Ranch hand advisory committee




НазваниеRanch hand advisory committee
страница17/23
Дата17.10.2012
Размер1.21 Mb.
ТипДокументы
1   ...   13   14   15   16   17   18   19   20   ...   23

DR. FAVATA: Yes, it is.

DR. STOTO:    under general health. General health itself is one of the dependent variables there.

DR. MICHALEK: Wait a minute. Hostility. Hostility does not have to be included. That's the first question. Hostility is now going to be just taken if necessary from a scale on whatever that exam is.

DR. FAVATA: Right.

DR. HARRISON: Okay, that's taken care of.

DR. FAVATA: Done. We don't need to adjust it for anything.

DR. HARRISON: So then there's the perception of health. You're raising that as a separate issue now.

DR. FAVATA: Yes.

DR. HARRISON: Why do we need a perception of health?

DR. MICHALEK: Well, perception of health is an endpoint in the study.

DR. HARRISON: And maybe the best answer is that we've always had a perception of health, and it is not a big deal.

DR. STOTO: And we also talked about the possibility of incorporating a more general    something like the SF36, to get at self-assessed health.

DR. HARRISON: Right, but then we would be adding a new questionnaire at the last cycle.

DR. STOTO: Right. I'm not saying we should do it; I think we should look into that possibility.

DR. MICHALEK: What I hear you saying is that the use of a hostility as a covariate in the analysis of general health, perception of health, is not necessary. And I accept that, that's your opinion, and we'll just scratch it from the Statement of Work.

DR. HARRISON: Okay, so the only work that Paul is left with is the short term question of is there another   .

DR. GRUBBS: I have a question.

DR. HARRISON: The other reason this was being raised is because short-term memory is one of the things that's been found. And yet there's only way of testing it within this psychological section, and this Mike was saying "Well, if that's something that's been found, shouldn't there be"   ?

DR. CAMACHO: Other tests.

DR. HARRISON: Yes.

DR. GOUGH: If there is one.

The other thing, Paul, is this paragraph here, which just seems to me to be    I brought it up with Joel about, how do you think the mixture of their dependent or independent variables, their codependent variables and things; and Joel said this needs attention.

DR. CAMACHO: See where the problem is    you've got page 32?

DR. GOUGH: Well, it's probably on your 33.

Here it is. You've got it. It's back there.

Could you look at that and could you consult with your friends and find out if this is a reasonable listing and    it's such a conglomeration of stuff.

DR. CAMACHO: This one.

DR. GOUGH: Yes.

DR. CAMACHO: 36?

DR. MICHALEK: 36443.

DR. GOUGH: Maybe actually that's the way psychology is.

DR. MICHALEK: And actually 36442, excluding AIDS-positive patients from this analysis.

DR. HARRISON: Well, their perception of health is probably different.

DR. GOUGH: If you could shed some light on that.

DR. HARRISON: Yes.

DR. GRUBBS: Dr. Harrison, a question on personality type. It's used as a covariate in three chapters: General Health, Cardiovascular, Endocrine. I think the main driver of personality type was this covariate adjustment factor for cardiovascular.

We searched the covariate associations. Anything that's significant shows more abnormalities in type B participants than type A participants, throughout. Maybe it's a fault of how they're filling it out, or whatever.

Should personality type or some surrogate, whether it's hostility or impatience or whatever, should that even be used as a covariate for cardiovascular anymore?

DR. STOTO: It seems to me we have an opportunity to look at something new, which is the component of the SL- whatever it is.

DR. GRUBBS: Is hostility an appropriate covariate for cardiovascular?

DR. STOTO: I think it's a well-known    psychological factors like that are well-known to be associated with cardiovascular outcomes. Right?

DR. HARRISON: I guess the first point is that you can't use the old test. That's really not us telling you   

DR. GRUBBS: Yes, it's impractical.

DR. HARRISON:    you're telling us that guys are playing games with it now. So it's out. What you'd propose to do instead was to add another hostility assessment, but yet it turns out there's a hostility assessment within the exam that's being given. So you don't really need to add, if hostility is what you're going to use, there are    my understanding is that there weren't any other tests to separate into type As and type Bs, so now the only question that remains is, when you do the analysis, are you still going to use    are you going to use the hostility section from this as a covariate?

And what Mike is saying is that you haven't used it before, so actually you got a good chance to add something new to the study.

DR. GRUBBS: And it's an appropriate covariate.

DR. MICHALEK: It's appropriate for cardio. Is it appropriate for endocrine?

DR. GRUBBS: It was done on thyroid.

DR. MICHALEK: Your face says no.

DR. GRUBBS: No.

DR. MICHALEK: Should not be.

So drop it from endocrine, but kept it in cardio, and keep it in general health.

DR. GRUBBS: Good.

DR. HARRISON: Where are strokes? Cardiovascular?

DR. GRUBBS: Yes.

MR. WEIDMAN: I guess I'm a little puzzled over here on a number of things. One is that --

DR. HARRISON: After this, though, we need to try and stick with our --

MR. WEIDMAN: This has to do with, you already know that heart conditions are secondary to PTSD in chronic acute stress and are already rated as such by VA, and although it may or may not be strictly stress, it may be effects on the endocrine system that lead to that and the triggering of the chronic acute stresses of firing over and over again of the adrenal gland, which may be related back to the thyroid and the parathyroid gland.

How would you tease out all of that in the tests that you do for the covariables, I guess is my question.

DR. HARRISON: Well, there are some things I don't think you can tease out. I think that the psychological tests themselves are probably better at identifying men who are stressed than any of the hormonal levels that we're going to find. If that's the question that you're asking.

MR. WEIDMAN: I guess, I mean there are many reasons why people would be hostile, as an example. And hostility at this point, I mean some people are just hostile towards the Air Force, you know, who have served in --

DR. STOTO: Hostility is not the dependent variable here. Hostility is the adjustment variable. If it so happens that the controls are more hostile than the    if the Ranch Hands are more hostile than the controls, that may be responsible for a higher level of cardiovascular problems in the Ranch Hand group. And you want to make sure that if you find a bigger difference, it's not attributable to --

MR. WEIDMAN: Oh, I see. So it's just comparison between those two.

DR. STOTO: It's to help even out the groups.

DR. HARRISON: You know, the problem with this type of a study is that there is no hypothesis to start with. If you knew at the beginning that, if your hypothesis was that exposure to dioxin causes more myocardial infarctions, then you can focus on that and you can do whatever you want. But when the study began, you didn't know what you were going to find.

In fact, I spent the first few years on this committee being beaten to death over something called a checkmark pattern that essentially meant that if you analyze enough people    well, this is how I would interpret it    if you analyze enough people with enough tests, you wind up with differences. You know, not because there are differences but because things happen. That's one of the realities of life.

MR. WEIDMAN: Mr. Chairman, thank you. We'll submit written things, and I appreciate all of your efforts.

DR. HARRISON: It's been a pleasure having you. You're not going to stick around and see the last part of this sausage turned out?

MR. WEIDMAN: I'm going to go get ready for my meeting with Dr. Olden on Thursday and with Tony Principi in Thursday. So thank you for arming me further.

DR. HARRISON: If you can remember, please tell him hello for me.

MR. WEIDMAN: I will, indeed.

DR. CAMACHO: Did you say you informed somebody about this meeting and they were coming here tomorrow?

MR. WEIDMAN: I didn't inform her, but Mary Ellen McCarthy was planning to come here tomorrow. If you're not going to meet tomorrow, I'd urge you to call her.

DR. HARRISON: Well, --

MS. JEWELL: Bob, we have no choice. It's a Federal Register notice and our meeting goes through noon tomorrow, and we have to meet tomorrow.

DR. HARRISON: Okay. That's no big deal. I'll be here tomorrow.

DR. MICHALEK: I'll be here, too.

DR. HARRISON: Not a problem.

MR. WEIDMAN: Thank you.

DR. HARRISON: Take care.

MR. WEIDMAN: Is it possible, that in addition to those letters that have already been sent to Dr. Olden and others, and to Secretary Shalala that the prospective letter could be routed in our direction?

DR. HARRISON: That's what I was saying.

MR. WEIDMAN: We'd be grateful, sir. Thank you.

Take care now.

[Mr. Weidman leaves the meeting.]

DR. HARRISON: Okay, where are we here now?

DR. CAMACHO: Are you ready for Mike?

A couple of questions. I got this today, when I walked in the door. I had some pre-s, but then we had some CD trouble, when I e-mailed her.

The Psychological Battery is 3.2.2, and it goes on a bit about what they're going to do. And then it's way down in 364 you're asking me to look at. I just want to make sure I'm not off on a wild goose chase here. 3 2 2 is related to this 3 6 4; it's got to be. Right? I mean, you're not doing these things out of the blue.

DR. GRUBBS: 3 2 2 talks about the conduct of the exams; 3 6 4 is in the Statistical Analysis section.

DR. CAMACHO: So this is the plan to collect the data; this is the plan to analyze that data.

DR. MICHALEK: That's right.

DR. CAMACHO: Okay. And you're asking me to comment on this so I have to look at this as well.

DR. MICHALEK: Yes. Whatever you decide will influence what goes in this 3 2 2 paragraph.

DR. CAMACHO: Okay.

DR. MICHALEK: You might decide to administer another questionnaire, so we would say so here.

DR. CAMACHO: All right. Because, pardon me, but I just came on the committee and I got    the whole history of this is a non    I don't have any history on it.

DR. MICHALEK: Right.

[Discussion about hotels.]

DR. HARRISON: Paul do you get anything else?

DR. CAMACHO: I know. I'm going to look at 3 2 2 for the deaths, the methodology of collecting the data and comment on 3 6 4 4 about the analysis techniques of the data, and you're back to him within four to six weeks.

DR. HARRISON: Okay so we're not going to discuss that any more now.

So now we have a discussion    and we're not going to have the discussion of dioxin and debates, we're going to have a discussion of report writing?

DR. MICHALEK: Yes.

DR. HARRISON: What we have left is discussion of report writing, and agenda, and future meeting dates.

DR. STOTO: We didn't do section 3.7 yet.

LTC BURNHAM: 7, and addendum A. Those are two.

DR. HARRISON: Oh, I'm sorry. Well, wait a minute.

DR. STOTO: I have a very insightful comment on 3 7.

DR. HARRISON: You're right, I am sorry. Here we go.

Somebody start off.

DR. STOTO: On page 45 of the new version of this, section 3.7.3, they talk about the summary tables. And remember, at some previous meeting some of us had some good ideas about how they should be presented in clinically meaningful terms. Like saying, you know, that there's a so-and-so increase in hemoglobin a1C and so on, not just whether they're statistically significant. And it was decided at that time that it was too late to do that. Well, now it's time, I think, that we should work that into this paragraph here that says a summary table should be presented in terms of clinically meaningful effect sizes.

DR. MICHALEK: So the idea would be to annotate the findings that are biologically plausible, or clinically meaningful?

DR. STOTO: No. Whichever ones that would have made the summary table before in terms of statistical significance, that you report not only the p-value but also whatever --

DR. MICHALEK: Direction.

DR. STOTO: No.

DR. MINER: And clinical relevance.

DR. STOTO: You know, it depends on what variable it is. If it's a continuous variable, you report the difference in the mean of the outcome.

DR. HARRISON: Mike, you're talking about 7.3, right?

DR. STOTO: Yes.

DR. HARRISON: Can you read it the way you think it should be written?

DR. MICHALEK: It says right there, 3 7 3, p-values together with corresponding group means and percent of normal.

DR. GRUBBS: Dr. Stoto, I think possibly your sense said that the relative risk was 1.5, and it was significant. You'd want to see the 1.5 rather than p=0.001.

DR. STOTO: That's right, that's part of it. Or if it's a continuous variable --

DR. GRUBBS: You'd like to see the effect, the difference--

DR. STOTO: Right. You know, these guys differ   

DR. GRUBBS: Increase of 3 grams.

DR. STOTO: 10 millimeters of mercury, or something like that.

These are values that are given in    whatever values are given earlier in analysis, they just need to be carried over to the summary table.

DR. HARRISON: The reason I'm asking because I -- this is the kind of stuff that I get very confused about. And you know exactly what it is that you would like, what you're trying to say, and it helps me to have you presented in complete sentences, you know.

DR. STOTO: I would say that the second sentence there --

DR. HARRISON: Main effect p-values. Let me take that sentence, let me give you a replacement for that sentence.

DR. HARRISON: All right.

- DR. STOTO: I would say the results should be presented in terms of clinically-meaningful effect sizes, and p-values.

DR. GOUGH: What does "clinically-meaning"   ?

DR. STOTO: Well, I was going to say, you could add "such as relative risks or difference in group means or regression slopes or something like that.

DR. HARRISON: So what you're saying is, you not only want them to say that the difference between Ranch Hand and comparison was significant, p=.005, but you want to say relative risk for this particular finding was 1.5.

DR. STOTO: If it's a dichotomous value, the clinical and meaningful thing was a relative risk. If it's a continuous variable, then it would be the group means. If it's a kind of regression analysis, it would be a slope; it would say "How much more does this go up when you increase the X by so-and-so?"

But whatever you used and was reported in the earlier tables --

DR. HARRISON: You want it in the summary table.
1   ...   13   14   15   16   17   18   19   20   ...   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