Page 53 - Veterinary Toxicology, Basic and Clinical Principles, 3rd Edition

20 SECTION | I General

VetBooks.ir TABLE 1.1 Bradford Hill (1965): Guidance for Interpreting Epidemiological Evidence for a Given Agent or Factor

Causing a Specific Disease
Strength: If there is a large effect size (a strong effect), this makes a true causal relationship more likely. There is no fundamental reason
why a strong association should be more likely to be causal than a weak association. However, Bradford Hill was thinking of possible
confounding factors and argued that if a strong association is not causal, an association with some other factor that varies closely with the
factor originally suggested must exist. If the association is strong, this covariable will be easier to recognize than if the association is
weak.
Consistency: Is there evidence for the same findings from more than one study, preferably from different settings?
Specificity: Is the effect specific to the exposure? In reality this rarely occurs, and some regard this as the “icing on the cake” when
considering causal evidence. However, the specificity of a particular mechanism, if seen across a range of health endpoints, would meet
the requirements of this component.
Temporality: This is the only absolute. Exposure must precede outcome.
Biological gradient: Is there a dose response relationship? In addition (although explicit in Bradford Hill’s original paper), is there a
threshold of exposure below which an effect is not seen?
Plausibility: Does this exposure-effect relationship make biological sense? Is there mechanistic evidence to support the likelihood that this
might occur given current knowledge? Bradford Hill pointed out that this feature should not be demanded: what is implausible today
may be entirely plausible tomorrow.
Coherence: Does the proposed causal association cohere with other findings? For instance, if we were concerned that a specific exposure
might be related to a health effect that was based on an inflammatory response, has a similar effect been seen with other inflammatory
conditions, or is an apparent effect on mortality accompanied by effects on morbidity?
Experiment: Is there evidence from experiments (involving either animals or humans, whether individually or as populations) that
removal of the exposure reduces the effect?
Analogy: Are there analogous situations that would tend to support the likelihood of a causal relationship? For instance, when
considering a possible teratogen, consideration of the effects of thalidomide on the developing fetus would come to mind.

Experimentation At this juncture, it is appropriate to note the
importance of animal welfare issues. The Animal Welfare
An additional option for acquiring information is to
Act (AWA), initially enacted in 1966 and amended in
conduct toxicological studies with controlled exposures in
1970, 1976, 1985, 1990, and 2002, is the principal federal
typical laboratory animal species. Such studies are the
statute in the United States governing the sale, handling,
cornerstone of research conducted to evaluate the safety/
transport, and use of animals. The AWA applies to
risk of newly synthesized agents whether they are
all species of warm-blooded vertebrate animals used
a potential new pharmaceutical, pesticide or herbicide,
for research, testing, or teaching excluding animals
a significant consumer product, or a new chemical or
used for agricultural research. The US Department of
intermediate chemical to be used in commerce. It is well
Agriculture’s Animal and Plant Health Inspection Service
recognized, certainly by veterinarians, that no single labo-
(APHIS) has responsibility for implementing the AWA.
ratory animal species is a miniature version of the human
The 1985 Amendments to the AWA clarified the impor-
species, i.e., 15 cm in height, weighing 180 g, and sharing
tance of humane care, minimization of pain and distress,
all of the common biological traits and diseases of
consideration of alternatives, the role of institutional
humans. However, it is fortunate that humans and labora-
animal care and use committees, the psychological well
tory animals do share many common biological traits and
being of primates, and exercise for dogs. The primary
diseases. Knowledge of the extent to which there are sim-
reference on animal care and use is the “Guide for the
ilarities and differences between humans and a given lab-
Care and Use of Laboratory Animals” prepared and
oratory animal species can be used to guide the selection
periodically updated under the auspices of the Institute
of a species to serve as a surrogate for humans in devel-
of Laboratory Animal Resources of the National
oping data for safety/risk evaluations for humans. It is
Academy of Sciences/National Research Council (ILAR,
encouraging that some veterinary medical schools are rec-
2010). All toxicologists involved with laboratory investi-
ognizing the importance of extending the range of species
gations should be familiar with the contents of the Guide
studied in the core curriculum from the usual companion
irrespective of the species they use for their research.
animal and domestic livestock species to include common
An additional matter the experimentalist should be
laboratory animal species.
aware of is the need for use of Good Laboratory Practices