Medicare Coverage Policy ~ MCAC
Executive Committee
Transcript for March 1, 2000 Meeting
(For more information,
contact the Executive
Secretary.)
Note: The language on this website comes directly from the
transcribed testimony taken at this panel meeting. The views and
opinions are those of each of the experts and not those of the
Health Care Financing Administration. HCFA does not edit these
transcripts and makes no assertion as to their accuracy.
00001
1
2
3
4
5 HEALTH CARE FINANCING ADMINISTRATION
6
7
8
9
10 Medicare Coverage Advisory Committee
11 Executive Committee Meeting
12
13
14
15 March 1, 2000
16
17
18
19 Health Care Financing Administration
20 Main Auditorium
21 7500 Security Boulevard
22 Baltimore, Maryland 21244
23
24
25
.00002
1 PANELISTS
2 Chairperson: Harold C. Sox, M.D.
3 Vice-Chairperson: Robert H. Brook, M.D.
4
5 Voting Members:
6 Thomas V. Holohan, M.A., M.D., F.A.C.P.
7 Leslie P. Francis, J.D., Ph.D.
8 John H. Ferguson, M.D.
9 Robert L. Murray, Ph.D.
10 Alan M. Garber, M.D., Ph.D.
11 Michael D. Maves, M.D., M.B.A.
12 Frank J. Papatheofanis, M.D., Ph.D.
13 Ronald M. Davis, M.D.
14 Daisy Alford-Smith, Ph.D.
15 Joe W. Johnson, D.C.
16
17 HCFA Liaison:
18 Hugh F. Hill, III, M.D., J.D.
19 Jeffrey L. Kang, M.D., M.P.H.
20 Consumer Representative:
21 Linda A. Bergthold, Ph.D.
22 Industry Representative:
23 Randel E. Richner, M.P.H.
24 Executive Secretary:
25 Sharon K. Lappalainen
.00003
1 TABLE OF CONTENTS
2 Page
3 Welcome and Introduction
4 Harold C. Sox, M.D. 5
5 Conflict of Interest Statement
6 Sharon Lappalainen 7
7 Opening Remarks
8 Jeffrey L. Kang, M.D., M.P.H. 8
9 Hugh F. Hill, III, M.D., J.D. 9
10 Subcommittee Report/Chairperson's Remarks
11 Harold C. Sox, M.D. 12
12 Open Public Session and
13 Scheduled Commentaries
14 Guido Tricot, M.D., Ph.D. 39
15 Richard Justman, M.D. 43
16 Morgan Downey 53
17 Donald S. Baim, M.D. 61
18 Wayne I. Roe 69
19 Vicki Gottlich, Esquire 80
20 Larry M. Weisenthal, M.D., Ph.D. 88
21 Sandy Sherman 98
22 Thomas Meskan 100
23 Recitation of Letter from ACP-ASIM
24 Sharon Lappalainen 110
25
.00004
1 TABLE OF CONTENTS (Cont'd)
2 Page
3 HCFA Response to Subcommittee Report
4 Jeffrey L. Kang, M.D. 115
5 Hugh F. Hill, III, M.D., J.D. 117
6 Open Committee Deliberation 124
7 Lunch Recess 164
8 Open Committee Deliberation 164
9 Open Public Comments
10 Stephen J. Northrup 258
11 Thomas Meskan 264
12 Steven Lascher, D.V.M., M.P.H. 266
13 Closing Comments
14 Jeffrey L. Kang, M.D., M.P.H. 271
15 Conclusion 276
16
17
18
19
20
21
22
23
24
25
.00005
1 PANEL PROCEEDINGS
2 (The Executive committee meeting was
3 called to order at 8:11 a.m., Wednesday, March 1,
4 2000)
5 DR. SOX: I'd like to welcome everybody to
6 this meeting of the Executive Committee of the
7 MCAC. The purpose of this meeting is to discuss
8 the recommendations of the subcommittee that
9 developed recommendations for all principles and
10 procedures for the panels, and we'll be hearing
11 from a number of representatives of the public
12 today as well as from HCFA as well as from the
13 subcommittee.
14 We're going to start off by introducing
15 the members of the Executive Committee who have
16 made it already. And I'll start on this side,
17 and hopefully people will show up before we get
18 around to the other side.
19 Randel, will you introduce yourself and
20 say where you're from.
21 MS. RICHNER: Randel Richner, Boston
22 Scientific, industry representative.
23 DR. BERGTHOLD: I'm Linda Bergthold,
24 and I'm the consumer representative.
25 DR. MURRAY: I'm Bob Murray from the
.00006
1 Laboratory and Diagnostic Services panel.
2 DR. HOLOHAN: Tom Holohan, Chief of
3 Patient Care Services, VA, headquartered in
4 Washington.
5 DR. HILL: Hugh Hill, HCFA.
6 DR. SOX: I'm Hal Sox. I'm from
7 Dartmouth Medical School and Chairman of the
8 Executive Committee.
9 Jeff, will you introduce yourself.
10 DR. KANG: Hi. Jeff Kang, Health Care
11 Financing Administration. I'll introduce myself
12 later on also. I apologize. I'm a little under
13 the weather here, as you can tell from my voice.
14 MS. LAPPALAINEN: Hello. I'm Sharon
15 Lappalainen with the Health Care Financing
16 Administration. I'm the Executive Secretary for
17 the panel.
18 DR. BROOK: Robert Brook from RAND,
19 UCLA.
20 DR. GARBER: Alan Garber, Department of
21 Veterans Affairs, Stanford University.
22 DR. DAVIS: Ron Davis from the Henry
23 Ford Health System in Detroit.
24 DR. PAPATHEOFANIS: Frank
25 Papatheofanis, University of California in
.00007
1 San Diego.
2 DR. SMITH: I'm Daisy Alford-Smith.
3 I'm the Director of the Summit County Department
4 of Human Services in Ohio as well as the
5 Chairperson of the DME panel.
6 DR. FERGUSON: I'm John Ferguson, Chair
7 of the Laboratory and Diagnostic Services panel
8 as a consultant in healthcare.
9 DR. SOX: Now we're going to hear from
10 Sharon with some procedural matters.
11 MS. LAPPALAINEN: Good morning and
12 welcome to the panel, chairperson, the Executive
13 Committee and members of the audience.
14 The committee is here today to hear
15 reports from its subcommittee and will discuss
16 and consider the levels of evidence and types and
17 presentation of information that it believes
18 should be considered by the medical specialty
19 panels at future MCAC meetings.
20 For the record, I will read the
21 conflict of interest statement for this panel.
22 Conflict of interest for the Executive
23 Committee meeting, March 1, 2000.
24 The following announcement addresses
25 conflict of interest issues associated with this
.00008
1 meeting and is made part of the record to
2 preclude even the appearance of an impropriety.
3 To determine if any conflict existed, the agency
4 reviewed the submitted agenda and all financial
5 interests reported by the committee participants.
6 The conflict of interest statutes prohibit
7 special government's employees from participating
8 in matters that could affect their or their
9 employer's financial interests.
10 The agency has determined that all
11 members may participate in the matters before the
12 committee today. With respect to all other
13 participants, we ask in the interest of fairness
14 that all persons making statements or
15 presentations disclose any current or previous
16 financial involvement with any firm whose
17 products or services they may wish to comment
18 upon.
19 And at this time I'll turn the panel
20 over to Dr. Sox.
21 DR. SOX: Thank you. First we're going
22 to hear some opening remarks from Dr. Jeffrey
23 Kang, who is Director of the Office of Clinical
24 Standards and Quality.
25 DR. KANG: Dr. Sox, thanks a lot.
.00009
1 Given my voice, I actually have some remarks that
2 I really want to make at 10:30, 10:40, and I'm
3 going to ask Hugh to read those for me.
4 I just want to say in addition to being
5 the director of the office, I am HCFA's chief
6 clinical officer, and coverage is one of several
7 responsibilities that I have. I am greatly
8 appreciative of the efforts of the Medicare
9 Coverage Advisory Committee on coverage
10 decisions.
11 DR. SOX: Thank you.
12 DR. HILL: If I can say Jeff's prepared
13 remarks, thank you. Good morning to you all.
14 And on behalf of him, I would welcome you all and
15 indicate that the office of clinical standards
16 and quality are the folks that this committee and
17 through you the other MCAC panels advise. He's
18 had a chance to meet many of you personally, but
19 he wanted to welcome you and the members of the
20 public that are here to the second meeting of the
21 Executive Committee of the Medicare Coverage
22 Advisory Committee.
23 Jeff wanted me to express our
24 appreciation to all those present for your
25 participation in this process, and on behalf of
.00010
1 HCFA's administrator Nancy-Ann DeParle, we want
2 to especially thank the members of the committee
3 for their service.
4 Involvement in the initial phase of
5 anything can be challenging and perhaps even more
6 so when the government makes a change. This
7 seems to be true even when that change is
8 universally applauded as an improvement in the
9 way HCFA fulfills its responsibilities to our
10 beneficiaries and the American public generally.
11 Since the Medicare program began a
12 little over a third of a century ago, some things
13 have changed, and many have stayed the same. We
14 continue to see our mission as beneficiary
15 focused. While we strive for leadership in
16 improving the health of all Americans, our goal
17 remains assuring access to healthcare for the
18 Medicare-eligible population as we increase our
19 concern for planning in the access of future
20 beneficiaries as well as today.
21 We have moved towards working with
22 providers of all types as customers and partners
23 in delivering care in recognition of the
24 continued central role of the care professional
25 in assuring our beneficiaries' health. My
.00011
1 office, Jeff's office, has important new tools
2 and programs for measuring and improving quality,
3 but our eyes remain firmly fixed on Medicare's
4 original and continued goal, better health.
5 Let me tell you -- myself as well as
6 Jeff would like to tell you -- although there are
7 those that would say otherwise, making good
8 beneficiary-focused coverage decisions is not a
9 new goal for HCFA. Yes, we've shifted from the
10 role of processor and payer to the role of
11 prudent purchaser. And yes, we are more attuned
12 to projections of future Medicare costs than we
13 were at the program's beginnings, but coverage
14 questions have been with us from the beginning.
15 Congress gave us some guidance in the
16 original statute. Told us not to pay for
17 anything that wasn't reasonable and necessary.
18 You are, I think, aware of our renewed efforts to
19 define what we think those terms mean. But
20 clearly, unarguably, science should have a role
21 when we decide whether or not something is
22 reasonable or necessary. We think science should
23 have the most important role.
24 We recognize that the critical
25 examination of the scientific literature is
.00012
1 complex in every case and difficult in many.
2 That's why we need your very expert help, and we
3 are profoundly grateful for it. Thank you.
4 DR. SOX: Thank you. The next agenda
5 item is the subcommittee report. I'm going to
6 deliver the subcommittee report, and if I could
7 ask for the first transparency, we can get
8 started.
9 First let me introduce the members of
10 the subcommittee, Randel Richner, Linda
11 Bergthold, myself, Bob Brook, Alan Garber, and
12 David Eddy was also a participant. Dr. Eddy,
13 because of the extreme press of other businesses,
14 had to resign from the MCAC, but he nonetheless
15 has substantial input into this document.
16 DR. BERGTHOLD: No, he hasn't.
17 DR. SOX: I beg your pardon?
18 DR. HILL: We're still talking.
19 DR. SOX: Oh. We're still talking?
20 DR. HILL: We're hoping to keep him
21 involved one way or another.
22 DR. KANG: He's resigned actually from
23 being a chair of the panel but would like to stay
24 on as a member of the MCAC.
25 DR. SOX: Wonderful. Thank you for
.00013
1 that correction. I appreciate that.
2 So our document has two purposes. The
3 first is to provide general guidance to the
4 panels in the form of suggestions -- general
5 suggestions, not detailed suggestions -- about
6 how to evaluate evidence and focus on two
7 characteristics of the evidence.
8 The first is is it adequate to draw
9 conclusions? And the second is how big is the
10 benefit of the intervention?
11 So in fact, we asked these two
12 questions. Is the evidence concerning
13 effectiveness in the Medicare population adequate
14 to draw conclusions about magnitude of the
15 effectiveness relative to other items or
16 services? And then secondly, if the evidence is
17 adequate, how does the magnitude of effectiveness
18 of the new medical item or service compare with
19 that of other available interventions?
20 Then the second major purpose of our
21 document is to suggest specific procedures that
22 the panels should follow in trying to draw
23 conclusions about the adequacy of the evidence
24 and the magnitude of the effect. And these
25 procedures are drawn from the collected
.00014
1 experience of the members of the subcommittee in
2 doing this sort of work in other venues.
3 So the goal basically of our document
4 is to make the evaluation process more
5 predictable for the proponents of technology so
6 they know what's going to happen and can prepare
7 for it and therefore avoid unnecessary delays in
8 getting an effective intervention through the
9 coverage process, to make sure that our panels
10 are consistent from one panel to the other and
11 from one technology to the other, to make our
12 decisions, or rather, our recommendations, more
13 understandable to the proponents of the general
14 public, and finally, to make sure that the panels
15 are accountable both to each other and the
16 Executive Committee for the quality of work that
17 they do, but also more accountable to HCFA and to
18 the public. So the whole notion is to try to
19 make this process more transparent so that both
20 proponents and the public understand the basis
21 for coverage decisions that HCFA would make based
22 on our assessment of the evidence.
23 So let's turn to the next transparency
24 where we deal with what is probably the most
25 difficult problem, which is deciding whether the
.00015
1 evidence is adequate. Our statement is that the
2 panels must determine whether the scientific
3 evidence is adequate to draw conclusions about
4 the effectiveness of the intervention in routine
5 clinical use in the population of Medicare
6 beneficiaries.
7 And that statement really can be broken
8 down into two substatements. The first is is the
9 evidence valid? Do the conclusions really
10 represent what actually happened? And secondly,
11 is the evidence applicable to Medicare
12 beneficiaries, the population of interest? So
13 let's spend some time talking about each one of
14 those.
15 Now, the first question you have to ask
16 when you're comparing the effects of a new
17 intervention to an old established intervention
18 is are the two populations of patients that
19 you're using to make that comparison truly
20 comparable so that the only difference between
21 them that might affect the outcomes that you're
22 trying to measure is the intervention itself? So
23 when we ask about bias, we ask whether the study
24 systematically overestimates or underestimates
25 the effect of the intervention because of
.00016
1 possible bias or other errors in assigning
2 patients to either the intervention group or the
3 controlled group.
4 An example might help here. Suppose
5 there's a surgical procedure of unknown
6 effectiveness, but pretty risky. It's the sort
7 of thing that you wouldn't do on somebody who was
8 real sick for fear that they would die
9 prematurely as a result of the intervention
10 rather than of the disease for which the
11 intervention is intended.
12 In an observational study in which you
13 try to compare the outcomes of using this
14 intervention with the previous intervention,
15 which is let's say less dangerous, but possibly
16 less effective as well, the problem would ensue,
17 when the surgeon looks at a patient and says this
18 patient is simply too sick to go through this
19 procedure, so I'm going to assign this patient to
20 the controlled group, it's not going to get the
21 procedure. And through a series of such
22 decisions, you end up with the study population
23 that gets the intervention, who's basically
24 pretty well because they're well enough to get
25 through the procedure safely, and the controlled
.00017
1 group, which are all the sick patients, who look
2 like they wouldn't be able to get through the
3 procedure.
4 So a year later when you look at the
5 outcomes, sure enough, the people who got the
6 procedure, many more of them are still alive and
7 functioning well as compared with the controlled
8 group, but because the two groups are very
9 different in their composition, you can't tell
10 whether it was the intervention that led to them
11 being more healthy after the intervention or
12 whether it was the fact they were healthier
13 before the intervention as a result of assignment
14 on the basis of their ability to survive the
15 procedure. So that's an example of biased
16 allocation of patients to intervention and
17 controlled group that could lead to a very
18 misleading interpretation of the outcomes at one
19 year.
20 So how do you avoid bias? Well, the
21 best way to avoid bias is simply to allocate
22 patients randomly to the controlled group or to
23 the intervention group. Random allocation
24 eliminates the type of systematic bias that I
25 described in my example, although it's still
.00018
1 possible that the two groups could be unbalanced
2 because of just the random allocation process,
3 which doesn't necessarily assign people to the
4 two groups in equal numbers if the numbers in the
5 two groups are relatively small.
6 Now, in an observational nonrandomized
7 study such as the one I described in my example,
8 it's often very difficult to decide whether the
9 results were due to bias or due to the
10 intervention. And so we're advising the panels
11 to be very alert to the possibility of systematic
12 allocation bias and observational studies by
13 considering, first of all, the comprehensiveness
14 of the available data, how the patients were
15 selected to receive the intervention and the
16 extent of disease in intervention and controlled
17 groups.
18 And it's possible, using statistical
19 methods, to control for the variables that you
20 know about if you've measured them carefully.
21 The big problem is that you can't control for the
22 variables you don't know about. And that's the
23 beauty of the randomized approach is that the
24 intervention and the controlled group are
25 equivalent, not just for the variables you know
.00019
1 about, but also for the variables you don't know
2 about. It's a very powerful idea,
3 randomization.
4 In some cases the panel may decide that
5 it can't draw firm conclusions about the
6 effectiveness of an intervention without
7 randomized trials. And you can see how that
8 might be the case from the example I described.
9 But in some other cases, perhaps many cases, the
10 panel will determine that observational evidence
11 is sufficient to draw conclusions about
12 effectiveness.
13 When they do that, it's really the
14 panel's obligation to describe potential sources
15 of bias that they perceive and to explain why
16 biased allocation as the result of those factors
17 doesn't account for the results. So in other
18 words, there's a substantial burden of proof on
19 the part of the panel to show that it was really
20 the intervention that made the difference rather
21 than some other difference in the two study
22 populations.
23 Finally, the subcommittee made, I
24 think, a very strong statement saying that a body
25 of evidence that consisted only of uncontrolled
.00020
1 studies, whether based on anecdotal evidence,
2 testimonials or case series or disease registries
3 without adequate historical controls, is never
4 adequate. So we really feel strongly there needs
5 to be some form of control even if it's only
6 historical controls.
7 So let's move on then to the question
8 of external validity basically asking the very
9 simple question, do the results apply to the
10 Medicare population? Do we expect that we will
11 see these results in the Medicare population if
12 they receive the intervention?
13 For a long time randomized studies
14 tended to deal with populations that did not
15 include the elderly. Part of the reason for that
16 is that the older people have other diseases that
17 may cause their death before the disease for
18 which the intervention that you're testing is
19 intended. And so it's much better if you get a
20 population of patients who have only the disease
21 that you're trying to evaluate as the potential
22 cause of death. And so as a result, until
23 relatively recently, elderly patients were not
24 included in randomized trials.
25 For example, there are no women over
.00021
1 the age of 75 in randomized trials of screening
2 for breast cancer despite the fact that the
3 incidence of breast cancer continues to rise
4 through the 70s.
5 Now, increasingly, randomized trials
6 are including elderly men and women. However, if
7 elderly men and women are included in those
8 studies only in proportion to their numbers in
9 the population as opposed to a study that's only
10 including elderly people, there may be too few
11 older people in the study to draw firm
12 statistical conclusions about the effect of the
13 intervention.
14 There's also a concern if the study
15 population is not the same as the general
16 population, the Medicare beneficiaries, then you
17 have to decide that results in a particular
18 subsection of Medicare beneficiaries apply to all
19 Medicare beneficiaries that might eventually
20 receive the intervention.
21 So we call upon the panel to explain
22 its reasoning in deciding that the findings of a
23 series of studies really apply to all Medicare
24 populations. And in fact, the panel might
25 conclude that they don't, and it would be up to
.00022
1 HCFA then to decide on coverage based on that
2 conclusion.
3 Finally, interventions vary from site
4 to site. What works at Johns Hopkins or at Mass
5 General may not work in a community hospital. So
6 the panel has to explain whether the results that
7 are published are going to apply to all
8 healthcare settings and explain why they think
9 that would be the case.
10 So far we've talked about how you
11 evaluate the adequacy of the body of evidence.
12 And the issues, again just to repeat them, are,
13 first of all, biased allocation of patients to
14 the intervention group and the controlled group
15 as something that interferes with the ability to
16 draw a conclusion about whether it's the
17 intervention that really made the difference,
18 and secondly, the general applicability of the
19 results to the Medicare population.
20 So let's now turn to talk about the
21 size of the health effect. And our statement is
22 that evidence from well-designed studies that
23 meet the first criterion -- that is to say
24 adequate evidence -- must establish how the
25 effectiveness of the new intervention compares
.00023
1 with the effect of established services and
2 medical items.
3 And we think that we've helped HCFA
4 with its assignment to make coverage decisions by
5 placing both the size of the effect and the
6 direction of the effect as compared with
7 established services or medical items into one of
8 these seven categories. And by the direction of
9 the effect, I mean is it better or is it the same
10 or is it worse?
11 So one category would be a breakthrough
12 technology. This is something that we all want
13 to see a lot more of, something that causes such
14 a large improvement in healthcare outcomes that
15 it becomes overnight standard of care.
16 The second category would be more
17 effective. The new intervention improves
18 healthcare outcomes by a definite significant,
19 albeit small, margin as compared with established
20 services or medical items.
21 The third category would be as
22 effective, but with advantages. So the
23 intervention has the same effect on healthcare
24 outcomes as established medical services or
25 items, but it has some advantages that would be
.00024
1 important to some if not all patients, such as
2 convenience, rapiditive effect, fewer side
3 effects and so forth. So some people might
4 prefer it over existing interventions.
5 Then there's a category called as
6 effective, but with no advantages, an
7 intervention that basically has the same effects
8 on healthcare outcomes as existing services and
9 doesn't have any substantial advantages.
10 A fifth category is less effective, but
11 with advantages. So it's certainly possible that
12 an intervention could be somewhat less effective
13 than existing alternatives, but it would have
14 some advantages that would be so important to
15 some patients that they might choose it even
16 though it might not have the same effect on their
17 health status as existing interventions.
18 The sixth category is less effective
19 with no advantages. The intervention is less
20 effective than established alternatives, but more
21 effective than doing nothing, and doesn't have
22 any significant advantages.
23 The last category is not effective.
24 The intervention has no effect or has deleterious
25 effects on healthcare outcomes when compared with
.00025
1 doing nothing, such as treatment with placebo or
2 patient management without the use of a
3 diagnostic test in the case of a diagnostic test.
4 So let's then move on from two
5 principles by which the panels can hopefully
6 provide consistent, understandable advice to HCFA
7 about the quality of the evidence and the
8 magnitude of the effect on healthcare outcomes.
9 Now we're going to get into operational
10 procedures, how the subcommittee feels the panel
11 should operate in order to provide consistent
12 results from panel to panel and from intervention
13 to intervention.
14 And the first basic principle is that
15 the panel must explain its conclusions in
16 writing. And this requirement is clearly aimed
17 at trying to improve the transparency of the
18 process and the accountability to the public as
19 well as to the proponents of the technology.
20 We've also put it in the hands of the
21 panel chair to be responsible for writing the
22 explanation of the panel's conclusions.
23 The next procedural recommendation has
24 to do with structuring the evidence so that the
25 panels can function effectively. So we recommend
.00026
1 that the panels should receive well-organized,
2 high-quality background information before they
3 begin their deliberations about the adequacy of
4 the evidence and the size of the effect. And we
5 recommend that the evidence should be summarized
6 in a report, which we call an evidence report,
7 not simply presented as a collection of data or
8 primary studies. And there's ample precedent for
9 this in the technology evaluation efforts of many
10 other organizations.
11 So our basic principle is the integrity
12 of the coverage decision process begins with
13 complete critical evaluation of the literature.
14 And we feel that the standard for HCFA should be
15 the best that's out there in other settings, such
16 as the private sector where Blue Cross Blue
17 Shield has a long track record of doing
18 evaluations of the evidence and making coverage
19 decisions in what is a process that's both
20 efficient and I think highly regarded by
21 professional organizations such as the ACP-ASIM
22 and by other federally sponsored panels. The
23 Agency for Health Research and Quality has a
24 series of evidence-based practice centers in
25 various universities, and I think there are a
.00027
1 couple of private settings around the country,
2 and they provide technical support for the U.S.
3 Preventive Services Task Force on which I serve.
4 Now, evaluating the evidence carefully
5 and providing a balanced, well-organized report
6 of it to the panels is a task that inevitably is
7 going to take some time. It's the opinion of the
8 subcommittee that it should be possible to do
9 these reports in six months or less. Those of
10 you who are experienced in doing this work know
11 that that's fast for doing an adequate evidence
12 report, but we think that HCFA should meet that
13 standard.
14 The next procedural recommendation is
15 basically that members of the panel should be
16 actively involved in the process of reviewing the
17 evidence, and that's based on quite a lot of
18 experience with other health technology
19 programs.
20 So for example, we think that the chair
21 of the panel and perhaps others -- but certainly
22 the chair -- should work with HCFA to establish
23 which are the most important questions that the
24 evidence report should address, and then
25 ultimately the panel must answer as part of its
.00028
1 deliberations.
2 Secondly, we feel that several members
3 of the panel should be active participants in
4 designing the evidence review and preparing the
5 evidence report that the panel will consider.
6 And that's based in part on what we feel is the
7 need to have real expertise on the panel on the
8 topic in question. And the best way to get that
9 expertise is to participate in the design of the
10 evidence review and the writing of the report.
11 Finally, we feel that it's very
12 important that each evidence report be given an
13 extremely careful review. We expect that all
14 members of the panel will read the report very
15 carefully, but we also recommend that one or two
16 members of the group be assigned to be what are
17 called primary reviewers, and we expect those
18 people to really dig into that report, do their
19 best to find any potential problems with the
20 report so that the panel will know that the
21 report has been given sort of the ultimate in
22 very close scrutiny.
23 Finally, we recommend that there be
24 expert review of the evidence report. To ensure
25 that the evidence report is complete and free
.00029
1 from bias, the Executive Committee recommends
2 expert review of the evidence reports. This is
3 going to mean in general subjecting the reports
4 to external review. And the purpose of that is
5 to assure everybody, the public, the proponents
6 and the panel, that the evidence report is
7 complete and that it's fair.
8 That external review should take place
9 before the panels meet, and the evidence report
10 as well as the comments of expert reviewers will
11 be part of the public record of the panel's
12 deliberations. We envision a relatively small
13 number of expert reviewers, perhaps a half dozen,
14 and we will require them to complete their review
15 in a timely fashion, within a month.
16 Now, the last transparency is not part
17 of our report, but it's based on what you could
18 read in the report as a possible time line for a
19 typical MCAC evaluation. So times zero is the
20 time that HCFA decides to go to MCAC for an
21 opinion about the adequacy of the evidence. Then
22 in the first month HCFA and the panel chair would
23 decide on what are the key questions that the
24 panel needs to address and what are the key
25 requirements of the evidence report. In
.00030
1 addition, HCFA would decide who would do the
2 evidence report.
3 Month two to seven would represent the
4 time during which the evidence report would be
5 prepared. And again, it might not be month two
6 to seven. It might be month two to five if the
7 topic was one that led itself to a more speedy
8 conclusion of the review of the evidence.
9 In month eight the report is out for
10 external review. It's out to members of the
11 panel for review. And at the end of that month
12 there's a meeting of the panel that leads to a
13 report to the Executive Committee. And certainly
14 in the ideal world, the timing of the Executive
15 Committee meetings would be closely tied to panel
16 meetings, so the Executive Committee could sign
17 off on the recommendations of the panel within a
18 month after the completion of the panel meeting.
19 And then it will be up to HCFA to decide on its
20 own time schedule about coverage policy.
21 So that concludes the report of the
22 subcommittee. And I think it would be good now
23 for members of the subcommittee to say anything
24 that they wish about my report to be sure that it
25 reflects the views of the members of the
.00031
1 subcommittee.
2 So would anybody on the subcommittee
3 like to comment at this point on my review?
4 MS. RICHNER: I have something.
5 DR. SOX: Randel, please.
6 MS. RICHNER: I actually wrote
7 something last night. I wanted to write them all
8 down so that I didn't forget anything. So excuse
9 me while I load up here to get something. If
10 anybody else has anything to say -- I didn't know
11 that this was my time to talk.
12 DR. SOX: Randel, is it okay if John
13 makes a few remarks?
14 MS. RICHNER: Sure.
15 DR. FERGUSON: Just a few. First of
16 all, I think that this is a very nice road map.
17 It's an idealistic road map in my view. And I
18 guess my overall view is although I think that
19 this is something that we all might like to shoot
20 for, that the end result following this totally
21 might tie the process so that it wouldn't work,
22 and I would not like to see that happen.
23 A couple of specifics. Point one on
24 the adequacy of the evidence.
25 DR. SOX: John, actually, if you don't
.00032
1 mind, I think I'm going to interrupt you. We're
2 going to have an opportunity later on in the
3 morning to present our concerns about the
4 report. I think maybe it would be better to do
5 that later and just have the members of the
6 subcommittee comment on whether I have given the
7 report as they think it is. Is that okay?
8 DR. FERGUSON: Sure. You meant from
9 the members of the subcommittee?
10 DR. SOX: Yes.
11 DR. FERGUSON: Excuse me.
12 DR. SOX: If you wouldn't mind holding
13 it.
14 DR. FERGUSON: That's fine.
15 DR. SOX: Has that given you enough
16 time to get your thing up on the computer?
17 MS. RICHNER: Once again, I'm sorry to
18 have to do it this way, but I decided to write
19 this on the computer last night, so I didn't have
20 any way to print it.
21 DR. KANG: We can print it for you.
22 MS. RICHNER: That's okay. I'll just
23 read it.
24 In my work to date with MCAC, I have
25 attempted to bring views on the impact of our
.00033
1 coverage and process recommendations on the
2 industry, on technology development and
3 innovation, and first and most importantly, of
4 the impact of these recommendations on patient
5 access to new technology.
6 My views are derived from years of
7 practical experience and applied research from
8 being a nephrology transplant nurse, public
9 health research background, including health
10 economics -- now comes research for the
11 pharmaceutical industry -- and most recently, as
12 the vice president of a large manufacturer of
13 minimally invasive technology.
14 I've always considered myself one who
15 comes from a scientific and clinical perspective
16 and passionate about what is important for the
17 patient. Having said this, I am certain that no
18 matter what I say, it will not be to the liking
19 of at least one if not several of the
20 constituencies represented here today.
21 While I was invited to participate in
22 the subcommittee who has drafted this document, I
23 can say that I am not completely satisfied with
24 the final output of this draft. First, I was
25 particularly concerned with the tone, which
.00034
1 implied a lack of flexibility in reviewing and
2 assessing the information that is available for
3 technology assessments. I feel that overall the
4 document assumes that new technology information
5 is innately flawed, or another way of saying it,
6 that all technology is guilty until proven
7 innocent and that it is HCFA's responsibility to
8 protect the public.
9 Second, we do not take into account the
10 availability and rigor of evidence that is
11 available over time for a technology. Depending
12 upon when the technology is referred to MCAC, the
13 life cycle of the technology can have a profound
14 impact on the level and the types of evidence to
15 be reviewed.
16 Third, our primary task was to describe
17 a process for which the panels could make
18 efficient decisions. I felt the draft was never
19 clear on the who, what and when directions for
20 the panels. I also was concerned that we have
21 added on time and many additional reviewers that
22 would make the overall process arduous for any
23 technology to overcome.
24 However, I must strongly support that
25 we, the industry -- and I assume that we're all
.00035
1 the industry in some ways -- have a
2 responsibility to the patient to ensure that the
3 technologies we develop and expect to be covered
4 and paid for will ultimately produce some
5 additional benefit to the Medicare patient. This
6 should be expected and demanded by consumers of
7 healthcare services and products.
8 Finally, I feel that HCFA should have
9 provided MCAC more guidance for the Executive
10 Committee on content and process. I feel that
11 the lack of published guidelines could have
12 provided clearer guidance on criteria for which
13 the technology should be assessed. They've
14 essentially left it de facto to the committee.
15 I'm very committed to the MCAC
16 process. We have an incredible resource of
17 dedicated, highly talented individuals from which
18 we can freely draw and use their expertise for a
19 technology assessment process that is workable,
20 doable, predictable and fair.
21 The committee should have had
22 instruction on the goal of coverage evaluations
23 in a divided, fragmented coverage and payment
24 system that no one can possibly understand who is
25 not intimately involved with the inner workings
.00036
1 of HCFA. I even wonder if those inside HCFA
2 really understand how one system affects
3 another. It's very important.
4 As a quick example, how many times have
5 I heard recently from very educated individuals,
6 why can't we simply get them, HCFA, to increase
7 the DRG payment to cover the new technology?
8 J&J did it with stents. I hear that one all the
9 time.
10 In conclusion, all the dialogue has
11 been particularly useful to move this to the
12 point where I believe we can now successfully
13 design a process and criteria that will work for
14 fair technology assessments. With some open and
15 frank discussions I expect we'll have today, I
16 hope that we can enable a definitive coverage
17 process for promising therapies and
18 technologies. Thank you.
19 DR. SOX: Thank you very much.
20 Would any other member of the
21 subcommittee wish to make any remarks?
22 Well, since there are no further
23 remarks from the subcommittee, it's now time for
24 us to go into open public session. And let me
25 just briefly lay out the ground rules. We have
.00037
1 nine people.
2 DR. BERGTHOLD: I'd just like to say
3 one thing for the record.
4 DR. SOX: Thank you very much.
5 DR. BERGTHOLD: I just wanted to
6 comment on the process of the subcommittee for
7 those of you who didn't have the opportunity to
8 be involved, including people here around the
9 table, and that is that Hal as chair was very
10 open to all kinds of our concerns about nuance,
11 word and tone, and I believe this went through at
12 least a dozen drafts and iterate of drafts trying
13 to be sure that the tone was clear.
14 And so while some may think that this
15 looks negative, I think it is incumbent upon
16 everyone, not only here, but in the audience, to
17 really carefully read this document. Almost
18 every word was discussed and talked about at
19 great length so that the tone would be clearly
20 that while there's a gold standard for evidence,
21 we understood, all of us, that not every new
22 technology will meet that standard.
23 So I just wanted to make that clear,
24 that we had this level of discussion at the
25 subcommittee level, and I wanted to thank Hal for
.00038
1 being very receptive and open to everybody's
2 comments. Thank you.
3 DR. SOX: Thank you very much.
4 Any other comments before we move on?
5 In that case we'll go into open public
6 session. The plan is to have five speakers in
7 the next hour, then take a 20-minute break, and
8 then come back for the last four speakers, then
9 move on to the HCFA presentation at approximately
10 a quarter to 11:00.
11 So five divided into 60 goes 12 minutes
12 per speaker. Excuse me.
13 Could you approach the mic if you have
14 to make a comment.
15 DR. WEISENTHAL: My name is Larry
16 Weisenthal, and I just have a protest concerning
17 the allocation of time to the speakers. I
18 noticed that your five speakers for the first 60
19 minutes have 12 minutes a piece, and that leaves
20 four speakers in 20 minutes for five minutes a
21 piece. So the first speakers get 12 minutes.
22 The second speakers get five minutes.
23 I paid $900 of my own money to fly from
24 California and miss two days of work, and I was
25 told in advance I'd have ten minutes. I can say
.00039
1 it in ten minutes, but I'd really like to have
2 12.
3 DR. SOX: Thank you very much.
4 Everybody's going to have the same amount of
5 time. Let's see. We've got basically an hour
6 and -- I think what we'll basically say is ten
7 minutes per speaker, which I guess is what you
8 were led to expect, and we'll just let the time
9 fall where it may.
10 So I'm going to ask you to stop at ten
11 minutes, and I will be impolite and tell you to
12 sit down if you try to go over, just so you
13 understand that's the way I am. And I'll raise
14 my hand with about a minute to go to give you a
15 chance to wrap up.
16 So let's start with Guido Tricot, who
17 is Director of the Myeloma Transplant Center at
18 the University of Arkansas. Welcome.
19 DR. TRICOT: Thank you very much for
20 giving me the time to bring up a few issues. My
21 name is Guido Tricot. I'm the director of the
22 myeloma program at the University of Arkansas.
23 The first issue I would like to bring
24 up is the age issue. Although we assume that
25 Medicare is mainly for patients over the age of
.00040
1 65, when we reviewed the records of patients who
2 had transplants for myeloma, approximately
3 one-third of the patients were under the age of
4 65. That's one issue.
5 The second issue about age is that most
6 of the reasons why age has become a problem --
7 MS. LAPPALAINEN: Could you bring the
8 mic closer to you? It's wireless, so you can
9 pick it up, if you'd like.
10 DR. TRICOT: -- why age has become a
11 problem is because of the comorbid conditions
12 that the patients may have. And in most studies
13 there are sufficient exclusion criteria to deal
14 with the comorbid conditions. And rather than
15 making age an issue, because we all know that
16 there is basically no difference between a
17 patient who is 64 years and 11 months and
18 somebody who is 65 years, and that there's a
19 difference between calendar age and biologic age,
20 I think exclusion criteria rather than age itself
21 should be the main thing to exclude comorbid
22 conditions.
23 A second point that I would like to
24 bring up is that in the explanation of panel's
25 conclusion, the panel chair is responsible for
.00041
1 writing the explanation of the panel's
2 conclusion. We need to make sure that there are
3 mechanisms in place that the report is a
4 reflection of the whole group of the panel and
5 not necessarily mainly a reflection of what the
6 chair's vision is.
7 A third point is the external review by
8 experts. Although it states that this will
9 become part of the public record, we need to make
10 sure that this becomes part of the public record
11 prior to the panel meeting and that there's
12 adequate time to review and comment at the time
13 that the proponents will make the report.
14 A smaller comment is on the randomized
15 studies. Although we all would like to have many
16 randomized studies all showing the same results
17 and going in the same directions, we also need to
18 be aware of the fact that once there is one
19 randomized study that shows that one treatment is
20 better than the other, it becomes difficult to do
21 further randomized studies. In principle you're
22 only supposed to do randomized studies if as a
23 physician you're not convinced that one treatment
24 is better than the other and that you have no
25 bias toward any of the treatment modalities.
.00042
1 There's also a problem with referral
2 patterns. We at the University of Arkansas have
3 tried to do randomized studies, but the patients
4 that are coming to our institution come from
5 everywhere, and they come because they want a
6 certain procedure done, and we have never been
7 able to do randomized studies because of that.
8 And the last point I would like to
9 bring up is that there is a tremendous time lapse
10 between initiation of the process and the point
11 in time the proponents are convinced that what is
12 proposed is better than what has been available
13 before and the ultimate approval. And it's going
14 to be at least nine months, and probably more
15 likely, 12 months or more. And I think there
16 should be a mechanism in place that provides
17 temporary approvals in between this 12-month
18 lapse and that a committee of experts can be
19 gathered to give temporary approvals until the
20 final decision by HCFA is made.
21 I think those are my main concerns.
22 Thank you very much for giving me this time.
23 DR. SOX: I should remind the members
24 of the Executive Committee that we're going to
25 have about an hour to ask questions of the people
.00043
1 who are going to speak. So take notes and be
2 ready to ask some questions during the hour that
3 will be reserved for discussion with them.
4 With that, we'll move on to Richard
5 Justman, who is medical director of United
6 Healthcare and the American Association of Health
7 Plans.
8 DR. JUSTMAN: Thank you. Good
9 morning. My name is Dick Justman, and I do not
10 have any financial connection to technology or
11 device manufacturers. In my current position
12 that would be very difficult.
13 My name is Dick Justman, and I'm the
14 national medical director of United Health Group.
15 DR. HILL: Excuse me, Dr. Justman.
16 Would you do the same thing with your
17 microphone? Folks in the back are indicating
18 they can't hear.
19 DR. JUSTMAN: Is that better?
20 DR. HILL: Thank you.
21 DR. JUSTMAN: I'm the national medical
22 director of United Health Group, and I'm here
23 today speaking on behalf of the American
24 Association of Health Plans. AAHP represents
25 more than a thousand health maintenance
.00044
1 organizations, preferred provider organizations
2 and other similar network-based health delivery
3 systems that provide healthcare to more than 150
4 million Americans. AAHP member health plans are
5 dedicated to the philosophy that we put patients
6 first by offering them benefit packages offering
7 coordinated comprehensive healthcare.
8 United Health Group, the company for
9 which I work, has 40 health plans around the
10 United States serving approximately 14 million
11 commercial enrollees in HMO, PPO point of service
12 and exclusive provider organization products. We
13 also have approximately 400,000 Medicare
14 enrollees.
15 As you may have read recently in the
16 newspapers, United Health Group has recently
17 embarked upon a program which we call care
18 coordination, and this is a model of healthcare
19 coverage which essentially allows physicians and
20 patients to make healthcare decisions with
21 minimal intrusion by the health plan subject only
22 to the limitations of benefit design. However,
23 we feel very strongly that for this endeavor to
24 work, we need to be covering procedures to
25 biases, treatments and drugs that we know
.00045
1 actually do work.
2 We strongly endorse a rigorous,
3 evidence-based approach to coverage
4 determinations. We applaud the establishment of
5 the Medicare Coverage Advisory Committee to
6 assist HCFA to evaluate the clinical evidence
7 about the relative effectiveness of new medical
8 devices, services and other technologies.
9 The report of the Executive Committee
10 working group to be discussed today will promote
11 systematic and consistent evaluation of the
12 clinical evidence by the panels that we believe
13 should meet the needs of all the stakeholders.
14 There is compelling evidence, including
15 evidence cited by President Clinton's own
16 advisory commission on consumer protection of
17 quality in the healthcare industry, that
18 Americans do not always receive the best possible
19 healthcare. In many instances they do not
20 receive important healthcare services that they
21 should, and yet in other instances they receive
22 services of uncertain value, and unfortunately in
23 yet other instances they receive services of
24 questionable quality.
25 Also, too often medical treatments are
.00046
1 widely disseminated before they have been proven
2 to be effective putting patients potentially at
3 risk of harm, and this also discourages for
4 further research.
5 Both of these problems, the variation
6 and the use and quality of healthcare services
7 and the proliferation of unproven treatments,
8 illuminate the importance of promoting a delivery
9 care that is based upon robust, scientific
10 evidence.
11 To give you an example, a recent study
12 showed that between 1987 and 1991, only 21
13 percent of eligible elderly patients were treated
14 with beta blockers for ischemic heart disease,
15 myocardial infarction and related disorders and
16 that the subsequent mortality rate for those who
17 did receive the treatment was 43 percent lower
18 than for those who did not receive the
19 treatment. This translates into, in that study
20 group, 18,000 potentially avoidable deaths that
21 would not happen because the appropriate
22 treatment was not given.
23 What is really stunning in this case is
24 that in the words of the American Medical
25 Association, beta blockers are one of the most
.00047
1 scientifically studied and substantiated medical
2 therapies. There is a plethora of published
3 evidence about them. The American College of
4 Cardiology and the American Heart Association
5 have brought guidelines and physician statements
6 promoting their use. And despite this and
7 despite voluminous evidence, there are many
8 eligible people who potentially would have
9 benefited from beta blockers who have not
10 received them.
11 A second problem undermining the
12 quality of care is the proliferation of
13 treatments that have been widely disseminated in
14 the absence of proof that they are effective. In
15 such cases patients may be harmed because they
16 forego a standard proven therapy in favor of a
17 treatment that may be less effective than the
18 standard one.
19 A most recent example is that of high-
20 dose chemotherapy and bone marrow transplantation
21 for women with breast cancer. An assumption was
22 made many years ago that if women are partially
23 responsive to standard dose chemotherapy, that
24 high-dose chemotherapy coupled with bone marrow
25 or peripheral stem cell rescue would be even more
.00048
1 effective. Unfortunately at the time this
2 assumption was made, there was little evidence to
3 support this, little robust scientific evidence.
4 And in fact, this became widely disseminated as a
5 treatment that women must have. Well-intentioned
6 advocacy groups promoted its use. Many states
7 actually passed laws mandating coverage for
8 this. And this essentially became a
9 self-fulfilling prophecy.
10 Women assumed that if states were
11 mandating coverage for this, this must be a
12 preferred and effective treatment. This
13 essentially made it very difficult for women to
14 randomize themselves into controlled trials
15 because women were afraid that if they were
16 randomized into the standard treatment group,
17 they would miss out on treatment that might be
18 effective. So in fact, there was circular
19 reasoning here.
20 And as you know, there has been recent
21 published evidence that says that if anything,
22 high-dose chemotherapy bone marrow
23 transplantation is no more effective than
24 standard chemotherapy for women with breast
25 cancer although the morbidity of high-dose
.00049
1 chemotherapy is substantially greater. So this
2 is a very stunning example of a situation in
3 which a therapy is rapidly proliferated in the
4 absence of scientific evidence, and it is very
5 difficult now to reverse that trend.
6 Another example of a less life-
7 threatening but equally pervasive disorder has to
8 do with low-back pain. Approximately a year ago
9 in a national news weekly, a device was
10 discussed, which presumably through a heat
11 treatment, reduces significantly diskogenic
12 low-back pain. This was widely reported, and
13 many providers in many regions of the country
14 began to promote this treatment.
15 At the time that this was done, there
16 was almost no scientific evidence published at
17 all. All the scientific evidence that was
18 available was available on a website.
19 To make matters worse, there were yet
20 other providers who began to use this device to
21 treat neuropathic pain, for which the FDA
22 indications never existed in the first place. So
23 this is yet another example where in the absence
24 of scientific evidence, there can be rapid
25 proliferation of technology that desperate people
.00050
1 will try to use.
2 Health plans have taken a prominent
3 role in promoting evidence-based care.
4 Increasingly, health plans are working with
5 physicians to reduce the variation in practice
6 patterns through the dissemination of chemical
7 profiling tools and processes of care that guide
8 physicians to provide their patients the right
9 care at the right time and in the right setting.
10 Health plans distribute and encourage
11 the use of evidence-based processes of care by
12 physicians and other healthcare providers.
13 Health plans also provide feedback to physicians
14 about how their treatment practice patterns,
15 including underutilization and overutilization,
16 compared to scientific evidence and also to the
17 practice patterns of their peers. Health plans
18 make scientific coverage determinations based
19 upon the best available evidence. Through these
20 and other activities, health plans actively
21 promote the use of evidence-based care.
22 Through technology assessment, health
23 plans are working to approve coverage of new
24 treatments supported by medical evidence and to
25 avoid the coverage of treatments for which there
.00051
1 is no scientific evidence and for which these
2 treatments may actually harm patients. In
3 technology assessment organizations gather and
4 evaluate the scientifically valid evidence
5 available, including, but not limited to,
6 surgical procedures, devices and drugs.
7 First, they determine whether the
8 evidence demonstrates that the treatment is
9 safe. Second, they evaluate whether or not the
10 evidence demonstrates that the treatment is as
11 effective or more effective than an existing
12 treatment if an existing treatment does exist.
13 Health plans use this information in
14 determining whether or not the treatment should
15 be a covered service. By implementing a
16 structured method for evaluating new or existing
17 treatments and not covering treatments not proven
18 to be effective, health plans are working to
19 reduce the proliferation of unproven and
20 potentially unsafe treatments.
21 However, health plans cannot solve this
22 problem alone. We need the help of others within
23 the system, including Medicare, Medicaid
24 providers, researchers and manufacturers.
25 Increasingly, the healthcare community and policy
.00052
1 makers recognize the importance of promoting
2 evidence-based care and are working to change the
3 current environment.
4 In addition to health plans, others in
5 the healthcare community understand the
6 importance of promoting and providing evidence-
7 based care, and in order to be valid, the
8 evidence itself must meet certain criteria.
9 We support very definitely the use of
10 the best possible scientific evidence, and we are
11 aware that randomized controlled trials ideally
12 are the best evidence. We recognize also,
13 however, that those are not always possible,
14 either due to the lack of availability of a
15 control arm, the size of the cohort or other
16 factors. However, we believe very strongly that
17 we must always seek the best scientific evidence
18 that is available and the best methodology
19 available in order to make coverage decisions.
20 In conclusion, I would like to stress
21 that the first goal of the healthcare system
22 should be to provide quality healthcare
23 services. In our current system too often
24 quality is compromised because the care delivered
25 is not consistent with the best available medical
.00053
1 evidence.
2 Health plans are committed to improving
3 quality care through reliance on medical evidence
4 when making coverage determinations, when
5 evaluating new therapies and in communicating
6 with providers. In order to improve the quality
7 for all patients, however, all stakeholders in
8 the healthcare system, not just the health plans,
9 must be actively committed to the process of
10 using evidence-based medicine. Thank you.
11 DR. SOX: Thank you very much. Just so
12 that the speaker knows when there's one minute to
13 go, I'm going to stand up, which hopefully will
14 catch your eye. Putting up my hand didn't seem
15 to work very well.
16 Our next speaker is Morgan Downey,
17 Executive Director of the American Obesity
18 Association.
19 MR. DOWNEY: Thank you, Mr. Chairman
20 and members. It's a pleasure to be here with you
21 this morning.
22 My name is Morgan Downey, and I am the
23 Executive Director of the American Obesity
24 Association. This association is about four
25 years old, and it was founded as an adequacy
.00054
1 organization to promote research, treatment,
2 prevention and intervention in the epidemic the
3 country is going through, obesity.
4 I'm very pleased to be able to address
5 the complex issues of obesity in the Medicare
6 program with you this morning. For the record,
7 the American Obesity Association is supported by
8 several major companies, including Amgen Hoffman-
9 LaRoche and all pharmaceuticals, Weight Watchers
10 International, in dues from professional and lay
11 members. To the best of my knowledge, no
12 supporter has a specific coverage issue before
13 the Medicare Coverage Advisory Committee at this
14 time.
15 At the outset I'd like to put our
16 current and immediately foreseeable situation on
17 the record. Over half of the United States
18 population is overweight, and about a quarter is
19 obese measured as their body mass index of over
20 25 and over 30 respectively. According to 1991
21 data, the percentages of the Medicare population,
22 with the BMI of over 27.8 percent for males and
23 27.3 for females, ranged from 23.8 percent for
24 white males to 48.7 percent for black females.
25 As you well know, obesity is a major
.00055
1 independent risk factor for conditions such as
2 Type II diabetes, hypertension, heart disease,
3 stroke, several cancers, arthritis, end stage
4 renal disease, gallbladder disease and sleep
5 apnea, to name a few of the 30 or so conditions
6 where associations have been found.
7 We know that obesity is increasing
8 rapidly in the population. Jeffrey Copeland,
9 Director of the Centers for Disease Control and
10 Prevention, has likened its spread to that same
11 in infectious diseases. According to a recent
12 article in JAMA in October, between 1991 and
13 1998, the prevalence of obesity measured as a BMI
14 over 30 among persons age 60 to 69 increased 44.9
15 percent. The prevalence among persons over 70
16 increased 28.6 percent. That is a rate of 6.4
17 percent per year at a BMI level of 30 and four
18 percent a year increase for a person over 70.
19 We also know that obesity is a major
20 generator of healthcare costs. According to a
21 study of the American Obesity Association
22 commission from the Lewin group last year, the
23 direct healthcare cost of obesity exceeded a
24 hundred billion dollars in 1999. This figure
25 does not include indirect costs or costs spent on
.00056
1 treating obesity itself. We did not ask for a
2 breakdown by payers, but I think it's fair to
3 assume that the Medicare program plays a
4 significant if not majority component of those
5 costs.
6 So it's not without substantial
7 justification that obesity is now listed as one
8 of the nation's ten leading health indicators, as
9 announced a few weeks ago by the surgeon
10 general.
11 We concede, therefore, that more and
12 more Americans are becoming obese, which will
13 dramatically increase their risk for diseases,
14 which Medicare will pay for. These people will
15 come into the Medicare program, both as they age,
16 and also as they become eligible for disability
17 under Social Security disability procedures.
18 The standards for the evaluation of
19 obesity under Social Security is currently
20 undergoing some changes, but we expect that the
21 current number of 137,000 persons who receive
22 Social Security disability under their obesity
23 listing will continue to increase. And as you
24 know, after two years on disability, these
25 individuals start receiving healthcare coverage
.00057
1 under the Medicare program.
2 Our interests today are twofold.
3 First, we propose that the committee consider
4 when evaluating new medical profits, be they
5 laboratory tests, diagnostic procedures,
6 preventative intervention or treatment, that a
7 large portion, a quarter to a half of the
8 Medicare population, is overweight or obese.
9 Questions might be asked were the
10 studies in support of the procedures conducted in
11 a representative sample of the current population
12 by weight? Can Medicare beneficiaries who are
13 obese access the new technologies?
14 As an example, there are recent studies
15 showing, for example, that obese women receive
16 pap smears and mammograms with less frequency
17 than do nonobese women.
18 Last fall the representative of HCFA,
19 speaking at a conference we had on public policy
20 implications of obesity, indicated that the bone
21 marrow transplantation protocols in this country
22 exclude persons with obesity without medical
23 justification.
24 Second, we propose that the committee
25 begin the process of clarifying Medicare coverage
.00058
1 of obesity. Paragraph 3526 of the coverage
2 manual states, quote, obesity itself cannot be
3 considered an illness. The immediate cause is a
4 caloric intake, which is consistent with a higher
5 than caloric output. Program commitment may not
6 be made for the treatment of obesity alone since
7 this treatment is not reasonable and necessary
8 for the diagnosis and treatment of an illness or
9 injury. Yet under paragraph 3540, obesity
10 surgery, bariatric surgery is covered if
11 medically appropriate and necessary to correct an
12 illness caused or aggravated by obesity.
13 Clearly these two paragraphs are
14 inconsistent. If obesity cannot be considered an
15 illness, the surgery to correct it can't be
16 covered. On the other hand, as a reduction of
17 weight can correct an illness or injury
18 aggravated by obesity, what possible
19 justification is there for covering exclusively
20 the most drastic and life-threatening
21 intervention when other equally effective and
22 less risky treatments are available? Clearly
23 3526 of the coverage manual is wrong and should
24 be considered an embarrassment to the Health Care
25 Financing Administration.
.00059
1 Illness is synonymous with disease.
2 Virtually every medical and scientific definition
3 define diseases as, for example, does Stedman's
4 medical dictionary, which is, one, an
5 interruption, cessation or disorder of body
6 functions, systems or organs, or two, a disease
7 entity characterized by at least two of these
8 criteria; one, recognized etiologic agent or
9 agents, two, an identifiable group of signs and
10 symptoms, three, consistent anatomical
11 alterations. Clearly obesity means all three of
12 these criteria.
13 Any analysis of the definitions of
14 illness and injury disorder will demonstrate that
15 obesity is considered an illness by the vast
16 weight of modern, scientific and medical
17 understanding. Therefore, we'd like to suggest
18 two issues for your consideration.
19 First, given the increase in the
20 overall Medicare population which is obese and
21 the increases in medical technology, we want to
22 be sure that all such advances are available to
23 the obese Medicare population. Therefore, AOA
24 suggests that all future subjects for Medicare
25 coverage determinations be evaluated with this
.00060
1 population in mind.
2 Second, we suggest the committee
3 establish a subcommittee or working group to
4 revise the current and incorrect coverage manual
5 paragraph 3526. There are many professional
6 guidelines for the treatment of obesity in adults
7 including that developed two years ago by the
8 National Institutes of Health, which relies on
9 literally hundreds of randomized controlled
10 clinical trials and other studies which would
11 meet the criteria earlier elucidated by the
12 chairman regarding the considerations of this
13 committee.
14 The American Obesity Association would
15 be pleased to provide whatever assistance or
16 support would be helpful to the committee in
17 these undertakings. Thank you.
18 DR. SOX: Thank you very much. Our
19 next speaker is Donald Baim.
20 DR. KANG: Hal?
21 DR. SOX: Jeff?
22 DR. KANG: Mr. Downey, on your second
23 issue, procedurally -- I think you got our April
24 notice last year -- you really need to submit a
25 coverage decision internally. MCAC gets only a
.00061
1 very small subset referred to by HCFA. This is
2 actually the first time I'm aware of that
3 coverage manual issue, and we'd be happy to look
4 at it, but maybe we can talk about that off line
5 how to get that done.
6 MR. DOWNEY: Okay.
7 DR. SOX: Thank you very much.
8 Our next speaker is Dr. Donald Baim,
9 Chief of the Interventional Cardiology Section at
10 the Beth Israel Deaconess Hospital, and he's
11 speaking today on behalf of the Health Industry
12 Manufacturers Association.
13 DR. BAIM: Thanks. It's my pleasure to
14 be down here. HIMA asked me to speak about some
15 of the real world applicability of technology
16 innovation and adoption in the interventional
17 cardiology area and specifically as it pertains
18 to the coverage decisions by this group.
19 Can I see the first overhead, please.
20 I think we all share common goals in terms of
21 encouraging industry to develop newer devices and
22 device improvements and facilitate the rapid
23 adoption of safe and effective new diagnostic and
24 therapeutic technologies in healthcare to improve
25 the well-being of our population. We more than
.00062
1 anyone endorse the use of robust-data-driven
2 approaches and avoiding technologies that are
3 less effective. And I'll talk a little bit about
4 where the FDA process has gone in interventional
5 cardiology.
6 But in reading the report of the
7 committee, I'm concerned that we preserve the
8 nimbleness and responsiveness of a system of
9 coverage decisions both to allow rapid adoption
10 of technology and avoid placing already strapped
11 hospitals in further financial jeopardy by
12 forcing them to buy effective new technologies
13 without offsetting reimbursement. And we'll talk
14 about an example of that next.
15 So I want to make three basic points in
16 this ten-minute slot. The first is that we
17 really need a variety of evidentiary sources,
18 randomized clinical trials being one of them, but
19 also including registries, equivalence trials and
20 OPCs to deal with different situations.
21 The second is to point out that the
22 trials that are currently being done for FDA
23 approval are large and very methodical and should
24 be the first points considered as new
25 technologies emerge from the FDA process and are
.00063
1 considered for coverage. I'll talk a little bit
2 about the fact that I do believe they're
3 sufficiently generalizable to apply to the care
4 of Medicare population by mainstream operators.
5 And third, that delayed HCFA coverage
6 approval restricts application of new and better
7 therapies and adds financial burdens to hospitals
8 with an expense reimbursement gap as well as
9 industry.
10 So I really want to cover that first
11 point, the variety, the spectrum of evidentiary
12 sources. At different points in the development
13 of new technology, pilot registries may be
14 valuable for proof of concept and device
15 refinement, although not for the coverage
16 decisions you're talking about here, but broader
17 registries that may contain thousands of patients
18 may be adequate for approval of certain well-
19 characterized devices.
20 Third, randomized equivalency trials
21 are now being used by FDA to approve new
22 generation stents that we'll talk about in a
23 second and demonstrate noninferiority relative to
24 other established therapies. The randomized
25 superiority trials that the guidance document
.00064
1 focuses on to establish superior outcomes or
2 cost-effectiveness of high-volume, high-cost or
3 high-risk procedures once they're mature versus
4 the prior standard of care are not the only sort
5 of valid evidence that needs to be considered in
6 the coverage decision.
7 And finally, the importance of post FDA
8 approval collection of population-based outcome
9 data to document the use, patterns and risk-
10 adjusted outcomes of competitive procedures for
11 certain conditions in the real world should not
12 be underestimated.
13 I just wanted to talk briefly about how
14 this whole interventional cardiology got here,
15 and it was through registries. The NHLBI PTCA
16 Registry 1, in 1977 to 1981, lead to the adoption
17 of this therapy, and the Registry 2, in 1985 and
18 '86, documented the improvement in devices and
19 technique. Katherine Detre from the University
20 of Pittsburgh and I, with NHLBI funding, set up a
21 third registry in 1989 that ended up enrolling
22 some 4500 patients with seven new interventional
23 devices and really still constitutes the largest
24 series of patients with core angiographic
25 laboratory evaluation of one-year follow-up for
.00065
1 many of these devices.
2 That type of registry approach,
3 however, was not sufficient to lead to the
4 approval of stents. So in 1993 the first stent
5 versus angioplasty randomized trials were
6 performed within the NACI registry that use
7 single indications, a full randomized clinical
8 trial machinery and lead to the approval of the
9 J&J stent in a rigorous FDA process in 1994,
10 making the United States the last of the
11 industrialized countries to receive approval for
12 this device. So it's a very slow process,
13 randomized trials. Particularly as new
14 technology becomes accepted, there's emerging
15 reluctance to randomize stentable patients to
16 conventional angioplasty, and that leads to a
17 very prolonged approval for the second stent to
18 try to go through this randomized comparison to
19 angioplasty.
20 So how have the variety of stents that
21 are now in interventional practice gotten through
22 this FDA process? It's really been by a change
23 in paradigm. And the change in paradigm that
24 took place in 1996 was really to say we don't
25 need to randomize stents versus angioplasty any
.00066
1 longer, that documenting equivalency to approved
2 stent designs would be also an acceptable
3 approach. And the last half a dozen stents to be
4 approved have been done in that format, usually a
5 thousand patients randomized to a new versus an
6 old stent. Recruitment is faster because
7 everyone gets a stent, and it's a good solution
8 to follow-on improvements and accepted
9 technology. It has the rigor of an RCT, but
10 without a placebo group. It can also monitor for
11 improvements in stent designs, but it's a
12 paradigm that's showing signs of age because
13 showing equivalency to a first generation stent
14 is probably not good enough, and it wastes the
15 money of reconfirming the performance of the
16 first generation stent in each successive trial.
17 So where we're headed in this new
18 device era in 2000 and beyond is to develop OPCs,
19 objective performance criteria, that will collect
20 registry data and document performance consistent
21 with the OPCs for stent performance. The reason
22 I go through this series of evaluation paradigms
23 is really we're right back now with registries,
24 and each of these different formats for evidence
25 collection has been appropriate for a different
.00067
1 point in the development of the technology. We
2 can't just fixate on randomized clinical trials.
3 I just wanted to show you what this new
4 device era has meant in our own practice, and
5 this one shows in stacked bars the different
6 therapies used in our program over the five years
7 from 1994, when the J&J stent was approved,
8 through 1998. Angioplasty is the bottom bar
9 shown in red, conventional balloon angioplasty,
10 which has now fallen to 21 percent in
11 interventions. Stenting over that period has
12 risen, the yellow bar, from 29 to 68 and now 79
13 percent last year in 1999 with two atherectomy
14 technologies accounting for the final quarter.
15 So this adoption of technologies has
16 really revolutionized our field. The J&J stent,
17 as we said, was approved in 1994. And Medicare
18 decision about coverage and assignment to DRG
19 116, however, did not take place until 1997. And
20 in those three years between FDA approval and
21 Medicare reimbursement coverage, the hospitals
22 were having to buy this effective technology from
23 manufacturers without any incremental
24 reimbursement, and it contributed in no small way
25 to the financial deneument of many of the leading
.00068
1 institutions.
2 Now, one could say this rapid adoption
3 of technology is just to appease technology-
4 crazed operators, but this shows the
5 corresponding incidence of major complications
6 over that same time period. And the adoption of
7 these technologies has in fact cut major
8 complications in half, so we need to keep
9 facilitating this rapid adoption process.
10 I just want to close by taking you
11 through one of the trials, a Boat trial and
12 atherectomy trial, to give you a flavor for the
13 generalizability of the Medicare population.
14 This trial enrolled a thousand patients over a
15 one-year time frame, actually 16 months, to
16 angioplasty versus atherectomy. This was done at
17 36 centers, and this shows that they are
18 geographically distributed, and they're both
19 active practice centers.
20 One concern is the age of patients, and
21 what I've shown on this is the cumulative
22 distribution in yellow of our own interventional
23 patients whose median age is 64 compared to the
24 age in pink, I guess, of 12 trials with 8,000
25 patients that have been run by our daily
.00069
1 coordinating center showing the median age of 63.
2 So the age distribution in the interventional
3 trials is representative of about half the
4 Medicare population of routine practice.
5 The issue about few golden operators
6 driving the results of these trials, I think, is
7 addressed here showing the center-by-center
8 performance in this trial. There's a wide
9 variety of operators and operator experience, and
10 as you can see in the DCA results shown in the
11 yellow bars, in terms of residual stenosis
12 there's a wide variety of practice patterns.
13 Thank you.
14 DR. SOX: Thank you very much. Our
15 next speaker is Wayne Roe, who is Chairman of
16 Covance Health Economics & Outcome Services in
17 Washington, D.C., and he's speaking on behalf of
18 the Health Industry Manufactures Association.
19 MR. ROE: Good morning. I'm glad to be
20 here. I'm actually speaking on behalf of
21 myself. I'm speaking at the behest of HIMA. I
22 have lots of reasons to have conquest in this
23 business, and I do a little bit of consulting in
24 the coverage policy area, very little bit from
25 the old days. I'm on the boards of six medical
.00070
1 start-up copies in the California area, involved
2 with three venture capital firms who fund life
3 sciences companies, all of whom will have things
4 that will come before HCFA someday, but maybe not
5 for three or four years.
6 I think HIMA asked me to be here
7 because I spent the last 15 years getting gray
8 hair by coming to HCFA and working on coverage
9 policies for probably over a hundred different
10 devices, drugs, diagnostic tests and surgical
11 procedures. I've learned a lot about the
12 process, got a lot of headaches through the
13 process, have a lot of respect for the people
14 doing coverage, and I think this group has its
15 work cut out for it. This is incredibly
16 complicated stuff, as you hear today. It's not
17 simple, it's not trivial, and it can be academic
18 and inherently judgmental no matter what you do.
19 I'll start out with just a few
20 comments. HIMA doesn't know what I'm going to
21 say because I wrote this last night when I was
22 helping my daughter do chemistry, having read
23 your paper several times. I want to commend the
24 MCAC. I think you've done some very thoughtful
25 work. I think in 11 or 12 or 13 pages there's
.00071
1 lots of good stuff in there. I'm not going to
2 try to wordsmith it at all. I congratulate you
3 on seven categories on the size of health
4 effects. I think those are pretty novel, pretty
5 creative. I think they really importantly
6 reflect the fact that most new technologies in
7 medicine, like it or not, are incremental. They
8 have a whole wide range of possible effects,
9 positive and negative.
10 Unfortunately, we believe there are too
11 few breakthrough technologies. It seems to be
12 the way things work. I wish we had more of
13 them. I think we want to encourage people to
14 have more of them. But I think having those
15 categories three or four that clearly ought to
16 lead to positive Medicare coverage decisions is
17 kind of a good way to kind of simplify the
18 world.
19 I spent the last ten years telling
20 medical developers I think they should stop
21 thinking about thinking about themselves -- and a
22 lot of this comes out of reading the work of Dr.
23 Brook and Hal Sox and David Eddy and so forth --
24 stop thinking about themselves as making tools or
25 making drugs, but think about themselves as
.00072
1 changing outcomes or changing the practice of
2 care. And if they don't do the right kind of
3 research or science to demonstrate a change in
4 how their product has an impact on how the
5 patient does or at least how the patient is
6 managed, then they shouldn't be bringing their
7 technologies to HCFA or Blue Cross Association or
8 anyone else.
9 I think by and large that kind of
10 admonition, which lots of people have been saying
11 is getting through in the overall level of
12 science, in the life sciences world, is a hell of
13 a lot better today than it was 10 or 12 years
14 ago. There's no question about it. No one even
15 thought about any kind of randomized study, even
16 controlled study, 12, 14, 15 years ago when I
17 entered the device industry and we had the old
18 National Center for Healthcare and Technology,
19 which said many of the same things we've said
20 that you are trying to say to today.
21 And I encourage you to appreciate
22 really that the document you're writing here is
23 going to be a sentinel of technology
24 gatekeeping. We don't like to think this
25 sometimes, but the bottom line is it's going to
.00073
1 get read by lots of people, the final document,
2 and it's going to be used by lots of people to
3 make decisions. It's a gatekeeping signpost.
4 Obviously HCFA is a gatekeeper, but you all are
5 the experts.
6 We have a luminary panel here, the best
7 and brightest we have in terms of doing outcomes
8 research, and I think it's appropriate and
9 important for you to encourage better science, to
10 challenge the innovators to do better scientific
11 work. And I think the tone of this should be to
12 do that. On the other hand, I think it would be
13 very bad to discourage them, to tell them well,
14 we want everybody to high jump eight feet, and
15 less than eight feet was never going to be
16 adequate, but you know, we really know behind the
17 scenes six, five or six, six is going to be
18 okay. I think that's a discouraging kind of
19 tone, and I encourage you to take a look at the
20 tone again.
21 HCFA staff and the care and medical
22 directors, as we're here today, to private
23 managed care medical directors, will read what
24 you say, and they'll use it. You don't want to
25 give them the excuse to hide behind it, to not
.00074
1 make decisions, to put everything on randomized
2 controlled trials, because the bottom line is
3 we're not going to have them all. We're never
4 going to have them all. And it would be kind of
5 an academic pipe dream to expect we're going to
6 have it. I don't think you should set the bar so
7 high for people to use that as an excuse not to
8 make tough decisions, not to allow progress in
9 medicine. So please be realistic. You can't be
10 academic in this exercise even though you want to
11 be.
12 I guarantee you I've been through
13 this. Somewhere in Menlo Park, California there
14 is someone sitting down making a decision to fund
15 $20 million for an Internet taco business versus
16 some promising technology that will gather up
17 plaque during cardiac endarterectomies that might
18 save one of our lives someday. You don't want to
19 discourage those people who might get the money
20 to do the atherectomy device or filtration
21 technology with the idea that you have to have
22 two huge randomized controlled trials in order to
23 get coverage. That is just a bad thing to send.
24 But those decisions happen all the time with
25 increasing frequency. You've got your capital
.00075
1 world and the pharmaceutical firms and so forth
2 who are going to read this document and look at
3 it, and they're going to look to you for some
4 guidance. Give them hope, give them a challenge,
5 but don't let them feel like it's hopeless
6 because they'll go and fund those Internet taco
7 businesses, and I don't think we need that as
8 much as we need things to deal with
9 endarterectomy.
10 Specific suggestions. First, I find it
11 quite amazing -- a little hyperbole in all of
12 this, of course -- that there's no mention
13 whatsoever -- maybe one mention -- of the FDA
14 standard of evidence or labeling in this
15 document. Everything goes through the FDA to
16 start. I know we all in the coverage policy
17 arena realize maybe it's not enough sometimes,
18 but every new technology is studied with the FDA
19 in mind. And the FDA has very good outcomes
20 researchers there, and they require sometimes
21 randomized trials, sometimes not randomized
22 trials, sometimes controlled trials, sometimes
23 not, depending upon the product. It seems to me
24 there ought to be some recognition that the FDA
25 is enough for certain things, particularly
.00076
1 pharmaceuticals.
2 The concept that people do
3 well-controlled randomized trials, two of them in
4 pharmaceuticals, for the purposes of
5 demonstrating safety and efficacy and they're
6 labeled to do and not to say hey, those things
7 we're not going to take a look at and do a report
8 on just seems to me to make your job more
9 difficult and question what we have the FDA for.
10 So I'd take a hard look what the FDA says.
11 I had these discussions years ago with
12 the Food and Drug Administration. For whoever
13 you talk to, the people I've talked to up there
14 say when we approve something, be it a device,
15 drug, diagnostic test, we're not approving it for
16 Stanford, Hopkins or Cleveland Clinic. We
17 believe that if we let it in the marketplace,
18 it's going to work when lots of people use it,
19 everybody uses it, the average physician who is
20 licensed and capable of using it. You may
21 question that, but the FDA doesn't say that. If
22 we think that only certain experts can use it,
23 it's going to be effective there, then we're
24 going to put that in the labeling and
25 restrictive. So take a look at that question.
.00077
1 You heard this before. The document in
2 places, I think it needs more tone editing. Far
3 too much weight on randomized controlled trials
4 as the desired level of evidence. We're going to
5 have them, we're going to have more of them, but
6 they're going to be rare. And we can't afford
7 them all. And we all know there are lots and
8 lots and lots of reasons why we can't do them.
9 And the FDA doesn't require them every time even
10 for drugs. So I think you have to recognize
11 that. There's lots of good science being done
12 far better than before. Overemphasis on
13 randomized controlled trials is going to make
14 other research seem inadequate, and I think it
15 will lead to some research not being done, some
16 good research not being done, and things not
17 being developed.
18 I think in the probably hundred things
19 I've taken to HCFA over the last 15 years for
20 national coverage evaluations or at least a peek
21 at the national level without decisions being
22 made to float down to the care level, maybe five
23 technologies had very good powerful two or three
24 randomized controlled clinical trials, but I
25 never brought anything up here that wasn't pretty
.00078
1 good scientific evidence that would lead someone
2 to believe this is something that should have a
3 good shot at being covered, and I'd say
4 two-thirds of the time they were. So I'd go back
5 and recognize that there's a pragmatic end to
6 this area, and if you put five or six clinical
7 experts in a room before you to develop a
8 technology, you can probably get to a scientific
9 result that will make people feel that there's a
10 benefit there.
11 I think there's a serious source of
12 bias in this document. The bias is against new
13 innovations. Effectively what you're saying here
14 is -- and Dr. Brook and others have published on
15 this -- ten percent or less of all medicine that
16 we have right now has any scientific controlled
17 studies done on it. This effectively says we're
18 grandfathering all the old stuff. We're not
19 going to take a look at what we're comparing it
20 to. We want you to compare it to the old stuff.
21 What if the old stuff's never been studied? To
22 me one of the biggest problems we have in
23 technology evaluation of coverage policies is we
24 can't get rid of the old stuff.
25 For example, if the HMOs feel that ABMT
.00079
1 for breast cancer is not any good, are they still
2 covering it today? We need to take a look at
3 this. We've got to get rid of the old stuff and
4 question that before we just say the bar's higher
5 now for everything new. The science behind
6 everything new is definitely better.
7 Timing. I worry about how long this is
8 going to take. Reports, consultants, et cetera,
9 there's no way this is a six-month deal. It's
10 hard to believe. There may not be enough top
11 flight people with time who aren't publishing and
12 doing research to be able to do this evaluation.
13 I think MCAC should seriously take a look at
14 talking with HCFA on provisional coverage. If
15 the data isn't quite right, but we think it's
16 promising, then let's think about a situation
17 where we set out these are the outcomes we'd like
18 to have you take a look at. We will cover for a
19 fixed time period and stick to it, six months, a
20 year. This technology and other things that are
21 being done require you, the person who's getting
22 the benefit of having the thing covered, to
23 collect the information, come back to us a year
24 later because the clock stops, the coverage stops
25 here till you give it to us. I think you need
.00080
1 some kind of innovative idea here which will
2 allow research to be done.
3 So in short, be realistic in what you
4 ask for. Use the FDA. They've got to have a
5 role here. Don't ask for what you can't have.
6 It's very discouraging. Question the old stuff.
7 Don't be advised against the new. And time is
8 money and opportunity. I think you can
9 incentivize better science with coverage, and
10 we're not doing enough of it now, and I think
11 that can be done even within the legal
12 parameters. Thank you.
13 DR. SOX: Thank you very much. At this
14 point we've earned ourselves a break of about 20
15 minutes. So be back at five minutes after 10:00
16 o'clock.
17 (Whereupon, recess taken -- 9:45 a.m.)
18 (Whereupon, after recess -- 10:05 a.m.)
19 DR. SOX: If I could call the meeting
20 back to order, please. The first speaker is
21 Vicki Gottlich, Center for Medicare Advocacy and
22 Healthcare Rights Project.
23 MS. GOTTLICH: I'm Vicki Gottlich, an
24 attorney with the Center for Medicare Advocacy
25 and their Healthcare Rights Project in
.00081
1 Washington, D.C. The center is about 15 years
2 old. Our organization represents low income
3 Medicare beneficiaries. We currently have about
4 60,000 open case files in which we're trying to
5 get Medicare to pay for medically necessary
6 services for our clients.
7 I appreciate the opportunity to speak
8 here today, and I particularly appreciate the
9 opportunity to be representing beneficiaries
10 before this committee.
11 It is imperative for our clients that
12 HCFA establish a mechanism for protecting the
13 rights and interests of beneficiaries to receive
14 medically necessary care and services authorized
15 by their doctors. The current processes
16 available to beneficiaries, the claims and
17 appeals process and the national coverage
18 determination process under discussion today do
19 not protect beneficiary rights. Our clients and
20 other beneficiaries have had limited success with
21 the NCD process often because that process has
22 not been open to them. Few patients know they
23 will need a procedure or technology when the
24 process is underway, and even if they have timely
25 knowledge, they generally do not have the
.00082
1 resources to participate in the process.
2 Of utmost importance, the current
3 process for evaluating new procedures and
4 technologies and for reevaluating previous
5 coverage determinations is too slow. Conditions
6 deteriorate, and beneficiaries die, and I really
7 want to emphasize that we have had clients die
8 while waiting for HCFA to decide to cover
9 services, technologies and devices covered by
10 other insurers, including private industry, the
11 Department of Veterans Affairs and state Medicaid
12 agencies.
13 We applaud the subcommittee for their
14 efforts to clarify the national coverage
15 determination process. We are greatly concerned,
16 however, that the process used by HCFA and under
17 consideration today exceeds the agency's
18 authority by depriving beneficiaries of services
19 prescribed by their physicians for extended
20 periods of time.
21 Let me explain. I really don't need to
22 describe to this group what the Medicare statute
23 says because you're all familiar with the
24 Medicare statute. And the statute provides that
25 services will be covered as long as they are
.00083
1 medically necessary or Medicare will not pay for
2 services that are not reasonable and necessary.
3 The key point to the exception that
4 HCFA will not cover services is a determination
5 by HCFA that a service is not reasonable or
6 necessary. In other words, Congress placed the
7 burden on the agency to overcome the presumption
8 that the service is covered. Congress did not
9 prohibit coverage of services prescribed by
10 beneficiaries' doctors simply because enough or
11 the right kinds of studies showing their positive
12 value have not yet been amassed. This
13 interpretation is in keeping with the prohibition
14 against controlling the practice of medicine or
15 the manner in which medical services are
16 provided.
17 But the proposals today follow HCFA's
18 practice of placing the burden of proof on the
19 proponent to show why a service or technology
20 should be covered and to produce evidence of a
21 certain type in standard that is not always
22 available or even appropriate to the
23 beneficiaries who actually need the service.
24 The proposals do nothing to assure that
25 beneficiaries will receive quick access to the
.00084
1 services their own physicians found reasonable
2 and necessary.
3 For example, the suggestion that
4 outside experts be used in certain situations to
5 evaluate the evidence exasperates the delay
6 problem. In addition to harming beneficiaries,
7 such delays cause further disparities between
8 Medicare and private insurance coverage and
9 result in carriers having to deny Medicare
10 coverage for services they cover in their own
11 private insurance practice.
12 The proposals also fail to address
13 adequately the needs of the over five million
14 beneficiaries under age 65. Many members of this
15 community are adversely affected by HCFA's
16 failure to include new devices and technologies
17 among Medicare's covered services. Delays in the
18 processing for approving devices and technologies
19 result in beneficiaries with disabilities losing
20 their independence or their ability to function
21 to their maximum capacity.
22 Beneficiaries with disabilities are
23 also adversely affected by national coverage
24 determinations that are based on evidence
25 applicable only to the population over age 65.
.00085
1 For example, the Office of Civil Rights
2 of the Department of Health and Human Services
3 last year worked on and assisted a Medicare
4 beneficiary in her mid 40s who was denied
5 coverage of a potentially life-saving cancer
6 treatment because of a national coverage
7 determination. The national coverage
8 determination was based on evidence that the
9 treatment was not efficacious for women over age
10 65. Ample evidence existed, however, that the
11 procedure was effective for younger women, and
12 the Medicare HMO in which the woman was enrolled
13 covered the procedure for its non-Medicare
14 population.
15 While the appeals process is not a
16 concern of this group, it is really an important
17 element for our clients because the appeals
18 process provides no recourse for beneficiaries
19 who seek to challenge the national coverage
20 determination or to get Medicare coverage of a
21 technology or device not yet approved by
22 Medicare. The Medicare statute makes it nearly
23 impossible to challenge a national coverage
24 determination rule upon which services were
25 denied by preventing consideration of the issue
.00086
1 at the administrative level. If the claim
2 reaches federal court, a federal judge who
3 determines that the record is incomplete or
4 insufficient to support the validity of the
5 national coverage determination must remand the
6 case for supplementation of the record. The
7 court may only determine that an item or service
8 is covered after review of the supplemented
9 record.
10 So the individual who was adversely
11 affected by the obesity ruling that was discussed
12 earlier today would have to go through the whole
13 national coverage determination process and
14 couldn't go through an appeals process in order
15 to change the ability to get coverage for
16 treatment for obesity. If the national coverage
17 determination process is as lengthy as the
18 appeals process, it is going to be years, and
19 that's why we are very concerned about the
20 delays.
21 In sum, we are not advocating that
22 Medicare pay for quack services, which have been
23 shown to lack medical value. We are advocating
24 for an efficient coverage determination process
25 that allows Medicare beneficiaries to receive
.00087
1 Medicare payment for services and procedures,
2 devices and technologies that have been approved
3 by the FDA where appropriately are being covered
4 by private insurers, the VA and Medicaid, and are
5 found by the beneficiary's own physician to be
6 reasonable and necessary for treatment of that
7 beneficiary's illness or condition.
8 We also seek an effective and
9 expeditious appeals process that will allow
10 beneficiaries to challenge a denial of coverage
11 based on an NCD that is no longer supported by
12 medical evidence and practice. And while that's
13 not within your jurisdiction, we do ask that you
14 consider an expedited process to consider NCDs
15 that don't have any support for them. And there
16 are a lot, as I'm sure that you are aware. Thank
17 you very much.
18 MS. LAPPALAINEN: Vicki, would you
19 state for the record whether you have any
20 financial interest in the --
21 MS. GOTTLICH: I'm sorry. Our
22 organization has no financial interest in any
23 medical devices, and neither do I. Thank you.
24 DR. SOX: Our next speaker is Larry
25 Weisenthal from the Weisenthal Cancer Group.
.00088
1 DR. WEISENTHAL: My name is Larry
2 Weisenthal. I'm a medical oncologist in private
3 practice, and I provide the service that I'll be
4 describing. I'm a medical oncologist from
5 Huntington Beach, California. I participated in
6 the Medicare Coverage Advisory Committee meeting
7 last November 15th and 16th. My experience
8 related to this meeting is what now compells me
9 to offer comments concerning the structure and
10 procedures for future MCAC reviews.
11 My specific concerns involve, one,
12 serious defects in the advanced draft outline of
13 the proposed review process, and two, a lack of
14 appreciation for special considerations related
15 to laboratory testing in a draft proposal which
16 seems exclusively directed toward the review of
17 direct therapeutic interventions.
18 Rather than speaking in a theoretical
19 sense, I would like to use my own experience with
20 the November MCAC meeting to convey my concerns.
21 The draft proposal places heavy emphasis on a
22 series of independent reviews by so-called
23 experts in the field. Essentially the process
24 would be centered around a collection of up to
25 six independent written reviews by these
.00089
1 experts. There would appear to be a relatively
2 small role for the proponents of the technology
3 under consideration as they would have no
4 opportunity to rebut these reviews in advance of
5 the meeting. One can easily project proponents
6 having to use their entire 15 or 20 minutes or
7 less of allocated time at the meeting just to
8 hurry through complicated rebuttals of complex
9 and misconstrued data.
10 The November MCAC meeting considered
11 the issue of human tumor assays, which involved
12 short-term cultures of fresh biopsies of human
13 tumors in the presence and the absence of
14 anticancer drugs. Following cell culture, drug
15 effects are assessed by one of two end points,
16 either cell proliferation or cell death.
17 Historically all work in this area was
18 effectively abandoned in American universities in
19 the mid-1980s. The only major academic group
20 continuing work in this area was the lung cancer
21 group at the National Cancer Institute. However,
22 the NCI investigators had a primary focus on
23 creating cell lines through passaging and
24 subculturing. I anticipated a major emphasis on
25 three public studies arising from this work, and
.00090
1 I quoted several pages of my proposal, submitted
2 two and one-half months in advance of the
3 November meeting, to a detailed rebuttal of this
4 work.
5 Fearful that this rebuttal would be
6 overlooked, I was also forced to devote precious
7 minutes of my oral presentation to this issue,
8 which gave me no time to take the committee
9 through the many important positive studies and
10 prestigious peer-reviewed journals, which were
11 included in my written proposal, but which were
12 ignored by all the reviewers chosen by HCFA.
13 The major reviewer of the cell death
14 technologies proposed for coverage by me was Dr.
15 Edward Sauceville, associate director of a
16 developmental therapeutics program at the
17 National Cancer Institute. Dr. Sauceville did
18 not attend the morning presentations by the
19 proponents and their supporters. This led to the
20 following embarrassing statement, quote, you can
21 tell a patient who has the unfortunate diagnosis
22 of pancreatic cancer that they're likely not
23 going to respond to a medicine chosen after
24 having gone through an additional test to obtain
25 tissue and then test it for assay resistance.
.00091
1 This statement was embarrassing because
2 one of the earlier speakers had been a pancreatic
3 cancer patient who has been in complete remission
4 for more than three years after presenting with
5 liver and kidney metastases and then being
6 treated with an assay-selective drug regimen,
7 which everyone agrees would never have been
8 chosen absent performing the test.
9 Dr. Sauceville was also either not
10 shown or did not bother to read my written
11 proposal submitted two and one half months in
12 advance of the meeting. He showed his complete
13 ignorance of the field by failing to even
14 mention, much less consider, 80 percent of the
15 studies, totalling more than 1500 patients,
16 confining his review almost exclusively to
17 studies published before 1987 and to the
18 irrelevant studies that the NCI lung cancer group
19 alluded to previously. Neither did he nor any of
20 the other HCFA reviewers review and describe most
21 of the many studies correlating assay results
22 with patient survival.
23 Again, all these data references were
24 provided to HCFA two and a half months in advance
25 of the meeting. Nonconsideration of these
.00092
1 studies led to the following remark at the
2 December Executive Committee meeting by one of
3 your members, Dr. Ferguson, who related, quote,
4 we had very little survival information. There
5 were some unsettled elements. I don't remember
6 that there were other ones.
7 This remark forced me to make the
8 following frustrated comment at the December
9 Executive Committee meeting, quote, there were
10 many misrepresentations made, such as the lack of
11 survival data. I showed a slide at the meeting.
12 There are 15 studies showing strong correlations
13 with survival. This is not just based on
14 response.
15 That the above assessment of the
16 inadequacy of the outside review process is not
17 just a figment of my imagination was shown by the
18 comments of the committee chairman Dr. John
19 Ferguson again at the prior meeting of this
20 Executive Committee in December. Quote, another
21 was that the NCI representative presented a paper
22 which in my view I was a bit disappointed in
23 coming from my former institution that it did not
24 seem to me to be up to date and lacked in that
25 aspect. Dr. Ferguson went on to say so I am not
.00093
1 certain that the protagonists were given all the
2 critiquing information. We didn't have it. We
3 tried to give the protagonists time to respond.
4 I think that that could have been done a little
5 bit better in the sense that if all the critiques
6 of presented papers could have been given to the
7 presenters in advance, they might have had time
8 to prepare some rebuttal in response to the
9 critiques.
10 Even more egregiously misleading than
11 Dr. Sauceville's inadequate review was the
12 horribly misleading review of HCFA's Dr. Burken,
13 which by objective evidence demonstrably and
14 unfairly damaged the case put forward by the
15 proponents. By way of background, one of the
16 technologies proposed for consideration of
17 coverage was the cell proliferation assay based
18 on measuring tritiated radionuclide incorporation
19 as an assay end point.
20 Data was presented to document the high
21 specificity of this assay in identifying drug
22 resistance. In his review of the literature, Dr.
23 Burken devoted considerable time to technologies
24 which had been abandoned 10 to 15 years
25 previously and which were not proposed for
.00094
1 Medicare coverage by anyone in the November
2 review. One of these abandoned technologies was
3 a radionuclide precursor incorporation assay
4 measuring the incorporation of tritiated
5 thymidine or uridine only three hours after the
6 addition of anticancer drugs to freshly
7 disassociate the tumor cells.
8 This contrasts with the technology
9 under MCAC consideration which measured thymidine
10 incorporation five days -- not three hours --
11 after drug administration. Whereas the five-day
12 assay predicted for drug resistance with very
13 high specificity, the three-hour assay gave very
14 poor results and was abandoned by its own
15 proponents in the 1980s. Yet Dr. Burken showed
16 four different slides detailing the poor results
17 with this assay. This demonstrably confused and
18 mislead the panel, as conveyed by the panel's
19 industry representative, who showed us a table
20 constructed and to specify the MCAC panel
21 depicting the negative predictive accuracy
22 reported in the various studies and prominently
23 including the four studies with the long
24 abandoned three-hour assay which showed such poor
25 correlations.
.00095
1 The verbatim transcripts of the MCAC
2 panel's deliberations revealed the damaging
3 effect which the inclusion of these irrelevant
4 studies had on the MCAC enthusiasm for coverage.
5 Although clear from the transcript that there was
6 overwhelming support for HCFA developing a policy
7 to include coverage of these assays in at least
8 some clinical situations, this support would have
9 clearly been less reserved in the absence of the
10 misleading presentations by the reviewers chosen
11 by HCFA. This is crystal clear in the
12 transcripts of the meeting.
13 But the purpose of my comments here is
14 not so much to complain about the past as to help
15 the Executive Committee develop a better process
16 for future reviews. To this end we must begin to
17 appreciate that we are working in a time when an
18 increasing number of important advances in
19 medicine are occurring outside the traditional
20 NIH and university research system.
21 In the case of human tumor assays,
22 there are no experts at all in either American
23 universities or at the NIH. No investigator at
24 these institutions has contributed in any way to
25 the literature in the field I represent of cell
.00096
1 culture drug-resistance assays with cell death
2 end points. In my 20 years of full-time work in
3 this field, I've talked with hundreds of
4 university and NIH-based investigators with an
5 opinion about this field. It's been more than
6 ten years since I last had a discussion with a
7 non-European and non-Japanese university-based
8 investigator to be able to discuss the subject
9 based on an intelligent understanding of concepts
10 and literature.
11 So HCFA must be very careful to ensure
12 a central role of the proponents of the new
13 technology in presenting and explaining data to
14 the MCAC panels.
15 Cutting to the chase, we propose the
16 following modification in the overall outline of
17 the proposed system. First, the process begins
18 with a formal request to HCFA for coverage
19 consideration. Once informed that HCFA agrees to
20 consider the issue, the proponents are
21 responsible for presenting a formal defense of
22 their proposal centered around a description of
23 technology and complete review of all relevant
24 data and literature. This proposal is then sent
25 to each of the outside reviewers. The outside
.00097
1 reviewers then prepare their own independent
2 reviews, which are then given back to the
3 proponents for rebuttal. The rebuttals go back
4 to the reviewers who are allowed to have the
5 final word in the pre-meeting written
6 presentations and reviews provided to the MCAC
7 panel. The proponents should also certainly
8 receive a copy of this final review while in
9 advance of the meeting.
10 The meeting itself could then take
11 place with all the complicated and contentious
12 issues having already been pre-argued. The
13 meeting itself would begin with relatively brief
14 summations by both proponents and reviewers,
15 followed by a devotion of most of the time to
16 open discussion by the committee with committee-
17 directed questions to both proponents and
18 reviewers. However, prior to final deliberations
19 and votings, both proponents and reviewers should
20 have the opportunity to make brief final remarks.
21 I've got one page here which I won't go
22 over the time, but could this be put into the
23 record?
24 DR. SOX: Sure. If you want to submit
25 something in writing.
.00098
1 DR. WEISENTHAL: Thank you.
2 DR. SOX: Our next speaker is Sandy
3 Sherman, Assistant Director of Division of
4 Federal Affairs & Outreach of the American
5 Medical Association.
6 MS. SHERMAN: Good morning. I just
7 have a brief statement from Dr. E. Radcliffe
8 Anderson, who's the Executive Vice President and
9 CEO of the AMA, regarding your discussion paper.
10 After the first MCAC Executive
11 Committee meeting in December, I wrote to
12 Nancy-Ann DeParle to say that the AMA was
13 impressed and gratified by the commitment of the
14 advisors and HCFA to ensure that MCAC
15 recommendations would be grounded in scientific
16 evidence of clinical effectiveness. I also said
17 that the meeting made it clear that she had
18 fulfilled her promise to create an open, timely
19 and accountable process for making national
20 coverage decisions.
21 The discussion paper that the committee
22 members prepared for today's meeting underscores
23 the observations we made in December. The
24 recommendations for evaluating evidence clearly
25 state the key issues to consider in assessing the
.00099
1 state of the knowledge regarding medical
2 interventions proposed for Medicare coverage. We
3 are pleased that in addition to recommending a
4 critical review of evidence from clinical trials,
5 the Executive Committee or the members who
6 prepared this proposal recommend that the
7 standard of excellence for the evidence report
8 include work developed by the national medical
9 specialty societies. We also commend the
10 advisors for recommending that panel members take
11 an active role in framing the questions to be
12 addressed by the evidence report, participate in
13 the report's preparation and seek external review
14 of the evidence reports.
15 Prior to the MCAC's formation, the AMA
16 had expressed concern that Medicare coverage
17 decisions might be driven to a large degree by
18 information presented by those with a vested
19 interest in coverage instead of by the available
20 scientific and clinical evidence. The discussion
21 paper developed by the advisors has allayed our
22 concerns in this regard, and we encourage
23 adoption of its recommendations.
24 DR. SOX: Thank you very much.
25 Our last speaker is Thomas Meskan,
.00100
1 president of Medical Alley.
2 MR. MESKAN: Good morning. My name is
3 Tom Meskan, president of Medical Alley. In terms
4 of your financial statement, obviously we have
5 members who pay dues to our association, and I
6 presume that a number of them have issues pending
7 before the agency.
8 For those of you who aren't familiar
9 with Medical Alley, we're a 15-year-old not-for-
10 profit trade association based in Minnesota who
11 has members from all aspects of healthcare. Our
12 members include health plans, medical device
13 manufacturers, hospitals, clinics, long-term care
14 organizations and academic health centers. Our
15 mission is to serve as a collaborative form which
16 promotes an environment to enhance innovation in
17 healthcare.
18 I appreciate the opportunity to share
19 our perspective and thoughts as they relate to
20 the discussion paper. We think that the MCAC
21 process is an important aspect of Medicare's
22 decision making and want to acknowledge and
23 express our thanks for the time and effort all of
24 the people, both you as panel members and agency
25 staff, are spending to try and make the MCAC a
.00101
1 valued component of Medicare decision making.
2 To help you get a sense of the
3 orientation of our organization, I will point out
4 that we believe that Medicare should be a prudent
5 purchaser of services, and we think that it is
6 important that the agency has appropriate levels
7 of resources to do its job. At the same time we
8 believe that the environment surrounding
9 Medicare, and for that matter, all of healthcare,
10 should be dynamic so that patient care improves
11 in a timely and continuous manner.
12 With regard to our principles on
13 generating evidence, they are that HCFA
14 preferences for how evidence is presented should
15 be transparent. Any approach to decisions about
16 coverage criteria should be administratively
17 feasible for both the agency and the
18 stakeholder. It is desirable that stakeholders
19 achieve the level of valid scientific evidence
20 necessary to demonstrate that a service should be
21 covered, and there should be a minimization of
22 potential for bias into conduct, reporting and
23 analysis of studies.
24 Our comments today fall into two
25 categories. First, we want to offer some
.00102
1 observations about the role of perceptions in the
2 success of your efforts. Second, we will offer
3 some specific reactions to some of the text in
4 the discussion document.
5 It is clear by looking at the names
6 which make up this committee and the impressive
7 roster of individuals that make up the MCAC
8 panels that there is a wealth of expertise
9 available to the agency. I had the opportunity
10 to introduce myself to Dr. Sox during the break,
11 and he, if I can paraphrase him, said what he
12 liked about his involvement in this committee is
13 its potential effect to a large number of human
14 beings and their health condition. And I think
15 that that's a very accurate statement. And the
16 most important point is we must make sure that
17 you guys do everything you can to maximize your
18 potential.
19 Obviously each of you are approaching
20 your MCAC responsibilities in good faith and with
21 a desire to achieve the goals of consistency and
22 accountability. Further, you have laid out the
23 recommendations in a manner which strongly
24 signals your interest in promoting the greatest
25 possible degree of rigor in the methods used to
.00103
1 generate evidence.
2 We too want to encourage the
3 development of a decision-making process that
4 will be informed, and we also support the
5 continued improvement in the way the supporting
6 data is collected and utilized. Nonetheless,
7 this committee, the agency and external
8 stakeholders must acknowledge the history of
9 coverage policy development so that whatever
10 process this committee decides upon enjoys
11 support of the largest possible percentage of
12 affected stakeholders. In this manner you can
13 ensure that your time and efforts are valuable.
14 In brief, that history suggests that
15 whatever approach is taken by the agency and
16 those who advise it to create greater detail on
17 the concept of reasonable and necessary will be
18 subject to extremely close scrutiny.
19 We know the examples, a coverage
20 regulation that has been kicked around since
21 1987, the fact that this committee is just
22 starting to get off the ground two years after
23 the GAO found the act to be in violation of FACA.
24 We also know that frequently in coverage decision
25 making it becomes subject to second-guessing by
.00104
1 Congress.
2 We raise this because we want to
3 encourage you to get this process off on the
4 right foot. We want the MCAC process to succeed
5 and be used. And while I heard Dr. Bergthold's
6 comments about the effort that you went towards
7 submitting this, it serves no one's interest if
8 your approach is perceived incorrectly or not as
9 so academically grounded that MCAC becomes
10 nothing more than another health policy center
11 which provides insights that have little life
12 beyond those who formulate and to make them
13 internally.
14 We believe it is fair to say that
15 outcomes research and technology assessment are
16 evolving disciplines. Further, while the
17 document does not say so, it is extremely rare
18 that data is ever perfect. Similarly, a number
19 of decisions faced by panels are likely to
20 inquire around one of the truisms that surround
21 healthcare. That is part art and part science.
22 Therefore, we encourage you to modify
23 your discussion document to acknowledge these
24 factors and create the opportunity for our
25 acceptance of your approach. Similarly, it will
.00105
1 enhance your opportunity to improve the
2 effectiveness of the panels.
3 We offer you the following language as
4 an example of a kind of statement that you might
5 make. Evidence presented to support a coverage
6 decision should be deemed acceptable if it is
7 ethically appropriate, administratively feasible
8 and if it meets the current generally accepted
9 used requirements for evaluation of a health
10 service typically found within a technology
11 assessment literature that were in place at the
12 time the study was undertaken. This is not to
13 say that the evidence is then accepted as meeting
14 a case for coverage, but rather reflects a common
15 sense approach to considering the practical
16 implementation issues which surround the
17 methodology options for generating data.
18 It is simply the case that a majority
19 of the people who are involved in generating
20 evidence for decision making are well-meaning
21 people who want to do the best job they can.
22 This does not mean that they are at all as
23 schooled and knowledgeable as you on the nuances
24 of evidence generation. Your document needs to
25 implicitly acknowledge these individuals and to
.00106
1 speak to them in a manner which allows them to
2 see clear, feasible pathways to being
3 constructive contributors to Medicare coverage
4 decision making.
5 We suggest that with that opportunity
6 comes an obligation. We would suggest that the
7 document be modified to express the interest of
8 panels in receiving from stakeholders the
9 rationale which drove such things as the study
10 design, data sources utilized, the rationale for
11 what the service is being compared to, the time
12 horizon that's chosen and the statistical
13 analysis methods used to address random events.
14 In addition, we think it's appropriate for
15 stakeholders to describe this data from
16 unpublished sources. This will provide useful
17 information to the panels as they seek to weigh
18 the value of the evidence presented.
19 Let me now move to our observations
20 about the specific aspects of the document.
21 First of all, we would note that the paper fails
22 to acknowledge those stakeholders who have
23 already completed or are currently in the process
24 of carrying out efforts to generate data for a
25 national coverage decision. The paper needs to
.00107
1 provide some guidance so that these stakeholders
2 and/or the panels do not feel that an
3 organization must necessarily go back to square
4 one in generating evidence because of this
5 document.
6 Moving to another area, while we
7 recognize the panel's purpose is to focus on
8 issues of science and evidence, it's somewhat
9 ironic that the words or concept of a patient do
10 not appear until page 6. While the document's
11 failure in this regard could be seen as semantic
12 window dressing, we believe it's important that
13 we all keep front and center in the end. This is
14 what we're all about.
15 That said, the committee has indicated
16 its interest in the panel's making conclusions
17 about health outcomes. We would ask that the
18 committee modify the text on page 7 or at least
19 my Internet version on page 7, item 3. This text
20 addresses the need for the panel to explain its
21 conclusions. We suggest that the committee ask
22 the panels to describe as specifically as
23 possible how each of the various health outcomes,
24 including, but not limited to, mortality,
25 morbidity, functional status, quality of life and
.00108
1 patient experience were factored into its
2 decision making. By making the reporting
3 requirements more detailed, the goals articulated
4 in this item will be better achieved.
5 We also believe that significant
6 thought should be put into the item on page 7
7 about the evidence reports provided to the
8 panels. Although the ability of this proposal to
9 operate in a timely manner is suspect, we are
10 also very concerned that the document does not in
11 any way provide affirmative action between the
12 stakeholder and MCAC on what materials will be
13 contained in the evidence report. We think the
14 document should provide a mechanism for dialogue
15 between stakeholders and the appropriate panel
16 representatives before submitting the report.
17 Another area of concern is found on
18 page 5, the last sentence dealing with bias. The
19 text can be read to require that the panels
20 describe why bias does not account for the
21 results. Conversely, the subjectivity, if you
22 will, in judgment calls which are involved with
23 these issues, we believe that the panel should be
24 empowered to describe why it's comfortable with
25 its conclusions.
.00109
1 Finally, on page 6, the last two
2 sentences on external validity, the terms typical
3 practice setting and general practice setting
4 appear to be used interchangeably. Because of
5 the importance that the agency puts on
6 appropriateness of making decisions, we believe
7 it would be valuable to clarify what the terms
8 typical and general mean.
9 In sum, we believe that all Medicare
10 stakeholders are benefited by the recognition
11 that improving the Medicare coverage decision-
12 making process is a long road. We believe the
13 MCAC process is an important resource for the
14 agency and for external stakeholders, but at
15 these early stages of this effort care must be
16 taken to create conditions for success. We know
17 that the talent, insight and good efforts exist
18 on this committee to achieve these conditions.
19 We stand ready to assist you in every way we can
20 and thank you for your attention and
21 consideration of our views.
22 DR. SOX: Thank you very much. Before
23 we go on to the HCFA presentation, Sharon's going
24 to read a letter that we just received today from
25 the ACP-ASIM on the same day that AMA commented
.00110
1 on our document.
2 MS. LAPPALAINEN: The letter is
3 addressed Dear Ms. Lappalainen, the American
4 College of Physicians-American Society of
5 Internal Medicine (ACP-ASIM), representing over
6 115,000 physicians who specialize in internal
7 medicine and medical students, wishes to offer
8 its comments and concerns on the draft report of
9 the subcommittee of the Medicare Coverage
10 Advisory Committee's Executive Committee
11 entitled, Recommendations for Evaluating
12 Effectiveness. ACP-ASIM is generally supportive
13 of these recommendations, but feels it critical
14 that the MCAC strike a healthy balance between
15 assuring a coverage review process which is
16 credible and defendable from a scientific
17 viewpoint, yet not so mired in technical detail
18 that final coverage decisions are unreasonably
19 delayed.
20 ACP-ASIM is very supportive of the
21 draft report's objectives; that important
22 clinical coverage decisions be reviewed on the
23 basis of sound and objective clinical evidence by
24 the MCAC's six medical specialty panels, and that
25 there be a standardized methodology and format
.00111
1 for panels to present their recommendations to
2 the MCAC Executive Committee, thereby allowing
3 the Executive Committee to make uniform,
4 high-quality and scientifically defendable
5 coverage recommendations to HCFA. We also
6 support the draft report's recommendation that
7 the MCAC only focus on the clinical and
8 scientific questions around the medical
9 effectiveness of new items and services and the
10 comparative effectiveness of new items and
11 services relative to existing alternatives, and
12 that the MCAC not address questions about dollar
13 costs of new items or services.
14 We are impressed with the amount of
15 scientific rigor the draft report proposes for
16 assessing the adequacy of clinical evidence
17 related to a new item or service and calculating
18 the magnitude of the health benefit such coverage
19 would have on the Medicare population. We do
20 wish to raise some technical concerns under the
21 draft report's section on Evaluation of
22 Evidence.
23 On page 3 the discussion of potential
24 sources of bias has some noteworthy ommissions,
25 including double-binding, perfect compliance,
.00112
1 adequate length of follow-up, distinct treatment
2 separation and inappropriate statistical
3 analysis. Imperfections in any of these would
4 permit bias to enter into a randomized controlled
5 clinical trial and thus make the results less
6 valid for the population under study and thus
7 difficult from which to generalize.
8 We also feel the draft report's
9 recommendation on page 4, that MCAC panels be
10 required to describe possible sources of bias and
11 explain why a panel decided that bias does not
12 account for the results, should be applied in all
13 coverage decisions, not just the limited
14 circumstance of uncontrolled studies described on
15 page 4.
16 Also, on page 5 where seven categories
17 of size of health effect are presented, there
18 appears to be one category omitted, which we
19 would recommend the addition of, more effective,
20 but with disadvantages.
21 In summary, ACP-ASIM believes it is
22 vital that coverage decisions remain in the hands
23 of the medical experts comprising the panels of
24 the MCAC and that the credibility of this body
25 will depend on striking a balance between
.00113
1 scientific rigor and decision making which is not
2 bogged down in process. Decisions reached by the
3 MCAC must be based on the best mix of objective
4 data and professional judgment possible and lead
5 to coverage recommendations that have a
6 compelling weight of evidence, yet are rendered
7 in reasonable time frames to avoid work backlogs
8 which might undermine MCAC effectiveness and
9 credibility.
10 ACP-ASIM supports the MCAC coverage
11 decision process and welcomes the opportunity to
12 contribute to its evolution. We believe the time
13 spent now will pay great dividends in the future
14 and that the MCAC's evidence-based decision-
15 making model will soon become one of which we can
16 all be proud. Sincerely, it is signed by Whitney
17 W. Addington, M.D., F.A.C.P, president. Thank
18 you.
19 DR. SOX: We'll now move on to the HCFA
20 presentation by Dr. Kang and Dr. Hill. Jeff, go
21 ahead. Well, Bob, you had something to say.
22 DR. BROOK: I don't quite understand
23 the transition here, and I'd like some
24 clarification on the process. Up to now we've
25 had a description of the subcommittee report and
.00114
1 then a public session with public comment. What
2 is this part?
3 DR. KANG: This is actually the HCFA
4 comment.
5 DR. BROOK: Is this the response to our
6 subcommittee report?
7 DR. KANG: Yes.
8 DR. BROOK: I'm wondering whether the
9 process we ought to -- I mean since we are an
10 advisory committee to HCFA, do we want to have
11 some discussion of the committee before we hear
12 what HCFA thought of the report in relationship
13 to the public report or is this a process that's
14 prescribed by law or something that we can't do
15 this? I'm just wondering which way we want to do
16 this since we're advisory to HCFA anyway. Do you
17 want us to put all this together when we try to
18 deliberate or just look at the public response
19 first?
20 DR. KANG: I'm actually okay either
21 way, quite frankly, because there's many of the
22 issues here which have been raised which I think
23 we can resolve through discussion. So if we want
24 to kind of cut to the chase here, that's fine
25 with me.
.00115
1 DR. HILL: In the sense that the
2 subcommittee asked for a comment and a report to
3 be given, when something's presented to the
4 panel, we also would like to be able to comment
5 about the subcommittee report at this point and
6 hope that you would take that into consideration
7 in your mix.
8 DR. SOX: Alan, do you have a
9 suggestion?
10 DR. GARBER: Just speaking for myself,
11 I would like to hear HCFA's comments before the
12 committee deliberates so we can deal with all of
13 the comments as a whole.
14 DR. KANG: I'm going to nix my
15 presentation then. I actually had only one
16 comment then. Dr. Hill has a bunch.
17 I wanted to note that when I was a real
18 doctor -- I guess I'm no longer a real doctor --
19 it's been awhile since I've practiced --
20 practicing geriatrics, I had to make very
21 difficult choices and/or recommendations for my
22 patients almost every minute of the day which
23 diagnostic test to order, should I recommend
24 hospitalization or home care, what treatment
25 options should I suggest et cetera. Usually this
.00116
1 involved choices amongst well-understood,
2 commonly utilized possibilities.
3 Sometimes, though, something new or
4 something new to me was as an appropriate
5 consideration. Usually in these situations I
6 turned to the medical evidence and the literature
7 to help me make a choice in this decision. I
8 think I did that largely in part because I wanted
9 to be sure before abandoning the old that using
10 the new would be better. I think in many ways
11 this is what we're wrestling with, and this is
12 what national coverage decisions are about that
13 we face frequently with new technology. What
14 does the evidence or science say about the new
15 technology?
16 In practice, though, I must admit I
17 also recall the patient's condition and the
18 availability of alternatives had a lot to do with
19 how I reviewed the evidence. If our patient was
20 in serious trouble and there was a lack of any
21 other beneficial alternatives, it actually made
22 me more likely to offer the service even if the
23 literature was suboptimal. I think this was
24 especially true if the risk of the service or
25 procedure was very small.
.00117
1 So I just ask in your deliberations
2 today that you discuss whether or not the
3 patient's condition, the availability of other
4 alternatives and the risks associated with the
5 service should affect how we actually view the
6 evidence.
7 That said, I applaud and thank you for
8 your efforts to deal with this in a consistent
9 manner for all panelists on how we read the
10 evidence. I believe that actually you're off to
11 a great start, and there's many things that can
12 be resolved today.
13 DR. HILL: Thank you. I'll be as brief
14 as I can. First of all, I want to say on behalf
15 of our group within HCFA that the subcommittee
16 report is both admired and appreciated by us.
17 Nothing that I will say should be taken as a
18 denigration or a disparagement of this important
19 contribution to HCFA's efforts to improve our
20 coverage decision-making process.
21 The report's recommendations for an
22 optimal process, speaking from the position of
23 the people who are going to have to carry this
24 out, appear to be well-challenging. It may be
25 that at least for some decisions, we will have to
.00118
1 commit to all the steps you outlined, but that
2 possibility causes us as well as others to have a
3 care for the time required.
4 This is the most open and accountable
5 process for making national coverage decisions in
6 the history of Medicare. When we designed and
7 started this new way of doing business, including
8 the MCAC, we knew that the period required to
9 reach a decision would often include required
10 minimum components and time periods because of
11 the steps. For example, announcing the planning
12 of MCAC panels' open public meeting means some
13 time is needed. As we talk today about how to
14 prepare for and get the best advice from MCAC
15 panels, we're thinking again about the time
16 required. But let me be plain. We were not
17 then, and we are not now, hiding behind the
18 process to delay coverage, to delay getting the
19 latest evidence-proven treatments to Medicare
20 beneficiaries, and we do not want anyone else to
21 either.
22 Our intentions and success in meeting
23 those intentions are and will continue to be
24 clear. We announce matters under consideration
25 for coverage decisions on the web with due
.00119
1 dates. If we can't meet our self-imposed
2 deadlines, we give our reasons, again posting
3 them publicly. This process must not be driven
4 back into a black box by criticism of that
5 process, including criticism of timing.
6 Our goal is to reach well-reasoned,
7 scientifically sound decisions as rapidly as can
8 be consistent with that level of quality. We
9 believe that this committee shares that goal with
10 us, and we appreciate its comments on how to keep
11 things moving.
12 Let me refer to a couple of specifics
13 in the subcommittee report that may raise
14 concerns for process duration. The suggestion
15 that each panel explain its conclusions in
16 writing should not in our view delay a decision
17 until a second panel meeting months later is
18 voting on that right. We should be able to
19 address this commendable desire for
20 accountability, as consistently expressed in this
21 suggestion, without more time than is already
22 contemplated to write up and post the summary of
23 that meeting. This is something we're already
24 going through.
25 The suggestions regarding the structure
.00120
1 of the evidence presented to the panel should not
2 delay. We are committed to presenting high-
3 quality and well-organized information as called
4 for in the subcommittee report and doing so
5 within the time frames previously contemplated.
6 We will get help doing this in a timely way when
7 necessary, and we are already doing this for the
8 next planned panels.
9 I'm pleased to see Dr. Deborah Zarin
10 from our well-respected sibling, the Agency for
11 Health Research and Quality, with us today in the
12 audience. Dr. Kang and I have met on multiple
13 occasions with AHRQ's leadership, and we look
14 forward to their involvement as an important
15 resource for us in examining evidence and
16 preparing for MCAC panels. We'll be talking
17 about the subcommittee's time frames with them.
18 Finally, on the time frame issues I
19 want to respond to the subcommittee's item number
20 6, expert review of evidence reports. At the
21 present time we are not planning to do this in
22 every case. Even if time were not an issue --
23 and it may not be if this added step can be
24 accomplished within current expectations -- we
25 still regard this as a quality control feature.
.00121
1 If we're doing a good job with the presentations
2 to the panels and the postings on the web, if the
3 process seems to be working without this step, we
4 do not presently intend to make additional
5 external review part of the routine.
6 The other major concern we have heard
7 about the subcommittee report -- you've heard it
8 too -- is that it seems to set some impossibly
9 high hurdle to bar every new technology without
10 any regard for type. We don't read your
11 statement that way, but this should not be a
12 concern regardless because we continue to explain
13 that we are not abrogating our responsibilities.
14 We understand that we have to make the coverage
15 decisions. You advise us, and we decide in part
16 basing our decision on your advice. So we want
17 to know the basis of your advice, your
18 recommendations, your thinking. We will want to
19 know what's behind the MCAC panel's inclusion
20 about evidence. We don't expect the panel to,
21 nor can we allow the panel to, decide for us
22 whether or not there's enough evidence to allow
23 us to cover it.
24 For example, when the subcommittee
25 report says uncontrolled studies are never
.00122
1 applicable, I read, in the context of that
2 section, that if a clinical experiment reported
3 in medical literature carries the possibility of
4 bias in selection of patients, we understand the
5 difficulties of explaining away that bias without
6 randomization or other forms of controls.
7 Dr. Sykes gave a good explanation of
8 bias in his presentation to the subcommittee
9 report. Does the risk of unaccounted for
10 selection bias mean that we shouldn't give the
11 experiments' results much weight in deciding
12 whether or not to cover the tested treatment?
13 Possibly. Does it mean we automatically refuse
14 to cover? No.
15 As the subcommittee report suggests,
16 observations alone may sometimes allow a panel to
17 make conclusions about effectiveness. Such
18 suboptimal evidence may allow us to conclude that
19 Medicare should cover the service. Deadly
20 diseases without alternatives come to my mind
21 immediately as such a situation, also logical
22 consistency with general medical science
23 understanding. The proof required to allow
24 applicability to the Medicare population might be
25 less where the application makes sense than when
.00123
1 it's counterintuitive or inconsistent, hard to
2 explain in the context of the rest of the
3 science.
4 I also see no credibility in the
5 assertion that the committee is threatening to
6 tell HCFA that one threshold fits all. No one
7 should take seriously the suggestion that we
8 might require unrealistic trials such as double-
9 blind tests of surgically implantable devices as
10 a dodge to avoid covering something. We said,
11 and I say again, that the sector-specific
12 guidance documents are purely of our
13 quality-oriented coverage plan, and they are the
14 next step after a coverage regulation proposal in
15 the federal register. We have already
16 demonstrated, in the coverage decisions made so
17 far under our new process, that we are aware of
18 and can properly include the flexibility
19 necessary for the variety of situations we face.
20 But the questions you ask are at least
21 potentially constant, and the important questions
22 you've asked of this document can't be ignored.
23 We still want to know whether studies that do not
24 focus on patients over 65 produce results that
25 can be applied to the Medicare population of that
.00124
1 age group. It's possible that the answer can be
2 no or even unsafe over 65, and we might consider
3 still covering, but only for our disabled and
4 ESRD beneficiaries who are within the age range
5 where medical benefit is shown by the evidence.
6 So to the subcommittee we say thank you
7 for this important contribution. Thank you for
8 these questions. To industry and those who want
9 to cover our product or service, we say let's
10 look together at these questions. We understand,
11 and you know we understand, that these questions
12 do not control HCFA's coverage decision making,
13 but they will help inform and improve the quality
14 of those decisions. And to our beneficiaries and
15 the public generally we say we will be faithful
16 stewards of your health and the health of the
17 future beneficiaries. We will ask these
18 questions. We will continue the work begun two
19 years ago, always listening to the medical
20 community, providers, consumers and manufacturers
21 and promoters, the work of improving Medicare's
22 national coverage decision process. Let's keep
23 going together.
24 DR. SOX: Thank you. We now go into an
25 open committee deliberation, and what I'd like to
.00125
1 suggest is that we start our deliberations and
2 perhaps spend as much of the next hour as it
3 takes to ask follow-up questions of people who
4 made presentations to us, both from the public as
5 well as HCFA, and then, again depending on how
6 much time it takes us, either proceed on to
7 starting a round table discussion of this
8 document and what we need to do to come to a vote
9 to recommend to HCFA.
10 So with that brief introduction, I'd
11 like to focus for now on trying to ask questions
12 of the various presenters and so forth. Bob?
13 DR. BROOK: Panel, can I raise a
14 process issue of what we're trying to accomplish
15 today? Let me tell you what I've heard. I
16 didn't hear anyone except maybe HCFA have a --
17 I'll retract that. I didn't hear anybody sort of
18 say the document is out of bounds. It should be
19 burnt and thrown away. I've heard a lot of
20 wordsmithing in some places, a lot of questions
21 about tone and other questions, but no wholesale
22 disregard for it.
23 The question I'm asking is should we
24 consider on this committee a bifurcated process?
25 We need something to help the next set of panels
.00126
1 get started with. We could say that we've gotten
2 there with this document as getting started, and
3 we could ask the people that presented as well as
4 other people to take the document we have and
5 actually instead of doing what we did here,
6 require them to do what we did ourselves, which
7 is to white out, edit, alter whatever they would
8 like in that document and provide a justification
9 and a reason for what they're trying to
10 accomplish by doing that and then take this so
11 that we would actually have a written record that
12 basically would allow us to look at this
13 paragraph by paragraph, sentence by sentence on
14 the belief that both the people at HCFA and the
15 people of the subcommittee and people of the
16 committee will disappear sooner than we can
17 probably imagine given our mortality.
18 And I wonder whether that kind of a
19 process would be one that we would then have a
20 written record of what people really would do to
21 this document if they were all part of the
22 subcommittee. And then the subcommittee would
23 then take those, produce a written record of how
24 we responded to that and in a document that then
25 we would do and produce as a second version and
.00127
1 continue to involve this process over time as we
2 get experience with it.
3 So the thought here is go with what
4 we've got now as advice to the committees to do
5 the next round of the panels, get written input,
6 continue to revise, continue to deal with this
7 kind of a document and make it an evolutionary
8 document with a history behind it so that we can
9 continue the process forward.
10 And as we get feedback, both from how
11 it worked in the panels, and what the public
12 believes about this feedback, we could then
13 continue to modify this document and do it as
14 sort of that kind of an approach as opposed to us
15 trying to ask questions, get off-the-cuff
16 responses, some of them well thought out, but not
17 sort of at the level of how would you change this
18 sentence? When you mean tone, okay, what do you
19 really want done here? So getting commitment in
20 writing to what people really want done.
21 I'm wondering whether that would be a
22 process that would get us further along.
23 DR. SOX: Let's discuss that. It's a
24 reasonable proposal. Let's have some serious
25 discussion.
.00128
1 MS. LAPPALAINEN: Right. We have the
2 document available for projection, and we are
3 prepared to have someone make edits now. For the
4 entire afternoon we have set aside a large amount
5 of time today for the committee to make those
6 kinds of suggestions to the document. Because
7 the subcommittee met in essence in private, the
8 deliberation and the review of the document needs
9 to be in public today in order to satisfy the
10 Federal Advisory Committee Act. And this is why
11 we have called the meeting today so that the
12 entire Executive Committee could deliberate and
13 review in open public format this document.
14 DR. SOX: Okay. Well, Bob, in essence,
15 I think, has said that we need to get rolling
16 with the process, that the document that we've
17 generated so far doesn't have any deadly flaws in
18 it, but at the same time we've had some very
19 useful comments and perspectives that might
20 strengthen the document if they were incorporated
21 into it.
22 And perhaps we could simply have a
23 two-part process, which we would decide whether
24 or not to use the document as it is now to help
25 the panels in their deliberations that are on the
.00129
1 schedule right now and meanwhile give the public
2 an opportunity for input into the document and
3 reframe it as seems appropriate, then come back
4 at our next meeting to present what we've come up
5 with for further discussion and options.
6 DR. BROOK: That's not what I said.
7 It's close, Hal.
8 DR. SOX: Thank you.
9 DR. BROOK: I think that we could have
10 open deliberation today at the level of a
11 committee about do we think this is good enough
12 to overcome some of the major problems with the
13 running of the next set of panels? And we ought
14 to confine our discussion to that for us at this
15 moment. But at the same process, I've heard that
16 there are people that really want significant
17 written changes in this document that we all may
18 think there's no problem with, and it would
19 improve the document.
20 And if we had a process of saying --
21 and I don't know the timing of this here, but you
22 have six weeks to take this document and to write
23 down, not just the edits, but just the reason you
24 want it changed, the justification, what you're
25 trying to accomplish, and then have the
.00130
1 subcommittee look at that and then try to
2 incorporate as much as this into a revised
3 document and bring it back to the Executive
4 Committee so that we get closer to what people
5 really want and go through the step before we
6 meet again as an Executive Committee of actually
7 looking seriously at those changes and
8 incorporating them, then we would have a written
9 reason, a written justification, and then we
10 could respond as a committee and say yes, we
11 agree with, no, we don't, for these reasons. And
12 this would be a different kind of a process.
13 DR. SOX: So we have comments. I was
14 looking this way. So Alan, why don't you take
15 the first one.
16 DR. GARBER: I'll be very brief. I
17 just wanted to remind everyone -- and correct me
18 if my memory is incorrect -- that at our last
19 Executive Committee meeting we said that the
20 subcommittee would produce a document that's
21 really intended to be interim to provide guidance
22 to the panels until HCFA issues its regulations.
23 So one thing to keep in mind, none of us, I
24 think, have the intention of producing something
25 that's going to be permanent. If this does
.00131
1 happen to coincide perfectly with the rules that
2 HCFA eventually develops, that would be great. I
3 don't think we have the expectation that that
4 will necessarily happen.
5 So this is indeed an interim document,
6 and I don't think the idea is to make this so
7 pristine and perfect that it never needs to be
8 changed because we are almost bound to change
9 this in the course of the next year, year and a
10 half, however long it takes.
11 The second point is that I think we
12 said at the previous meeting that we hoped that
13 we would more or less wrap this up at this
14 meeting, and I think it's premature to talk about
15 longer term changes until we've heard from the
16 members of the Executive Committee, who did not
17 yet have an opportunity to comment on the
18 document, to get some sense of whether this is
19 very close to the right ballpark and just needs
20 some technical revisions that can be handled
21 today or if it needs very extensive revisions.
22 So I think we need to discuss ongoing
23 revision only after we've heard from the
24 Executive Committee has a whole.
25 DR. SOX: So Alan, let me understand
.00132
1 you correctly. Are you saying that we can't act
2 on Bob's proposals until we discuss the document
3 as it currently stands looking at it as an
4 interim document that's going to help us get off
5 the ground in the next 12 months or so?
6 DR. GARBER: Exactly.
7 DR. SOX: That certainly seems like a
8 reasonable suggestion. But why don't we see if
9 there are any other comments.
10 Jeff, did you have your hand up?
11 Leslie?
12 DR. FRANCIS: I wanted to comment that
13 I think that we should go actually section by
14 section with the idea of whether or not there are
15 things in this document, using it as a general
16 framework, that we think are problematic even on
17 an interim basis. One example might be the
18 implication in the generalizability section to
19 the Medicare population, that the Medicare
20 population is only the elderly.
21 DR. KANG: Yeah. I would actually
22 agree with that. I think we need some minor
23 tweaks here and more along the line of tone or
24 clarification, and I don't think we're that far
25 apart.
.00133
1 Listening to the comments, I read this
2 document in a completely different way than many
3 of the commenters are reading it, and that really
4 suggests that we have somewhat of a problem.
5 The first is I did not read in this
6 document that there's an implication that
7 everyone has to have a randomized controlled
8 trial. What this document in my mind says is
9 that's the gold standard, but to the extent that
10 you deviate from the gold standard, you have to
11 explain biases, how you dealt with it et cetera.
12 So clearly a case controlled trial
13 where the biases let's say against device or
14 service or whatever, someone can say well, that's
15 okay. All the biases are against it. That's a
16 good trial.
17 The second observation I had was the
18 same as Dr. Francis', and this really actually
19 dealt with, I think, the Medicare beneficiary
20 rights testimony and a couple of other
21 testimonies. I think we do have to clarify that
22 the results associated with the study population
23 are the results associated with the study
24 population. Now, it so happens that the study
25 population excluded people under the age of 65,
.00134
1 and if you want to broaden that coverage, you
2 actually have to deal with whether you can get
3 there or not.
4 As it turns out, as the doctor with
5 multiple myeloma from Arkansas was saying, if in
6 fact the study didn't have age exclusion but
7 actually had another exclusionary criteria, then
8 the age probably goes away. You just actually
9 write a coverage decision that had the
10 exclusionary criteria.
11 The whole point, though, is you look at
12 the study population, and you agree with the
13 results. And then to the extent that you want to
14 cover beyond the study population, you actually
15 have to justify why it had reason to do that and
16 explain why that's an okay thing to do.
17 So I would actually see that those two
18 minor tweaks -- and maybe they're not minor, but
19 I think what Bob is suggesting is they still
20 require a fair amount of wording, but I think
21 that gets to most of the problems that have
22 actually been identified by the presenters that
23 there are some process problems.
24 DR. DAVIS: Well, I agree with a lot of
25 the comments that have been made. And to pull
.00135
1 them together, what I would like to see is I
2 agree with Leslie that a section-by-section
3 review would be appropriate today. We're not
4 going to do all the things that need to be done
5 to the document, but we can do a lot to fix
6 this. So I think a section-by-section review
7 would be good, and then by the end of the day
8 approve it with the fixes that the committee
9 agrees to, and then approve it as work in
10 progress, then give it to the panels as a
11 framework to guide their work in the coming
12 months, and then continue to come back to the
13 document and refine it as necessary, especially
14 considering that when panels begin to use it,
15 that will represent a pilot test, if you will, of
16 how appropriate and practical the document is,
17 but again coming back to it over time refining it
18 as necessary. And also, I'm sure we'll want to
19 take into consideration more detailed comments
20 from the public and from various stakeholders.
21 DR. SOX: Ron, maybe you could also
22 speak briefly to the concept Bob has advanced
23 about getting public input to this document. To
24 me it's kind of an attractive idea that we would
25 really seek broad input. We would have to make
.00136
1 the final call on the wording, but it would give
2 us an opportunity to make some changes in tone,
3 and if it seems appropriate to do so, that may be
4 very difficult to accomplish in the short-term.
5 What do you think of the overall
6 strategy of getting public input?
7 DR. DAVIS: Well, we've obviously had
8 some already today, we had some before we came
9 here today, and we'll have more later on this
10 afternoon. So my sense is let's try and improve
11 it today. Maybe we can go section by section and
12 allow people to propose improvements, and maybe
13 those can be approved as we go along by the
14 committee or disapproved, then hear some more
15 public comment from 3:15 to 3:30 or whenever that
16 happens as listed on the agenda, and then leave
17 the final approval by the committee to the end
18 of the day as the agenda indicates. Then there
19 will be more detailed commentary after we adjourn
20 today, and we'll take that into account when we
21 reconvene in a couple of months.
22 DR. SOX: Other comments about the
23 process? I would like to advance a notion and
24 see how it flies with you. I'm a little worried
25 that we're going to get into wordsmithing over
.00137
1 tone that's going to kind of bog us down and
2 would like to propose that we try to focus more
3 on technical content and less on tone during our
4 discussion, explicitly recognizing that we're
5 going to get a fair amount of public input
6 hopefully in writing, I would suggest, on how we
7 alter the tone in a useful way.
8 My guess is that as long as this
9 document continues to be an interim working
10 document in the next few months, these issues of
11 tone probably aren't central to getting on with
12 that work.
13 Does that feel pretty comfortable to
14 you all that we focus on technical content and
15 recognize we have a process for modifying the
16 tone in response to public comment both here and
17 that we may receive later on? Alan?
18 DR. GARBER: Well, I want to make sure
19 I understand the implications of what you're
20 proposing. I just know my panel, medical surgery
21 panel, is meeting in a little more than a month,
22 and I suspect that members of my panel won't care
23 much about the tone of the document and will care
24 a great deal about content. And if by technical
25 issues, you mean the content -- that is how are
.00138
1 you going to evaluate the evidence and so on --
2 that's great. That's what we need. And I agree
3 the wordsmithing about tone is not going to be
4 the number one concern of our panel.
5 So if we could end today with the
6 consensus about content as in what are the
7 specific directions that the panels will receive.
8 And let's not forget that although this is a
9 public document, its primary purpose is to guide
10 work for the panels. So that's really what we
11 should be focusing on.
12 If we can come to some consensus today,
13 that would be extremely helpful to us and I
14 suspect all the other panels.
15 DR. SOX: Bob, did you want to
16 comment?
17 DR. BROOK: From a process perspective,
18 I believe that the question we ought to ask the
19 committee, as a guide for the first panel
20 meetings, is there anything you find in the
21 document that's objectionable that would allow
22 you not to want to give this to the panel as
23 guidance for the first meeting?
24 If we limit ourselves to that question,
25 then I think we could do the task that people
.00139
1 have talked about, going section through
2 section. If we do anything else, I don't think
3 we're going to succeed.
4 I think that, however, this is
5 basically not a technical document, but a
6 political document written by a technical group,
7 and I would urge that we view it as such and
8 therefore insist that before we finally approve
9 the document, I think we can say to the panels
10 use it as a guidance for the first thing, that we
11 get absolutely specific written comments from
12 anyone in the public who wants to give it to us
13 with a justification for what they're trying to
14 achieve by that comment so that we can explicitly
15 respond in writing, do the same thing we're
16 asking the panel to do, to explicitly respond in
17 writing why we believe that this word ought to
18 stay the same, this word ought to change or that
19 we consider this other thing, and then do this as
20 an evolutionary process.
21 So my concern is do we have enough
22 discipline to hold ourselves for this
23 conversation around the table to say what's in
24 here that really the chair should not use at the
25 first set of panel meetings, not what you think
.00140
1 about the tone and structure and everything, what
2 we think this eventual document will look like?
3 DR. SOX: So it's partly objectionable,
4 but it's also unclear and confusing. I mean if
5 you don't understand the document, you can't
6 instruct the panel about problems. We've got to
7 deal with those problems as well. Okay. I think
8 we're all together. Bob?
9 DR. MURRAY: I'd like to comment that I
10 think it's inevitable that this is a guidance
11 that is titled recommendations. It's filled with
12 words like should, it's expected to, would
13 normally. It's only a guideline. It's not a
14 prescriptive legal statute.
15 Secondly, it's inevitable that it's
16 going to be treated as such because we have only
17 a month or six weeks before the next panel
18 meeting, and one of the provisions calls for a
19 six-month or anticipates a six-month time line in
20 order to get to the panel meeting. Well, of
21 course, you're not going to squeeze six months'
22 work into six weeks.
23 My feeling is that we should approve it
24 as is or with minor modifications because it's a
25 guideline. It's a recommendation.
.00141
1 DR. SOX: I think we're all clear. My
2 suggestion is that we take it section by section
3 and we take a few minutes before starting the
4 discussion for people to go back over and if they
5 haven't already identified concerns, to do so.
6 I'm not sure everybody has a comment.
7 Have most people already marked it up?
8 Great. In that case we can go right into it.
9 DR. HOLOHAN: Since we're switching our
10 agenda a little bit, we're going to ask questions
11 or make comments on some of the public
12 statements, there are a couple of things I'd like
13 to comment on before we start just to get them in
14 the public record. The written comments that
15 were supplied are, I presume, in the public
16 record, and I think a few things have to be
17 clarified.
18 One is HIMA has a statement that says
19 the six months that are suggested in the document
20 is the length of the life cycle of some
21 technologies. I find that very difficult to
22 believe. So it doesn't square with Mr. Roe's
23 interest in people investing money into a --
24 stent versus medical technology.
25 Secondly, there's a HIMA statement that
.00142
1 says technologies have improved laparoscopic
2 cholecystectomy -- would have difficulty in
3 clearing the evidentiary hurdle. Laparoscopic
4 cholecystectomy was actually decided as a
5 coverage issue by Medicare on the basis of the
6 request for review by the U.S. Public Health
7 Service. Their standard, arguably lengthy
8 procedure, that was extant in the early 1990s,
9 and HCFA was able to make a coverage decision in
10 a period of four months. So it's in the public
11 record, but it's not entirely true.
12 The only other comment I'd like to
13 make, Ms. Gottlich mentioned again VA coverage.
14 I'm perhaps oversensitized to this because it
15 came up four times at our panel discussion on
16 treatment of multiple myeloma.
17 I think, as the only VA representative
18 here, it's inappropriate to make comparisons
19 between benefits provided by Veterans Health
20 Administration and benefits provided by Medicare
21 for two reasons. The major one is that HCFA's
22 statutory requirements and the VA's statutory
23 requirements are considerably different. The
24 Veterans Administration is required by law to
25 provide clinical care to patients to do research,
.00143
1 to provide medical education to medical students
2 and house officers and to act as a backup for the
3 Department of Defense, and I think it is
4 misleading to see VA provision of medical care as
5 some kind of a federal imprimatur about safety
6 and effectiveness in part because of the fact
7 that research is part and parcel of what VA
8 does.
9 The second is that the VA benefits
10 package extends far beyond medical care to things
11 that HCFA doesn't cover, for example,
12 modification of vehicles for patients with spinal
13 cord injury, modification of homes, a much more
14 expansive long-term care program. So I think
15 it's simple to say well, since the VA does
16 provide high-dose chemotherapy and stem cell
17 support for some patients with multiple myeloma,
18 that it's ipso facto or important to VA for the
19 safe and effective therapy, and Medicare, as
20 another federal program, should follow suit.
21 It's deceptively simple, but it's in fact not the
22 case.
23 DR. SOX: Let's begin. Let me suggest
24 some ground rules that you want comments on
25 elements of the text that seem objectionable as a
.00144
1 basis for your panel proceeding or the text is so
2 unclear that you feel that you can't proceed, it
3 doesn't give you instructions you can understand.
4 I'd like to suggest that people who
5 have a problem with it try to identify the
6 problem, if possible propose a solution, and the
7 process for getting agreement is going to be
8 mostly me looking around the room and seeing nods
9 or asking if there's objections. Try not to take
10 votes unless we go into something that's real
11 controversial.
12 DR. DAVIS: Hal, can I ask a process
13 question?
14 DR. SOX: Go ahead.
15 DR. DAVIS: I think what you've just
16 outlined is fine, but I wonder if we go through
17 it section by section and stick to the issues
18 that you mentioned a few moments ago, and if we
19 have time perhaps we can go back section by
20 section and address tone again if there's time.
21 Would that fit in with what you're
22 trying to do?
23 DR. SOX: I agree with separating the
24 two, and if we have time, it would be reasonable
25 to address tone. I'm mindful of the fact that
.00145
1 there may be a few members who are going to have
2 to leave a little early. So I'm hoping we can
3 get done a little bit before it was scheduled for
4 the end of the meeting so we have everybody here
5 at the end. So I qualify it I guess.
6 MS. RICHNER: On that note I was
7 wondering if it's possible to do process first.
8 I think that's a critical component of what our
9 mandate is here. A lot of this is so theoretical
10 in the sense that we may get bogged down, and I'm
11 very concerned that one of the huge issues is the
12 evidentiary reports, and that whole section is
13 very unclear, and I would love to be able to
14 focus on that first.
15 DR. SOX: How do other people feel
16 about that?
17 DR. GARBER: I guess although I think
18 it's very important to get there, I think we
19 should proceed in order. I think that there are
20 two big issues that were raised overall, if I
21 could summarize what the commentators said in the
22 public testimony.
23 One of them had to do with the
24 impression some had that -- trials would be
25 necessary, and the other issue was timeliness.
.00146
1 So the first is in the first part of the
2 document, and the second is in the process part
3 of the document. I think we need to get through
4 both, so that will be the responsibility of Hal
5 to get us through this in a timely manner.
6 DR. SOX: Responsibility on all of us.
7 Jeff?
8 DR. KANG: Mr. Chairman, if I could
9 just add, as Dr. Hill was suggesting, the process
10 in many ways, a lot of the timing is HCFA's
11 responsibility, and we really have to work out
12 the logistics et cetera. And during the
13 presentation this is the first time I saw the
14 time frame, and I quite frankly think we can do
15 much better. So to the extent that we don't get
16 there, I really just wanted to signal that we
17 will very work very aggressively with the MCAC to
18 speed up the time frames et cetera.
19 MS. RICHNER: Preparation of the
20 evidentiary reports was another issue as well as
21 the reviewers.
22 DR. KANG: I think we can do that
23 faster. A lot of that responsibility, quite
24 frankly, falls to HCFA because it's staff
25 preparation. So I just want to send that message
.00147
1 loud and clear to the extent that we get bogged
2 down. I actually think we should get to the
3 content of guidance. And we are committed to
4 working on the process issue and getting things
5 done faster.
6 DR. SOX: I think we ought to focus on
7 issues that seem really important to the panel
8 chairs and co-chairs. So perhaps there won't be
9 any comments on the preface since it's not
10 procedural.
11 DR. BERGTHOLD: I would like to make a
12 suggestion that we consider what we heard from
13 the public today, which I thought was a very good
14 point, and that we put explicitly up front in the
15 preface, even though we all understand that, that
16 this is for the Medicare beneficiaries to better
17 serve them, so something like after the first
18 sentence, provide advice regarding coverage so
19 that Medicare beneficiaries can be better
20 served. I can't make a vote, but if someone else
21 would carry that vote.
22 DR. SOX: That's a tone thing.
23 DR. BERGTHOLD: I don't think it's a
24 tone thing. I thought about that really hard. I
25 think it's a substantive thing that we missed.
.00148
1 DR. SOX: Anybody have any problem with
2 now saying observing Medicare beneficiaries?
3 DR. FRANCIS: I'd like to add an
4 invitation to the panels -- this will be on the
5 last paragraph in the preface -- to convey back
6 to us concerns about the document as they work
7 with it.
8 MS. LAPPALAINEN: Just a matter of
9 helping our typist, when the committee makes a
10 suggestion to modify the document, you can then
11 ask yourself if it's all right. If then the
12 committee agrees that that change is fine, if the
13 person could then dictate slowly, and we can make
14 that change. We don't have to necessarily do a
15 vote for each individual change. We're hoping to
16 have the document modified and that at the end of
17 the day the entire document can be endorsed, if
18 you will. Thank you.
19 DR. FRANCIS: My suggestion might be
20 you just add the paragraph of the interim
21 document a work in process. We invite panel
22 comments about your impressions of the document
23 and what changes they might recommend to the
24 Executive Committee.
25 DR. SOX: Let's go down to the next to
.00149
1 last paragraph. So you want some wording that
2 might go on to have that paragraph, the last
3 sentence, continue to say and in response to
4 suggestions from the panel based on experience,
5 something like that?
6 DR. FRANCIS: Sure. The Executive
7 committee invites comments from the panels based
8 on their experience with this interim document.
9 DR. BROOK: Why don't we just say we
10 will modify these recommendations in response to
11 panel feedback and as needed to respond to the
12 HCFA final rule -- in response to feedback from
13 panel members or something like that. We will
14 modify these recommendations as reflected by
15 input from the panelists and as needed in
16 response from the panel members.
17 DR. FRANCIS: Alan, are you clear that
18 that's an open invitation to your panel to give
19 us feedback on how it will work?
20 DR. GARBER: Yes.
21 DR. SOX: Okay. Any other changes to
22 the preface? No objections? Okay.
23 Let's go on to Evaluation of Evidence.
24 I'd like to suggest we basically go through it
25 paragraph by paragraph so we're not jumping
.00150
1 around, and it will make it easier for the person
2 who's trying to make the changes in the permanent
3 record.
4 Any problems with the first paragraph?
5 The second paragraph?
6 DR. DAVIS: We're talking about
7 substantive process, right?
8 DR. SOX: We're talking about
9 objectionable for the basis of panel action or
10 unclear.
11 DR. DAVIS: Fine.
12 DR. SOX: So first paragraph? Second
13 paragraph? What about the statement in boldface
14 about the adequacy of the evidence, does that
15 tell you what you need to know?
16 DR. MURRAY: This is one of the few
17 places where the word must appears, and perhaps
18 this is tone, but in the prior paragraph the word
19 should is used.
20 Would this be inconsistent to change
21 must to should or must to is expected to? I'm
22 trying to address some of the concerns heard in
23 the comments that this is overly prescriptive.
24 DR. SOX: Anybody have any problem with
25 substituting should for must? Go ahead, Alan.
.00151
1 DR. GARBER: Well, I think this is the
2 sine qua non of what panels do. Details are
3 shoulds, but I can't see how a panel will
4 discharge its duty if it does not determine
5 whether the scientific evidence is adequate. So
6 this is one place where I feel the word must is
7 used advisably.
8 DR. MURRAY: We must use must? I
9 really don't have any objection to that.
10 DR. SOX: Any problem with using must
11 here? Other comments on adequacy of the
12 evidence? John?
13 DR. FERGUSON: Just a comment, and that
14 is that it was my understanding that HCFA
15 wouldn't send anything to the MCAC panels unless
16 they had some pretty good indication that there
17 was enough evidence. Now, that doesn't abrogate
18 the panel's responsibility for judging it, but I
19 think HCFA has said in their previous generation
20 that they would not send things to the panel
21 unless there was some clear evidence base.
22 DR. SOX: Do you have a wording change
23 suggestion?
24 DR. FERGUSON: I would say probably in
25 the paragraph before, the quality of the evidence
.00152
1 from these sources will vary, and the panels
2 should weigh the evidence according to its
3 quality, a portion of that weighing has been done
4 by HCFA prior to sending the request to the
5 panels or something like that.
6 DR. BROOK: Can we stay away from
7 that? We don't know how HCFA will want to use
8 this process in the future. Why don't we just
9 write a document on what the panel should do, and
10 HCFA can determine what it will do.
11 DR. KANG: I think that's correct. You
12 can't presume what will happen here.
13 DR. SOX: That process isn't written
14 down.
15 DR. KANG: Quite frankly, I think that
16 the, quote, slam dunks, we'll just deal with
17 administratively. And the reality is that on
18 your broad shoulders we'll be getting the plain
19 ones that are somewhat controversial, so I think
20 that we have to be very careful there. I would
21 just encourage you to just go ahead and do what
22 you think is right.
23 DR. SOX: Anybody here who doesn't find
24 Alan and Jeff's point compelling?
25 Other comments on the boldfaced
.00153
1 adequacy of evidence? Any specific wording
2 changes? I don't hear them.
3 So let's move on to the first paragraph
4 under comment. I'm just going to expect you to
5 holler.
6 Let's go on to the second paragraph,
7 the one that says many forms of evidence.
8 Third paragraph, when several such
9 well-designed trials, any changes to this?
10 How about the next one, the Executive
11 Committee believes? Jeff?
12 DR. KANG: I hate to say that this is a
13 tone also, but we say here in considering the
14 evidence from any study, whether they're
15 randomized clinical controlled trials or any
16 other trials or whatever, you could say the MCAC
17 now should try to answer these two main
18 questions.
19 DR. DAVIS: Where are you?
20 DR. GARBER: It's the last paragraph
21 before bias. You want to insert whether
22 randomized controlled clinical trial or
23 observational study?
24 DR. KANG: Or other controlled trials.
25 DR. GARBER: Or other controlled study?
.00154
1 DR. KANG: Yeah.
2 DR. SOX: So it's really any controlled
3 study. It wouldn't apply to a noncontrolled
4 study.
5 DR. KANG: Right. Any controlled study
6 including randomized controlled trials because
7 you do want to deal with bias, and even in an RTC
8 it's possible.
9 DR. SOX: So the suggested wording is
10 that after any, we would put any controlled
11 study, including randomized controlled trials.
12 MS. RICHNER: What about the issue of
13 registries again? I think that limits this.
14 DR. SOX: We speak later on to the
15 issue of registries without any form of control.
16 DR. GARBER: Well, there are some
17 changes we might want to make later on, but I
18 think we have to make it clear that registries
19 can be controlled, and they can be uncontrolled,
20 and I have some suggested wording later.
21 MS. RICHNER: But this wouldn't then
22 negate evaluation of that type of evidence later
23 on?
24 DR. GARBER: Right.
25 DR. SOX: If there was a control, then
.00155
1 it would fall into this.
2 MS. RICHNER: Okay. I see what you're
3 saying.
4 DR. BROOK: Jeff, just to be clear,
5 you've made this more limiting than it was
6 before. The purpose by inserting all that
7 nonsense, the purpose of this sentence, was
8 basically to say this is not a rigid
9 restriction. This is a general. And now by
10 stating controlled trials in it, you've made it
11 much more rigid. Study is very vague.
12 DR. KANG: I agreed with that point,
13 but I was surprised by the comments that we were
14 getting.
15 DR. SOX: Actually I think there's a
16 logical reason for sticking it in there because
17 the bias to controlled group and intervention
18 group doesn't apply to a noncontrolled study. So
19 in other words, the remark about bias isn't an
20 issue unless you're comparing groups. So I think
21 it makes much more sense.
22 DR. MAVES: Hal, that may be true, but
23 I again like the way it was worded beforehand
24 because it was more open, and it was broader and
25 less sort of proscriptive. Unless Jeff has a
.00156
1 good reason for putting it in there.
2 DR. BROOK: What about this? In
3 considering the evidence from any study, whether
4 randomized or not, the MCAC should try to answer
5 these two main questions. There can be bias in a
6 randomized trial study. So why don't we say
7 considering the evidence from any study, whether
8 randomized or not.
9 DR. KANG: That's fine.
10 MS. RICHNER: Thank you. That's
11 better.
12 DR. SOX: Is that compromise agreeable
13 with everybody? Okay. Any other comments on
14 that paragraph?
15 How about the next paragraph, the one
16 that defines effectively bias? Then we have a
17 real long paragraph coming up, many opportunities
18 for finding fault here. Anybody want to make
19 suggestions about how to change this next
20 paragraph on potential sources of bias?
21 DR. HOLOHAN: The investigators cannot
22 be sure that they have measured all of the ways
23 in which treated patients differ from untreated,
24 do you really want to put in the word measure?
25 DR. SOX: Can you tell us where that
.00157
1 is, please.
2 DR. HOLOHAN: The fourth line down.
3 It's talking about observational studies. The
4 investigators can't be sure that they have
5 measured all the ways --
6 DR. BROOK: Are you saying measure to
7 assess?
8 DR. HOLOHAN: Measured implies a
9 quantitative evaluation which may not be possible
10 in many instances.
11 DR. MAVES: How about considered?
12 DR. SOX: Alan?
13 DR. GARBER: The operational issue here
14 is has it been recorded in some way that it can
15 be incorporated into a study design? And to
16 observe is not sufficient. To consider is not
17 sufficient. It has to be recorded. Measure does
18 not necessarily mean quantified in continuous
19 terms. It can mean it's a binary variable.
20 Doesn't necessarily mean quantitative. Measured
21 means observed and recorded.
22 DR. HOLOHAN: Why don't we just say
23 observed and recorded.
24 DR. GARBER: Well, fine. I wouldn't
25 have any objection to that.
.00158
1 DR. BROOK: That sounds fine. Observed
2 and recorded.
3 DR. SOX: Great. Other comments on
4 this paragraph?
5 Now we turn to the one paragraph that
6 starts random allocation of patients. Any
7 objections to this paragraph for lack of clarity?
8 Then let's go on to the next paragraph,
9 in an observational, nonrandomized study.
10 Remember now we've got to focus on issues that
11 are objectionable for the basis of panel action
12 or unclear. Ron?
13 DR. DAVIS: I guess some of these
14 comments could address interpretation by panels,
15 so maybe I'll offer this comment which could be
16 tone, could be interpretation.
17 At the very end where we say clinical
18 trials of treatments for cancers that have an
19 unpredictable natural history, for example, have
20 repeatedly demonstrated that the results of
21 observational studies are misleading, I wonder if
22 we should say are often misleading.
23 DR. SOX: Yeah. They aren't always.
24 Fair?
25 DR. BROOK: It's not that they're
.00159
1 misleading. They're overly optimistic of the
2 value of the therapy.
3 DR. SOX: How about frequently
4 overestimate the size of the treatment effect?
5 DR. BROOK: That would be better.
6 DR. SOX: The results of observational
7 studies frequently overestimate the size of the
8 treatment effect, and delete often misleading,
9 and go back to the --
10 DR. BROOK: Remove repeatedly at the
11 first part of that sentence.
12 DR. SOX: One more wordsmithing change
13 in that sentence, repeatedly on the left hand
14 side, delete that. Okay. Good.
15 Next paragraph, to detect important
16 bias. This one has a lot of operational
17 implications. Does it really do it for you?
18 Okay.
19 Next paragraph, although a body of
20 evidence.
21 DR. HOLOHAN: Can I suggest that the
22 phrase is never adequate be clarified a little
23 bit? And I think what was meant by the
24 subcommittee was that it would never reach to the
25 reliability of a probably done randomized
.00160
1 controlled trial, but not that it is ipso facto
2 inadequate.
3 DR. SOX: Alan, do you want to respond?
4 DR. GARBER: Well, I realize this is
5 not a flash point, and I think we should be --
6 the issue here that is I believe perhaps a
7 semantic one -- I'm not certain -- and that is
8 what do we mean by uncontrolled? And from
9 hearing the comments today, I think that some of
10 the people may have been under the impression
11 that what was meant by uncontrolled is not
12 randomized controlled, and that's not the case.
13 And I actually got some suggested
14 rewording, and I don't know if this will do it.
15 And Tom, I particularly appreciate your opinion
16 about this. That is the first sentence of the
17 paragraph would begin although they do not have
18 randomized controls, all well-designed
19 observational studies include some form of
20 control. They may consist of an implicit or
21 explicit controlled group or statistical
22 controls, that body of evidence consisting only
23 of uncontrolled studies. And I think that's
24 intended to make it clear that registries are
25 probably assigned observational analyses,
.00161
1 probably assigned controls, and the issue truly
2 uncontrolled study, I think it's strictly true.
3 If it is uncontrolled, it is not valid evidence
4 by itself, yet there are plenty of studies that
5 could have valid controls that are not
6 randomized, and I would hate for the readers of
7 this document to think that this paragraphs means
8 you have to have randomized controlled trials.
9 In fact, I was struck that some of the
10 public comments seem to suggest that this
11 document meant only randomized controls would be
12 suitable. We put a great deal of effort on the
13 part of the subcommittee to try to make it clear
14 that observational data would often be -- well,
15 at least would sometimes be adequate, and it
16 really depends on the characteristics of the
17 studies that were being done.
18 MS. RICHNER: I still think that's
19 missing the mark in a sense because I think why
20 this is so controversial in a sense is that once
21 again when you're looking at the technology curve
22 when you have very little evidence in the very
23 beginning of adoption, it's rare that you're
24 going to have the kind of rigorous studies that
25 you're interested in. So I think what this does
.00162
1 is we want to make sure that you're looking at
2 the composite of all possible data that's
3 available. And this doesn't allow that.
4 Essentially looking at perhaps unpublished data
5 that might be available that would be
6 interesting, case studies, et cetera, et cetera,
7 and somehow this tone of this paragraph limits
8 all of that.
9 DR. SOX: We've got to have something
10 to vote on and some wording to vote on.
11 MS. RICHNER: And unfortunately I had
12 wording that I sent to you that I thought was
13 appropriate on e-mail that would have addressed
14 that as well. Unfortunately my computer has now
15 just died.
16 DR. BROOK: Can I suggest some
17 wording? I want to suggest an alternative
18 wording before we vote.
19 DR. SOX: I'm thinking that maybe what
20 we need to do is to get -- this is a really an
21 important issue, and that perhaps an approach
22 would be that we delay the vote on this. We can
23 move on without this. Each of you submit your
24 wording that we get it up there and we actually
25 wordsmith out.
.00163
1 DR. BROOK: Can I suggest an approach
2 to this background before we do that? I would
3 like to suggest that we're limiting everything up
4 to in some cases, and we start by saying in most
5 cases given the current state of scientific
6 evidence, panels will determine that well-
7 collected observational evidence -- and then I
8 think we ought to list in there what we mean by
9 that -- will be sufficient to draw conclusions
10 about effectiveness, and I think that that's the
11 tone you want in this paragraph.
12 MS. RICHNER: Yes, that's much better.
13 DR. BROOK: Because with a large part
14 of the technologies, that's what's going to
15 happen. So that's how I would alter that
16 paragraph. And I would then spell out in detail
17 what we think are well-controlled observational
18 kinds of studies, registries with historical
19 controls, quasi experimental designs, et cetera,
20 et cetera. And I think I'd even add the point
21 that Jeff came up with. This would be especially
22 true when we have breakthrough technologies and
23 technologies dealing with people with severe
24 diseases with no other recourse.
25 DR. KANG: That's good.
.00164
1 DR. BROOK: I think that's what the
2 panels are going to do, and I think we might want
3 to say it.
4 DR. KANG: May I make a suggestion
5 since we're almost at lunch? I don't think we're
6 that far apart. It actually strikes me that
7 maybe Bob, Alan and Randel sit down at lunch and
8 hack it out. I hate to infringe on your lunch
9 period.
10 DR. SOX: I think that's actually a
11 very good suggestion. We'll appoint a committee
12 of three, and if any member of that committee is
13 not satisfied with what you come up with, then
14 that person will submit an alternative, and we
15 can vote on it. Does that sound reasonable?
16 We've got about five minutes to 12:00. Should we
17 give ourselves a break at this point? And we'll
18 come back at 1:00 and continue the process.
19 (Whereupon, recess taken -- 11:55 a.m.)
20 (Whereupon, after recess -- 1:10 p.m.)
21 DR. SOX: Alan, do you have a report
22 of the work group of the subcommittee?
23 DR. GARBER: We weren't able to locate
24 one of the members of our subcommittee. Randel
25 and I went over some language that I think we
.00165
1 agree on. So if I could read that to the
2 committee and the audience.
3 DR. SOX: Should we perhaps have it --
4 DR. GARBER: Let me read it once first
5 because there's a lot of changes. Okay. This
6 refers to the bottom of that page. It's right
7 above the subheading external validity, the last
8 paragraph, and it currently starts although a
9 body of evidence.
10 The new language is as follows.
11 Although if they do not have randomized controls,
12 all well-designed observational studies include
13 some form of control. Controls may consist of an
14 implicit or explicit controlled group or
15 statistical controls. A body of evidence
16 consisting solely of studies with no controls
17 whatsoever, whether based on anecdotal evidence,
18 testimonies or case series, is never adequate.
19 And then the last sentence reads, now that
20 there's a change in the last part, when these
21 circumstances apply, the panel must describe
22 possible sources of bias and explain the basis
23 for its decision that bias does not account for
24 the results.
25 Randel, does that reflect what we
.00166
1 said?
2 MS. RICHNER: Yeah. The key issue here
3 is that any of the case series studies or
4 composite of any of those sort of testimonials,
5 anecdotal studies combined, can never constitute
6 the proper evidence if it's only those types of
7 studies.
8 DR. GARBER: Only studies without
9 controls.
10 MS. RICHNER: Right. Without some type
11 of control. So even in an observational study,
12 you can use a statistical methodology in which to
13 observe or have a control as part of that. And
14 that works. What do you think, Bob?
15 DR. BROOK: My fault. I didn't go to
16 lunch, so I couldn't find you guys. So my
17 fault.
18 DR. FERGUSON: Can that be written down
19 and circulated?
20 DR. GARBER: I just wanted to get it
21 done in general first.
22 DR. BROOK: In general terms I don't
23 believe a document ought to ever use the word
24 never.
25 MS. RICHNER: Then never is a problem.
.00167
1 I still don't like the never.
2 DR. BROOK: There is not a single
3 testimonial that couldn't be put into historical
4 context by some historian. Whether you choose to
5 do it or not makes it adequate or inadequate, but
6 there is no case series that could not be put in
7 some historical context no matter how bad. And
8 the panels are going to be left to judge how much
9 effort and how good these controlled efforts have
10 been. That's why I would have simplified this
11 just to say -- I mean that's their job in terms
12 of what's going on. That's okay. It's my fault,
13 as I said, for not being there.
14 DR. SOX: Okay. Alan, do you want to
15 read that one more time? Then we can have
16 discussion of it and maybe start to get it on the
17 document as well.
18 DR. GARBER: Should I read this line up
19 to it? Insert at the beginning of the paragraph
20 the following.
21 DR. BERGTHOLD: No. She's just going
22 to type it separately for now.
23 DR. GARBER: Oh, okay. Fine. Although
24 they do not have randomized controls, all well-
25 designed observational studies include some form
.00168
1 of control. Controls may consist of an implicit
2 or explicit controlled group or statistical
3 controls. And then the next up is -- do you want
4 to just retype the remainder of the paragraph?
5 THE TYPIST: Would that be here at the
6 end?
7 DR. GARBER: It goes to the although.
8 It's now the next sentence. The word although is
9 struck and then a body of evidence. So you
10 struck that. The body of evidence consisting
11 solely, and then strike only, and then strike
12 uncontrolled. And then after studies insert with
13 no controls whatsoever. And then after case
14 series strike and disease registries without
15 adequate historical controls. Then it stays the
16 same is never adequate. And then insert however
17 before in. This is something I didn't mention
18 that we changed also. Strike some and replace it
19 with many. In many cases. Then it goes to the
20 last part of the paragraph. Strike why it
21 decided and insert the basis for its decision.
22 MS. RICHNER: Bob, you certainly still
23 have a chance to comment.
24 DR. SOX: Well, it's time for comments
25 or questions. Actually I have a question.
.00169
1 Statistical controls, could you explain what that
2 means?
3 DR. GARBER: In other words, it's an
4 observational study where they can collect data
5 on a number of variables and basically look at
6 patterns of outcomes, how they're explained by
7 things like say age et cetera. That can be a
8 form of statistical control.
9 DR. SOX: Is that multivariant analysis
10 essentially?
11 DR. GARBER: Yes.
12 DR. KANG: This is different or the
13 same? You do multivariant plus sensitivity
14 analysis?
15 MS. RICHNER: I actually have some
16 literature that is very recent from the
17 pharmaceutical industry of which they do this
18 type of methodology. And once again, I can't
19 articulate it well, but there are methods to do
20 this in using observational data that is well-
21 grounded. I mean McMasters has done a lot of
22 work at that.
23 DR. KANG: Could you take another
24 attempt at trying to explain to me?
25 DR. GARBER: Let me tell you about some
.00170
1 of the work we've done using Medicare claims
2 files. Let's say that you want to have an idea
3 of whether revascularization in post MI improves
4 outcomes. You can take Medicare claims files
5 which have extensive information about discharged
6 diagnoses, age, location and a number of other
7 individual characteristics, and there are various
8 statistical methods you can use to determine
9 whether the people who have treated with
10 revascularization did better. So you'll have Bob
11 Brook saying that's all very hokey, but that's
12 what statistical controls are, and the panels
13 have to decide whether this type of evidence is
14 adequate or not.
15 DR. HOLOHAN: It's retrospective.
16 DR. GARBER: Well, it's actually
17 historical prospective. The point is we're not
18 going to determine right now whether any
19 particular study in science is adequate. The
20 point is that there are methods, and there are
21 cases where you can use that kind of a controlled
22 group -- that is implicit statistical control --
23 to draw conclusions. The panels may decide yes,
24 this is convincing or they may decide it's not on
25 a case-by-case basis.
.00171
1 DR. SOX: Any other questions or
2 comments about this? Ron?
3 DR. DAVIS: Well, I like it. I just
4 wanted to suggest one other small change at the
5 end. Instead of saying that bias does not
6 account for the results, to say that bias is
7 unlikely to account for the results. I think the
8 panel would more likely say we don't think bias
9 accounts for the results. I don't think they'd
10 say bias does not account for the results.
11 DR. SOX: Does that sound reasonable to
12 you guys?
13 MS. RICHNER: We had that discussion as
14 well. Are you comfortable with that?
15 DR. GARBER: Yeah, I think that's
16 fine.
17 DR. SOX: Any other comments? So it
18 goes. We now go on to external validity, first
19 paragraph.
20 DR. FRANCIS: There's a replacement
21 effort.
22 DR. KANG: If you don't mind, Dr.
23 Francis and I, in going through it ourselves as a
24 group of two, took another crack at this. So
25 this is under external validity. And maybe we'll
.00172
1 read it.
2 DR. SOX: Is this suggested as a
3 substitute for the paragraph?
4 DR. FRANCIS: Yeah. For the first
5 paragraph.
6 MS. LAPPALAINEN: I'll read it out
7 loud. Issues of external validity related to the
8 study of population. Medicare beneficiaries
9 include elderly, nonelderly, and disabled
10 people. The Medicare population also may or may
11 not include patients with comorbid disease. That
12 said, historically many controlled trials
13 unfortunately excluded older men and women,
14 people with disabilities and people with comorbid
15 disease. This means that even when a trial has
16 adequate statistical power for the study
17 population, that its results may or may not be
18 generalizable to some portions or all of the
19 Medicare population. If the requester is asking
20 for, or the panel is advising, coverage beyond
21 the clinical and demographic characteristics of
22 the study population, the panel should state that
23 they believe the results of the trials are
24 applicable to a broader population, define what
25 that population is and explain its reasoning
.00173
1 why.
2 DR. SOX: So Leslie and Jeff, perhaps
3 you could explain what lead you to make this
4 change so we all understand what's behind it.
5 DR. FRANCIS: One thing that was behind
6 it was the recognition that Medicare population
7 is not just the elderly. And at least the way
8 the myeloma panel was set up, the question that
9 was posed to the panel was we've got a lot of
10 data in there under 65s. Can we extrapolate from
11 65s and over? And we wanted to take away any
12 implication that that's the way stuff should be
13 set up rather than focus on the question of what
14 were the inclusion and exclusion criteria in
15 studies and what that says about what are all
16 portions of the management population coverage
17 recommendations we are aiming for. So that's
18 what we're trying, however inartfully, to
19 capture.
20 DR. KANG: Part of the problem with the
21 tone of this paragraph is it assumes that all
22 Medicare coverage decisions are for the general
23 population. We are now -- practically all of our
24 coverage decisions are limited in some way, have
25 exclusion criteria or inclusion criteria, and a
.00174
1 lot of times we do it for the study population.
2 That is something, quite frankly, that's been
3 new.
4 So I really think the issue here is is
5 it a statistically valid study population -- then
6 a request is for that study population. And we
7 should cover for that study population. And if
8 it so happens we only have three beneficiaries,
9 that's okay. It's still covered for those three
10 beneficiaries. That's more or less what we were
11 trying to get to.
12 DR. GARBER: Well, Jeff, I guess you
13 correctly guess that my concern is the last part
14 of this.
15 DR. KANG: That's correct.
16 DR. GARBER: And the problem is
17 probably semantic, but as I read this revision,
18 it could be applicable to a broader population,
19 but it doesn't necessarily mean it could pass
20 that criterion and still not necessarily be
21 applicable to any defined population of Medicare
22 beneficiaries. So the original language -- I
23 mean I completely agree with the intent of this
24 and with the rest of it, but the original
25 language, just to remind people, is if the study
.00175
1 population in the available trials is not the
2 same as the general population of Medicare
3 beneficiaries who would be candidates to receive
4 the intervention, the panel must state whether
5 the results of the trials apply to typical
6 Medicare patients and explain its reasoning.
7 And that language was really saying
8 does this generalize to the relevant population
9 of beneficiaries? And I'm not sure the language
10 that you proposed at the end actually gets at
11 that. So I would propose something like an
12 amendment to the original language for the last
13 part, and instead of saying typical Medicare
14 patients, maybe two defined populations of
15 Medicare beneficiaries so you cover ESRD,
16 disabled et cetera.
17 DR. BROOK: Can I suggest changing
18 broader population to the results of the trial
19 applicable to any group of patients covered by
20 Medicare? So that would then allow you total
21 flexibility since we're writing this for
22 Medicare.
23 MS. RICHNER: Results in the study too
24 rather than trials.
25 DR. BROOK: The results of the trials
.00176
1 are applicable to any population covered by
2 Medicare or can be applied to any population
3 covered by Medicare. Define what the Medicare
4 population is and explain its reasonings why or
5 what part of the Medicare population it applies
6 to and explain its reasonings why.
7 DR. KANG: I'm not sure that gets it.
8 I'm okay with it.
9 DR. GARBER: I like my wording better,
10 which is defined populations of Medicare
11 beneficiaries so you can say this is effective
12 for ESRD beneficiaries, and this is effective for
13 elderly Medicare beneficiaries, and this is for
14 the disabled. But the point is that the panel
15 should explicitly say which population of
16 beneficiaries if any they believe the results of
17 these trials apply to.
18 DR. SOX: Alan, are you proposing we go
19 back to the wording of that last sentence?
20 DR. KANG: Alan, I'm not sure I
21 understand that because we actually -- our
22 coverage decisions are now running like this is
23 effective for ESRD patients who don't have heart
24 failure or whatever it is.
25 DR. GARBER: That's what we're saying,
.00177
1 that the panel should say what the trials apply
2 to, some population like that. Now, you could
3 tell us look, we'll decide. We don't want the
4 panels to get in the business of determining
5 whether the trials apply to populations of
6 beneficiaries. I think you'd be better off using
7 panels to try and evaluate the evidence and see
8 whether they think they can extrapolate from the
9 trials to some population of interest to
10 Medicare.
11 DR. FRANCIS: Why don't we just change
12 the last sentence to say to populations or to
13 groups covered by Medicare, define what those
14 groups are, and explain the reason why.
15 DR. GARBER: Could you say the exact
16 words?
17 DR. FRANCIS: Believe the results of
18 the trials are applicable to some groups covered
19 by Medicare, define what those groups are and
20 explain its reasons why.
21 DR. BROOK: Define it in clinical terms
22 if you want to.
23 DR. GARBER: No. I think that's fine.
24 DR. BERGTHOLD: Does that allow
25 Medicare to make, sort of, fine, sort of,
.00178
1 distinctions within those populations though?
2 Because that almost sounds like if you're an ESRD
3 person, you get this treatment even if you do
4 have heart failure or whatever. No? That
5 doesn't mean that?
6 DR. BROOK: No.
7 DR. KANG: No.
8 MS. RICHNER: The other question I
9 would have here about define it in terms of just
10 trials, wouldn't you want to make it a little
11 broader in terms of studies? Because the whole
12 part before was describing we're going to be
13 looking at lots of different kinds of evidence,
14 so therefore we don't want to limit ourselves to
15 trials here.
16 DR. KANG: I was concerned this study
17 has to be statistically -- so you could say --
18 MS. RICHNER: Well, yes, but that's
19 covered in the part before.
20 DR. KANG: Okay.
21 DR. HILL: I don't think you meant,
22 Leslie, to say that if the requester is asking,
23 the panel should state. That first phrase is in
24 the alternative. You only mean if they agree.
25 DR. FRANCIS: Right.
.00179
1 DR. HILL: So you state whether or not
2 they believe.
3 DR. FRANCIS: Whether they believe.
4 DR. HILL: This way it's grammatically,
5 if the requester asks, that they are being
6 requested by you to state that they believe.
7 DR. FRANCIS: No. They should state
8 whether they believe.
9 DR. FERGUSON: I have a question. Is
10 it true that the sentence that says the study
11 population results may or may not be
12 generalized -- wait a minute. If the requester
13 is asking for, or the panel is advising,
14 coverage, is HCFA comfortable with our panel's
15 advising coverage? Are coverage questions going
16 to be asked specifically?
17 DR. HILL: We've answered that as we've
18 gone along and repeatedly said that we understand
19 we have the responsibility for deciding coverage.
20 So I take that to mean if you want to clean that
21 language up, I'd be grateful, but I don't want to
22 slow you down.
23 DR. FERGUSON: Safe and effective or
24 some other words.
25 DR. KANG: See, this is tough because
.00180
1 by our federal register notice we are asking the
2 requester to specify the population that they're
3 seeking coverage for. We get that with varying
4 degrees of success.
5 Maybe one of the ways we do that is to
6 clean that up and really demand, before it gets
7 to the panel, that they are very clear about what
8 population they're looking for. Then the panel's
9 decision is whether or not the evidence supports
10 that.
11 The only thing that we get into
12 somewhat of a problem is if it doesn't support
13 it, then there's the question of well, what would
14 it support?
15 DR. FERGUSON: But advising coverage
16 and advising that the evidence supports coverage
17 might be --
18 DR. HILL: May I suggest if the
19 requester is asking for coverage or the panel
20 concludes that medical benefit can be --
21 DR. SOX: I'd like to suggest -- I
22 think we know what we're going to say here.
23 Rather than try to wordsmith this thing in
24 detail, I'd like to suggest that we take it down
25 and somebody work on some language that doesn't
.00181
1 have us recommending coverage, but still allows
2 the requester to request coverage. I think we
3 know what we want to say.
4 DR. KANG: I think, John, advising
5 support for will be okay. Let's just get it over
6 with.
7 MS. RICHNER: And the other part about
8 trials versus studies.
9 DR. KANG: We took care of that.
10 DR. BERGTHOLD: It doesn't apply
11 above.
12 MS. RICHNER: That sentence,
13 historically many controlled trials
14 unfortunately --
15 DR. GARBER: Yeah. But that's true.
16 It's much more common trials and observational
17 studies to --
18 MS. RICHNER: Okay. I see what you're
19 saying.
20 DR. KANG: That's correct. That's the
21 ages within our society.
22 DR. SOX: I'd like to turn it over to a
23 wordsmith to clean it up a little bit and make
24 sure we're happy with the wording. Who would
25 like to volunteer to be the wordsmith? Ron?
.00182
1 DR. KANG: I want to make sure you're
2 okay with it. I don't think this violates your
3 original intent.
4 DR. GARBER: I think it's probably
5 fine. It's certainly not worth struggling over.
6 DR. SOX: Okay. Let's move on. We'll
7 give this to Ron, he'll work on it, and we'll
8 move on to issues of external validity also apply
9 to the intervention. Any objections or
10 clarifications required here?
11 MS. RICHNER: This paragraph we also
12 discussed at lunch briefly. One of the issues
13 here -- and I don't know if this example is the
14 appropriate example in here. I mean I guess we
15 can go ahead and use it, but I'm concerned about
16 the interpretation of this. Certainly, once
17 again, the technology, this skill of the surgeon
18 over time improves, and the outcomes associated
19 with time improve as well. But once again, this
20 is an example of external validity.
21 DR. SOX: Gets over the concept I think.
22 MS. RICHNER: Yeah, I think we're
23 okay.
24 DR. SOX: Any other questions about
25 this one?
.00183
1 DR. SMITH: I guess now that we have
2 somewhat talked about the elderly and nonelderly
3 and disabled, I guess my concern is I read where
4 you have like demographics. Have we lost or does
5 that encompass let's say racial and ethnic
6 inclusions or should there be, can there be, some
7 consideration given to that particular area?
8 DR. SOX: Are you talking about --
9 DR. KANG: She's talking about the
10 previous.
11 DR. SOX: -- the previous paragraph?
12 DR. SMITH: The previous one. I mean
13 it seems as if it's getting lost.
14 DR. KANG: Yeah. I think the reason
15 why -- and I'm not sure I'm aware of a trial with
16 racial exclusion, but I could be completely wrong
17 on this. But I would not have any problems, I
18 don't think, adding racial inclusion to the
19 extent that it occurs.
20 DR. SMITH: I thought about it. It may
21 even be something that could be stated in the
22 preface rather than just in one specific area,
23 and then that automatically would speak to it
24 with some consistency throughout the document.
25 DR. KANG: Actually this would be the
.00184
1 place to deal with it I think.
2 DR. SOX: We need specific wording
3 suggestions. Daisy, do you want to take a look
4 at this paragraph after Ron gets done with it and
5 suggest some language? Not all of us completely
6 understand.
7 DR. SMITH: So when you have concerns,
8 you just keep quiet, right?
9 DR. SOX: No. We need something to
10 look at so we know whether we like it or not.
11 DR. HOLOHAN: Just as an editorial
12 comment, the best example I can think of recently
13 of a trial that was dramatically racially
14 imbalanced are the studies of -- and hepatitis C
15 patients. The patients tested do not represent
16 the population of patients with hepatitic C in
17 the United States today.
18 DR. KANG: So then probably we should
19 add it along, and that would be the easiest way
20 to deal with it.
21 DR. GARBER: Just to make maybe a
22 substantive point because there will be a lot of
23 interested parties here, we don't intend to imply
24 that every study has to have adequate sample
25 sizes of various ethnic groups and so on to draw
.00185
1 conclusions. Just the panel needs to decide
2 whether they think the results of the studies
3 apply to those populations. We don't want to
4 send a message gee, you're going to have to have
5 an adequate number of Hispanics, adequate number
6 of Asian Americans and so on. That would be
7 impossible.
8 MS. RICHNER: As a matter of fact,
9 there's one more point I wanted to make about
10 this, and that's foreign data. I don't know how
11 you're going to address that, but certainly there
12 are many studies that are done outside the U.S.
13 And how does that apply to Medicare populations?
14 And in turn, we run across this all the time.
15 The FDA now accepts foreign data. So that is
16 going to be an issue associated with this as
17 well.
18 DR. KANG: By this language we're not
19 excluding foreign. This language says if it's
20 foreign, then say that I believe this is
21 generalizable to the American population for
22 these reasons.
23 MS. RICHNER: As long as we're talking
24 about methodology and study design, et cetera,
25 and evidence.
.00186
1 DR. HOLOHAN: The issue is can the
2 panels extrapolate?
3 DR. FRANCIS: One of the things that
4 was very striking about the myeloma discussions
5 was that although the incidence of the disease is
6 much higher in African-Americans, the actual
7 apparent access to the therapy in the testimony
8 of the patients, who were all white, there were
9 obvious issues of access that underlay the whole
10 discussion, and I wonder whether there's a way to
11 go back to the preface and put in something about
12 equity and the importance of equity in the
13 coverage process.
14 DR. SOX: Is that something that we
15 could deal with after today and still operate
16 as --
17 DR. KANG: We can.
18 DR. SOX: I want to move on now to Size
19 of Health Effect. Any problems with the way that
20 is stated?
21 DR. FRANCIS: I have a clarification
22 and a question. The clarification is I want to
23 be sure that category 2, more effective --
24 DR. SOX: You're getting ahead of us.
25 We're going paragraph by paragraph. First, just
.00187
1 the stuff that's in boldface. Any problems with
2 that? John?
3 DR. FERGUSON: Must we have must
4 instead of should?
5 DR. GARBER: Yeah. Because I think
6 we're saying there's going to be a standardized
7 way of reporting. Each panel reports the
8 evidence into these same set of seven categories.
9 And if there's any reason these seven categories
10 aren't right, we should probably change the
11 categories now rather than saying should.
12 MS. RICHNER: Well, there was a
13 suggestion by the audience for an additional
14 category that was from one of the letters. Not
15 only that, I remember in our conference call that
16 we had David Eddy suggested that there were
17 perhaps 15 different categories. So I think we
18 do have to think carefully.
19 DR. FERGUSON: I withdraw my comment
20 because I think what you're saying is the
21 comparison is the must, and that's clear.
22 DR. SOX: Okay. So we've dealt with
23 the stuff in boldface. Now let's go on to the
24 first part of the comment, just that first couple
25 of sentences. Then we'll go through the seven
.00188
1 categories. No problems? Then let's go to the
2 seven categories.
3 I'd like to suggest that modifying
4 these may be the sort of thing that we do after
5 we have a chance to use them a little bit, and we
6 may find that these categories need to be
7 expanded in order to deal with circumstances that
8 will come up only when we actually do a study and
9 try to classify its effect size and find we
10 really can't do it properly. It may work better
11 than trying to wordsmith these categories or at
12 least change significantly the categories right
13 now. John?
14 DR. FERGUSON: Just a comment. And I'm
15 sort of asking this. One of the advantages might
16 be cost, something would cost less. And maybe we
17 shouldn't put that in there, but it's certainly
18 something that I would hope we sometimes are
19 presented with as an advantage. Is that a no
20 no? Can we list that as an example?
21 DR. SOX: Basically it's a no no.
22 DR. KANG: For the time being.
23 DR. FRANCIS: Can I just ask you about
24 category 2? Does that include small benefits for
25 lots of people as well as relatively significant
.00189
1 benefits for small numbers, but we don't know how
2 to sort those out into identifiable subsets?
3 DR. SOX: Alan, do you want to respond?
4 DR. GARBER: The question comes down to
5 whether they are prospectively identifiable
6 categories of people who get substantial benefit.
7 If they are identifiable, I would have
8 interpreted this to mean they go in category 1
9 and category 2 for the other groups. And if they
10 aren't identifiable, it's irrelevant. There's
11 always some people who will benefit, but you
12 don't have any way to sort them. You just have
13 to go with the average benefit.
14 So the question is can you identify a
15 category with greater benefit? Obviously if you
16 give an intervention that's slightly better, what
17 that usually means is that there's some people
18 like you're measuring mortality, more people
19 live, but you don't know for sure who's who.
20 That's what subgroup analysis --
21 So the other just quick comment, the
22 ACP-ASIM talked about more objective, but some
23 disadvantages. I think that we discussed that in
24 the conference call, and that would have gone
25 into category 2. So what they're talking about
.00190
1 is subdividing category 2. And the subcommittee
2 was trying to get the smallest number of
3 categories that we thought would do a good job of
4 classifying people. So it's up to the Executive
5 Committee whether you think that should be
6 expanded or not.
7 DR. SOX: I think we also want to get a
8 sense from HCFA about whether those categories
9 are likely to be beneficial to them in trying to
10 make coverage decisions. That's certainly the
11 principle purpose of this system of categories.
12 DR. KANG: I actually think it would
13 be helpful, yeah. I mean obviously this is the
14 place, quite frankly, where our final coverage
15 criteria will interact, but at this point I think
16 the better strategy is to go for more categories,
17 whatever we can think of, and then to the extent
18 that we're collapsing categories in the future --
19 DR. SOX: Debbie Zarin made the
20 suggestion we've really got a three by three
21 matrix for everything except for the breakthrough
22 technologies, which would basically include every
23 possible combination of effective on the three-
24 point scale and advantages, no advantages or
25 disadvantages. So maybe we should simply use
.00191
1 that and then collapse those categories if you
2 find they're not useful. Alan?
3 DR. GARBER: I guess my experience
4 regarding the technologies per Blue Cross Blue
5 Shield is that the vast majority of technologies
6 have some advantages and some disadvantages, and
7 I think that we would be telling the angels how
8 to repent if we tried to decide whether or not
9 they were more or less advantageous. I mean some
10 of these technologies have fewer side effects for
11 the initial treatment, shorter duration of
12 benefit. Some have greater convenience, but less
13 effectiveness. And sometimes they trade off one
14 side effect for another. So I like our original
15 classification because I thought this
16 classification keeps us from spending too much
17 time pondering the imponderable.
18 DR. KANG: I'm going to withdraw. I've
19 run into the same problems and gotten paralyzed
20 from inaction. So I like this just fine.
21 DR. SOX: We could in our explanation
22 say why we put it in a particular category and
23 actually list any factors that led us to do that,
24 and that might be more valuable to you than the
25 category itself for making a judgment.
.00192
1 DR. KANG: I think that's correct.
2 DR. SOX: Randel?
3 MS. RICHNER: I wanted to ask the
4 overall panel if anyone has any concerns about
5 how to identify what the established service and
6 medical item is that you're going to be comparing
7 the technology to or the item to. Is that going
8 to be an issue? That's a question I have for
9 everyone. We've talked about that at length in
10 the subcommittee about what an established
11 medical service or item is and how do you
12 determine what that is. Is that going to be an
13 issue?
14 DR. HOLOHAN: Can you be more explicit
15 in what you mean by how do you determine --
16 MS. RICHNER: What's usual care, what's
17 usual practice. How are you going to decide that
18 this technology -- what are you comparing it to
19 for benchmarking this?
20 DR. HOLOHAN: You mean the term
21 established services?
22 MS. RICHNER: Right.
23 DR. SOX: Originally we had it already
24 covered, and we thought that would be too
25 limiting.
.00193
1 MS. RICHNER: It is.
2 DR. KANG: Having thought about this
3 problem a lot, I would actually suggest we're not
4 going to be able to resolve this one today. I
5 think that we ought to wrestle with this as we go
6 on and refine this one. This really is a tough
7 question.
8 MS. RICHNER: It's a tough question,
9 but I think that the tumor assay issue sort of
10 stems from all of that in terms of what is the
11 comparison and what is the benchmark?
12 DR. SOX: I wonder whether it will vary
13 from instance to instance. And part of this
14 series of things that you do during that first
15 month when you're trying to get the chart set up
16 is to decide what the comparison technology is
17 going to be.
18 DR. KANG: This is actually why Dr.
19 Hill referred to sector-specific guidance
20 documents. The reality is this is best addressed
21 by the panels almost because this is going to
22 vary from the sector that your talking about.
23 Maybe we could indicate that the panel can at
24 least in their context think about what the
25 comparisons ought to be. But this at this level
.00194
1 is not a solvable problem.
2 DR. SOX: Or the panel chair in
3 collaboration with HCFA staff is setting up the
4 charts. So I think Jeff has withdrawn his
5 proposal that we expand the number of categories,
6 and we can probably take that matrix down for
7 now. So we're still at 7, and we're going to
8 stay with established.
9 Are there any other comments about the
10 categories before we move on?
11 Hearing none, we'll move on to
12 Suggestions for Panel Operations. The first one
13 is Explanation, A panel must explain its
14 conclusions in writing. I think the basic reason
15 for this, the rationale is pretty clear, probably
16 not likely to cause much push back, but maybe the
17 implementation is an issue.
18 DR. FERGUSON: A comment, and I'm not
19 sure how the wording needs to be changed, but the
20 panel's conclusions will still be established by
21 voting; is that correct?
22 DR. SOX: That's correct.
23 DR. FERGUSON: Those who oppose a
24 motion are supposed to say why. Those who vote
25 yes, they presumably don't have to do that; is
.00195
1 that correct, don't have to say why they're
2 voting yes?
3 MS. LAPPALAINEN: Right. The
4 individual panel chair has the discretion at each
5 panel meeting to go round robin after the vote is
6 taken. Generally a no will invoke a question of
7 why you said no in order to make sure that any
8 minority response gets to the record. And the
9 other is, of course, the majority of the vote.
10 But this does not preclude the members from
11 expressing their opinion or even a dissension in
12 writing.
13 DR. FERGUSON: Okay. So then the panel
14 chair is responsible for summarizing the thought
15 that went into the yes or no votes I guess.
16 And again, it's a common question, how
17 to handle it. Maybe in case the panel chair, who
18 does not vote unless there's a tie, would be
19 responsible for writing this conclusion, and I
20 might disagree with the conclusion, which has
21 already happened once --
22 DR. SOX: It's the panel chair's
23 responsibility to write the conclusion that
24 reflects the majority regardless of his or her
25 own preference.
.00196
1 DR. KANG: Or, John, you could delegate
2 to a majority. I mean it's really at your
3 discretion.
4 DR. SOX: I would hope that panel chair
5 is capable of writing a strong piece on something
6 they disagree with. That's part of the job.
7 DR. DAVIS: I wanted to propose a
8 change on this last sentence which picks up on
9 this issue we're discussing. I wanted to suggest
10 that we change it as follows. The panel chair is
11 responsible for drafting the explanation of the
12 panel's conclusions, which should be circulated
13 to panel members for their comments and/or
14 approval. I just don't think it should be solely
15 in the hands of the chair without the opportunity
16 of the panel members to see it.
17 DR. SOX: Sharon, did you want to say
18 something?
19 MS. LAPPALAINEN: I just wanted to
20 clarify something. A summary of what happened at
21 the panel meeting is required by the Federal
22 Advisory Committee Act. That summary is
23 certified to by the executive secretary and the
24 panel chair. That is a legal requirement that we
25 will continue to do, and this is in addition to
.00197
1 that.
2 DR. SOX: Alan?
3 DR. GARBER: Well, I think Ron's
4 suggestion kind of comes down to what this report
5 of the conclusions is supposed to be in, and I
6 guess in the course of our subcommittee's
7 deliberations I had in mind saying it's going to
8 be much more rapid, something like a one-page
9 document that is approved at the time of the
10 meeting.
11 I think we have to be very sensitive to
12 the ways that we might unintentionally create in
13 this process, and I thought we should be brief
14 and very rapid in summarizing the results of the
15 meeting so that the panel can in real time
16 approve the chairman's summary of the conclusions
17 and the reasoning for the conclusions.
18 I think in most cases this is only a
19 summary. It does not have to be an exhaustive
20 review of what happened at the meeting because
21 after all, transcripts will be available and the
22 other materials that Sharon was talking about.
23 So I had in mind something like a one-page report
24 that is done at the meeting and wrapped up.
25 MS. RICHNER: But that's not clear.
.00198
1 DR. GARBER: I agree. So I guess Ron
2 has a much clearer way of stating the one model,
3 which is a longer process, but my intent had been
4 we do something in real time.
5 DR. SOX: I like Ron's approach better
6 because I think it's very difficult to write a
7 one-pager that is really good on the fly. Maybe
8 you can, Alan, but most of us can't.
9 And the alternative would be to require
10 the panel chair to write it, get it out for
11 comment, and if you don't hear from somebody in
12 48 hours, then you would assume to send and have
13 a requirement basically that it be back in HCFA's
14 hand in a week. That would give a little bit
15 more time to advise carefully and would give an
16 opportunity for thoughtful review of what's been
17 written. And I would think of it not so much as
18 approval, but comment. And ultimately it's the
19 responsibility of the chair to, in a just and
20 fair way, take into account comments. So that's,
21 I guess, more an attempt to telescope it out in
22 the interest of clarity.
23 DR. GARBER: Can I make a proposal that
24 we approach with discretion and collect some
25 experience? Because it sounds like we're
.00199
1 planning to adopt different approaches.
2 DR. SOX: But I think we ought to have
3 a sense of the group. Something ought to be back
4 in HCFA's hands in ten days.
5 MS. RICHNER: It went from 48 hours to
6 seven days to ten days. That's too long.
7 DR. SOX: Does a week seem reasonable
8 to get this done?
9 DR. GARBER: My concern is that there
10 are discrepancies in the comments. There's no
11 problem if the only differences are points of
12 clarification where there's no disagreement. But
13 as we've seen in some of these issues, there can
14 be considerable disagreement. And if you as the
15 panel have to adjudicate between two members that
16 say directly contradictory things, it's very hard
17 to resolve that without having a conference call
18 or face-to-face meeting. And I assume things
19 really have to be public.
20 MS. LAPPALAINEN: Presumably that would
21 fall under an operational aspect because we had
22 the public meeting, and the public transcripts
23 are available. The putting together of this
24 document would be operational, so we could have
25 another meeting to talk about the route.
.00200
1 DR. KANG: I actually have to agree
2 with Alan. While I'm sensitive to actually
3 Daisy's concerns, we do want to try and make sure
4 that the process does not slow down. I think
5 forcing a summary at the end actually forces
6 people to agree on what they can agree on and
7 disagree on what they can disagree on and
8 actually get it up there. The transcripts are
9 available to HCFA and its staff, and the whole
10 richness of the discussion is available. And
11 quite frankly, we would factor that in and look
12 at that also and look at the summaries together.
13 So I think forcing the summary before you go home
14 is the way to go.
15 DR. SOX: Any other comments?
16 DR. BROOK: I can tell you what's going
17 to happen here. People will reach agreement and
18 have very different reasons why they got there.
19 And the chair will only figure out what he
20 thought he heard, and it will not be what each of
21 the individual panel members voted yes or the
22 majority opinion agreed. So we are stuck with
23 either the panel chair trying to summarize the
24 evidence saying they voted already and to
25 summarize the reason for it or to ask each member
.00201
1 of the panel before they go home to write a
2 one-pager in support of their position, which
3 then would be the summary.
4 Instead of having this long transcript,
5 you could have a situation where the panel chair
6 is not responsible for summary, but each person
7 who votes is responsible for defending their vote
8 yes or no, and therefore, that would be part of
9 the evidence that goes with the vote. And nobody
10 would try to reconcile that this person believed
11 this because he liked that study, and this person
12 believed this because that person wore a green
13 tie, and this person believed this because they
14 were tuned out and daydreaming.
15 I mean it would motivate each panelist
16 to pay a little bit more attention -- I think
17 everyone would anyway -- but to pay a little more
18 attention to the process if they knew at the end
19 of it they would have to justify their vote.
20 So I would change this to say that not
21 only would this thing be voted on, but each
22 panelist is responsible for explaining in writing
23 at the panel's conclusion their individual vote.
24 DR. BERGTHOLD: There are seven
25 questions sometimes or ten to answer, Bob, and
.00202
1 that means that's a lot of stuff to write.
2 MS. RICHNER: I'm thinking of the FDA
3 advisory panel process. I mean it's been awhile
4 since I've been there, but you decide that day,
5 and you give your vote, and you say your
6 explanation as to why you gave your vote, and
7 it's on the transcripts, everybody knows it, you
8 can use that data later on, and you don't leave
9 that room until that's finished.
10 Sharon, correct me, but that's what I
11 remember.
12 MS. LAPPALAINEN: You're right. The
13 FDA process is as follows. Any primary reviewers
14 that are assigned to review prior to the panel
15 meeting, those written recommendations are part
16 of the administrative file for the particular
17 matter in front of the committee. At the
18 advisory committee transcripts are taken,
19 summaries are written and certified to.
20 Panelists will often think about what happened at
21 the panel, and subsequent to the panel meeting,
22 will send to FDA in writing, if they feel
23 compelled to do so, or if they feel that they had
24 a minority opinion that was not properly brought
25 forward. Those things that are in writing are
.00203
1 also part of the administrative record of what
2 happened at the panel.
3 DR. SOX: Our goals here, I think, are
4 twofold, mostly to serve HCFA's needs, and
5 secondly, to turn out a product that you can
6 understand and reads well. And it seems to me
7 that going around the room and explaining your
8 vote really deals with Bob's issue, puts that on
9 the record for HCFA to look at and say whoa,
10 actually this person has a point, we'll do it
11 this way instead of that way. So I think it
12 deals pretty well with that issue.
13 I'm still, frankly, troubled, Jeff,
14 with whether you're going to get the really good
15 prose that you want to put on the Internet from
16 trying to do it at the end of a long afternoon,
17 but we'll try it and see how it goes the best.
18 DR. KANG: Sharon, I'm not familiar
19 with the FDA process. On the FDA panels do they
20 actually try to do what Alan is suggesting with
21 the one page?
22 MS. LAPPALAINEN: Well, if I can have a
23 long response, the FDA asks particular questions
24 regarding particular matters that come in front
25 of the committee, and the panelists generally go
.00204
1 round robin on those questions during the open
2 committee deliberation. However, the vote for
3 either premarket approval in the device world or
4 new drug application in the drug world or
5 licensing application in the biologics world is
6 actually approved. And the panel has three
7 choices, to approve, to approve upon a condition
8 or to not approve. And so the ultimate vote is
9 really only on that issue and not the individual
10 questions.
11 DR. SOX: Bob?
12 DR. MURRAY: I'm a bit concerned about
13 trying to do it too quickly or in too frank a
14 fashion. Several points.
15 Number one, if the purpose is to form a
16 body of case law, then it has to be reasoned and
17 organized, and I think doing it on a very short
18 deadline before you leave in the afternoon would
19 not serve that purpose.
20 Secondly, I don't think it would serve
21 the purpose of giving a concise, logical document
22 to be used by other committees, by other panels
23 or by the same panel subsequently, if instead of
24 a single document, you had 10 or 15 separate
25 opinions each scribbled hastily.
.00205
1 And thirdly, if I were assigned to
2 write the summary, I would like to look at the
3 transcript because I would not want my summary,
4 the words I use in my summary, to come back to
5 haunt me if at a later meeting somebody had the
6 transcript and were able to argue that I did not
7 accurately summarize the expressions or the
8 reasons for the vote.
9 So I think that for the purposes that
10 we're intending this summary to serve, we're
11 simply going to have to live with a longer time
12 line, that it will have to be a week or more than
13 a week, at least until the transcript is
14 available, so that we can have the document that
15 will meet the needs that we've set forth.
16 DR. HILL: For our purposes, we're
17 going to have to take some responsibility in the
18 questions that we ask the panel because what we
19 need more than why you voted like you did is what
20 your scientific reasoning is. So the points at
21 which there's consensus of the panel and the
22 recommendations, that's going to be most helpful
23 to us, not so much the details of the dissent,
24 but the whys. And I think we can get at that
25 with the questions.
.00206
1 So Sharon's two-step process, we can go
2 ahead and fulfill our obligations under the
3 federal law with a summary. And do you want to
4 set a time frame today for how long you expect
5 the panel to turn it around?
6 And one last question, if I may. May I
7 take it that the Executive Committee is telling
8 the panels that if the chairman of the panel
9 disagrees as an individual with the findings of
10 his or her panel, that they are tasked with
11 writing or cooperating in the writing of the
12 summary in the most favorable possible way
13 against their own call, but in keeping with their
14 panel's decision, rather than delegating?
15 DR. SOX: Well, that was my opinion,
16 but others may disagree. I just think we're
17 professionals, and we ought to be able to do
18 that.
19 Jeff, can you give us a signal? Your
20 voice is saying, and your face is saying, you're
21 not sure whether a week or the same day is really
22 going to serve us well.
23 DR. KANG: I'm not sure I understand
24 the recommendation that's on the table or on the
25 floor or whatever. I guess what I'm hearing is a
.00207
1 summary that discharges our responsibility under
2 FACA, but then a formal, kind of, more thought-
3 out, well-reasoned document following it?
4 DR. SOX: Maybe we can do two things.
5 DR. HILL: I'm suggesting that in most
6 cases I think we're going to be able to go ahead
7 and use the committee's recommendations on the
8 basis of the preliminary thing, and if people
9 want to get their statement on the record for the
10 record, to further the record later on, I don't
11 think we're going to have to wait.
12 DR. HOLOHAN: I think you've just
13 confused me. You can make your decision on the
14 basis of want while waiting for a more formal
15 explanation, which makes it seem like the
16 explanation is ipso facto redundant.
17 DR. HILL: No, sir. I'm sorry. I
18 didn't mean to say that. Thank you for pointing
19 that out. I mean to suggest that we can begin
20 the process of working with the results of the
21 panel's findings, getting it into a form that we
22 can use. We don't sit down the next day and say
23 okay, that was it, here's the decision, and issue
24 it. We've got to go through some more work with
25 it ourselves. So if you take ten days, it's not
.00208
1 going to slow us down. We're going to begin our
2 work right away.
3 DR. BROOK: Can I just make a
4 suggestion here? We have three things on the
5 table. Maybe we're just going too far. There's
6 to be a vote at the end of this. There's this
7 transcript. Sharon said that HCFA has to write a
8 summary of it. Maybe we just ought to leave it
9 at that and allow panelists the opportunity to
10 submit within a couple of days any justification
11 for their vote, if they so choose. And then we
12 get away from the chair having to summarize
13 opinion without voting and doing all this kind of
14 stuff. But basically that they will have a vote
15 on the issue.
16 They're already going to have a summary
17 of the transcript that HCFA has to prepare and
18 which presumably is going to be done technically
19 competently. That will leave us with only the
20 option that a panelist could offer, if they would
21 like to explain their vote in writing, they could
22 do it or not do it.
23 DR. SOX: The problem with leaving it
24 as an option is that --
25 DR. BROOK: Then you come back to the
.00209
1 answer that you require each -- I don't see how
2 you can avoid requiring each panelist within a
3 reasonable time period -- doesn't affect the
4 vote -- to add anything else they want to add to
5 the record. That's what you're asking them to do.
6 DR. SOX: I think giving each panel
7 member the opportunity or the obligation to say
8 why they voted is going to help HCFA to --
9 DR. BROOK: So the panelists either
10 orally or in writing will be given the
11 opportunity, both orally or in writing, to
12 indicate why they voted on a particular issue.
13 And that discharges their responsibility. And
14 the panel chair's responsibility is to arrive at
15 a vote on this subject, not to write the summary.
16 And it's HCFA's responsibility, going over the
17 transcript under whatever this law is, to
18 basically write the summary. And then we don't
19 have a lot of redundancy.
20 And I don't think any chair, believe it
21 or not, is going to spend the next two days after
22 getting the transcript reading the -- it takes
23 two days to read it, right -- to read through the
24 transcript to summarize it while HCFA is doing
25 the same thing. That doesn't seem to make a lot
.00210
1 of sense.
2 DR. SOX: Alan?
3 DR. GARBER: Well, I want a stab at
4 this. I think a lot of this discussion is based
5 on some unstated assumptions maybe I don't
6 share. I think unlike the two panels that met
7 already, the way the future recommendations in
8 this report are implemented, it will be a highly
9 structured evidence review. The issues the panel
10 will have to deal with will be very sharply
11 focused. The staff has done its job in preparing
12 these reports. And it will boil down to a
13 limited number of issues that the panel will have
14 to make decisions about.
15 And frankly, I don't think it's that
16 difficult to write a brief summary in real time
17 that talks about those issues. It does not mean
18 that you redo the work of HCFA staff as part of
19 the report. And I have the sense that people are
20 talking about a very extensive redredging of the
21 information and the arguments and so on, and I
22 would suspect that will almost never be necessary
23 if a good evidence report structured on the
24 guidelines that this document states is
25 available.
.00211
1 I think this is actually pretty
2 simple. We're talking about what might amount to
3 a handful of bullet points, to summarize it. And
4 I think a longer report, given all the other
5 materials will be issued, is not going to be
6 particularly useful.
7 DR. SOX: Maybe what we should do is to
8 require a brief summary and then leave it up to
9 the chair, if he or she wishes, to write
10 something that would be somewhat longer, that
11 would be literate, logical and so forth, and then
12 just see what happens, what feels right once he
13 or she has some experience with that.
14 DR. HOLOHAN: One of the purposes of
15 this is to get uniformity and consistency, and it
16 sounds like we're now drifting away from that
17 again.
18 DR. SOX: But on the other hand, we're
19 in a mode of trying to learn by doing. And if we
20 have an understanding that we're going to reach
21 some final decision on this in a year, then we
22 can have our cake and eat it too.
23 DR. MAVES: I actually support Bob's
24 opinion on this. And I think if you do want to
25 write a summary, if the chair wants to do it, I
.00212
1 think it would be fine as long as it was
2 contemporaneously done, as Alan has indicated.
3 I'd be very concerned about a report written a
4 day or two after. You sort of go home on the
5 airplane, you think about this, you do the
6 inevitable Monday-morning quarterbacking, and the
7 report that Sharon writes and the report that the
8 chair writes may have a little different spin or
9 a little different angle. Not much. But that
10 could be very, very important as time goes on.
11 So I agree with Bob. I think we've got
12 two or three sort of summaries. You have a
13 transcript. You have a HCFA-put-together summary
14 of the meeting. You have the testimony of the
15 individual panel members as they give their votes
16 on each one of these things. I think that record
17 should stand as is, and I think to do otherwise,
18 except for perhaps a contemporaneously written
19 document by the chair that's there, that we can
20 see, that we can look at just like this, I would
21 be very, very concerned about both panel members
22 and the chair writing something after the fact
23 that would potentially cause us more problems
24 than fix them.
25 DR. SOX: It's pretty clear we're not
.00213
1 going to reach a consensus on this, so I think we
2 should have a motion and have a vote and move
3 on. Anybody want to make a motion so we can get
4 off this one? Mike, please.
5 DR. MAVES: I would move that the
6 deliberative process that we use consists of the
7 transcript, which is already being done by HCFA,
8 the summary, which will be prepared by HCFA
9 staff, the oral comments of the panel members as
10 they testify, and that those three pieces of
11 evidence suffice as the work product of the
12 panelists.
13 DR. JOHNSON: Second.
14 DR. SOX: Any discussion of that motion?
15 DR. KANG: I have a modification.
16 DR. SOX: Please. A friendly amendment?
17 DR. KANG: A friendly amendment. I
18 think what we want here is a summary, we want the
19 transcript, and then we want the opportunity for
20 dissent or whatever, which could always occur
21 later.
22 The summary could be done either way.
23 I would suggest it could either be a HCFA-done
24 with, as I understand, FACA, with agreement with
25 the chair, or they can go ahead and do it right
.00214
1 there and leave it up to the panel to figure it
2 out. But the end result is the summary, either
3 HCFA-done, with approval of the chair, or Alan,
4 kind of contemporaneously with whoever is doing
5 it right there at the panel, the transcript, and
6 then finally an opportunity for written further
7 dissent, comments or whatever.
8 DR. DAVIS: And a vote tally.
9 DR. KANG: And a vote tally.
10 DR. SOX: Michael, is that acceptable?
11 DR. MAVES: I'll accept that.
12 DR. SOX: Okay. Do we second it? Any
13 other comments?
14 DR. FERGUSON: Wait a second. So let
15 me understand this. So this summary then,
16 instead of being written by the chair, will be
17 either written by HCFA and/or with the chair's
18 input?
19 DR. KANG: The way FACA runs is by HCFA
20 with approval of the chair. So essentially the
21 chair is delegated to represent the whole
22 committee, or in fact, given the tone and
23 everything, they can just go ahead and write it
24 right there.
25 DR. FERGUSON: Done at the end of the
.00215
1 meeting so that it's seen by all those present at
2 the meeting; is that correct?
3 DR. KANG: Right. Either one. Up to
4 the chair. Either way would be acceptable.
5 DR. SOX: It wouldn't have to be done
6 at the meeting.
7 DR. KANG: Right.
8 DR. FERGUSON: So that this third thing
9 on our proposal here is sort of nixed at this
10 point? 3. Explanation: A panel must explain
11 its conclusions in writing. We're now doing
12 this --
13 DR. SOX: We've now operationalized
14 that. We probably should add this to the
15 document, add Mike's motion to the end of this
16 just to make it operational.
17 DR. FERGUSON: The transcript is done
18 anyway. That's a given.
19 DR. KANG: Right.
20 DR. SOX: Ready to vote? Bob?
21 DR. MURRAY: I would just like to make
22 a comment. I'm inclined to vote against the
23 motion. The reason is that when we had our
24 executive meeting in December, I believe we got
25 all of the documents that were discussed. We had
.00216
1 the HCFA summary of what had happened at the
2 previous two panel meetings, and we had
3 voluminous information. But in all of that, what
4 I found most valuable was Tom's summary of his
5 view. And what I hear happening is that we would
6 not ordinarily get that unless somebody like Tom
7 chose to do that.
8 I would rather see a reasoned summary
9 done after the fact because that would make our
10 job as an Executive Committee much easier and I
11 think more effective.
12 DR. SOX: Other comments?
13 DR. BROOK: There are two ways of
14 writing a summary. You've now recalled my
15 memory. I believe the HCFA summary was day one
16 began with this. These people testified. That's
17 not a summary.
18 When Sharon said that HCFA is required
19 to write a summary, I understood that to be an
20 executive summary of the 400 pages like the chair
21 did up to now, which says here's the evidence,
22 here's the major evidence discussed, here's the
23 opinions, and here's the results, and that the
24 panel would actually look at this 20-page
25 executive summary of these 400 pages of material
.00217
1 or maybe 30 pages of these 400 pages and have
2 that kind of a document. But if that's not the
3 case, then somebody has to write that document.
4 MS. LAPPALAINEN: But the summary must
5 reflect the agenda and what happened that day, so
6 the construction of the last three summaries,
7 which should have been available -- as a matter
8 of fact, we have one as a public handout now --
9 followed the agenda of December 8th.
10 DR. BROOK: But it didn't have any
11 summary of the issues. It did not have anything
12 that said the scientific evidence was presented.
13 The panelists basically looked at it. The
14 scientific consisted of this kind of information.
15 In other words, it wasn't a contents summary. It
16 was a process summary.
17 And I agree with you. Somebody should
18 write for the record a 20-page or so contents
19 summary of this voluminous amount of material
20 that only a very few people are going to read.
21 DR. GARBER: That's the evidence
22 report. I think looking back on the last panel
23 meeting, this is a misleading -- because we will
24 have evidence reports in place. That is assuming
25 that the recommendation goes forward. So a lot
.00218
1 of this would be superfluous.
2 DR. KANG: I think that's absolutely
3 correct. We can't look at the last two meetings
4 as -- these are all interactive. The reality is
5 the first half of this discussion was setting
6 guidance. It's saying that these are the
7 questions they'd have to answer. Then the fact
8 that -- good evidence report, this really tees up
9 the issues, and I think that we're learning,
10 quite frankly, as we're going along, and I really
11 don't think that the first two will be
12 representative of --
13 DR. SOX: I think we have a motion on
14 the table. We've had some discussion. Is there
15 anybody else who would like to offer discussion
16 before we vote?
17 DR. FERGUSON: I guess my discussion is
18 a question again. The last sentence here, the
19 panel chair is responsible for writing the
20 explanation of the panel's conclusions, modified
21 with what Dr. Davis did, that's different than a
22 summary, as Dr. Brook said. So we're not voting
23 on whether or not the panel chair or a designee
24 should write a summary of the panel's
25 conclusions. We're voting on something else.
.00219
1 DR. SOX: Why don't we vote on this,
2 and then it seems to me that vote implies we
3 ought to cross that out.
4 DR. KANG: How about the panel chair is
5 responsible for writing the executive summary?
6 DR. SOX: But according to the motion,
7 apparently not approved, it could be the panel
8 chair or it could be HCFA staff with the panel
9 chair.
10 DR. KANG: So HCFA staff or panel chair.
11 DR. SOX: I think we can basically
12 delete that sentence and substitute the process
13 that we voted on.
14 DR. FERGUSON: Delete this last
15 sentence? Is that what you're saying?
16 DR. SOX: That would be implied if we
17 vote this in. Any other questions?
18 DR. HOLOHAN: Can I ask for a
19 restatement of the motion?
20 DR. SOX: Restatement of the motion,
21 please.
22 DR. MAVES: I'll try. The motion was
23 that the operational documents that would result
24 from the panel meetings would be -- the
25 transcript will be number one.
.00220
1 DR. SOX: Not quite so fast.
2 MS. LAPPALAINEN: Operational
3 documents.
4 DR. MAVES: From the panel meetings
5 would be the transcript of the meeting, the
6 summary of the meeting -- and I think you could
7 put in parentheses prepared by HCFA staff -- and
8 the explanation of each member's votes for the
9 deliberations or the questions that are asked by
10 folks.
11 MS. LAPPALAINEN: With an opportunity
12 for dissension?
13 DR. MAVES: With an opportunity for
14 dissension.
15 DR. DAVIS: If I could ask a question.
16 If there are seven questions posed to the panel,
17 then you'll have to go around the table and get
18 an explanation from every panel member for each
19 of the seven questions?
20 DR. MAVES: Yes. And I think that
21 mirrors the practice that goes on at the FDA, if
22 any of you have been out there.
23 MS. LAPPALAINEN: As I have it written,
24 operational documents from the panel meeting will
25 consist of the transcripts, the summary that is
.00221
1 prepared by HCFA and signed off by the panel
2 chair, an explanation of each panel member's
3 votes with an opportunity for panel member
4 dissension.
5 DR. MAVES: Yes. I want to make sure
6 my seconder is here. Jeff, you're comfortable
7 with that?
8 DR. KANG: John, you were about to say
9 something.
10 DR. FERGUSON: I'm not sure that's
11 different than what we did before. I mean we
12 went around and voted on each question, and we
13 were obliged to say why we voted against
14 something, not really obliged for why we voted
15 for, and that was all captured in the transcript
16 and then HCFA's summary.
17 DR. MAVES: The reason for this is my
18 sense was that we're getting to a point where
19 we're going to have a third document, which would
20 be the chair or his designee's interpretation
21 being done at some point afterwards. And my key
22 concern about that was that you could have two
23 different, if you will, interpretations of the
24 same meeting. And rather than that we have this
25 as much as possible, either contemporaneously
.00222
1 recorded and transcribed or as needs be done
2 apparently through FACA, the HCFA summary of the
3 meeting done as well so that we don't have
4 situations -- and I think we had a little bit of
5 that last time where the interpretation of the
6 meeting and the HCFA document and the chair's
7 recommendation or the chair's interpretation of
8 the summary were two different things.
9 DR. SOX: I think there was one more
10 comment.
11 DR. BROOK: I want to just be clear
12 about the HCFA thing. Sharon, when you write the
13 HCFA summary, the last part of this is you're
14 going to have the up-front evidence report, then
15 you're going to have the explanation of the
16 votes. So you're going to look at this, the two
17 pieces of this stuff. Other than the process of
18 the agenda, you're going to summarize something
19 from the evidence report, a summary of the
20 evidence report, what's available going in, and
21 then the common themes across those whatever
22 number of panel votes for each of those votes.
23 So if Alan said the reason I voted yes
24 on this was because there were six controlled
25 trials and seven of these, the benefit was this,
.00223
1 and I believe it could be extended, you're going
2 to look at how they come across all the
3 individual panelists and then summarize that in a
4 factual manner so that it would be an aggregated
5 factual summary across the vote. That's the key
6 of what would have to happen. It would be
7 factual, but the aggregate across the votes is
8 based on reading the transcripts.
9 Is that what I understand this summary
10 is going to be? If everyone has agreed or said
11 the same thing, it could be one page?
12 MS. LAPPALAINEN: Right. The
13 requirement for this motion -- and that is an
14 explanation of each member's votes -- will be
15 added to the agenda as an agenda item for each
16 panel, and that will be included in the summary
17 if that is a required agenda item for each
18 panel.
19 DR. BROOK: There are two issues here.
20 You have ten people each saying a paragraph of
21 stuff. Somebody's going to look at the common
22 themes and write a summary of that. That's the
23 key fact that has to be done. And you're going
24 to do that. HCFA's going to do that.
25 DR. SOX: And the chair is going to
.00224
1 approve it.
2 DR. BROOK: Now, does the chair have a
3 right, if they're nonvoting, to actually give his
4 or her summary on the record when you go around?
5 After you've taken the vote, can we modify the
6 process so that the chair just doesn't sit there,
7 let's say at the end of this or at some point in
8 this process, and say here's how I would have
9 voted or something like that and here's my
10 explanation? Can that be done legally?
11 MS. LAPPALAINEN: Right. Presumably
12 after the voting period on the agenda and the
13 agenda item, which has been added, which is the
14 explanation of the vote, this also includes at
15 the end of that an opportunity for the chair to
16 express his or her opinion after the vote.
17 DR. BROOK: Why don't we require that.
18 Why don't we state that the chair should on the
19 record, after the vote has been taken, explain
20 his or her explanation for what he would have
21 voted or she would have voted, if he had the
22 opportunity to vote, so that it becomes part of
23 the record and part of the summary that you
24 write.
25 DR. SOX: Does that sound reasonable?
.00225
1 DR. BROOK: So we don't get the problem
2 with the chair saying something later because he
3 or she never had the opportunity like happened
4 last time.
5 DR. SOX: In just a minute Sharon's
6 going to read the motion, but first, since there
7 has not been a motion to vote, there's still an
8 opportunity for people to comment if they wish
9 to. Hearing none, Sharon?
10 DR. HOLOHAN: I don't want to
11 redundantly overclarify, but the HCFA summary
12 will in fact be what's written in this paragraph
13 as a -- and I'm quoting -- written explanation?
14 DR. BROOK: Yes.
15 DR. HOLOHAN: Okay.
16 DR. SOX: Ready to vote? And you're
17 going to reread it and then say who's eligible to
18 vote and who isn't.
19 MS. LAPPALAINEN: The motion which we
20 have on the table -- and we have a second I
21 believe -- is the operational documents from the
22 panel meeting will be the transcripts, the HCFA
23 summary, including an explanation, an explanation
24 for each panel member's votes at the panel
25 meeting, with an opportunity of dissension. The
.00226
1 chair after the vote will provide their opinion.
2 DR. SOX: Ready to vote? All in
3 favor?
4 DR. KANG: I think it's a summary of
5 the votes. It's an aggregate explanation with an
6 opportunity for dissension. The point is it's
7 got to be a content summary. It's got to say we
8 took a vote, here was 8 to 3, and on average this
9 is why it went this way.
10 DR. BROOK: It could say in voting yes,
11 that there were adequate controlled trials, three
12 said there was this, and two said this, but you
13 have to take that two paragraphs of that -- or
14 that two minutes of what that person says and
15 write a thoughtful summary. And we're giving the
16 HCFA staff the responsibility to do that with the
17 chair's approval, with the chair looking over
18 that part of the transcript, which will be much
19 shorter than the bigger thing, to do that.
20 MS. LAPPALAINEN: The motion is
21 operational documents from the panel meeting will
22 be the transcripts, the summary that HCFA
23 prepares, including a summary of the content and
24 explanation of each member's votes at the
25 meeting, with an opportunity of dissension. The
.00227
1 chair after the vote will provide their opinion
2 as well.
3 For today's meeting the members that
4 are eligible to vote on this motion are Thomas
5 Holohan, Leslie Francis, John Ferguson, Robert
6 Murray, Alan Garber, Michael Maves, Frank
7 Papatheofanis, Ron Davis, Daisy Alford-Smith, Joe
8 Johnson and Robert Brook.
9 Dr. Sox will vote in the case of a tie
10 vote.
11 DR. SOX: All those who are in favor,
12 please raise their hand and keep it up long
13 enough for Sharon to tally the vote.
14 DR. SOX: Two against. Abstentions?
15 One abstention.
16 MS. LAPPALAINEN: I'm going to read the
17 vote back. For the motion we have eight for, two
18 against and one abstention.
19 DR. BERGTHOLD: Now you need a written
20 explanation of that.
21 DR. SOX: We now need to move on to
22 talk about structure of the evidence provided to
23 the panel.
24 DR. FERGUSON: We don't have to explain
25 our no votes here?
.00228
1 DR. HOLOHAN: I think you should be
2 free to express why.
3 DR. SOX: Why did you vote no?
4 DR. FERGUSON: I voted no because of
5 some confusion on my part as to the timing of
6 when these documents will occur. My
7 understanding is that the transcript doesn't
8 occur to be finished until a week or more later.
9 The summary before wasn't finished at the time of
10 the meeting so that we could all look at it. And
11 I can't imagine that summary occurring at the end
12 of the meeting in a fashion that can be seen by
13 all of us. So since I was not clear on when that
14 could occur in a way that I could conceive of, I
15 had to vote no.
16 DR. SOX: Ron, do you want to explain
17 your abstention?
18 DR. DAVIS: I thought it was confusing
19 and awkwardly written, and I liked the original
20 with the amendment that I proposed.
21 DR. SOX: And Leslie, your no vote?
22 DR. FRANCIS: I would have preferred
23 just the requirements in the Federal Advisory
24 Committee Act and let panels explain it.
25 DR. SOX: Thank you very much.
.00229
1 DR. HOLOHAN: Could I explain why I
2 changed my vote? I thought that Bob Brook really
3 nailed down the content, and I was comfortable
4 with that.
5 DR. SOX: Does anybody else want to
6 explain a positive vote? Hearing no other
7 comments, let's move on to number 4, structure of
8 evidence provided to the panels.
9 I guess before we get into this, I'd
10 like to note that we have not at this point said
11 what ought to go in those evidence reports. And
12 presumably if we approve this section, then we're
13 going to have to get a group to get together
14 perhaps to work in collaboration with HCFA to
15 decide what will be the requirements for
16 whoever's going to write the evidence report. I
17 think maybe that would be better to not try to do
18 that together, but rather to do that off line
19 since it's really in the area of operations.
20 If anybody disagrees with that, I'd
21 like them to speak up, but that's my take on it
22 given the time.
23 Alan, do you think that's reasonable to
24 do it off line?
25 DR. BERGTHOLD: Mr. Chairman, it's
.00230
1 2:35, and we had a break scheduled for 2:15. I
2 just want to check. This next thing is going to
3 be actually I think complicated, or maybe not.
4 MS. RICHNER: Yes.
5 DR. BERGTHOLD: So I was wondering
6 could we take our break now?
7 DR. SOX: We're hard at work, and we've
8 shown our ability to talk for quite awhile in
9 trying to solve some of these operational issues.
10 So my suggestion is if there are members of the
11 panel who need to excuse themselves, they should
12 do so, but I think we ought to just work straight
13 on through.
14 Okay. So now we have number 4,
15 structure of evidence provided to the panels.
16 And what we're interested in hearing is -- again,
17 just to remind you of objections to this as a
18 basis for the panel's operations or lack of
19 clarity that's going to interfere with your
20 ability to work with your panel. And if you have
21 a problem with the language, we'd like you to
22 propose a change so we'll have something specific
23 to work on.
24 With that, I'll open the discussion.
25 Jeff?
.00231
1 DR. KANG: Could I just ask a question
2 as to your opening question? Because I missed
3 the first MCAC meeting of executive counsel.
4 I actually had thought the whole
5 purpose of the preceding four or five pages,
6 quite frankly, posed the evidence questions that
7 the evidence may support needs to think about
8 with this, so I was -- that was my -- and you're
9 thinking now that that's not adequate?
10 DR. SOX: I guess for myself, I'm
11 thinking that it provides the framework, but it
12 will be my, for example, wanting to talk to the
13 folks who are running the U.S. Preventive
14 Services Task Force to find out what their charge
15 has been to the evidence-based practice, for
16 example, what their deliverables are, and then
17 modify that as appropriate to meet the needs of
18 this group. I really think we need to define the
19 deliverables of whoever's going to provide these
20 reports, and those are specific.
21 DR. KANG: Let me suggest a strategy
22 because that second issue you raised is more of a
23 logistical issue. It's an issue as to --
24 whatever we want exists already or can we get it
25 -- I'm kind of wrestling with why it would be
.00232
1 fund -- I know that it meets the -- but why
2 wouldn't we want the evidence-based report to
3 take the first stab at answering the questions
4 that we have posed here in the heart of the first
5 five pages of this document?
6 DR. SOX: We might have some opinions
7 based upon our expertise about what they would
8 actually have to do to answer those questions
9 operationally.
10 DR. KANG: But there you're trying to
11 do deal with that in number 5. Whoever's working
12 on the evidence-based report who wanted
13 interaction with the panel members back and
14 forth, these things could get created -- some
15 interaction back and forth.
16 DR. SOX: Maybe it would be useful for
17 Alan, who's at least peripherally involved in --
18 comment on what sort of things go into their
19 report just so we have an idea of what we're
20 really talking about.
21 DR. GARBER: I think actually if I can
22 go to the prior question first, I think Jeff and
23 Hal are talking about this real important
24 operational issue, should the Executive Committee
25 give a lot of detail about how the evidence
.00233
1 should be structured to HCFA or should HCFA staff
2 proceed. And my relevant experience is actually
3 as chair of the med-surg panel where we've been
4 going over the agenda for our upcoming meeting,
5 and I've seen the first draft of what would be an
6 evidence report, and it's occurred to me from
7 seeing that, which I might add so far seems to be
8 very well done, that we might want to build up
9 some experience with HCFA staff doing these
10 before we make recommendations.
11 So I actually think that what they've
12 done so far is exactly the kind of thing that
13 this committee would recommend anyway. And maybe
14 because there are some areas that are a little
15 different, like diagnostic technologies, we might
16 want to gain some experience before we the
17 Executive Committee make any more specific
18 recommendations.
19 I'm actually very sympathetic to what
20 Jeff has just said based on my experience in
21 trying to prepare for our upcoming meeting. That
22 is it may not be suitable for us at this point to
23 give very detailed information about what things
24 like evidence tables should look like because
25 right now what they're doing for the urinary
.00234
1 incontinence studies is exactly what any EPC
2 would do.
3 DR. KANG: Let me add. It's
4 unfortunate because I don't think the first two
5 issues are representative. Reality is we can
6 contract out for some of this stuff, and we can
7 have, whether it's the tech assessment group over
8 at AHRQ or whatever, and then there's no reason
9 why the panel member can't interact with
10 whoever's doing that and interact in a fashion,
11 take a quick look, say no, you forgot to ask this
12 question or whatever.
13 I don't think we should get into the
14 logistics of how to do this. I think we should
15 just stick with we want an evidence-based report,
16 here is the list of issues and concerns we are
17 concerned about right now at this point, and
18 start working to answer those questions, and then
19 have number 5 there as an interaction to the
20 extent that they're things that are coming up
21 that we didn't anticipate.
22 MS. RICHNER: One of the discussions
23 that we had in our conference call, if you'll
24 remember correctly, was that the panels would
25 have an opportunity to pose the questions for the
.00235
1 evidence report before they were originally
2 conceived. So is that still the issue? I mean
3 that's still going to occur then?
4 DR. SOX: That will be number 5.
5 MS. RICHNER: My other problem and
6 question once again, how does this fit? I still
7 don't understand how and where the evidence
8 report fits in this Medicare coverage process
9 that has been published. So where and how is it
10 triggered and where does it fit in terms of the
11 panel receiving it? I still don't understand
12 it.
13 DR. HILL: In that flow chart you'll
14 see where we have the opportunity to refer things
15 to the Medicare Coverage Advisory Committee when
16 we take in issues as part of the process of
17 preparing the information for that committee
18 between the time the intention is arrived at to
19 send something to the panel and an accurate
20 amount of time before the panel. So they can be
21 able to digest it, we either create or get some
22 help in creating this evidence table.
23 MS. RICHNER: So the evidence reports
24 as we know take approximately six months to do.
25 DR. HILL: Not always. We've had
.00236
1 indications from AHRQ that in some cases, many
2 cases, they'll be able to do something for us a
3 little faster than that. We're working on our
4 own process internally trying to gear ourselves
5 up to be able to do those things faster.
6 If you're concerned about the time
7 frame that's involved, that's not stated on
8 there, and it wasn't -- so this doesn't change
9 that.
10 MS. RICHNER: You see, if we have the
11 evidence reports being -- there has to be
12 something written in here that when you, HCFA,
13 trigger this to MCAC, the evidence report and the
14 questions that need to go into the evidence
15 report have to be decided by the panel at that
16 particular moment. You have to have some
17 mechanism for the panel to get together to say
18 these are the seven things I want the evidence
19 report to reflect, and that doesn't say that in
20 here. I'm really grappling with this.
21 And then you have the six-month time
22 period where the evidence report would be
23 prepared approximately four to six months. Then
24 it would come to MCAC. We would then get the
25 evidence report and review it and have all this
.00237
1 time associated with reviewing it. I mean I'm
2 not tracking it with this document.
3 DR. SOX: Let me try to recall the end
4 of my talk this morning. HCFA decides to refer
5 something to MCAC. In that first month they work
6 with the chair of the appropriate panel to define
7 the questions, and that's a process that could
8 include other members of the panel if the chair
9 so designated. And they decide who's going to do
10 the piece of work and perhaps on the basis of the
11 nature of the problem --
12 MS. RICHNER: Who's they decide?
13 DR. SOX: The chair and HCFA.
14 MS. RICHNER: Decide who it's going to
15 be referred to?
16 DR. SOX: The decision about who's
17 going to do it -- HCFA decision.
18 MS. RICHNER: Whether it's going to be
19 ACRI or AHCPR or whoever is going to --
20 DR. HILL: Or internal.
21 DR. SOX: And then after that sort of
22 month of preliminary work, whoever is going to do
23 it gets the job, and they spend four to six
24 months doing it. They produce a report. And
25 then that goes out to the members of the panel to
.00238
1 prepare for a meeting that will occur
2 approximately a month after the report is
3 completed.
4 MS. RICHNER: None of that is reflected
5 in here. You know that, right? None of those
6 times.
7 DR. HILL: That's correct. As I said
8 earlier, we didn't state those times.
9 DR. SOX: Leslie?
10 DR. FRANCIS: This is a clarification
11 question. As a member of the panel, I would want
12 to get copies of the studies as well as the
13 evidence report, right? I don't want to just get
14 somebody's summary of it.
15 DR. GARBER: You may have 200 studies.
16 Again this is patterned on well-established other
17 technology evaluation processes.
18 And really, Randel, your questions are
19 getting into point number 5. But anyway, the
20 idea is that combination of staff and the chair
21 will identify interested panel members with
22 appropriate expertise and will involve them in
23 the process of helping to advise HCFA staff about
24 the scope of the evidence report or advise the
25 contractors or whoever it may be.
.00239
1 And this is intended to make sure that
2 the evidence report is the most suitable document
3 for the panel's deliberations. That means not
4 the entire panel is involved. The attempt is to
5 bring in all the really interested members of the
6 panel. And if by some chance that group of
7 people -- that is, the chair and the interested
8 panel is identified to assist in setting the
9 parameters on the evidence review, if by chance
10 they really goof up and they give directions that
11 some important studies were neglected or the
12 scope of it was wrong, that would come up during
13 the panel meeting. And then perhaps the panel
14 will conclude they didn't have the evidence they
15 needed.
16 But generally speaking, this kind of
17 system works where you get all the interested
18 parties to give input early in the process, and
19 you don't have to go through actually convening
20 two panel meetings, one to set up the evidence
21 report and another to evaluate it.
22 DR. SOX: Leslie?
23 DR. FRANCIS: I'm not asking for two
24 panel meetings. I just would not feel, as a
25 panel member, that I was in a position to
.00240
1 evaluate the evidence unless I both had the
2 studies on which the evidence report is based and
3 the evidence report as an analytic summary of
4 those studies. What concerned me with the
5 myeloma panel was that I had about 30 studies and
6 nothing else.
7 DR. SOX: My take is that if an
8 individual panel member wanted those studies and
9 had the time to do it, they could get them and
10 that the evidence report, if it focused on two or
11 three really key studies, that those might be
12 included as an appendix to the report so you
13 could read it.
14 DR. HILL: Our intention at this point
15 is when we identify, or the panel chairman
16 identifies, key studies that should be sent to
17 all panel members, they will be. And when you
18 get to the table, if there's something you read
19 off there, then we'll send it to you. And if you
20 tell us ahead of time that you're the one person
21 who wants to get the whole five crates, we'll
22 talk about it.
23 DR. SOX: Any other comments about this
24 section before we move on? Jeff?
25 DR. KANG: I had to step out of the
.00241
1 room, but I wanted to comment to Randel's issues
2 on timing. I said this earlier in the morning.
3 The last slide we had, which was some
4 time frames, was actually -- I don't know how to
5 say -- was kind of Medicare Coverage Advisory
6 Committee centric I guess. The reality is that
7 staff is really responsible for the logistics of
8 the timing and the flow.
9 I really would encourage you all in
10 your deliberations to consider what is desirable,
11 what do you want. We then are responsible for
12 the timing and the logistics and meeting what we
13 said we were going to meet in the federal
14 register notes. And we're committed to trying to
15 make that work.
16 Now, it may turn out what you all
17 believe is desirable is physically humanly
18 impossible, and then we may have to rethink this.
19 But I actually, quite frankly, think it is
20 possible. And this guidance that you've given
21 staff is extraordinarily helpful because it will
22 lead quickly to evidence-based reports to answer
23 the questions or up front there's this
24 interaction in step number 5. I think this is
25 very doable and still meeting the time frames
.00242
1 that we said in the federal register notes.
2 That's a commitment. But our issue is to try to
3 sort out the logistics, and we will do that.
4 MS. RICHNER: One more question. The
5 data. There was a point when we had our
6 orientation for the panel members. As an
7 industry representative, anyone can give me
8 information that I would share then with the
9 panelists. That's one of my roles, which could
10 be unpublished literature, it could be white
11 papers, it could be FDA information, that would
12 not have been provided in the evidence report.
13 Where and how does that get
14 considered? I know that they may not be
15 controlled studies, but it's information that can
16 go into the decision-making process. So where
17 does that fit?
18 MS. LAPPALAINEN: Right. The industry
19 representative's role on the panel is to
20 represent industry. And if you believe that
21 information needs to get to the panel, you need
22 to give that to us at HCFA, not directly go to
23 the panel and have the panel interact.
24 MS. RICHNER: But how does that fit
25 into this? Is it the only thing you receive is
.00243
1 the evidence report?
2 DR. KANG: It's part of the evidence
3 report.
4 MS. RICHNER: So that means I would
5 have to give it to ACRI or AHCPR?
6 DR. KANG: You'll give it to us, and
7 we'll figure it out.
8 DR. BROOK: The only problem with that
9 is if the information is proprietary, then you're
10 going to have a hard time because the evidence
11 report, you're job should be -- everybody's job
12 should be to get to the person at HCFA everything
13 under the sun. And that person should summarize
14 that in an unbiased manner. And so published,
15 not published, we ought to be beating every drum
16 we can find to get good information. But if you
17 send along a tag you can't use it or publish it
18 because it's proprietary, then it won't be used.
19 MS. RICHNER: Of course.
20 DR. KANG: Randel, it's really not your
21 responsibility. It is the requester's
22 responsibility.
23 MS. RICHNER: Right. But it's not
24 reflected here in this process, and so I just
25 wanted to make sure that -- maybe the public now
.00244
1 is aware that that is part of the process, that
2 information can be provided to your industry or
3 consumer representative that should be given to
4 the panel or to HCFA as part of the evidence
5 report.
6 DR. KANG: We'll make that clear, but I
7 don't think this is the document to make it
8 clear.
9 DR. HILL: We already do invite those
10 sendings in our announcements.
11 DR. SOX: Any other comments on this
12 section before we move on?
13 The next section is about panel member
14 involvement, the chair up front with appropriate
15 other members of the panel, in framing the
16 questions, and several panel members should be
17 participants in the evidence review as a way of
18 gaining familiarity with data and expertise on
19 the topic, and finally, there should be a couple
20 of primary reviewers whose responsibility would
21 be to spend a lot of time going over the evidence
22 report prior to the meeting and be in a position
23 to summarize their take on the evidence as
24 reflected in the report.
25 So those three aspects of panel member
.00245
1 involvement are now open for discussion.
2 DR. HOLOHAN: I think I'm asking this
3 for Leslie. It says panel members should take an
4 active role in reviewing the evidence, a word
5 that I believe is distinct from the evidence
6 report.
7 DR. FRANCIS: It's not the evidence
8 report. It's the evidence.
9 DR. GARBER: I don't think that's
10 realistic in some of these areas; that is to say
11 to review all the evidence. I mean this is
12 basically reviewing all the evidence. You do a
13 serious job of it even without writing it up.
14 It's a several-week, full-time job.
15 DR. HOLOHAN: I understand that. I'm
16 simply saying it --
17 DR. GARBER: Oh, okay.
18 DR. SOX: What do you think would be
19 good language there? Reviewing the evidence
20 report?
21 DR. KANG: Preparing the evidence
22 report.
23 DR. SOX: So it would be an active role
24 in preparing the evidence report?
25 DR. FRANCIS: The reason I had made
.00246
1 that is not equal to the evidence is I'm not
2 going to know how to vote as a panel member
3 unless I think I've been able independently to
4 come to my own judgment. I'm not around here to
5 rubber stamp an evidence report. An evidence
6 report and other people's comments on it are
7 helpful to me in trying to reach my own judgment,
8 but if it's all just laid out and I can't in any
9 way try to exercise my own critical judgment, I
10 don't have any business being here.
11 DR. BROOK: First of all, there's no
12 rubber stamp on this. An evidence report just
13 puts the evidence together. And then you need to
14 produce the judgment, based on the evidence, what
15 to do.
16 Now, if you're saying you want to redo
17 the evidence report, what I think Alan and I are
18 doing, having done a lot of these evidence
19 reports, be it as it may, in areas which have
20 lots of literature, we've reviewed 10,000 titles
21 to come up with 300 articles to summarize, and
22 we're struggling to get this done in six months.
23 There is no question that HCFA, if
24 you'd like, should be able to provide you all the
25 original material that we work from, but I will
.00247
1 tell you that unless you're the most
2 extraordinary individual under the sun, you will
3 not have the time to redo this what to do, but
4 you ought to have the right to do it.
5 And certainly I think any panel member
6 ought to have the right to get the original
7 evidence, and it ought to be stored in a manner,
8 put together in a manner, and that ought to be
9 sent out. But you ought not to expect the
10 average panel member to do that. We ought to
11 expect the average panel member to believe that
12 the evidence has been synthesized correctly, and
13 now you have to make a judgment about how it
14 should be used and what it means.
15 DR. SOX: Do you want to add to that?
16 DR. GARBER: I think Hugh put it really
17 well about how this would work. I think, Leslie,
18 the issue for us is going to be we have to look
19 at the original data for some key studies, and
20 all the panelists should get those key studies,
21 but not the huge volume that Bob was alluding to
22 that we usually start with. So that's why this
23 will never be -- I doubt that this will ever be a
24 rubber stamp. The panelists are going to read
25 some studies, but they have to be whittled down
.00248
1 somehow. And that's all we're saying is be
2 selective about it.
3 DR. SOX: I'd like to hear from the
4 panel if there's objections to the concepts that
5 are imbedded in the boldface number 5. Does this
6 look reasonable for panel members? That's
7 great.
8 DR. BROOK: Under the first boldface it
9 should insert report.
10 DR. SOX: Panel members should take an
11 active role in, I thought we said, preparing the
12 evidence report.
13 DR. KANG: Preparing the evidence
14 report.
15 DR. SOX: Not reviewing. Change it to
16 preparing and insert report after evidence.
17 So let's discuss this section trying to
18 pick out -- we don't have a lot of time now, so
19 we've got to kind of focus again on problems with
20 clarity, pieces that are objectionable. John?
21 DR. FERGUSON: Just a suggestion. This
22 number 5 might better be put on one of the first
23 under Suggestions for Panel Operations because
24 this sort of explains what the beginning of the
25 process is, which Randel was questioning about.
.00249
1 Because this is the first part. The evidence
2 report, the panel chair and others working with
3 the evidence, we're going to do the evidence
4 report. And the first thing you start out with
5 is actually the end result.
6 MS. RICHNER: That would help a lot,
7 just moving it to the first.
8 DR. SOX: The problem is that it does
9 talk about the evidence report, which is defined
10 in the immediately preceding section. It can
11 certainly be --
12 DR. FERGUSON: It sort of operationally
13 comes after the fact.
14 DR. SOX: I think we can probably move
15 the first one till later because it comes later
16 in sequence. That would work. Okay.
17 Other comments on this section? In
18 that case we're going to move on to number 6.
19 MS. RICHNER: One more thing. To me,
20 once again, it's the timing, but that's going to
21 be clarified by HCFA and not by us?
22 DR. KANG: We're on the hook for
23 timing.
24 DR. HOLOHAN: If I can offer an
25 unsolicited comment. There's been a lot of
.00250
1 concerns first about how long these take. If
2 nothing else, the experience with the first two
3 panels should have instructed us that doing it
4 the right way is the fastest way.
5 DR. SOX: Just to quickly follow up on
6 John's suggestion, consistent with John's
7 suggestion that we try to get these operational
8 things consistent with sequence, everybody happy
9 with moving the first one, which is now number 3,
10 a panel must explain its conclusions in writing,
11 make that the last one? Okay.
12 Then let's move on to number 6, which
13 is expert review of evidence reports.
14 DR. KANG: Before we discuss this, can
15 I ask the subcommittee to explain why this is
16 here just so I can understand?
17 DR. SOX: The opinion of experts is the
18 best way to assure everyone that the evidence
19 report is complete and fair. So it's a notion
20 getting said people that are competent experts to
21 look at it and say it didn't miss anything, the
22 report didn't distort the clinical facts as we
23 know it. It's something that, at least on all
24 the other panels I've been involved with, outside
25 review has been a key part of it if only to
.00251
1 establish the credibility of the process, to say
2 look, we've given the people who have an axe to
3 grind the chance to sling their strongest arrows.
4 DR. KANG: Well, maybe I didn't read
5 this closely enough. The assumption here that
6 the evidence-based reports are being done by
7 nephrologists, internists or whatever and that
8 the final analysis, if it's about some surgical
9 procedure, that you'd also like to show it to a
10 couple of surgeons? Is that the issue here?
11 DR. SOX: That strategy -- namely,
12 having evidence-based clinicians prepare the
13 report, then have it reviewed by competent
14 experts -- seems to really work well on the other
15 side.
16 DR. GARBER: Jeff, one of the explicit
17 precedents here is the evidence-based practice
18 center's review process in which the external
19 reviews come from actually a wide range of types
20 of expertise ranging from pure methodology to
21 pure clinicians. I'd just like to emphasize the
22 language here is committee recommends expert
23 review. I think we recognize this could be
24 onerous in some circumstances and maybe not
25 always is necessary in some circumstances as
.00252
1 others. So this is really truly advisory, but we
2 do feel it's very important to do it to ensure
3 the highest quality.
4 DR. SOX: Any other comments about this
5 section?
6 DR. FRANCIS: I apologize for
7 continuing to beat what's probably a dead horse,
8 but I really do think whatever else you do, in
9 addition to the evidence report, you ensure that
10 panel members have the key studies. It's okay
11 from the evidence report to identify key studies,
12 but I want to see them too.
13 DR. SOX: Ron?
14 DR. DAVIS: I actually scribbled out a
15 sentence to address that, and if the Executive
16 Committee feels it's important to make that
17 explicit, panel members will have the evidence
18 report at their disposal and will have the right
19 to obtain any primary sources upon which it's
20 based. But I don't think there should be an
21 affirmative obligation on behalf of HCFA staff to
22 send us all those primary resources.
23 DR. SOX: That suggestion seems
24 consistent with the discussion we've had. Any
25 objections to it? Okay. Then we have to go back
.00253
1 and --
2 DR. KANG: I'm sorry. I heard what
3 Leslie was saying was key articles be part of the
4 report and then that she also has access to the
5 10,000 if she wants.
6 DR. FRANCIS: Exactly. That's what I
7 want.
8 DR. SOX: We've gone through the
9 document once. Now, let's start over.
10 Ron has been given responsibility for
11 marking up the transparency that Jeff and Leslie
12 prepared during lunch. Do you have a report to
13 make?
14 DR. DAVIS: It's over there on the
15 transparency.
16 DR. SOX: Okay. Let's look at it and
17 see if we like it.
18 DR. DAVIS: I tried to cut down words.
19 The first line and a half goes in italics, I
20 guess, because it's a subheading. Should I read
21 it now?
22 DR. SOX: Yeah.
23 DR. DAVIS: Medicare beneficiaries
24 include elderly, nonelderly and disabled people.
25 The Medicare population also may or may not
.00254
1 include patients with comorbid disease.
2 Historically many controlled trials unfortunately
3 exclude older men and women, people with
4 disabilities and people with comorbid disease.
5 Thus these studies may have had adequate
6 statistical power for the study population, but
7 the results may or may not be generalizable to
8 some portions or all of the Medicare population.
9 If the requester is asking for coverage or if the
10 panel believes there is a medical benefit beyond
11 the clinical and demographic characteristics of
12 the study population, the panel should state
13 whether it believes the results of the studies
14 are applicable to some groups covered by
15 Medicare, define what those groups are, and
16 explain its reasoning.
17 DR. SOX: Anybody have any changes
18 they'd like to make to that masterful piece of
19 rewriting?
20 DR. FRANCIS: Thank you.
21 DR. SOX: Great. Thank you very much,
22 Ron.
23 DR. DAVIS: Alan just suggested at the
24 end to change it to say define the groups, and
25 then we'll say --
.00255
1 DR. SOX: Daisy, are you ready with
2 some suggestive language for the preface
3 regarding --
4 DR. SMITH: Yes. In fact, if you'll
5 recall, initially we had discussed the
6 possibility of inserting it under external
7 validity. And at that time when I was in that
8 mindset, I thought we were going to say although
9 the panel recognizes that adequate representation
10 of every study may not be possible, consideration
11 should be given to the applicability including
12 race and culture when appropriate and necessary.
13 Then I thought that would get into too much, but
14 that's what I was charged to do in terms of the
15 insertion.
16 But instead I chose to suggest that we
17 put it in the preface and add it to the amendment
18 that had already been added, which stated so that
19 Medicare beneficiaries -- you know, we said
20 something about that. I think Linda started.
21 Then I just added to that -- can be better served
22 regardless of race, ethnicity or socioeconomic
23 status. And that's a generalized statement
24 without attempting an insert with limitations.
25 DR. SOX: Any objections to the way
.00256
1 this is done? Then we have one more suggested
2 change.
3 DR. BERGTHOLD: Mr. Chairman, when will
4 these changes be on the books? I didn't take
5 down every one. Maybe I should be asking Sharon.
6 DR. HILL: By next week. Maybe even
7 sooner.
8 DR. SOX: So Ron, what is it that he
9 just gave you?
10 DR. DAVIS: It's the sentence that I
11 mentioned earlier about having the opportunity to
12 review any of the primary sources upon which the
13 evidence report is based.
14 DR. SOX: I'd like to move on now. I
15 think we're ready, are we not, Sharon, to have
16 open public comments before we vote?
17 DR. KANG: I actually have one
18 modification, that we put in a phrase that says
19 based on feedback from the panels, this is a
20 living document basically. This has been
21 modified. I just wanted to say maybe it's
22 feedback from the panels and other stakeholders.
23 Obviously we have public comment period. So it
24 really is maybe the other way to say feedback
25 from everyone. It could come from public, could
.00257
1 come from the advisory committee, the Executive
2 Committee itself.
3 DR. SOX: Do you recall where that
4 language was?
5 DR. KANG: It would be the last
6 paragraph before Evaluation of Evidence, the
7 section that begins Evaluation of Evidence.
8 DR. SOX: We're running out of time
9 because some of our members are going to have to
10 leave at 3:30, and I'd like, if possible, to have
11 as many people here for the vote on this. So we
12 will put your suggestion in, Jeff. Sounded like
13 everybody was happy with it.
14 We now have a 15-minute period when
15 anybody who wishes to make a comment may do so.
16 In order to assure that there be equitable
17 distribution of the 15 minutes, I'd like anybody
18 who wishes to make a comment to please raise
19 their hand so I'll know how many people want to
20 make a comment, and then I can decide how much
21 time each person will be allotted.
22 In the event that only a few people
23 want to make a comment, I would hope that they
24 could keep their remarks short because we would
25 like to have a whole committee here if possible
.00258
1 for final vote on this document.
2 DR. KANG: I apologize. Just one
3 procedural issue. You recall earlier this
4 morning I actually gave up some time to try to
5 get on with the meeting. I have an announcement
6 I'd like to make with regard to coverage criteria
7 for the public, and unfortunately I think the
8 appropriate time would be right after the vote.
9 I just wanted to alert people that I did want us
10 to have maybe some closing remarks.
11 DR. SOX: Excellent. We look forward
12 to those.
13 So Mr. Northrup is one person who's
14 scheduled.
15 Anybody else who wants to make a
16 comment? A total of four. Three minutes each.
17 Mr. Northrup?
18 MR. NORTHRUP: I want to thank you for
19 this opportunity. This is about as close to the
20 last word as anybody outside the government ever
21 gets on a public policy issue, so thank you very
22 much. I do want to thank all of you for what you
23 are doing for Medicare beneficiaries, and that's
24 why we're all here. Before you close, I do want
25 to also reiterate why what you're doing and why
.00259
1 you're doing it --
2 DR. SOX: Excuse me. I didn't
3 introduce you.
4 MR. NORTHRUP: I'm Steve Northrup,
5 Executive Director of the Medical Device
6 Manufacturers Association in Washington, D.C.
7 Again, I want to point out why what
8 you're doing and how you do it is so important to
9 the medical devices community and the patients
10 we're trying to serve.
11 A way of a little background, our
12 association, MDMA, was created in 1992 by a group
13 of medical technology entrepreneurs to represent
14 and serve medical technology entrepreneurs. And
15 I do want this committee to keep in mind, and you
16 probably already know it, but please keep in mind
17 the foundation of innovation in medical
18 technology is the entrepreneurial sector. Most
19 of the innovation in this industry comes from
20 entrepreneurs, and in fact, I read recently one
21 of the CEOs of a large medical technology company
22 said that 60 percent of all the medical products
23 sold in this country are less than 12 months old.
24 And that seems like an impossible number, but
25 that's the nature of innovation in this industry.
.00260
1 It's incremental innovation fostered by
2 entrepreneurs, entrepreneurs with lots of ideas,
3 but limited time and limited cash. And we need
4 to be sensitive to that, and we talk about the
5 type and amount of evidence HCFA's going to
6 require and this committee is going to require
7 and the amount of time it's going to take to
8 reach a decision.
9 And that brings me to the points I'd
10 like to make briefly about evidence and about
11 time. With respect to evidence, I do appreciate
12 the steps you've taken today to make some of
13 these guidelines, I think, more reasonable with
14 respect to evidence. And ultimately HCFA's
15 coverage criteria, which Dr. Kang will talk
16 about, will provide that, quote, unquote, road
17 map that manufacturers are looking for.
18 Manufacturers are willing to jump over
19 a reasonable bar, but if it's unreasonably high
20 where we can't even see it, a lot of us smack our
21 heads right into it. And most importantly for
22 your purposes, please keep in mind that most of
23 the advances in medical technology that your
24 panels will be considering are incremental
25 advances and don't necessarily require a de novo
.00261
1 review. So when you're looking at incremental
2 advances, let's look at the incremental
3 evidence.
4 With respect to time, still somewhat
5 concerned -- and I do appreciate Dr. Kang's
6 comments along these lines -- that some of the
7 things you're considering will slow down the
8 process of coverage decision making
9 unnecessarily, and that will in turn slow down
10 the pace of innovation in our industry. The
11 government will never be able to keep up with the
12 pace of innovation in this or any other industry.
13 That's just the nature of the beast. But we need
14 to try to keep the gap between innovation and the
15 government's pace as small as possible. And with
16 respect to the comment that was made earlier
17 about doing it the right way is the fastest way,
18 to borrow a phrase, I'd like to say it depends on
19 what your definition of right is, and we need to
20 focus on doing it the best way.
21 I do want to thank you for your time,
22 and ultimately I'm not asking you to be sensitive
23 to our companies or their needs or how they
24 conduct their business. That's my job and not
25 yours. What I would ask you to do is make sure
.00262
1 that your actions and decisions don't hinder or
2 discourage medical technology entrepreneurs from
3 innovating because innovation is the key to
4 improving the health of Medicare beneficiaries.
5 Thank you.
6 DR. SOX: Thank you very much, Mr.
7 Northrup. Who's going to speak next? Yes, sir.
8 Please introduce yourself.
9 MR. COOK: My name is Ken Cook.
10 MS. LAPPALAINEN: Do you have any
11 financial interest in any service?
12 MR. COOK: I have no financial
13 interest. My name is Ken Cook, and I'm a
14 facilitator for a cancer support group at the
15 University of Maryland Medical Center. I just
16 want to make a comment on two issues on the
17 external validity issue and Medicare patient
18 participation.
19 Not only are Medicare patients excluded
20 sometimes from clinical trials because of age,
21 but because also of the financial problem. Since
22 Medicare will not pay for experimental protocols
23 and since probably most patients or most
24 beneficiaries of Medicare do not carry separate
25 insurance, unless they are financially
.00263
1 independent, they're basically precluded from
2 being in the population that is undergoing a
3 clinical trial. So it's a catch 22. You can't
4 get into the clinical trials because you don't
5 have the money. That's the first item. And so I
6 would like to point that out for your
7 consideration.
8 The second issue is on the number of
9 patients involved in any disease study. If you
10 are studying prostate cancer, there are many,
11 many patients available for clinical trials.
12 There is sufficient research money available.
13 But if you are a patient with let's say multiple
14 myeloma, which is less than one percent of the
15 population, there is very little research and
16 very little research money available or public
17 interest in that issue. And to try to require
18 the same degree of rigidity in proof and making
19 sure that the protocols are as perfect as can be
20 possible may not be appropriate.
21 So like there is the orphan drug law, I
22 think that as we consider the various types of
23 protocols and how they're applicable to the
24 different groups, the same measures are not
25 applicable to all. One size does not fit all.
.00264
1 Thank you ever so much.
2 DR. SOX: Thank you very much, sir.
3 DR. KANG: Chairman Sox, could I just
4 respond to the first point? I think that's a
5 real issue. I'm very aware of the IOM report. I
6 just wanted to say that I don't think this is the
7 venue, the MCAC, but I just want to assure you
8 that the issue on payment for -- clinical trials
9 is very much on our screen and being reviewed
10 here at HCFA.
11 MR. MESKAN: I'm Tom Meskan, Medical
12 Alley. The committee at one point was discussing
13 its willingness to take comments about tone
14 and/or substance of the document, and I tried to
15 listen to the conversation closely, but never
16 heard a complete resolution of whether you wanted
17 to accept those remarks that would have any
18 value, and what's your orientation for us.
19 DR. SOX: I think the sense of the
20 group is that when we put this thing back on the
21 website in its modified version, it will call for
22 public comment very much on the spirit that Dr.
23 Brook suggested of specific wording that we might
24 change, specific changes in the wording that
25 might improve the tone. And it will, of course,
.00265
1 be up to the committee to decide to accept those
2 suggestions. But I think that's the sense of the
3 group.
4 MR. MESKAN: As it relates to tone, are
5 you open to substantive changes or do you feel
6 that where your document is now is kind of where
7 it is and yes, it's an interim document that will
8 be ongoing, but should we bother to spend the
9 effort to make our points again in perhaps more
10 compelling ways on substance?
11 DR. SOX: I think it would probably
12 serve our group and your ideas best if you came
13 back to us with them as we reconsider the
14 document on a periodic basis. Given the time, it
15 probably isn't going to get the attention that
16 maybe it deserves.
17 DR. KANG: I think I did hear Bob say
18 though -- and I thought it was appropriate --
19 that substantive changes would be considered, but
20 then you have to justify why the substantive
21 changes should be in that item.
22 DR. SOX: We want suggestions about
23 tone and substance, and we'll take them up in due
24 time, but we won't ignore them.
25 The last speaker, please introduce
.00266
1 yourself.
2 MR. LASCHER: Steve Lascher,
3 epidemiologist of the Maryland College of
4 Physicians. I have no financial affiliation.
5 DR. SOX: What organization?
6 MR. LASCHER: ACP-ASIM. Related to the
7 overhead that was written related to the
8 generalizability, I just wanted to mention that
9 statistical power was mentioned, and perhaps in
10 that respect it wasn't the appropriate term since
11 statistical power relates to type two error, and
12 perhaps you were thinking about sample size, and
13 it might lead to some misunderstanding.
14 DR. SOX: Thank you. It's now time for
15 the committee to take a vote. Sharon?
16 MS. LAPPALAINEN: At this time Dr. Sox
17 would call for a motion, and he will be asking
18 the voting members of the panel to vote
19 concerning whether the report of the subcommittee
20 should be ratified or ratified with modifications
21 or not ratified.
22 For today's panel, a forum is present,
23 and the voting are Dr. Thomas Holohan, Dr. Leslie
24 Francis, Dr. John Ferguson, Dr. Robert Murray,
25 Dr. Alan Garber, Dr. Michael Maves, Dr. Frank
.00267
1 Papatheofanis, Dr. Ronald Davis, Dr. Daisy
2 Alford-Smith and Dr. Joe Johnson. Dr. Robert
3 Brook is absent.
4 The panel vote may take one of three
5 forms, ratification with no other modifications,
6 ratification upon condition, for example,
7 resolution of some clearly identified
8 deficiencies which have been cited by you or by
9 the HCFA staff. Examples of deficiencies could
10 include resolutions of some of the questions of
11 wording or issues that you believe are necessary
12 or you would like to see implemented.
13 If you believe that modifications are
14 necessary, then your recommendation should
15 address the following points; the reason or
16 purpose for the modification and the information
17 that's required to change it. And for
18 nonratification, if you believe that the
19 subcommittee report should not be ratified, we
20 ask that you state for the record your reasons
21 why the report should not be ratified and to
22 identify those measures that should be taken in
23 order for you to ratify it in your opinion.
24 Thank you.
25 DR. SOX: Sharon, am I correct in
.00268
1 saying that the only people that can participate
2 in the discussion now are voting members?
3 MS. LAPPALAINEN: Yes.
4 DR. SOX: I've asked Ron to prepare a
5 motion, and I'll read it on behalf of him. Then
6 there can be a second, then there can be an
7 opportunity for discussion, and then amendment.
8 Motion, that the Executive Committee
9 approve the subcommittee's report and
10 recommendations as amended and that the Executive
11 Committee revisit the report and revise it as
12 needed in response to comments from panel members
13 and the public.
14 So that's now open for a second.
15 DR. GARBER: Second.
16 DR. SOX: Second?
17 DR. FERGUSON: Second.
18 DR. SOX: Is there a discussion or
19 modification?
20 DR. FERGUSON: The only modification
21 that I would recommend on that would be to state
22 the document be as it's approved, that it be used
23 as an interim document so that HCFA could move
24 forward in their process, that it be used as an
25 interim document, recognizing that it is dynamic.
.00269
1 And I would suggest that with the comments that
2 we're getting from the public and from the panel
3 members, that as part of the two-day meeting that
4 we have scheduled next, that we make this an
5 agenda item to revisit, at least at some point
6 during that two-day meeting, part of the comments
7 on this document.
8 DR. SOX: Why don't we get the wording
9 up there. And then it would be nice, if we could
10 get the wording up there, then you can suggest
11 how to --
12 DR. FERGUSON: It's essentially the
13 same thing except that it would be approved as an
14 interim document would be the only other addition
15 with that and that we specifically make it
16 revisited in the two-day meeting that's planned
17 next.
18 DR. GARBER: I guess I have a
19 question. I agree with everything you said, but
20 I take Ron's wording as meaning that it's interim
21 when he says it should be revised and revisited.
22 Is that acceptable?
23 DR. FERGUSON: As long as that's
24 understood, yes. I have no problem with the word
25 interim not being in there as long as it's
.00270
1 understood that HCFA's got something they can
2 move forward with now as part of the process
3 rather than having to wait.
4 DR. SOX: Would you like to say
5 something to the effect of a new sentence perhaps
6 that the panel shall consider possible revisions
7 to the document at its next two-day meeting or
8 something like that? Would that capture the
9 sense of what you'd like to have? That would
10 make it -- to do it --
11 DR. FERGUSON: Sure.
12 DR. SOX: -- as an agenda item. I
13 guess as the Executive Committee, right? That's
14 offered as a friendly amendment, Ron?
15 DR. DAVIS: Accepted.
16 DR. SOX: Any other comments or
17 additional amendments? It's now time for a vote.
18 All those who are in favor, please
19 signify by raising your hand. Hold it up so the
20 counter can tally the vote. It's unanimous.
21 MS. LAPPALAINEN: Except for an
22 absentee.
23 DR. SOX: We're now going to turn to
24 hear briefly from Jeff with some announcement
25 and benediction or something like that.
.00271
1 DR. KANG: Actually I believe these
2 comments were made for the public and will be
3 available outside. They were meant as opening,
4 and they're closing now.
5 I would just, Chairman Sox, like the
6 opportunity to reinforce and expand on HCFA's
7 preface to the subcommittee's, which has now been
8 the adopted subcommittee's recommendations as
9 amended. If people have that preface in front of
10 them, I'd actually like to refer to the third and
11 fourth paragraphs and just for the record read
12 them in. Actually now it's the current document
13 below.
14 We view the current document or the
15 voted-in document as a list of suggested topics
16 that should be considered and addressed to assure
17 full and consistent discussion of issues by the
18 MCAC panels. HCFA itself will not view this
19 report as a prescription of criteria by which we
20 are to determine coverage or even an absolute
21 standard by which we may judge the adequacy of
22 evidence.
23 In short, this document is a list of
24 suggested topics that the MCAC and its panel
25 should consider and address in evaluating
.00272
1 clinical evidence in rendering advice to HCFA.
2 Based on the advice in the record, HCFA will make
3 its coverage decision. We are confident that the
4 MCAC and its process will be an enhancement, not
5 a barrier -- the new document that you've all
6 voted in -- not a barrier to the fair and open
7 consideration HCFA will give to proposals for
8 coverage.
9 In summary, I think that we are
10 interested in how good is the clinical evidence,
11 what does it say, and what conclusions can be
12 drawn from it? And that's really what the
13 evaluation of evidence is all about.
14 Furthermore, as I stated in the fifth
15 paragraph of that preface, we are not interested
16 in asking the MCAC for advice on cost issues.
17 You are really the clinical scientific experts,
18 and that's what we're seeking your advice on.
19 Finally, with regard to coverage
20 criteria -- that's in the sixth paragraph here --
21 we are diligently working on publishing a
22 coverage criteria to further explain and
23 interpret what reasonable and necessary means in
24 discriminating cover from noncoverage services.
25 I actually do want to point out today
.00273
1 that today's effort deals with what is the
2 evidence, what does it say and what conclusions
3 can be drown from it, how we read it and how we
4 interpret it. That is distinctly different from
5 criteria.
6 Scientific evidence is in many ways the
7 yardstick or the measuring stick while criteria
8 is really how far you have to go, whether you
9 have to go one foot or three feet or ten feet to
10 get covered. The evidence really is the
11 measuring stick or the yardstick.
12 To further the analogy to our current
13 situation, HCFA could interpret in a rule that
14 reasonable and necessary means many things. For
15 example, we could interpret it as meaning just
16 safety, we could interpret it that a service has
17 to be safe and effective, or we could interpret
18 it as it has to be more effective, or we could
19 interpret it as benefits must outweigh the risks,
20 or it could be interpreted as being cost-
21 effective, or we could interpret it as being
22 cost-beneficial. And there are other variations
23 of the theme.
24 The point here is irrespective of what
25 we finally end up as criteria in the final rule,
.00274
1 it should not change your work regarding what is
2 good evidence, how do we read it, how do we
3 interpret it, what does it say, and what
4 conclusions can we draw from it? Thus, your work
5 is distinctly separate from our coverage criteria
6 and can certainly go on in the absence of the
7 criteria. Of course, in the final analysis,
8 today's work only guides your activity and your
9 advice, and HCFA will be the final decision maker
10 of what should be covered or not.
11 Now, I would like to take this
12 opportunity to briefly update you with where we
13 are on the coverage rule. On a personal note, to
14 my chagrin, I've now figured out why the agency
15 has struggled for over ten years to publish a
16 rule. However, the good news is that we actually
17 do have criteria in mind and a framework for how
18 they would be applied. However, it does raise
19 several operational and implementation
20 questions.
21 Given what is at stake and the
22 considerable interest in this rule, I am pleased
23 to report that we are expecting to publish soon a
24 notice of intent for rule making in advance of a
25 proposed rule. In this notice we will share our
.00275
1 current thinking and framework for coverage
2 criteria, how it would work, and we would also
3 raise some of the implementation questions that
4 we are wrestling with internally. Such a notice
5 will provide ample opportunity for the public and
6 other stakeholders or all stakeholders to have
7 adequate input and assist us in our deliberations
8 before we even propose a rule.
9 And on that, today is not about
10 coverage criteria. I thought I'd take the
11 opportunity to talk to you about coverage
12 criteria. But I would like to thank the advisory
13 committee, the Executive Committee, for all of
14 your efforts today to deal with, in a consistent
15 manner for all panels, how we read the evidence.
16 And I assure you we're working diligently on the
17 coverage criteria, and I believe that you are off
18 to a great start with regard to how we read and
19 interpret evidence.
20 DR. SOX: Thank you, Jeff.
21 Before we adjourn, for the record we
22 had one absence. Dr. Brook had to leave a few
23 minutes early. He left this note.
24 I am happy with the report. I would
25 like to see the revised Section 6, signed Dr.
.00276
1 Brook.
2 Is there anything else that we need to
3 do before we adjourn?
4 MS. LAPPALAINEN: Just to conclude
5 today's panel meeting, I'd like to remind you
6 that the next meeting of the Executive Committee
7 is tentatively scheduled for June 6th through
8 7th, the year 2000. Please call the HCFA
9 advisory committee line at 1-877-449-5659, which
10 is toll free, or for local calls, 410-786-9379,
11 and specify the Medicare Coverage Advisory
12 Committee, or you may check our website for
13 up-to-date information. And again, I'd like to
14 thank the committee.
15 DR. SOX: Before adjourning, I'd like
16 to point out that copies of Dr. Kang's remarks
17 are available on the table outside the door. We
18 want to thank everybody on the panel for their
19 hard work and the audience for their patience.
20 Thank you.
21 (Whereupon, at 3:40 p.m. the meeting
22 was concluded.)
23
24
25
|
