January 30, 2009

Answers to EPA HSRB Questions on Spatial Repellents

Provided by Daniel Strickman, Ph.D.

USDA Agricultural Research Service

National Program Leader, Veterinary, Medical, and Urban Entomology

  HYPERLINK "mailto:Daniel.Strickman@ars.usda.gov" 
Daniel.Strickman@ars.usda.gov 

These answers are given from the professional estimation of the author
and do not intend to represent the position or policy of USDA.

Environmental Issues

 

1. What are the environmental (temperature, wind, time of day, humidity,
proximity to water/plants, size and type of space) and human factors
(height/weight; gender, age, ethnicity, density of humans in space) that
can affect insect behavior and repellent efficacy relevant to space
treatment studies?  

Environmental factors have a strong effect on avidity of biting
arthropods. The effects will depend greatly on the species of arthropod.
For the best known arthropods, the parameters of maximum avidity are
known and could be replicated in the lab or sought in the field.  Also,
use of controls will assure adequate avidity has been achieved.  One
environmental factor that is particularly important to the function of
most spatial repellents is wind.  Even light air movement will affect
performance of products that are based on dispersal of a chemical in a
large volume of air.

Humans vary a great deal in attractiveness to biting arthropods. This
variation is affected by gender and diet, but these relationships are
not well understood.  Conservatively, the amount of variation is
four-fold (attracts four time more mosquitoes) and translates to
approximately 25% decrease in duration of DEET topical repellent.
Presumably, this could affect performance of behavior-altering active
ingredients but not that of lethal active ingredients.

2. What factors need to be considered for test spaces with respect to
size of area in which the test is conducted? How is the most appropriate
test area determined? 

The test area should exceed the label claim for area protected. An ideal
test would evaluate the area protected by measuring the occurrence of
pests across a transect.

3. Does the number of human subjects within testing environments of
different sizes affect insect activity? Does the number of subjects in a
given area affect product efficacy or the measurement of product
efficacy?

In general, results from field tests are unaffected by the number of
subjects as long as the density of pests is sufficient. Often, the local
population of pests is so high that additional traps or subjects do not
affect results from each device or person. In the laboratory or large
cage trials, where populations are limited and controlled, additional
traps or human bait subjects might reduce the number collected by each
trap or person. 

4. Are there any other special considerations regarding insect behavior
in such studies that require inclusion in protocols?

If the product affects a specific behavior or a particular odorant
receptor, then the behavioral status of the pest may affect performance.
 For example, if a chemical blocks a carbon dioxide receptor but the
pest is not seeking a blood meal, then the product may appear to fail
(e.g., pest goes through the repellent to a light trap) even though it
would work well for its purpose (to prevent bites).

Statistics and Study Design 

1. How is the location of open spaces typically selected? How many
different or similar types of sites are appropriate to assess
generalizability?

Locations are usually selected for the species and abundance of the
pest. There is a danger that sites with low numbers of pests will be
selected intentionally to make the product look more effective. In
practice, there is probably a greater tendency to choose sites with a
large number of pests in order to reduce the number of trials required
in order to attain statistical significance.

The location should also be typical of the intended use site. For
example, a mosquito coil or candle would be appropriately tested on an
actual patio representing typical use site. If the product is registered
for outdoor use only, it should be tested outdoors in a field or
semi-field (large cage and/or intentional release of pests) situation.

The number of different sites is not as important as the number of kinds
of pests in order to achieve generalizable results. These kinds of
products are also very sensitive to wind, which might influence the
number of trials required and selection of weather conditions.

What are common spatial dispensing devices? How are they related to the
nature of the product dispensed (e.g., gas, suspended liquid, smoke)?
What are design or measurement challenges for different dispensing
devices and products?

Current products registered in the US include candles, mosquito coils,
heated paper mat devices, reservoirs of active ingredient with a fan
above them, granular products impregnated with active ingredients,
liquid sprays, and no-pest strips.  All attempt to suspend an active
ingredient in the air, creating a gaseous or fine particle barrier
between the human and the pests, or to create a bubble of active
ingredient around the human. Only the no-pest strip is currently
registered for indoor use. The product is designed to distribute the
active ingredient at an even rate over a defined period of time. Heated
devices probably deliver the most constant rate, since they are
independent of ambient temperature. Passive systems like granules,
sprays, and no-pest strips would release more active ingredient during
higher temperatures and when there is more wind. These factors will make
evaluation of duration more difficult.

3. What type of dosimetry data is required to determine amount of
product application used in testing? How is discharge time determined?
What are the relative design merits of the experimenter or subject
discharging the repellent? 

For the most part, the rate of release from spatial repellent products
is not adjustable. Where the amount applied can be adjusted or the
number of devices used simultaneously can be changed, the tests should
follow proposed label rates. Discharge time is not a factor for
currently designed devices. Installation and discharge by the
experimenter may have a small advantage over that by the subject in
order to reduce trial to trial variation.

4. How are outcomes measured in these studies? How are insect knockdown
and mortality effects measured? Are both knockdown and landings/bites
usually measured in the same study? What is the difference in knockdowns
vs. bites in terms of information regarding product
efficacy/effectiveness? 

 There is no standard method of measuring outcome. Mosquito coils were
most commonly evaluated by measuring knockdown in cages, but this is
probably inappropriate. The purpose of spatial repellents is to
interrupt the pest on its way toward a host and that is what should be
measured. A test of movement toward a host or attractant should be
sufficient, eliminating the need to expose subjects to biting. An
example would be current trials at USDA ARS Gainesville testing efficacy
of proprietary devices inside a tent; about 18 small traps inside the
tent to detect distribution under treated and untreated conditions; each
trial is conducted with tent inside large cage with known number of
mosquitoes.

5. What is the difference with respect to measurement in assessing
efficacy of the active ingredient and effectiveness of the formulation?

The purpose of these tests is to compare or document products rather
than active ingredients. What is more, formulation is likely to have a
large influence on effectiveness. Therefore, tests of non-formulated
active ingredients would be largely irrelevant.

Sample Size and Statistics

1. Depending on the outcome measure, what are best practices with
respect to human sample size? What is the sample size norm in the field?
How is determination of sample size related to square feet of test area?
What is the best way to determine power for these studies? 

 Psychodidae) and Ochlerotatus caspius (Diptera: Culicidae) in
Sanlıurfa Province, Turkey, J. Med. Entomol., 40, 930, 2003.

R. A. Wirtz, J. D. Turrentine, Jr., and L. C. Rutledge, Mosquito area
repellents: Laboratory testing of candidate materials against Aedes
aegypti, Mosq. News, 40, 432, 1980.

R. A. Wirtz, J. D. Turrentine, Jr., and R. C. Fox, Area repellents for
mosquitoes (Diptera: Culicidae): identification of the active
ingredients in a petroleum oil fraction, J. Med. Entomol., 18, 126,
1981.

2. What are best practices with respect to statistical analysis?  How is
censored data handled?  

I am not aware of any best practices. Please refer to the book chapter.

3. What are the pros and cons of various endpoints (e.g. ending the
study after a set number of hours, waiting until the first landing/bite,
other) to assess product efficacy (e.g. to meet assumptions for
appropriate statistical analyses)?

Since human exposure to bites is not necessary, there is no reason not
to test products to failure in order to develop a complete chronological
curve of effectiveness. Many of these products do not achieve complete
protection, implying that the best measure is a profile of percentage
repellency over the period of time to complete failure.  

Human Subjects 

1. Why are human subjects necessary for such studies if the outcome
measures are knockdowns or mortality?

Human subjects are not necessary unless the product is designed to
disrupt particular insect behaviors. If humans are needed as bait, they
do not need to be exposed to any bites. Knockdown and mortality are not
the best measures of effectiveness. Even with a lethal ingredient like
transfluthrin or allethrin, a certain number of mosquitoes at low
dosages will probably be stopped from approaching a host or trap even
though they are not killed (true repellency); therefore measuring only
mortality underestimates benefit.

2. What are the potential risks to treated subjects (e.g. inhalation,
dermal effects)?  What are exclusion criteria in subject selection to
avoid such risks?   How is the degree of risk related to dosage,
ingredient, formulations, and aerosol pressure? 

The risk of the product itself to the subject should be minimal and
acceptable based on toxicological evaluations by EPA. Very basically,
safety is determined by studies of exposure and risk. Risk is determined
by a large battery of acute and chronic tox tests required by EPA. EPA
makes a judgment whether the intended use exposes the public to a
significant risk, with appropriate safety factors. These evaluations are
the foundation of safe use of pesticides in the US.

3. What is the methodological rationale for continuous versus
intermittent exposure? How does human risk differ for these types of
exposures?  Will exposure start at the beginning of the test period
immediately after release of the product?

Where humans are used as bait, there should be no additional risk from
continuous exposure. The most accurate results would be obtained from
studies with continuous exposure because the products are designed for
continuous use during the protection time claimed. These studies would
require continuous or interval counts of pests as they fly into a screen
trap surrounding the human bait. These approaches do not work for
topical repellents because the distances over which they work are so
small; whereas, spatial repellents are intended to work over a large
area.

4. If the test agent has properties to repel or destroy an insect, what
is the relationship (if any) to a related mechanism of action to humans?

Ù

Ú

7

8

 h…

 

Page   PAGE  1  of   NUMPAGES  5 

