Ranch hand advisory committee




НазваниеRanch hand advisory committee
страница18/23
Дата17.10.2012
Размер1.21 Mb.
ТипДокументы
1   ...   15   16   17   18   19   20   21   22   23

DR. STOTO: I want in the summary table. Now new analyses.

DR. GRUBBS: Or the significant ones.

DR. STOTO: Right. So the effects that go into the summary table are the same; just a little bit more information about those effects.

DR. HARRISON: So a little more detail in the summary.

DR. STOTO: I wouldn't say detail, I would say more information. Because you're not talking about more    I'm not saying put the interactions in the summary or anything like that; it's just a little more information.

DR. HARRISON: Mike?

DR. GOUGH: I had some related question. I would like to see, for all the things, point out the significance between the Ranch Hands, the difference between the Ranch Hands and controls.

DR. HARRISON: In the summary table, now?

DR. GOUGH: In the summary table. Then whether or not the measured effect is outside the normal range of values, is it of any clinical significance in the judgment of somebody, and what are the levels that a physician will have to see before he would refer the patient to a specialist, prescribe medication or suggest alterations in diet or life style.

I would like to see this put on some sort of clinical indicator basis. That's a lot of work.

DR. STOTO: I think needs to be discussed, but it's a complicated issue.

Like for instance, if you did a very large study that showed that you can lower blood pressure by 2 millimeters of mercury, 2 millimeters of mercury doesn't make a difference to any given individual.

DR. GOUGH: But clinically it makes a difference to the person.

DR. STOTO: No, it doesn't.

DR. GOUGH: Your prescription will say   

DR. HARRISON: Not 2. He's saying --

DR. STOTO: 2 millimeters is within the measurement area.

DR. HARRISON: He's deliberately choosing something that's --

DR. GOUGH: All right. You mean there are error measurements in this stuff?

DR. STOTO: But if everybody in the population lowered their blood pressure by 2 millimeters of mercury, they would have a noticeable effect on cardiovascular mortality.

DR. HARRISON: If the truth be known, you probably would need several million subjects in each group in order to show that.

DR. STOTO: Right. But this is a point; that there's a difference between what's meaningful for an individual and what's meaningful in terms of a population.

DR. HARRISON: One of the questions that Mike asked resonates with me, and that is that you can see differences between groups when both groups are normal. And that's a point that I think gets forgotten sometimes.

DR. STOTO: I think that that kind of    my suggestion would give you the information you need to think about that, but wouldn't be in a position of dismissing things that might be important on a population basis.

DR. GOUGH: I don't care about population. I think this is --

DR. HARRISON: You weren't saying to exclude anything; you were saying you wanted it extended. You were saying that he's extended --

DR. GOUGH: I wanted it to be related to what it means to an individual.

DR. STOTO: I don't think that's the right perspective for analyzing an epidemiological study.

DR. HARRISON: Yeah. I --

DR. GOUGH: I do.

DR. HARRISON: Okay, well --

DR. STOTO: That information is in the report as it has been done in the past.

DR. GOUGH: It's very, very hard to find that information.

DR. HARRISON: So there's two levels that you're talking about here. One level, it seems to me, points out that there are population differences that are seen that all occur within what we think of as a normal range. And you can argue there whether there's any physiological relevance there.

But then you're asking, "Well, even if I show a population difference, is it sufficiently different that a physician would worry about it?"

DR. GOUGH: No, there it was not the population difference, but say    yes.

DR. HARRISON: Because you know    if we take the diabetics, those patients who have a two hour postprandial greater than 140, their physicians don't worry about that.

DR. GOUGH: Yes.

DR. STOTO: So this is not an issue for the dichotomous variables; it's an issue for the continuous variables.

DR. HARRISON: That's true.

DR. STOTO: And I think that if we said "Oh, well, we find a difference of so-and-so in a1C hemoglobin" but for an individual that would be meaningless, that would tend to dismiss the whole thing when that was inappropriate.

DR. MICHALEK: What you really want to know is were the results of the dichotomized continuous variables, what was the percent abnormal? So you have a percent, the number of abnormally high glucoses, for example. That has real clinical meaning.

DR. HARRISON: Right. You don't need to know how high, just that they are high.

DR. MICHALEK: And that's already in the report.

DR. HARRISON: Let me ask you this, Mike: Why should this study, with 2000 and some men in it, covering 20 years of experience, try and determine whether a value is significant    try to determine in its report whether a value has significance for an individual. They're already reporting this stuff back to the primary care physician. So if they find anything, the airman is being taken care of.

DR. GOUGH: Yes.

DR. HARRISON: What do you see reporting it in that way would serve?

DR. GOUGH: What I see is    if this study ever gets a lot of public attention again, there'll be newspaper reports saying they were elevated levels of so-and-so. And I would like for it to be right there, whether or not those elevated levels might have clinical significance or whether or not they're within the range of normal values. That's all. And I think that's important for the individual, too.

DR. STOTO: The range of normal values is something that applies to individuals, not to population studies.

DR. GOUGH: Excuse me?

DR. STOTO: The range of normal values is something that applies to individuals, not to population studies.

DR. GOUGH: It's what's tolerated as acceptable among the population, isn't it?

DR. STOTO: No. The range of normal values for blood pressure is 30-40 millimeters of mercury, something like that. Okay? If you found a 2 millimeter difference that you can show was statistical significant, on average, that would make a big difference to the population. And to say it all, 2 millimeters is within the 40 millimeter range.

DR. GOUGH: In this report it says normal blood pressures are 90 over 140 or below. If you find a lot of blood pressure variation, whether it's high in one group or another, but it's all    it's occurring, where the average in these two groups is below 90 over 140, I don't think that's a significant result in that group.

DR. STOTO: Well, it is. It could be statistically significant, and that's what's in the report.

DR. GOUGH: What does it mean?

DR. STOTO: It means that dioxin is associated with this health outcome.

DR. HARRISON: If one group's average diastolic pressure is 85, and the other group's average diastolic pressure is 89, on a population basis that will show increased strokes, increased myocardial infarction, increased incidence of renal failure.

On the other hand, if you go into the clinic and they use one of those automated blood pressure cuffs on you, you can get about a 4 millimeter mercury difference in an individual measurement. If you're in a parking lot and you walk up a ramp or two before you go into the doctor's office, you can get a 4 millimeter difference in mercury measurement.

If what you're after is to present these data or have a presentation of these data that would serve the purpose for the popular press, if this were to get out, that's not something that has to be included in the blue book. The popular press isn't going to look at the blue book anyway, with all due respect.

DR. GOUGH: No.

DR. STOTO: And I think that if you do what they've already been doing plus what we suggested just a moment ago, the information will be there for someone to make a judgment about that.

DR. HARRISON: I would like to see him include, or I'd at least like to see the notation made of whether the variations are still within the normal range.

DR. MICHALEK: That's not appropriate, and I think Mike has already expressed that opinion.

DR. STOTO: Normal range has to do with individuals. It would be inappropriate to talk about population averages in those terms.

DR. MINER: We've gone round and round --

DR. HARRISON: Actually, you know something, I'm    my sleep deprivation is starting to get to me. Really. I'm sorry, I didn't mean that, I take it back.

DR. MICHALEK: A summary that might be of use would be to summarize how many of those letters we give to our study subjects. After they leave the clinic, we send them a letter saying "These measurements were abnormal. Go see your doctor right away." We send them these letters to spell out all their abnormalities and give them advice about seeing their family doctor.

DR. GOUGH: That would be very useful --

DR. MICHALEK: Now, those letters, we can summarize in the document; tell how many we sent and for what reasons.

DR. STOTO: And how many of the controls, too.

DR. MICHALEK: Same thing; Ranch Hand and control. Would that help?

DR. GOUGH: Yes.

DR. MINER: The piece we're missing, I think, though, is    that we've gone round and round about, is a subclinical effect.

DR. MICHALEK: Exactly. We see things statistically that we've never seen in an individual.

DR. GOUGH: Well, I've just been told that there's no such thing as subclinical effect.

DR. HARRISON: No, no. On a population basis    that's why I'm sitting here embarrassed. On a population basis, of course.

DR. CAMACHO: But why did you say here's this whole pool of Ranch Hand people, they're units of analysis, each one of them; and we can't make an inference to the group as a population?

DR. STOTO: No, I didn't mean to say that.

DR. CAMACHO: All right. See, that's where    and you said you went round and round. That made my head go round and round.

DR. HARRISON: We're talking about sub --

DR. MINER: No, it was how to display a difference in the means on a continuous measurement when both were, let's say, within the normal range. Should we just blow that off and say "Well, that's within the normal range, it doesn't mean anything"? Or are we missing an important piece that might be a subclinical effect?

DR. HARRISON: Actually, though, within which normal range, though? Are you talking about within the lab's normal range, or are you talking about within the normal range that you can define for the comparison group?

DR. STOTO: I think that --

DR. HARRISON: Because if you're showing a statistical difference, that's saying then that that's not within the normal range.

DR. MICHALEK: Let's put it this way, if the Ranch Hand mean were outside the normal range, you would have at least 60 to 70 percent of the Ranch Handers abnormal. In other words, it is not appropriate to argue that these data are useless or not informative, because the means are within a normal range yet.

DR. GOUGH: I'm not saying they're useless.

DR. STOTO: I think that the subclinical confuses it, because when there's a small difference in the mean, that means that there's an increased proportion of the people who have the condition, however defined. And that proportion is important, because    maybe it's 3 people, maybe it's 30 people, have this condition, more than you would expect.

DR. HARRISON: But let's --

DR. STOTO: That's not meaningless to those people; it's--

DR. GOUGH: No.

DR. CAMACHO: So what you're saying is the control group has a curve that looks like this, and ours, while it's still in the mean, is shifted more by so many points.

DR. STOTO: Mike, can I draw something on the board here?

DR. CAMACHO: Fine, because if that's the case then I would say that you're right. If Group B is within the normal range but they're at the very high end of it,    is at the high end of this normal range, and Group A means yes.

DR. MICHALEK: That's right.

DR. STOTO: This is Ranch Hand and the other one is controls. So if you say the difference of the averages is pretty small, and that's well within the normal range, the difference of the averages. But if you make a cutoff over here, there's many more people in the Ranch Hand than the control.

It's the kind of variable that you measure with a continuous measure; you know, the statistics have to do with this, but there may not be a natural cutoff, but whatever cutoff you choose out here in the tail there's more people with the condition than --

DR. HARRISON: Now Mike, if you were to draw another standard curve that was wider --

DR. CAMACHO: You mean broader?

DR. HARRISON: [Going to easel] Let's suppose that because of the way life works, this is the normal. That's what comes with the kit to the lab and so on, okay? Because this is what I think some of us are talking about.

So in this case, both of these are within the normal    normal for the United States population, but if you define normal by the comparison group, then there's no confusion; they're abnormal in both cases. In other words --

DR. GOUGH: You mean both   ?

DR. HARRISON: In other words, this is abnormal because normal range has been redefined by the range of the comparison group. If we accept the contractor's laboratory normal range, we may think that we're talking about stuff that's just fluctuating --

DR. STOTO: But that has to do with a discrete analysis.

DR. HARRISON: No, this is a continuous analysis, right?

DR. STOTO: But you would only use normal range from either source when you have a discrete analysis. Where you first say, "Does this person have an abnormal value?" And then say "Is there a greater proportion of people with abnormal values in the Ranch Hand as compared to the control?"

DR. HARRISON: If you take the serum calcium, serum calcium in most labs runs from like 8-1/2 to 10-1/2 whatever, milliequivalents per deciliter.

DR. STOTO: Okay.

DR. HARRISON: If you were to analyze the comparison group, let's just say for example it would turn out that the normal range for the comparison group was from 9 to 10.2.

DR. STOTO: Well, see, there's two ways to analyze a laboratory value like that. One is, you say "What proportion of the people have high levels?" And high could be define different ways depending on what you decide is the normal population.

And then --

DR. HARRISON: No, what you do is you measure all these calciums, you plot them on here, and what falls between whatever it is, 97.5 is normal, and what falls outside of that is abnormal. That's the way most laboratory values are defined.

DR. STOTO: But the way that it's going to be done here, if I understand it, is you figure out what's high, what number is high, and there's a number of ways we discuss of doing it. But then you say, "Are more people high in the Ranch Hands than in the controls?"

DR. HARRISON: I didn't think that was how we said we were going to do it. I thought --

DR. STOTO: And that corresponds to this, and you might be saying "Well, maybe this should be set over there rather than here," but in any case, that's a discrete analysis, and then you just report the proportion for --

DR. HARRISON: See, the thing that's worrying me is that when a laboratory gives you a normal range, they've got the entire population that they're talking about. We're talking about a subset of the normal population, because none of these guys are younger than what --

LTC BURNHAM: Average is 60.

DR. HARRISON: So you can't use the SGOT, the SGPT, you can't use any of those values just out of hand from the --

DR. STOTO: I understand that, and we deal with that--

DR. MICHALEK: We already talked about that.

DR. STOTO: We already talked about that, but that's dealt with in terms of defining what the cutoff value should be.

DR. HARRISON: Those are defined by the comparison group. DR. STOTO: We talked about doing it both ways, and I'm happy with either one.

DR. HARRISON: Okay.

DR. HARRISON: But they both result in, either people are above it or below it. They result in a discrete outcome.

DR. GOUGH: And then you tell me that none of these numbers have any significance, right?

DR. STOTO: No, I never said that.

DR. GOUGH: Well, what I was saying is that there may be a number whereby above it a physician would pay attention, below it a physician wouldn't.

DR. STOTO: Right.

DR. GOUGH: You draw just such a thing on a curve.

DR. HARRISON: Let me actually make a suggestion here, that's actually not mine, but I won't reveal the source because he might be stoned.

This is one of the areas that elicited some of the more heated comments, has elicited some of the more heated comments in past meetings. Am I not correct?

DR. GOUGH: Oh, yes, that's correct.

DR. HARRISON: And yet there really is very little change in it. And the committee has suggested stuff before. I mean, you're listening to it all over again here.

I think we need a rewrite. I think instead of us trying to go word by word through say the summary tables here, I think what we should say is, "We think you need to rewrite this and send it back to us."

DR. STOTO: Well, the issue about what to use for the cutoffs is dealt with someplace else other than paragraph 3.7.3.

DR. HARRISON: I'm not disagreeing with that; you say it's in there, and    okay. But this    still, how this gets presented is something that has    in fact, didn't Landrigen have concerns about it as well and   ?

DR. MICHALEK: Let me try to --

DR. HARRISON: What I'm suggesting is --

DR. MICHALEK:    you would like a rewrite. We'll do a rewrite and I'll e-mail it to all.

DR. HARRISON: Make it a little more explicit, too.

DR. MICHALEK: There's a fundamental conflict here that we've touched on many times. The conflict goes like this: We as statisticians and epidemiologists are trying to test the hypothesis that two means are the same. Fundamentally that's the hypothesis we want to test.

Or else, we want to test the hypothesis that two proportions of disease are the same in two groups. And we do that by testing the hypothesis that the relative risk is one. Okay, that's what we do. That's all we do in this 3,000 page report, is we test hypotheses. We are not in the business of deciding whether a particular individual is sick and whether he should go see his doctor. That's a different line of thought, and that's all handled through correspondence with our study subjects.

There's a fundamental conflict there between testing hypotheses and drawing inference about parameters and populations, and making decisions about individuals. They're two fundamentally different ideas, okay?

DR. HARRISON: I don't think there's any disagreement about that.

DR. MICHALEK: They're not saying that either one should be done; they're both done in the study. The report is summarizing statistical summaries of these analyses of association.

DR. HARRISON: I'm getting back to the suggestion that Mike kicked off --

DR. MICHALEK: Which is?

DR. HARRISON:    about changes in the --

DR. MICHALEK: Effects side. Show me the change. I agree.

LTC BURNHAM: Is that the only way you want it to be rewritten? We need some guidance.

DR. HARRISON: You guys have had guidelines.

DR. MICHALEK: I understand what Mike is saying; I made a note of the change, and I'm going to change it in the text.

And you will see that in the revision, which will come in a few weeks.

DR. HARRISON: Yes, Paul?

DR. CAMACHO: 3 1, the statement of opinion. I take it those are end notes? Or they're just collected up there.

DR. GOUGH: 3.7.

DR. CAMACHO: Yes, 3.7.3.1, Statements of Opinion.

DR. MICHALEK: Statements of Opinion --

DR. CAMACHO: I'm writing along, I know your syntax and the language and everything, and I'm writing along and whether I use a certain adjective or adverb has all this    so at the end of that sentence you're saying we're going to get bang, an endnote number, if he's implying an opinion or he's leading to an opinion; and then that's what you're saying here, right? Is that what you're saying? And then at the end of each chapter.

I mean, are they going to just sit out there? They're going to be refer to special sections, end.

DR. MICHALEK: A discussion section, that's where the opinion will go. We will make no value judgments about the findings until we get to discussion, then we exercise value judgments, all in one place, separate from the data.

DR. STOTO: But it will be part of the text but clearly labeled.

DR. MICHALEK: Yes. It will be labeled Opinion or Discussion.

DR. CAMACHO: And you're telling the story all over again from the opinion side.

DR. MICHALEK: Right. Just like in an article.

DR. CAMACHO: It's not an endnote.

DR. MICHALEK: It's written as in an article.

DR. CAMACHO: Okay.

DR. MICHALEK: Introduction, Methods, Results, Discussion.

DR. HARRISON: And just to back up one more, the clinical discussion, you've already mentioned it once; but you're saying explicitly now the contractor is going to provide 12 experts. That should be interesting.

[Simultaneous discussion]

LTC BURNHAM: And a medical editor.

DR. MICHALEK: And a medical editor smoothing out the writing.

DR. HARRISON: Well, rather than    it almost stands to reason that you're going to have a pulmonologist write the pulmonary chapter; but the point we were making was that it shouldn't just be someone who is just a pulmonologist, it should be someone with demonstrated expertise. That would mean to me that you want someone who is published and who has shown ability to get his or her writing through review.

- What about these other associations between covariates and exposure group or dioxin. Tabular results of all    tabular summaries of all results. I mean, when these things start coming out, you guys are going to spend two days wrangling over how this stuff is presented.

There's no other comments to be made on this by the committee? By the statistical experts?

DR. STOTO: I understand that has to do with things like the association between BMI and cardiovascular illness, which is not the focus of this study.

DR. HARRISON: But a covariate issue.

DR. STOTO: These are covariates; these are not exposure or outcome. These are things that --

DR. HARRISON: Okay, what about Ranch Hand initial dioxin analysis summaries? Summaries shall contain: A) continuous, B) binary, 2, polychotomus?

LTC BURNHAM: Polychotomus.

DR. HARRISON: Polychotomus.

LTC BURNHAM: Excellent, good, fair, poor. There.

DR. HARRISON: A country boy from Mississippi can't hardly even talk in this   .

DR. HARRISON: It's another new word for you; you learned dichotomous.

VOICE: Trichotomous.

DR. MINER: Just wait until tomorrow.

DR. STOTO: I think that this whole thing here basically says they're going to do it the way they've been doing it, at least in the previous study --

DR. MICHALEK: Unless you tell us something different, yes.

DR. STOTO:    and we didn't have general problems with that the last time around.

DR. GRUBBS: That's the road map, if you will, for the 171 page Appendix G that gets every number, seemingly, in the report, summarized in one appendix. And that just specifies what should be done. Included in that appendix.

DR. GOUGH: But you are going to redo 3.7.3?

DR. MICHALEK: Yes.

DR. HARRISON: 3.9.1. I think you all should start using eTickets. Seriously, that's transportation.

You don't want to be delivering these tickets to these guys, do you?

MS. JEWELL: They want us to use eTickets, and I'd be glad to send you them, but if you want to make a change at the last minute, you're not going to be able to do it. We've had our travelers stuck because they have to stand in line and go to the original electronic issuing person, get a piece of paper to allow them to go someplace else to re-book themselves.

DR. STOTO: You can't change airlines.

MS. JEWELL: Exactly. I'd be glad to do electronic; it's a lot easier.

DR. MINER: It was still applicable.

DR. HARRISON: I was wondering why you were getting off down there   

(Laughter)

DR. HARRISON: Okay. Does anybody -- Oh.

Addendum A, General Physical Examination. No. 8.

What are PBs per minute?

DR. MICHALEK: Premature beats? Or --

DR. FAVATA: Where's, is that related to your respiratory rate?

DR. HARRISON: No, no. You're telling me that these are premature beats per minute? That are assessed in the physical exam?

DR. CAMACHO: Where are you?

VOICE: Sixties.

DR. HARRISON: Addendum A, General Physical Examination. No. 8 says PBs per minute.

DR. MICHALEK: We will find out, and we will spell it out.

DR. HARRISON: Addendum A.

DR. MICHALEK: The first addendum.

DR. GOUGH: What table is it, Joel?

DR. MICHALEK: It's in Table 2.

DR. HARRISON: I hear a suggestion that there should be a glossary.

DR. MICHALEK: Or else no abbreviations at all.

DR. HARRISON: Yes, that sounds    so we agree that, one or the other.

DR. MICHALEK: Right.

DR. HARRISON: On the EKG --

MR. COENE: The glossary still would be a good idea only because you will probably use the abbreviations somewhere else.

DR. HARRISON: Or you can always define clean.

DR. MICHALEK: A glossary. We'll add a glossary.

DR. HARRISON: Under Addendum A, Item 6, electrocardiogram, where it says follow the last sentence, it says follow the mounted tracing and rhythm strip if obtained, to the diagnostician. I just suggest that you should indicate that you're trying to follow the reading from the cardiologist as well. Or if those are going to be back in time.

DR. MINER: That's not the one.

DR. HARRISON: Am I reading off the wrong one?

DR. MINER: You may be looking at a slightly older version.

DR. GOUGH: This is the one we got yesterday.

DR. HARRISON: No, no, this is one I did at home.

DR. MICHALEK: That's what mine says.

DR. MINER: Then you edited the wrong one.

DR. MICHALEK: No, I didn't touch that section.

DR. FAVATA: What section are we on?

DR. HARRISON: I'm on addendum A, Item 6, electrocardiogram, and all I'm just saying is, they say they're going to send the tracing to the diagnostician. Earlier they say they're going to have the tracing read. Somewhere they said they were going to have this read.

DR. FAVATA: Yes, will be interpreted by the cardiologist.

DR. HARRISON: And I'm just saying that the tracing and the reading should go to the diagnostician. That's kind of nickel and dime stuff.

DR. MICHALEK: Tracing and the interpretation.

DR. HARRISON: And the interpretation.

No. 11, doppler testing of peripheral pulses. Procedure will be conducted after a minimum of four hours' abstinence from smoking. Isn't it customary to have abstinence for a certain time from caffeinated beverages as well?

DR. MICHALEK: We apparently have two or three different versions of this. Do your versions say anything about abstinence from caffeine, on the doppler?

Bill Grubbs.

DR. GRUBBS: I'm looking.

DR. STOTO: The part about the travel here says that you have to give coffee to the wise but only water to the men.

DR. MICHALEK: No caffeine?

DR. HARRISON: So it's just as soon that they are caffeine-free for their three day    I can't believe that.

DR. MICHALEK: Oh, no. I think it just doesn't say anything; that's all. It says nothing.

DR. HARRISON: Well, you know, it says they can't smoke, and I'm just wondering why it doesn't say they can't have --

DR. MICHALEK: That could be an oversight on our part.

There's no mention of caffeine?

DR. MINER: Not in either of the versions we have. But I don't know if that's the --

DR. MICHALEK: Are you suggesting that we should abstain from --

DR. HARRISON: Yes.

DR. MICHALEK: For four hours?

DR. HARRISON: I think that affects the pulses as much as smoking, or more.

DR. STOTO: For four hours before that test?

DR. MICHALEK: Four hours is sufficient?

DR. HARRISON: I think four hours probably is.

DR. STOTO: And Doppler, as well?

DR. HARRISON: Well, actually I was referring to the Doppler.

DR. MICHALEK: What about EKGs?

DR. HARRISON: Okay, EKG it says, abstinence from smoking, food, and liquid intake. Unless you're going to chew on instant coffee.

We already talked about measurement of height and weight. You want the measurement of the hip. And I think that's really defined as just the hip at its widest.

If you all are going to draw dioxins again, and I guess this is probably the same protocol you have been doing all the time; but it's probably a good idea to label that bag when you're starting to use it, not at the end. I mean, if people get distracted or something --

DR. MICHALEK: Yes, there's some text I forgot to edit right there.

DR. HARRISON: There's 4.1 --

DR. MICHALEK: Section 14.

DR. HARRISON: Addendum B.

DR. MICHALEK: Oh, I'm backing up to Section 14, Laboratory Procedures. There's a paragraph in there about 650 for dioxin --

DR. GOUGH: I wondered about that.

DR. MICHALEK: It's not going to be 650.

Now you were on Addendum B.

DR. HARRISON: Although Addendum B, 4.1 in the procedure, after    basically it's saying that you draw the blood, you clamp the tubing, and then you label the bag.

DR. MICHALEK: We would probably label the bag first.

DR. HARRISON: I would say that surely it's somewhat--

MR. OWENS: We have state laws that govern the protocol, in California, and it's done according to that. It's labeling first.

DR. HARRISON: Okay. I would agree with the state law.

DR. GOUGH: Will there be electricity during most of these exams?

(Laughter)

DR. MICHALEK: Let's hope so.

DR. HARRISON: For some reason, I have a note    I have a note that you need a better description of your computer, and the modem is missing from the specs.

DR. MICHALEK: We will edit that computer description.

DR. HARRISON: I guess the thing that I    first of all, I don't know what ADPE is.

DR. MICHALEK: That's old-fashioned Automated Data Processing    that's 1970s technology. We'll drop that.

DR. HARRISON: At any rate    it doesn't matter a whole lot, but you did say that you were going to need a modem and I could see where you'd want a modem to get back and forth to Brooks. And we just didn't have it down here and I just --

DR. MICHALEK: We'll fix that.

DR. STOTO: Or were you going to do it over the Internet.

DR. MICHALEK: We need the Internet connection.

DR. HARRISON: Well, either that or you need a network connection; one way or another. A network card.

Anybody else have anything?

DR. FAVATA: Just respiratory rate on general physical exam.

DR. MICHALEK: Where is that?

DR. FAVATA: Page 60.

DR. MICHALEK: Respiratory rates?

DR. FAVATA: Yes.

DR. MICHALEK: Are you saying that should be added or?

DR. FAVATA: I don't see it there, it should be added.

DR. HARRISON: Is there any difference in the temperature you get with that little ear thing? Isn't that the same temperature you get orally?

DR. FAVATA: I imagine they've done studies.

DR. HARRISON: Okay, Agenda and Future Meeting Dates.

What is the committee's will?

DR. GOUGH: I have a question to Joel about the addenda. A number of places in the last couple days you've said "I'm going to take care of that" or "I'll rewrite that." When should we expect that and when do you have to have it back?

DR. MICHALEK: Four weeks.

DR. GOUGH: The turnaround?

DR. CAMACHO: Oh, you want it from me in four weeks.

DR. MICHALEK: Well, no. We want something from you in four weeks. When do you think we should --

DR. MINER: The RFP is supposed to go out on the 2nd of March, is that right?

VOICE: That's six weeks.

DR. MICHALEK: So our deadline is it has to go out, so to speak, on the 2d of March. That means we have to have a revision -- we have to have something pretty much firmed up within four weeks. It means I have to receive feedback from Paul Camacho, I have to put all these changes in that you suggested, and get it all cleaned up; and what I could do after I get your feedback is put that in and then send the whole thing back to the committee. And then we're going to need a quick turnaround from you within two weeks.

DR. STOTO: When you do that, would you highlight the changes that have been made since this draft?

DR. MICHALEK: Sure.

MR. COENE: Just one set of changes at this point.

DR. STOTO: I'd say everything that's been changed up until now; treat that as fixed.

MR. COENE: Except --

DR. STOTO: The things that were discussed at this meeting, and you said you would change. Highlight those.

LTC BURNHAM: That changes Paul's deadline.

DR. MICHALEK: So anything Paul gives me will be in a different color font.

DR. CAMACHO: I'm going to try to    you're not expecting anything dramatic unless I check with my friends and somebody trips over somebody, somebody says something. Bing. So I'll try to do it within three weeks.

DR. HARRISON: You're talking about the Psych section, right?

DR. CAMACHO: Yes.

DR. HARRISON: So nobody's going to disagree with Paul about the Psych section. So there's no sense in making it different from the rest of the --

DR. CAMACHO: I just want to check with a couple of people.

DR. HARRISON: I understand. But I don't see where we're going to --

LTC BURNHAM: So if we sent next version out within two weeks; and we got it back within two weeks of that, then we'd have it in a month.

DR. HARRISON: Well, why don't we    why don't you send it out with one of those    well, how do we want to do this? We want everyone to send suggestions back directly to the Air Force?

DR. MICHALEK: You've already given us. You mean on the next revision?

DR. HARRISON: In other words, what we're saying is, maybe something got misunderstood and missed in the translation.

DR. MICHALEK: Right. So the attitude could be "Well, here it is, are there any objections?"

DR. STOTO: Ideally we should do that through the staff. But is there going to be a staff?

MR. COENE: Barbara and Len.

DR. GOUGH: You mean send it to them before we're gone?

DR. STOTO: Yes.

DR. HARRISON: So --

DR. STOTO: Or at least copy them.

DR. GOUGH: Yes. I agree.

DR. HARRISON: So we're going to    you're going to send a corrected copy back to staff in two weeks.

DR. MICHALEK: Yes.

DR. HARRISON: Staff is going to then Fedex some more CD-ROMs out to us with a two week deadline. If we find something that we don't like, we get back to you.

DR. MICHALEK: And if you say nothing, that's mean's it's approved.

DR. HARRISON: If you say nothing --

DR. MICHALEK: That's it.

DR. HARRISON: It means that everything is just hunky-dory.

DR. MICHALEK: It's a done deal.

LTC BURNHAM: And whoever wants hard copy should tell Len now rather than    you know, if you had CD problems or --

DR. HARRISON: Well, I'm not sure those weren't problems with the    I don't know.

DR. MICHALEK: We think it was a problem with our CD cutter, our CD burner. That's why, we think that was what happened.

DR. HARRISON: Well, mine wasn't readable, either.

DR. MICHALEK: Understood. We can get around that.

DR. HARRISON: I'm not sure that it's any big deal to just send it as an attachment.

DR. GOUGH: That's how I got mine.

DR. HARRISON: Because it's not huge.

DR. MICHALEK: No.

MR. COENE: It's 2 meg. Some e-mail --

DR. MICHALEK: Won't take that big a file? I could split it up into two files. I could put the main statement of work and all the addendums as a separate file. That gets it down to 1 meg.

MR. SCHECHTMAN: You can make two attachments to one e-mail.

DR. STOTO: No, that may have the same problem.

MR. SCHECHTMAN: Then have two separate e mail messages; Part A and Part B.

And then everybody can use Microsoft Word?

DR. HARRISON: yes.

MR. SCHECHTMAN: It's a common format.

DR. HARRISON: Why don't you do this? If you're going to send them directly, why don't you send them directly and why don't you send a separate e mail message saying "I just send you your file, let me know if you didn't get it."

DR. MICHALEK: For that I need a complete and current list of everyone on the committee and all their e mail addresses. I do not have that.

MR. COENE: It's on the Internet; it's on our web site.

VOICE: Turnabout   

(Laughter)

DR. STOTO: Where is the web site?

MR. COENE: It's nctr.fda.gov, and you go in there and there's a thing about committees, and you'll find both committees and you'll find the committee makeup.

LTC BURNHAM: We're just sending them to Len, though, Joel.

DR. HARRISON: No, He's going to send them directly.

DR. MICHALEK: I'm just repeating back what he told me.

DR. HARRISON: That's what I was suggesting. If you're going to be sending CD-ROMs and the CD-ROMs are going to    that might make sense, but if it's e mail messages, I don't see why e mail messages have to come here to go, come to me.

MR. SCHECHTMAN: Everybody, plus us, will all be included on the e mail message, right?

DR. MICHALEK: Right.

MR. SCHECHTMAN: Then everybody will wire back their comments to you. That would be the most efficient way as a reply to that e mail document. Then you'll include all those comments, adjustments, changes, and finalize it.

DR. STOTO: I think that we need to copy you.

DR. HARRISON: I'm    yes, I   .

MR. SCHECHTMAN: Right; cc everything for the record, to us.

DR. HARRISON: Just because some of us are forgetful and some of us aren't here, make sure in your message that you spell out what we're supposed to do. Reply to all, da-da-da, all that sort of stuff.

DR. STOTO: What are we going to do now, and tomorrow?

DR. HARRISON: Well, the question of the timeline for report writing has come up, and Mr. Grubbs wanted to say something about that. That is part of our agenda here.

DR. GRUBBS: The Air Force asked me to just briefly give an overview of how we put together these blue books or the entire report.

DR. MINER: Wait a minute. We lost --

DR. MICHALEK: Should we stop or should we wait?

DR. MINER: Wait.

DR. HARRISON: Well, we've lost two-fifths of the committee here.

DR. MINER: Well, this was a big area of contention last time, and we want to give adequate time for --

DR. HARRISON: Well, I agree.

DR. MICHALEK: How about a five minute break?

DR. HARRISON: Well, yes.

[Recess]

DR. HARRISON: You can go ahead.

DISCUSSION OF REPORT WRITING

DR. GRUBBS: The Air Force asked me to give a brief summary of our report writing process, the blue books, if you will, and how do we go about accumulating all these blue books. It resulted in about an 1800 page report.

So I've got a basic layout here; it's not exact for all 20 chapters or some huge Gantt chart there, but I'd like to go over that. And this is also reviving the review of the report for the '97 report, and also giving the new members some idea of the process it goes through.

In '97 we put together the report of 20 chapters, 10 of which were clinical. It was 1,790 pages in length. The clinical chapters, there are ten of that we've gone through today, they were from 57 to 235 pages in length, gastrointestinal, if you'll remember, was the big one there.

We had six appendices. In each analysis, we did four statistical models, both an adjusted and an unadjusted analysis, and there were a total of 266 endpoints. So when you put all the things together, this is an overview of what you came up with.

DR. STOTO: Things like diabetes, where you measure it a couple ways and some continuous, some discrete, that's --

DR. GRUBBS: That's Two. That's two endpoints.

DR. STOTO: Well, it's more than two for diabetes, because it's --

DR. GRUBBS: Diabetics are very different. Right. Any analysis,

DR. STOTO: Ten or twelve.

DR. GRUBBS:    really any table, if you will   .

[Slide]

Just to quickly go through the different chapters, these 20 chapters, the first column here and through the covariate associations we call, if you will, the up-front chapters. They're not the clinical chapters. They give an overview of the questionnaire methodology, the physical exam methodology, a summary of which participants attended by group, and reasons for refusal. Joel can comment, he's important in the bias, examining the bias of the participants coming to the study. Quality control methods, the statistical methods section and covariate associations.

Then we started    that went through Chapter 8, I believe, and then Chapter 9 was the general health assessment. We stepped through ten clinical chapters; the committee reviewed all of them, but the majority of the time was spent on the clinical chapters, starting with General Health.

General Health, Neoplasia and Neurology/Psychology, Gastrointestinal, Cardiovascular, Hematology, Endocrine, Immunology, Pulmonary. This time there will be two additional chapters that were done in 1992, the Dermatology and the Renal, for a total of twelve clinical chapters.

At the end there is a Conclusions chapter, which summarizes the results found across all the clinical chapters. There was a one-page Future Directions; Future Directions may be how do we dispose of the data or transfer the data, this being the last follow up; and then appendices that went along with it. Summary tables, the Addendum A and Addendum B that you just looked over, they were part of the report. The Examiner's Handbook, copies of forms, things like that.

[Slide]

Each clinical chapter typically would include the following: A review of the literature, then we'd summarize previous Air Force Heath Study results in the next section. We'd put which endpoints were to be analyzed, the statistical methods that we used. We would go through and do this statistical analysis: We'd have a clinical discussion of the findings, a statistical summary, and we've talked about summary tables of the findings, and then an overall conclusion.

So those are some of the general big sections you'll see in a chapter.

[Slide]

To put together a report of this magnitude in a    I don't want to say fairly short time, but a limited time, what we have to do to get this going is get chapters going simultaneous. I may have three analysts working for me, working on three different chapters. They may be in different stages. Last time I started one of my analysts off on the gastrointestinal chapter right at the very beginning. We didn't get everything done until April or May of the next year; it was just so big.

So we'll have these chapters staged, and they'll be going in stages and delivered to the Air Force in stages for their review. We call this the rolling report. At the last cycle we delivered two drafts of each chapter, and each appendix before, we even got to the draft final report. And that process took 13 months. It started on 26 August '98 and went through 26 September '99. Add five years to each of those and you get a general time frame of when it's going to be happening this time, 2003, 2004.

There's multiple chapters going on at once, there's many functions within the chapters going on at once, and what we've done here is lay out a chart, if you will. I just did three chapters here and I just said "Okay, three medium-sized clinical chapters" and this is just a bogey, if you will, as to the process that we'd use.

First chapter, we might start building the analysis database    might take us a couple weeks. Then six weeks or so to analyze the chapter, have them go through our editor, another couple weeks; give it one final review before we deliver draft one to the Air Force. Then give the Air Force two, three weeks to review it, that's the comments; and then get their comments back, resolve them, and deliver draft two, which might take another two to three weeks there.

As you can see, I laid that out for the first chapter, but you're seeing the second and third chapter going on simultaneously. So one chapter might be in editing and the other we might be starting to build the analysis database.

I personally put my hands on each of the chapters; I don't do all of the analysis for the chapter, but I read over everything before it's sent to the Air Force, and I can be reading multiple chapters within the same week, at different stages there.

[Chart]

So we call this our rolling schedule. I laid it out here for three chapters; you can imagine it for 20. I've got that over eight or nine months, but tried to visualize that, 20 chapters over a period of 13 months.

One of the main purposes of this is we need to involve the committee and get your review also. In cycles before 1997 we had only delivered one draft to the Air Force. We delivered draft one to the Air Force, we got comments, and we resolved comments and then we put it off to the side; we didn't deliver a second draft back. They would get that in the draft final report.

What we found last time, and it served a number of purposes; but I think it helped the advisory committee to see the Air Force's comments incorporated before you all even saw the report; and that's why we did a second draft.

So the first appropriate time for the Air Force, for the Advisory Committee to get involved would be after this delivery of draft two. I say that from a formal standpoint; the time for the Advisory Committee to get involved, we're doing it of course yesterday and today in terms of getting your comments in the statement of work.

So we have a document and a road map and a direction to go in. And that's why I'm very interested in these discussions here, to know what the statement of work is going to look like for three years down the road here.

But one of the options for the Committee is to get involved right after the second draft here, on each of the chapters. Now again, imagine this with 20 chapters culminating in a draft final report, 13 months past when we start. That it would be another opportunity for the Advisory Committee to get involved is, okay you've got the draft chapters and now you've got all the draft chapters, you've got the entire report, or what it's going to look like.

There were some comments last time that "Well, to make a final assessment on gastrointestinal, I may need to see the endocrine chapter, or some chapters interrelate, or I might come up with some ideas as to well, could we look at the relationship between cardiovascular and diabetes, for example. So that's a point to be made for seeing all the chapters, and then making a final assessment.

What I'm here to emphasize is, speaking for SAIC, we can do it however it needs to be done, but we need to know when the Advisory Committee primarily would like to become involved in the process so we can lay out the schedule, the overall schedule as to    really comes down to the endgame; how do we take it from the draft final report to the final report?

For example, if the committee agrees that they can make their comments on the General Health chapter after draft two, getting back to the Air Force getting back to us when we do the draft three, if you will    that's one option to do, and then we're stacking those up as we come to the draft final report.

If the committee decides "Well, I need to see the entire report and I can't make a definitive comment until the draft final report comes out" that's another option. However, that adds more time at the end of the report between the draft and the final, in case additional analysis needs to be done or things need to be reworked.

So SAIC has done it, worked with the Air Force both ways in this, and mainly we wanted to lay out the process and say "Here are opportunities for the Advisory Committee to review" and had some discussion on that.

DR. STOTO: Can I make a suggestion on that? This has been very helpful. I think it would work the best if there were three meetings; one where we reviewed about half of the chapters, draft two. Another one where we reviewed the other half of the chapters, draft two, recognizing that the first batch may be more weighted toward the front of the book stuff, and so on. And then a third meeting where we reviewed your response to our first comments and how the overall thing fits together.

DR. HARRISON: As I recall, the problem and the complaints from the last cycle was that the time frame was compressed enough so that we got a lot of draft one's, I think. And we found ourselves correcting stuff that was then    it was almost like the Statement of Work business this time, where you had two drafts out and that was problem.

Is it feasible within this schedule to have the three meetings like Mike suggests? Is that doable?

The other issue is, there has to be a certain time elapsed from the delivery of the chapters to when we meet so there's opportunity for us to actually do our review, so we actually have to build that in.

DR. GRUBBS: That's correct. If there was a meeting scheduled right here, this chapter might not make it into that, but previous ones would, for example.

DR. HARRISON: Right. Right.

DR. GRUBBS: That's correct.

DR. HARRISON: So is that --

DR. GRUBBS: And if I may add, I believe that was the intent back for the '97 study. The timing just didn't work out. I think there had been an attempt to schedule meetings in May or June and August to the first review that you would talk about. And for some reason, the timing just didn't work out, because their meetings ended up being in August for about half the chapters, and then October for the draft final report. There really should have been three meetings.

DR. STOTO: We really had the first and the third of the meetings I spoke about, and the first one late.

DR. GRUBBS: Yes.

DR. MICHALEK: I think the problem has been, when you decide up front we're going to have a meeting in September, you always have problems trying to get this committee to meet in a particular month, because everybody's busy. And so you try and then you fail, and then so the meeting slips a month or two, and that's --

DR. HARRISON: I agree; I think, though, one of the advantages that we have this time, and I don't think we've done this like this before is that we're talking about three years--

DR. MICHALEK: Down the road.

DR. HARRISON: So we can really start to plan meetings    hopefully I won't be here, but we can start planning meetings now for when   .

One other thought crossed my mind, and that is, am I not remembering that there is essentially software now that's used to do collaborative writing? You know, you all are doing Microsoft Word, but as I recall, I remember reading packages where you're the author so the author's copy is locked, but the collaborator's copies are open to change and somehow the software reconciles those changes?

MS. YEAGER: You can do that in Word as well.

DR. HARRISON: I guess I might have to get Word. I'm an anything-but-Microsoft person.

But that might be something else to consider; and that is, instead of pulsing the chapters, that we'll work out the assignments and the chapters will roll; when you deliver draft two, you'll deliver draft 2 to the Air Force and to the Committee reviewers.

DR. STOTO: Is that in addition to the meetings?

DR. HARRISON: Yes. In addition to meetings. I'm not suggesting    in fact, I'm saying that we're starting to plan well enough that we should be able to have the meetings and have them in the right time frames with lead time.

That's very helpful, thank you.

Any other comments on scheduling this kind of analysis?

No. Surely we can get a few consultants back during that time.

All right. What else? Can we quit for today?

DR. STOTO: When would be the next meeting.

DR. HARRISON: Here's what I was going to suggest.

DR. MICHALEK: When and where?

DR. HARRISON: First of all, let me make this suggestion: Why don't we discuss that in the morning?

DR. MICHALEK: Great.

DR. HARRISON: The only person who for sure isn't supposed to be here is Dr. Favata, who is not coming to the next meeting. Look at the smile on her face.

VOICE: Mary's not going to be here, either.

MS. JEWELL: It's been a delight.

LTC FOX: And I won't be able to attend; I've got a research conference that begins tomorrow morning.

DR. HARRISON: What kind of research conference?

LTC FOX: Gulf War Veterans Illness Research Conference.

DR. GOUGH: This has to be relaxing compared to that.

DR. HARRISON: That sounds --

DR. STOTO: What do we have on the agenda for tomorrow?

DR. HARRISON: Well, first of all we have to    and so I don't see any reason for us to beat ourselves over the head on when our next meeting date's going to be when we are going to be here tomorrow, and have something to discuss.

DR. STOTO: That's fine.

DR. HARRISON: I'm not opposed to doing it now, if you all want to. It's just, to walk in here to sit around for a few minutes to see if Ms. McCarthy, is it, shows up. You know    it'll save something for us to do.

There is that and there's something else that you mentioned.

MS. JEWELL: Writing the letters.

DR. HARRISON: Yes, I'll give a try at    or Len can give a try at putting a draft letter together. He took all the notes; and then tomorrow morning we can sit here and shove that around a little bit.

DR. STOTO: Okay.

DR. GOUGH: The other thing is, is the Air Force clear about the recommendations we made?

DR. MICHALEK: I'm clear.

DR. GOUGH: All right.

DR. MICHALEK: I don't think I'm confused.

DR. HARRISON: Also the Statement of Work is going to be in your hands in two weeks. And if it turns out the Joel has misunderstood something, I'm sure that   .

Is that satisfactory to everyone, that we will meet here tomorrow morning? What time.

DR. STOTO: Say 11.

(Laughter)

DR. MICHALEK: Same time?

DR. HARRISON: 8:30. Meet here at 8:30. Okay.

Oh, just before the tape gets turned off, we should have one last expression of appreciation to Dr. Favata, the only polymath on the Advisory Committee.

Thank you.

DR. FAVATA: Thank you all.

[Applause.]

MR. COENE: And for the record we have -- along with other problems with the Secretary's office    a certificate of appreciation. Unfortunately, it didn't get --

MS. JEWELL: It's coming, but it's because of the transition, and you don't expire until the 31st, so she's not the Secretary then, so --

MR. COENE: Mr. Thompson will sign it.

MS. JEWELL: You kind of got caught in the middle there, so it is coming.

DR. FAVATA: Not a problem.

MS. JEWELL: I will Fedex it.

DR. STOTO: It's your term that expires, not you.

(Laughter)

[Whereupon, at 4:01 p.m., the meeting concluded for the day, to resume at 8:30 a.m. the following day, January 24, 2001.]

- - -


1   ...   15   16   17   18   19   20   21   22   23

Похожие:

Ranch hand advisory committee iconAdvisory committee on immunization practices

Ranch hand advisory committee iconMedical Devices Advisory Committee

Ranch hand advisory committee iconVeterinary medicine advisory committee

Ranch hand advisory committee iconNational Vaccine Advisory Committee (nvac)

Ranch hand advisory committee iconExternal Advisory Committee on Cities and Communities

Ranch hand advisory committee iconWildlife Diversity Policy Advisory Committee

Ranch hand advisory committee iconPeer reviewed by the Arizona Department of Commerce Economic Research Advisory Committee

Ranch hand advisory committee iconFood and drug administration national institutes of health advisory Committee on: transmissible spongiform

Ranch hand advisory committee iconAdvisory Committee, Cuyahoga Valley School-to-Career Consortium, Broadview Heights, Ohio 1996-2002

Ranch hand advisory committee iconJane D. Siegel, md; Emily Rhinehart, rn mph cic; Marguerite Jackson, PhD; Linda Chiarello, rn ms; the Healthcare Infection Control Practices Advisory Committee

Разместите кнопку на своём сайте:
Библиотека


База данных защищена авторским правом ©lib.znate.ru 2014
обратиться к администрации
Библиотека
Главная страница