U.S. ENVIRONMENTAL PROTECTION AGENCY PRIVATE  

+ + + + +

HUMAN STUDIES REVIEW BOARD (HSRB)

+ + + + +

PUBLIC MEETING

+ + + + +

FRIDAY,

OCTOBER 26, 2007

+ + + + +

		The meeting convened at 8:30 a.m. in the Conference Center at One
Potomac Yard, 2777 South Crystal Drive, Arlington, Virginia, Celia B.
Fisher, PhD, Chair, presiding.

HSRB MEMBERS PRESENT:

CELIA B. FISHER, PhD, Chair

STEPHEN BRIMIJOIN, PhD, Vice Chair

GARY L. CHADWICK, PharmD, MPH, CIP, Member

JANICE CHAMBERS, PhD, DABT, Member

SUSAN S. FISH, PharmD, MPH, Member

SUZANNE C. FITZPATRICK, PhD, DABT, Member

DALLAS E. JOHNSON, PhD, Member

KYUNGMANN KIM, PhD, CCRP, Member

KANNAN KRISHNAN, PhD, Member

MICHAEL D. LEBOWITZ, PhD, Member

LOIS D. LEHMAN-MCKEEMAN, PhD, Member

REBECCA PARKIN, PhD, MPH, Member

SEAN M. PHILPOTT, PhD, Member

RICHARD B. SHARP, PhD, Member

HSRB STAFF:

PAUL I. LEWIS, PhD, Designated Federal

	Officer

EPA STAFF:

JOHN CARLEY, Office of Pesticide Programs

WILLIAM JORDAN, Office of Pesticide Programs

KEVIN SWEENEY, Office of Pesticide Programs

PUBLIC COMMENT:

NIKETAS SPERO, PhD, Principal Investigator,

	ICR, Inc.

ROBIN TODD, PhD, Director, ICR, Inc.

	TABLE OF CONTENTS

AGENDA ITEM	PAGE

Convene Meeting	4

ICR Repellency Efficacy Protocol A117	8

Public Comments	59

Board Discussion	82

EPA Update of AEATF and AHETF

	Research Programs

EPA Presentation	124

William Jordan (OPP, EPA)

Public Comments	n/a

Board Discussion	139

Adjourn

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

8:36 a.m.

		DR. FISHER:  Okay, so John or whoever is -- did you want to follow up
on anything, Bill?  Okay, so John, we're on ICR.

		MR. CARLEY:  We are on ICR protocol which we know familiarly and for
short as A117 because the full designation is about a block long.  

		Good morning, everybody.  Welcome back.  There it is.  So we're going
to review Protocol G0590607001 A117.  Next one, please.  This protocol
was submitted in early August on behalf of Avon Products by their agent
toXcel, so there are three entities involved here.  ICR is the
laboratory, Avon is the sponsor and toXcel is Avon's agent, the
regulatory agent, and it was through them that it was submitted to us.

		It proposes a laboratory study by Insect Control and Research,
Incorporated, which we'll call ICR from -- in the rest of this, and
they're going to check the repellant efficacy in the laboratory against
Culex genus mosquitos  of two registered repellant products that contain
10 percent picaridin.  Our science and ethics review of September 24th
was based on reviews of this August 8th submission and looking back at
the labels that we had previously accepted for these products.

		The August 8th submission fell short of the standard of completeness
defined in 40 CFR 26.1125 in several ways that I'll talk about
specifically in my ethics review, but these didn't compromise our review
of the protocol.  Some refinement is still needed to address the
deficiencies that we noted in both science and ethics, but we think that
this protocol is right for and will benefit from review by this board.

		In a submission on October 17th, responding to our review, the
submission came from toXcel, the investigators through toXcel promised
to revise protocol and consent forms to address most of the comments
that we had made in our review.  It's noteworthy in this that ICR and
Avon, the sponsor, want to use the same one gram per 600 square
centimeters standard dose and the same time to first controlled bite
rather than landing as the end point for consistency of both dosing and
end points in this study with -- and the earlier field studies of these
products. 

		This proposal requires HSRB review.  In terms of the standard of
completeness, there are some important differences in our view and that
of ICR about how to apply that standard of completeness.  They thought
they'd covered everything.  We didn't and I just wanted to make these
general observations about the basis for the difference.  

		We don't consider all of those items listed, protocol drafts, consent
forms, application forms, MSDSs and so forth, we don't consider that to
be IRB correspondence.  We think that that's a much smaller and more
focused category and it makes it a lot easier to find if you use the
narrower definition but we also feel that consent forms, by their very
nature as documents intended to be -- to communicate with subjects,
don't address the requirements of 1125(a)(1) through (5) for a
discussion of risk and risk minimization and benefits and alternatives
and risk benefit balance.  We're expecting separate discussions in the
protocol or in a covering document or something rather than just saying,
"Oh, here are the page references for the consent document".  

		And that's why we -- that's where we saw the gaps in completeness. 
The missing elements were acceptable discussions of the nature and
magnitude of all expected benefits and to whom they would accrue, that's
1125(a)(3).  The balance of risks and benefits, 1125(a)(5) and an
acceptable description of the circumstances and methods for presenting
information and seeking a consent, 1125(b).  

		As a passing note, the submission said that the IRB procedures were
available for inspection only at the EIRB site in New Jersey.   EIRB has
sent them to us directly.  As in the case that we were looking at
before, they're covered by a claim of confidentiality but we have them
and we have reviewed them so, that statement was not correct and
shouldn't be a cause of concern.  And now Kevin is going to talk about
the science assessment of this study.

		MR. SWEENEY:  Good morning.  The purpose of this study is of course,
to determine the efficacy of the subject repellant against Culex
quinquefasciatus mosquitos which are West Nile Virus factors and this is
a laboratory study.  And we're going to determine the mean protection
time.  In this case the end point is going to be bites and not landings
as proposed provided by the test articles under laboratory conditions to
confirm the hypothesis and essentially these data are being submitted to
amend the label to support the claim of mosquitos that may or which can
vector West Nile Virus.   Next slide.

		Both the products being used here are 10 percent picaridin and they're
both registered and they're also both owned by Avon and they're both
sprays.  Next.  Test conditions, laboratory maintained at 80 degrees
Fahrenheit, plus or minus 15 degrees with a 70 percent relative
humidity.  There will be six cages in this test, each measuring two by
two by two feet.  They're screened on all sides except the bottom. 
There's a mirror on the bottom so you can watch the underside of the
arms for landing or biting mosquitos and as I'll show you in a picture
in a minute, I think we talked about this in the second HSRB meeting,
there's two sleeve ports on each opposite sides of the test cage.

		There will be two subjects per cage and four arms per cage.  And this
will be a one-day test that will last up to nine hours.  Okay, the cage
itself is shown here.  This is two foot by two foot by two foot.  We
showed a similar cage in a, I think it was June 2006 HSRB meeting and in
a brief demo we gave on the lab test.  Sleeve ports on either side as
you can see, they're muslin cloth.  There's a wood bar in the middle
that the handle rests on.  I'll show you a picture of that in a minute,
and you see the mirror image.  So anyway, we've got the mirror on the
bottom, the rest of the screen -- cage is screened and the subjects will
sit on chairs on either side of this cage.  Next slide.

		The test mosquitos are female only, of course, quinquefasciatus
mosquitos.  They're three to eight days old.  As I was reminded
yesterday by Dr. Gupta about the age of the species.  These are fairly
newly emerged mosquitos that have not received prior blood meal.  In
fact, they're going to be fasted for 24 hours before the testing begins
so they'll be plenty hungry when they're actually exposed to the arms.  

		There's 100 mosquitos per cage as proposed, 100 more per cage, fewer
than five lands per five minutes, in untreated control.  This colony has
been lab colonized for many years and is not ever exposed to the
outside.  You know, it's in a chamber inside of a building.  It's been
there for a number of years, and all mosquitos are destroyed after a
single test and as a result we're not going to have any worry about
transmission of blood borne diseases.

		Okay, this is another picture of the test.  Okay, the mosquitos are
released into the case by a technician with a gloved hand through the
sleeve and they're precounted into the container, released into the
cage.  There will be 100 in there.  As you can see, they already start
resting immediately when they come into that cage.

		In terms of the exposure, next slide, they must expose their untreated
arms to the caged mosquitos just to establish attractedness and
generally I can tell you from personal experience, in a cage of 100
mosquitos, there's usually no issue with getting personal attractedness,
but some people, there could be a rare person who's not, you know, found
to be very  attracted to these mosquitos.  There will be 10 treated
subjects plus two treated alternates to insure a sample size of at least
10.  The treated subjects are treated with one formulation each arm,
expose both arms to the caged mosquitos for five minutes at 30 minute
intervals.  So essentially, you've got two replicates in each cage, a
total of four arms.  And there will be one untreated subject selected by
lot that monitors the aggressiveness of the caged mosquitos, for one
minute at 30-minute intervals.  This is just to make sure that they are
still actively seeking a blood meal.

		Okay, in terms of the dosing regime, it's a little bit different from
what we've seen in the other studies that we talked about with Dr.
Carol.  The proposal is to use a standardized rate that's generally used
in a lot of these tests of one gram of product per 600 square
centimeters of skin surface, which is equivalent to 1.67 milligrams of
product per centimeter square.  And the treatment is applied to a 250
square centimeter area and I'll show you that in a second, on each
forearm of each treated subject.  And as a result the total dose on both
arms is 835 milligrams of product, which is equivalent of 83 milligrams
picaridin.  Based on the WHO end point for dermal NOAEL of 5,000
milligrams per kilogram, the MOE is going to be greater than 1800, so
there's really no issue at all with exposure to the chemical.  At this
point, it's pretty minimal and only the subjects will be blind in this
study.  Next slide.

		Okay, this is just a picture of the actual test as its conducted. 
You've got gloved hands on the subjects.  They're taped.  Again, you've
got more of these Ace-type bandages here, taped on and they're layered
so the mosquitos cannot bite through them, and as a result, you have
this much skin surface applied and you have to look underneath here to
see if mosquitos landed on the bottom to bite and for the comfort of the
subjects just rest their arms on the cage.  And you'll have one from
this side as well. 

		And one thing I just wanted to point out there, too, is that the
mosquitos do land right up to the edge of this, okay, as you can see
here.  And we talked about yesterday about whether or not there was
enough vapor pressure or an area wide effect from repellants. 
Generally, it's not generally the case in these cage tests that you see
that type of effect.  The area that's covered, the mosquitos land on the
area that's not covered.  They will only stay off it as long as they're
repelled.  So you don't have any kind of area effect.  They'll come
right up to the edge.  So I just wanted to mention that.  And it's not
as if there's an aura round this arm, the mosquitos will stay away from
the subjects.  They're going to come right up to the edge.

		Okay, in terms of the end points, again, they're similar to what we
talked about before.  The subject limb dimensions, in this case, of
course, we're only using arms.  And the time post-treatment of all bites
on treated subjects and the landing rate on untreated control arm. 
There was no form proposed for recording results of the attractancy
tests to establish whether or not the subjects would qualify.  The
duration of efficacy measured as time from treated to first confirmed
bite or eight hours, whichever occurs first.  And as a quote from the
study itself states, "ICR prefers to evaluate repellency based on
protection from bites rather than landings while conducting laboratory
studies".  Okay, and this is their proposal.  

		A proposed analysis plan, and I'll just mention something about the
statistics briefly, is they're non-parametric statistics Cochran's Q
Test with the cross tabs and Fischer's Exact Test.  And I just want to
mention that I'm not overly familiar with this test myself and I did ask
ICR to provide more of an explanation or more detail about the testing
and some of the assumptions related to it.  I think it relates to
proportions of subjects that have been bitten.  

		They'll report the mean time to first confirmed bite with the standard
deviation 95 percent confidence interval.  And then the median time to
first confirmed bite and confidence interval by percentile if
significantly skewed distribution.  Again, I also asked for a little
more explanation about that.  

		Untreated control will not be used for comparison, only to establish
whether the mosquitos are biting or not.  And there will not really be a
comparison between the two formulations.  Rather they're going look at
the ability of the formulations to repel Culex mosquitos.  The method of
handling the data from subjects who withdraw early is also not specified
in this protocol.  

		The rationale for the sample size, and this was also mentioned -- at
least this publication was referred to yesterday by Dr. Gupta and this
is the 1999 publication he published with Louis Rutledge about the
Tascosa sample sizes, the confidence interval, standard deviations, et
cetera and I think we also mentioned that we would provide this to the
Board, so we will get a copy of that.  But based on that study and some
of the assumptions made in that study, the study therefore, will use 10
treated test subjects.

		And of course, there's also two alternates that will be treated and
will help insure a minimum of ten and will aid in protecting the privacy
of any dropouts.  

		Now, some of the deficiencies we noted, I think in my mind, one of the
bigger ones was that the EPA guideline called for a density of 200
mosquitos per cage and so the proposal to use 100 per cage is not really
explained or justified.  It may be a preference of the sponsor.  And I
think, I just want to comment that in my opinion and also from personal
experience, the density of mosquitos in a cage does have some effect on
their behavior.  I mean, they tend to be more voracious, it just seems. 
I don't know if they're competing with one another but, I mean, the more
mosquitos in a cage, the more aggressive they tend to be. 

		In terms of the explanation of diagnostic testing for the normality of
distributions or the analysis of non-normal data, we don't really
consider that complete as I said already, so we'd like some more
explanation about that.  We generally don't see this test being used too
much in these types of repellency studies.  And of course, the products
are EPA approved and any label should be appended to the protocol.  

		And as far as compliance with scientific standards, this test is
fairly standardized frankly.  I think we can resolve some of the
deficiencies related to the statistical analysis and related
discussions.  I think it will be fine.  Of course, then there's also the
discussion of whether we should be using landings versus bites as an end
point and I guess we'll have the discussion here today and I think John
may  make mention of that as well, which is significantly different than
what we talked about before.  So other than that, that's pretty much it.
 I'll hand it over to John.

		DR. FISHER:  Thanks, Kevin, that was really helpful.

		DR. CLARKE:  Any clarifying questions for Kevin about that?

		DR. FITZPATRICK:  Kevin, I think you've pretty much answered this
question but one thing that was in my mind was whether or not two arms
in the cage would be close enough to compound the data on one another or
the other person in the cage and you're suggesting, I think a few
minutes ago that there isn't enough vaporization of this product to
carry over to the next arm or the next person; is that correct?

		MR. SWEENEY:  Yes, I think that's correct and especially if the
density were increased to 200, I don't think it's going to be a problem
at all for sure.  I just don't -- it's just not that aromatic and their
vaporization is not that great.

		DR. FITZPATRICK:  The arm question was mine at first but my other
question is, if all the subjects make it to eight hours do you do any
kind of statistics at all or do you just put eight hours on the label?

		MR. SWEENEY:  In other words, if there's no failures at all?  

		DR. FITZPATRICK:  Right.

		MR. SWEENEY:  Then, yes, I think we're going to assume that the --
that it lasted for eight hours.

		DR. FITZPATRICK:  Okay.

		MR. SWEENEY:  Okay, and these are already registered products that had
field tests done on them already that supported an eight-hour claim but
there weren't -- they weren't a West Nile Virus claim now and you have
to really test QX which is the major vector to get that claim.  And we
didn't want to do it in the field for obvious reasons, so we asked them
to do a laboratory study.  

		DR. FITZPATRICK:  Well, another question from me then is, did they do
a comparable laboratory study just on other species before or did they
just go directly into the field studies?

		MR. SWEENEY:  In this case, they did just a field study on these two
products.

		DR. FISHER:  Gary?

		DR. CHADWICK:  Two things that you sort of touched on but just for
clarification.  The mosquito age, is there a difference in the different
species as far as biting pressure?  We heard yesterday that five to 15
days was the optimal age range and this group is using three to eight. 
Is that species dependent or -- I mean, is it -- 

		MR. SWEENEY:  No, I don't think it's species dependent.  I mean, I
think that at least in my experience and I've reared a lot of mosquitos
as well and in lab studies, I mean, for a lot of species, you can use
three days.  I mean, once they emerge, in a day or two they're generally
ready to start blood feeding.  So, I mean, five might be a little bit
better, I mean, between three and five, I don't personally think there's
a big difference.  I mean, Dr. Gupta or Dan Strickman, I can't remember
who mentioned that but in my opinion, 15 days to me seems a little bit
long.  Within the first 10 days I mean, they tend to be pretty -- I
mean, not only pretty voracious but pretty healthy, frankly, when you're
rearing them in the lab, and if you go without feeding them blood for
over a week, I just don't think that's a great idea.  I've just given
them sugar water alone.  They need some of that.  The females need blood
for nutrition as well, so I mean, they want to blood feed by the end of
-- you know, by the end of a week.

		DR. CHADWICK:  Yes, okay, the other question I had was, and you
touched upon it, the mention about bites versus landings and ICR saying
they prefer to do this.  On repellents, is there a scientific -- I mean,
is it better to do bites.  Do you approve -- 

		MR. SWEENEY:  Well, there's opinions on both sides of that, okay, as
to whether or not you should use landings or bites based on science
alone.  I mean, a lot of people say a bite is a bite.  I mean, when
they're biting, they're biting.  There's no discussion about whether
they it or not.  And then of course, versus a landing  being a landing. 
Now, again, a landing is a more conservative measure.  Once they land on
that arm that has a repellent, the assumption could be made that they're
going to try to bite or at least they're seeking to bite.  Somehow
they've managed to attack the arm, okay, even though the repellent is on
it.  So it's a more conservative measure to use landings versus bites.  

		I think when you're in the lab studies when you've got hundreds of
mosquitos in the cage, though, you know, you've got to be watching that
mirror the whole time and if one just jumps off, you know, will you miss
it or will you not?  Generally not, if you've got two people or one or
two people watching the cage.  So you know, I don't have a personal
preference either way.  I mean, I think landings are more conservative
than bites. I mean, I'll just say that.

		DR. LEHMAN-MCKEENMAN:  Can you give me your perspective on the utility
of the untreated control in this study?  The data aren't going to be
compared to that subject.  If I understand correctly, that subject gets
to put his or her arms into each of the six test cages and yet, each of
the subjects will also put an untreated arm into a cage.  So I'm not --
am I confused?

		MR. SWEENEY:  Well, I think they initially put their untreated arm in
the cage just to establish they're attracted -- the subjects this is,
but once they've done that, then both arms are going to be treated.  So
the subjects are not going to be exposed untreated again.  Generally, I
think this is just my opinion, is that this may have actually come over
time from some of the sponsors frankly, wanting to make sure that
whatever mosquito colonies were being used in these tests were going to
be continually interested in blood feeding over time and therefore,
using one person's arm to establish their attractiveness is where that's
come from.  Now is it necessary in every case?  I mean, frankly, if the
mosquitos have been established -- you know, if they're landing on the
subjects, I don't know that it's really necessary to do that, in my
personal opinion.  Okay, but I mean, that's what the lab has proposed to
do.  That's part of their general protocol.

		MR. CARLEY:  A small clarification that may help here.  The subjects
who are going to be treated use an untreated arm to establish
attractancy.  But the subjects who are not treated use an untreated arm
to establish continuing mosquito pressure.  It's the same as the role of
the untreated controls in the field studies we've looked at.  They're
making sure that there continues to be pressure, landing pressure,
biting pressure, that meets a predefined threshold.  And if it doesn't,
unlike a field study, they have the option of releasing more mosquitos
into the cage, so that they can get a fresh batch and get the avidity
back up to a level that provides a fair test of repellency.

		MR. SWEENEY:  And just one other comment, too, I think when you have a
business where you're testing a lot of repellents for a living,
sometimes when you test certain compounds, going back to the area-wide
effect we were talking about before which we don't really see a lot of
but it could happen, or there could be some attributes to a repellent
that's being tested that could cause mosquitos maybe not to behave
properly.  It could be a behavioral, you know, effect and this kind of
thing might detect that.  

		In this case, though, with the caradom, we generally don't see that. 

		DR. FISHER:  KyungMann and then Suzanne.

		DR. KIM:  One quick question.  In one place you mentioned one
nine-hour day of testing and in other place I saw that the subject will
be exposed for up to eight hours.  What is the -- 

		MR. SWEENEY:  I think -- well, I think the exposure period will be
eight hours and I'm not -- I think that maybe, and I don't know if I
think Dr. Todd or they may make a public comment, but I think that maybe
they're just talking about the subjects being there for a nine-hour day,
for eight hours, right.

		DR. FITZPATRICK:  My question was just will you -- will EPA accept a
lower number of mosquitos in the cage?

		MR. CARLEY:  It's an irrelevancy.  If you look at the toXcel
supplemental submission, they said they'll go to 200.  

		DR. FISHER:  Richard and then Jan.

		DR. SHARP:  Just a quick follow-up on Dr. Kim's comments, so are
breaks structured into this and for the subjects and so forth?  How are
they arranged for, you know, necessary needs and so forth?

		MR. CARLEY:  Exposure is intermittent.  It's five minutes of 30
protocol so there's an opportunity to take care of personal needs and
get something to eat at lunchtime and so on.

		DR. SHARP:  So the eight hours is eight total hours, not eight hours
of actual exposure in the cage.

		MR. CARLEY:  Yes, there's not a one-hour gap in the middle.  The
cyclic intermittent exposures would continue according to the schedule
throughout the day up to eight hours.

		DR. FISHER:  Jan?

		DR. CHAMBERS:  There are two alternates in case somebody drops out. 
Are the data from -- if everybody participates and nobody drops out,
will the data from those two alternates also be used?

		MR. CARLEY:  That isn't clear.

		MR. SWEENEY:  It's not clear.  It would be nice, but it's not clear.

		DR. FISHER:  Sean?

		DR. PHILPOTT:  Just a point of clarification because after reading so
many mosquito protocols, I can't keep them straight any more.  Yes, I've
gone buggy, thank you.  I'm  hearing from a comment that Lois made and
sort of that seemed to be confirmed by you that the -- even the treated
subjects will put an untreated arm into the cage to determine
attractiveness prior to -- okay, in your -- well, and maybe this is a
question then, that can wait for the ethics review, but I don't see that
in the consent documents, that description to the participants, so maybe
you can touch upon that in your review, please.

		DR. FISHER:  I just had a few questions of clarification, based on our
talks yesterday which were so informative.  So from what I understand
from yesterday, continuous is better than intermittent from a scientific
perspective and bite versus landing, at least from Dr. Gupta's
perspective, is more informative, although you gave an opposite
explanation in the sense it's conservative.  And they've already done
some kind of field test.  So I'm just trying to figure out what the --
how the results are used here in terms of how they inform field testing
versus field testing in forming this.  How is this going to be used when
the methodologies are so different and so -- what does this add?

		MR. SWEENEY:  This really just adds whether or not there's a
repellency against Culex mosquitos that can repel West Nile Virus,
that's it, although it will be interesting to see how the results
compare, frankly.  

		MR. CARLEY:  Also bear in mind that the pattern of intermittent
exposure, the pattern of 250-square centimeter area of treatment, the
use of bites as the end point and the first confirmed bite as the
calculated end point, all of those are parallel to the previously
conducted field tests.  So this is not precursor to field testing.  This
is icing on the earlier field testing cake.  This is just adding a
little bit because there were not Culex mosquitos present during those
earlier field tests.  

		DR. FISHER:  And I'm just wondering, given this is for pathogen
carrying -- I mean, the purpose is for pathogen carrying mosquitos. 
Obviously, they're not pathogen carrying in the lab, but given that it
sounds like the Army which is interested in pathogen carrying mosquitos,
appears to be using continuous and all of our speakers were talking
about the advantage of continuous.  I'm just trying to figure out, I
understand -- and it was very informative when John was speaking about
yesterday how in some sense moving toward the intermittent was based on
the human subjects regulation than an increased sensitivity to human
subjects issues although those issues are not the same in the lab.  

		So I'm just wondering as we're moving forward, and not as a critique
of this study but just as a way of looking at laboratory studies, is it
more important to have a laboratory study identical in procedure,
intermittent versus continuous because we want to protect people in the
field or is it more informative to use these other types of methods, it
seems as if our consultants, at least said were better in the lab when
it's safe?

		MR. CARLEY:  I think that's an open question, which is the better way
to go?  The argument that's being made by the proponents of this study
is that parallelism to the earlier study is appropriate when what you're
trying to do is extend the range of inferences, if you will, regulatory
inferences if you prefer, that you can draw from this body of work.  It
makes sense to me intuitively.  Bear in mind also that the purpose -- as
Dr. Schofield pointed out more than once yesterday, the purpose of the
research that's done by American and Canadian militaries is a public
health purpose.  It's not a regulatory purpose and there are distinct
differences in what it makes sense to do in those two arenas. 

		DR. FISHER:  Thank you.  That's helpful.  Any other questions on the
science?  John?  Oh, yes, John.

		MR. CARLEY:  Okay, all right, let's do the ethics assessment.  This
study would test -- this is a bit of a recap but I think it's worth
focusing on just for clarity.  This tests the efficacy of two already
registered products against mosquitos of the genus Culex in the
laboratory.  There were previous field tests that showed efficacy
against other genera of mosquitos but there weren't any Culex mosquitos
present and EPA requires testing that shows repellency against Culex
mosquitos to support the proposed label claims of efficacy against
mosquitos which  can vector West Nile Virus.  So that's the background,
that's why the study is being conducted.  

		And there is some potential societal benefit in identifying repellents
that are effective against potential vectors of West Nile Virus without
exposing subjects to wild mosquitos.  Even if the odds of infection are
very low, no exposure at all is lower still.

		The subject selection here is analogous to what we saw before from the
Carroll-Loye protocols but it's a little bit different.  What's similar
is that primary recruiting is done among a group of previous subjects
and people that they've referred.  The size of the pool from which ICR
draws is comparable.  In scale, it's a few score people.  I don't
remember it precisely.  They assert that this pool is as representative
of the potential repellent users as we are able to make it.  There are
some clear exclusions.  Anyone under 18, in this case they propose to
cap it at 55, the same point that Dr. Fisher raised yesterday about
there not being a clear reason to cap it at 55, in a laboratory setting
where the live pathogen isn't an issue applies here and we commented on
that. 

		Consistent with our regulations, they exclude pregnant and nursing
women.  They exclude people who have a known sensitivity to repellents
or to mosquito bites, those in poor health or physical condition, those
unable to speak and understand English and any employees or relatives of
employees of ICR, of the sponsor or of any other interested party.  

		If you remember, there was an issue about this exclusion in the ICR
protocol that we looked at, at the last meeting.  They've changed it. 
It is much more broadly exclusive of people who might potentially have
an interest or a vulnerability from the pool of subjects.

		So with that change, my conclusion is that no subjects would be drawn
from populations vulnerable to coercion or undue influence.  The
eligibility factors are inappropriately defined in my judgment.  They
say to be eligible to participate you must follow the requirements of
the study.  You must abstain from tobacco.  You must abstain from
scented products before and during the study.  You must wear specified
clothing.  Well, those aren't eligibility criteria that can be applied
before the fact.  Those are behavioral rules that apply during the
testing and are potentially grounds for removing a subject but you can't
apply them as eligibility factors.  They make perfectly good sense. 
They're just sort of placed under the wrong heading and I think clarity
would be improved if they were viewed as behavioral standards for
participation.  And the solution to that problem is to say subjects must
agree to follow the directions, they must agree to abstain from tobacco
and so forth.  And that can then be applied as an eligibility factor.

		Protective clothing, ICR will protect the untreated parts of the
forearm with bandages and hands with gloves as you saw in the picture
earlier, and I should add our thanks to ICR for providing those very
good pictures that showed what was going on there.  		The protocol calls
for them to wear their own blue jeans, heavy socks and long sleeved
shirts.  The consent form promises but doesn't provide an explanation of
proper protective clothing and my view before I saw their response was
that blue jeans and heavy socks didn't seem needed for a cage test in
the laboratory.  I thought it was a carry-over from a field test
protocol.  Their response was that because mosquito occasionally slip
through the sleeve, this protected them from miscellaneous stray bites. 
		I went to a demonstration at the ICR lab a few weeks ago and one of
their staff was patrolling the room with a small net.  He was the one
who had bred all of these mosquitos and he called them his babies.  And
he was sad as he caught his babies and killed them, but he told me he'd
raised millions and millions and millions of mosquitos.  Next one,
please.

		Let's talk about the risks from the repellents.  There are some
discrepancies in the way this is characterized.  One of the products is
registered with a Tox Category 3 caution label and the other with a Tox
Category 2 warning label based on eye irritation.  A reminder, the lower
the number of the Tox Category, the higher the hazard associated with
the product.  But the protocol misleadingly refers to the Tox Category 4
rating for inhalation and skin irritation.  It's misleading because the
way the Tox Categories work is that the highest hazard rating dominates.
 It doesn't matter if it's Tox Category 4, four inhalation, if it's Tox
Category 2 for eye irritation, that's what we treat as -- it's sort of
like the point of departure.  It's the end of the distribution of
possible values.  That's what we want.  So saying that it's Tox Category
4 misleadingly suggests that it's not as hazardous as it really is. 

		The consent form talks only about the hazard of picaridin, the active
ingredient, not about the test products which include other ingredients
that make it more irritating to the eye specifically.  So that needs to
be reworked so that the characterization of the risk is specific to the
materials that the subjects are actually going to be exposed to.

		The risks of allergic or irritation response from the test materials
are minimized by excluding known -- the candidates with known
sensitivity, limiting the area treated and monitoring the subjects
closely.  The risks of mosquito bites, the risk here is not of disease,
but of a reaction to a probe or bite and remember the point that was
made by the consultants yesterday, that a probe is -- can be just as
irritating, have the same kinds of effects as a bite.  

		It's not the loss of blood that causes the problem but it's the
anticoagulants and the other materials that are injected in the course
of a probe.  The risk is correctly described as having a broad range
from trivially minor effects to very serious effects.  The risk is
reduced in the appropriate ways by excluding sensitive candidates. 
Encouraging the untreated control subjects to shake off landing
mosquitos before they can bite.  

		If you've got 200 hungry mosquitos in this small space and you stick
your bare arm in there, you're going to be doing a whole lot of shaking
and the more you shake, the harder it is to count what's going on.  So
I'm a little uncomfortable about that.  

		They also expose the untreated control only long enough to confirm
continued pressure.  The rate is five per minute, I think, but if you
get five in the first 15 seconds, you're done.   Your job is done.  

		They also expose only a small area of treated skin intermittently and
they pair subjects.  When I saw this demonstration, I appreciated that
when you've got subjects sitting on opposite sides of that cage, the
easiest way to see the under-side of an arm is to look at the under-side
of our partner's arms rather than your own.  It's just the angle, the
reflection off that bottom.  So it's helpful to have the pairs working
and keeping track.

		The risk could be further reduced, this is directional.  Nobody can do
a very a good job quantitatively here but risk could be further reduced
by using landings as the end point for efficacy breakdown.  That's a
point that needs to be discussed with ICR.  I just did an absent-minded
trick of turning off my mike instead of changing the slide.  

		The mosquitos are pathogen free as Kevin pointed out, so the risk of
disease as characterized in the protocol in various ways, sometimes
zero, sometimes minimal, it is in any event, very, very low.  The
subject age, as I mentioned a moment ago, is unnecessarily limited  to
under 55 or not greater -- I've forgotten, not greater than 55 but that
doesn't make very good sense in the context of the lab study.  There are
no direct benefits to the subjects.  The primary beneficiary is the
sponsor.  

		The discussion in the protocol of societal benefit is about the
benefit of bringing new repellents to market.  That's irrelevant to this
case.  The products are already in the market and what we're talking
about is broadening the range of label claims.  So that is the means by
which a societal benefit might be realized.  There is some potential
societal benefit in identifying repellents effective against vector
species without exposing the subjects to potential disease but it's
important to recognize that the potential societal benefit here is
different from whatever would be associated with bringing new products
to market.

		There are some opportunities to further reduce risk, particularly by
considering landings as the end point.  If minimized, the residual risks
to subjects would be very low and I guess I should add if they are not
minimized, they're pretty darn low.  I mean, if it isn't changed to
landings, it's still pretty darn low. 

		DR. FISHER:  John, could you give us an estimate on the control arm or
the -- how many mosquito bites would they get?

		MR. CARLEY:  The control that is used to establish continued pest
pressure uses landings, not bites.  Now, the control puts one arm into
each cage at the beginning of each exposure cycle to see how many
mosquitos land in a very short period of time.  Now, it's possible that
some of those mosquitos may bite during that period of time.  This is
where that shaking them  off thing -- I mean, the alternative is --

		DR. FISHER:  The picture just looked like two arms, so I was confused.

		MR. CARLEY:  Well, the picture was of a test in progress not of a
control subject testing the waters.  What was I going to say?  Yes, they
could get bitten.  The alternative to the shaking thing that might be
worth considering would be remember the approach that's used in the
Carroll-Loye Laboratory where they are teaching people to aspirate
mosquitos.  You put two arms into the cage, one of which is gloved and
sleeved and fully protected and holds an electric aspirator.  

		You put the other bare arm in there and when a mosquito lands, you
suck that little sucker right up.  And it may not be possible to
aspirate enough mosquitos fast enough under these conditions and so you
know, the practicality of relying on that as a protection against bites
is unclear.  It's probably not going to be less effective than shaking
which is the proposed alterative.  It could be done in conjunction with
shaking, although shaking vigorously while trying to get the end of the
aspirator right on that little bug, I don't know.  

		DR. FISHER:  So this has never been done before that we could estimate
how many would bite?

		MR. CARLEY:  The design is that none would bite, but I suggest you ask
that question of the investigators when they make their comments.

		DR. FISHER:  Thank you.

		MR. CARLEY:  Where are we?  Yes.   Next page, please.  The Essex
Institutional Review Board Incorporated of Lebanon, New Jersey reviewed
this package.  They conditionally approved the protocol and consent
forms calling for some minor revisions on July 30th.  A week later they
approved Amendments 1 through 8 and revised consent forms.  The
amendments that they called for were entirely editorial.  The IRB is
independent of the sponsors and investigators.  It's registered with
OHRP.

		Dr. Chadwick clarified for me something about its status with PHRP. 
They were accredited by PHRP before it went defunct.  They were among
the first accredited by PHRP and because the accreditation expires on
the anniversary of the initial event, they were among the first to fall
off the role.  I hadn't appreciated that so the comments in my memo that
I couldn't find them mentioned the PHRP website, it's still true,
they're not mentioned there.  They weren't three months ago, six months
ago.  But apparently, they were until some time when they fell off and
PHRP is, of course, a defunct organization no longer accrediting people.

		They say that they are in the process of seeking accreditation from
AAHRPP but that organization doesn't report on the status of pending
accreditation so that's all we know about it.  They have sent us their
procedures under CBI claim.  We've reviewed them and they meet the
requirements of -- that are incorporated into 26.1125 for procedures
covering those following topics.

		The consent process is described inconsistently in different places
and all of the descriptions were pretty sketchy.  They need to be
clarified, reconciled between occurrences.  I'm concerned that there is
this sort of graduated scheme where you have somebody that you first
contact and then they express interest and then they gradually make a
commitment to the thing and then they finally come in and sign a consent
form and I don't like that.  I see only two statuses.  You are a
candidate or you are a consented subject and imputing commitment to the
project as some intermediate point as a consequence of some intermediate
event, I think they need to be more careful about that.

		DR. FISHER:  I think it can work both ways, John.  I think that if
they're consented on the phone to something, they may feel less likely
that they can leave.  So it's not necessarily a negative.  It can work
both ways and so I'll just say that.

		MR. CARLEY:  And as I mentioned in a risk discussion, that hazards of
the test products are misleadingly described and that needs to be fixed
and the protocol says there's no risk of disease because these are
laboratory reared, pathogen free mosquitos.  The consent form is silent
with respect to any risk of disease and that should be cleared up.  	
The methods proposed for managing information about prospective and
enrolled subjects will generally protect their privacy.  One of the data
collection forms includes inappropriate provision for the subject's full
signature in spite of the assertion that only coded identifiers would be
used.  Subjects would be free to withdraw at any time.  Medical care for
research related injuries will be provided at no cost to the subjects.  
	This is a proposal for third party research involving intentional
exposure with the intention of submitting the results to us under the
pesticide laws so it's subject to 46 -- 40 CFR 26, Subparts K and L. 
Our review of how it addresses those points in detail has been provided
to you.  I wanted to add in this context that in the protocol, they
correctly cite -- they make a commitment to comply with Subparts K and L
but they do this in a context, this is on pages 17 and 18, of asserting
that ICR policies comply with our regulations because they submit all
necessary documentation to an independent institutional review board.  

		And of course, far more than submitting paperwork to an IRB is
required to be in compliance with these regulations and they're
correctly committing to comply but I'm a little concerned that they seem
to take a pretty narrow view of what that might involve.  		I noted the
following deficiencies.  The discussions about societal benefit of new
products aren't relevant to this proposal.  They should be replaced by a
more appropriately targeted discussion of the benefits of adding claims
to registered labels.  There is no discussion of the relation of risks
and benefits.  It's not a big deal in this case.  The risks are pretty
low but the benefits are also not great and to they need to be linked
together and discussed together.  

		Risks from mosquito bites could be further reduced by treating
landings and probes as evidence of efficacy failure.  Remember that
they're already considered -- landings are already considered evidence
of pest pressure.  So it's possible to do that.  The -- to expand on
something that Kevin said a minute ago, the bite is an unambiguous end
point.  When a mosquito's abdomen is engorged with blood, you know
you've been bitten.  

		A landing is a more ambiguous -- much more ambiguous end point and
there are different ways of looking at it.  You're familiar with Scott
Carroll's landing with intent.  Lots of people have made lots of
comments about how difficult it is to discern the intent of a mosquito. 
But Scott Carroll has also explained that there -- that in the lab he
teaches the subjects to recognize the steps that a mosquito goes through
landing, preparing to bite, you know, placing their proboscis against
the skin and so forth.  In some -- I've forgotten frankly, whether it's
in this protocol or another one from ICR that I was looking at, they
were describing a landing as something that lasts three seconds.

		These are different ways of trying to differentiate between the casual
drop-by or drive-by landing and a serious landing that represents and
event that we ought to be concerned about.  That problem isn't going to
go away, so if somebody wants to use landings, we would expect them to
define pretty carefully, exactly what they think a landing -- a
countable landing is and we would be looking for something that made
that pretty straightforwardly as objective as possible rather than just
a mosquito real near my skin.  		If you also visualize a mosquito
landing on a person with a very hairy arm, they could land and be a
quarter of an inch away from the skin, a distance from which they can't
very well bite.  So that, you know, depending on the arm, depending on
the hairiness, it could take quite different amount of time.

		DR. FISHER:  Yes, yes, I think we understand.

		MR. CARLEY:  In the supplemental submission, of October 17th that came
from toXcel, they promised to revise their discussion of societal
benefits, but they didn't mention the need for a discussion of risks and
benefits and their balance.  ICR and their sponsors say that bites are
more rigorous and reliable as an end point than landings but they don't
explain probes as an intermediate point.  Certainly, a probe is as good
as a bit, although it may be a little hard to identify, harder than a
bite.  

		The incomplete descriptions of the consent process need to be
clarified.  They need to be careful about the suggested prior
commitment.

		DR. FISHER:  I guess, John, was there anything in the letter that
changed what you had told us?

		MR. CARLEY:  What I told you was what the protocol said.

		DR. FISHER:  No, I'm just asking, is there anything that is not an
issue now based on their response to you?

		MR. CARLEY:  They said that they would make many changes.  They did
not make any of those changes.

		DR. FISHER:  Okay, so I think --

		MR. CARLEY:  So I don't think -- the short answer is that it's
interesting to know what they promised to do but it doesn't alter my
assessment of what they've done.

		DR. FISHER:  Okay, and I'm just trying to move things along because
we've all read the letter and I think you're -- you know, you've done a
very good job in highlighting what the deficiencies were and if they
haven't remediated any of those clearly, then I don't think we have to
go through the letter and their response. 

		MR. CARLEY:  All right.  Now, how would you like me to proceed?

		DR. FISHER:  I'd like you to proceed with the charges.

		MR. CARLEY:  Fine.  Please go to the charge question.  If the proposed
research described in ICR's proposed prepared protocol is revised as
suggested in EPA's review, does the research appear likely to generate
scientifically reliable data  useful for assessing the efficacy of the
test substances for repelling mosquitos of the genus Culex and does the
research appear to meet the applicable requirements of 40 CFR Part 26,
Subparts K and L?

		DR. FISHER:  Thank you.  Are there any questions concerning the
ethics?  Suzanne?

		DR. FITZPATRICK:  What's EPA's feeling that they insist on bites?  I
mean, is that a deal breaker for you guys or -- 

		MR. CARLEY:  The guidelines that are out there, granting that they're
undergoing a lot of thought and will be revised, the guidelines speak
only of bites.  We don't talk about landings as an end point in the
draft guidelines and we will accept the -- we will accept studies on
bites.  What I feel is that it takes more than an expressed preference
to justify using bites.  There may be circumstances in which bites are
justified.  Certainly, it makes more sense to use bites as an end point
in a cage study than it does in the field. 

		DR. FISHER:  Are there any other questions?  Kannan?

		DR. KRISHNAN:  You indicated a concern about the lack of the
description on the eye irritation, one of the hazards.  The proponents
described mostly the low inhalation and dermal hazards and Dr. Sweeney's
calculation, the margin of exposure is based on one of those, I think
the dermal.  So I'm just asking whether there's any concern that the
margin of exposure be different if you consider that higher hazard?  I
know it's been around for awhile.

		MR. CARLEY:  Well, the highest hazard from these products is from eye
irritation and given that the application is done by technicians to a
delineated area on the arm, this doesn't seem to be a relevant concern. 
Bill points out that when a treated subject pulls his or her arm out of
the sleeve, it will -- the stuff will still be there and, you know, they
could touch it.  They could get it in their eye.

		It's important to tell the subjects that eye irritation is the biggest
problem with this.  It's also important to tell the subjects as they are
told in these kinds of experiments, not to rub their treated arm, not to
abrade it.  Not to rub it up against furniture or things like that. 
When you're all strapped and buckled into that thing and treated you're
very much aware of it.  You can't just sort of absent-mindedly stick it
in your eye.  I don't think that eye irritation would be the appropriate
basis for calculating an MOE on this.

		DR. KRISHNAN:  But I think it certainly would be important to include
in the consent form where the information would be.

		DR. FISHER:  Any other questions?  Okay, we're going to have public
comments.  Mr. Spero and I don't know if it's a Ms. or Mrs. Todd, Dr.
Todd.  Thank you.  You are together?  Okay.  

		DR. SPERO:  Hi, I'm Nick Spero, Associate Director of ICR, a contract
lab located in Baltimore.  I'd like to thank the Board and EPA for the
consideration of this protocol.  I'd like to clarify a few point on our
wanting to use bites as an end point.  ICR conducted several field
repellency studies that justified EPA registration with these currently
registered products.  

		At EPA's request, we're conducting a lab study to get a West Nile
Virus claim under safe conditions where these mosquitos are going to be
pathogen free.  But by conducting this lab study, we want to adhere to
the same parameters that we used in these field studies, specifically
the same application rate, and the same end point.  The end point used
in all of these field studies happens to be bites.  So we want to use
that same end point and we want to use the same application rate
specifically for this particular lab study.  And that's our rationale in
this study.

		Regarding control subjects, the number of bites that they would
receive, our hope and intention is that they would receive none.  Our
criteria for verification of the landing rate in these cages is that we
have an active population of mosquitos and the criteria we set forth is
five landings within 60 minutes.  The landings will occur, as Kevin
eluded -- I'm sorry, 60 seconds, 60 seconds.  As Kevin eluded to, these
mosquitos are very aggressive in these cages and these five landings
should occur in a very short period of time.  And this control person
strictly puts their arm in.  There are handlers on either side.  They
verify that the received this number of landings and that arm will be
pulled out immediately.  So John's instance of putting out -- or
suggestion of putting an aspirator in there would probably complicate
and confound things.  In our experience, this control person receives no
bites.

		So hopefully, I've addressed that concern.  As far as statistics, we
have approached the statistician and he's willing to provide analysis
for these data and we're going to analyze data in two different ways. 
We just wanted to present it in two ways.  One, if the data is normal,
the data are normal, we would do a Z test to check for normality.  If
the data are normal, then we would present the mean time of protection
and a confidence level, 95 percent confidence level and standard
deviation.

		The data proved to be non-normal and in this instance we expect to see
non-normal data with a negative skewness.  In this instance we would
correct the data.  We would reflect the data and do a transformation, a
logarithmic transformation and provide the median protection time and
the upper and lower confidence level for that.  

		The other analysis was to try to present regression or digression of
events that occur over time.  Each subject potentially will have 16
observation points.  Each subject is being evaluated at half hour
intervals and we would plot these events that occur over each half hour
period whether they received a bite or no bite, plot this data and show
the regression of protection time in this fashion.  So that's our intent
with the statistical package that we've proposed. 

		DR. FISHER:  Any other comments?  Okay  I have a question and you
know, I think we all appreciate the issue of the landing versus the bite
and you're putting good thought in it and I guess we'll all discuss it,
but it's certainly good thoughts.  But what I'd like to know is John had
mentioned a number of suggestions that EPA had made about the protocol
and about -- not the science part because I think we'll discuss that
and, you know, some of the issues are interesting to discuss.   But you
know, he also said that in terms of your response letter, there were
some deficiencies.  

		And it's difficult for us because we don't want to just say, "Well,
there are deficiencies, we don't know if you're going to do it".  So it
would be easier for us if you could address whether or not you're going
to respond and how to some of these issues that are in the protocol and
some of the -- you know, where basically they are minor issues but if
they're not addressed then we don't know if they're addressed.  It puts
us in a really difficult situation.

		DR. SPERO:  We have every intention of addressing all of the issues
that EPA brought up, the ethical constraints with the consent form.  We
have every intention of clarifying all of the deficiencies in there.  I
think we're on the same page with EPA.  We want to address all things in
the protocol that are unclear or conflicting and we have every intention
of taking EPA's advice regarding the protocol and the consent document.

		DR. FISHER:  And so was there a reason in the response letter that the
details of that wasn't included?  Was it a time constraint?

		DR. SPERO:  It sort of was a time constraint.  This was the last
minute response but we have every intention of complying with EPA's
instructions.

		DR. FISHER:  Okay, and I would just suggest in the future, not just
for you but other respondents, that that will make our life and
everybody's life a lot easier if -- you know, I understand there's not
enough time to redo.  You have seven days and it comes out, and you
know, we all understand that.  But it probably would be very helpful to
say we intend on following this, this, this, this, and you know, we all
understand that but it probably would be very helpful to say we intend
on following this, this, this, this.  I mean, John does a lot of work,
and --

		DR. SPERO:  Absolutely.

		DR. FISHER:  -- it's difficult for us to be in this netherworld of
whether you're going to do it or not.  So I'm happy to hear and I wanted
to give you the opportunity to say you were going to do it and just
advice in the future that we haven't had time to revise the informed
consent but we are going to do exactly what was suggested, anything else
that was suggested, et cetera, et cetera.

		Okay, KyungMann and then Sean and -- 

		DR. KIM:  I have a question about the Q test that you talked about. 
You said that you're going to count the number of -- I mean, the
incidents of landing over the 60 and 30-minute period.  What happens
when somebody has a bite?  Are they still kept in the experiment or are
they withdrawn?

		DR. SPERO:  Yes, the criteria for breakdown is two bites within a half
an hour or one-five minute period or one bite in each of two successive
periods.

		DR. KIM:  And after that what happens, the patient drops out of the --


		DR. SPERO:  That person has broken down, so that -- 

		DR. KIM:  That means you cannot do that Q test.  

		DR. SPERO:  Well, each person could have up to 16 --

		DR. KIM:  Right, so when a subject drops out, that subject does not
belong to the denominator in the Q test.  So your Q test is invalid.

		DR. TODD:  Dr. Kim, we retain the statistician, Dr. Ralph Piedmont
from Loyola College up in Baltimore and he is going to handle -- we are
not statisticians.  He's going to handle the statistical analysis, so
we're relying upon his expert advice in this.  So our familiarity with
the nuts and bolts of the Q test is limited. 

		DR. KIM:  The other question I have is with eight hours of exposure
what happens if there's no bite until the eight-hour period?

		DR. SPERO:  That wasn't addressed in the protocol.  If in the event
that someone goes eight hours without a bite, we will assume nothing
beyond eight hours.  We will claim the protection time to be eight hours
for -- 

		DR. KIM:   How are you going to do the normal analysis?

		DR. SPERO:  Pardon me?

		DR. KIM:  How are you going to do the normal analysis that you
indicated?  How are you going to handle the censored observation?

		DR. FISHER:  I don't think they know.

		DR. SPERO:  That's the truth.  I'll have to check with the
statistician on this.

		DR. FISHER:  Right.

		DR. KIM:  So protocol as it currently  it is -- 

		DR. FISHER:  Right, and that's what we're going to have in the
discussion.  This is targeted for their questions and they do not know. 
And we'll discuss what might be appropriate or inappropriate during the
discussion.  Sean?

		DR. PHILPOTT:  Just a real minor point of clarification, in the --
with respect to the untreated controls when they stick their arms in the
cages to look at landing pressure, in the protocol it says that they
will shake their arms.  In the consent document, it says that the
mosquitos will be brushed away by staff.  Which is it?

		DR. SPERO:  They'll be shaken away and we'll remove that inconsistency
between the protocol and the informed consent.  

		DR. LEHMAN-MCKEEMAN:  I have a question about rational and then just a
minor detail question.  As I understand it, you want to use bites
because that's how the field test was done. 

		DR. SPERO:  That's correct.

		DR. LEHMAN-MCKEEMAN:  But as I also understand it, Culex wasn't in the
field test, this genus of mosquito.  So what exactly are you comparing
between the lab study and the field study?

		DR. SPERO:  The same end points and the same application rate are two
of the main things that we want to keep consistent between the lab and
the field study. In the field studies we were consistent with our
application rate in all field studies and the end point that we used for
protection time was bite.  So we want to use the same two things in the
lab study. 

		DR. FISHER:  With respect to that, again, just execution now, in
looking at the products at least one of them is identified as a
repellent spray but the protocol indicates that the agents will be
applied with a syringe onto the skin.  So was the field test done that
way as well?

		DR. SPERO:  That's correct.

		DR. LEHMAN-MCKEEMAN:  And so then the product isn't being applied
precisely as it would be used.

		DR. SPERO:  That's correct.  We had an application amount that we
wanted to cover, a certain area, 250 square centimeters.  We dispensed
it into a receptacle and then applied a known amount to that area.  

		DR. FISHER:  And does that mean the formulation is not important with
this product, I mean, the delivery system?

		DR. SPERO:  I can't answer that if the delivery system is important. 
We wanted to 

-- we didn't check for the delivery system.  We checked to verify that
we applied a known amount to a known area.  

		DR. SHARP:  So the environment that the subjects are in is of
increased heat.  It's about 80 degrees higher humidity and I would
imagine that it's a bit stressful for folks to for the first time throw
their arms into this cage containing all these insects and that you
would encounter certain folks that would be a little light-headed or
faint or anything of that sort.  Is that a common reaction in your
experience?

		DR. SPERO:  It's an uncommon reaction in my experience.  Most of the
test subjects that we have used had been in lab studies many times and
we've never had that comment from any of them.  They're free to use the
facilities as needed to receive or make important phone calls if
necessary but for the most part, we want them all to be in the same
environment, that would be in the insectary.  Never had a complaint that
the heat was too oppressive or the humidity.

		DR. SHARP:  What sorts of complaints have you had in terms of subject
burden or anything?

		DR. TODD:  Possibly boredom, tedium, they run out of jokes and after
they've read the newspaper a couple of times, they're glad to get out at
the end of the study but I think boredom is the main complaint and the
humidity.  It is very humid there.  

		DR. FISHER:  Jan and Rebecca.

		DR. CHAMBERS:  In your experience are the mosquitos pretty consistent
in their biting pressure batch to batch, these lab reared mosquitos?

		DR. SPERO:  These lab-reared mosquitos would be for all intents and
purposes sisters.  It's the same cohort of mosquitos that are used in
the study and we find that they are pretty consistent from cage to cage,
yes.

		DR. FISHER:  Rebecca?

		DR. PARKIN:  If they're sisters, no wonder they're fighting.  I have
one question about your volunteer database because I'm just not familiar
with how companies like yours compile them but it strikes me that an
individual to be eligible for this study has to have two arms.   Is that
an eligibility requirement to be in your database?  It's never stated
anywhere.

		DR. SPERO:  It's a requirement that's just kind of assumed.  It's not
stated.  All of our subjects do currently have two arms.

		DR. PARKIN:  Would that exclude them from your volunteer data base?

		DR. SPERO:  Yes, it would exclude them, yes it would.

		DR. PARKIN:  Because, I mean, I think this is a generic issue that we
saw yesterday too, that it just struck me after I went home, I thought,
"Gee, people have to have two legs in order to participate or they
wouldn't be able to randomize the legs for treatment.

		DR. FISHER:  Let's not -- I don't think we have to discuss why that's
an issue, okay?  I have a question for clarity.  This is an improved
product.  Your -- correct?

		DR. SPERO:  Yes, correct.

		DR. FISHER:  And you're basically   -- I guess the goal is a label
goal to see whether or not the time period is extended in terms of
protection, is that -- 

		DR. SPERO:  The goal is to get a West Nile Virus claim.

		DR. FISHER:  A West Nile Virus claim.  Okay.  Now, you've chosen to do
the syringe to be equivalent to the field.  Right?  		DR. SPERO: 
Correct.

		DR. FISHER:  As opposed to this is in spray form that the consumer
gets?

		DR. SPERO:  That's correct.

		DR. FISHER:  Okay, so how is this relevant in some sense to the label.
 How have you equated whatever the syringe amount is and its consistency
across a particular surface and it's concentration across a particular
surface with the actual way that it would be applied and actually spread
using the product?  And what kind of dosimetry did you use if any, to
number one, estimate what amount and how it's going to be applied and in
that sense, what's the generalizability to the actual use of the
product?

		DR. SPERO:  We did not dose range finding for the field studies.  We
used the current EPA guidelines and we used their recommended rate for
application.  To be consistent we applied -- we wanted to know what was
applied to the arm so we dispensed inter-receptacle and applied a known
amount to a known area and this was prior to the role, prior to this
dose range finding schedule and that's why in this particular lab study
which is a verification for a West Nile Virus claim, we want to use the
same parameters that we used during the field studies.

		DR. FISHER:  Would it be helpful to -- I understand, I think that's a
good idea to equate field and lab.  I think it's an excellent idea but
would it be of value to also do the dosimetry to understand what this
equation means in terms of the use of the product?

		DR. SPERO:  I think in the future we will do a dose range finding. 
For this particular thing compare lab to field, I'm not sure there's a
value because we didn't do that in the field studies to know what -- 

		DR. FISHER:  I understand that, but I'm just -- I'm wondering and we
were talking to some people, in terms of the usage and the value of the
usage for EPA, it doesn't sound like dosimetry is that hard or
time-consuming or expensive to do, so I'm wondering if there's a value
to add a dosimetry just so it could provide more information about the
generalizability of this data to the actual product use itself.

		DR. SPERO:  We'll consider that.  I don't have an answer for you at
this point.  Thank you.  

		DR. FISHER:  Okay, thank you so much.  Thank you very much.  Are there
other public comments?  All right, I wanted to make a general point that
is not directed at this particular study but I think it's very important
and should be required that either the statistician that people are
using is present at the meeting or the statistician has consulted prior
to a protocol being submitted.  We don't want to put sponsors in a
position in which a study once it's conducted, is inefficient because of
analyses that are either inappropriate or there wasn't enough power, et
cetera, et cetera, and once again, it kind of places the board in this
kind of netherworld to say, "We have a statistician, it's going to be
analyzed.  We don't know why -- you know, we don't know why they said
it's okay and we haven't consulted with them yet." 		So, I mean, this is
critical and EPA, you know, when you see that in a protocol, we really
need explicit statistical analysis plan in the future.  Michael?

		DR. LEBOWITZ:  Yes, I'm wondering in fact, given the way things are
scheduled whether the sponsors or their agents don't have time to
respond sufficiently to EPA for us to see those responses before we have
to deal with these issues to evaluate whether in fact, that they've met
adequate scientific and medical criteria.  I mean, it's a concern to me
and I would ask Mr. Carley to repeat again when the comments from EPA
were submitted to toXcel and/or ICR and so as to know whether they had
enough time to reply to us.

		MR. CARLEY:  I don't recall the exact date but it was within 24 hours
of our completion  of the review that we passed it onto them.  

		DR. FISHER:  Months, how many --

		MR. CARLEY:  Twenty-four hours, Dr. Fisher.

		DR. FISHER:  No, no, no, not your response.  We're just trying to
figure out from today how long ago was it?

		MR. CARLEY:  The review -- the date on the review is September 24th. 
They had this by September 25th.

		DR. FISHER:  Yes. No, no, we just wanted to get a sense of how long
they had to respond.

		MR. CARLEY:  Do you now have that sense?

		DR. FISHER:  Yes, a month, right, which is fine.  KyungMann?

		DR. KRISHNAN:  I -- I mean, I've been serving on this board since
April of last year but one of my most frustrating point is that EPA
presents the review of the studies with a lot of statistical deficiency
and none of these are addressed when EPA gave a presentation, and I'm
given the charge to assess whether it's scientifically  sort of
sufficient to go ahead.  And with EPA not addressing any of these
deficiencies, I'm sitting here finding all these difficulties,
deficiencies.  It's been a very, very awkward position.  And I still
don't understand why EPA cannot have a statistician to get the protocols
and provide the appropriate set of a particular protocol before it comes
to hear.

		DR. FISHER:  I also think that sometimes -- I think we understand as a
board that some studies are seasonal and when there is this kind of
three months short time span, it may be that it's important for the
sponsor, for us to get the protocol even though they may not have the
time, in terms of their written response to have responded to all the
criticisms.  But that should be rare and this is not a case.  This is
the laboratory study and I think some of us are wondering why this was
brought to us.  When you know, John and Kevin identified deficiencies,
those deficiencies may be addressed but there's no record that they will
be addressed.  There's no statistical analyses plan and I think one of
our questions is, why -- it's not a question really and this is a
dynamic process.  So it's not even a criticism.  

		I think we're -- we don't want to be turning down studies simply
because there's inadequate information.  We would prefer that everybody
on board understands the kind of information that we need so that we're
not in a position that we have to turn it down because of absence of
information or turn it down because the analyses wasn't presented to us
the first time around and then the whole study, everybody put in all
this effort to do the study and it's unanalyzable.  And I think going
forward -- and I'm not sure how we -- you know, we'll discuss how we
want to handle today but going forward, if there's no statistical
analysis plan, then we should not see the study.  

		Board comments?

		DR. CHADWICK:  Amen.  Well said.

		DR. FISHER:  Michael?

		DR. LEBOWITZ:  I absolutely agree and I think we've already talked
that if the appropriate information hasn't been supplied to us at least
the week before we have a meeting then we really can't deal with it.  

		DR. FISHER:  Okay, so I guess we're going to have the science
discussion and we'll see where we are with that.  Jan, could you begin,
please?

		DR. CHAMBERS:  Sure.  This is a hypothesis based study that it would
provide eight hours of protection.  I found it very clearly written.  I
don't really think there's a lot of need to go through a lot of the
detail.  It seems like there are just a few critical points that need to
be discussed.  One is the number of insects per cage and I believe the
sponsor has decided to up that to 200 to be consistent with EPA's
guidelines, so that's taken care of.

		One of the other points that's on the table, I guess, is the dosage
and dosimetry.  I think at risk of sounding inconsistent with some of
the things that I've said before, I think their rationale for using the
same dosage which is the standard guideline dosage, makes sense.  They
are trying to compare a registered product to get another label claim of
mosquitos that were not available during the field studies and I think
that this lab study needs to match that just as well as possible so that
that comparison is valid.  So I think that the dosage that they are
suggesting, which is what they used previously, is appropriate.

		With respect to the suggestion you made a little while ago, should
they do a dosimetry.  I'm not really sure what that information would be
used for at this point and it seems like that would be at this point,
rather a waste of effort. 

		Now, in terms of future guideline adjustments and that sort of thing,
it's probably that's something that should be factored in as we had
suggested in the past.  But, again, this is not an independent study. 
This is to match up with something that's already been done, so I really
think that makes some sense.

		Consistent with that then, they used bites in the previous study and
they are suggesting bites again to be consistent with this laboratory
based study so that they can expand the label of an existing product. 
So that makes sense to me, too.  Of course, the animals -- these
mosquitos are not disease-carrying mosquitos, so there is no danger from
the vector here.  So you know, the bites that people who do informed
consent and understand that they're going to be bitten and if they're
sensitive, they wouldn't participate and that sort of thing.  So they
would have a limited number of bites.  Presumably the controls would
shake them off and would not be bitten.  So I think again, for the sake
of consistency, the bits make sense.  So from the standpoint of the main
sticking points here of whether it's bites or landings, in this case, I
think bites makes sense.  The dosage, the standard dosage makes sense
from the standpoint of consistency with the field study that already
exists for the label.

		DR. FISHER:  Thank you, Jan.  Lois?

		DR. LEHMAN-MCKEEMAN:  I guess in a moment of levity, I'd just like to
say that as I read this protocol, I was reminded of Jerry Lee Lewis and
I think this is going to be a whole lot of shaking going on. 
Nonetheless, I think that my -- the only thing that I would add to what
Jan has said from a scientific point of view and although I don't have a
really strong feeling about this, I tend to disagree on the dosimetry
relative to what Jan has just stipulated.  As I understand the objective
of this particular study, it's specifically to support a claim against
the genus that carries the West Nile Virus and so it's not clear to me
how this is directly comparable to the field study given that that
particular genus wasn't in the field study.  And I don't know and I
cannot tell what the comparison is to the field study and so if the
claim is, in fact, to be made specifically against a mosquito that can
carry a potentially serious disease where I come out is the guideline
itself may be sufficient. 

		So I think, you know, my perspective is this notion that there needs
to be some kind of direct comparison to the field study is not
persuasive to me.  And, in fact, I think in some respects this study
should be looked at almost in a stand-alone way and that is to say, it's
a laboratory study.  For all the right reasons, it's a laboratory study.
 And under that condition, it could be perfectly acceptable to simply
follow the guidelines of the study.  So I  think what I'm quibbling with
is really what the sponsor has said its objectives are and its rationale
for using the paradigm that its provided.  

		So I think ultimately if the goal is to develop a claim against this
particular genus scientifically, I don't see any issues with this study
actually generating data that are meaningful.  I am simply not compelled
by the comments that this needs to be directly comparable to a field
study.

		DR. FISHER:  Thank you, Lois.  Suzanne?

		DR. FITZPATRICK:  I don't have much to add.  I just wanted to say, I
agree, I think that even though the sponsor is the benefit -- ultimate
beneficiary, there is -- there are consumers that would really like to
have a -- some type of insect repellent with this type of claim on it. 
I think that if you ask consumers that they would agree that that's
something they would like to see.  I agree with Jan, I don't think --
although I appreciate Lois' comments, I don't think they really need to
do the dosimetry for this particular study.  The standard dose is
probably what's in the EPA guidelines, isn't it?  Using that, so I'm
assuming that they're going by that.  

		I don't think -- we talked a little bit about whether it should be
continuous or intermittent.  I doubt you could do a continuous study
with people having their arms in these cages for eight hours
continually.  So, intermittent is fine.  I don't have any problem with
landing versus bites.  As long as the subjects are adequately consented
and they really understand what they're getting into, then they're
making the choice of whether they want to be in the study where there's
a possibility for getting bitten and that's their choice to make, you
know.

		So as long as the consent adequately describes the actual risk.  I
think that's the only other thing.  I do want to comment that generally
when any regulatory agency, like EPA writes a deficiency letter, even
though we may not see that, that the sponsor, you know, adequately puts
all deficiencies.  They generally do.  Usually when you get a deficiency
letter, especially things that are easily corrected, the sponsor will
respond to them.  I think that we need to have a little faith that EPA
when they write a deficiency letter, make sure that they, you know,
adequately puts -- you know, addresses all of those.  We might not get a
chance to see it.

		DR. FISHER:  Well, I just want to respond.  It's presented to us as
they did not respond.  So that was not -- it was not presented to us as
we assume they will respond.  It was presented to us that they failed to
respond.  So I think this has to be kind of clarified, because it's not
up to us to assume EPA has faith in the sponsor.  They have to
articulate that.  Rebecca?

		DR. PARKIN:  I don't have any other key point to add to what the
others have already said except to say that I think my position aligns
more with what Jan has said.  I'm not particularly concerned that this
is a different mosquito.  It has to be a different mosquito to
demonstrate what they're trying to demonstrate and I don't -- I'm not
concerned about a particularly different protocol, just because the
mosquito species has changed, so I'm comfortable with -- I would make
the same comments that Jan made.

		DR. FISHER:  Okay, discussion, Mike and KyungMann?

		DR. CHAMBERS:  He was just being gracious.  I think just to defend
again, what I'm saying about consistency with the field studies.  As I
understand this, the label will be the same as it is right now with just
the added efficacy against Culex statement and as long as the label is
not going to change in any other way, then I feel like the experimental
conditions in terms of dosage and that sort of end point needs to be the
same.  That was my perspective.

		DR. LEBOWITZ:  Yes, may concerns have to do with the choice of one
member of the genus Culex and we know that other members like tarsalis,
in fact, do carry West Nile Virus and we know, in fact, that other genus
carry West Nile Virus, including Aedes, different members of the Aedes
genus.  And I don't know whether all of those other mosquito -- types of
mosquitos were out in the field to insure that this product will,
indeed, protect against all the different mosquitos and all the
different conditions in which, you know, they want to have a claim,
vis-a-vis, West Nile Virus, and in fact, in terms of matching the field
study, I also want to know whether the temperature and relative humidity
conditions in the lab are going to match the range of temperatures and
relative humidities that they found in the field to -- for this specific
study to insure, in fact, that what they saw in the field would apply to
this member of the genus Culex under the same conditions.  

		Now, then of course, there's the generalization again whether the lab
study will also mimic other weather conditions in which various
mosquitos carrying West Nile Virus are active and so my take on it is
that the answer to those questions would be in the negative so that the
lab study will neither mimic the field study sufficiently, nor cover all
the other types of mosquitos that carry West Nile Virus under various
weather conditions and since the objective is to be able to change the
label to say it protects against mosquitos carrying West Nile Virus, I
have a great deal of concern from that standpoint.

		DR. FISHER:  KyungMann?

		DR. KIM:  My reading of the protocol states that the hypothesis of the
study is such that the repellent samples are expected to provide
eight-hour personal protection from the  -- this species of mosquitos
and I presume that means sort of an average of eight hours, right?  But
if you have an observation period of up to eight hours, how are you
going to verify that?  I'm not so sure whether that study can provide
that data because, your observation period is limited by eight hours. 
If there's any event occurs before that observation termination period
you will have failed, right?

		DR. FISHER:  You're looking at me?  I don't know.  

		DR. KIM:  Right, right?

		DR. SPERO:  I think those such questions are better addressed to
Kevin.

		DR. KIM:  And the other part is that -- I mean, this information is
particularly sort of interesting in light of the fact that they justify
the sample size based on having a standard deviation not greater than
two hours after eight hours of testing.  If you test only up to eight
hours, you cannot -- I mean, you cannot get that kind information
because of the limitation of the observation period and currently it's
completely ignoring this answering and I asked the sponsor
representation earlier how they're going to handle this answering which
is not provided in the protocol.  That means sample size calculation
itself is in question.

		DR. FISHER:  Let me -- my understanding and this is probably not
correct, but first of all I thought they were going to handle the
sensoring by having alternates and therefore, there wouldn't be
sensoring.  If somebody leaves, then they'll have an alternate, so that
only people who have completed the eight hours will be included. 
Assuming -- but I think -- are you saying that what if one of the people
that decides to leave was bitten in the interim?

		DR. KIM:  I mean, people withdrawing early, you should never just
remove them from the analysis.  They have information about the
protection time.  The other parties that even if somebody has followed
the entire duration of eight hours but if they don't have any bite, they
are observations, which is different from dropping, although from a
statistical analysis perspective, they contribute in the same way.

		DR. FISHER:  Right but my -- 

		DR. KIM:  So --

		DR. FISHER:  Ask your question, because I have a social science
background, so I'm just trying to equate it to that.  We use Likert type
scales and let's say we have a score of one to eight, so we are able to
determine a standard deviation on a scale of one to eight, so I'm just
-- I'm not -- you know, I may be totally off-base but I'm just not clear
why if nobody is making claims above eight hours, why if there's been no
bites, it becomes zero and not eight.  

		So I guess, I'm just not understanding this and why -- I understand in
the purest form but I guess maybe Kevin, if you want to -- 

		MR. SWEENEY:  Can I make a comment?  Okay.  First I'll comment
regarding your comments about, I think the temperature and relative
humidity are fairly similar to the field -- the field tests were done in
Southeastern United States, I think in Georgia and Florida, if I
remember correctly.  In terms of the species present, there was a
variety of genera present but none of them were at least principal West
Nile Virus factors, but there were other genera presences, so I think
that there's been established repellency against those genera of
mosquitos, you know, and as far as testing other vectors of West Nile
Virus, frankly, that would be nice but we generally don't require them
to do more than one vector of a disease, generally speaking.  

		That's just a policy we've had.  In other words, we feel comfortable
using Culex quinquefasciatus as an indicator or Culex tarsalis as an
indicator of West Nile Virus transmission as the two principal species
that they could test.  

		The other issue is if you're doing a lab test is lab rearing and
availability of lab colonies.  Quinquefasciatus tends to be more easy to
rear and just easier to get.  There are other vectors of West Nile Virus
in the United States but really the principal vectors are those two
Culex species and so we try to go with the major vectors, if you will,
when we do the testing.

		I mean, you know, you could have sponsors testing against all kinds of
species but I mean, we try to use indicator species to do that.  Then,
going back to the statistical part of it, the issue about why is the
test only running up to eight hours, your point is well-taken.  They've
chosen to run this test for eight hours and they're banking on the fact
that they're going to have 100 percent success at eight hours.  Okay?  I
don't know that to say to that.  I mean, that's their choice when
they're doing the test.  

		I understand your point.  I go back to your point about censoring of
data; you know if somebody fails within that eight hours, then what are
you going to do with the data point and can you get the correct mean or
derive a median even that would be adequate?  I mean, those are good
questions and I mean, I can't really as a sponsor, respond but the point
is well-taken that the test period might have to be longer if there are
failures.  The point is also well-taken and again I'm not an expert on
the Q test either, so frankly the use of proportions and all to me was
-- I mean, I guess if you have everybody in there and they run the whole
gamut, you're good and if they don't, then you may have issues with
variability, you know, I don't know.  I mean, so that's -- I mean, so
that point is well-taken as well.

		DR. KIM:  If I may, I see Q test as a sort of secondary, supportive
analysis.  The primary endpoint in this case is timed to efficacy
failure and so they chose eight hours.  If no one fails, then this will
be a success because that established the minimum protection time of
eight hours.  If any of the subjects were to have a bite before that
eight hour day is over, this study has failed in terms of establishing
the claim.

		DR. FISHER:  So if that's the case, is that -- when it came back to us
if we saw that there was a bite prior to the eight hour, within the
eight hour period, we would say that it's not used, that conclusions of
its efficacy is a failure?

		DR. KIM:  No.  When I said failure, it's that their claim is going to
be a failure.

		DR. FISHER:  Right.

		DR. KIM:  But the study is perfectly legit.

		DR. FISHER:  Okay.

		MR. SWEENEY:  In other words, I think what he is saying is it might
only last six hours.  So therefore, the recommendation would probably be
you can't claim eight for that vector.  You would have to claim less.

		DR. FISHER:  Okay.

		DR. JOHNSON:  So is the current claim for eight hours?

		MR. SWEENEY:  The current claim is for eight hours.

		DR. FISHER:  Michael.

		DR. LEBOWITZ:  Yes, so one could say up to eight hours.  The
difference between this and a discrete scale is that this is a
continuous time.  It could actually be 10 or 12 or whatever, but they
can't say, you know, eight hours plus or minus or even six hours plus or
minus if there are people who weren't bitten by the time, by the end of
eight hours.  They can say up to eight.

		DR. FISHER:  And if one out of ten is bitten, do they say 90 percent
effective up to eight hours?

		MR. SWEENEY:  Generally, we don't have percentage claims on the label,
no.

		DR. FISHER:  It's all or nothing.

		MR. SWEENEY:  At this point in time, it's all or nothing.  We are
discussing some labeling modifications in the future.  There's another
work report on that.

		DR. FISHER:  Okay.  So it sounds like there's a way to interpret this
data, that it's important to know that a single less-than-eight-hours
would mean that the claim could not be met, that anybody bitten within
that eight hour period, they have to be included as data.  If they're
leaving, they're still included as data because they were bitten and
that's where it's still going to be a confirm bite.

		MR. SWEENEY:  Any bites would be recorded.

		DR. FISHER:  Okay.

		MR. SWEENEY:  But a confirm bite would remove a subject.  But if
somebody left before they got a confirm bite, then that data would be
included as well within the dataset and accounted for.

		DR. FISHER:  Okay.

		MR. SWEENEY:  You know, again I think the study can be run and be a
valid study.  It's just a matter of the endpoint that you're looking
for, whether or not the calculation, the analysis.

		DR. FISHER:  And does the label claim say that it's protective against
some species that carry West Nile or will the label claim say it's
protective against West Nile?

		MR. SWEENEY:  The label claim generally, the kind of claims that we
allow would say this product will repel against vectors that may
transmit West Nile Virus.

		Excuse me.  Vectors or mosquitos that may transmit West Nile Virus.  I
mean that's generally the way the claim is made.  So it doesn't qualify
which vector.  We're using the culex as an indicator.  We have data on a
lot of other mosquito species.  So we're comfortable with that statement
and it's also a statement the consumer understands.

		DR. FISHER:  Thank you.  Dallas, did you have --

		DR. JOHNSON:  Yes.  Some of these tasks that they propose were not the
same ones that they spoke about when they were talking over here.  The
Fishers' exact test is kind of a test like a binomial proportion and
generally these tests are not very powerful when you have small sample
sizes and it seems like there are small sample sizes here.  And I was --
one of the things I liked about some of the other studies is this
Kaplan-Meier statistic and it seemed to give good information.  And so I
think maybe they might want to look into that or ask their statisticians
to look into a Kaplan-Meier type test as to getting some estimates and
it might be able to handle some of the censoring part  as well.

		MR. SWEENEY:  I just want to make one comment just about the
statistics in general and your comment about reviewing this
statistically.  I think just going back to the beginning of this process
frankly, I mean, one deficiency that has been noted consistently has
been the fact that even when there are protocols and I don't care who
submitted the protocol, we haven't seen a full blown explanation of how
you're going to analyze the data and I agree it just needs to happen. 
They really need to go through and state the possibilities and how
they're going to treat the data and if this happens and the whole bit. 

		Generally, I think what we've seen researchers liking to do is they
say, "Okay. We're going to analyze the data using maybe this test or
that test," until they see the data and then maybe they'll want to
transform it somehow or whatever.  But I think that really just needs to
be explained up front so we're comfortable with how they're going to
approach the analysis and that the design will match how they're going
to do the analysis.  So you're going to get the data you need to do the
analysis.

		DR. KIM:  Related to that comment is that when you peak the analysis
method after you look at the data, that's what is known as a data-driven
analysis which always leads to biased analysis.  So an analysis plan has
to be up front, probably specified to avoid these kind of --

		DR. FISHER:  Because it also seems like they say we'll collect the
data and then we'll give it to our statistician.  So I think what the
Board is saying is we do not want to see protocols unless there's a
statistical analysis presented, okay, an analytic plan.

		MR. SWEENEY:  A plan.

		DR. FISHER:  We do not want to see it.  It's premature.  Okay.  So I
don't hear any lack of approval for this study.  I think there are a few
things that could be firmed up.  But it seems as if the -- all the
issues that we've raised are not limitations within the constraints of
how this is going to be used.  So if that's the case then we answer A by
saying it is likely to generate scientifically reliable data for the
purposes that we've been told it will be used.  Any -- otherwise, okay,
so we'll go on to Ethics.  All right.  Sean.

		MR. PHILPOTT:  Okay.  In general, once again, thank you, Mr. Carley,
for your very detailed review and I agree with your factual observations
of the strengths and weaknesses of the study as detailed in your and Mr.
Sweeney's report and believe that once those changes are incorporated
into the protocol that the proposed research will meet the applicable
requirements of subparts K and L.  I'm loath to wade back into the bites
versus landing issue because after hearing all of the presentations of
the last three days I'm probably more confused than ever.  But because
this is a laboratory study involving pathogen-free mosquitos, I am a lot
less concerned about the desired use first confirmed bite versus landing
with intent to bite for this particular protocol.

		In general, I find that the risk to study participants are minimal and
justified by the likely benefits and as with all of the studies that we
appear to have been reviewing lately the risks to participants are
really threefold: a reaction to the test materials themselves, exposure
to biting arthropods and then the exposure to arthropod-borne diseases. 

		In the case of the first risk, the active ingredient of these
formulations and it's my understanding that this is already a
commercially-available product.  This is simply research to support a
labeling change.  So this has already been on the market for awhile.

		I do agree with Mr. Carley's recognition of the mischaracterization of
the toxicological risks in the informed consent document and recommend
that those changes be made.  But since this is already on the market,
it's been previously reviewed and approved and that volunteers with
known allergic reactions to insect repellents and common cosmetics are
excluded from participating.  You have a limited amount of skin being
treated.  It is unlikely that enrolled participants will be at increased
risk of experiencing adverse side effects of pond exposure to the test
materials.  There are also clear stopping rules and plans for medical
management of any side effects or adverse events associated with product
exposure.

		The endpoints of the study protocol, of course, do require two
confirmed mosquito bites to document breakdown of repellent
effectiveness and there is also the risk to the untreated controls of
bites even through they're apparently going to be doing a lot of
shaking.  But reactions to mosquito bites are usually mild and easily
treated with steroidal creams and these will be on hand as will other
treatments like calamine and rubbing alcohol to alleviate these minor
symptoms and by excluding individuals who have a history of severe skin
reactions to bites I think it further minimizes the risks of a
participant experiencing a severe physical reaction to a bite.

		With respect to arthropod-borne illnesses, obviously we're using
laboratory bred and raised mosquitos that are considered to be pathogen
free, thus minimizing the risk of vector borne disease.  Mr. Carley
noted that this is not listed in the informed consent document.  Because
it's not a risk, I'm not sure that you would want to put it in the risk
category as you may -- or at least as your presentations seem to imply. 
I would include it in the informed consent document more to alleviate
participant worries because West Nile Virus is the thing that's in the
news a lot.  So I would actually put it in the informed consent
document, but not characterizing it as a risk.

		And as has already been noticed, there's no clear justification for
excluding individuals over the age of 55.  That seems to be a leftover
from the field studies where because West Nile does adversely impact
people over the age of 55 more that you do want to exclude those
individuals.

		I do believe that there needs to be a more detailed explanation of
study recruitment.  We've discussed this before in reviewing ICR
protocols.  But as currently written from what I can evaluate,
compensation for study participation doesn't seem to be so high as to be
unduly influential and the exclusion of employees and contractors of the
laboratory and the sponsor and other interested parties as well as
family members also helps alleviate concerns about coercion and undue
influence.

		You're excluding minors and pregnant women with confirmatory,
over-the-counter pregnancy tests and I'm assuming that the two
alternates will also be used in a manner to help protect the
confidentiality of these results if a woman has to be excluded from a
study because of a pregnancy test that you will be able to use the
alternate system as a way of dropping her from the study without
compromising her privacy in this regard.

		A couple of really minor points that need to be corrected.  I'm glad
to see that the differential risks to untreated control subjects are
listed in both the protocol and the informed consent document.  But
there's this misleading statement about the shaking in the protocol
versus the brushing off by ICR staff of the mosquitos that land on the
untreated control arms.  I do not see in the informed consent document
anything talking about the attractiveness component of the study whereby
the treated subjects will stick an untreated arm into the cage before
the efficacy study begins to demonstrate that mosquitos do like them and
that needs to be added to the informed consent document.

		And one of the benefits that's listed in the informed consent document
is bringing a new repellent to market.  This isn't bringing a new
repellent to market.  The repellent is on the market.  It's a labeling
change and I think that that should be modified accordingly.

		DR. FISHER:  Thank you, Sean.  Gary.

		DR. CHADWICK:  I agree with both Mr. Carley's analysis and Sean's.  I
am, I think, comfortable with the argument that ICR makes regarding
landings versus bites and again I think if they minimize the potential
allergic reactions from mosquito bites that this would be truly minimal
risk and ethically acceptable.

		I do think there needs to be some addressing in the consent form of
the West Nile Virus purpose of the research as well as its risk
categorization and I don't think you can say that there is no risk but
they certainly can say that it's not expected and that sort of thing and
I do think it does provide useful information for potential subjects.

		As far as the age of 55, I also wonder if that is in there for
comparability  to the field study.  I agree that in a lab study there's
no reason for it, but I just wonder whether or not that's an issue that
if the other group was limited to 55 whether or not there's any and
that's a scientific question I'll leave to EPA and others.

		And the other potentially scientific question that came up was you
were discussing the amount of arm hair in this case or leg hair or
whatever.  Is that something that ought to be exclusionary?  And again
that's more of a scientific question than an ethics.  But I'm done. 
Thank you.

		DR. FISHER:  Thank you, Gary.  Richard.

		DR. SHARP:  I won't even go there as a follow-up.

		Two smallish comments.  First is a comment about minimization of risk.
 Unlike some of the other studies we've seen, it seems to me that the
primary risk to subjects here actually are not related to the
application or even to the exposure to insect biting.  I actually think
the primary risk here may be associated with the environment and being
asked to sit in a room with slightly elevated heat and so forth for that
amount of time.  And there's language in the protocol that's a bit
ambiguous with regard to how the investigators understand that
particular type of risk.  So they stated at one point that eligible
subjects have to be able to "withstand the rigors of testing."  So they
do acknowledge that there must be some physical burdens associated with
that.  They do have the restriction of individuals above the age of 55
which may actually be a reflection of the fact that they want to
minimize that risk as well.

		On the other hand, when you look at the medical monitoring that's in
place, it simply consists of having a first aid kit available and cell
phones to be available to place an emergency call.  So I would suggest
that given that ambiguity that's there, the investigators provide a
little bit more detail with regard to the medical monitoring that
they're going to put in place and I would suggest that they have a
clearer plan of attack in place to deal with what are likely to be very
rare events, somebody passing out or something like that but nonetheless
something that may come up and for those of you in the room that have
done field based work and have done any work really probably related to
this if something bad can happen occasionally it does.  So we need to be
ready for that or they need to be ready for that.

		The second point is coming back to the point that Mr. Carley raised
having to do with the pre-enrollment changes of behavior that are being
requested here and these are changes to clothing and hygiene that are
being requested prior to the actual consenting of the research subject. 
I wanted to come back and mention that I don't think that that's
problematic in this case.  I don't think that it's atypical either of a
whole range of different research studies that take place. But were they
asking subjects to make more extensive changes in their behavior, I
think that's something we ought to be concerned about.  But in this
particular case, my personal opinion is that that's not something that I
find troubling.

		DR. FISHER:  Thank you.  Any other comments?

		Okay.  So it seems as if there are some changes that need to be made
but none of them, we are confident that they will be -- yes, Sean.

		MR. PHILPOTT:  Actually, so just to get back to Gary's question, it
sounded like he did throw it back to the science group about the 55 age.

		DR. CHAMBERS:  I don't have a clue.

		DR. FISHER:  I think -- well, I think we have kind of mixed opinions
about whether the comparison is even legitimate to the 55.  But I also
don't see the harm in keeping a ceiling on 55 given that's what was used
in the field and given we have no evidence that there are differences
that the data should be generalizable.  We're not orphaning the over 55
population.

		MR. PHILPOTT:  That's fine.  I just wanted to make sure in the
consensus statement and in the write-up then.  I completely -- I'm fine
with at least acknowledging that there may be a reason to keep a 55 cap.

		DR. FISHER:  Good.  Good.  So just to summarize then just so we're all
on the same page, the untreated arm baseline needs to be added to the
informed consent.  The purpose has to be changed.  It's not to bring a
new product.  It's for a labeling change.  How the pathogen is -- the
absence of pathogen is explained needs to be reframed, although there
also may be risks of heat or whatever it is but in terms of their other
kind of environmental minor risks that should be included.  A clear plan
of attack, not just having a cell phone but who you're calling and how
they're getting there even though we hope that will never happen but
better safe than sorry.  I think we're in agreement that for this
particular study pre-enrollment changes do not have to be part of the
informed consent as long as, I guess what I would suggest would be as
long as people who are given cab fare to go home or travel fare to go
home if they decide not to be in it once they get there, they're not
stuck when they're taking informed consent which are the typical
procedures.

		So other than that, I think we're done with this.  So before we take a
break, two things, okay.  So one is to confirm that the HSRB does not
want to see any protocols to which there is not a planned statistical
design incorporated into the protocols.  In addition, we don't -- to the
extent possible, when EPA has responded to a sponsor or an applicant
with respect to modifications that they've identified need to be made,
we don't want to see the protocol if, in fact, there is no information
about whether those changes are going to be made.  Simply the absence is
just not sufficient for us and I understand there's that seven-day
period, but I think what would be helpful is that -- and I know that
that's a real tight frame and is that in regulations, seven days?

		DR. LEWIS:  No, what -- this is Paul Lewis.  What we do is we ask
people, the public, to submit their comments within seven days and then
acknowledge that.  If it's given to us with less than seven days, it
will be difficult for the Board to review the materials and we note that
in terms of the timeframe in the case.

		DR. FISHER:  And like for this we're not picking this instant, but it
looks like they'd have at least three weeks before the seven days.  But
once again, there could be -- they could put in the letter we are
changing the informed consent but within seven days we don't have time
to do it, I mean, but we need to know that they've paid attention and
what they are going to comply with and what they feel may be problematic
because sometimes they may have a very good reason that they want to
present in their public comments about why they want to or don't want to
adhere.  But simply the absence of knowing puts us in a very awkward
position and as we move on, I don't think we want to be in that
position, understanding also that there are time constraints and that we
can be flexible but that this is what our priorities are and our
preferences.

		Then the next thing is so we'll take a 15 minute break.  Then we'll
come back and I want to alert everybody that I don't think, if this
lasts an hour that will be a lot.  It will probably be a half hour.

		I want to apologize.  The working group we had at NEPA.  We had hoped
for more extensive information that did not come in which was why we
kind of packed yesterday and we wanted to make sure people weren't
leaving at 5:00 p.m. today and we will try to be more receptive to
short-term changes in that so that we won't have something like
yesterday.  But we really appreciate your tolerance about that.  

		Okay.  So let's come back at around five to eleven. 

		(Whereupon, at 10:41 a.m., the above-entitled matter recessed and
reconvened at 10:57 a.m.)

		DR. FISHER:  Okay.  One thing I do want to mention.  The Board has
requested that at our next meeting when the work group who helps with
the agenda gets together we would like to have a period at the next
Board meeting with some statistical consultants in terms of, and maybe
KyungMann will certainly be on that work group, designs models that are
familiar with the constraints of some of these studies.  You can't run
them for 20 hours but how they can be analyzed and maybe we can begin to
be more informative and not just say this can't be analyzed but have
some models to which they can be.  So we'll seek KyungMann's help and
when we have our planning conference call we will add those and anybody
who thinks of good consultants let us know.

		But this is also for the insect repellents.  It's not just -- I think
the Ag issues are a much larger scale and I know that EPA we've had
those discussions and they're looking for consultants on that.  So this
is in particular on these kind of small sample sized studies that have
both human subjects constraints as well as hours constraints, too, so
that we can move positively toward the future rather than just saying it
can't be analyzed.

		Okay.  Bill.

		MR. JORDAN:  Thank you, Dr. Fisher.  This agenda item is the one that
I talked about, it seems like a week ago when I was reviewing the June
meeting and my purpose is to provide a brief report.  I think it will
take me five to ten minutes to summarize where we are with regard to the
sampling strategy issue in the handler exposure research area.

		When this was initially put on the agenda, both I and Dr. Fisher and
the work group hoped that we would be able to provide a fairly in-depth
discussion of this question and I am a little disappointed that I don't
have more to say.  What I will do is quickly review the context of this
issue and then provide you with a description of what we have done and
where things stand.  But I'm afraid that the in-depth discussion of this
issue is going to have to wait until the January meeting assuming that
the task forces do decide to go ahead with the research.  But I'm
getting a little bit ahead of myself.

		Let me say that since the June meeting we have been working on the
issues raised by the HSRB regarding sampling strategies, working closely
with the task forces and keeping Dr. Fisher and Paul Lewis informed
about those discussions.

		In June, EPA asked the Board to review extensive documentation
relating to research that was proposed by two industry groups, the
Agricultural Handlers Exposure Task Force and the Antimicrobial Exposure
Assessment Task Force.  The purpose of their research was to generate
data on the exposures received by people while carrying out very
specific tasks involving the mixing, loading and/or application of
pesticides and among other issues, among other elements of their
proposals, the task forces described a sampling strategy approach that
they called purposive diversity sampling.  PDS will be the shorthand
that I use for that.

		And the HSRB reviewed, I think, fairly unfavorably the PDS strategy,
stating that EPA couldn't draw any statistically defensible inferences
about the larger pesticide handler exposure population from data
collected from that kind of PDS sampling strategy and the Board
recommended a sampling strategy that used a random selection of test
subjects from the target population.  You said a lot more than that, but
that was sort of the kernel as we understood it.

		After the June meeting, we at EPA and the task forces talked about
that recommendation in considerable depth and those conversations have
taken place over a number of months and again I'll attempt to summarize.
 The task force has raised three main concerns.  The first is that they
continue to believe that PDS strategy will give reliable data because of
the manner in which the subjects are selected and they trust the
judgment of the expert selection of the conditions to monitor to be in
their view confident that they will skew the data where it is known
toward the conditions that are most likely associated with higher
exposure and where the influence of factors on exposure is unclear, they
will on purpose assure a selection across a range of factors.  So if
it's a geographical factor that's influencing exposure, they'll select
across a range of geographic locations.  The second -- so they argue in
effect that the data will be relevant and that EPA can make inferences
about exposure of the larger handler population from a dataset developed
through purposive diversity sampling.

		The second point that the task force raised in their conversations
with us had to do with timing and they said, in effect, if, this is more
true I would say for the agriculture handler exposure research than it
is for the antimicrobials, if the research is to be carried out during
calendar year 2008 the task force needs to know what EPA's position is
sufficiently in advance so that they can prepare their protocols, put
them through IRB review, put them through EPA review, put them through
HSRB review, make adjustments and then execute the protocols in the
field during the seasons when the activities occur and they developed
and showed to us calendars which basically led us to believe and we
accept this as being legitimate that they need to know essentially in
early November what they have to do in terms of a sampling strategy in
order to be able to get into the field to do the research next year.

		We are at EPA very interested in getting these data because we believe
the data resulting from this research will represent a significant
improvement over the quality of the information that we now have to use
in exposure assessments.  And so from our point of view, it was
important to try to push ahead as quickly as we could in order to try to
resolve the difference in view between the task force and the advice
that we got from the Board.

		The third concern, generally speaking, that the task forces raised
related to feasibility and cost of acting on the Board's recommendation
to move toward a random selection of test subjects, the task forces
argued with us or claimed to us that in order to have a database that
was randomly selected and would support the kind of statistical
inferences that they understood the Board was seeking would require
considerable revision of their sampling strategy and would cost a whole
lot more money and the task force folks have seen that the changes
they've made in the design of the overall research program in response
to feedback that they have heard from EPA and from the Board has led
them to project considerably increased expenses in terms of the overall
research program and they have asserted to us that they really can't
afford to spend much more money.  And so they are objecting to the idea
of doing random sampling on a cost basis saying that they would either
need to abandon doing some of the 33 scenarios that they project in the
Ag Handler or 15 scenarios or 18 scenarios, I forget which, in the
antimicrobial exposure task force and that if they move away from doing
scenarios, then they'll lose members and if they lose members, then they
don't have the funding and if they don't have the funding, then they
can't do the research.  So for them, the cost implications are a very
significant factor as well.

		So we in the Office of Pesticide Programs have attempted to work
through these issues.  We think they are very important and we wanted to
do the best job we could of understanding the implications, both
scientifically as well as financially and to that end, we asked the
Secretariat of the Scientific Advisory Panel, the group of EPA employees
whose job it is to find experts to serve on SAP panels to deal with
specific issues.  Some of you have been recruited to work with the SAP
and know how that process works.  They are very effective at locating
experts in particular scientific disciplines and getting them cleared
through a process that involves ethics reviews and hiring them to work
with EPA.  So we worked with the Secretariat to get someone to help us
deal with the statistical implications of various sampling strategies.

		I was rather discouraged about how hard it was.  I think a significant
factor was the timing.  We needed to get somebody in to work with us in
September and early October and we were only able to line up one expert
to work with us.  Fortunately, however, we were able to get Dr.
Tapabrata Maiti.  I'm sure I'm not pronouncing his name correctly.  Dr.
Maiti is an associate professor at Iowa State University in the
Department of Statistics, a colleague of Dr. Carriquiry with particular
expertise in survey sampling and we asked Dr. Maiti to review a huge set
of material.  I think he may have had to pay extra money to bring the
papers here to Washington.  We gave him the reports from the HSRB, from
the SAP, the documents prepared by the task forces as well as some other
background documents that we supplied from the EPA files.

		And he came to Washington, met with the staff from the Office of
Pesticide Programs.  We filled him in on what we thought the issues
were.  Then we invited members of the two task forces to join on us for
a better part of two days last week, October 17th and 18th, at which we
dug into these issues in considerable depth.  We alerted the task forces
to the issues that we wanted them to cover in their presentations and
they made a series of presentations that stretched over I think about
four hours covering our specific questions about how the PDS sampling
strategy would work, how particularly the cost issues would arise, what
were the feasibility questions that were of particular concern to them.

		I think it was a very useful conversation for us at EPA.  I hope that
the task forces felt the same way.  Certainly, Dr. Maiti came away with
a much greater appreciation of the circumstances that we're dealing with
in terms of doing this research and he was able throughout the meeting
during the sessions while the task forces were around and while we took
breaks and went off to our corner and the task forces did their
discussion in their corner to give us a lot of very valuable independent
assessments and insights into the implications of the task forces'
presentations.

		There were frankly some questions that we asked the task forces that
they couldn't answer, particularly regarding the detailed cost breakdown
of what it would take to do a random sampling strategy.  They basically
said when I listened to their discussions and Dr. Maiti's affirmed, I
think, that this is reasonable is that in order to figure out what it
costs you have to figure out exactly what you would do and to figure out
exactly what you would do would be in effect to do it or at least to do
a large part of it and so we were working based on the judgments of the
various sampling design strategists involved, both the people from the
task forces, Dr. Maiti's experience which is not in the area of
agricultural research.  He's done a lot of work in other fields and
sampling, that diversity of background helped informed us but he was
quite clear that he couldn't say for sure that it would cost 30 percent
or 70 percent more or you could get most of what you wanted with a 15
percent increase in cost.  But we did try to get as best we could our
arms around the issue of how much more it would cost.

		So that meeting was last week.  We at EPA got together and have talked
and we've identified some fairly specific questions that we will be
sending to Dr. Maiti and asking him to provide a written report which we
hope to get in about ten days and to review that report and make a
decision about what we're going to do.

		As I indicated, the task forces have to make a judgment about whether
it's going to be feasible to get into the field in 2008, particularly
Agricultural Handlers, and in order to do that, they need a decision in
early November.  So that means we have about two weeks left to sort
through this issue and make a decision and inform the task force of our
decision.

		Once we do that, the task forces then need to make a decision as to
whether or not they're going to go ahead with the research.  If they
decide to do the research, it's my understanding they will probably aim
to have a protocol ready in January and I would hope that we would see
protocol from both of the task forces, although I'm not absolutely sure
that will be feasible.  And we would obviously have made our decision
about what our position is at EPA regarding this issue and assuming that
the task forces are going ahead with the research, we'd be prepared to
present an explanation for why we decided what we did.

		I've talked to Dr. Fisher about the agenda for the January meeting and
she suggested and I think this is a very sensible idea of putting aside
a specific time to address this specific question even in advance of
looking at the protocols so that we would be able to tackle that
question discretely and have a full in-depth kind of discussion.

		So that's where we are with regard to this and as I said, I wish we
had more to say but that's as much as I feel we can report at this
point.

		DR. FISHER:  Thank you, Bill.  Just a little background, I had very
much wanted and I think we all did to be able to discuss these issues of
random sampling versus purpose of sampling independent of the protocols
coming to us because to be fair to all.  The working group had suggested
that we invite our own consultants so that we could have had that
discussion today.  Unfortunately, EPA thought that they would -- had
rejected that and thought that they would have their consultant and
therefore have some material today.  So I do want to express that I'm
disappointed, but I totally understand that there was nothing that Bill
could do about it.  But I am disappointed that we were discouraged from
inviting our own consultants because I think we end up in these
situations where what if we don't hear -- hopefully, everybody will be
on the same page and whatever the decision made in November is with
respect to purposes versus randomization.  It would be lovely if we're
on board with that. However, we don't know and there will no feedback to
the sponsor or to EPA about how we view that issue until January and we
may be getting the protocol at that time.

		So I'm disappointed from a process view because I just think we could
have been more help and also we don't want to put sponsors into that
situation to which there's so much uncertainty about how the Board would
respond to different procedures and I just don't think there's anything
we can do about that.  I assume that the Ag Handler is seasonal so that
if they were going to do it it would have to come in January.

		MR. JORDAN:  There are different activities that are covered by the
different scenarios, at least some of which need to happen early in the
summer and in the spring and to execute those protocols, they would need
to have been reviewed in January.  There may well be some research
that's done in the latter part of the growing season in July or August
that could be reviewed at the April meeting.

		DR. FISHER:  I wish I could be creative and think of what to do during
the interim after you've made your report, although I'm not sure.  I
don't know if there are any options of having a working group that
reviews the recommendation.  I mean I really don't know what the options
are other than having this potential tension.  Hopefully, there's no
tension, but there is a potential and I'm sad about that.  Yes?

		DR. LEBOWITZ:  Could a working group of ours look at the
deliberations, the statements by your statistician and then the
deliberations that occurred at the meeting last week and do that in the
next month or so prior to the holidays so that we have our own as it
were preview of the discussion that might occur in January.

		DR. FISHER:  Let me extend that.  The model I was thinking and I just
don't know regulatorily if we're allowed to do it. I was thinking about
the CBI model where we have a work group that reviews the material to
see if it's adequate for us to make the decision and I don't know if
that's feasible and helpful.  But given you're going to have this
information and EPA is going to make a decision by November and we know
the people who are expert on our Board in terms of the statistics and
the agricultural and whether or not, I just don't know if we're allowed
to do that because that would really be helpful to have a preview of
where -- it just might be helpful and not create a situation that we
don't want.

		DR. LEWIS:  Let me just make one comment and I'll let Mr. Jordan
continue the discussion.  In terms of process, if you have a work group
that looks at the materials, provides some commentary, again that
recommendation is from the work group and not from the Board.  So I just
want to make sure everyone understands that.

		DR. FISHER:  Right.

		DR. LEWIS:  So the recommendation that comes to the work group can't
be representative and say this is what the Board thinks.  It has to go
from the work group through the Board.  I just want to put that in
context.

		DR. FISHER:  First, if it's possible, then we have to see if the Board
thinks it's a good idea.  So, Bill, is this possible and is it of help
to you if it was possible?

		MR. JORDAN:  Here's what I think is certainly possible.  At the
meeting on October 17th and 18th we received at EPA materials from the
task forces and I see no reason why we can't give those materials to you
and so if you want them we could do that.  In addition, as I indicated,
we expect a written report from Dr. Maiti and I see no reason why we
shouldn't or couldn't give that to you as well.  So with that material
what I anticipate is that we will have discussions inside of EPA and it
will likely be that those discussions are oral as opposed to a written
memorandum with an explanation of all the thinking that we've expressed
in our conversations with our senior management that lead us to a
decision.

		What I was anticipating is that between when we make our decision in
November and are getting ready for the January meeting we will translate
those oral discussions into written or PowerPoint presentations that
would explain to the Board where we have come out on those issues.  I'm
not sure that we will have time between now and in November to commit
that to paper beyond the kinds of documents that I've described.

		DR. FISHER:  Let me -- my concern is once again it's not my area of
expertise, but obviously there is this kind of tension between the very
real financial issues that come into play with randomized sampling
supposedly, although it would have been nice, I think, some people had
recommended there are people who know how to do it somewhat less
expensively.  But I don't really know if that's true and the kind of
statistical issues that randomization might be somewhat better.  And I
think what -- I guess my understanding is that once you have made these
oral decisions, the Ag Handler Group is going to be acting on them and
that we would potentially see something in January.

		So I understand you're always in a time crunch and we are so sensitive
to that, but it might be helpful, at least it would relieve me, if maybe
we could have a work group that would see the materials that you're
seeing, Maiti, is that his name, Dr. Maiti, and the other materials that
you've collected to give an overview of any major concerns or lack
thereof that would derive from this composite.  So it just is additional
information for you in terms of how the Board might be looking at this
and for the Ag Handlers as we move forward, if that's appropriate,
because we all want a success and we don't want -- if possible and we
also don't want undue work if it's not going to go anywhere and I'm just
trying to -- within the incredible time constraints that your office and
everyone else is always under.

		So what does the Board think about that idea?  Is that a feasible
idea?  Is it something that those who are in the know would -- or is it
just my tensions?  Rich.  Rich, of course.  Rich and Dallas and I don't
know if KyungMann, if his expertise goes to this.  Mike.  So I mean
definitely Rich and Dallas I know are interested in Ag Handler things. 
So I'm volunteering you and KyungMann and Michael we can talk to if you
feel your expertise is in there.

		DR. SHARP:  Which Rich?

		DR. FISHER:  I'm sorry.  Rich Fenske.  Yes, that's what I was
thinking.

		DR. SHARP:  Alicia.

		DR. FISHER:  And Alicia if she's around.  Absolutely.  I think she's
in Chile.  So I'm not sure.

		DR. KIM:  One question I have is have we seen the SAP's comment on the
issues.

		MR. JORDAN:  Yes, you have.

		DR. KIM:  We have.  Because SAP has a member who expertise is in
survey sampling who is Dr. Steve Heeringa.

		DR. LEWIS:  Dr. Heeringa who presented in May if you may recall.

		DR. FISHER:  So we might -- is he allowed to be a consultant to the
work group?  Well, first, let me just -- is this a good idea or is it
not because I have a particular view as chair about how I would like the
thing to go but that doesn't mean the Board is on -- yes, Sue.

		DR. FISH:  I think this is a great idea.  I think one of the
frustrations I'm sure for your office, Bill, and one of the frustrations
for me is that people work really hard to bring their protocols here. 
You guys work really hard to prepare for it and then you come here and
if we have a different opinion, that sends the whole thing back to
square one.  If there's a way to get our concerns, if there's a
regulatorily acceptable way to get our concerns out on the table during
the development process, which is what's going on, it seems it would
smooth the process without compromising our role.  So I think the
earlier that you and the Ag Handlers get input from our experts in
whatever it is, whether it's the actual science or the design and
statistical issues, or whatever that the smoother this process is and
less confrontational, more collaborative and less confrontational and
still meeting our charge of the HSRB.

		DR. KIM:  I second that suggestion and support it.  I think it's a
great idea.  But it has its purview under the regulatory constraint.

		DR. BRIMIJOIN:  I just want to express pessimism.  I think we're
actually struggling to find some positive way of responding to this
very, very difficult situation and I think the Board should be commended
for making the effort and I guess I think any effort to save a major
project, anything that could be done now to save a really important
project is warranted.

		But I am pessimistic.  Unless a group of statisticians can come up
with a brilliant way to do random sampling that's easy to implement and
going to be essentially the same cost or cheaper which seems to me
extremely unlikely, I think at the end of the day we're going to be
faced with a group who is limited, but I have some sympathy for them,
who feels that we're faced with doing something that will delay our
progress substantially in time and cost substantially more and we just
won't do it.  And if on the other hand, neither the statisticians nor
the Board are prepared to recommend any use of data collected by a
nonrandom method, we will have, in fact, a total catastrophe.  But I
fear that our choice is going to be to accept purposive sampling or see
this not happen.

		DR. FISHER:  This was one of the reasons that I had wanted to have
consultants here to see whether or not there's a third path of analysis.
 I think once again it's not my area of expertise, but I think the kinds
of things we've been discussing over the last couple of days with
respect to the data will be good for this, but is limited for that, I
think, I agree with you.  We've all agreed with the EPA about the
necessity of a new dataset.  There is no doubt in anybody's mind.

		At the same time, we don't want a dataset that's not useable.  It's a
lot of expense for everybody involved, a lot of time, a lot of effort, a
lot of commitment and I'm hoping it's not going to be an all or none. 
So it seems to me that even saying where the weaknesses might lie in
terms of planning, even a heads-up on that, both in terms of what we're
prepared to know and a heads-up on where we find the weaknesses could be
very beneficial in terms of everybody's expectancies when we get to the
meeting.

		To have it adversarial is just really not what we're all here for. 
We're all here for it to help promote good science and ethics and the
kind of data that EPA wants.  So this is just one strategy that might
help that and I think we're hoping that your scenario doesn't take place
and I don't know anything else.

		DR. BRIMIJOIN:  Me too.

		DR. FISHER:  I know and I don't know anything that can be done.  Who? 
Which one?  Dallas.

		DR. JOHNSON:  I was just going to say that maybe rather than make a
decision today as to whether this small group will get together, working
group or whatever it is, it would be good to get the reports that Bill
talked about from the consultant at Iowa State, Maiti, and then once we
have those in hand, then we could get some idea as to whether it would
be worth having such a meeting or not.

		DR. FISHER:  That's certainly --

		DR. JOHNSON:  And could do it on the basis of that.

		I also might note that when it came to this idea of purposely versus
random sampling that Alicia -- I was a consultant at that meeting and
Alicia and I didn't quite agree on that issue.  And I think Dr. Kim was
more in line with Alicia than perhaps with me.  But I also served on the
SAP that looked at that study and there were three or four statisticians
there and I'm not sure we all agreed either.

		So I don't know whether you can get a consensus of statisticians and
it would be interesting to see what Dr. Maiti says because if he says it
has to be random sampling, then I don't think we have any place to go.

		DR. FISHER:  That's an interesting point.

		DR. JOHNSON:  If there might be something else that could be done that
would be acceptable, then we have something to talk about.

		DR. FISHER:  And in terms of what the data means.

		DR. JOHNSON:  Right.

		DR. FISHER:  That there might be some meanings that could be attached
to it and  others that it's limited for.  Mike.

		DR. LEBOWITZ:  Yes.  I think this is a good point and I think Dallas'
suggestion is good, that the working group can tell from what they
receive whether it's deserved of follow-up and to what extent or whether
it's a dead issue.

		But in my mind if statisticians disagree, then that's fine also
because it generates hopefully some kind of consensus even to the extent
of saying we don't know which in fact would not be detrimental to the
process and for AHETF or EPA and would allow it to go forward with
certain suggestions that might make data more useful once they were
collected.  So even to that extent, I think it would be worthwhile.

		DR. FISHER:  The critical issue, one of the critical issues for us is
not just the work put in for the design, but it has to come back -- it
cannot go forward until it comes back to us.  And so we also want to
work preventively not to have people put in all this money into a
project to which we say most of it is not useable.  So there's a -- it's
not just the protocol that we're looking at and saying we don't know. 
We're also trying in some sense to say if you do this, we will at it as
this is the range of useful data we will look at it.  I don't think we
want to wait either to see how it turns out.  There's something of in
some sense an agreement among people if we reach it that when it came to
us if done as suggested we would be looking at the data as valid for the
area that we said it would be valid for.

		DR. LEBOWITZ:  Yes.  In this regard, you see, I think the discussion
we had on the first day of the NERL document, and it was very useful in
the fact that we heard clearly that a lot of their studies,
observational studies, are small and in fact do not use randomized
populations and how they're used for risk assessment and so forth.  And
I think that's useful to have because even now given the language would
not have occurred but for and including for the subjects' participation
in the study actually indicates to me there's a fair amount of overlap
including between the NERL funded studies of farm workers, etc., that
may or may not have been random.  So there are things that we know we
can learn from that and what they've done in their smaller studies of
the size we're talking about for each scenario.  That's very useful and
we can for our purposes blur the distinction between observational and
intentional vis à vis what's been done before by other parts of the
agency, what's going to be done and how the data are used.  So I think
looking at it that way also would be of benefit and then Fenske and I
can contribute to that from that side of it if not from the statistics.

		DR. FISHER:  So what I would suggest then is that Fenske and Mike and
Dallas and probably Kim if Alicia can't be on it, but we're getting
different viewpoints, you will read the materials.  We'll get the
materials.  We'll have a work group conference call with me and Paul and
Bill to say whether we go forward or not in terms of our own work or
just whether or not it's a clear whatever because this feedback is
important for the Ag Handlers.  It really is.  They don't want to come
here for no reason, you know, after all this hard work in January to
have an a priori negative result.  That's just not fair and that's my
goal.

		And once again, this is not a decision of the Board.  This is merely a
kind of a priori looking at it, making an estimate of what we think the
issues the Board would probably identify, positive and negative, in
terms of what we get.  Okay?

		So that will be sometime in December I guess because we'll get it
after the November whatever date that Bill has, and Bill, of course, is
always so helpful and informative on the calls.  So we welcome that and
anybody else you want to be on the call.  If you want Dr. Maiti on the
call, that would be great too.

		Okay.  So is there anything else?  Paul or Bill.  I'm sorry, Bill.  Is
there anything else?

		MR. JORDAN:  Nothing else.

		DR. FISHER:  Okay.

		DR. LEWIS:  Just a few comments.  Again, I would like to thank members
of the Board and our new members for all the hard work and Dr. Fisher
serving as chair and also members of the public that have observed the
operation deliberations of the Board the past several days.

		Just a couple of announcements about future Board activities.  On
November 13th, we're having a teleconference from 1:00 p.m. to 3:00 p.m.
Eastern Time.  There's a Federal Register notice that published actually
this week announcing this teleconference for review of the Board's
report from its June 2007 meeting.

		And in terms of the Board's next face-to-face meeting that is
tentatively scheduled for January 15 to 18, 2008 in the Potomac Yard
facility.  As we've done in the past, we schedule four days and we will
let the Board and the public know in terms of whether we'll need all
four days or a shortened amount of time and we'll be announcing in the
Federal Register more specifics about the meeting in terms of topics and
public opportunity to make some comments.  Thank you.

		DR. FISHER:  Thank you everybody.  It's been a pleasure as usual.  

		(Whereupon, at 11:41 a.m., the above-entitled matter was concluded.)

 

 

	NEAL R. GROSS

	COURT REPORTERS AND TRANSCRIBERS

	1323 RHODE ISLAND AVE., N.W.

(202) 234-4433	WASHINGTON, D.C.  20005-3701	www.nealrgross.com

	 page \* arabic 1 

