los Angeles County, that is not what I see.

I mean, I know a lot of the university settings. I know that City of Hope not only compares their Category IVs with their Category Os and, I mean, I think that is outstanding, commendable, but it's not done everywhere else.

CHAIR HARVEY: Yes, Dr. Ikeda?

DR. IKEDA: This is Dr. Ikeda. I helped write the test, the COMISA test, and I don't know how many hours are available.

MS. PURA: Seven.

DR. IKEDA: Seven hours are available for something that we wrote over 10 years. We started, I think, in '92, '91, '92. I have been on that committee. Hey, listen, I have been on a committee over 10 years now, so I wonder what type of self-assessments, if Congress mandates it that it must be done, those are very valid questions.

Number one, I think if, again, not discoverable, but then would that self-assessment on the CD-ROM fulfill the requirement and are there other self-assessments that the oncologists or medical oncologists or internal medicine docs and surgeons also use as tools for this?

I think if you are in a setting at a university, perhaps we could use one of these other self-assessments where we look at things and reassess ourselves and get CME credit for it, but for facilities, like you said, in Los Angeles, they may not be able to do that. But are there other tools available?

CHAIR HARVEY: Yes, Dr. Harrison?

DR. HARRISON: Dr. Harrison. As we have this discussion, I think we can shape what this animal that we're discussing right now is, and that is the use of the word test or self-assessment. Now, this is semantics, folks. Come on, it's a test. Somebody gave you a grade and you flunked because of it, but our self-assessment is exactly that. It is to allow any individual surgeon the opportunity to compare yourself to other surgeons in the United States, so that you are practicing the standard of care in 2003 now, so that no, the score is not discoverable, but yes, you are compared to every other surgeon that took that test and there are only 37 percent that take the test.

So in a real sense, yes, it's a test. No, I don't get a grade. I don't have a sense for whether I passed or failed, but I do know that when compared to those 37 percent of surgeons in the United States that took that self-assessment exam, I knew 90 percent of the questions and only 10 percent of the people knew more than me.

So it's a test, but it's not really a grade. It's not really. So I think it's semantics that we're talking about. And if we are truly in the position to do what I think, then we can shape the animal that gets invented.

DR. FINDER: Yes. This is Dr. Finder. I agree with you 100 percent. We don't know what is going to come out of Congress, but the chances are whatever they are going to write is going to be fairly general, and then it's going to be up to FDA with the help of this committee, and that is why we're trying to get the input now in the concept that they are going to say something fairly general, so that we can move ahead and not wait for another meeting.

But I would assume that whatever language is going to come out, and this is just my assumption, it's going to be non-discoverable. It's going to be self-assessment. They may not use the word test at all.

DR. HARRISON: Yes, it's semantics.

DR. FINDER: And that we then are going to have to try and implement that, and from our standpoint, from an inspection standpoint -- well, let's put it this way. There are at least two ways to look at this. One is from the inspection side and one is from the physician's side.

From our inspection side, we can make it fairly simple to test, you know, to inspect against. We can, there has been some talk in Congress about this, accept an attestation from the interpreting physician that five of the 15 hours was in self-assessment. They may even put in language like that. I can't be sure, but they might. If they do, it makes it easier for us to write the regulation.

However, you can be assured that we are going to get questions from the facilities about what does it mean to be self-assessed? What would you accept? And we have to be prepared to answer those questions and if somebody tells, you know, if this committee recommends and we decide to implement that it be a test, then we can go back and say it has to be a test. It has to be measured against a national not a local level or it can be a local. These are the types of things that we need to discuss, so that when the questions come in, we can respond.

And right now, I agree, we have a fair amount of flexibility. First of all, they haven't written anything and for all we know, they may not even include it as a requirement whatsoever, but if they do, we would like to try and be prepared, so we can move ahead. And I did hear some talk about it doesn't have to be a test. Doesn't it have to be a test? Can it be something else more based on more like the audit and a review of the audit? That is a possibility, obviously.

If it's going to be some type of testing with some type of scoring, whether you get a grade or not, is there a concept that you have to get a certain score? Do you have to get the passing score? Otherwise, the five CMEs doesn't count or is it just a matter that you took this course and whether you got a zero on it or not doesn'