Page 486 - The Veterinary Laboratory and Field Manual 3rd Edition

Wildlife health and disease surveillance 423

stems from several practical and logistical chal- Such data are potentially valuable long into the
lenges. Some of the ‘special problems’ that limit future.
the detection of diseases in wildlife include, for There are, however, issues and biases to
example, difficulties in defining and achieving consider when relying solely on passive surveil-
an adequate sample size, selection bias in the lance for disease identification in wild animals.
sampling methods, difficulty in obtaining good For example, in the wild, carcasses of dead ani-
quality samples and measurement bias in the mals tend to be removed or cached by scavengers
diagnostic tests used (that is, lack of validated very quickly (within hours as opposed to days).
tests for wild species). This means that passive surveillance is likely
Wildlife surveillance can be conceived as a to detect only a fraction of the total individu-
designed survey (active or targeted surveillance), als that have died from a disease, if any at all
for example, with the objective of identifying one (that is, in the case of small-sized species and
or more diseases and/or understanding disease carcasses in remote/sparsely populated regions).
prevalence. This relies on the active collection Additionally, larger or more charismatic animals
and analyses of data and biological samples that are found dead are more likely to be reported
from live and/or dead animals (for example, by than other species. Conversely, there are rare
capturing live animals, by obtaining samples cases in which passive surveillance might lead
from harvested animals or through field disease to overestimation of certain diseases. For exam-
investigation activities). In local and traditional ple, rabid animals are less likely to avoid danger
knowledge systems, participatory epidemiol- and have greater risk of being killed by vehicles.
ogy (PE) tools can be also used as a form of The prevalence of rabies in road-killed animals
disease surveillance. In PE, the epidemiological might, therefore, be higher than the prevalence
intelligence on wildlife populations, including of rabies in the general wildlife population of an
population health status and the presence of dis- area. Selection bias and measurement bias both
eases, is gathered in the form of ethnoveterinary need to be carefully considered during analyses
knowledge by interviewing key informants (for of wildlife data from passive or active surveil-
example, hunters, trappers). This process relies lance. For more information, refer to Wobeser
on the use of participatory tools that allow for (2007) in the bibliography.
both qualitative and quantitative assessments The design of a wildlife disease survey will
(see example in section 11.3). depend on the disease to be studied, the expected
More commonly, wildlife health and disease prevalence of the suspected disease in the targeted
surveillance relies on routine reporting, as opposed population and the size of the population(s) to
to the active investigation, of disease. This is be sampled. The principles of survey design are
called passive surveillance (also known as general well outlined in various textbooks but it is prefer-
or scanning surveillance). For example, an animal able to discuss any intended survey programme
is found dead and is submitted to a central labora- with a statistician, veterinary epidemiologist and
tory where standardized necropsy, examination of other professionals, as well as the laboratories
samples and testing are carried out. A diagnosis or involved in the testing of the samples. When PE
suspected diagnosis is determined and a detailed tools can be applied, the expected prevalence of
report is written, which is then usually entered the targeted disease(s) can be estimated through
into a database. Epidemiologists and others can participatory techniques and the data used for
access the central database to generate informa- sample size calculations.
tion on characteristics, occurrences and patterns When collecting biological samples and data
of disease at local or larger geographical scales. from live wild animals, it is necessary to work

Vet Lab.indb 423 26/03/2019 10:26