Clinical Toxicology (2008) 46, 93100

Copyright Informa Healthcare USA, Inc.

ISSN: 1556-3650 print / 1556-9519 online
DOI: 10.1080/15563650701664905

REVIEW
LCLT

Animals as sentinels of chemical terrorism agents:

An evidence-based review
PETER RABINOWITZ, M.D., MPH1, JAMES WILEY, M.D., MPH1,2, LYNDA ODOFIN, DVM, MSPH1,
MATTHEW WILCOX, M.S.1, and F. JOSHUA DEIN, VMD, M.S.3
Animals as sentinels of chemical terrorism

Yale University School of Medicine, New Haven, Connecticut, USA

University of Connecticut School of Medicine, Farmington, Connecticut, USA
3
USGS National Wildlife Health Center, Madison, Wisconsin, USA
2

Objectives. The goal of this systematic review was to identify evidence that animals could serve as sentinels of an attack with a chemical
terrorism agent. Methods. The biomedical literature was systematically searched for evidence that wild or domestic animals exposed to
certain chemical weapons of terrorism had either greater susceptibility, shorter latency period, or increased exposure risk versus humans.
Additionally, we searched for documented reports of such animals historically serving as sentinels for chemical warfare
agents. Results. For a small number of agents, there was limited evidence that domestic and/or wild animals could provide sentinel
information to humans following an airborne attack with chemical agents, usually related to increased potential for environmental
exposure. Some of this evidence was based on anecdotal case reports, and in many cases high quality chemical terrorism agent evidence
regarding comparative susceptibility, exposure, and latency between humans and sentinel animal species was not
found. Conclusion. Currently, there is insufficient evidence for routine use of animals as sentinels for airborne chemical warfare agents.
At the same time, Poison Center surveillance systems should include animal calls, and greater communication between veterinarians and
physicians could help with preparedness for a chemical terrorism attack. Further analysis of comparative chemical warfare agent toxicity
between sentinel animal species and humans is needed.
Keywords Animal sentinels; Chemical warfare agents; Evidence-based medicine; Comparative medicine; Nerve agents; Terrorism

Background
As part of preparedness for a possible terrorist attack using
biological or chemical weapons, the CDC Strategic Planning
Working Group (1) has called for prompt diagnosis of
unusual or suspicious health problems in animals. The CDC
report recommended establishing criteria for investigating
and evaluating suspicious clusters of human and animal disease or injury and triggers for notifying law enforcement of
suspected acts of chemical terrorism.
It is conceivable that domestic or wild animals living in
proximity to human populations would also be sensitive to
many of the agents that are potential chemical weapons and
could therefore serve as sentinels for a chemical terrorism
attack, much as coal miners in the UK and the US used canaries to provide early warning of deadly mine gases (2). The
US population lives in often intimate contact with pets and
other animals; according to American Veterinary Medical

Received 4 May 2006; accepted 2 July 2007.

Address correspondence to Peter Rabinowitz, M.D., MPH, 135
College
Street,
New
Haven,
CT
06510,
USA.
E-mail: peter.rabinowitz@yale.edu

Association estimates, more than half of American households

own at least one pet, while the population of dogs and cats has
increased by 16% between 19962001 to 62 million and 68
million, respectively (3). Additionally, human-wildlife interaction is increasing (4) and livestock are potential targets of
attack with secondary ramifications for human health.
Historically, it has been noted that during chemical warfare
attacks, animals near the battle lines may also be affected. A
newspaper account during World War I provides vivid details
of the effects on both animals and humans of an unspecified
type of gas attack (5):
Among other features, the effect of gas attacks upon animals
has been carefully noted, the Germans also having studiously
investigated the same subject. Results show that horses suffer
much from the noxious fumes, and are subsequently thrown
into a state of nervous terror on again scenting them. Mules
are more inclined to stand their ground, and appear as if trying not to breathe. Gas helmets of a kind have been successfully tried for both these animals. In the trenches are many
animals kept by the soldiers as pets. Of these, cats quickly
scent the gas, and run about howling. Guinea-pigs are the
first to succumbRats and mice emerge from their holes, and
are found dead in quantities, which as the soldiers say, is the
only advantage of a gas attack by the enemy.

94

P. Rabinowitz et al.

While in hindsight it is not surprising that animals so close

to the site of an attack could show symptoms, the writer of
this account took note of the fact that some animals more
than others appeared sufficiently prone to the effects of the
gas that they could provide warning to humans.

fore searched the medical literature for scientific evidence

supporting the occurrence of these phenomena in wildlife or
domestic (companion or livestock) animals.

Methods
. . . Poultry of all kinds are useful for giving warning, ducks
and fowl becoming agitated 10 minutes or so before the
oncoming gas clouds. Many kinds of wild birds are greatly
excited, and the usually unruffled owl becomes, as it were, half
demented. Only the sparrow seems to disregard the poisonous
vapor, and sparrows chirp on where horses are asphyxiated,
and bees, butterflies, caterpillars, ants, and beetles die off in
great numbers. The gas at once kills snakes, and earthworms
are found dead in their holes many inches below the ground.

During World War II, crates of rabbits were placed on the

cargo deck of the ships transporting nerve gases, and crewmen were instructed to watch for sudden animal mortality
that could signal a gas release (6).
Even today, despite the development of sophisticated biosensor technology to detect minute concentrations of chemical warfare agents in the environment (7), the use of animals
as sentinels of chemical warfare hazards continues to be
explored. Two modern examples come from the aftermath of
the Tokyo sarin attack and the recent United States invasion
of Iraq. During raids on the Aum Shinrikyo cult compounds,
Japanese policemen carried canaries in cages with them to
serve as a warning of poison gas release (8). During the initial
invasion of Iraq in 2003, the US militarys plan to use chickens as sentinels for nerve gas agents received wide publicity
(9). The idea was that if caged chickens remained alive following a warning alert of a possible gas attack, it would be
safe for soldiers to remove their gas masks. The hope was
also that the use of live animals would avoid possible false
alarms that had been attributed to ion-mobility spectrometry
biosensors. Yet several weeks after this announcement, such
plans were suspended, due to concerns that chickens in fact
were not likely to show effects of a nerve agent before nearby
humans (10,11). The EPA, through its Safe Building Program, has stated its intention to consider evidence regarding
animals as sentinels of chemical threats (12).
These anecdotal reports do not provide convincing evidence for the use of animals as sentinels of airborne chemical
exposures. At the same time, they emphasize the need for
properly designed studies or adequately described incidents
that would support the use of animals as useful sentinels of
specific chemical exposures.
It has been proposed that animals could act as sentinels if
signs of toxicity in the affected animals would be easily recognizable either before the emergence of human illness, or
sufficiently contemporaneous to aid in the timely recognition
of human illness. This could happen for any of four reasons:
either the animals could have greater susceptibility to a particular toxin relative to humans, a shorter latency time from
exposure to onset of signs of apparent toxicity, and an
increased level or different routes of exposure (13). We there-

Selection of chemical agents

The CDC publication entitled Biological and Chemical Terrorism: Strategic Plan for Preparedness and Response (1)
contains a list of chemical agents that could potentially be
used in a terrorist attack. Chemical agents were included in
this list because of their potential to cause major morbidity or
mortality, as well as social panic and disruption.
We focused for the purposes of this review on four different
classes of chemical agents that have been and could again be
used in a terrorist attack, according to a classification scheme
developed by Kales and Christiani (14). These include cholinesterase inhibitors (e.g., organophosphorus nerve agents),
blistering agents or vesicants (e.g., sulfur mustard), asphyxiants (e.g., cyanide), and respiratory tract irritants (e.g., chlorine). These agents are included in the MEDLINE Medical
Subject Heading (MesH) term chemical warfare agents.
Search strategy
We searched the bibliographic databases MEDLINE, CAB,
and AGRICOLA as well as medical and veterinary textbooks
and subject reviews for reports of animal exposure to these
chemical agents. Our search methodology involved crossing
the term chemical warfare agents with terms for domestic
animals, companion animals, wild animals, and zoo animals.
Table 1 shows the results of the MEDLINE search.
Selection of studies
From the studies identified as above, we selected those which
provided at least one of the following types of evidence
regarding sentinel potential, based on a scenario of a widespread, airborne release of a chemical warfare agent.

Table 1. MEDLINE search strategy for studies of animals serving

as sentinels for chemical terrorism agents
Search terms
1

2
3

exp birds/ or exp Sheep/ or exp cattle/ or

exp swine/ or exp goats/ or exp horses/ or
exp cats/ or exp dogs/ or exp raccoons/ or
exp rabbits/ or exp sciuridae/ or exp
mephitidae/ or exp fishes/ or exp animals,
domestic/ or exp animals, wild/ or exp
animals, zoo/
exp Chemical Warfare Agents/
1 and 2

# Citations
1269923

17129
1892

95

Animals as sentinels of chemical terrorism

1. Studies providing data on the susceptibility of a potential
sentinel animal species to a chemical warfare agent, in terms
of a clinical response at a certain dose level (LD50, etc).
2. Studies reporting on latency period (time from exposure to
onset of adverse effect) for particular agents in sentinel
species.
3. Studies providing evidence of relative exposure risk to
chemical agents for sentinel animals versus humans.
4. Reports of actual exposures to both naturally occurring
animal and human populations from particular chemical
warfare agents that potentially provided evidence for animals giving early warning to human populations.
Of the 1892 studies identified in the MEDLINE search, 38
met at least one of the above criteria. To identify additional
studies, we searched bibliographies of these 38 papers, as
well as review articles on chemical terrorism, and also
searched the CAB and Agricola for the terms chemical terrorism. This process identified an additional four sources.
In situations where the dose-response estimates for humans
were based on animal studies (due to a lack of human data), it
was not possible to reach conclusions about interspecies susceptibility differences between animals and humans (15). In a few
cases, however, human susceptibility data not based on animal
studies were available. To explore whether animals could be at
increased risk of exposure for a particular toxin, we looked for
evidence of differential physiology affecting exposure routes
such as skin and respiratory, as well as time spent in particular
environments. In order to identify reports of animals providing
early warning, we searched for historical reports of actual exposure events that provided data regarding the effects of the chemical on animals as well as human populations.
Classification of studies and grading of evidence
Articles located were classified according to species, study
methodology, evidence regarding susceptibility, latency,
exposure, and early warning, and the classification data were
entered into an online database of animals as sentinels of
human environmental health hazards (16). For each evidence
category (susceptibility, latency, exposure risk, and early
warning) we reviewed studies appearing in the database, and
then graded the strength of evidence for each category on a
three-point scale based upon published classification schemes
(17,18) that grade evidence in the medical literature depending on the rigor and design of individual studies. Table 2
displays the grading scheme used in this review.

Results
Table 3 displays the results of the search and analysis of evidence related to animals as sentinels of chemical agents. Only
six articles were found that supported animals as possible
sentinels of an airborne chemical terrorist attack.

Table 2. Grading scheme for levels of evidence

Level of evidence*
1
2
3

Type of study
Experimental studies, cohort studies
Case-control studies, ecologic
(aggregate) and cross-sectional surveys
Professional consensus statements,
textbooks, and descriptive case reports

Nerve agents
A number of organophosphorous compounds are listed as
potential nerve agents, including the G series compounds
tabun (GA), sarin (GB), soman (GD), and cyclosarin (GF), as
well as the V series, Ve, Vg, Vm, and Vx. These compounds inhibit cholinesterase and can produce acute effects in
animals and man including rhinorrhea, miosis, difficulty
breathing, excessive sweating, drooling, nausea, vomiting,
cramps, twitching, confusion, seizures, loss of consciousness
and respiratory arrest. Susceptibility to nerve agents varies
widely between species. Rats, for example, may be less susceptible than humans, since they possess blood enzymes (aliesterases) that can bind to and reduce the toxicity of certain
nerve agents such as GA (24). These aliesterases are not
present in humans. Furthermore, carboxylesterase activity in
other species, such as guinea pigs and rabbits, may confer
protection from Soman, relative to humans (25). At the same
time, there are interspecies differences in RBC cholinesterase
activity with many species, including pigs, sheep, dogs, rabbits, and cats having less activity than humans (15,26). Table
4 shows these differences.
Lower RBC acetylcholinesterase activity in humans has
been associated with increased susceptibility to nerve agents
(27). Lower activity within a given species may indicate
greater susceptibility to nerve agent exposure if that species
does not have an alternative system to combat cholinesterase
inhibition by nerve agents. According to Osweiler (1985),
50% of the total blood ChE in humans exists in plasma (30),
while Wills (1971) has reported that in dogs, plasma ChE
comprises about 40%, and in sheep horses and cows only
10% (19). Therefore, a number of biological differences
could leave animals either at greater or lesser susceptibility to
nerve agents relative to humans.
While precise knowledge of the human lethality for a number of nerve agents is not available, the LCt50 (50% lethal
concentration via inhalation route) for military personnel has
been estimated from animal studies. The LCt50 for GB has
been estimated at 35 mg.min/m3, while estimates of LCt50 for
other compounds are 70 mg.min/m3 for GA, 35 mg.min/m3 for
GD, and 35 mg.min/m3 for GF and 15 mg.min/m3 for VX (15)
The measured lethal concentrations of these agents in some
animals for similar times of exposure are mostly based on
studies conducted between 1940 and 1980. For GA, there are
LCt50 values of 320 mg.min/m3 for dogs, 960 mg.min/m3 for

96

P. Rabinowitz et al.

Table 3. Evidence for animals serving as sentinels for selected chemical terrorism agents
Species with increased
susceptibility

Agent
Nerve agents (G Series, V Series)

Species with
shorter latency

Species at increased risk

of exposure

Vesicant/blistering agents
(sulfur mustard, etc.)
Respiratory tract irritants

Level 1 evidence:rabbit
(miosis) (19)*
No evidence found

No evidence found Level 3 evidence:sheep,

horses, cattle (5,6,20)
No evidence found No evidence found

No evidence found

Asphyxiants (cyanide, etc.)

Level 1 evidence: dogs (21)**

No evidence found No evidence found

Level 3 evidence: wildlife,
No evidence found
livestock (22,23)

Anecdotal reports
of early warning

No evidence found
No evidence found

*After callaway 1971.

**Single observation only.

Table 4. RBC AchE activity in different species (26)*

Species
Human
Monkey
Pig
Goat
Sheep
Mouse
Dog
Guinea pig
Rabbit
Rat
Cat

RBC AchE activity

(umol/ml/min)

Optimum substrate
concentration (M)**

12.6
7.1
4.7
4.0
2.9
2.4
2.0
2.7
1.7
1.7
1.5

2 103
2 103
1 103
2 103
2 103
2 103
2 103
2 103
5 103
5 103
5 103

*After callaway 1971.

**Acetylthiocholine iodide concentration for maximum RBC AchE activity.

rabbits, and 450 mg.min/m3 for rats. There is a report of an

LCt50 of 9.2 mg.min/m3 for goats exposed for 10 minutes to
VX, based on experiments carried out in 1960 (19).
Using a model based on interspecies relationships between
body size/weight/shape/physiology and toxicology to estimate human LCt50, an excellent correlation has been found
between the minute respiratory volume to body weight ratio
(MV/BW) and the observed toxicity of GB. This model estimated the LCt50 for GB for humans (based on 10 minute
exposures) as 57 mg.min/m3 (MV/BW =0.223) compared to
actual LCt50 measures derived from compilations of prior
studies in multiple species: 145 mg.min/m3 for rabbits (MV/
BW=0.416), 90 mg.min/m3 for cats (MV/BW=0.335), 122
mg.min/m3 for dogs (MV/BW=0.328), and 316 mg.min/m3
for mice (MV/BW=1.076). The lower LCt50 for humans
compared to non-human animals corresponds to the lower
MV/BW ratio (28).
While the lethal nerve agent exposure level for animals
may, according to such models, be higher than that for
humans, it is still possible that animals could exhibit nonlethal effects of nerve agents sooner and more noticeably than
humans. Callaway (1971) demonstrated that rabbits develop

90% miosis at a lower inhaled concentration of GB compared

to humans (2.71 mg.min/m3 versus 13.85 mg.min/m3 [19].
Another possibility for animals providing early warning is via
the sense of smell. There are a number of significant differences in the anatomy and biochemistry of the nasal passages
between animals and humans, which can lead to significant
interspecies differences in the ability to sense low gas concentrations (29). The G series of nerve agents is reported to
have a faintly fruity or spicy odor. VX, however, is considered to be odorless. Reported odor thresholds for humans
for these agents range between 1.5 and 7 mg/m3 (19). We
were unable, however, to locate published studies of relative
animal odor thresholds for nerve gases in order to compare
thresholds in animals versus humans.
Nerve agents have a short latency period in humans, with
acute effects noticeable in a number of minutes. No studies
were found indicating a shorter latency time for animals compared to humans.
Regardless of the general lack of evidence regarding
increased susceptibility or decreased latency period for nerve
agent toxicity in animals compared to humans, it is possible
that in the event of a widespread release of a nerve agent
aerosol, animals may experience greater exposure than
nearby human populations. Vapor clouds of the G and V
series of nerve agents are heavier than air, and would therefore tend to settle close to the ground level (30,31). Animals
with a lower breathing zone could therefore potentially have
greater respiratory exposure. However, animals such as
rodents that are nose breathers are capable of partially detoxifying nerve agents in the nasal pathways, presumably by
hydroxylation and other mechanisms (19). Two other exposure pathways could be important in an acute release. Percutaneous exposure due to vapor droplets could be significant
for some agents. However, we did not locate studies showing
a lower LCt50 for percutaneous exposure in animals vs.
humans. In fact, it has been hypothesized that due to protective fur or feathers, animals may have decreased dermal
absorption compared to humans: when the US military
recently withdrew a plan to use chickens as sentinels for airborne nerve agents, the protective quality of the chickens
feathers was one reason provided (11). Finally, ingestion of

97

Animals as sentinels of chemical terrorism

food containing chemical residues from airborne deposition
could be another route of exposure. However, the majority of
oral lethal dose studies for these agents have been performed
in rats, which appear to have a dose-related oral susceptibility
to nerve agents that is similar to humans (32).
There are a number of instances where shared exposures to
nerve agents have been reported for both animals and
humans. After an apparent inadvertent release of two different nerve agents (one of which may have been VX) near a
chemical and biological warfare center in Utah, a band of
sheep grazing near the base was noted to be acting crazy in
the head, and thousands in the band died less than 24 hours
later (20). At the same time, humans, cattle, dogs, and horses
living nearby were not reported to have symptoms. Cholinesterase testing of the sheep in the area showed severe
depression, as did cattle and horses, while cholinesterase testing of dogs and humans was apparently normal. This phenomenon of greater effects in sheep (and lesser evidence of
toxicity in horses and cattle) may have been due to these livestock receiving higher exposures through ingestion of vegetation contaminated by the agent, or spending more time in the
vicinity of the chemical release.
In summary, there is limited evidence that rabbits and
goats may show some increased susceptibility to nerve
agents for certain non-lethal endpoints compared to
humans. Anecdotal reports such as the sheep kill in Utah,
however, suggests that differential exposure scenarios
(based more on geographical and behavioral rather than
physiological differences) could still produce effects in animals more than humans. For the practice of using captive
chickens or canaries, however, in instances where human
and animal exposures would be more similar, the evidence
seems to be lacking, even though hens have been used as a
model to study delayed neuropathy due to low dose GA
exposure (33).
Blistering/vesicant agents (sulfur mustard and nitrogen
mustard)
Sulfur mustard (also known as agent HD, dichloroethyl sulfide, or mustard gas) is a chemical vesicant or blistering agent
that attacks epithelial surfaces including skin and the mucus
membranes of the eyes and respiratory tract. Due to its low
solubility in water, it tends to persist in the environment. It
has a garlic-like odor, but olfactory fatigue can develop after
several minutes (34). The eyes are considered to be the most
sensitive target organ. Unlike nerve agents, the latency of
clinical effects may be as long as several days, although ocular and upper respiratory tract responses have a shorter
latency (several hours) compared to dermal and systemic
responses (19).
Lethality levels for human due to acute inhalation exposure
are based on animal data and have been estimated at 1500
mg.min/m3 and 10, 000 mg.min/m3 for inhalation and percutaneous exposure, respectively (15). Several animal species

have been studied for acute lethality. Rosenblatt, found that

(1975) in dogs the LCt50 was 600 mg.min/m3, while in the
cat it was 700 mg.min/m3 (15), and in the guinea pig
reported estimates range from 800- 1700 mg.min/m3. The
human odor threshold has been reported between 0.15 mg/m3
and 0.6 mg/m3(19). No data could be located regarding animal odor thresholds. The exposure level required to produce
severe ocular irritation in humans has been reported as an
effective concentration 50% level (Ect50) of 100 mg.min/m3
for severe ocular effects (19). Similar estimates for animals
could not be located.
The latency period for effects of sulfur mustard appears to
be related to the concentration and duration of exposure. We
were unable to locate precise data regarding interspecies
differences in latency at similar exposure levels.
In the case of skin exposure, in vitro experiments with sulfur mustard have indicated that the it moves through pig epidermis three times faster than it moves through human
epidermis, possibly because of differences in membrane constituents (35). However, another study found pig skin to have
similar permeability characteristics to human skin and to be a
better model than rat skin with respect to a variety of chemicals (36). In general, animal model studies of skin exposure
to nitrogen mustard, sulfur mustard and Lewisite indicate that
the potential of these agents to cause skin lesions with direct
contacts is greatest in hairless animals such as pigs, and less
in fur-covered species (37).
We were unable to locate any case reports or other epidemiological studies of animals exposed to sulfur mustard and
providing early warning to a human population. Our review
of existing reports therefore found little to no evidence that
animals are either more susceptible to an airborne exposure to
blistering agents or that they could serve as sentinels for an
airborne attack with such agent.
Respiratory tract irritants (including phosgene and
chlorine)
For the class of respiratory tract irritants that includes chlorine and phosgene, there is evidence that animals develop
pulmonary pathology after airborne exposure that is comparable to humans (38). For chlorine, 50% lethal concentration
(LC50) estimates are 250 ppm and 100 ppm in regular and
vulnerable human populations, respectively. This is within
the same order of magnitude as LC50 estimates for other animal species: dog 650ppm, rat 293 ppm, and mice 137 ppm
(39). While phosgene does have a musty hay odor that can
act as a warning (40), no data were located regarding relative
odor thresholds for animals versus humans. Although most
animals develop the characteristic delayed pulmonary edema
after phosgene exposure, dogs seem prone to bronchospasm
that is manifest as acute bronchiolitis (38). For chlorine we
could not identify studies demonstrating that certain species
were more susceptilde than humans to direct effects. With
phosgene, due to its low solubility, there is a latency period

98
of 224 hours in humans, and no evidence could be found
suggesting a significantly shorter latency in animals.
No studies were located documenting cases of shared
exposure between animals and humans to phosgene or chlorine gas. Evidence is currently lacking that any animal species would have greater susceptibility to respiratory tract
irritants or could serve as sentinels to an airborne chemical
attack that utilized these agents.
Asphyxiants
For the asphyxiant agent, hydrogen cyanide there are significant interspecies differences in the ability to metabolize the
toxin (41), with dogs appearing to be more susceptible to cyanide poisoning relative to humans, due to low levels of
endogenous rhodanese, the hepatic enzyme (42) that catalyzes the sulfuration of cyanide to thiocyanate. Indeed, a
report of a controlled exposure suggests that dogs could be
more susceptible than humans. In a case of simultaneous,
controlled (albeit a one time experiment that was never
repeated) exposure to hydrogen cyanide gas, a 70 kg man and
a 12 kg dog were exposed in an airtight chamber to HCN
concentrations between 500 and 625 ppm. The dog became
unsteady at 50 seconds, unconscious at 75 seconds, and convulsive at 90 seconds. One second after this, the man walked
out of the exposure chamber with no symptoms, although
over the next ten minutes developed some transient nausea
and difficulty concentrating (21). While it is possible that the
dog experienced a higher exposure than the human, HCN
tends to be lighter than air, arguing against this possibility
and instead indicating that the dog was indeed more susceptible.
Therefore, cyanide appears to be an agent where certain
species (dogs) may have increased susceptibility relative to
humans based on physiological differences that would place
them at greater risk in a scenario of shared exposure. However, the evidence on susceptibility of dogs rests largely on
one report and needs to be borne out in more rigorous studies.

Discussion
This evidence-based assessment of the utility of animals as
sentinels of chemical terrorism agents found, in general, little
evidence that animals were intrinsically more susceptible or
capable of developing acute effects more readily than
humans. A notable exception was the possibly increased susceptibility of dogs to cyanide gas. At the same time, anecdotal case reports related to accidental releases of chemical
warfare agents provide some limited evidence that in the
event of a chemical terrorism attack, animals could exhibit
mortality or acute symptoms sooner than humans, as a result
of increased environmental exposure. However, the many
gaps in the evidence about animals as sentinels, and the interesting, although anecdotal case reports suggesting that at
times animals could sicken before humans argue for continued

P. Rabinowitz et al.
exploration of comparative toxicological risk in animals living near humans, including companion animals, livestock,
and peridomestic wildlife. Such additional research could
include analysis of surveillance and other observational data
as well as standard toxicological exposure studies.
Linking animal and human health surveillance data has
inherent challenges. At present, there is no consistent way for
public health authorities to monitor disease events in animals.
In the absence of a nationwide surveillance system for animal
diseases, the responsibility for detecting clusters of unusual
symptoms in animal populations falls to a diverse group of
professionals including farmers, agriculture officials, individual animal owners, veterinarians, animal control officers,
wildlife rehabilitators, (animal) poison control centers, and
the lay public. Few of these agents operate under a mandate
to examine the human health implications of an animal disease event. Recent awareness of zoonotic diseases such as
West Nile Virus and Avian Influenza have led, however, to
greater tracking of wild bird and other animal mortality
events by public health authorities. Additionally, data from
pilot animal health surveillance projects such as the syndromic surveillance system recently implemented in a large veterinary hospital chain (43) could provide additional
information about the value of animal sentinels.
Data from Poison Control Centers could provide a unique
insight into the human health sentinel value of animal poisoning events. According to the American Association of
Poison Control Centers, poison control centers received
131,336 calls regarding animal exposures in 2005 with
88.6% of all animal calls involving dogs (44). Currently, the
case data recorded in the AAPCCs Toxic Exposure Surveillance Systems (TESS) includes only the limited exposure
data available around the time an incident occurs. AAPCCs
TESS data are analyzed in near real time for unusual occurrences or clusters as part of an ongoing sentinel surveillance
system. If cases of animal deaths or illness suspected to be
secondary to chemical exposures were reported to regional
poison control centers, these events could be included in the
nationwide AAPCC sentinel surveillance system. Possible
linkages between animal and human health disease events
could then be analyzed, and our understanding of the value
of animals as sentinels of human health hazards could be
enhanced.
Our search for evidence regarding animals as sentinels
of chemical terrorism agents revealed numerous knowledge
gaps in linking animal health and human health events.
Research is needed to determine relative susceptibilities
and differences in exposure pathways between animal species and human populations. While such measures could
have potential value in the early detection and management
of an attack with chemical terrorism agents, the enhanced
communication between animal and human health professionals could also improve our ability to manage emerging
zoonotic disease outbreaks, as well as aid in the early
detection and prevention of human illness from other environmental hazards.

99

Animals as sentinels of chemical terrorism

Acknowledgments
Funding for this project came from the National Library of
Medicine, grant no. 1 G08 LM07881-01. Some of this material was presented as a poster at the annual conference of
North American Congress of Clinical Toxicology (NACCT),
September 2005 in Orlando, FL. The authors thank Drs.
Jonathan Borak and Fred Henretig for providing helpful critique of article drafts.

19.

20.
21.
22.
23.

References
24.
1. Anonymous. Biological and chemical terrorism: Strategic plan for preparedness and response. Recommendations of the CDC Strategic Planning
Workgroup. Morbidity and Mortality Weekly Report 2000; 49:114.
2. Burrell G, Seibert F. Experiments with small animals and carbon monoxide. The Journal of Industrial and Engineering Chemistry 1914;
6(3):2414.
3. Wise JK, Heathcott BL, Shepherd AJ, Avma. Results of the 2002
AVMA survey of US pet-owning households regarding use of veterinary services and expenditures. Journal of the American Veterinary
Medical Association 2003; 222(11):15245.
4. Daszak P, Tabor GM, Kilpatrick AM, Epstein J, Plowright R. Conservation medicine and a new agenda for emerging diseases. Annals of the
New York Academy of Sciences 2004; 1026:111.
5. Anonymous. Animals in Gas Attacks. Corydon Republican 1918 July 4;
Sect. 7.
6. Brankowitz WR. Chemical Weapons Movement: History Compilation.
Aberdeen Proving Ground, MD: Office of the Program Manager for
Chemical Munitions; 1987 April 27.
7. Paddle BM. Biosensors for chemical and biological agents of defence
interest. Biosensors & Bioelectronics 1996; 11(11):1079113.
8. Biema DV. Prophet of poison. Time 1995 April 3:2733.
9. Ember L. Chemical weapons: Chickens on alert in Kuwait. Chemical &
Engineering News 2003; March 10:12. Available at <pubs.acs.org/cen/
topstory/8110/notw3.html>
10. American Forces Press Service. Defending against Iraqi chemical, biological threats, 2003. Available at <http://www.defenselink.mil/news/
Mar2003/n03032003_200303034.html>
11. United States Department of Defense, 2003. Briefing on Chemical and
Biological Defense Readiness March 03, 2003. <http://
www.defenselink.mil/transcripts/2003/
t03032003_t0303chembio.html<, (accessed 8 March 2007).
12. U.S. Environmental Protection Agency. Sentinels detection methods
and chemical decontamination health risk key components of EPAs
Safe Building Program. <http://www.epa.gov/scienceforum/PDFs/
security/hatch_g.pdf>, (accessed15 April 2006).
13. Cottrell T, Morgan E. Animal surveillance in NBC defensive operations.
J R Army Med Corps 2003; 149:225229.
14. Kales SN, Christiani DC. Acute chemical emergencies. N Engl J Med
2004; 350:800808.
15. Anonymous. Review of the U.S. Armys health risk assessments for
oral exposure to six chemical-warfare agents. Introduction. J Toxicol
Environ Health Part A 2000; 59:281526.
16. The Canary Database: Animals as Sentinels of Human Environmental
Health Hazards. <http://www.canarydatabase.org>, (accessed 9 March
2007).
17. Ebell MH, Siwek J, Weiss BD,Woolf SH, Susman J, Ewigman B,.
Bowman M. Strength of recommendation taxonomy (SORT): a patientcentered approach to grading evidence in the medical literature Am
Fam Physician 2004; 69:548556.
18. Phillips B, Ball C, Sackett D, Badenoch D, Straus S, Haynes B. Levels
of Evidence and Grades of Recommendation. Oxford, UK: Oxford

25.

26.

27.
28.

29.
30.

31.

32.

33.

34.

35.

36.
37.

38.

39.

40.

Center of Evidence Based Medicine; 2001 <http://www.cebm.net/

levels_of_evidence.asp#levels>, (accessed 9 March 2007).
National Research Council: Subcommittee on Acute Exposure Guideline Levels, Committee on Toxicology. Acute Exposure Guideline Levels for Selected Airborne Chemicals. Volume 3. Washington, DC: The
National Academies Press; 2003.
Boffey PM. 6000 sheep stricken near CBW center. Science
1968;159:1442.
Barcroft J. The toxicity of atmospheres containing hydrocyanic acid
gas. Journal of Hygiene 1931; 31:134.
Koenig R. Environmental disasters. Wildlife deaths are a grim wake-up
call in Eastern Europe. Science 2000; 287:17371738.
Aslani MR, Mohri M, Maleki M, Sharifi K, Mohammadi GR, Chamsaz
M. Mass cyanide intoxication in sheep. Vet Hum Toxicol 2004;
46:186187.
Fonnum F, Sterri SH. Factors modifying the toxicity of organophosphorous compounds including Soman and Sarin. Fundam Appl Toxicol
1981; 1:143147.
Maxwell DM, Brecht KM, ONeill BL. The effect of carboxylesterase
inhibition on interspecies differences in soman toxicity. Toxicol Lett
1987; 39:3542.
Ellin RI. Anomalies in theories and therapy of intoxication by potent
organophosphorus anticholinesterase compounds. Gen Pharm 1982; 13:
457466
Leng, G.L., J., Role of individual susceptibility in risk assessment of
pesticide. Occup Environ Med 1999; 56: 44953.
Bide RW, Armour SJ, Yee E. GB toxicity reassessed using newer techniques for estimation of human toxicity from animal inhalation toxicity
data: new method for estimating acute human toxicity (GB). J Appl
Toxicol 2005; 25:393409.
Dalton P. Upper airway irritation, odor perception and health risk due to
airborne chemicals. Toxicol Lett 2003; 140141, 239248.
Agency for Toxic Substances and Disease Registry. Facts about VX.
<www.bt.cdc.gov/agent/VX/basics/facts.asp>, (accessed 9 March
2007).
Facts about Soman. Agency for Toxic Substances and Disease Registry.
<www.bt.cdc.gov/agent/soman/basics/facts.asp>, (accessed 9 March
2007).
MacGregor JA, Plunkett LM, Youngren SH, Manley A, Plunkett JB,
Starr TB. Humans appear no more sensitive than laboratory animals to
the inhibition of red blood cell cholinesterase by dichlorvos. Regul
Toxicol Pharmacol 2005; 43:150167.
Wilson BW, Henderson JD, Spencer PS. Clinical effects of low-level
exposures to chemical warfare agents in mice and chickens. Drug Chem
Toxicol 1998; 21(Suppl 1):183190.
Agency for Toxic Substances Disease Registry. Toxicological Profile for Sulfur Mustard (Update) U.S. Dept of Health and Human
Services, 2003.
Chilcott RP, Jenner J, Hotchkiss SA, Rice P. In vitro skin absorption
and decontamination of sulphur mustard: comparison of human and
pig-ear skin. J Appl Toxicol 2001; 21:279283.
Dick IP, Scott RC. The influence of different strains and age on in vitro
rat skin permeability to water and mannitol. Pharm Res 1992; 9:884887.
Smith KJ, Casillas R, Graham J, Skelton HG, Stemler F, Hackley BE, Jr.
Histopathologic features seen with different animal models following
cutaneous sulfur mustard exposure. J Dermatol Sci 1997; 14:126135.
National Research Council : Subcommittee on Acute Exposure Guideline Levels, Committee on Toxicology. Acute Exposure Guideline Levels for Selected Airborne Chemicals. Volume 2. Washington, DC: The
National Academies Press; 2002.
National Research Council: Subcommittee on Acute Exposure Guideline Levels, Committee on Toxicology. Acute Exposure Guideline Levels for Selected Airborne Chemicals. Volume 4. Washington, DC: The
National Academies Press; 2004.
Karalliedde, L., et al., Possible immediate and long-term health effects
following exposure to chemical warfare agents. Public Health,
2000;114: 238248.

100
41. Sousa AB, Manzano H, Soto-Blanco B, Gorniak SL. Toxicokinetics of
cyanide in rats, pigs and goats after oral dosing with potassium cyanide.
Arch Toxicol 2003; 77:330334.
42. Aminlari M, Gilanpour H. Comparative studies on the distribution of
rhodanese in different tissues of domestic animals. Comp Biochem
Physiol 1991; 99B:673677

P. Rabinowitz et al.
43. ORourke K. National pet health surveillance system in the works. J
Am Vet Med Assoc 2003; 222:13351336.
44. Lai MW K-SW, Rodgers GC, Abrams JY, Haber DA, Bronstein AC,
Wruk KM. 2005. 2005 Annual report of the American Association of
Poison Control Centers National Poisoning and Exposure Database.
Clin Toxicol 2006; 44:803932.